Principles of Molecular Medicine [1st ed.] 9780896035294, 0-89603-529-8

Citation preview

PRINCIPLES OF

MOLECULAR MEDICINE 䉬 EDITED BY

J. Larry Jameson SECTION EDITORS

Dennis Ausiello Joseph B. Martin Andrea Ballabio Michael J. McPhaul Michael J. Holtzman Charles B. Nemeroff Ethylin Wang Jabs James C. Reynolds Laurence Kedes Anthony Rosenzweig Thomas Kupper Swee Lay Thein Ralph C. Williams, Jr. HUMANA PRESS

CONTENTS

PRINCIPLES OF MOLECULAR MEDICINE

i

ii

CONTENTS

SECTION EDITORS DENNIS AUSIELLO, MD

MICHAEL J. MCPHAUL, MD

DEPARTMENT OF MEDICINE MASSACHUSETTS GENERAL HOSPITAL BOSTON, MA

DEPARTMENT OF INTERNAL MEDICINE UNIVERSITY OF TEXAS SOUTHWESTERN MEDICAL CENTER DALLAS, TX

ANDREA BALLABIO, MD TELETHON INSTITUTE OF GENETICS AND MEDICINE (TIGEM) MILANO, ITALY

MICHAEL J. HOLTZMAN, MD DIVISION OF PULMONARY CRITICAL CARE MEDICINE ST. LOUIS, MO

AND

ETHYLIN WANG JABS, MD CENTER FOR MEDICAL GENETICS JOHNS HOPKINS UNIVERSITY SCHOOL OF MEDICINE BALTIMORE, MD

LAURENCE KEDES, MD INSTITUTE FOR GENETIC MEDICINE AND PROGRAM FOR GENE THERAPY UNIVERSITY OF SOUTHERN CALIFORNIA SCHOOL OF MEDICINE LOS ANGELES, CA

THOMAS S. KUPPER, MD DERMATOLOGY DIVISION BRIGHAM AND WOMEN’S HOSPITAL BOSTON, MA

JOSEPH B. MARTIN, MD, PhD FACULTY OF MEDICINE HARVARD MEDICAL SCHOOL BOSTON, MA

CHARLES B. NEMEROFF, MD, PhD DEPARTMENT OF PSYCHIATRY AND BEHAVIORAL MEDICINE EMORY UNIVERSITY SCHOOL OF MEDICINE ATLANTA, GA

JAMES C. REYNOLDS, MD DIVISION OF GASTROENTEROLOGY AND HEPATOLOGY ALLEGHENY UNIVERSITY OF THE HEALTH SCIENCES PHILADELPHIA, PA

ANTHONY ROSENZWEIG, MD CARDIAC UNIT AND CARDIOVASCULAR RESEARCH INSTITUTE MASSACHUSETTS GENERAL HOSPITAL BOSTON, MA

SWEE LAY THEIN, MRCP, FRCPath MRC MOLECULAR HAEMATOLOGY UNIT INSTITUTE OF MOLECULAR MEDICINE JOHN RADCLIFF HOSPITAL, HEADINGTON OXFORD, UK

RALPH C. WILLIAMS, JR., MD DIVISION OF RHEUMATOLOGY DEPARTMENT OF MEDICINE UNIVERSITY OF FLORIDA GAINESVILLE, FL

iii

CONTENTS

PRINCIPLES OF

MOLECULAR MEDICINE EDITED BY

J. LARRY JAMESON, MD, PhD NORTHWESTERN UNIVERSITY MEDICAL SCHOOL CHICAGO, IL

FOREWORD BY

FRANCIS S. COLLINS, MD, PhD NATIONAL HUMAN GENOME RESEARCH INSTITUTE BETHESDA, MD

HUMANA PRESS

TOTOWA, NEW JERSEY

iv

CONTENTS

© 1998 Humana Press Inc. 999 Riverview Drive, Suite 208 Totowa, New Jersey 07512 For additional copies, pricing for bulk purchases, and/or information about other Humana titles, contact Humana at the above address or at any of the following numbers: Tel: 973-256-1699; Fax: 973-256-8341; E-mail: [email protected] All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording, or otherwise without written permission from the Publisher. Due diligence has been taken by the publishers, editors, and authors of this book to assure the accuracy of the information published and to describe generally accepted practices. The contributors herein have carefully checked to ensure that the drug selections and dosages set forth in this text are accurate and in accord with the standards accepted at the time of publication. Notwithstanding, as new research, changes in government regulations, and knowledge from clinical experience relating to drug therapy and drug reactions constantly occurs, the reader is advised to check the product information provided by the manufacturer of each drug for any change in dosages or for additional warnings and contraindications. This is of utmost importance when the recommended drug herein is a new or infrequently used drug. It is the responsibility of the treating physician to determine dosages and treatment strategies for individual patients. Further it is the responsibility of the health care provider to ascertain the Food and Drug Administration status of each drug or device used in their clinical practice. The publisher, editors, and authors are not responsible for errors or omissions or for any consequences from the application of the information presented in this book and make no warranty, express or implied, with respect to the contents in this publication. Cover design by Thomas B. Lanigan and Patricia F. Cleary. This publication is printed on acid-free paper. ∞ ANSI Z39.48-1984 (American National Standards Institute) Permanence of Paper for Printed Library Materials. Photocopy Authorization Policy: Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by Humana Press Inc., provided that the base fee of US $8.00 per copy, plus US $00.25 per page, is paid directly to the Copyright Clearance Center at 222 Rosewood Drive, Danvers, MA 01923. For those organizations that have been granted a photocopy license from the CCC, a separate system of payment has been arranged and is acceptable to Humana Press Inc. The fee code for users of the Transactional Reporting Service is: [0-89603-529-8/98 $8.00 + $00.25]. Printed in the United States of America. 10 9 8 7 6 5 4 3 2 1 Library of Congress Cataloging-in-Publication Data Principles of molecular medicine / edited by J. Larry Jameson foreword by Francis S. Collins. p. cm. Includes index. ISBN 0-89603-529-8 (alk. paper) 1. Medical genetics. 2. Pathology, Molecular. 3. Molecular biology. I. Jameson, J. Larry. [DNLM: 1. Genetics, Medical. 2. Molecular Biology. QZ 50 P9573 1998] RB155.P695 1998 616'.042—dc21 DNLM/DLC for Library of Congress 98-17729 CIP

CONTENTS

v

Foreword

Until recently, medical genetics and molecular medicine were considered the exclusive province of academic specialists in tertiary-care medical centers. Queried about their familiarity with molecular genetic aspects of clinical medicine, most primary-care providers only a few years ago would have responded that such matters were irrelevant to their daily practice. Yet few could say that today. Few internists or general practitioners have not prescribed recombinant insulin, tPA, or erythropoetin; few pediatricians have not gone through the molecular evaluation of a child with dysmorphology or learning disability; few obstetricians have not performed amniocentesis or CVS for couples at increased genetic risk; and few general surgeons have not faced penetrating questions about the role of genetic testing or prophylactic surgery from women with a strong family history of breast or ovarian cancer. This level of emergence of molecular genetics into clinical medicine is still quite modest, however, compared to what is coming. As the human genome project hurtles toward completion of the sequence of a reference human genome, and the identification of all human genes, by 2005, the pace of revelations about human illness will continue to accelerate. Until recently, most disease-gene discoveries have related to single-gene disorders (cystic fibrosis, fragile X syndrome, and so on) or to Mendelian subsets of more common illnesses (BRCA1 and BRCA2, the hereditary nonpolyposis colon cancer syndromes, and so on). But with

the initiation in 1998 of an aggressive new genome project goal, cataloging all common human sequence variations, it is expected that the weaker polygenic contributors to virtually all diseases will begin to be discerned. Many consequences will result. Individualized preventive medicine strategies, rooted in the gene-based determination of future risk of illness, will become part of the regular practice of medicine. New designer drugs, based not on empiricism but on a detailed understanding of the molecular pathogenesis of disease, will appear. Pharmacogenomics, wherein the efficacy and toxicity of a particular drug regimen can be predicted based on patient genotype, will become a standard component of designing optimum therapy for the individual. And gene therapy, fed by a wealth of disease-gene discoveries, will mature into a significant part of the physician’s armamentarium against disease. As we watch this train coming down the track, this is an ideal time to collect information about molecular medicine into one authoritative text. Principles of Molecular Medicine aims to do just that, bridging the current gap between basic science and the bedside. It will thus be useful to researchers and clinicians alike. With more than 100 chapters covering a wide variety of topics, its distinguished cohort of section editors, and its abundant tables and illustrations, it provides an accessible and much needed manual to the present and the future of molecular genetics and medicine. Francis S. Collins

v

vi

CONTENTS FOREWORD

CONTENTS

vii

Preface

For most physicians, molecular medicine and genetics have not traditionally played a major role in day-to-day clinical practice. However, new insights into the molecular basis of disease are being generated at an ever-increasing rate, resulting in a transformation in our understanding and management of diseases. This explosion of information has been ignited by a variety of technological advances, and has been fueled by the rapid progress of the human genome project. It is widely recognized that molecular biology is now causing a paradigm shift in the teaching and practice of medicine. The promise of molecular medicine is exciting, but there are also great challenges in the integration of this rapidly advancing field into our understanding and treatment of disease. Principles of Molecular Medicine attempts to close the gap between traditional textbooks of medicine and the burgeoning database of new knowledge that has been generated by molecular biology. The book’s title, Principles of Molecular Medicine, reflects our effort to translate the advances provided by genetics and molecular biology into each of the major specialties of medicine. By analogy with traditional textbooks of medicine, the book is organized according to major organ systems. This format is familiar to most medical readers and thus conforms to the specialty areas of many authors and readers. We believe that this book will be of value to a broad audience that includes sophisticated students, specialists who are seeking updates in their own areas or on topics that they have not followed closely, and practicing physicians who remain vitally interested in learning about the remarkable changes in molecular medicine. Each of the various specialty sections of Principles of Molecular Medicine have been edited by experts in their respective fields. The book opens with a series of introductory chapters, and each specialty section contains additional background overview chapters that address issues of molecular pathophysiology specific to their respective organ systems. Even though the field of molecular medicine is evolving quickly, the book contains up-to-date reviews of the genetic basis of diseases, with an emphasis on principles that should allow the reader to integrate basic knowledge with all the latest breakthroughs. Our authors have aspired to clarify complex topics

with many lucid figures that depict cellular and genetic pathways. And we have placed particular emphasis on the molecular mechanisms of disease and on those new concepts resulting from application of the tools of molecular biology. An especially exciting dimension of the book is the translation of the many recent advances in research into clinically useful information; each of our authors has attempted to project the future implications of recent developments in their specialty areas. Beyond the insights afforded into the pathophysiology of disease, it can be argued that genetics plays a role in virtually every medical condition. It has been estimated that the human genome contains between 50,000 and 80,000 genes. Though many diseases are caused by mutations in critical genes, it is becoming increasingly evident that one’s “genetic background” can result in predisposition to many diseases or can modify the host response to environmental events. In some cases, such as hypertension and cardiovascular disease, the genetic contributions are polygenic, and we are still in the early stages of identifying the many genes that contribute to these conditions. In other instances, such as hemophilia, cystic fibrosis, and hundreds of other disorders, the responsible genes have already been well-characterized. From one perspective, the onslaught of new information can seem daunting and difficult to assimilate, particularly because much of the technology and terminology is new. Ironically, the new insights provided can in fact greatly simplify areas that were previously mysterious. For example, several different genetic defects can cause peripheral neuropathies, but disruption of the normal folding of myelin sheaths appears to represent a common final pathway. Likewise, several genetically distinct forms of Alzheimer disease appear to share a common final pathway that involves the formation of neurofibrillary tangles. Identification of the nature of the defective genes (e.g., dystrophin, CFTR, FGF-receptor) can pinpoint the pathway that is involved in key physiologic processes. Similarly, transgenic and gene “knockout” models can reveal the physiologic function of genes. One of the surprises in this new field of molecular medicine is its already pervasive impact in every specialty of medicine. For example, cardiologists are unraveling the molecular basis of inherited cardiomyopathies and ion chanvii

viii

CONTENTS PREFACE

nel defects that predispose to arrhythmias. Neurologists have identified a startling number of genes in which mutations lead to neurodegenerative disorders. Not surprisingly, hematology has progressed rapidly from the classic genetic descriptions of hemoglobinopathies to define the molecular basis of other disorders, including red cell membrane defects, clotting disorders, and thrombotic disorders. The genetic causes of leukemias and lymphomas have created important paradigms for understanding mechanisms of neoplasia. The identification of the cystic fibrosis transporter allows molecular diagnosis of this disease, and gene therapy protocols are underway at several centers. Inherited variations in the immune system impact upon susceptibility to infectious agents, as well as an array of inflammatory and autoimmune disorders. Our understanding of such viral infections as hepatitis and HIV has been aided greatly by the use of recombinant DNA technology. In endocrinology, new insights into hormone action, sexual differentiation, and mechanisms of endocrine neoplasia have been gained by elucidating the nature of genetic defects. Moreover, the field of hormone and signal transduction has revealed a remarkably intricate network of interacting pathways that mediate cellular responses to external signals. In nephrology, a cause of polycystic kidney disease has been identified and the molecular pathogenesis of ischemic and inflammatory disorders are becoming clearer. In dermatology, there has been remarkable progress in mechanisms of neoplasia, the molecular pathogenesis of pemphigoid diseases, and a variety of other autoimmune disorders. Psychiatry too is an area in which the role of genetics is rapidly emerging as highly important; the foundations for this development are outlined in our chapters on behavior, schizophrenia, affective disorders, and alcoholism. Molecular mechanisms have also now been elucidated for many “classic” genetic disorders. Studies of Prader-Willi, Angelman, and Beckwith-Wiedemann syndromes have revealed a critical role for imprinting. The impact of nucleotide repeat disorders has been documented in the fragile X syndrome, as well as for such neurologic disorders as Huntington disease. In most cases, our new knowledge of these disorders has improved the feasibility and accuracy of diagnostic testing, enhanced our understanding of pathophysiology, and is beginning to unmask new avenues for therapy, including gene therapy.

We have been fortunate to enlist the expertise of a renowned international group of section editors and authors for Principles of Molecular Medicine. They have generously taken time from their full palette of research, teaching, and clinical activities to prepare chapters in a manner that is useful to individuals both within and outside of their specialty areas. Although the book contains more than 120 chapters, it is not all-inclusive. Rather, we have focused on disorders in which there has been substantial recent progress and for which there appear to be gaps in coverage elsewhere. The reader is encouraged to seek out further sources for additional information. A particularly valuable resource is the on-line version of Mendelian Inheritance of Man (OMIM), which can be accessed at www3.ncbi.nlm.hih.gov./ OMIM. The molecular basis of metabolic disorders has been well-covered in textbooks of internal medicine and in Scriver’s, The Metabolic and Molecular Basis of Inherited Disease. We anticipate that Principles of Molecular Medicine will evolve with the field, retaining an emphasis on clinical molecular medicine. In addition to the dedication and scholarship of our section editors and authors, a great many individuals have contributed to the success of this book. The idea for the text was conceived in discussions with Victoria Reeders. Mary Kay McMahon provided enormous effort toward the timely completion of this project, superimposed upon her many other responsibilities. Kristina Stanfield, Patty Kalan, Joanne McAndrews, and William Lowe helped to ensure that last-minute updates were included. I am also grateful to the many colleagues who took the time to proofread and critique chapters. Many thanks are owed to Thomas Lanigan at Humana Press for his enthusiastic support and for ushering this concept to completion. I am grateful to many mentors and students for stimulating my interest in molecular medicine. Among these individuals, I especially acknowledge Joel Habener and William Crowley, with whom I have shared the excitement that comes with new insight and discovery. On a personal note, my wife, Michele, and my children, Ryan, Christina, and Jimmy, provided the encouragement that has allowed me to work in the mode of a perpetual student. J. Larry Jameson

ix

CONTENTS

Contents

FOREWORD PREFACE CONTRIBUTORS

XIII

LIST OF COLOR PLATES

XIX

12

V

1

VII

13

Recombinant DNA and Genetic Techniques

14

Transcriptional Control of Gene Expression

15

Transmission of Human Genetic Disease

16

3

The Human Genome Project

17

9

The Cell Cycle

18

25

Oncogenes and Tumor Suppressor Genes Molecular Diagnostic Testing

19

59

Genetic Counseling

65 20 73

Transgenic Mice as Models of Disease

21

83

22

89

Cardiovascular Gene Therapy

161

SWEE LAY THEIN Hematopoiesis: Growth Factors and Mechanisms of Regulation

171

Disorders of Hemoglobin Structure and Synthesis

179

Disorders of the Red Cell Membrane

191

Red Cell Enzymopathies

197

Lucio Luzzatto and Rosario Notaro

23

97

Coagulation Disorders

209

Martina Daly, Anne Goodeve, Peter Winship, and Ian Peake

24

II. CARDIOLOGY ANTHONY ROSENZWEIG Molecular Cardiology: An Overview

157

Jean Delaunay

T. Rajendra Kumar and Martin M. Matzuk

11

Cardiac Arrhythmias

Swee Lay Thein and Jacques Rochette

Beth A. Fine

10

145

Andrew Haynes and Nigel Russell

C. Sue Richards and Patricia A. Ward

9

Hypertension

III. HEMATOLOGY

43

J. Larry Jameson

8

141

Giuseppe Vassalli and David A. Dichek

Lynda Q. Nguyen and J. Larry Jameson

7

Endothelium-Derived Nitric Oxide and Control of Vascular Tone

Barry London

J. Larry Jameson

6

133

George Koike and Howard J. Jacob

Peter Kopp and J. Larry Jameson

5

Coronary Atherosclerosis

Santiago Lamas and Thomas Michel

Wade Johnson and J. Larry Jameson

4

127

Robert E. Gerszten and Anthony Rosenzweig

Marcus Grompe, Wade Johnson, and J. Larry Jameson

3

Inherited Cardiomyopathies Christine E. Seidman, Calum MacRae, and J. G. Seidman

Sarah H. Elsea and Pragna I. Patel

2

117

Alvin J. Chin

I. INTRODUCTION TO MOLECULAR MEDICINE ANDREA BALLABIO AND J. LARRY JAMESON Organization of the Human Genome, Chromosomes, and Genes

Congenital Heart Disease

Thrombotic Disorders

219

Robin J. Olds, David A. Lane, and Swee Lay Thein

113

25

Anthony Rosenzweig

Paroxysmal Nocturnal Hemoglobinuria Bruno Rotoli and Khedoudja Nafa

ix

227

x

CONTENTS

26

Leukemias

233

43

27

Lymphomas

241

44 45

IV. IMMUNOLOGY RALPH C. WILLIAMS, JR. Regulation of Humoral Immunity Molecular Regulation of Cellular Immunity

251

Cytokines

259

46

The HLA Complex

47 273

Inherited Immune Deficiency

283

Human Immunodeficiency Virus and Acquired Immune Deficiency Syndrome (AIDS) 293 Autoimmune Diseases

50

Allergic Diseases: Asthma as a Model

299

51

309

52 53

Cystic Fibrosis

54

Pulmonary Emphysema

319 55

Surfactant Deficiency

Disorders of the Parathyroid Gland Congenital Adrenal Hyperplasia Adrenal Diseases Multiple Endocrine Neoplasia Type 2

329 339

Pamela M. Thomas, Gilbert J. Cote, and Robert F. Gagel

56

451 459 475 481 495

57

Lung Cancer: The Role of Tumor Suppressor Genes 357

Regulation of Reproduction

513

519

Disorders of Sex Determination and Differentiation

527

Charmian A. Quigley

58

Steven Jay Weintraub

505

Michael J. McPhaul

349

Jeffrey A. Whitsett and Timothy E. Weaver

40

Thyroid Disorders

Molecular Mechanisms of Hypoglycemia Associated with Increased Insulin Production

Steven D. Shapiro and Robert M. Senior

39

Growth Hormone Deficiency Disorders

Robert F. Gagel, Sarah Shefelbine, and Gilbert Cote

Daniel B. Rosenbluth and Steven L. Brody

38

443

Constantine A. Stratakis and George P. Chrousos

Michael J. Holtzman, Dwight C. Look, Michael F. Iademarco, Douglas C. Dean, Deepak Sampath, and Mario Castro

37

Pituitary Function and Neoplasia

Robert C. Wilson and Maria I. New

V. PULMONOLOGY 36

433

Andrew Arnold and Andrew F. Stewart

Susan M. MacDonald

MICHAEL J. HOLTZMAN Asthma

Diabetes Mellitus

Peter Kopp and J. Larry Jameson

N. Lawrence Edwards

35

419

Joy D. Cogan and John A. Phillips III

John W. Sleasman and Maureen M. Goodenow

34

MICHAEL J. MCPHAUL AND J. LARRY JAMESON Mechanisms of Hormone Action

Shlomo Melmed

49

Richard Hong

33

407

William L. Lowe, Jr.

48

Robert W. Karr

32

Small- and Large-Bowel Dysfunction

William L. Lowe, Jr., Richard G. Pestell, Laird D. Madison, and J. Larry Jameson

267

William L. Lowe, Jr., and Barbara A. da Silva

31

401

VII. ENDOCRINOLOGY

Eric Sobel

30

Pancreatic Exocrine Dysfunction

Deborah C. Rubin

Ralph C. Williams, Jr.

29

387

David Whitcomb and Jonathan Cohn

Finbarr E. Cotter

28

Viral Hepatitis and Liver Disease Sanjeev Gupta and Michael Ott

Jeffrey E. Rubnitz and Ching-Hon Pui

Sex Chromosome Disorders

561

Andrew R. Zinn

VI. GASTROENTEROLOGY 41

JAMES C. REYNOLDS Hepatology

59

Inherited Liver Disease Juan Ruiz and George Y. Wu

569

Karen D. Bradshaw and Charmian A. Quigley

365

60

Defects of Androgen Action

581

Michael J. McPhaul

Piet C. de Groen and Nicholas F. LaRusso

42

Disorders of Pubertal Development

375

61

Testicular Diseases Marco Marcelli

587

xi

CONTENTS

62 63

Elizabeth A. McGee and Nicholas A. Cataldo

The Dystrophic Forms of Epidermolysis Bullosa

Breast Cancer

Jouni Uitto and Angela M. Christiano

Ovarian Diseases

611

78

625

B. Genetic Mutations That Predispose to Cancer

Melora D. Berardo, D. Craig Allred, and Peter O’Connell

79

VIII. NEPHROLOGY 64

DENNIS AUSIELLO Renal Development Mechanisms of Leukocyte Extravasation

80

Ischemic Acute Renal Failure

635 81 641 651

Joseph V. Bonventre

67

Potassium Secretory Channels in the Kidney Alport Syndrome

82

83

Karl Tryggvason and Pirkko Heikkilä

69

Nephrogenic Diabetes Insipidus

669

Dennis Brown and Dennis A. Ausiello

70

Polycystic Kidney Disease

675

Gregory G. Germino and Luiz F. Onuchic

71

Renal Neoplasms: Wilms’ Tumor and Renal-Cell Carcinoma

84

685

85 86

88

89 695

Angela M. Christiano, Daniel B. Dubin, and Thomas S. Kupper

Psoriasis Atopic Dermatitis and Atopy Pemphigus Foliaceus and Pemphigus Vulgaris

90

Cutaneous Lupus Erythematosus

Epidermolytic Hyperkeratosis

91

707

801

811

821

829

X. MUSCULOSKELETAL 713

92

LAURENCE KEDES Muscle Development and Differentiation

841

Eric N. Olson

719

93

Skeletal Muscle Structure and Function

851

Henry F. Epstein

Lowell A. Goldsmith and Ervin Epstein, Jr.

Junctional Forms of Epidermolysis Bullosa 723 Angela M. Christiano and Jouni Uitto

793

Grant J. Anhalt and Diya F. Mutasim

Edwin A. Smith and E. Carwile LeRoy

Darier’s Disease and Hailey-Hailey Disease

789

Bullous Pemphigoid, Cicatricial Pemphigoid, and Pemphigoid Gestationis 817

Yiu-mo Chan and Elaine Fuchs

699

Amy S. Paller

77

Melanoma Genetics

Scleroderma (Systemic Sclerosis) and Morphea

Mosaicism and Epidermal Nevi

785

Richard D. Sontheimer

John J. DiGiovanna, Sherri J. Bale, and Peter M. Steinert

76

Basal- and Squamous-Cell Carcinomas

Janet A. Fairley, Xiang Ding, George J. Giudice, and Luis A. Diaz

A. Selected Epidermal Gene Mutations

75

ACQUIRED DISEASES OF CUTANEOUS TISSUES Acquired Diseases of Cutaneous Tissues: Introduction 783

Donald Y. M. Leung and Larry Borish

CONGENITAL DISEASES OF CUTANEOUS TISSUES

Epidermolysis Bullosa Simplex

775

James T. Elder and John J. Voorhees

THOMAS S. KUPPER

74

The Skin as a Vehicle for Gene Therapy

Daniel B. Dubin and Saumyen Sarkar

IX. DERMATOLOGY

73

749

Paul Nghiem and Thomas S. Kupper

87

Introduction to Selected Epidermal Gene Mutations

Xeroderma Pigmentosum and Related Disorders

Thomas S. Kupper

Kim E. Nichols and Daniel A. Haber

72

745

Soosan Ghazizadeh, Tadeusz M. Kolodka, and Lorne B. Taichman

659 665

Basal Cell Nevus Syndrome

W. Clark Lambert, Hon-Reen Kuo, and Muriel W. Lambert

Steven C. Hebert

68

737

Ervin Epstein, Jr.

M. Amin Arnaout

66

Oculocutaneous Albinism Jean L. Bolognia

Vikas P. Sukhatme

65

729

94

Muscular Dystrophies Eric P. Hoffman

859

xii

95

CONTENTS

Rhabdomyosarcomas

865

109 110

XI. NEUROLOGY 96

JOSEPH B. MARTIN Molecular Neurobiology

Schizophrenia

989

Ming T. Tsuang and Stephen V. Faraone

Stephen J. Tapscott

Affective Disorders

995

Francis J. McMahon and J. Raymond DePaulo, Jr.

111 871

Alcoholism

1005

Eric J. Devor and Arthur Falek

Joseph B. Martin and Frank M. Longo

97

Huntington’s Disease

891

Marcy E. MacDonald

98

Molecular Genetics of Alzheimer’s Disease 901 P. H. St. George-Hyslop

99

Amyotrophic Lateral Sclerosis and Related Motor Neuron Diseases 907 Meret E. Cudkowicz and Robert H. Brown, Jr.

100

Spinocerebellar Ataxia and Other Disorders of Trinucleotide Repeats 913 Huda Y. Zoghbi

101

Charcot-Marie-Tooth Disease and Related Peripheral Neuropathies

XIII. GENETIC BASIS OF CONGENITAL MALFORMATIONS 112

Molecular and Genetic Basis of Prion Diseases Stanley B. Prusiner

103

Genetic Basis of Mitochondrial Disease

941

Malignant Hyperthermia and Central Core Disease

949

David H. MacLennan and Beverley A. Britt

105

Retinoblastoma

955

Joan M. O’Brien

106

Neurofibromatosis: Type 1 and Type 2

115

107

Brain Tumors

Fibroblast Growth Factor Receptor-Related Skeletal Disorders: Craniosynostosis and Dwarfism Syndromes 1029 Aarskog-Scott Syndrome

1039

Jerome L. Gorski

116

Beckwith-Wiedemann Syndrome

1047

Ellen R. Elias, Michael R. DeBaun, and Andrew P. Feinberg

117

Prader-Willi and Angelman Syndromes

118

Fragile X Syndrome

1053

Robert D. Nicholls

1063

David L. Nelson

963

119

971

120

Jaime O. Claudio and Guy A. Rouleau

1021

Maximilian Muenke, Clair A. Francomano, M. Michael Cohen, Jr., and Ethylin Wang Jabs

Donald R. Johns

104

Greig Cephalopolysyndactyly Syndrome and Limb Disorders Karl-Heinz Grzeschik

114

921

927

1015

Andrew P. Read

113

James R. Lupski

102

ETHYLIN WANG JABS Waardenburg Syndrome

Down Syndrome

1069

Stylianos E. Antonarakis

Mark A. Israel

The 22q11 Deletion: DiGeorge and Velocardiofacial Syndrome

1079

Deborah A. Driscoll and Beverly S. Emanuel

XII. PSYCHIATRY 108

CHARLES B. NEMEROFF Molecular Mechanisms and Regulating Behavior Paul M. Plotsky and Charles B. Nemeroff

121

Orofacial Clefting

1087

Jacqueline T. Hecht and Susan H. Blanton

122 979

Molecular Genetics of Hearing Disorders

1093

William J. Kimberling

Index

1099

CONTENTS

xiii

Contributors

D. CRAIG ALLRED, MD, Department of Pathology, University of Texas Health Science Center, San Antonio, TX GRANT J. ANHALT, MD, Department of Dermatology, Johns Hopkins University, Baltimore, MD STYLIANOS E. ANTONARAKIS, MD, Division of Medical Genetics, University of Geneva Medical School, Geneve, Switzerland AMIN ARNAOUT, MD, Massachusetts General Hospital, Renal Unit, Charlestown, MA ANDREW ARNOLD, MD, Center for Molecular Medicine; Division of Endocrinology and Metabolism, University of Connecticut Health Center, Farmington, CT DENNIS A. AUSIELLO, MD, Department of Medicine, Massachusetts General Hospital, Boston, MA SHERRI J. BALE, PhD, Genetic Studies Section, Laboratory of Skin Biology, NIAMS, National Institutes of Health, Bethesda, MD MELORA D. BERARDO, MD, Cytopathology Center, Johnson City, TN SUSAN H. BLANTON, MD, Department of Pediatrics, University of Virginia, Charlottesville, VA JEAN L. BOLOGNIA, MD, Department of Dermatology, Yale Medical School, New Haven, CT JOSEPH V. BONVENTRE, MD, PhD, Massachusetts General Hospital, Renal Unit, Charlestown, MA LARRY BORISH, Department of Pediatrics, National Jewish Hospital, Denver, CO KAREN D. BRADSHAW, MD, Department of Obstetrics and Gynecology, University of Texas Southwestern Medical Center, Dallas, TX BEVERLEY A. BRITT, Department of Anesthesia, University of Toronto, Toronto General Hospital, Toronto, Ontario, Canada STEVEN L. BRODY, MD, Washington University School of Medicine, Division of Pulmonary and Critical Care Medicine, St. Louis, MO DENNIS BROWN, MD, Renal Unit, Massachusetts General Hospital, Charlestown, MA ROBERT H. BROWN, JR., MD, Day Neuromuscular Laboratory, Massachusetts General Hospital, East, Charlestown, MA MARIO CASTRO, Division of Pulmonary and Critical Care Medicine, Washington University School of Medicine, St. Louis, MO NICHOLAS A. CATALDO, Department of Ob/Gyn, Stanford University, Stanford, CA

YIU-MO CHAN, Genetics Division, Children’s Hospital, Boston, MA ALVIN J. CHIN MD, Cardiology Division, Children’s Hospital of Philadelphia, PA ANGELA M. CHRISTIANO, PhD, Department of Dermatology, Columbia University, New York, NY GEORGE P. CHROUSOS, MD, FAAP, FACP, NIH Clinical Center, Bethesda, MD JAIME O. CLAUDIO, MSC, Center for Research in Neuroscience, McGill University and Montreal General Hospital, Montreal, Quebec, Canada JOY D. COGAN, Department of Pediatrics, Division of Genetics, Vanderbilt University School of Medicine, Nashville, TN M. MICHAEL COHEN, JR., DMD, PhD, Dalhousie University, Halifax, Nova Scotia, Canada JONATHAN COHN, MD, Department of Medicine, Duke University, Durham, NC GILBERT COTE, PhD, University of Texas, MD Anderson Cancer Center, Houston, TX FINBARR E. COTTER, Molecular Haematology Unit, LRF Centre for Childhood Leukemia, Institute of Child Health, London, UK MERET E. CUDKOWICZ, MD, Day Neuromuscular Laboratory, Neurology Service, Massachusetts General Hospital, Charlestown, MA MARTINA DALY, PhD, Division of Molecular and Genetic Medicine, Royal Hallamshire Hospital, Sheffield, UK DOUGLAS C. DEAN, Division of Pulmonary and Critical Care Medicine Washington University School of Medicine, St. Louis, MO MICHAEL R. DEBAUN, Genetic Epidemiology Branch, National Cancer Institute, Bethesda, MD; Present Address: Department of Pediatrics, Division of Hematology and Oncology, Washington University School of Medicine, St. Louis, MO JEAN DELAUNAY, MD, PhD, Department of Biochemistry and Molecular Biology, Faculte de Medicine Grange-Blanche, Genetique Moleculaire Humaine, Institute Pasteur de Lyon, France J. RAYMOND DEPAULO, JR., MD, Department of Psychiatry and Behavioral Sciences, Johns Hopkins University School of Medicine, Baltimore, MD ERIC J. DEVOR, PhD, Department of Psychiatric Research, University of Iowa College of Medicine, Iowa City, IA

xiii

xiv

CONTRIBUTORS CONTENTS

LUIS A. DIAZ, MD, Dermatology Department, Medical College of Wisconsin, Milwaukee, WI DAVID A. DICHEK, MD, Gladstone Institute of Cardiovascular Disease, University of California, San Francisco, CA JOHN J. DIGIOVANNA, MD, Division of Dermatopharmacology, Department of Dermatology Brown University School of Medicine, Rhode Island Hospital, Providence, RI XIANG DING, PhD, Department of Dermatology, Medical College of Wisconsin, Milwaukee, WI DEBORAH A. DRISCOLL, MD, Children’s Hospital of Philadelphia, PA DANIEL B. DUBIN, MD, Division of Dermatology, Brigham and Women’s Hospital, Boston, MA N. LAWRENCE EDWARDS, MD, Department of Medicine, University of Florida, Gainesville, FL JAMES T. ELDER, MD, PhD, Dermatology and Radiation Oncology (Cancer Biology), Department of Dermatology, University of Michigan Medical Center, Ann Arbor, MI ELLEN R. ELIAS, New England Medical Center Hospitals, Boston, MA; Present Address: Children’s Hospital, Boston, MA SARAH H. ELSEA, PhD, Department of Neurology, Baylor College of Medicine, Houston, TX BEVERLY S. EMANUEL, PhD, Children’s Hospital of Philadelphia, PA ERVIN EPSTEIN, JR., MD, Department of Dermatology, San Francisco General Hospital, San Francisco, CA HENRY F. EPSTEIN, MD, Baylor College of Medicine, Houston, TX JANET A. FAIRLEY, MD, Department of Dermatology, Medical College of Wisconsin, Milwaukee, WI ARTHUR FALEK, PhD, Department of Psychiatry and Behavioral Science, Emory University School of Medicine, Atlanta, GA STEPHEN V. FARAONE, PhD, Department of Psychiatry, Massachusetts Mental Health Center, Boston, MA ANDREW P. FEINBERG, MD, MPH, Johns Hopkins University School of Medicine, Baltimore, MD BETH A. FINE, MS, Graduate Program in Genetic Counseling; Obstetrics & Gynecology, Northwestern University Medical School, Chicago, IL CLAIR A. FRANCOMANO, MD, Medical Genetics Branch, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD ELAINE FUCHS, PhD, Department of Molecular Genetics and Cell Biology, Howard Hughes Medical Institute, University of Chicago, IL ROBERT F. GAGEL, MD, Section of Neoplasia and Hormonal Disorders, University of Texas MD Anderson Cancer Center, Houston, TX GREGORY G. GERMINO, MD, Division of Nephrology, Johns Hopkins University School of Medicine, Baltimore, MD ROBERT E. GERSZTEN, MD, Cardiac Unit and Cardiovascular Research Institute, Massachusetts General Hospital, Charlestown, MA SOOSAN GHAZIZADEH, Department of Oral Biology and Pathology, SUNY at Stony Brook, NY GEORGE J. GIUDICE, PhD, Department of Dermatology, Medical College of Wisconsin, Milwaukee, WI

LOWELL A. GOLDSMITH, MD, University of Rochester School of Medicine and Dentistry, Rochester, NY MAUREEN M. GOODENOW, Department of Pediatrics, University of Florida College of Medicine, Gainesville, FL ANNE GOODEVE, PhD, Division of Molecular and Genetic Medicine, Royal Hallamshire Hospital, Sheffield, UK JEROME L. GORSKI, MD, University of Michigan Medical School, Ann Arbor, MI PIET C. DE GROEN, MD, Center for Basic Research in Digestive Diseases, Mayo Medical School, Clinic, and Foundation, Rochester, MN MARCUS GROMPE, MD, Department of Molecular and Medical Genetics, Oregon Health Sciences University, Portland, OR KARL-HEINZ GRZESCHIK, PhD, Institute for Humangenetik und Humangenetik Polikinik, Bahnhofstrasse, Marburg, Germany SANJEEV GUPTA, MBBS, MD, MRCP, Liver Research Center, Albert Einstein College of Medicine, Bronx, NY DANIEL A. HABER, MD, PhD, Massachusetts General Hospital Cancer Center, Charlestown, MA ANDREW HAYNES, DM, MRCP, MRCPath, Department of Haematology, Nottingham City Hospital and University of Nottingham, UK STEVEN C. HEBERT, MD, Renal Division, Department of Medicine, Brigham and Women’s Hospital, Boston, MA; Present Address: Division of Nephrology, Vanderbilt University Medical Center, Nashville, TN JACQUELINE T. HECHT, PhD, Department of Pediatrics, University of Texas Medical School, Houston, TX PIRKKO HEIKKILÄ, MSc, Division of Matrix Biology, Department of Medical Biochemistry and Biophysics, Karolinska Institute, Stockholm, Sweden ERIC P. HOFFMAN, PhD, Department of Molecular Genetics, University of Pittsburgh School of Medicine, Pittsburgh, PA MICHAEL J. HOLTZMAN, MD, Division of Pulmonary and Critical Care Medicine, Washington University School of Medicine, St. Louis, MO RICHARD HONG, MD, Genetics Laboratory, University of Vermont, Burlington, VT MICHAEL F. IADEMARCO, Division of Pulmonary and Critical Care Medicine, Washington University School of Medicine, St. Louis, MO MARK A. ISRAEL, MD, Department of Neurological Surgery, Preuss Laboratory for Molecular Neuro-oncology, Brain Tumor Research Center, University of California, San Francisco, CA HOWARD J. JACOB, PhD, Department of Physiology, Medical College of Wisconsin, Milwaukee, WI J. LARRY JAMESON, MD, PhD, Division of Endocrinology, Metabolism, and Molecular Medicine, Northwestern University Medical School, Chicago, IL DONALD R. JOHNS, MD, Division of Neuromuscular Disease, Beth Israel Hospital, Harvard Medical School, Boston, MA WADE JOHNSON, Division of Endocrinology, Metabolism, and Molecular Medicine, Northwestern University Medical School, Chicago, IL ROBERT W. KARR, MD, Searle Discovery Research, Monsanto, St. Louis, MO WILLIAM J. KIMBERLING, PhD, Boys Town National Research Hospital, Omaha, NE

CONTRIBUTORS CONTENTS

GEORGE KOIKE, Medical College of Wisconsin, Department of Physiology, Milwaukee, WI TADEUSZ M. KOLODKA, Department of Oral Biology and Pathology, SUNY at Stony Brook, NY PETER KOPP, MD, Division of Endocrinology, Metabolism, and Molecular Medicine, Northwestern University Medical School, Chicago, IL T. RAJENDRA KUMAR, PhD, Department of Pathology, Baylor College of Medicine, Houston, TX HON-REEN KUO, PhD, New Jersey Medical School, Newark, NJ THOMAS S. KUPPER, MD, Dermatology Division, Brigham and Women’s Hospital, Boston, MA SANTIAGO LAMAS, MD, PhD, Department of Protein Structure and Function, Center for Biological Investigation, Madrid, Spain MURIEL W. LAMBERT, New Jersey Medical School, Newark, NJ W. CLARK LAMBERT, MD, PhD, New Jersey Medical School, Newark, NJ DAVID A. LANE, Department of Haematology, Charing Cross & Westminster Medical School, London, UK NICHOLAS F. LARUSSO, MD, Center for Basic Research in Digestive Diseases, Mayo Medical School, Clinic, and Foundation, Rochester, MN E. CARWILE LEROY, MD, Chairman, Department of Microbiology and Immunology, Medical University of South Carolina, Charleston, SC DONALD Y. M. LEUNG, MD, Department of Pediatrics, National Jewish Hospital, Denver, CO BARRY LONDON, MD, PhD, Division of Cardiology, University of Pittsburgh Medical Center, Pittsburgh, PA FRANK M. LONGO, MD, PhD, Veterans Affairs Medical Center, San Francisco, CA DWIGHT C. LOOK, Division of Pulmonary and Critical Care Medicine, Washington University School of Medicine, St. Louis, MO WILLIAM L. LOWE, JR., MD, Division of Endocrinology, Metabolism, and Molecular Medicine, Northwestern University Medical School, Chicago, IL JAMES R. LUPSKI, MD, PhD, Molecular and Human Genetics; Pediatrics Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, TX LUCIO LUZZATTO, Department of Human Genetics, Memorial Sloan Kettering Cancer Center, New York, NY SUSAN M. MACDONALD, Department of Medicine, Johns Hopkins Asthma and Allergy Center, Baltimore, MD DAVID H. MACLENNAN, PhD, FRS, Banting and Best Department of Medical Research, University of Toronto, Charles H. Best Institute, Toronto, Ontario, Canada CALUM MACRAE, Brigham and Women’s Hospital, Boston, MA LAIRD D. MADISON, MD, PhD, Division of Endocrinology, Metabolism, and Molecular Medicine, Northwestern University Medical School, Chicago, IL MARCO MARCELLI, MD, Veteran Administration Medical Center, Houston, TX JOSEPH B. MARTIN, MD, PhD, Dean, Faculty of Medicine, Boston, MA

xv

MARTIN M. MATZUK, Departments of Pathology, Cell Biology, and Molecular and Human Genetics, Baylor College of Medicine, Houston, TX MARCY MCDONALD, PhD, Molecular Neurogenetics Unit, Massachusetts General Hospital, Charlestown, MA ELIZABETH A. MCGEE, MD, Department of OB/GYN, University of Texas Southwestern Medical Center, Dallas, TX; Present Address: Department of OB/GYN, Stanford University Medical Center, Stanford, CA FRANCIS J. MCMAHON, MD, The Johns Hopkins University School of Medicine, Baltimore, MD MICHAEL J. MCPHAUL, MD, Department of Internal Medicine, University of Texas Southwestern Medical Center, Dallas, TX SHLOMO MELMED, MD, Cedars Sinai Medical Center, Los Angeles, CA THOMAS MICHEL, MD, PhD, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA MAXIMILIAN MUENKE, MD, Division of Human Genetics and Molecular Biology, Children’s Hospital of Philadelphia, PA; Present Address: Medical Genetics Branch, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD DIYA F. MUTASIM, MD, Department of Dermatology, University of Cincinnati College of Medicine, Cincinnati, OH KHEDOUDJA NAFA, Department of Human Genetics, Memorial Sloan Kettering Cancer Center, Memorial Hospital, Sloan Kettering Institute, New York, NY DAVID L. NELSON, PhD, Baylor College of Medicine, Houston, TX CHARLES B. NEMEROFF, MD, PhD, Department of Psychiatry and Behavioral Medicine, Emory University School of Medicine, Atlanta, GA MARIA I. NEW, MD, Department of Pediatrics, The New York Hospital, Cornell Medical Center, New York, NY PAUL NGHIEM, MD, PhD, Dermatology, Brigham and Women’s Hospital, Boston, MA LYNDA Q. NGUYEN, Division of Endocrinology, Metabolism, and Molecular Medicine, Northwestern University Medical School, Chicago, IL ROBERT D. NICHOLLS, DPhil, Department of Genetics, Case Western Reserve University, Cleveland, OH KIM E. NICHOLS, MD, Massachusetts General Hospital Cancer Center, Charlestown, MA ROSARIO NOTARO, Department of Human Genetics, Memorial Sloan Kettering Cancer Center, New York, NY JOAN M. O’BRIEN, MD, Department of Ophthalmology, University of California, San Francisco, CA PETER O’CONNELL, MD, Department of Pathology, University of Texas Health Sciences Center, San Antonio, TX ROBIN OLDS, Department of Pathology, University of Otago, Dunedin, New Zealand ERIC N. OLSON, PhD, Department of Biochemistry/Molecular Biology, MD Anderson Cancer Center, University of Texas, Houston, TX LUIZ F. ONUCHIC, MD, PhD, Division of Nephrology, Johns Hopkins University School of Medicine, Baltimore, MD

xvi

CONTRIBUTORS CONTENTS

MICHAEL OTT, MD, Albert Einstein College of Medicine, Bronx, NY AMY S. PALLER, MD, Children’s Memorial Hospital, Chicago, IL PRAGNA I. PATEL, PhD, Department of Neurology, Baylor College of Medicine, Houston, TX IAN PEAKE, Division of Molecular and Genetic Medicine, Royal Hallamshire Hospital, Sheffield, UK RICHARD G. PESTELL, MD, FRACP, PhD, Department of Medicine and Developmental and Molecular Biology, The Albert Einstein Cancer Center, Albert Einstein College of Medicine, Bronx, NY JOHN A. PHILLIPS, MD, Pediatrics & Biochemistry, Vanderbilt University School of Medicine, Nashville, TN PAUL M. PLOTSKY, PhD, Department of Psychiatry and Behavioral Sciences, Emory University School of Medicine, Atlanta, GA STANLEY B. PRUSINER, MD, Departments of Neurology and of Biochemistry and Biophysics, University of California, San Francisco, CA CHING-HON PUI, St. Jude’s Children’s Research Hospital, Memphis, TN CHARMIAN A. QUIGLEY, MBBS, Riley Children’s Hospital, Indianapolis, IN ANDREW P. READ, MA, PhD, Department of Medical Genetics, St. Mary’s Hospital, Manchester, UK JAMES C. REYNOLDS, MD, Division of Gastroenterology and Hepatology, Allegheny University of the Health Sciences, Philadelphia, PA C. SUE RICHARDS, PhD, Department of Molecular and Human Genetics, Baylor DNA Diagnostic Laboratory, Baylor College of Medicine, Houston, TX JACQUES ROCHETTE, BM, DPharm, DSc, MRC Molecular Haemotology Unit, Institute of Molecular Medicine, John Radcliff Hospital, Headington, Oxford, UK; Present Address: Medical Genetics, Faculte de Medecine, Universite Jules Verne, Amiens, France DANIEL B. ROSENBLUTH, MD, Division of Pulmonary and Critical Care Medicine, Washington University School of Medicine; Adult Cystic Fibrosis Clinic, Barnes-Jewish Hospital, St. Louis, MO ANTHONY ROSENZWEIG, MD, Cardiac Unit and Cardiovascular Research Institute, Massachusetts General Hospital, Charlestown, MA BRUNO ROTOLI, Hematology, Federico II University Medical School, Napoli, Italy GUY A. ROULEAU, MD, PhD, Montreal General Hospital, Montreal, Quebec, Canada DEBORAH C. RUBIN, MD, Division of Gastroenterology, Washington University School of Medicine, St. Louis, MO JEFFREY E. RUBNITZ, St. Jude’s Children’s Research Hospital, Memphis, TN JUAN RUIZ, MD, PhD, Centro de Investigaciones Biomedicas, Facultidad de Medicina, Universidad de Navaria, Pamplona, Spain NIGEL RUSSELL, MD, FRCP, FRCPath, Department of Hematology, Nottingham City Hospital and University of Nottingham, UK DEEPAK SAMPATH, Division of Pulmonary and Critical Care Medicine, Washington University School of Medicine, St. Louis, MO

SAUMYEN SARKAR, PhD, Brigham and Women’s Hospital, Boston, MA CHRISTINE E. SEIDMAN, MD, Department of Genetics, Harvard Medical School, Boston, MA J. G. SEIDMAN, PhD, Brigham and Women’s Hospital, Boston, MA ROBERT M. SENIOR, MD, Pulmonary and Critical Care Medicine, Washington University School of Medicine, St. Louis, MO STEVEN D. SHAPIRO, MD, Washington University School of Medicine, St. Louis, MO SARAH SHEFELBINE, University of Texas, MD Anderson Cancer Center, Houston, TX BARBARA A. DA SILVA, MD, Endocrinology, Metabolism, and Molecular Medicine, Northwestern University Medical School, Chicago, IL JOHN W. SLEASMAN, Department of Pediatrics, University of Florida, Gainesville, FL EDWIN A. SMITH, MD, Division of Rheumatology and Immunology, Department of Medicine, Medical University of South Carolina, Charleston, SC ERIC SOBEL, MD, Division of Rheumatology, Department of Medicine, University of Florida, Gainesville, FL RICHARD D. SONTHEIMER, MD, Department of Dermatology, University of Texas Southwestern Medical Center, Dallas, TX ANDREW F. STEWART, MD, Division of Endocrinology and Metabolism, West Haven VA Medical Center, West Haven, CT; Present Address: Division of Endocrinology, University of Pittsburgh Medical Center, Pittsburgh, PA P. H. ST. GEORGE-HYSLOP, MD, FRCP, Centre for Research in Neurodegenerative Diseases, University of Toronto, Ontario, Canada PETER M. STEINERT, PhD, Laboratory of Skin Disease, National Institutes of Health, Bethesda, MD CONSTANTINE A. STRATAKIS, Section on Pediatric Endocrinology, National Institutes of Health, Bethesda, MD VIKAS P. SUKHATME, MD, PhD, Renal Division, Beth Israel Deaconess Medical Center, Boston, MA LORNE B. TAICHMAN, PhD, Department of Oral Biology and Pathology, SUNY at Stony Brook, NY STEPHEN J. TAPSCOTT, MD, PhD, Fred Hutshinson Cancer Center, Seattle, WA SWEE LAY THEIN, MRCP, FRCPath, MRC Molecular Haematology Unit, Institute of Molecular Medicine, John Radcliffe Hospital, Headington, Oxford, UK PAMELA M. THOMAS, MD, University of Michigan Medical Center, Ann Arbor, MI KARL TRYGGVASON, MD, PhD, Division of Matrix Biology, Department of Medical Biochemistry and Biophysics, Karolinska Institute, Stockholm, Sweden MING T. TSUANG, MD, PhD, DSc, Department of Psychiatry, Massachusetts Mental Health Center, Boston, MA JOUNI UITTO, MD, PhD, Department of Dermatology and Cutaneous Biology, Thomas Jefferson University, Philadelphia, PA GIUSEPPE VASSALLI, Gladstone Institute of Cardiovascular Disease, University of California, San Francisco, CA JOHN J. VOORHEES, MD, Dermatology Department, University of Michigan Medical Center, A. Alfred Taubman Health Center, Ann Arbor, MI

CONTRIBUTORS CONTENTS

PATRICIA A. WARD, MS, Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, TX ETHYLIN WANG JABS, MD, Center for Medical Genetics, Johns Hopkins Hospital, Baltimore, MD TIMOTHY E. WEAVER, Division of Pulmonary Biology, Children’s Hospital Medical Center, Cincinnati, OH STEVEN JAY WEINTRAUB, MD, Departments of Internal Medicine and Cell Biology and Physiology, Washington University School of Medicine, St. Louis, MO DAVID WHITCOMB, MD, PhD, University of Pittsburgh, PA JEFFREY A. WHITSETT, MD, Division of Pulmonary Biology, Children’s Hospital Medical Center, Cincinnati, OH RALPH C. WILLIAMS, JR., MD, Division of Rheumatology, Department of Medicine, University of Florida, Gainesville, FL

xvii

ROBERT C. WILSON, PhD, Department of Pediatrics, Cornell University Medical College, New York, NY PETER WINSHIP, DPhil, Division of Molecular and Genetic Medicine, Royal Hallamshire Hospital, Sheffield, UK GEORGE Y. WU, MD, PhD, GI Division, University of Connecticut School of Medicine, University of Connecticut Health Center, Farmington, CT ANDREW R. ZINN, MD, PhD, McDermott Center, University of Texas Southwestern Medical Center, Dallas, TX HUDA ZOGHBI, MD, Howard Hughes Medical Institute, Department of Pediatrics, Molecular and Human Genetics, and Neurology, Baylor College of Medicine, Houston, TX

xviii

CONTENTS

CONTENTS

xix

Color Plates

Color plates appear as an insert following p. 684.

Plate 6 (Figs. 1–4 from Chapter 76). Fig. 1. Hyperkeratotic irregular papules in Darier’s disease on sunexposed sites. Fig. 2. Hemorrhagic vesicle-papules in kindred with Darier’s disease. Fig. 3. Red and white linear longitudinal bands in nails of patient with Darier’s disease. Fig. 4. Papules and erosions characterizing HaileyHailey disease. Plate 7 (Fig. 1A–C from Chapter 73). Clinical pictures of EBS patients. Plate 8 (Fig. 7 from Chapter 92). Expression pattern of a myogenin-lacZ transgene in an 11.5-day mouse embryo. Plate 9 (Fig. 2 from Chapter 92). Activation of MHC expression in fibroblasts expressing exogenous MyoD. Plate 10 (Fig. 3 from Chapter 92). Schematic representation of a MyoD/E12 heterodimer.

Plate 1 (Fig. 4 from Chapter 68). Expression of βgalactosidase in porcine kidneys following in vivo perfusion with an adenovirus containing the β-galactosidase reporter gene under the cytomegalovirus promoter. Plate 2 (Fig. 2B from Chapter 75). Lesional skin from the epidermal nevus, showing normal basal cells, but hyperkeratosis and cytolysis in suprabasal cells. Plate 3 (Fig. 1 from Chapter 12). Fetal circulation in the human; neonatal circulation in the human. Plate 4 (Fig. 3 from Chapter 14). Expression of VCAM-1 at sites of atherogenesis. Plate 5 (Fig. 2 from Chapter 29). The “open ends” of class II molecule.

xix

1

CHAPTER 1 / HUMAN GENOME, CHROMOSOMES, AND GENES

INTRODUCTION TO MOLECULAR MEDICINE SECTION EDITORS:

ANDREA BALLABIO AND J. LARRY JAMESON

I

2

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

CHAPTER 1 / HUMAN GENOME, CHROMOSOMES, AND GENES

3

1 Organization of the Human Genome, Chromosomes, and Genes SARAH H. ELSEA AND PRAGNA I. PATEL INTRODUCTION

length. One can imagine that such an enormous amount of DNA would be difficult to contain in the nucleus of a single human cell. Indeed, an elaborate system of packaging/folding (or supercoiling) is necessary for replication, transcription, recombination, and maintenance of the DNA from one generation to the next. As illustrated in Fig. 1-1, during this very orderly process, basic proteins (called histones) provide a core around which the DNA is wrapped, forming nucleosomes. Nucleosomes are not formed randomly but rather are phased along the DNA such that the position of a nucleosome along any given stretch of DNA is the same from cell to cell. These nucleosomes are further organized into higher order solenoid structures, which are packed into loops attached to a nonhistone protein scaffold. As the cell cycle progresses into prophase, the loops thicken into chromomeres, which can be seen as densely staining regions of the chromosomes. This thickening process continues and eventually the DNA condenses into the highly organized ultrastructure of mitotic chromosomes, thus compacting the DNA ~8000-fold at metaphase (Figs. 1-1 and 1-2). This higher order packaging process not only compacts the DNA within the nucleus but is also important for gene regulation and protection of the DNA from nucleases. The higher order structure of the chromatin is fluid and dynamic, with condensation/decondensation of the DNA occurring throughout the cell cycle (Fig. 1-2). There are several classes of DNA within the human genome. In general, less than 10% of the genome actually encodes genes. The more densely packed regions of the chromosomes, particularly near the centromeres, are termed heterochromatin and are not transcribed, whereas transcriptionally active sequences are located within less densely packed regions termed euchromatin. Approximately 75% of the genome is made of unique or single-copy DNA, whose sequences are represented only once per haploid genome (haploid refers to half the complement of chromosomes; in other words, the DNA donated by either a sperm or an egg). In contrast, the remaining portion of the genome consists of several classes of repetitive DNA, whose sequences are present anywhere from two to as high as 107 copies per haploid genome. While it is not known what role repetitive DNA plays in the cell, it is thought to either function in maintenance of chromosome structure or gene regulation, or, in fact, it may have no function at all. Even though unique DNA makes up the majority of the sequences in the genome, only a small proportion of this DNA actually codes for proteins (coding regions of genes). Some of the

DNA is the chemical basis of heredity. Within its sequence is the information necessary for cells to live, grow, differentiate, and replicate. It is the DNA that provides both consistency (all humans generally look the same) and variability (height, eye, and hair color) among organisms. While the human genome is extremely polymorphic and the majority of deviations in DNA sequence are thought to be benign, variations in its sequence can and do lead to genetic disorders. Although we currently understand the roles of only a small percentage of the total number of genes, great strides are being made toward elucidating the physical and molecular structure and function of the human genome. Through this knowledge we can more fully appreciate the complex physiology of the human organism. With the evergrowing number of diseases thought to have a genetic basis, it is important to understand the structure and function of DNA and chromosomes, as well as the genes they encode. This knowledge will enhance determination of the gene(s) underlying a particular disease as well as design of potential treatments and cures for the disease. This chapter will provide a basis for general understanding of the principles behind the field of molecular genetics.

GENOME ORGANIZATION Human DNA is packaged in 23 pairs of chromosomes, specifically 22 homologous pairs of chromosomes (autosomes numbered 1–22) and two sex chromosomes, XX in females and XY in males; this represents the diploid genome. The human genome consists of approximately 6 billion bp of DNA and is estimated to contain 50,000–100,000 genes. For a chromosome to function properly, it must be able to replicate, segregate, and maintain itself from one generation to the next. These fundamental processes require that a chromosome contain at least one origin of replication, a centromere for proper segregation of daughter chromosomes, and a telomere for complete replication and maintenance of the DNA (Fig. 1-1). Each chromosome is comprised of a single linear duplex DNA molecule that is complexed with numerous proteins. This DNA– protein complex is termed chromatin during interphase of the cell cycle (Fig. 1-2). If the DNA contained in a single cell were stretched from end to end, it would measure about two meters in From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

3

4

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Figure 1-1 Chromatin structure. Chromatin is a complex of DNA and proteins. DNA is wrapped around histones and other nonhistone proteins in an orderly fashion to form nucleosomes, which are then further organized into solenoid structures and finally form the highly condensed metaphase chromosome.

Figure 1-2 Condensation/decondensation of DNA in the cell cycle. The interphase nucleus contains chromatin that is dispersed throughout the nucleus. In prophase, the replicated chromosomes begin to condense, and, by metaphase, the chromatin has condensed ~8000-fold.

DNA functions in the regulation of these genes (noncoding sequences within the gene). Long stretches (>25 kb) of unique DNA are rare, and these single-copy sequences are often found interspersed between stretches of repetitive sequences. Repetitive DNA can either be clustered within or dispersed throughout the genome. Clustered repeat sequences comprise 10–15% of the genome and comprise arrays of short repeats organized in a head-to-tail fashion. These tandem repeats are collectively called satellite DNAs. These repeats can vary in length from a few nucleotides to several million. Their location varies as well, with some repeats found only in the heterochromatin of the centromeres or telomeres and some found only on particular chromosomes. Minisatellite DNA repeats are generally 15–465 bp in length but can run up to ~20 kb. These repeats are typically found along the length of each chromosome and are used as DNA markers for mapping of the genome as well as for chromosome analysis. Another class of repeat sequences includes related sequences that are dispersed throughout the genome (sometimes within genes), constituting approximately 15% of the total DNA. While several subclasses of these repeats exist, two are of particular importance. SINES are short interspersed sequences. The best characterized of these elements are those belonging to the Alu family, so named because these repeats contain an AluI restriction

site near the middle of the sequence. SINES, while not precisely conserved, are of similar sequence, ~300 bp in length, and are present in about 300,000–500,000 copies per haploid genome, accounting for ~3–6% of the genome. Although the function of these repeats is not known, the sequence is related to that of 7SL RNA. In addition, Alu sequences are of medical importance because these repeats may act as transposable elements, causing aberrant recombination between different Alu family members resulting in mutations that may lead to genetic disease. A second major class of repetitive DNA is termed LINES, or long interspersed sequences of DNA. A well-represented member of this family is L1, which is found in ~20,000–50,000 copies per human genome, runs about 6 kb in length, and makes up ~3–4% of the total DNA in a cell. A portion of the L1 sequence is similar to the viral enzyme reverse transcriptase. These sequences have also been implicated in the cause of several mutations in hereditary disease due to their potential function as transposable elements.

GENE STRUCTURE AND ORGANIZATION As discussed, DNA is the hereditary material of the cell. Its sequence encodes genes which, in turn, code for RNAs and ultimately, proteins. Basically, DNA is transcribed into RNA, which is then processed and translated into protein (Fig. 1-3A).

CHAPTER 1 / HUMAN GENOME, CHROMOSOMES, AND GENES

5

genes. As a result, many genes are related to others and belong to a larger family of genes with related sequence, structure, and/or function. There are several examples of gene families found either dispersed throughout or sometimes in gene clusters in the human genome. These include genes for globin, myosin, visual pigment, seven-transmembrane-spanning receptors, small nuclear RNAs, and immunoglobulins. Just as functional genes are found in the genome, so are related genes that either have lost their function or were not fully functional from the start. These sequences, referred to as pseudogenes, are thought to be byproducts of evolution and can arise as a result of incomplete duplication events, mutations, or from processed or unprocessed transcripts becoming incorporated into the DNA. They are abundant throughout the genome.

CYTOGENETICS

Figure 1-3 How does a gene work? (A) DNA is replicated and transcribed into RNA, which is then processed to remove introns. Translation then produces a functional protein. (B) Diagram of the basic aspects of a gene necessary for proper transcription and translation of the DNA.

Proteins were originally thought to be encoded by continuous segments of DNA, but actually very few genes contain uninterrupted coding regions. The vast majority of genes are interrupted by one or, typically, several noncoding regions called intervening sequences or introns (Fig. 1-3B). Introns are transcribed into RNA but are removed when it is processed into its mature form (mRNA); therefore, the sequences contained in the intron are not represented in the translated product. The coding sequences, called exons, actually dictate the protein sequence. A schematic representing the features of a typical gene are shown in Fig. 1-3. Genes have been found that range in size from ~100 bp to ~2.3 Mb, but most genes are thought to range between 1 and 200 kb. As illustrated in Fig. 1-3, genes not only include the coding region but also include surrounding sequences that may help in regulation of that gene’s expression. These sequences include: “start” and “stop” signals for both transcription and translation; the promoter region, which lies 5' to the coding region and helps regulate transcription; the 5'-untranslated region (5'-UTR), which may function in regulation; and the 3'-untranslated region (3'-UTR), which contains the polyadenylation signal important for maturation of mRNA and also may be involved in other aspects of RNA processing, transport, degradation, and translation. New genes arise during evolution as a consequence of duplication, divergence, recombination, and mutations in old or existing

While the distribution of genes and repetitive sequences across the genome is not uniform, chromosomes do form consistent and interesting patterns in their organization. Chromosomes can be stained and examined microscopically to reveal distinct light and dark banding patterns. For example, it is thought that metaphase bands that stain lightly with Giemsa (termed G-banding) harbor more housekeeping and tissue-specific genes, SINES, and GC-rich regions of DNA, whereas the darker-staining Giemsa bands tend to contain fewer genes, more LINES, and have a lower GC content. Chromosomal abnormalities contribute significantly toward congenital malformations and are responsible for >50% of all spontaneous abortions/miscarriages. In addition, ~0.7% of newborns (and ~2% of live births in women over 35 years of age) have significant chromosomal abnormalities. Furthermore, many cancers result from these types of aberrations as well. Chromosome analysis is indicated under several circumstances, including multiple miscarriages, fertility problems, hematological malignancies, multiple congenital anomalies, sexual differentiation disorders, and/or if there is a known or suspected abnormality, such as a family history of mental retardation in males. Cytogenetic analysis of chromosomes in peripheral blood lymphocytes has facilitated the identification of many chromosomal abnormalities. In these analyses, the lymphocytes are blocked at metaphase and stained with Giemsa. This process results in the light and dark staining mentioned above. G-banding gives ~350–550 bands per haploid set, where 1 band is equal to ~5– 10 million bp, representing a few to many genes. Alternatively, a prophase block can give a higher resolution of ~850 bands. Basic chromosome morphology is illustrated in Figs. 1-4 and 1-5. The short arm is termed “p” for petite and the long arm is designated “q.” Chromosomes are acrocentric if the centromere lies near the end of the chromosome, metacentric if the centromere lies near the middle, and submetacentric if the centromere is located between the center and the end of the chromosome. The numbering of the bands on each chromosome is from the centromere to the telomere on each the short and the long arms. The chromosomes in a single cell are arranged by number into karyotypes, which provide a visual representation of an individual’s chromosomes (Fig. 1-4). Karyotypes may reveal deletions, duplications, and/or rearrangements and can help to localize the region of DNA responsible for the genetic disorder in question. The nomenclature used for describing chromosome morphology is detailed in Table 1-1.

6

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Figure 1-4 A human karyotype of G-banded chromosomes. A normal female (46, XX) karyotype is shown. Chromosomes are placed in groups based on their size and structure. (Karyotype kindly provided by Lisa Shaffer, PhD.) Table 1-1 Karyotype Terminology 1–22 X,Y cen del der dic du fra i ins inv mar mat p pat q r rcp t ter + or –

Figure 1-5 Fluorescence in situ hybridization to metaphase chromosomes. Pictured is a metaphase spread of a patient with Smith-Magenis syndrome. This syndrome results from an interstitial deletion on chromosome 17p11.2. Controls probes for 17q are present on each chromosome 17, while the diagnostic probe for Smith-Magenis syndrome is present on only one chromosome at 17p11.2.

: :: /

Autosome numbers Sex chromosomes Centromere Deletion Derivative of a chromosome Dicentric chromosome Duplication Fragile site Isochromosome Insertion Inversion Marker chromosome Maternal origin Short arm of chromosome Paternal origin Long arm of chromosome Ring chromosome Reciprocal translocation Translocation Terminus (i.e., pter or qter) Placed before a chromosome number, indicates addition or loss of entire chromosome, as in +21 for Down syndrome. Placed after the chromosome number, indicates gain or loss of a part of that chromosome, as in 5q- for Cri-du-Chat. Breakage Breakage and reunion Mosaicism (i.e., 46/47 indicates a mixed cell population in which some cells have 46 and others have 47 chromosomes)

CHAPTER 1 / HUMAN GENOME, CHROMOSOMES, AND GENES

There are several relatively common types of chromosomal abnormalities. The most common aberration is aneuploidy, which can mean either an extra chromosome or a deleted chromosome. Normal human genotypes have 46 chromosomes, including two sex chromosomes (46, XX for females; 46, XY for males). Down syndrome (47, XX, +21) would be an example of aneuploidy or trisomy 21, while Turner syndrome (45, X) is an example of monosomy. Other gross abnormalities include uniparental disomy (both chromosomes in a pair come from a single parent), marker chromosomes (extra chromosomal pieces or supernumerary chromosomes), ring chromosomes (partial chromosomes containing centromeres resulting from the deletion of the telomeric ends of the p and q arms), and mosaicism (where a fraction of the cells in the body have normal chromosomes, while the remaining cells carry an aberration; this may or may not produce a phenotype). Structural problems can also result from deletions, duplications, translocations, inversions, and unequal crossing-over.

MOLECULAR CYTOGENETICS Preparations of intact chromosomes can also be used for molecular genetic diagnosis, analysis of gene deletions and duplications, and mapping of genes within the genome. Fluorescence in situ hybridization (FISH) uses fluorescently tagged

7

DNA probes and fluorescence microscopy to determine the relative location of that particular DNA sequence (Fig. 1-5). This allows for localization of the DNA probe to subchromosomal regions of the genome that G-banding cannot determine. Standard FISH analysis with metaphase chromosomes provides resolution to within 5–10 Mb and serves to bridge the gap between standard cytogenetic analysis and detailed molecular analysis using purified DNA. This technique is rapidly improving. With the use of multicolor probes and interphase chromosomes (where the DNA is less compact), resolution of probes can be obtained to within 100 kb. Additionally, sets of markers can be applied that effectively “paint” the chromosomes so that it can be determined if small segments of DNA have been rearranged and to diagnose aneuploidies. FISH is fast becoming important for long-range genome mapping and ordering of markers within a segment of DNA.

SELECTED REFERENCES Alberts B, Bray D, Lewis J, Raff M, Roberts K, Watson JD. Molecular Biology of the Cell. 2nd ed. New York: Garland, 1989. Dracopoli NC, Haines JL, Korf BR, et al. Current Protocols in Human Genetics. New York: Wiley, 1995. Lewin B. Genes VI. New York: Oxford University Press, 1996. Scriver CR, Beaudet A, Sly WS, Valle D. In: Metabolic and Molecular Bases of Inherited Disease. 7th ed. New York: McGraw-Hill, 1995.

CHAPTER 2 / RECOMBINANT DNA AND GENETIC TECHNIQUES

2

9

Recombinant DNA and Genetic Techniques MARCUS GROMPE, WADE JOHNSON, AND J. LARRY JAMESON

INTRODUCTION

sequence-specific manner. Biologically, the function of these enzymes is to “restrict” the entry of foreign DNA by cleaving it at recognition sites that do not exist in the host bacterium. Each enzyme has a specific sequence that it recognizes, binds, and cuts. These recognition sites usually consist of a code of 4–8 bp, and the restriction endonucleases will only cut sequences that perfectly match this code. For example, the recognition sequence for the enzyme EcoRI is GAATTC. All sites with this sequence will be cleaved by the enzyme (Fig. 2-1), but even a single base mismatch, such as CAATTC, will prevent DNA digestion. The chance of finding a specific restriction site in a given piece of DNA is largely dependent on the number of bases in the recognition site. Based on the probability that the four nucleotides (GATC) will create a characteristic restriction site, a 4-base cutter should cut human genomic DNA every 2.5 × 102 bp, whereas an 8-base cutter would cut on average of every 1.0 × 106 bp. Restriction enzymes can cut DNA in three basic ways: they can cut the recognition site asymmetrically to leave ends with 5' or 3' overhangs, or they can cut symmetrically to leave blunt ends (Fig. 2-1). Restriction enzymes are named according to the bacteria from which they were isolated (e.g., EcoRI from Escherichia coli). Hundreds of different restriction endonucleases are commercially available today. In practical terms, the restriction enzyme is incubated with DNA in a compatible buffer solution and cuts the DNA into pieces of a defined size. Because the cleavage is sequence-specific, different DNAs will produce a characteristic digestion pattern (Fig. 2-2). These unique DNA fingerprints can be recognized after separation of the different size fragments by gel electrophoresis. The production of such restriction maps represents one of the major uses of these enzymes. A second major use of restriction endonucleases is to “cut and paste” DNA in order to generate novel DNA sequences that are used in a variety of methods. DNA LIGASES DNA ligases are enzymes that attach two pieces of DNA covalently to one another. T4 DNA ligase is the most commonly used enzyme. The reaction occurs between the 3' hydroxy group of one DNA strand and the phosphate group of the partner strand (Fig. 2-3). ATP provides the energy for the reaction. DNA ligation can occur efficiently only between compatible ends of DNA fragments, and this specificity gives the investigator some control over how different DNA fragments ligate together in a reaction mix. The most common use of ligation is to incorporate a piece of DNA into a cloning vector.

Although DNA was recognized as the chemical substrate of heredity in the early 1940s, studies to examine its structure and function were historically quite difficult because of the lack of tools for its isolation and experimental manipulation. In the early 1970s, a series of technological breakthroughs began that dramatically changed our ability to study the role of DNA in biology and genetics. The impact of methodologic advances in the fields of molecular biology and genetics cannot be overemphasized. Since the mid-1970s, eight Nobel prizes have been awarded for research that led directly or indirectly to major methodological advances in addition to their profound new insights into biology and chemistry. Examples include the discovery of reverse transcriptase, restriction enzymes, plasmid cloning vectors, DNA sequencing, PCR, and others (Table 2-1). The methods available for research in molecular biology and genetics continue to increase rapidly. Although many of these techniques are relatively complex, a basic knowledge of a limited number of basic procedures will be sufficient for the general reader of this book. In practical terms, many common laboratory protocols are now available as “kits” that are provided by biotechnology companies. These kits typically provide detailed instructions as well as relevant controls. Despite the ready availability of such reagents, it is still important to understand the biochemistry of the reactions and the limitations of protocols if success is to be achieved. A number of excellent molecular biology manuals have been published, many devoting several hundred pages to individual protocols (see Selected References). The methods have been divided into six general groups: 1. Important chemical and enzymatic manipulations of DNA and RNA. 2. Methods for the production of DNA and RNA for various applications. 3. Techniques for analysis of DNA and RNA. 4. Expression of recombinant proteins. 5. Determination of protein–DNA interactions. 6. Identification of protein–protein interactions.

ENZYMATIC MANIPULATIONS OF DNA AND RNA RESTRICTION ENDONUCLEASES Restriction endonucleases are bacterial enzymes that cut double-stranded DNA in a From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

9

10

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Table 1-1 Recent Nobel Prizes and Molecular Genetic Technology Year

Scientist(s)

Discovery

1975

D Baltimore R Dulbecco HM Temin W Arber D Nathans HO Smith P Berg W Gilbert F Sanger NJ Jerne GJF Kohler C Milstein JM Bishop HE Varmus R Roberts P Sharp M Smith KB Mullis

The interaction between tumor viruses and the genetic material

Reverse transcriptase

Restriction enzymes and their application to molecular genetics

Restriction enzymes

Biochemistry of nucleic acids Determination of base sequences in nucleic acids

Plasmids DNA sequencing

Principles for the production of monoclonal antibodies

Monoclonal antibodies

The cellular origin of retroviruses

Reverse transcriptase

Genes are split into exons

Structure of genes

PCR and oligonucleotide-directed mutagenesis

PCR Site-directed mutagenesis

1978

1980 1980 1984

1989 1993 1993

Technique derived

Figure 2-1 Restriction enzyme cleavage. Three enzymes are shown which digest DNA in three separate ways. EcoRI, 5' overhang; SmaI, blunt end; KpnI, 3' overhang. Arrowheads indicate the site of cleavage within the restriction site for each enzyme.

DNA POLYMERASES DNA polymerases are enzymes that use a single strand of DNA as a template for the synthesis of a second strand. All DNA polymerases require a primer, and all DNA polymerases synthesize new DNA in a 5' to 3' direction. A primer is a small piece of DNA that is complementary to the strand to be copied. Polymerases attach to the primer sites and then extend the primer in a 5' → 3' direction using individual deoxynucleotides as building blocks (Fig. 2-4). DNA polymerases have multiple uses. Among these are the production of radioactively labeled DNA, the polymerase chain reaction, and DNA sequencing.

Figure 2-2 DNA digestion pattern. Restriction digestion pattern of DNA with EcoRI and HindIII as well as the two enzymes combined. After electrophoresis, the DNA smaller fragments migrate through the gel faster than larger fragments. Basepair sizes are indicated on the left side of the gel.

CHAPTER 2 / RECOMBINANT DNA AND GENETIC TECHNIQUES

Figure 2-3 Ligase reaction. T4 DNA ligase will catalyze the joining of complementary ends of DNA by joining the 5' phosphate group to the 3' OH group of two strands of DNA. This reaction requires ATP as well as compatible ends of DNA.

REVERSE TRANSCRIPTASE Reverse transcriptase (RT) enzymes are a subtype of DNA polymerase that utilize RNA as a template to produce DNA. They also require a primer and synthesize in the 5' → 3' direction. Reverse transcriptases are isolated from retroviruses, which have the ability to reverse copy their RNA genome for their life cycle in mammalian cells. The major use of RT is to copy mRNA into DNA, which is then called “complementary DNA” or cDNA. The reaction of RT on an RNA template produces a hybrid RNA-DNA molecule. The DNA in this hybrid is called “first-strand cDNA” and can be used to generate another complementary copy of DNA to create double-stranded DNA (dsDNA). In this manner, cDNA is used to generate libraries representing the expressed genes of a tissue (see below) and as a substrate for polymerase chain reaction (PCR) in the sequence analysis of RNA species. RADIOLABELING Radioactive DNA is an important reagent in molecular biology, as it serves as a valuable hybridization probe. There are two basic ways to incorporate radioactivity into DNA: internal labeling and end-labeling. For internal labeling, DNA is denatured into single strands by boiling, and DNA polymerase is used to copy the strand in the presence of a radioactively labeled nucleotide (most commonly 32P-dCTP). The newly synthesized strand is therefore highly radioactive. Polynucleotide kinase is used for end-labeling. This enzyme transfers the radioactivity from an ATP molecule to the 5' end of single-stranded DNA. Only the most 5' base is labeled, and no radioactivity is incorporated internally in the DNA molecule. Several nonradioactive methods are also available for “labeling” DNA. For example, biotinylation allows detection with avidin-based systems, and digoxygenin can be detected with enzyme-linked assays.

ISOLATION OF GENOMIC DNA, TOTAL RNA, AND POLY(A) RNA In order to be accessible for manipulation, diagnostic studies, and therapeutic uses, DNA and RNA must be isolated in suitable

11

form and quantity. The best method of DNA or RNA isolation, and propagation depends on the intended use. Although most techniques are labor-intensive and require large amounts of starting material, the use of isolation kits from a variety of biotechnology companies have made the isolation of DNA and RNA a relatively simple matter. Both classical isolation and kit isolation methods utilize the basic principles outlined below. ISOLATION OF TOTAL GENOMIC DNA Genomic DNA is double-stranded and can be isolated from any kind of cell, including bacteria, yeast, plants, and animal cells. For most uses, it is desirable that the DNA remain as undamaged and the highest molecular weight as possible. The process of purifying DNA tends to introduce strand breaks and to reduce its molecular weight, but proper precautions can minimize shearing. There are many variations in the details of DNA isolation protocols, but all involve lysis of the cell and nucleus (if the cell has a nucleus) as a first step. A suspension of the DNA-containing cells is incubated with detergents and/or lytic enzymes for this purpose. The next step involves dissociation of the DNA from any attached proteins (such as histones). Most protocols utilize a strong protein degrading enzyme (proteinase K) for this purpose. This step also inactivates any DNA degrading cellular enzymes (nucleases). The partially degraded proteins are removed from the solution by phenol/chloroform extraction, salt precipitation, or the solution being passed over a column, which retains the DNA. DNA is precipitated from the remaining solution by ethanol precipitation. High-molecular-weight DNA (such as human genomic DNA) will form threads when precipitated in this way and can then be “spooled” from the solution with a glass rod. The DNA is then dried and redissolved in an aqueous buffer for further use and is stable in this form for several years. ISOLATION OF TOTAL CELLULAR RNA Several different forms of RNA exist in cells, including messenger RNAs (mRNA), ribosomal RNAs (rRNA), transfer RNAs (tRNA), and others. For many types of studies (Northern blots, RT-PCR), the isolation of total cellular RNA species is acceptable. In contrast to DNA, RNA is single-stranded and relatively unstable. During RNA isolation and storage, constant caution should be taken to avoid degradation by RNase enzymes, which are ubiquitous. For RNA isolation, cells, or tissues are homogenized in a solution that contains RNase-inhibiting compounds, such as the chaotropic chemical guanidinium isothiocyanate. Many different methods are available for the next step, which involves separating the RNA from genomic DNA. As in DNA isolation, the purified RNA is precipitated with ethanol during the final step of all procedures. Aqueous RNA solutions should be stored frozen. ISOLATION OF mRNA For certain applications, total cellular RNA is inadequate. Ribosomal RNA represents the bulk of total cellular RNA and only 1–3% of all RNA is mRNA. In order to make libraries of expressed sequences or to detect rare mRNA species on Northern blots, it is important to isolate mRNA as the starting material. Most mRNAs are poly-adenylated (poly[A]RNA), and this feature can be exploited for their capture. Total cellular RNA is passed over a column or resin of polydeoxythymine (oligo d/T column, poly-dT) bound to a solid phase. Poly(A) RNA binds to the poly-dT and is immobilized on the solid phase. The remaining unbound RNA is then washed off, and the poly(A) RNA is eluted in a subsequent step.

12

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Figure 2-4 DNA polymerase-mediated synthesis of DNA. DNA polymerase will attach free nucleotides to an existing primer in the 5' to 3' direction in a complementary fashion. The dashed line indicates the complementary bases.

CLONING DNA In the DNA isolation method described above, the genome of cells is the source of DNA; therefore, all sequences will be represented in the same abundance as they exist in the cell. Depending on the genome size of the organism, individual genes of interest represent only a minute fraction of the total DNA in a cell. For example, the average single gene in humans (~100,000-bp size) represents only 0.003% of the total genome. Thus, individual genes are much too dilute in such samples to be useful for detailed studies (such as DNA sequencing). This problem is solved by the cloning of DNA, which allows the preparation of large and pure quantities of a single DNA species. All methods of DNA cloning involve the ligation of a sequence of interest (such as a human gene) into a cloning vector. Cloning vectors are DNA sequences that exist inside a cell (in most cases the bacterium E. coli) but are not part of the genome of that cell. The cloning vehicle (and the cloned DNA within it) replicates inside the host cell, producing large quantities of a specific DNA sequence that can be isolated for further study. E. Coli PLASMIDS Cloning into E. coli plasmids remains the most important method for preparing large quantities of pure DNA. Plasmids are small circles of double-stranded DNA (2000– 5000 bp) that replicate inside bacterial cells (Fig. 2-5). Plasmids have their own origin of replication and can propagate themselves with the help of the endogenous cellular machinery. Cloning plasmids contain a selectable marker, usually a gene that confers resistance to an antibiotic. E. coli cells containing this antibiotic-selectable plasmid are then able to grow in media containing the respective antimicrobial agent. Many plasmids contain the β-lactamase gene, allowing growth in ampicillin. Low- and high-copy-number plasmids are distinguished. Pure plasmid DNA can be isolated from E. coli genomic DNA, because plasmid DNA is much smaller in size than E. coli genomic DNA (the E. coli genome is ~1 × 106 bp). The plasmid-containing E. coli cells are grown in a large volume of liquid culture media containing the antibiotic of choice. The cells are then pelleted by centrifugation and lysed, either with alkali, or using detergents. All plasmid preparation methods use size selection to remove the high-molecular-weight genomic DNA and RNA leaving only the small circular plasmid DNA. Milligram quantities of pure plasmid DNA can be prepared from a 1-L culture of E. coli cells containing a high-copy-number plasmid.

Modern E. coli plasmids have been engineered to contain a region with multiple useful restriction endonuclease sites (polylinker sites). In order to clone a gene of interest, the following procedure is followed (Fig. 2-5): DNA from the parental plasmid is cut with a restriction endonuclease, thus opening the ring and converting it to a linear molecule. A linear double-stranded DNA sequence of interest is generated with matching ends and is then ligated into the gap to regenerate a circular plasmid. The ligated plasmid is introduced by a process called transformation into a strain of E. coli that is sensitive to the antibiotic resistance marker carried by the plasmid. Transformation of bacteria can be accomplished by either heating the DNA-bacteria mix at 42°C for 1 min or by a quick pulse of electricity (electroporation). Both methods move the plasmid DNA through the bacterial wall and into the bacterial cell. After transformation, the bacteria are grown on antibiotic-containing plates. Only cells that have taken up a circular plasmid are able to survive and form colonies in the presence of the antibiotic. Individual bacterial colonies are also called “clones,” because every cell in the colony is genetically identical. Individual clones can be expanded in liquid media for the preparation of large amounts of plasmid DNA. Since bacterial cells are immortal and can continue to divide without limit, plasmid-containing clones represent a permanent source of DNA. The cells can be stored frozen indefinitely, and clones can be mailed and shared with other laboratories. Modern high-copy-number plasmids are about 3 kb in size and can accommodate inserts of up to 15 kb. Most DNA cloned in plasmids, however, is less than 10 kb in size and larger sequences are frequently unstable. Thus, only relatively small pieces of DNA can be cloned in plasmid vectors. For this reason, multiple cloning vectors capable of harboring larger DNA sequences have been developed. The most important of these will be briefly described below. PHAGE CLONING VECTORS Bacteriophage λ is a virus that infects E. coli cells and has been used extensively for cloning. Its structure is very different from plasmids. Virus particles consist of a protein envelope and core and a linear double-stranded DNA genome. For cloning purposes, much of the wild-type genome has been removed to allow replacement by exogenous DNA. In contrast to plasmids, inserts of up to 23 kb in size are stable in phage. Cloning in phage (Fig. 2-6) is performed in the following way: A large internal portion of the wild-type phage genome is removed by restriction digestion leaving a long (or left) and short (or right)

CHAPTER 2 / RECOMBINANT DNA AND GENETIC TECHNIQUES

Figure 2-5 Subcloning in bacteria. pUC 19 cloning vector is used to propagate cDNA. The plasmid carries a bacterial origin of replication (ORI). Also, a resistance gene for selection in ampicillin media (Amp), as well as a gene that can be used to detect for insertion of cDNA (lacZ). After opening the plasmid and introduction of the cDNA through restriction enzyme digestion and ligation, the construct is transformed into bacteria (E. coli) through either heat shock or electroporation. After transformation, bacteria are plated on resistant media, allowing the bacteria to grow only if they have taken up the plasmid containing the resistance gene. Growing the colonies with X-gal, a substrate for B galactosidase, clones that have been inserted into the lacZ gene and disrupted its reading frame, will no longer be able to metabolize X-gal and will appear white on the media plate. However, those colonies that may have religated and still contain a functional lacZ gene will metabolize X-gal and have a blue appearance.

arm. Neither of these two arms alone contains enough sequence to produce an infectious phage nor are the arms large enough to be packaged. The DNA sequence of interest is then ligated between the two arms, restoring a single large DNA genome. A “packaging extract” containing viral structural proteins isolated from wildtype phage is then used to package the recombinant phage and to infect E. coli cells. After infection of a bacterial cell, λ phage replicates to high copy numbers within the cell and then lyses it. Phage particles are released and can infect other E. coli cells. When propagated on agar plates, phage will then lyse bacterial cells and leave a hole or “plaque” in the lawn of bacteria. Each plaque represents a clone of DNA, because all the phage in a plaque are genetically identical.

13

When a liquid culture of E. coli is infected with phage, most of the cells will lyse and a high density of phage particles is attained. Phage particles can then be purified and their DNA isolated in large quantities. Recently, one other E. coli bacteriophage has become an important cloning vehicle. Phage P1 can stably harbor DNA inserts of up to 100 kb and is therefore used to clone long stretches of genomic DNA. COSMIDS Cosmids harbor inserts of ~40 kb and are very useful cloning vectors. Their properties are similar to plasmids in many aspects. Cosmids can propagate in E. coli and do not lyse their host cell. Antibiotic resistance genes (usually ampicillin) are used to select for cosmid-containing cells, which form colonies on plates. Inserts are ligated into convenient restriction sites. BACTERIAL ARTIFICIAL CHROMOSOMES Bacterial artificial chromosomes (BACs) are a special E. coli plasmid that can stably replicate very large inserts of genomic DNA. BAC inserts can be as large as 180 kb, and complete libraries of the human genome are now available in BACs. In contrast to YACs (see below), BAC inserts are usually pure (nonchimeric), and the BAC DNA is very easy to isolate. YEAST ARTIFICIAL CHROMOSOMES Yeast artificial chromosomes (YACs) have become very important in the human genome project, because they can harbor by far the largest inserts (up to 2000 kb). Because of this, a single YAC can cover a relatively large genomic region. In contrast to all other cloning vehicles described here, YACs are not grown in bacteria, but in the yeast Saccharomyces cerevisiae. YACs consist of a artificial centromere, the insert, and an artificial telomere, thus literally representing an artificial chromosome. Both the artificial centromere and the telomere contain different selectable marker genes. In order to generate a YAC, large pieces of DNA are ligated to the centromere on one end and the telomere on the other. This construct is then introduced into yeast, which is incubated on selective media plates. Only those yeast cells that contain both the telomere and centromere (and therefore both selectable marker genes) will survive and form colonies. Individual colonies (clones) can then be picked, grown in liquid culture, and YAC DNA can be isolated from the yeast cells. YACs have the advantage of containing very large inserts, but they have the disadvantage of frequently being chimeric. Chimeric YACs contain inserts from more than one region of the genome, which can limit their experimental utility. For example, a chimeric YAC may contain an insert from chromosome 3p fused to a portion of chromosome 15. LIBRARIES Cloning vectors such as those described above can be used to replicate a DNA sequence of interest and purify large quantities. But how are genes of interest found in the first place? The process by which gene(s) are isolated from the complex mix of DNA sequences found in cells of higher organisms will be described briefly. Most gene-isolation strategies have traditionally involved the generation of a library of DNA or RNA sequences. The beginning material for the library is a complex mix of DNA known to contain the sequence of interest. For example, the starting material for the isolation of a human gene would be total genomic DNA isolated from human cells. This DNA is then broken down into fragments of the appropriate size by mechanical shearing or by restriction endonuclease digestion. The appropriate size depends on the cloning vector to be used. For λ phage libraries, the average fragment

14

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Figure 2-6 Construction of a cDNA library. mRNA is isolated and reverse-transcribed by reverse transcriptase. The RNA strand is digested away leaving single-stranded DNA to which oligo d/T primers and DNA polymerase are added. These primers will anneal to the poly (A) tail of the cDNA and are extended by DNA polymerase. The double-stranded cDNA is ligated into λ phage arms, and bacteria are infected. Filter lifts are performed on the lysed bacteria and hybridized with the appropriate probe. Following hybridization and autoradiography, the appropriate plaques may be isolated and the corresponding phage purified and characterized. Adapted with permission. (Jameson JL. Applications of molecular biology in endocrinology. In: DeGroot LJ, ed. Endocrinology, 3rd ed., Philadelphia, PA: WB Saunders, 1995; pp. 119-147.)

size should be 8–25 kb. This mixture of fragments is ligated into a cloning vector and the ligation reaction is used to infect or transform E. coli bacteria. The biology of cloning vectors assures that each individual bacterium will contain only one cloning vector with an insert. Thus, this procedure divides the initial complex mix of DNA into millions of individual clones, each of which contains only a single DNA sequence, a library. Libraries can be produced from any source of double-stranded DNA. In order to identify the protein-coding regions of genes, however, it is necessary to generate libraries from expressed sequences. This type of library is called cDNA library and the starting material is messenger RNA (Fig. 2-6). To isolate the gene for a liver metabolic enzyme, for example, mRNA would be purified from homogenized liver. This material contains all mRNAs that are present in the liver. Some mRNAs will be very abundant, because they encode proteins present in high amounts in liver. Others transcripts will be rare. Reverse transcription is then used to copy the mix of mRNAs into first-strand cDNA. The cDNAs are then converted to double-stranded DNA and ligated into a cloning vector to produce the cDNA library. It is important to note that cDNA libraries are tissue-specific, because each tissue expresses a different set of genes. Similarly, cDNA libraries are specific for a developmental stage of the organism. Once a library of clones has been generated from the nucleic acid source of interest, the challenge is to find the specific individual clone with the gene among the millions of clones in the library. This is achieved by library screening. The bacteria containing the library are spread on agar plates at a defined density, which is low enough to permit the isolation of individual clones.

A phage library produced from human genomic DNA, for example, would be plated a density of 50,000 plaques per plate. At this density, 20 plates represent about 1,000,000 individual clones. Part of each colony/plaque on a plate is then transferred, as an exact copy, onto a nitrocellulose or nylon filter in a procedure termed “filter lift.” These filters are then processed to covalently bind either the DNA or the expressed proteins of the transferred plaques/colonies. The filters are incubated with a radioactively labeled DNA probe or specific antibodies, which will bind to the position on the filter occupied by a individual clone containing the desired DNA or expressing the sought-after protein. The corresponding clone is then picked from the original plate based on the position marked on the filter. DNA in the clone can then by amplified and isolated as described above, usually by subcloning it into plasmid vectors.

CHEMICAL PROPAGATION OF DNA Large amounts of pure double-stranded DNA can be produced chemically by a method called the polymerase chain reaction (PCR) rather than by propagating cloning vectors inside living cells. Relatively small quantities of single-stranded DNA can be synthesized from individual nucleotides by oligonucleotide synthesis. OLIGONUCLEOTIDES A prerequisite for the advent of the now ubiquitously used polymerase chain reaction was the ability to chemically synthesize short pieces of DNA of defined sequence (oligonucleotides). This process begins with the attachment of the most 5' base in the oligonucleotide sequence to a solid polymer support. Each new base in the sequence is then added to the 3' OHgroup of the attached bases in a stepwise fashion until the entire desired sequence has been synthesized. During each cycle of syn-

CHAPTER 2 / RECOMBINANT DNA AND GENETIC TECHNIQUES

15

deoxynucleotide triphosphates (dNTPs) are provided. Each of the originally present strands of the target DNA is copied and the amount of DNA is doubled with each cycle. Thus, after 25–30 cycles, the desired sequence can be enriched over a millionfold. This allows direct visualization as a distinct band in an agarose gel, and this amount of DNA is more than adequate for cloning or DNA sequencing. The sequential temperature changes required for PCR are performed by machines called thermocyclers. A complete PCR typically takes about 1–3 h and is therefore the most rapid method to produce a specific DNA sequence for further study. Limitations of PCR include misincorporation of nucleotides by the Taq polymerase (Taq errors) and the inability to amplify relatively large lengths of DNA (several kilobases). Modified versions of Taq polymerase with higher fidelity can reduce the rate of PCR errors and increase the length of fragments that can be amplified. PCR has numerous uses that are reviewed in detail in the attached references. As examples, PCR is used to: (1) amplify specific regions of DNA for genetic engineering strategies that do not depend on the presence of restriction enzymes; (2) amplify genomic DNA for diagnostic studies; (3) amplify DNA for sequencing protocols; (4) to insert site-directed mutations; and (5) in combination with reverse transcriptase, to quantify to the amount of starting mRNA (RT-PCR).

SITE-DIRECTED MUTAGENESIS

Figure 2-7 PCR reaction. DNA template, primers A and B, and DNA polymerase (Taq) are combined in a reaction that will cycle through denaturation, annealing, and extension temperatures, allowing the amplification of DNA between the primer pairs.

thesis, only one new base is added, providing strict control over the sequence of the completed oligonucleotide. Conventional oligonucleotides contain one of the four bases contained in DNA at each position. However, it is also possible to customize the procedure and to utilize unusual nucleotides (e.g., inosine), modified nucleotides (tagged with fluorescent dyes or digoxygenin), or more than one base in a single position (degenerate oligonucleotides). POLYMERASE CHAIN REACTION PCR permits the investigator to generate large quantities of a specific DNA sequence from a complex mix of DNA within a very short period of time and without cloning. The procedure uses a thermostable DNA polymerase and two synthetic oligonucleotides (Fig. 2-7). The DNA sequence of the gene to be amplified needs to be partially known, so that two oligonucleotide primers corresponding to the ends of this target can be synthesized. The procedure then consists of repeated cycles of three steps: (1) denaturation of dsDNA, (2) primer annealing to DNA, and (3) primer extension. During denaturation, heat (94–98°C) is used to separate the two strands of the target DNA into single strands. These single strands are then available to hybridize to the matching oligonucleotide during the annealing step if the temperature of the reaction is lowered below the melting point for the oligonucleotide-DNA hybrid (usually 50–68°C). Once a primer has annealed, the DNA polymerase (Taq polymerase from the bacterium Thermophilus aquaticus) can copy the remainder of the single strand at 72°C, if the required

PCR has greatly enhanced the ease with which site-directed mutagenesis can be performed (although other procedures are also available for mutagenesis). Site-directed mutations have many valuable applications, including structure-function studies of receptors, transcription factors, enzymes, and other transcription factors. Site-directed mutations are also introduced into the regulatory elements of genes to alter transcription factor binding sites and assess the effects on promoter function. Site-directed mutations also facilitate genetic engineering by allowing the introduction of new restriction sites. Mutagenesis typically involves a single nucleotide substitution, although it is also possible to change several basepairs at once (including small deletions or insertions). One commonly used protocol is referred to as overlap extension mutagenesis (Fig. 2-8). In this procedure, mutagenesis is performed by creating an oligonucleotide primer that is complementary to the normal DNA sequence except for the mutant base, which is generally positioned near the 5' end of the oligonucleotide to assure adequate priming. Using PCR, the mutant primer is incorporated into the newly synthesized DNA. Subsequently, the DNA products of this first PCR (now containing the mutation) are used to prime a second PCR in conjunction with a new oligonucleotide primer. After a few PCR cycles, a long mutant template is created that can be amplified further by using traditional PCR. Several variations of this strategy are reviewed in the listed references.

ANALYSIS OF DNA AND RNA GEL SEPARATION The size-based separation of DNA and RNA fragments by electrophoresis in various gels systems is a very important basic technique in molecular biology and is outlined briefly. At neutral pH, DNA and RNA are negatively charged and will migrate toward the anode in an electrical field. If this migration takes place within a polymer matrix (gel), small fragments move more rapidly than larger fragments (Fig. 2-2). Thus, electrophoretic migration through a gel gradually separates a mixture of DNA fragments of different sizes into distinct bands. Many

16

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Figure 2-8 Site-directed mutagenesis. A mutant primer (asterisk) and normal 3' primer are used to amplify a cDNA template. The product of this reaction is used to amplify the template in conjunction with a normal 5' primer. This reaction results in a mutagenized cDNA that can be amplified further using normal primers.

different gel matrices can be used for nucleic acid separations, but agarose and polyacrylamide gels are used most commonly. Agarose gels are suitable for the separation of DNA fragments in the 0.1- to 20-kb range, whereas polyacrylamide are used to separate small DNA fragments in the 0.025- to 2-kb range. The inclusion of denaturants such as urea allows single-base resolution in polyacrylamide gels (e.g., for DNA sequencing). A specialized kind of agarose gel electrophoresis termed pulsed field gel electrophoresis (PFGE) is used for the separation of very large fragments (up to 1000 kb). Different techniques are used to visualize the DNA fragments after separation by gel electrophoresis. If the fragment is relatively abundant, gels can be stained during or after electrophoresis, and the DNA can be directly visualized. Ethidium bromide will bind to DNA and lead to red fluorescence under UV light. If there is too little DNA for direct visualization, radioactivity is used for detection (see hybridization, below). NUCLEIC ACID HYBRIDIZATION One of the most commonly used analytical methods is nucleic acid hybridization. This technique is used for Southern blotting, Northern blotting, and for screening libraries. The goal of this method is to visualize a specific nucleic acid (DNA or RNA) sequence in the background of a complex mixture of other sequences. The technique takes advantage of the fact that the two complementary strands of nucleic acids will bind to each other (hybridize) with very high specificity. In contrast, noncomplementary nucleic acid sequences do not bind very efficiently, because nucleotide mismatches lower the melting temperature. To detect a particular species of DNA or RNA in a complex mixture, the constituents are first immobilized on a membrane and converted into a single-stranded form (see Southern and Northern blotting, below). A hybridization solution containing

Figure 2-9 Southern blot. Genomic DNA is isolated and digested with the appropriate restriction enzymes. DNA is then run on an agarose gel and transferred to a membrane support to which the DNA is covalently attached. Following transfer, the DNA is hybridized to the appropriate 32 P-labeled probe. Specific interactions between the DNA and labeled probe are detected by exposing the membrane-DNA to autoradiographic film. Adapted with permission. (Jameson JL. Applications of molecular biology in endocrinology. In: DeGroot LJ, ed. Endocrinology, 3rd ed., Philadelphia, PA: WB Saunders, 1995; pp. 119–147.)

radioactively labeled single-stranded probe is added and the radiolabeled probe allowed to hybridize under defined conditions. The temperature and salt concentration of the hybridization fluid determine the specificity of binding. At high temperature and low salt concentration, the probe will bind only to a perfectly complementary target sequence. After hybridization, the excess nonbound radioactivity is washed off, leaving a radioactive signal to be detected only at the location of the specific target sequence. The radioactive signal is visualized by exposure to an X-ray film yielding a black spot or band at the site of the radioactivity. SOUTHERN BLOTTING When genomic DNA is digested with restriction endonucleases and separated by gel electrophoresis, individual fragments cannot be visualized, even if large quantities of DNA are used. Because of the complexity of genomic DNA, restriction digests will invariably result in a “smear” of thousands of DNA fragments when visualized with ethidium bromide staining. In order to detect specific fragments within this smear, hybridization with radioactive probes is used in a procedure termed Southern blotting (after Ed Southern) (Fig. 2-9). Southern blotting begins with the separation of restriction-digested

CHAPTER 2 / RECOMBINANT DNA AND GENETIC TECHNIQUES

17

Table 2-2 Summary of Blotting Procedures Procedure Southern blot Northern blot Western blot Southwestern blot Farwestern blot

Substance detected

Probe

Major application

DNA RNA Protein Protein Protein

Nucleic acid Nucleic acid Antibody DNA Protein

Gene structure Gene expression Protein levels DNA–protein interactions Protein–Protein interactions

Adapted with permission. (Jameson JL. Applications of molecular biology in endocrinology. In: DeGroot LJ, ed. Endocrinology, 3rd ed., Philadelphia, PA: WB Saunders, 1995; pp. 119–147.)

Figure 2-10 Sequencing of DNA. Template, primer, and polymerase are added to a reaction in which both dideoxy and deoxynucleotides are present. Four separate reactions are used in which ddATP, ddTTP, ddCTP, and ddGTP are used individually. Each of these reactions is run on a polyacrylamide gel. Alternatively, the sequencing reactions can be carried out using fluorescently labeled nucleotides (or primers) to allow detection by a laser. The DNA sequence is then downloaded into a computer. Adapted with permission. (Jameson JL. Applications of molecular biology in endocrinology. In: DeGroot LJ, ed. Endocrinology, 3rd ed., Philadelphia, PA: WB Saunders, 1995; pp. 119–147.)

DNA by gel electrophoresis. After the separation, the DNA is denatured (separated into two single strands) by treatment with alkali, and the single strands are transferred onto a membrane by capillary transfer. The DNA binds covalently to the membrane and is immobilized on this solid phase, creating a replica of the fragments in the gel. Specific DNA fragments can then be identified on the membrane using hybridization probes that are specific for the gene fragment of interest. NORTHERN BLOTTING Gel separation and nucleic acid hybridization can also be used to analyze RNA in a procedure termed Northern blotting (Table 2-2). There are several important differences in comparison with Southern blotting. First, RNA is more susceptible to degradation than DNA. Electrophoresis is therefore carried out in a buffer that contains protective chemicals (usually formaldehyde). Second, RNA is already single-stranded and requires milder denaturation. Third, specific RNA species already have a defined size, and therefore enzymatic digestion is not needed in order to obtain a band pattern. The two procedures are similar, however, in that the RNA is transferred onto a membrane via capillary diffusion after electrophoresis. Usually ultraviolet light is used to crosslink the RNA onto the membrane to immobilize it.

DNA SEQUENCING The determination of the actual nucleotide sequence represents the most detailed level of DNA analysis. Several different techniques are available for DNA sequencing, but the dideoxy chain termination method originally developed by Sanger is now used the most widely (Fig. 2-10). DNA must first be denatured and separated into single strands by heating. A single radioactively labeled oligonucleotide primer is then added to the reaction and anneals to its matching sequence on the target DNA. DNA polymerase is then used to copy the single-stranded DNA. In the presence of saturating quantities of all four deoxynucleotide triphosphates (dATP + dTTP + dGTP + dCTP = dNTPs), a large extension product radioactively labeled at its end would be generated, whereas no sequence information would be generated. The addition of small quantities of dideoxy nucleotide triphosphates (ddNTPs) to the dNTP mix, however, leads to sequence information. Dideoxynucleotides are incorporated into the 3' end of the newly synthesized strand, but the DNA polymerase cannot add new bases onto the ddNTP. Thus, the incorporation of a ddNTP leads to chain termination. By adding appropriate ratios of dNTP

18

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

and ddNTPs, it is possible to achieve conditions in which chain termination occurs randomly at each nucleotide position. For example, if a primer extension was performed in the presence of dATP, dTTP, dGTP, and ddCTP, the polymerase will synthesize a new strand of DNA until it has to use ddCTP (e.g., when the complementary base is G). The ddCTP will be incorporated, but the polymerase cannot extend beyond this point. The length of the radioactive extension product therefore defines the position of the first G in the strand that is being copied. In order to determine the position of not only the first G, but additional Gs in the DNA, real sequencing reactions would be performed in the presence of a mix of dCTP and ddCTP at a molar ratio of ~200:1. In this setting there is a ~1:200 chance for a chain termination to occur when a G is present in the strand being sequenced. Extension products of multiple lengths will be generated and can be visualized after electrophoresis in a polyacrylamide gel. Based on its length, each fragment precisely defines the position of a G. In order to determine the positions of all four bases, four separate sequencing reactions are performed for each sample. In each case, a mixture of a given dNTP and its corresponding ddNTP is used in combination with saturating amounts of the three other dNTPs. The four reactions are then electrophoresed in adjacent lanes in a sequencing (denaturing polyacrylamide) gel, thereby permitting direct reading of the DNA sequence. Despite the relative complexity of chain termination theory, DNA sequencing in practice is relatively straightforward. Modern technology has allowed partial automation of DNA sequencing. For large-scale projects, robotics can be used to prepare sequencing reactions. Perhaps more importantly, instrumentation allows real-time “reading” of DNA sequencing gels and automated entry into computer databases. In addition to the reduction in human effort, such automation reduces errors inherent in reading and entering DNA sequence manually. Currently, most automated sequencers use fluorescent dyes rather than radioactivity. The dyes can be incorporated into either the sequencing primers or into the incorporated nucleotides. As with manual sequencing, gel electrophoresis (or capillary electrophoresis) is used to separate DNA fragments by size. However, with automated sequencers, detection of the fluorescent DNA fragments occurs via a laser beam, and the signal is processed by a computer. Other methods of automated sequencing are in development, including the use of DNA chips. In this strategy, a large ordered array of oligonucleotides are attached to DNA chips. Hybridization of DNA fragments to the chips allows the detection of overlapping sequences that can be converted into a contiguous sequence of DNA. This technology is particularly promising for the detection of polymorphisms and mutations, because a known sequence can be applied to the chips with specific variations at each nucleotide.

DETECTION OF DNA SEQUENCE POLYMORPHISMS DNA sequence polymorphisms play an important role in molecular genetics. A DNA polymorphism is a DNA sequence alteration that is found at a frequency of >1% in a given population. In contrast to mutations, these sequence alterations do not adversely affect the function of genes. They can be thought of as neutral variations in DNA sequence. Polymorphisms are important, because they permit the tracking of gene loci and attached regions in pedigrees. Many different classes of DNA polymorphisms and methods for their detection exist. The most important of these will be reviewed here.

Figure 2-11 Restriction fragment length polymorphism (RFLP). Genomic DNA is isolated and primers are added with DNA polymerase to amplify the targeted region of DNA. The two alleles differ in a EcoRI restriction site, the first allele having a EcoRI site (GAATTC), the second allele having a C for a G base change (CAATTC). Following amplification and digestion with EcoRI, the products are run on an agarose gel along with uncut samples. Individual 1 is homozygous for allele 1, individual 2 is heterozygous for the alleles, and individual 3 is homozygous for allele 2.

RESTRICTION FRAGMENT-LENGTH POLYMORPHISMS The simplest kind of DNA polymorphism is a single-base change. If the altered base lies within a recognition site for a restriction endonuclease, it may destroy the recognition sequence of this enzyme and abolish the site (Fig. 2-11). Alternatively, the polymorphism may create a new site. Single-base polymorphisms of this nature will alter the size of DNA fragments resulting from a digestion with this enzyme. The altered size bands can be detected after agarose gel electrophoresis and Southern blotting. If a restriction site was destroyed, the polymorphic band will be larger, and if a new site was generated, it will be smaller than the prevalent band in the general population. Single-base polymorphisms have been estimated to be present approximately every 1 kb in the human genome. About 1/6 of these changes will be detectable as a restriction fragment-length polymorphism (RFLP). Although RFLPs are relatively abundant in the genome, their use is limited by the fact that there are only two possible alleles for each polymorphism. The restriction site is either present or absent. In addition, Southern blotting requires relatively large amounts of DNA for analyses (5–10 µg).

19

CHAPTER 2 / RECOMBINANT DNA AND GENETIC TECHNIQUES

Table 2-3 Techniques Used for the Detection of Mutations Method Cytogenetics (FISH) Southern blot RFLP VNTR PCR DNA sequencing Mismatch cleavage OSH DGGE SSCP PTT

Gene deletions

Gene rearrangements

+

+

Loss of heterozygosity

Linkage

Point mutations

+

+

+

+ + +

+ + +

+ + + + + + +

FISH, fluorescent in situ hybridization; RFLP, restriction fragment length polymorphism; VNTR, variable number tandem repeat; PCR, polymerase chain reaction; OSH, oligonucleotide specific hybridization; DGGE, denaturing gradient gel electrophoresis; SSCP, single-stranded conformational polymorphism; PTT, protein truncation test. Adapted with permission. (Jameson JL. Applications of molecular biology in endocrinology. In: DeGroot LJ, ed. Endocrinology, 3rd ed. Philadelphia: WB Saunders, 1995; pp. 119–147.)

VARIABLE NUMBER TANDEM REPEATS Another important class of polymorphisms are the variable number tandem repeats (VNTRs) (minisatellites), used for forensic applications and paternity testing. VNTRs consist of reiterated repeats of small DNA sequences (0.1–10 kb) in multiple copies. The number of repeats in a VNTR is highly polymorphic (e.g., very variable between individuals in a population). To detect the polymorphism, a restriction endonuclease is used, which cuts outside of the repeat. The fragments obtained from such a digest will vary depending on the number of repeats and are readily distinguished by agarose gel electrophoresis and Southern blotting. In contrast to RFLPs, VNTR polymorphisms can have many different alleles (in some cases over 100), resulting in many different fragment sizes. However, VNTRs are relatively rare polymorphisms, and their density in the human genome is insufficient to permit linkage analysis for all human chromosomes. MICROSATELLITES Microsatellite repeats are similar to VNTRs in that a simple sequence is reiterated in multiple copies, and the alleles are distinguished by their sizes. However, the repeats in microsatellites are much more simple. The most commonly used repeat sequences are dinucleotide, trinucleotide, and tetranucleotide repeats. Dinucleotide repeats are common in mammalian genomes, particularly the dinucleotide CA. Microsatellite repeats are detected by PCR. Two primers flanking the repeat are used to amplify the region, and the products of the PCR reaction are then analyzed in a denaturing polyacrylamide gel (see Chapter 5). If a dinucleotide repeat differs from the population by a single repeat unit, this will give rise to a PCR product that is differs in size by 2 bp. Differences of this magnitude, however, are detectable in polyacrylamide gels. Similar to VNTRs, microsatellite repeats usually have many different alleles (repeat sizes). Because their detection is PCR-based, and because they are common in the genome, this class of polymorphisms has become the most important for gene mapping and diagnostic linkage analysis (see Chapters 4 and 5).

DETECTION OF MUTATIONS Mutations are DNA sequence alterations that change the function of a gene or the protein encoded by the DNA sequence. Many different kinds of changes can disrupt gene function. These include: gene rearrangements, deletions of DNA (large and small),

insertions, and single-base changes. The optimal technique for the detecting mutations is dependent on the nature of the mutation (Table 2-3). Large chromosomal rearrangements, insertions, and deletions (megabases in size) are best analyzed by cytogenetic techniques, including fluorescent in situ hybridization (FISH). FISH uses fluorescently labeled probes to hybridize to chromosomes. A catalog of fluorescent probes is available for mapping each chromosome as well as typical inversions and deletions. Southern blotting is used to detect sequence changes that involve hundreds to thousands of bases. More subtle mutations, including insertions and deletions of a few base pairs, and single-base changes, require specialized techniques for screening and characterizing mutations. There are two basic starting materials for mutation analyses: mRNA and genomic DNA. In comparison to genomic DNA, the analysis of mRNA has the advantage that the entire transcript can be analyzed in one piece (the introns have been spliced out). The disadvantage of mRNA analysis is that the gene of interest has to be expressed in accessible tissue. For example, a gene responsible for a brain disorder may not be expressed in skin or blood cells. On the other hand, it is sometimes possible to extract mRNA from leukocytes to characterize a gene that is expressed predominantly in another tissue (e.g., thyroid-stimulating hormone receptor mRNA). In order to analyze mRNA for mutations, reverse transcriptase is used to produce first-strand cDNA. This cDNA is then specifically amplified by PCR. This procedure is termed reverse transcriptasePCR (RT-PCR). Genomic DNA is present in all cells regardless of whether or not the gene is expressed. However, the “interesting” protein coding sequences of the gene are dispersed on multiple exons, which can be thousands of base pairs apart (see Chapter 1). Therefore, in order to analyze genomic DNA for mutations, individual exons are amplified by PCR for analysis. Several methods exist to detect mutations on PCR products derived from mRNA or genomic DNA. These include a variety of screening methods that detect altered DNA secondary structure including denaturing gradient gel electrophoresis (DGGE), single-stranded conformational polymorphism (SSCP), and several mismatch cleavage techniques. These techniques are most useful for screening large genes in which mutations occur in a widely distributed manner. The protein truncation test (PTT) is also useful for detecting frameshift or premature stop mutations in large genes. In this case, the mRNA

20

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Table 2-4 Selected Uses of Recombinant Proteins Purpose of recombinant protein expression Unlimited amount of biologically active drug Abundant source of pure enzyme Protein structural studies Structure–function studies Drug screening Raise antibodies Detect protein–protein interactions

Examples Insulin, growth hormone, erythropoietin Restriction endonucleases, Taq polymerase X-ray crystallography, NMR Transcription factors, enzymes Receptors, enzymes Protein antigen Yeast two-hybrid assays, farwestern screening

is converted to cDNA, transcribed, and translated in vitro, allowing the detection of mutations that truncate the protein. For genes that have “hot spots” for recurring mutations, oligonucleotidespecific hybridization (OSH) can be used to detect specific mutations that allow mutant oligonucleotides to hybridize to mutated genomic DNA. As noted above, DNA chip technology is likely to play an analogous role in mutation detection in the future. DNA sequencing is the most definitive technique for detecting mutations and will determine the exact nature of the genetic change. Using a combination of PCR and direct DNA sequencing, it is practical to screen many small genes for mutations. Improvements in the protocols for automated DNA sequencing now allow reliable detection of heterozygous mutations (dominant or compound heterozygotes). These protocols allow about 500 bp to be sequenced reliably in a single run.

EXPRESSION OF RECOMBINANT PROTEINS The ability to express large quantities of functional recombinant proteins is one of the many advantages conferred by the cloning of cDNAs. There are many practical uses for recombinant proteins (Table 2-4). They provide unlimited sources for reagents such as restriction enzymes and represent an invaluable source for hormones and growth factors, such as human insulin and erythropoietin, that are difficult to isolate in abundance from natural sources. The availability of abundant amounts of recombinant proteins has greatly facilitated biophysical studies and structurefunction analyses of proteins. IN VITRO TRANSLATION OF RECOMBINANT PROTEINS There are now numerous methods and hosts for expressing recombinant proteins, only some of which will be reviewed here. The simplest form of expression involves in vitro translation using reticulocyte lysates. In this method, mRNA (usually transcribed in vitro from a cDNA cloned into a plasmid) is translated using crude preparations of ribosomes present in reticulocyte lysates. Although the amount of protein is not large, it is possible to label it by including radioactive amino acids (e.g., 35S-methionine). In vitro translated proteins can be used as substrates for enzymes, in catalytic reactions, immunoprecipitations, DNA–protein interaction assays, and so on. E. Coli EXPRESSION OF RECOMBINANT PROTEINS E. coli were an early host for expression and continue to represent an important source for producing recombinant proteins. Expression in E. coli works best for relatively small, intracellular proteins that do not require extensive posttranslational modification for func-

tion. Proteins are expressed using plasmid expression vectors, many of which can be induced by treatment with reagents such as IPTG (relieves promoter repression), to cause high-level expression before purification. Attaching proteins of interest to other molecules such as β galactosidase can sometimes stabilize a protein that is otherwise subject to degradation in E. coli. It is also common to add a purification tag to the expressed protein to facilitate its recovery after expression. Common tags include glutathione-S-transferase (GST) for isolation on glutathione columns or a polyhistidine tag for isolation on nickel columns. Advantages of E. coli expression include the relative ease of the recombinant DNA manipulations (similar to plasmid cloning) and the rapid selection and expression process. Disadvantages include its inability to perform complex processing like glycosylation, and the fact that some proteins are labile (or toxic) in this host. Also, large proteins are not usually produced or folded efficiently. BACULOVIRUS-MEDIATED EXPRESSION OF RECOMBINANT PROTEINS Baculovirus-driven expression in insect cells is another very common strategy for recombinant protein expression. Using transfer plasmids containing the cDNA of interest, recombination occurs with the baculoviral DNA in Sf9 cells such that the DNA sequence encoding a major viral coat protein, polyhedrin, is replaced with the protein of interest. This results in very high expression of the recombinant protein. Sf9 cells can carry out glycosylation (although structurally distinct from mammalian cells), and it is also possible to coexpress more than one protein subunit (e.g., immunoglobulins). The recombinant protein often assumes a relatively normal cellular localization (e.g., receptors on the membrane; transcription factors in the nucleus), although the high level of overexpression ultimately distorts this pattern. The high level of expression facilitates purification unless aggregation occurs. However, the requirements of recombination and selection make this system somewhat cumbersome if a large array of mutant proteins need to be investigated. EXPRESSION IN MAMMALIAN CELL LINES For many proteins, it is desirable to perform expression in mammalian cell lines. These systems, while they rarely produce proteins as efficiently as E. coli or Baculovirus, have the advantage of processing complex polypeptides using intracellular machinery adapted for this purpose. In addition to requirements for precise peptide processing, folding, or glycosylation, the purpose of producing many recombinant proteins requires their expression in an homologous or tissuespecific system. For example, studies of certain receptors or transcription factors require their expression in cells that contain appropriate cofactors or targets. In the case of mammalian cell lines, expression of a cDNA is usually driven by a strong viral promoter. Alternatively, inducible expression vectors have been developed that allow selection of a clonal cell line followed by induction of the promoter by the addition (or removal) of reagents (e.g., tetracycline) that modulate promoter activity. Proteins with appropriate leader sequences may be secreted into the media. Although this results in dilution, there are many fewer contaminants in the extracellular media compared to the intracellular contents. EXPRESSION BY ADENOVIRAL VECTORS A variety of viruses have been used for expression in mammalian cells, taking advantage of the natural tropism of viruses and their ability to induce alterations in the intracellular protein synthesis machinery that results in high-level expression. Adenoviral vectors have been of interest because they infect a wide spectrum of mammalian cells and because the adenoviral genome has been successfully altered

CHAPTER 2 / RECOMBINANT DNA AND GENETIC TECHNIQUES

to create replication-deficient viruses that can harbor mammalian gene sequences. Adenoviruses have been used to express recombinant proteins in cell lines and tissues that are otherwise difficult to transfect with high efficiency. However, their use as vectors for gene therapy has made them a particularly attractive system for gene expression. By deleting key adenoviral genes, the modified virus can function as a carrier of mammalian genes. Introduction of the viral vector containing the gene of interest into E1a-expressing 293 cells allows the virus to be packaged and harvested, but it does not replicate in other cells. Adenovirus vectors have been used to express normal genes in defective organs (e.g., expression of CFTR in the lungs of patients with cystic fibrosis), to express toxic genes in tumors (e.g., thymidine kinase in brain tumors), and to target cell-cycle arrest genes in proliferating vascular cells. Limitations of adenovirus include an inability to accept foreign DNA greater than about 10 kb, relatively short-term expression in vivo (weeks to months), cytotoxicity, and induction of immune responses. Several other viral delivery systems are under investigation including Adeno-associated virus, Herpes virus, and retroviruses (which integrate in the host genome). Vaccinia viruses have also been used to express recombinant proteins in cell lines, such as HeLa cells, and provide an efficient mechanism for high-level expression of genes in mammalian cell lines. OTHER SYSTEMS FOR RECOMBINANT PROTEIN EXPRESSION Although some of the guidelines noted above can help to direct the choice of an expression system, often this requires empiric trials of different systems to determine which one yields the highest levels of expression, the best bioactivity, or is most cost-effective in terms of reagents or labor. Thus, it is valuable to have an array of other alternative expression systems. Discussion of each of these is beyond the scope of this chapter, but they are described in the references. Other expression systems are as diverse as production in animal milk, Drosophila, yeast, and plants.

PROTEIN INTERACTIONS WITH DNA Studies of protein–DNA interactions usually involve transcription factors that bind to regulatory elements in the promoters of genes (see Chapter 3). Characterization of these interactions can be accomplished by a variety of footprinting assays and more direct studies of protein binding to DNA, such as electrophoretic mobility shift assays. The goal of these studies is to identify the protein binding region of a gene, to characterize the transcription factor, and to correlate these results with functional studies of the promoter regulation. DNA FOOTPRINTING Footprinting techniques refer to the fact that proteins can protect DNA from digestion by a nuclease, thereby leaving a footprint in the position where the protein was bound (Fig. 2-12). Footprinting assays can be performed in vivo and in vitro, and these approaches often give different results, suggesting that the interactions (or proteins involved) may be different in the intact cell in comparison to naked DNA that has been mixed with nuclear protein extracts. For in vitro footprinting, DNA is labeled on one of the two strands and mixed with nuclear proteins. After allowing the sequence-specific proteins to bind to DNA, the nuclease DNase I is added at a concentration that will digest the DNA in a nearly random manner. The DNase I concentration used should cause somewhat less than one digestion per template to produce a ladder

21

of labeled fragments. Where protein is bound to DNA, there will be an area that is protected from DNase I digestion. As a control, digestion of the same DNA fragment is carried out in the absence of nuclear proteins, or better still, using extracts that lack specific transcription factors. Digestion reactions are run on a DNA sequencing gel, such that a ladder of fragments is obtained. The areas that were protected by proteins are undigested compared to control DNA and appear as gaps. In vivo footprinting is similar except that permeabilized cells, or intact nuclei, are exposed to DNase I before isolation of DNA. The pattern of protection and digestion can be determined by using ligation-mediated PCR. Some footprinting procedures detect the ability of chemical modifications of specific residues (e.g., methylation) to modify access to DNA. For example, cells can be treated with dimethyl sulfate (DMS), which methylates guanine residues on the N7 position. DMS moves freely into the cell and methylates about 1 in 50 guanine residues. If a protein contacts a specific guanine residue, it will protect it from methylation and subsequent cleavage by piperidine. In addition, DMS treatment and piperidine cleavage can be combined when ligation-mediated PCR with gene-specific primers is used to amplify the gene of interest. Although they are valuable for defining protein-binding regions, footprinting assays have several shortcomings. It is only possible to analyze about 150–200 bp at a time. Although one may identify a protected region of a gene, little information is gained concerning the number or nature of the bound proteins. The sensitivity of the assay is relatively low and detects primarily abundant, high-affinity proteins, since any free DNA in the reaction will partially obscure a footprint. The resolution of the binding site is relatively low, as DNase I cleavage may occur several basepairs away from a binding site and is somewhat sequence-specific. Last, it is relatively difficult to carry out detailed mutagenesis studies of the protein interaction site. For these reasons, other procedures such as the electrophoretic mobility shift assay (EMSA) have gained favor. ELECTROPHORETIC MOBILITY SHIFT ASSAYS (EMSA) The EMSA is based on the fact that proteins bound to DNA shift its mobility during gel electrophoresis, forming new low-mobility complexes (Fig. 2-12). Whereas footprint analyses provide information about a broad region of interaction (150–200 bp), the labeled DNA used for EMSA is generally 10–50 bp in length. EMSA is highly sensitive and, because mutant oligonucleotides can be readily synthesized, the assay is particularly useful for determining the exact nucleotides necessary for protein binding. The changes in mobility of the DNA-protein complexes are somewhat proportional to the molecular mass of the protein complexes. For this reason, EMSA is useful for studies of protein dimerization and for resolving oligomeric protein complexes. EMSA also allows further characterization of the protein of interest by performing competition studies with unlabeled oligonucleotide and by using antibodies to supershift specific proteins. The assay is performed by labeling an oligonucleotide with 32P and performing incubations with proteins in nuclear or wholecell extracts. It is usually necessary to include nonspecific oligonucleotides or carrier DNA to reduce the nonspecific binding to the labeled probe. The reactions are then subjected to nondenaturing polyacrylamide gel electrophoresis and exposed to autoradiographic film. The unbound DNA and DNA–protein complexes appear as distinct bands, reflecting their original mobility in the gel. Excess unlabeled oligonucleotide can be used as a

22

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Figure 2-12 DNase I footprinting and electrophoretic mobility shift assays (EMSA). A variety of techniques can be used to detect DNA–protein interactions. (A) Schematic illustration of a transcription factor bound to DNA. (B) DNase I digestion of labeled DNA generates a ladder of labeled fragments. Protein bound to DNA protects it from digestion with DNase I, creating the appearance of a footprint. (C) EMSA refers to a delayed migration of labeled DNA caused by the binding of protein to DNA. The mobility shift of labeled DNA in the presence of protein is depicted in lane 2. Addition of unlabeled competitor DNA displaces bound protein (lane 3). Addition of an antibody (Ab) against the protein causes a supershift of the complex (lane 4).

competitor to establish the specificity of binding to the labeled probe. By using distinct oligonucleotide sequences that are known to bind particular proteins, it is possible to provide evidence for the nature of the bound proteins. As noted above, it is possible to further define the identity of binding proteins by using antibodies against candidate proteins. Antibodies can cause two different outcomes. If the antibody recognizes the protein in such a way that it alters DNA binding (e.g., recognizes the DNA binding domain), it may inhibit binding. Alternatively, an antibody may interact with the protein while it is bound to DNA, causing a supershift because of the addition of the high-molecularmass antibody. DNA AFFINITY COLUMN Isolation of unknown proteins can sometimes be accomplished by the use of an oligonucleotide affinity column. Using the methodologies described above, specific protein-binding DNA sequences can be identified along with mutations that disrupt protein binding. These oligonucleotides (usually multimerized) are biotinylated and attached to an avidin-containing column matrix. After partial purification, relatively crude protein extracts can be passed over the DNA affinity column that will retain specific binding proteins. It is sometimes useful to apply the eluted proteins to a mutant oligonucleotide column to eliminate nonspecific binding proteins. Several cycles of affinity chromatography can greatly enrich the specific activity of binding proteins. A common goal in this procedure is to obtain enough protein with sufficient purity for peptide sequencing to allow identification and cloning (using degenerate oligonucleotides) of the cognate transcription factor. Other uses include partial purification of transcription factors for in vitro transcription assays.

PROTEIN–PROTEIN INTERACTIONS Protein–protein interactions can be identified by a variety of methods. Three methods will be discussed below: farwestern, immunoprecipitation, and yeast two-hybrid analyses. FARWESTERN In a farwestern analysis, a protein is labeled and used to detect interacting proteins in a manner somewhat analogous to the use of antibodies in a Western blot (Table 2-2). The protein of interest must be expressed and labeled. If the protein is a phosphoprotein, in vitro phosphorylation with labeled 32P-ATP and an appropriate kinase can be used. Alternatively, a protein kinase A tag may be added to the cDNA, allowing in vitro phosphorylation of the protein with protein kinase A. In vitro translation with radioactive amino acids such as 35S-Met can also be used to label proteins that are to be used as probes. The assay is performed by subjecting whole cell extracts or nuclear extracts (or known proteins) to denaturing gel electrophoresis and allowing the proteins to renature during the transfer to a membrane. The labeled protein is then used to detect specific proteins on the membrane, providing an indication of the number of protein interactions and the approximate molecular weights of interacting proteins. An analogous procedure can be used to screen expression cDNA libraries. Usually, farwestern blots of protein extracts are used to demonstrate that the protein probe works well and to establish specificity before a library is screened. This strategy was used, for example, to clone CREB binding protein (CBP) using 32P-labeled CREB (cAMP response element-binding protein). IMMUNOPRECIPITATION Immunoprecipitation involves the use of a specific antibody directed against a protein to precipi-

CHAPTER 2 / RECOMBINANT DNA AND GENETIC TECHNIQUES

tate it (along with other interacting proteins) from crude extracts. Currently, many antibodies (either polyclonal or monoclonal) are raised against recombinant proteins. Alternatively, if an antibody is unavailable, an epitope tag can be added to a cDNA to allow immunoprecipitation of the expressed protein. In many cases, immunoprecipitation is carried out using cells that have been metabolically labeled with 35S-methionine to allow detection of relatively small amounts of protein. After metabolic labeling, cells are lysed and incubated with the antibody to a specific protein. Precipitation is often performed using Staph A to recognize antibody complexes. After denaturing gel electrophoresis, precipitated proteins can be visualized by autoradiography. As noted above, in addition to the protein that the antibody is directed against, immunoprecipitation may also bring down proteins that are associated with the protein of interest. It is important to distinguish candidate-associated proteins from proteins that are precipitated nonspecifically. YEAST TWO-HYBRID The yeast two-hybrid method takes advantage of the modularity of proteins. Construction of a hybrid molecule between the DNA-binding domain of a yeast transcription factor, Gal 4, and the protein of interest, creates a “bait” with which one can search for protein–protein interactions. The other construct is created by fusing a candidate-interacting protein to the transcription-activating domain, VP16. In this manner, a successful protein–protein interaction will bring the VP16 transactivation domain to the Gal 4 DNA binding domain, inducing transcription of genes that contain Gal 4 target sequences (Fig. 2-13). A variation of this procedure inserts a library of cDNA sequences into the VP16-containing construct to test for clones that interact with the bait protein, thereby inducing the transcription of selectable genes in yeast. In a common version of this procedure, yeast cells are transformed with the cDNA-VP 16 fusion genes and transferred to restricted media. The gene necessary for growth on the restricted media is under the control of the Gal 4 transcription factor. The Gal 4 is fused with the protein-interaction domain of interest and should be designed to lack a transactivation domain. In this manner, for the yeast to grow on restricted media, protein–protein interactions must occur that bring the VP16 activation domain to the Gal 4 DNA-binding domain. From yeasts that are selected, clones of the VP16 fusion proteins can be isolated and analyzed further.

ASSAYS OF GENE TRANSCRIPTION Transcription assays are used to identify pathways that induce gene expression at the level of transcriptional initiation. In this manner, they extend the information gained by studies of steadystate mRNA levels, which are usually determined by techniques such as Northern blots, RNase protection assays, or RT-PCR. Using reporter genes as a reflection of the activity of transfected promoter–reporter fusion genes, it is possible to dissect the DNA regulatory elements within the promoter. An overview of nuclear run-on transcription assays and reporter gene assays will be presented in this section. NUCLEAR RUN-ON ASSAYS Having demonstrated regulation of a mRNA by Northern blot or alternative methods, it is often important to know whether alterations in steady-state mRNA levels are because of changes in transcription of the gene or changes in mRNA stability. The nuclear “run-on assay” is classic approach for studies of transcriptional regulation. In this procedure, nuclei are isolated after a specific stimulus and allowed to

23

Figure 2-13 Yeast two-hybrid assay. A cDNA library is inserted into the VP16 fusion construct and introduced into yeast cells carrying the Gal 4 fusion construct, bait. A Gal 4 construct is created by fusing the Gal 4 DNA binding domain and the gene of interest. If protein–protein interactions take place between the gene of interest and the random cDNA-VP16 fusion, expression of the leucine gene will allow for the growth of yeast on leucine minus media.

carry out mRNA synthesis in vitro, because the amount of mRNA synthesis under these conditions reflects the number of transcripts initiated prior to isolation of nuclei. The initiated transcripts are elongated in the presence of radiolabeled ribonucleotides, and the labeled mRNA is hybridized to specific clones of immobilized DNA to allow quantitation. TRANSIENT GENE EXPRESSION AND REPORTER GENE ASSAYS IN TRANSFECTED CELLS Transient gene expression studies provide an alternative technique for examining transcriptional control (Fig. 2-14). This method also allows detailed mutagenesis of cloned promoter sequences prior to their introduction into cells. In this procedure, the promoter sequences of genes are fused to a reporter gene that can be assayed readily. Common reporter genes include chloramphenicol acetyltransferase (CAT) and luciferase (LUC). In each case, these reporter genes represent enzyme activities that are not normally found in eukaryotic cells. Thus, in the absence of gene transfer, the background activity of these reporter enzymes is negligible. A host of commercial vectors are now available to allow the insertion of promoter or enhancer elements into various reporter gene constructs. CAT assays catalyze the transfer of the acetyl group from acetyl-CoA to the substrate, chloramphenicol, and can be monitored by thin-layer chromatography, enzyme-linked immunosorbent assay (ELISA), or by liquid scintillation counting.

24

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

β galactosidase, growth hormone, alkaline phosphatase, and green fluorescent protein genes. The promoter–reporter fusion gene constructs are introduced into cells using a process referred to as transfection. Common transfection protocols include Ca2+PO4, DEAE-dextran, lipidbased, and electroporation. Transfected genes are transcriptionally active over 24–72 h, allowing relatively rapid analyses of promoter function. Alternatively, transfected genes can be stably introduced into cells by selecting for a resistance marker such as neomycin or dihydrofolate reductase. The principal goal of transfection experiments is to define DNA sequences that are required for promoter function or that respond to a specific extracellular signal or second messenger pathway. A variety of methods are now available to allow deletion or site-directed mutagenesis of promoter elements (Fig. 2-8). This approach has been critical for defining DNA sequences that regulate tissue-specific expression, basal promoter activity, and responses to intracellular signaling pathways.

SELECTED REFERENCES

Figure 2-14 Luciferase reporter assay. Reporter plasmid is transfected into cells. Reporter plasmids mediate the transcription of the luciferase gene, which is then translated into luciferase protein. Six to forty-eight hours later the cells are lysed and assayed for luciferase activity. Adapted with permission. (Jameson JL. Applications of molecular biology in endocrinology. In: DeGroot LJ, ed. Endocrinology, 3rd ed., Philadelphia, PA: WB Saunders, 1995; pp. 119–147.)

Luciferase assays use the luc gene from the firefly, Photinus pyralis, which catalyzes the reaction of D-luciferin and ATP in the presence of O2 and Mg+2, to result in light emissions than can be measured using a luminometer. For both assays, the amount of enzyme activity is proportional to the amount of mRNA expressed, which, in turn, reflects the activity of the promoter being used. Luciferase assays are 30–100 times more sensitive than CAT assays. A variety of other reporter genes are also used, including

Adams MD, Fields C, Venter JC. Automated DNA Sequencing and Analysis. New York: Academic, 1994. Ausubel FM, Brent R, Kingston RE, et al. Short Protocols in Molecular Biology, 3rd ed. New York: Wiley, 1995. Boultwood J. Gene Regulation: A Eukaryotic Perspective, 2nd ed. Totowa, NJ: Humana, 1997. Harwood AJ. Basic DNA and RNA Protocols. Totowa, NJ: Humana, 1996. Innis M, Gelfand DH, Sninsky JJ. PCR Strategies. New York: Academic, 1995. Kneale GG. DNA-Protein Interactions: Principles and Protocols. Totowa, NJ: Humana, 1994. Landegren U. Laboratory Protocols for Mutation Detection. Oxford, UK: Oxford University Press, 1996. Latchman DS. Transcription Factors: A Practical Approach. Oxford, UK: IRL/Oxford University Press, 1993. Leitch AR, Shwarzacher T, Jackson D, Leitch IJ. In Situ Hybridization. Oxford, UK: BIOS Scientific, 1994. Lewin B. The extraordinary power of DNA technology. In: Genes V. Oxford, UK: Oxford University Press, 1994; pp. 633-656. Tijssen P. Hybridization with Nucleic Acid Probes (Parts I and II). Amsterdam, North Holland: Elsevier, 1993. Trower MK. In Vitro Mutagenesis Protocols. In: Methods in Molecular Biology, vol. 57, Totowa, NJ: Humana, 1996. Tuan RS, Recombinant Gene Expression Protocols. Totowa, NJ: Humana, 1996. Watson JD, Gilman M, Witkowski, Zoller M. Recombinant DNA, 2nd ed. New York: WH Freeman, 1992. White BA. PCR Cloning Protocols: From Molecular Cloning to Genetic Engineering. Totowa, NJ: Humana, 1997. Wu R. Recombinant DNA Methodology II. New York: Academic, 1995.

CHAPTER 3 / TRANSCRIPTIONAL CONTROL

3

25

Transcriptional Control of Gene Expression WADE JOHNSON AND J. LARRY JAMESON refer to the regulatory DNA elements that control a gene. In contrast, trans-acting mutations refer to a diffusible factor, often a transcription factor. The conceptualization of trans-acting factors interacting with cis-elements in target genes helps to transfer genetic concepts into biochemical terms. The field of gene regulation has a long history of creating models that predate detailed biochemical explanations. The idea of regulatory cascades required accepting the notion of diffusible factors before there was direct evidence for their existence. The cis-acting regulatory elements in promoters have often been characterized by mutagenesis before obtaining direct evidence that they bind a transcription factor. In practice, this is almost always true. Now that transcription factor interactions with DNA can be demonstrated more readily using sensitive techniques such as DNA-protein mobility shift assays, a new set of questions has arisen concerning how these DNA-bound proteins interact with the basal transcription apparatus. There is now a large body of evidence for the presence of adaptor or coactivator proteins, and the ability to use transcription factors to clone interacting coactivator proteins has greatly expanded our understanding of this family. This chapter will focus largely on the principles of transcriptional control in higher eukaryotes. We acknowledge that many lines of thinking have arisen from studies of prokaryotes, yeast, and viruses. Indeed, the conservation of protein structure and regulatory pathways is very useful. Reflecting this conservation, many mammalian transcription factors are functional in yeast and vice versa. Nevertheless, space limitations necessitate a more limited review. Although the major focus is on transcriptional regulation, gene expression ultimately integrates many distinct regulatory steps, including transcription termination, mRNA stability, and the control of mRNA translation. Discussion of these other topics can be found in the listed references. We also hope to link information on transcriptional regulation to other topics covered in this book, including disease states that involve mutations in transcription factors. The function of normal and mutant genes is influenced by their levels and patterns of expression as well as by the properties of the expressed proteins. Some disorders are caused by mutations in the regulatory regions of genes (e.g., some thalassemias) or more commonly because of mutations in transcription factors (see below). In disorders that involve imprinting, the clinical manifestations largely reflect diminished gene expression from one of the parental alleles (e.g., Prader-Willi Syndrome). At least 20% of expressed genes are involved in gene and protein expression. By

INTRODUCTION The concepts of gene function and transcriptional control are intimately intertwined. Early ideas about genes were based on specific traits that could be transmitted in a predictable manner through generations. From Gregor Mendel’s studies of peas, the principle of using statistical approaches to understand genetics was established. His studies revealed the ability to predict the proportions of peas that would be yellow or green or exhibit smooth or wrinkled traits. The notion that these or other transmissible features were conveyed by genes was made more concrete by the studies of T. H. Morgan, who studied genetic transmission in Drosophila. Morgan demonstrated that certain traits were inherited together (or linked), reflecting their locations on the same chromosome. He also found that genes underwent recombination in a manner that reflected their distance from one another on a chromosome. Thus, genes that are widely separated have a greater statistical chance to undergo recombination than do genes that are immediately adjacent. These concepts of genes predated, of course, any understanding of the physical structure of a gene. Pivotal concepts in the field of gene expression were articulated by Francois Jacob and Jacques Monod. They formulated the idea of a “messenger,” now known to be mRNA, that linked genes to their biochemical effectors, such as enzymes and structural proteins. Inherent in the proposal of a messenger is the need to control gene expression. Thus, Jacob and Monod postulated an “operator,” which would serve as a control switch for gene expression. This model proposed that repressors acted on the operator element and that expression could be activated by blocking the repressor. The idea that gene expression (mRNA synthesis) is a regulated event is linked to the genetic concepts of alleles that act in cis and trans. Initially, these terms described the results of genetic complementation tests that were designed to determine whether recessive mutations reside in the same (cis) gene or in different (trans) genes. More recently, these terms have been adapted to describe other features of gene expression. Specifically, it has been convenient to conceptualize the promoter regions of genes in terms of cis and trans. For example, cis-acting mutations in an operator, normally controlled by a repressor protein, result in constitutive activation. Such a mutation affects only the genes that are directly linked to, and regulated by, the operon. Thus, the cis-acting sequences often

From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

25

26

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

comparison, it has been estimated, based on the composition of expressed sequence tagged (EST) genes from many different tissues, that cell structure and motility involve 8%, cell signaling involves 12%, and cell division involves 4% of the expressed genes. Although these results are preliminary (based on 7000 of an estimated 50,000–80,000 expressed genes), and the categorization is somewhat arbitrary, it is clear that a major fraction of genes are devoted to the control of gene expression. In addition to severe mutations in transcription factors, it is likely that many polygenic disorders will involve alterations in the levels of expression of certain genes or the sensitivity of a promoter to environmental effects.

STRUCTURE OF GENES STRUCTURE AND FUNCTION OF THE NUCLEOSOME The formation of chromatin is a characteristic feature of eukaryotic DNA (see Chapter 1). Chromatin is composed of histones, nonhistone proteins, and DNA. The histones are highly conserved proteins involved in the packaging of DNA into nucleosomes (Fig. 3-1). Most DNA (~ 95%) is associated with nucleosomes. The nucleosomes function at several levels to provide hierarcheal organization of DNA. The first order of organization consists of 10-nm “beads on a string” in which nucleosomes and DNA are arranged in a linear fashion. The second order (30-nm fiber) arises when the nucleosomes are compacted into an helical array in a process that requires nonhistone proteins. The third order occurs when the fibers are packed on themselves to form euchromatin (low packaging) or heterochromatin (high packaging). Histones 2A, 2B, 3, and 4 (H2A, H2B, H3, and H4) form the core nucleosome. Two molecules of H2A, H2B, H3, and H4 form the octameric core of the nucleosome. Histone 1 (H1) protein is probably not involved in the formation of the core nucleosome, as 10-nm fibers can be constructed without H1. However, secondorder packing into the 30-nm fiber requires H1, which resides on the outside of the DNA-nucleosome complex and stabilizes the higher order structure. Based on digestion patterns generated by monococcal nuclease, it has been established that approximately 200 bp of DNA are associated with each nucleosome. DNA (80 bp) wraps around each core nucleosome twice like thread around a spool. A spacer (40 bp of DNA) links nucleosomes together to create the “beads on a string” appearance seen using electron microscopy. In addition to playing a structural role in DNA packaging, it is likely that nucleosomes also participate in gene regulation. During the cell cycle, the structure of nucleosomes changes. In S phase, when DNA replication occurs, nucleosomes are acetylated. Acetylation occurs on lysine residues at the carboxyterminus of histone proteins and may alter their interactions with DNA. During G2 and mitosis, the nucleosomes are deacetylated and highly condensed. Transcription factors have been shown to modulate the acetylation of histones, potentially modifying nucleosome structure as a mechanism of controlling gene transcription. For example, histone acetylation may create a more open chromatin configuration that allows other transcription factors to gain access to the regulatory regions of a gene. The transcriptional coactivator, CREBbinding protein (CBP), possesses intrinsic histone acetyltransferase (HAT) activity. CBP also recruits other HAT proteins such as P/CAF. It has been speculated that histone acetylation may be an important mechanism by which CBP increases gene transcription. Other transcription factors (often repressors) recruit histone deacetylases. MAD/MAX and Mxi1/MAX heterodimers bind Sin3, which acts as a scaffold protein to bind histone

deacetylases 1 or 2 (HDAC 1, 2). Myc transformation of tissue culture cells can be blocked by MAX/Sin3 fusion proteins, but this inhibition is not seen using Sin3 proteins that do not bind to HDAC 1 or 2. Nuclear receptor co-repressors such as NCoR and SMRT also bind Sin3 and HDAC1 and 2 (see below). STRUCTURE OF THE TRANSCRIPTION UNIT A gene refers to an individual transcription unit that encodes either a single protein or a protein subunit. Genes are divided structurally into 5' and 3' regulatory regions that flank either side of the exons and introns (see Chapter 1). Exons refer to the portion of genes that are eventually spliced together to form messenger RNA. Introns refer to the spacing regions between the exons that are spliced out of precursor RNAs during RNA processing. The upstream 5' flanking regions of genes typically contain hormone response elements and other regulatory regions, including sequences involved in the initiation of transcription (see below). This regulatory region of the gene is also referred to as the promoter, although this term is generally reserved for the proximal 100–300 bp upstream from the transcriptional start site. The minimal core promoter consists of a TATA box (which binds TBP, TATA-binding protein) and initiator sequences that enhance the formation of an active transcription complex. Transcriptional termination signals reside at the 3' end of the gene. In some instances, regulatory elements also reside downstream of the gene or in introns. Specific signals such as the AAUAAA sequence at the 3' end of the mRNA are involved in designating the site for polyadenylation (poly[A] tail). With the exceptions noted above, the regulatory regions of genes generally reside in the 5' flanking DNA sequence of the gene. There is, however, great variation in the amount of DNA sequence that is required for normal expression of a gene. Almost all genes contain numerous (10–20) regulatory elements within the first 300 bp of the promoter (although comprehensive studies may be required to identify all elements). Most genes also contain several enhancer elements that are located further upstream. In some cases, such as the globins and the immunoglobulin genes, enhancers are located at great distances (>5 kb) from the remainder of the gene. An enhancer element was traditionally defined as a regulatory element that could operate over great distances and in an orientationindependent manner. However, further characterization of these elements, and the transcription factors that they bind, reveals considerable functional overlap with other promoter regulatory elements. Therefore, the distinction between enhancers and promoter regulatory elements has become blurred. The properties of these regulatory elements are strongly influenced by their locations in the gene and by the nature of surrounding DNA sequences. The organization of a prototypical promoter is depicted in Fig. 3-2. The promoter contains a TATA box located about 30 bp upstream from the site of transcriptional initiation. Initiator sequences typically surround the transcriptional start site. The initiation of translation starts at the first ATG triplet codon within eukaryotic mRNA. The sequences prior to this ATG comprise the 5' untranslated region (5' UTR) and the most 5' base within the 5' UTR constitutes the transcriptional start site (designated as +1). The 5' UTR varies greatly in length in different genes, and it may contain several exons. For this reason, it can be challenging to identify the location of the promoter, even when the coding region of the gene has been found. Several techniques are used to characterize the transcriptional start site(s) and, thereby, the location of the promoter. The S1 nuclease protection assay uses a single-stranded radio-

CHAPTER 3 / TRANSCRIPTIONAL CONTROL

27

Figure 3-1 Role of nucleosomes in gene transcription. DNA is wrapped around nucleosomes, resembling thread on a spool. Actively transcribed genes exhibit an open chromatin structure in which nucleosomes have been disrupted or shifted to a different location along DNA. Altered nucleosome positioning allows the binding of transcription factors which, in turn, recruit coactivators and other components of the basal transcription apparatus (TBP, GTFs, Pol II). GTFs, General Transcription Factors; TBP, TATA-binding protein; Pol II, RNA Polymerase II.

Figure 3-2 Structure of the regulatory region of a gene. A prototypical gene is depicted with an array of DNA regulatory elements. The core promoter includes the TATA-box that binds TATA-binding protein (TBP) and an initiator (INR) at the site of transcriptional initiation. The proximal promoter contains a CAAT-box that can bind several different transcription factors including CAAT-box/enhancer-binding protein (C/EBP), a cAMP response element (CRE) that binds members of the CREB/AP-1 family, an Sp-1 site (binds Sp-1), and an AP-2 site (binds AP-2). A composite enhancer is shown in the more distal 5'-flanking region of the promoter. It contains several closely spaced regulatory elements including a tissue-specific element (TSE), an E box (binds members of the bHLH family), and an AP-1 (c-Jun/c-Fos heterodimer) site. The composite nature of the enhancer allows combinatorial interactions of several different transcription factors to result in highly specific regulation. An adjacent hormone response element (HRE) is a binding site for members of the nuclear receptor family. Note that some regulatory elements, such as the AP-1 site, might reside in either enhancer or proximal promoter (e.g., CRE) locations.

labeled DNA probe that spans the 5' UTR and the transcription start site. S1 nuclease digests single-stranded RNA and DNA, leaving the protected DNA-RNA hybrid undigested. The length of the protected fragment allows the position of the start site to be deduced. An alternative approach uses primer extension in which a radiolabeled primer and reverse transcriptase are used to extend the mRNA to its 5' end. The promoter sequence adjacent to the identified transcriptional start site often contains characteristic sequences such as a TATA box and other DNA regulatory elements. The TATA box is the site of binding for TATA-binding protein (TBP), which is a key component of transcription factor II D (TFIID). The proximal region of the promoter is generally densely packed with transcription factor binding sites, only some of which are shown. This region often contains sites for ubiquitous proteins such as Sp-1, CAAT box/enhancer-binding protein (C/EBP), cAMP response element-binding protein (cAMP), or Activator Protein-1 (AP-1). However, factors involved in cell-specific expression may also bind to these sequences. Examples include interactions of bHLH proteins with E boxes in the myogenic genes (see Chapter 92), steroidogenic factor-1A (SF-1) for the steroidogenic enzyme promoters (see Chapter 57), or Pit-1 for the growth hormone promoter (see Chapter 49). At greater distances from the promoter, additional regulatory elements are often clustered in such a way that they form composite enhancers. These elements may be small, or they may contain several repeats of

different transcription factor-binding sites dispersed over several hundred basepairs. In some cases, these distant enhancers are important for developmental expression (e.g., globins), cell-specific expression (e.g., immunoglobulins), or hormonal regulation (e.g., glucocorticoid response elements [GREs]). Although enhancer and promoter elements are usually depicted in a linear manner along DNA, it is likely that DNA looping occurs to allow interactions of enhancer-bound transcription factors with components of the basal promoter (Fig. 3-3). When active, these enhancer sites may confer DNase I sensitivity, reflecting the absence of nucleosomes, and an open configuration that allows increased access of DNase I (and other transcription factors). The anatomy of a promoter is usually defined by a combination of gene transfer experiments to assess the effects of promoter mutants and studies of protein–DNA interactions (see above). In the gene transfer experiments, various promoter regions are linked to a reporter gene, such as the gene encoding luciferase. Sequential deletions of the promoter provide a gross delineation of the locations of regulatory elements, but this strategy may miss functional elements that act in combination with sequences that have been deleted. Ideally, point mutations are introduced into candidate regulatory elements to assess their role in the context of the native promoter. It is advantageous to simultaneously search for transcription factor binding sites, either using DNase I footprinting or gel shift assays. DNase I footprinting is most useful for screening 100- to 300-bp regions of DNA for the locations of protein-bind-

28

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Figure 3-3 Enhancer interactions with proximal promoter elements. Many enhancers reside a great distance from the site of transcriptional initiation. DNA bending, or looping, has been proposed to allow protein–protein interactions. In the model shown, an enhancer binds a dimeric transcription factor that interacts with the coactivator protein, CREB-binding protein (CBP). CBP integrates the actions of many different transcription factors, including phosphorylated cAMP response element-binding protein (CREB). CBP, or its homolog, p300, interacts with other proteins including histone acetyltransferase (HAT) and general transcription factors (GTFs) such as TFIIB, and so on to induce transcription by RNA Polymerase II (Pol II).

ing sites. After the identification of protein binding sites by footprinting, the electrophoretic gel mobility shift assay (EMSA) is useful for more detailed analyses of protein–DNA interactions. EMSA is a very sensitive measure of protein–DNA interactions and is relatively simple to perform from a technical perspective. Competition studies using unlabeled DNA or supershift studies using antibodies directed against candidate transcription factors can help to verify the nature of the bound transcription factor.

CLASSIFICATION AND FUNCTION OF TRANSCRIPTION FACTORS Because of the large number and diverse functions of transcription factors, it is helpful to group them into several classes based on their roles in gene transcription. In this chapter, transcription factors are divided into general transcription factors (GTFs), DNA sequence-specific transcription factors, and transcriptional coactivators and corepressors. The general transcription factors include a large number of proteins that are involved in the assembly of the basal transcription apparatus. As their name implies, the sequencespecific transcription factors include proteins that bind to DNA regulatory elements in the enhancers or promoters of genes. Coactivators and corepressors bind to other transcription factors through protein–protein interactions and regulate transcription by altering chromatin structure or by making contacts with the basal transcriptional machinery. GENERAL TRANSCRIPTION FACTORS Eukaryotic cells contain three RNA polymerases that catalyze the transcription of DNA into RNA. Polymerase I transcribes ribosomal RNAs (rRNA), polymerase II transcribes protein coding genes (mRNAs), and polymerase III transcribes tRNAs. Although RNA polymerase II and its core proteins are able to catalyze RNA synthesis, they are

insufficient for regulating gene-specific transcription. A large number of proteins are required for polymerase II (Pol II)-mediated transcription. Sequences associated with the core promoter (TATA box, Initiator) and gene-specific enhancers each serve to bind additional general transcription factors (GTFs) including TFII A, B, D, E, F, and H, as well as RNA polymerase II and associated proteins such as suppressor of RNA polymerase B (II) (SRB). Purification of the RNA polymerase holoenzyme, which is transcriptionally inactive, with TFII B, E, F, and H already bound, suggests that two major complexes of GTFs may be present in the cell: RNA polymerase holoenzyme and TFIID. TFIID refers to a protein complex composed of TBP and several TBP-associated factors (TAF II’s). TBP is highly conserved through evolution. Drosophila TBP and mammalian TBP can be substituted for one another in vitro transcription assays. TBP binds specifically to the TATA box, binding in the minor groove of DNA, creating a 90° bend in the DNA. Some promoters do not contain a consensus TATA box, but TBP is still necessary for transcription. In the case of these TATA-less promoters, other GTFs apparently tether TBP and TFIID to the promoter. Many of the TATA-less promoters are characterized by multiple transcriptional start sites, suggesting that one role of the TATA box is to specify a precise position for transcriptional initiation. In addition to TBP, TFIID contains at least eight TAFIIs that recognize various promoters and establish contacts with other transcription factors. TFIIA contacts TBP and stabilizes the TATA box interaction. TFIIB contacts DNA on both the 5' and 3' side of the TATA box, specifies the transcriptional start site, and it is important for enhancermediated stimulation of transcription in vitro. TFIIH contains helicase and kinase activities that are necessary for the initiation of transcription. As noted later, mutations in TFIIH are one cause

CHAPTER 3 / TRANSCRIPTIONAL CONTROL

29

Figure 3-4 Classes of transcription factors. The structures of selected classes of transcription factors are depicted schematically. In general, the factors are classified according to their DNA-binding domains. Examples of transcription factors are shown at the left and the class of transcription factor is shown at the right. Some of the identified functional domains are shown. Q1 and Q2, represent activation domains; KID, kinase inhibitory domain; HLH, helix loop helix; SRF, serum response factor; SP/TP, serine, threonine phosphorylation; HMG, high mobility group; POU, Pit-1/Oct-1/Unc.

of xeroderma pigmentosa, which is characterized by DNA repair defects in response to UV irradiation. TFIIE recruits TFIIH to the preinitiation complex, and modulates the activity of TFIIH. TFIIF recruits RNA polymerase II to the preinitiation complex. In higher eukaryotes, the role that TAFs play in transcription is only partially understood. The composition of TAFIIs associated with the TFIID-TBP complex provides a degree of specificity for RNA polymerase II. TAF function has been examined for a few transcription factors such as Sp1, which interacts with dTAF110 in vitro. Transcription factor-TAF interactions likely serve as a bridge between enhancer regions and the basal transcription apparatus. Transcriptional activation via TAFs is influenced by the DNA sequence surrounding the TATA box and the Initiator. Hence, the core promoter architecture plays an important role in gene-specific regulation by recruiting the appropriate basal transcription factors. DNA SEQUENCE-SPECIFIC TRANSCRIPTION FACTORS The ability of the basal transcriptional apparatus to respond to cellspecific and environmental signals is accomplished through DNA sequences that are specific to a given gene. These DNA regulatory sequences often reside in enhancer elements that are capable of conferring functions to a broad array of minimal promoters. Transcription factors that bind to enhancer regions are usually modular in structure and contain a DNA-binding domain and one or more activation domains. A partial compilation of various classes of sequence-specific transcription factors is shown in Fig. 3-4. This

classification is based on the domain that binds to DNA and includes bZip, bHLH, homeodomain, and zinc-finger-containing classes of transcription factors. This classification is somewhat arbitrary, because some factors contain more than one type of domain or DNA-binding regions that are poorly defined. Examples of the functional roles of some of these sequence-specific factors are discussed later in this chapter. It is striking that a relatively small number of DNA-binding structures have been used to generate tremendous diversity among the thousands of sequence-specific transcription factors. A transcription factor database on the Internet (http://transfac. gbf.de/) contains hundreds of different transcription factors and allows a promoter region to be searched for recognition sites for sequence-specific transcription factors. However, empirical determination of factor binding and function is ultimately necessary, because many consensus sequences are not particularly selective and transcription factor interactions can be dramatically influenced by surrounding DNA sequences or DNA-bound proteins. TRANSCRIPTIONAL COACTIVATORS AND COREPRESSORS This class of transcription factors has been identified relatively recently. Because these proteins do not interact directly with DNA, it has been challenging to isolate these factors. However, largely based on techniques that detect protein–protein interactions (e.g., yeast two-hybrid, farwestern techniques), the list of identified members of this family is growing rapidly (Table 3-1).

30

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Table 3-1 Transcriptional Coactivators and Corepressors Cointegrators CBP/p300

Coactivators ARA70 GRIP 95, GRIP 120, GRIP 170 p140 (ERAP 140, RIP 140) p160 (ERAP 160, RIP 160) SRC-1/ Nco-A1 Sug 1/Trip 1 TIF 1 TIF 2/GRIP 1 Corepressors NCoR 1 SMRT aSee

Interacting proteins Nuclear Receptorsa: (RAR, RXR, ER, TR, PR, GR) CREB, c-Fos, c-Jun, JunB, YY1, E1a, SV40 Tag, c-Myb, SRC-1, Sap 1a, STAT2, MyoD, E2F-1, TBP, TFIIB, p/CAF, pp90RSK, Tax, SREBP-2, p53 Nuclear receptors AR GR ER, RAR ER, RAR PR, ER, RAR, RXR, TR, GR TR, RAR, RXR ER, RAR, RXR PR, ER, RAR, RXR Nuclear receptors TR, RAR TR, RAR

Fig. 3-9 for abbreviations.

Coactivators have been identified primarily by using yeast two-hybrid screening assays. This assay takes advantage of the modularity of proteins. A yeast transcription factor, Gal 4, is fused to the protein of interest to create a “bait” with which to test for protein–protein interactions. By introducing a library of expressed cDNAs linked to a known transactivation domain (e.g., VP16), it is possible to select (by transcription of selectable genes) for interactions between the bait and proteins expressed from the cDNA library. An alternative to the yeast two-hybrid approach is to use farwestern analysis to screen for direct protein–protein interactions. The protein of interest is expressed, radiolabeled, and used to detect interacting proteins either from cell extracts (analogous to a Western blot) or from expression libraries. This method allowed phosphoCREB to be used to isolate clones for CBP. It is likely that most sequence-specific transcription factors will be shown to interact with transcriptional coregulators. These proteins provide a mechanism for integrating the effects of several different transcription factors, and they also provide a linkage between enhancer-binding proteins and general transcription factors. Examples of regulation by coactivators are discussed later in this chapter using CBP/p300 as an example of an activator and the interaction of the thyroid hormone receptor with NCoR (nuclear receptor corepressor) as an example of a repressor.

OVERVIEW OF MODELS FOR TRANSCRIPTIONAL CONTROL There is enormous variability in mechanisms of transcriptional control. In fact, every gene is controlled uniquely, whether in its spatial or temporal pattern of expression, or in its response to extracellular signals. Examples of signaling pathways that activate transcription factors are depicted in Fig. 3-5 (see Chapter 46). Thus, in addition to the diversity of transcription factors that are available, there is substantial diversity in the pathways that can activate transcription factors. Despite the myriad pathways and mechanisms for transcriptional activation (and repression), it is useful to consider fundamental models that recur for different genes.

TRANSCRIPTIONAL ACTIVATION Transcriptional activation can be divided into three main mechanisms (Fig. 3-6): (1) altered chromatin structure, (2) posttranslational modifications such as phosphorylation, (3) displacement of a repressor protein. These models are not mutually exclusive and some transcription factors participate in several different types of regulation. The topic of chromatin reorganization was discussed above. Regulation of the MMTV promoter by the glucocorticoid receptor appears to involve alterations in chromatin structure. In this case, the glucocorticoid receptor is proposed to displace nucleosomes in a manner that allows the binding of additional transcription factors such as NF-1. CBP, which transduces the actions of many transcription factors (see below), may also function in part through alterations in chromatin as several of the proteins that it recruits possess histone acetyltransferase activity. Phosphorylation and dephosphorylation are key steps in the covalent modification of many transcription factors. Phosphorylation of the aminoterminal δ domain of c-Jun by Jun N-terminal kinase (JNK) stimulates the transcriptional activity of c-Jun. However, phosphorylation by mitogen-activated protein kinase (MAPK) near DNA binding domain of c-Jun inhibits transactivation. Hence, for maximal activation of c-Jun, both phosphorylation and dephosphorylation must occur in different regions of the protein. The role of phosphorylation in the control of CREB transcription has also been studied extensively. In this case, phosphorylation of a key serine residue (Ser 133) by one of several different kinases is required for transcriptional activation. Phosphorylation induces conformational changes in CREB that are necessary for interactions with CBP. Ets-1 is phosphorylated by MAPK. When binding to its consensus DNA sequence, Ets-1 undergoes a conformational change, which induces its full activation. However, Ets-1 must interact with another factor to stabilize its interactions with the DNA. This type of synergism is exemplified by the c-fos promoter, where the binding of serum response factor (SRF) to the serum response element (SRE) recruits ternary complex factors (Ets-1). The synergistic interaction between SRF and Ets-1 accounts in part for rapid induction of the c-fos promoter in response to serum.

31

CHAPTER 3 / TRANSCRIPTIONAL CONTROL

Figure 3-5 Signal transduction and transcriptional control. A diverse array of signaling pathways activate transcription (see Chapter 46). A highly simplified schema is shown to illustrate some of these pathways. Steroids (or other nuclear receptor ligands) diffuse through the cell membrane to act on nuclear receptors. For some of these receptors (e.g., RAR, TR), the ligand displaces a nuclear corepressor (NCoR) and allows interactions with a steroid receptor coactivator, SRC-1. Growth factors activate a variety of tyrosine kinase receptors that stimulate kinase cascades including mitogen-activated kinase (MAPK) and c-Jun kinase (JNK). These kinases activate transcription factors such as AP-1 (c-Jun/c-Fos) and Ets-1. Hormones (e.g., catecholamines, peptide hormones) can interact with seven transmembrane G-protein-coupled receptors to activate adenylyl cyclase (AC), resulting in cAMP production and stimulation of protein kinase A (PKA). This pathway activates transcription factor CREB, which can also be stimulated by other kinase pathways, including Ca2+-mediated activation of CaMK (Ca2+-mediated kinases). Cytokines stimulate the Janus kinases-signal transducers and activators of transcription (JAK-STAT) pathway, resulting in translocation of phosphorylated STAT transcription factors to the nucleus. Several cellular stress pathways (UV, ultraviolet light; LPS, lipopolysaccharide; TNF, tumor necrosis factor) activate NFκB (p50/p65) by dissociating the inhibitor, IκB. CREB-binding protein (CBP) binds to many different transcription factors, providing a mechanism for integration (or competition) among transcription factors. In other cases, posttranscriptional modifications influence the cellular localization of a transcription factor (Fig. 3-5). NFκB is found in a cytoplasmic complex that consists of two subunits, p50 and p65, along with IκB. IκB binds to the two subunits and prevents their nuclear localization by masking the nuclear localization signal (NLS) within NFκB. Activation of NFκB requires the ubiquination and subsequent degradation of IκB, which allows translocation of NFκB to the nucleus. Signal transducers and activators of transcription (STAT) proteins are also found in an dormant state in the cytoplasm. Activation of cytokine receptors stimulates Janus Kinase (JAK), which phosphorylates the receptors, allowing recruitment of the STAT proteins (see Chapter 46). Once localized to the cytokine receptors, the STATs are phosphorylated, inducing dimerization and translocation to target genes in the nucleus.

Figure 3-6 (right) Mechanisms of transcriptional activation. Several models of transcriptional activation are shown (see text for details). (A) An activator protein induces an alteration in nucleosomes, thereby changing chromatin structure and allowing the recruitment of additional activators to DNA. (B) An activator is stimulated by a posttranslational modification such as phosphorylation, leading to the recruitment of additional activator proteins. (C) An activator displaces a repressor protein that binds to the same, or to an overlapping, site.

Figure 3-6

32

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Figure 3-7 Mechanisms of transcriptional repression. Several models of transcriptional activation are shown (see text for details). (A) A repressor protein induces an alteration in nucleosomes, allowing the recruitment of additional proteins that silence transcription. (B) A repressor protein competes for coactivator proteins. This mechanism is sometimes referred to as transcriptional “squelching.” (C) A repressor protein displaces an activator that binds to the same, or an overlapping, site. Another model for transcriptional activation involves displacement of a repressor by an activator. This model applies in particular to developmental cascades, in which the induction of a new transcription factor may be sufficient to stimulate transcription, or it may displace a pre-existing repressor. The orphan nuclear receptor, COUP-TF, acts as a repressor of many genes that contain nuclear receptor binding sites. Under circumstances in which other nuclear receptors are induced or activated, they can displace COUP-TF and activate transcription. TRANSCRIPTIONAL REPRESSION In general, mechanisms of transcriptional repression have not been studied to the same extent as mechanisms of transcriptional activation. In part, this reflects the technical challenges involved in measuring inhibition from a basal state of transcription. In certain respects, models for transcriptional repression are the reciprocal of activation, but there are also differences. As shown in Fig. 3-7, some repressors act by causing histone deacetylation (as opposed to acetylation in activation). As described, this pathway has been investigated recently for the MAD/MAX proteins and nuclear receptors (RAR, TR). In these cases, the transcription factors bind Sin3 and recruit histone deacetylases. For nuclear receptors, transcriptional silencing is seen in the absence of ligand, and this is reversed at ligand binding (see below). Also analogous to mechanisms of activation is the possibility that repressors can act by displacing activators. A good example

of this pathway involves expression of inducible cAMP early repressor (ICER), which can compete for the binding of CREB or CREM to cAMP response elements. ICER contains the bZip DNA binding domain of CREM, but does not contain the transactivation domains. Thus, when it occupies the DNA binding site, it is inactive and blocks access of other transactivators. The functional role of ICER is discussed later as a model for autoregulation of gene expression. A distinct mechanism for transcriptional repression involves the inhibition of transcriptional activators, either by direct interactions between the repressor and an activator or by competition for a coactivator. An example of direct inhibition involves Id inhibition of bHLH factors. In this case, Id lacks a DNA binding domain, but it retains a dimerization domain, allowing interactions with bHLH proteins such as Myo D, E12, and E47. Dimerization of these proteins with Id forms an inactive complex and blocks the action of these transcription factors. Another example of direct interactions is illustrated by inhibition of NFκB or c-Jun by the glucocorticoid receptor. The glucocorticoid receptor has been shown to interact directly with these transcription factors to inhibit their action. However, it has also been suggested that the activated form of the glucocorticoid receptor may compete for the coactivator, CBP, which is also shared by c-Jun. Thus, negative regulation by the glucocorticoid receptor may occur by direct interactions and by competition for coactivators. Competition for rate-limiting coactivators may prove to be a relatively common model for transcriptional repression. For example, most cases of transcriptional squelching likely involve competition for limiting amounts of transcriptional coactivators.

PRINCIPLES OF TRANSCRIPTIONAL CONTROL TRANSCRIPTION FACTORS ARE COMPOSED OF MODULAR FUNCTIONAL DOMAINS Soon after transcription factors were cloned, structure–function studies revealed a remarkable property: The functional domains of transcription factors are often separable and transferable. For example, domains involved in DNA binding can be localized to a relatively limited region of the protein and even swapped or exchanged into other transcription factors. Similarly, domains involved in dimerization, nuclear localization, phosphorylation, ligand binding, and transcriptional activation can also be localized and transferred to other proteins. These features suggest that for many transcription factors, functional domains are localized and are not created by complex threedimensional interactions between distant parts of the protein. Another implication of this finding is that functional domains are often conserved during evolution. These “evolutionary modules” diverge, but retain enough similarity that homology is readily detected. Homeodomains provide an example of a conserved functional region that has become greatly diversified during evolution. The homeodomain proteins are typically involved in developmental events such as segmentation or cell lineage. The homeodomain itself is involved in DNA sequence recognition. Two examples of transcription factors with well-characterized modular domains are shown in Fig. 3-4. The transcription factor, CREB, is representative of the basic leucine zipper (b-Zip) family of transcription factors. The carboxyterminal end of CREB contains a highly basic region adjacent to a series of repeated hydrophobic residues (primarily leucines) that are part of an amphipathic α-helix. The rotation of the α-helix positions the hydrophobic residues on the same surface of the protein. This feature has led to

CHAPTER 3 / TRANSCRIPTIONAL CONTROL

33

Figure 3-8 b-Zip family of transcription factors. b-Zip transcription factors are characterized by a domain comprised of a basic region adjacent to a leucine zipper. This domain mediates dimerization and DNA binding. The family can be subdivided into the CREB/ATF and Jun/Fos family. Different members of the family preferentially form homodimers or heterodimers. CREB, cAMP response element binding; CREM, cAMP response element modulator; ATF, activating transcription factor; CREBP, cAMP response element-binding protein. the description of a “leucine zipper,” in which a series of leucines (or other hydrophobic amino acids) are aligned with their hydrophobic side chains protruding from the protein. The hydrophobic amino acids create a dimerization interface that results in a coiledcoil of the two α helices. The adjacent basic residues are available to make contacts with negatively charged DNA. The DNAsequence recognition sites for this family of transcription factors are palindromic in nature. Thus, there is twofold symmetry (e.g., TGAC • GTCA) in the DNA sequence, and the structure of the DNA element reflects in the symmetry of the transcription factor dimer. The binding affinity of the dimer is much greater than the protein monomers. This probably reflects conformational changes that occur after protein dimerization, as well as the additional interactions with DNA that are conferred by having two proteins make contacts with DNA. Members of the b-Zip family can form homodimers (e.g., CREB-CREB) or heterodimers (e.g., Jun-Fos), providing a mechanism for diverse interactions (Fig. 3-8). Subtle variations in the DNA recognition site provide specificity for the binding of different dimers. In addition to its DNA-binding and dimerization domains, CREB also contains well-characterized transcriptional activation domains in its central and aminoterminal regions. These regions contain numerous sites for phosphorylation by enzymes such as protein kinase A and Ca2+-dependent kinases. These phosphorylation events alter protein conformation and result in transcriptional activation. The modularity of this domain has been revealed by its ability to transfer transcriptional activation to the DNA binding domain of GAL4. The GAL4-CREB fusion protein recognizes a GAL4 target gene, but it is transcriptionally activated by protein kinase A. In this experimental paradigm, it has been possible to mutate specific serine residues in CREB to demonstrate that they are involved in phosphorylation-dependent transcription. As described below, phosphorylation of CREB allows the recruitment of a transcriptional coactivator, CBP.

The modular structure of the TR is also shown in Fig. 3-4. This receptor is representative of other nuclear receptors such as the glucocorticoid receptor, estrogen receptor, and retinoic acid receptor (Fig. 3-9). This nuclear receptor superfamily also includes a number of so-called orphan nuclear receptors for which a specific ligand has not been identified. Each member of this family contains a centrally located DNA-binding domain that is formed by two zinc fingers. The modular properties of this domain have been demonstrated in swapping experiments. For example, exchanging the DNA-binding domains of the thyroid hormone and glucocorticoid receptors switches their DNA recognition properties. Thus, a gene that is normally a target for the thyroid hormone receptor can be activated by glucocorticoids when the DNA binding domain from the thyroid hormone receptor is inserted into the glucocorticoid receptor. DNA-binding specificity is provided primarily by a stretch of amino acids (P-box) at the carboxy-terminal base of the first zinc finger. This region was identified because, in evolutionary comparisons of the zinc fingers in different receptors, it represented an area of hypervariability. Thus, substitution of this motif in the glucocorticoid receptor with that from the estrogen receptor switches its DNA-binding specificity to that of the estrogen receptor. The carboxy-terminal region of the nuclear receptors contains several functional domains, including nuclear localization, dimerization, ligand binding, repression, and activation domains. The locations of these domains have been demonstrated by swapping protein region experiments to create chimeric nuclear receptors as well as by performing site-directed mutagenesis. The ligand-binding properties of these receptors require most of the carboxy-terminus. Consistent with this, the X-ray crystal structures demonstrate that the ligands bind deep in a pocket formed by several α-helical loops. During ligand binding, the receptor undergoes conformational changes, including repositioning of the transcriptional activation domains that contact coactivators.

34

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Figure 3-9 Comparison of nuclear receptors. Nuclear receptors share a highly conserved central DNA-binding domain (DBD) that consists of two (C2H2) zinc fingers. The carboxy-terminal region contains several functional domains including ligand binding, dimerization, transcriptional repression, and transcriptional activation. The amino-terminal domain is highly variable in length and contains additional transactivation domains. The lengths (in amino acids) are shown at the right side of the figure. MR, mineralocorticoid receptor; PR, progesterone receptor; AR, androgen receptor; GR, glucocorticoid receptor; ER, estrogen receptor; RAR, retinoic acid receptor; TR, thyroid hormone heceptor; VDR, vitamin D receptor.

Somewhat analogous to the b-Zip family of transcription factors, the nuclear receptors also form dimers. In this case, the dimerization domain is less clearly defined than the leucine zipper, but probably also involves hydrophobic interactions. In many cases, the nuclear receptors form heterodimers as well as homodimers. For example, the thyroid hormone receptor can homodimerize, but it also forms a heterodimer in combination with the retinoid X receptor (RXR). A second dimerization interface exists in the zincfinger DNA-binding domain. The zinc fingers thereby orient the two monomers by interacting with specific DNA sequences. Thus, the target DNA site participates in a ternary complex of DNA and two protein subunits. In the case of the thyroid hormone receptor, the DNA recognition sites are surprisingly complex. Although most sites conform to a conserved half-site (AGGTCA) that interacts with one of the receptor monomers, the orientation of the two half-sites relative to one another is quite variable. For example, the half-sites can be arranged head-to-head (as a palindrome), headto-tail (direct repeat), or tail-to-tail. These distinct orientations would predict that the carboxy-terminal dimerization domains would be positioned very differently or even on opposite sides of the DNA. However, these receptors appear to have a remarkably flexible hinge domain between the DNA binding domain and the carboxy-terminal dimerization domain. This hinge allows dimerization to occur irrespective of the alignment of the zinc-finger domains. In this way, the hinge has preserved the modularity of

functional domains, while allowing great diversity in the DNA target sequences. Transcriptional activation domains in the nuclear receptors have been localized in the carboxy-terminus, but also in the aminoterminus. The amino-terminal activation domains are constitutive or ligand-independent. This region of the nuclear receptors varies greatly in length and in sequence (Fig. 3-9). The activation domain in the carboxy-terminus is usually located near the extreme end of the protein. For many nuclear receptors, this function localizes to an amphipathic α helix that undergoes a marked conformational change during ligand binding. This region of the receptor contains a signature motif, LXXLL, that appears to be involved in interactions with coactivators (Table 3-1). The coactivators interact with the carboxy-terminal domain in a ligand-dependent manner, and it is thought that the recruitment of coactivators leads to interactions with the basal transcription apparatus, resulting in increased transcription. The nuclear receptor carboxyterminal domain also interacts with corepressors such as NCoR and SMRT (Table 3-1). The corepressors bind close to the nuclear receptor hinge domain. In the case of the thyroid hormone receptor, the corepressors interact only in the absence of ligand. Thus, the ligand acts as a transcriptional switch (Fig. 3-10). In its absence, corepressors bind to the receptor and suppress transcription. Addition of the ligand dissociates the corepressor and allows the recruitment of coactivators. In physiologic terms, these properties allow thyroid hormone to induce a dynamic range of responses in its target genes. DIMERIZATION PROVIDES A MECHANISM FOR DIVERSITY USING A LIMITED REPERTOIRE OF TRANSCRIPTION FACTORS Dimerization is common for many classes of transcription factors. The ability to dimerize accomplishes several goals: 1. Dimers allow the generation of diverse combinations of related factors. 2. Dimers provide the opportunity for variations in DNA sequence to specify interactions with different combinations of related factors. 3. Dimers can increase the affinity of transcription factor binding to DNA by contributing the energy (and conformational changes) induced by protein–protein interactions as well as by making additional contacts with DNA. 4. Dimers create an opportunity for pivotal transcriptional switches if some partners activate, whereas others inhibit it. Many of these principles of dimerization have already been illustrated in the discussions of the modular domain of the b-Zip and nuclear receptor families of transcription factors. The basic helix-loop-helix (bHLH) family of transcription factors also emphasize the important role of dimerization. The helix-loop-helix motif was first identified as a DNA binding domain in λ-phage repressors. Variations of this structure are also found in homeodomain proteins and in the bHLH proteins. In the latter group, amphipathic helices that are connected by the loop create a dimerization surface and the basic regions of the protein partners make contacts with DNA. The DNA consensus sequence is typically an imperfect palindrome (CANNTG), reflecting the symmetry of the dimer. This family of transcription factors is often involved in cell lineage and growth regulation. Some of the factors such as E12 and E47 are expressed ubiquitously, whereas other factors (e.g., MyoD) are expressed in a tissue-specific manner (see Chapter 92). The ability to form homodimers and heterodimers is variable. For

CHAPTER 3 / TRANSCRIPTIONAL CONTROL

Figure 3-10 Transcriptional silencing and activation by the thyroid hormone receptor. (A) The thyroid hormone receptor (TR) silences transcription in the absence of its ligand, T3. Addition of the ligand reverses transcriptional silencing and induces transcription above the initial basal level of transcription. (B) Schematic depiction of transcriptional silencing based on the ability of TR to recruit a nuclear corepressor, NCoR. NCoR, in turn, recruits additional proteins (not shown) such as Sin3 and histone deacetylases, which may function to alter chromatin structure. (C) Depiction of transcriptional activation. T3 dissociates NCoRs and allows the recruitment of coactivators such as SRC (steroid receptor coactivator) to induce transcriptional activation. example, E47 forms homodimers efficiently, but MyoD does not, and it preferentially forms heterodimers with one of the ubiquitous factors (e.g., E12/MyoD). The array of ubiquitous and cell-specific factors creates a combinatorial code that leads to selective gene expression. The bHLH family of dimers is also notable for the presence of a class of inhibitory domain partners, (Id). The Id proteins retain dimerization but lack DNA binding, thereby blocking the action of their protein partners. The availability of proteins such as Id provides an alternate mechanism for controlling target genes. For example, overexpression of Id can prevent MyoDdirected myogenesis. An analogous inhibitory pathway may also exist in the b-Zip proteins. In this case, inhibitory proteins such as CHOP or CREM (isoforms α, β, γ), can form inhibitory complexes with selected other members of the b-Zip family. In the nuclear receptor family, proteins such as COUP-TF dimers usually function to block target DNA sites, and a splicing variant (lacking the

35

transactivation domain) of the thyroid hormone receptor, TRα2, binds RXR to form an inactive heterodimer. UNIQUE COMBINATIONS OF TRANSCRIPTION FACTORS PROVIDE MECHANISMS FOR TISSUE-SPECIFIC EXPRESSION The issue of cell-specific expression has fascinated molecular biologists since the first Northern blots revealed that some genes are only expressed in one or a limited number of cell types. Among some of the early human cDNAs that were cloned, several exhibited highly restricted patterns of expression (e.g., globin, growth hormone, insulin, chorionic gonadotropin). Initially, investigators analyzed the promoter regions of these genes in search of cell-specific transcription factors in order to account for this property. Although such factors exist, it has been discovered repeatedly that the expression and function of many promoters are controlled by multiple regulatory elements and their cognate transcription factors. In the case of growth hormone, for example, the tissue-specific transcription factor, Pit-1, has been identified. Pit-1 is a member of the POU-homeodomain family, expressed specifically in the pituitary gland. The growth hormone promoter contains multiple binding sites for Pit-1. When coexpressed with the growth hormone promoter in nonpituitary cells, Pit-1 enhances growth hormone expression in the “ectopic” locus. However, in the absence of other enhancer binding factors, Pit-1 activates the growth hormone promoter rather weakly. When combined with other ubiquitous factors such as the thyroid hormone receptor, there is synergistic stimulation of the promoter. Synergistic interactions of transcription factors are commonly observed, particularly when an element contains adjacent cell-specific and ubiquitous factor-binding sites. Mechanisms of cell-specific expression have expanded to embrace the concept of a “combinatorial code.” This model holds that combinations of several different transcription factors allow unique patterns of transcription, including cell-specific expression. This idea is illustrated for several genes that are uniquely expressed in the thyroid gland, such as thyroglobulin and thyroid peroxidase. An array of transcription factors that bind to these promoters have been identified. TTF-1 (thyroid transcription factor-1) is highly expressed in the thyroid, but it is also expressed in the lung and brain. Another factor, PAX-8 (a paired box member), is expressed in thyroid, but also in the kidney. Although each of these factors are expressed selectively in the thyroid, they are also expressed in other tissues. However, their combined expression is unique to the thyroid gland. The absence of expression of thyroglobulin in other locations such as lung or kidney emphasizes the need for the combinations of transcription factors to achieve highlevel tissue-specific expression. Transgenic studies provide additional evidence that highly stringent mechanisms control gene transcription. Although transient expression studies often suggest that a limited region of a promoter is sufficient for high-level expression in a cell-type specific manner, transgenic studies frequently show different results. It is not uncommon to find that the proximal promoter sequence alone results in relatively low expression in the expected target tissue, or that leaky expression occurs in other cell types. In many cases, additional 5'-flanking sequences, or the inclusion of other regions of the gene (exons, introns, 3'-flanking sequences), allow recapitulation of the native pattern of expression (e.g., proopiomelanocortin, vasopressin). Presumably, control elements that either enhance cell-specific expression or suppress ectopic expression reside in these other regions of the gene.

36

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Studies of c-fos promoter elements dramatically demonstrate the importance of combinatorial regulatory elements for normal function of the promoter in vivo. Transfection studies delineated a number of c-fos promoter elements including a sis-inducible element (SIE), SRE, an AP-1 site, and a Ca2+-cAMP response element (CARE). Each of these elements functioned independently in transfection assays and only partially decreased c-fos promoter activity when mutated. In contrast, a mutation in any one of these elements was sufficient to preclude c-fos expression in transgenic mice, indicating strong interdependence in vivo. TRANSCRIPTION FACTORS THAT INTERACT DIRECTLY WITH DNA SERVE TO RECRUIT ADDITIONAL REGULATORS THROUGH PROTEIN–PROTEIN INTERACTIONS The methods used to identify transcription factors have generally relied on interactions with DNA. For example, proteins that interact with enhancer elements can be identified by DNase I footprinting or gel shift assays. Using these and other strategies for transcription factor isolation, a large array of sequence-specific DNA-binding transcription factors have now been identified (Fig. 3-4). Although identification of these transcription factors represents a major advance, important questions remain concerning how the DNAbound transcription factors act to initiate transcription. Several lines of evidence have suggested that additional proteins serve as intermediates between enhancer binding factors and components of the basal transcription machinery. Genetic studies in yeast have identified a set of genes (SWI1, SWI2/SNF2, SWI3, SNF5, SNF6) that are involved in transcriptional activation and analogous proteins have been identified in Drosophila. Some of these proteins, such as SWI2/SNF2, appear to function by converting chromatin into a transcriptionally active, open form. A variety of viral proteins also appear to function as coactivators. For example, the adenovirus E1a protein can activate or repress transcription that is mediated by a variety of cellular genes, even though E1a does not bind directly to DNA. Attempts to reconstitute transcription systems in vitro also supports the existence of non–DNA-binding coactivators. Early studies were able to reconstitute basal transcription in vitro using proximal promoter regions and initiator elements. However, the addition of upstream regulatory elements that have major effects on transcription in transfected cells, or in transgenic models, rarely induces marked activation, even when cognate transcription factors are added. Although there have been some success in the reconstitution of enhancer-mediated transcription (e.g., progesterone receptor), most such experiments still employ relatively unpurified nuclear extracts. The presence of coactivators in these systems is suggested by the loss of transcriptional activation with additional purification steps and by the ability to inhibit transcription by including excess amounts of transcription factors. Alternatively, it has been possible to demonstrate potentiation of in vitro transcription by supplementing the reaction with rate-limiting factors. Complementation of HeLa cell extracts with fractionated proteins from B cells was shown to enhance the activity of Octdependent transcription from the immunoglobulin promoter. An additional line of evidence suggesting the presence of coactivators involves overexpression experiments in transfected cells. In this paradigm, overexpression of many DNA-binding transcription factors is inhibitory. This “squelching” phenomenon has been proposed to involve the titration of rate-limiting coactivators (Fig. 3-7B). Moreover, when different transcription factors are used, there is often crosstalk of the squelching effect, suggesting that

some coactivators are shared by several different transcription factors. For example, overexpression of an estrogen receptor can inhibit transcription by the glucocorticoid receptor, suggesting that they share a common coactivator (this is now known to be SRC-1, among others). In other cases, the squelching phenomenon appears to be less specific. For example, overexpression of a strong transactivator, such as the viral protein, VP16, can inhibit transcription by most genes. A key issue has been to elucidate whether these effects are nonspecific or actually serve as a bioassay for coactivators. Several experimental approaches now provide direct evidence for coactivators. Using a yeast one-hydrid approach, a B-cell coactivator, Bob-1, that interacts with the octamer-binding protein, was isolated. This factor contains an interaction domain that contacts the POU domain of Oct-1 or Oct-2 and can modify its DNA binding specificity as well as enhancing transcriptional activation. As noted, a number of coactivators for the nuclear receptors have been isolated (Table 3-1), primarily by using a yeast two-hybrid strategy, in which the activation domain of the receptor is used as bait. The coactivators interact directly with the nuclear receptors without making contacts with DNA. They interact with the receptor in a ligand-dependent manner, implying a conformational change in the receptor. One class of the steroid receptor coactivators, SRC, interacts with the carboxy-terminal activation domain through a hydrophobic motif, LXXLL. The coactivator contains additional sequences that are involved in transactivation. Exactly how the coactivators facilitate transactivation remains unclear, but it is likely that additional intermediates connect these factors with the basal transcription machinery. The CREB-binding protein (CBP), was identified based on direct protein–protein interactions. CREB was shown to detect a high molecular weight protein in farwestern blot assays, but only after phosphorylation of CREB at a residue (Ser 133) known to be required for transactivation. Based on this observation, radiolabeled phosphorylated CREB was used to screen cDNA expression libraries, allowing identification and cloning of CBP. CBP is postulated to function as an integrator of transcription, because in addition to CREB, it appears to mediate the actions of a large array of transcription factors and other coactivators (Fig. 3-11). A homolog of CBP, referred to as p300, was identified initially as target of adenovirus E1a protein. CBP contains distinct interaction domains for various classes of transcription factors. For example, the nuclear receptors interact with the aminoterminal region of CBP, whereas CREB interacts with a more central motif. CBP has been shown to enhance nuclear receptor transcription, particularly when combined with other coactivators such as SRC. Some experiments suggest that limiting amounts of CBP may provide a mechanism for transcriptional crosstalk between pathways. For example, AP-1-mediated transcription is inhibited by the activated glucocorticoid receptor, but this inhibition is reversed by the addition of excess CBP. CBP contains endogenous histone acetyltransferase (HAT) activity, and it also recruits a histone acetyltransferase to the promoter. One mechanism by which CBP may function is by altering histone acetylation and nucleosome structure. As noted at the beginning of this chapter, initial concepts of gene regulation were based on models of repressors. In addition to repressors that bind directly to DNA, there is now good evidence for corepressors that act to silence gene transcription through interactions with DNA-bound proteins. This group of factors is exemplified by the nuclear receptor corepressors, silencing

CHAPTER 3 / TRANSCRIPTIONAL CONTROL

37

Figure 3-11 Domains of CBP. CREB-binding protein (CBP) is a large (270 kDa) protein that interacts with a many different transcription factors. Its domains are shown above the figure and interacting transcription factors are shown below the figure. RAR, retinoic acid receptor; ER, estrogen receptor; TR, thyroid Hormone receptor; RXR, retinoid X receptor; GR, glucocorticoid receptor; PR, progesterone receptor; STAT2, signal transducers and activators of transcription; CREB, cAMP response element binding; SV40 TAg, Simian virus 40 large T antigen; TBP, TATAbinding protein; TFIIB, transcription factor IIB; SRC-1, steroid receptor coactivator. mediator of retinoid and thyroid receptors (SMRT) and nuclear receptor corepressor (N-CoR). These proteins were also cloned by a yeast two-hybrid approach. However, rather than binding to the receptors in the presence of ligand, they only interact in the absence of ligand. In addition to transfection studies, the ability of these factors to silence transcription has also been demonstrated using in vitro transcription assays, in which relief of silencing can be demonstrated by the addition of excess unliganded receptor (opposite of squelching). The mechanism of transcriptional silencing appears to involve Sin3A, which is a homolog of the yeast transcriptional repressor, Sin3p. In addition, a histone deacetylase interacts with Sin3A to form a multisubunit repressor complex. Thus, modifications of the state of histone acetylation, and the nature of nucleosome phasing near the regulatory region of genes, may represent a common pathway for control by transcriptional repressors and activators.

FEEDBACK MECHANISMS CONTROL THE TRANSCRIPTION OF MANY GENES Although much work has focused on mechanisms of transcriptional activation, it is also important to switch genes off in a controlled manner. In some cases, this is a matter of reversing the step that leads to activation. For example, transcription factors that are activated by phosphorylation can be dephosphorylated by specific phosphatases. For a few genes, a pathway of autoregulation has been identified in which the product of a gene acts to regulate its own promoter. Autoregulation can be positive to result in rapid induction, or it can be negative to inhibit transcription once a certain level of the gene product has been achieved. An example of both positive and negative feedback control occurs for the Pit-1 gene. The Pit-1 promoter contains regulatory elements that bind Pit-1, resulting in a positive feedback loop. However, high levels of Pit-1 are inhibitory, apparently because they bind to low-affinity sites downstream of the transcriptional start site. Thus, low levels of Pit-1 act positively, but high levels are inhibitory.

Figure 3-12 Autoregulation of transcription. Control of cAMP response element modulator (CREM) gene expression is shown as an example of autoregulation. The CREM gene contains two promoters, P1 and P2. The P1 promoter drives expression of a fulllength protein that contains both activation domains (Q1, P-Box, Q2) and DNA-binding domains (DBD). The P2 promoter is located within an intron and drives expression of a small protein that only contains the CREM dimerization and DNA-binding domains. The P2 promoter contains a series of cAMP response elements (CREs) that are activated by the adenylyl cyclase (AC)-cAMP-protein kinase A (PKA) cascade. This pathway stimulates CREB and CREM, which induce the activity of the P2 promoter as well as other cAMP-responsive genes. Activation of the P2 promoter results in expression of inducible cAMP early repressor (ICER), which functions as repressor of CREs. Thus, ICER terminates its own production from the CREM gene and also inhibits a subset of other cAMP-responsive genes. Another type of feedback inhibition is illustrated by control of CREM gene expression (Fig. 3-12). Like CREB, CREM is a phosphoprotein that regulates cAMP-responsive genes. However, CREM has multiple isoforms that act as either activators or repressors of transcription. CREM isoforms α, β, and γ are repressors, whereas isoforms τ, τ1, τ2, and τα are activators. Each of these isoforms are splicing variants derived from the P1 promoter. The CREM gene also contains an intronic promoter, P2, which transcribes ICER. The P2 promoter contains several cAMP response elements. Thus, activation of the cAMP pathway and CREM expression results in rapid induction of ICER. ICER is a 120-amino acid (13.4-kDa protein) that contains only the DNA binding and dimerization domains of CREM. Because it binds to DNA, but lacks transactivation domains, ICER functions as a powerful repressor of cAMP-induced transcription. ICER may also inhibit CREB and CREM by forming inactive heterodimers. Physiologically, the induction of ICER appears to terminate or dampen activation by the cAMP pathway. During spermatogenesis, CREMτ is initially induced in mature germ cells in response to stimulation by follicle-stimulating hormone (FSH). CREMτ

38

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

expression leads to an increase in CRE-driven promoters, including the P2 promoter that expresses ICER. Increasing levels of ICER feed back to attenuate the duration and amplitude of cAMPresponsive gene transcription. Hence, the increase in ICER protein downregulates its own transcription to terminate the cycle. This pathway also plays a role in the control of circadian rhythms. In the pineal gland, ICER levels vary dramatically during the 24-h cycle, with peak levels occurring at night. The enzyme serotonin Nacetyltransferase (NAT), which is involved in melatonin synthesis, is increased in CREM-deficient mice. Characterization of the NAT promoter revealed an ICER binding site, suggesting that increased expression results from the absence of ICER inhibition in the CREM-deficient mice.

ROLE OF TRANSCRIPTION FACTORS IN GENETIC AND ACQUIRED DISORDERS As noted, a substantial fraction of expressed genes encode transcription factors. It is therefore not surprising that transcription factors have been associated with a number of different genetic disorders (Table 3-2). Transcription factors are obvious candidates to cause genetic and neoplastic disorders because of their roles as regulators of developmental pathways and as switches that respond to environmental signals. Therefore, it is possible to eliminate the expression of a gene, not only by a direct mutation in that gene, but also as a result of a mutation in a transcription factor that controls expression of the gene (in trans) or by preventing the development of the cell lineage in which the gene is expressed. Pit-1 mutations illustrate these points. Pit-1 is a POU-homeodomain protein that is selectively expressed in certain pituitary gland cell types. Pit-1 directly regulates the expression of pituitary genes including the growth hormone and prolactin genes (see Chapter 49). Mutations in Pit-1 cause deficiencies in growth hormone and prolactin and also impair the development of the somatomammotrope cell lineage that produces these hormones. In this manner, the transcription factor mutation results in clinical manifestations that resemble mutations in the target genes (growth hormone, prolactin). However, in this case, the phenotype is broadened (includes thyroid-stimulating hormone) because the transcription factor participates in a cascade that controls the development of cell lineage in the pituitary. Several other examples of transcription factors that control developmental cascades are summarized in Table 3-2. Genes in the paired box homeotic (PAX) family cause a wide variety of development disorders, which reflect their restricted patterns of distribution (see Chapter 112). The bHLH (basic helix-loop-helix) group of transcription factors have also been identified in many genetic disorders (e.g., Twist, MITF, MXI-1). This family of factors, which typically bind to DNA as heterodimers, are frequently involved in cell lineage development and tissue-specific expression (see Chapter 92). In addition to disorders that have already been identified, gene knockout experiments in mice predict an important role for these factors in other developmental disorders. The high-mobility group (HMG) proteins, SRY and SOX-9, comprise part of a pathway that leads to normal male sexual differentiation (see Chapters 57 and 58). Mutations in SRY cause XY sex reversal, and mutations in SOX-9 cause sex reversal in addition to skeletal abnormalities. The exact function of the HMG proteins remains unclear, but they are thought to induce DNA bending, which may facilitate the transcription of target genes. The homeodomain proteins, HNF-4α, HNF-1α, and IPF-1, participate

in a pathway that leads to normal pancreatic islet cell development and function (see Chapter 47). Mutations in HNF-4α and HNF-1α cause a form of late-onset diabetes mellitus referred to as maturity onset diabetes of the young (MODY). Homozygous mutations in IPF-1 cause pancreatic agenesis, whereas heterozygous mutations appear to cause another form of MODY. These genes illustrate how mutations in several different steps in the same developmental pathway can result in phenotypically similar disorders (nonallelic genetic heterogeneity). Mutations in the steroid hormone superfamily of nuclear receptors cause hormone resistance syndromes. For example, androgen receptor mutations (X-linked) cause androgen insensitivity. Severe mutations result in a syndrome referred to as testicular feminization, in which there is absolute resistance to androgen action, and the sexual phenotype appears female, even though inguinal testes are present and high levels of testosterone are produced (see Chapter 60). Similarly, estrogen, glucocorticoid, and vitamin D resistance syndromes result from mutations in their respective nuclear hormone receptors. Although these phenotypes can be predicted in part, the naturally occurring mutations have helped to elucidate the functional roles of the hormones and their receptors. For example, the estrogen receptor mutation prevented epiphyseal closure in the long bones, revealing that estrogen action is required (and cannot be substituted by androgens) for this event. In contrast to the other nuclear receptors, mutations in the thyroid hormone receptor illustrate a distinct property of some transcription factor mutations: the ability to function in a dominant negative manner. Thyroid hormone receptor mutations are transmitted in an autosomal dominant manner and cause partial end organ resistance to thyroid hormones (see Chapter 50). The basis for dominant transmission has been shown to involve the binding of mutant thyroid hormone receptors to target genes to block the function of the receptor from the normal allele. Consistent with this model, these receptor mutants retain the ability to form dimers and to bind to DNA, but they lack the ability to bind hormone or to initiate transcription. Analogous dominant negative mutations have been described for Pit-1, and are likely to be relatively common among transcription factor disorders, because they are manifest in the heterozygous state, and the modular nature of transcription factors lends itself to mutations that selectively inactivate one functional domain while preserving other features such as dimerization or DNA binding. The relevance of modular functional domains in transcription factors in disease is also illustrated by the formation of chimeric factors in neoplasia (see Chapter 7). Particularly in leukemias and lymphomas (see Chapters 26 and 27), but also in soft tissue tumors (see Chapter 95), translocations combine different transcription factor domains or bring a transcriptional activation region under the control of a heterologous gene. Several examples of such translocations are listed in Table 3-2. In one case, the t(1;19)(q23; p13) translocation in pre–B-cell acute lymphoblastic leukemia fuses the PBX homeodomain gene to the transactivation domain of the E2A transcription factor gene. Similarly, in promyelocytic leukemia, a translocation t(15;17)(q22;q21) combines a zinc finger ring domain (PML) with the ligand binding and transactivation domain of the retinoic acid receptor. In this case, the chimeric receptor retains responsiveness to retinoic acid, which has been shown to induce differentiation of leukemic cells. The FUS-CHOP translocation (12;16)(q13;p11) may combine features of chimeric proteins and dominant negative activity and is associated with myxoid

39

CHAPTER 3 / TRANSCRIPTIONAL CONTROL

Table 3-2 Examples of Transcription Factor Mutations and Rearrangements That Cause Disease Transcription factor Developmental/multi-organ Brn-4a PAX-2 PAX-3 PAX-6 Ptx-1 Twist MITF CBFα1 GLI-3 CBP Endocrine Pit-1 SRY SOX-9 HNF-4 HNF-1α IPF-1 TRβ AR ERα GR VDR DAX-1 Cancer/hematology Rb p53 WT1 MXI-1 TFIIH VHL XH2 HMGI-C PTH-Cyclin D1 FUS-CHOP Pax-3,7-FKHR EWS-ATF-1 E2A-PBX MLL-ELL CBFβ-MYHII MLL-AF9 PML-RAR PDGFR-ETS TCR-Hox-11 TCR-TTG-1,2 TCR-TAL-1,2 MLL-AF4 TEL-AML1 E2A-HLF Ig-MYC RFX-A

Class of factor

Disorder

POU domain Paired box homeotic Paired box homeotic Paired box homeotic Bicoid homeodomain bHLH bHLH Runt domain Kruppel Bromodomain coactivator

X-linked deafness; stapes fusion Colobomas optic nerve/ renal hypoplasia Waardenburg syndrome type 1 Aniridia; Peter’s anomaly Rieg syndrome Saethre-Chotzen syndrome Microphthalmia; Waardenburg 2A Cleidocranial dysplasia Greig cephalopolysyndactyly Rubenstein-Taybi syndrome

POU-homeodomain HMG HMG Nuclear receptor Homeodomain Homeodomain Zinc finger nuclear receptor Zinc finger nuclear receptor Zinc finger nuclear receptor Zinc finger nuclear receptor Zinc finger nuclear receptor Nuclear receptor-like

Pituitary insufficiency Sex reversal Campomelic dysplasia; sex reversal MODY1 MODY3 Pancreatic agenesis; MODY 4 Resistance to thyroid hormone Androgen insensitivity Estrogen resistance Glucocorticoid resistance Vitamin D-resistant rickets type IIA Adrenal hypoplasia congenita

Cell cycle HLH-like; cell cycle Zinc finger bHLH General transcription factor Transcription elongation Chromatin remodeling HMG Translocation; cell cycle Trans. factor-b-Zip Paired box homeotic-forkhead Translocation; b-Zip Transc. factor-Homeodomain Translocation; elongation Translocation; MYHII Translocation; AF9 Translocation; Retinoid action Translocation; ETS Translocation; Homeodomain Translocation; LIM Translocation; Twist Translocation; AF4 Translocation; AML1 Translocation; b-Zip Translocation; bHLH RFX

Retinoblastoma; other cancers Li-Fraumeni syndrome; other cancers Wilm’s tumor; Denys-Drash syndrome Prostate cancer Xeroderma pigmentosa; Cockayne syndrome Von-Hippel-Lindau syndrome; Renal Ca Mental retardation; α thalassemia Benign mesenchymal tumors Parathyroid adenoma Mixed liposarcomas Alveolar rhabdomyosarcoma Ewing’s sarcoma Acute lymphoblastic leukemia Acute myeloid leukemia Acute myeloid leukemia Acute myeloid leukemia Acute promyelocytic leukemia Chronic myelogenous leukemia T-cell leukemia T-cell leukemia T-cell leukemia Leukemia Leukemia Leukemia Lymphoma; leukemia Bare lymphocytes; MHC class II deficiency

aMany transcription factor abbreviations are now standard nomenclature. Selected abbreviations include: Brn-4, Brain-4; PAX, paired box homeotic gene; Ptx, pituitary homeo box; MITF, microphthalmia associated transcription factor; CBF, core-binding factor; GLI, amplified in Glioblastoma; CBP, CREB (cAMP responsive element binding)-binding protein; Pit, pituitary specific transcription factor; SRY, sex determining region Y; SOX, SRY box; HNF, hepatocyte nuclear factor; IPF, insulin promoter factor; TR, thyroid hormone receptor; AR, androgen receptor; ER, estrogen receptor; GR, glucocorticoid receptor; VDR, vitamin D receptor; DAX, dosage-sensitive sex-reversal-adrenal hypoplasia congenita critical region on the X-chromosome; Rb, retinoblastoma; WT, Wilm’s tumor; MXI, MAX-interacting protein; VHL, Von Hippel-Lindau; HMG, high-mobility group; PTH, parathyroid hormone; FUS, fusion; EWS, Ewing’s sarcoma; ATF, activating transcription factor; CHOP, C/EBP homologous protein; FKHR, Forkhead related; PBX, pre–B-cell leukemia; MLL, mixed lineage leukemia; ELL, elongation factor homologous to MLL; MYH, myosin heavy-chain polypeptide; PML, promyelocytic leukemia; RAR, retinoic acid receptor; PDGFR, platelet-derived growth factor receptor; HOX, homeobox; TCR, T-cell receptor; TAL, T-cell acute leukemia; TEL, translocation ETS leukemia; RFX-A, regulatory factor associated.

40

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

liposarcomas. The CHOP protein is normally a dominant inhibitor of members of the b-Zip family of transcription factors. However, in combination with FUS, which contains a transactivation domain, its inhibitory properties may be converted to activation. In other cases, the translocation results in ectopic expression of a transcription factor. This property is illustrated by translocations such as TCR-TAL, in which the T-cell receptor gene results in targeted overexpression of the twist-like protein, TAL. Similarly, the fusion of the Ig promoters to Myc causes overexpression of Myc in certain lymphomas and leukemias, and the PTH promoter has been shown to drive expression of cyclin D1 in a subset of parathyroid adenomas (see Chapter 51). Transcription factors that are linked to cell-cycle control, such as Rb and p53, play major roles in neoplasia (see Chapter 7). Rb was discovered as a result of mapping the locus for the childhood disorder, retinoblastoma (see Chapter 105). This disorder established the principle of the so-called two-hit model for certain dominantly transmitted cancer syndromes. In this case, an inherited mutation in one copy of Rb creates a circumstance in which a second somatic mutation results in the inactivation of both alleles, leading to clonal expansion of neoplastic cell. Subsequent studies have shown that somatic Rb mutations occur in many different kinds of cancers (e.g., osteosarcoma, brain tumors, breast, genitourinary), some of which overlap with the tumors that develop in patients with germline mutations in Rb. The principle of the two-hit model for germline mutations is also illustrated by p53, which causes Li-Fraumeni syndrome (early breast cancer, soft tissue sarcomas, brain tumors, adrenal tumors, and so on). p53 activates a number of genes involved in cell cycle control including p21. p53 acts at the G1/S checkpoint and also governs the ability to arrest cell growth and to permit cells to undergo apoptosis (see Chapter 6). In addition to germline mutations in p53 in Li-Fraumeni syndrome, somatic mutations in p53 are among the most common tumor suppressor gene mutations in cancer (see Chapter 40). It is surprising that mutations in factors such as the RARs do not have even more severe effects on cellular function. The phenotypes of the mice with knockouts of individual RAR isoforms (e.g., α, β, γ) are relatively subtle (occasional skeletal defects). The absence of a more severe phenotype has been attributed to the possibility of redundant function by the other isoforms. Other examples are emerging in which mutations in factors presumed to play a fundamental role in the function of the cell (and therefore assumed to be embryonic lethal) actually cause relatively specific postnatal syndromes. For example, a mutation in a component of the general transcription factor, TFIIH, is one cause of xeroderma pigmentosa. Similarly, the Von Hippel-Lindau gene product, which is involved in the control of transcriptional elongation, causes a syndrome consisting of predisposition to renal cancers, pheochromocytomas, retinal angiomas, and hemangiomas of the central nervous system (see Chapter 71). Another example is provided by the protein, CBP, which serves as a transcriptional integrator that links other transcription factors to the basal transcription factors. CBP is a coactivator for an enormous number of transcription factors including CREB and AP-1, nuclear receptors, c-myb, and STAT proteins (Fig. 3-11). Haploinsufficiency of CBP causes a severe, and characteristic, syndrome referred to as RubensteinTaybi. These patients have mental retardation, characteristic facies, broad thumbs, and great toes. In addition to its remaining normal allele, CBP function may be complemented to some

degree by a homologous protein, p300. In view of the fraction of the genome that is dedicated to expression of transcription factors, we can expect the list of disorders caused by transcription factor mutations to grow further.

SELECTED REFERENCES Alland L, Muhle R, Hou H. Role for N-CoR and histone deacetylase in Sin3-mediated transcriptional repression. Nature 1997;387:49–55. Aso T, Shilatifard A, Conaway JW, Conaway RC. Transcription syndromes and the role of RNA polymerase II general transcription factors in human disease. J Clin Invest 1996;97:1561–1569. Bannister AJ, Kouzarides T. The CBP co-activator is a histone acetyltransferase. Nature 1996;384:641–643. Baron MH. Transcriptional control of globin gene switching during vertebrate development. Biochim Biophys Acta 1997;1351:51–72. Chiba H, Muramatsu M, Nomoto A, Kato H. Two human homologues of Saccharomyces cerevisiae SWI2/SNF2 and Drosophila brahma are transcriptional coactivators cooperating with the estrogen receptor and the retinoic acid receptor. Nucleic Acids Res 1994; 22:1815–1820. Choy B, Green MR. Eukaryotic activators function during multiple steps of preinitiation complex assembly. Nature 1993;366:531–536. Glass CK, Rose DW, Rosenfeld MG. Nuclear receptor coactivators. Curr Opin Cell Biol 1997;9:222–232. Gstaiger M, Knoepfel L, Georgiev O, Schaffner W, Hovens CM. A B-cell coactivator of octamer-binding transcription factors. Nature 1995;373:360–362. Guarente L. Transcriptional coactivators in yeast and beyond. Trends Biochem Sci 1995;20:517–521. Heinzel T, Lavinsky RM, Mullen TM, et al. A complex containing NCoR, mSin3 and histone deacetylase mediates transcriptional repression. Nature 1997;387:43–48. Hill CS, Treisman R. Transcriptional Regulation by Extracellular Signals: Mechanisms and Specificity. Cell 1995;80:199–211. Judson HF. The Eighth Day of Creation: Makers of the Revolution in Biology. New York: Simon and Schuster, 1979; pp. 1–686. Kamei Y, Xu L, Heinzel T, Torchia J, Kurokawa R, Gloss B, et al. A CBP integrator complex mediates transcriptional activation and AP-1 inhibition by nuclear receptors. Cell 1996;85:403–414. Kingston RE, Green MR. Modeling eukaryotic transcriptional activation. Curr Biol 1994;4:325–332. Kwok RP, Lundblad JR, Chrivia JC, et al. Nuclear protein CBP is a coactivator for the transcription factor CREB. Nature 1994;370:223–226. Laherty CD, Yang WM, Sun JM, Davie JR, Seto E, Eisenman RN. Histone deacetylases associated with the mSin3 corepressor mediate mad transcriptional repression. Cell 1997;89:349–356. Latchman DS. Transcription-factor mutations and disease. N Engl J Med 1996;334:28–33. Lewin B. Genes V. Oxford; Oxford University Press, 1994; pp. 1–1272. Liu F, Green MR. Promoter targeting by adenovirus E1a through interaction with different cellular DNA-binding domains. Nature 1994;368: 520–525. Paranjape SM, Kamakaka RT, Kadonaga JT. Role of chromatin structure in the regulation of transcription by RNA polymerase II. Annu Rev Biochem 1994;63:265–297. Ptashne M, Gann A. Transcriptional activation by recruitment. Nature 1997;386:569–577. Rabbitts TH. Chromosomal translocations in human cancer. Nature 1994;372:143–149. Rhodes SJ, Chen R, DeMattia GE, et al. A tissue-specific enhancer confers Pit-1-dependent morphogen inducibility and autoregulation on the pit-1 gene. Genes Dev 1993;7:913–932. Robertson LM, Kerppola TK, Vendrell M, et al. Regulation of c-fos expression in transgenic mice requires multiple interdependent transcription control elements. Neuron 1995;14:241–252. Roeder RG. The role of general initiation factors in transcription by RNA polymerase II. Trends in Biochem Sci 1996;21:327–335. Sassone-Corsi P. Transcription factors responsive to cAMP. Annu Rev Cell Dev Biol 1995;11:355–377.

CHAPTER 3 / TRANSCRIPTIONAL CONTROL

Sauer F, Hansen SK, Tjian R. DNA template and activator-coactivator requirements for transcriptional synergism by Drosophila bicoid. Science 1995;270:1825–1828. Shikama N, Lyon J, La Thangue NB. The p300/CBP family: integrating signals with transcription factors nad chromatin. Trends in Cell Biology 1997;6:230–236. Smale ST. Transcription initiation from TATA-less promoters within eukaryotic protein-coding genes. Biochim Biophys Acta 1997;1351:73–88. Smith CL, Onate SA, Tsai MJ, O’Malley BW. CREB binding protein acts synergistically with steroid receptor coactivator-1 to enhance steroid receptor-dependent transcription. Proc Natl Acad Sci USA 1996;93:8884–8888.

41

Tjian R, Maniatis T. Transcriptional activation: a complex puzzle with few easy pieces. Cell 1994;77:5–8. Torchia J, Rose DW, Inostroza J, et al. The transcriptional co-activator p/ CIP binds CBP and mediates nuclear- receptor function. Nature 1997;387:677–684. Wade PA, Wolffe AP. Chromatin: histone acetyltransferases in control. Curr Biol 1997;7:R82–R84. Wolffe AP. Transcription: in tune with the histones. Cell 1994;77:13–16. Wolffe AP. Transcriptional control. Sinful repression. Nature 1997;387:16, 17. Zwicker J, Muller R. Cell-cycle regulation of gene expression by transcriptional repression. Trends Genet 1997;13:3–6.

CHAPTER 4 / TRANSMISSION OF GENETIC DISEASE

4

43

Transmission of Human Genetic Disease PETER KOPP AND J. LARRY JAMESON

INTRODUCTION

sequence of a gene are first transcribed into mRNA. The mRNA is translocated from the nucleus into the cytoplasm, where it is translated by ribosomes into a string of amino acids (protein synthesis).

Molecular DNA analysis has become an integral part of all medical specialties. Although the immense size and complexity of the human genome is daunting, recent progress in the Human Genome Project has greatly increased our knowledge of the genetic basis of many human diseases (see Chapter 5). The impact of molecular DNA analysis and genetics in clinical medicine includes new approaches for the diagnosis of diseases, the detection of pathogens, screening for disease predisposition, genetic counseling, drug development, pharmacotherapy, and, in selected cases, gene therapy. More than 3000 human diseases are known to be caused by defects in single genes and to follow a Mendelian mode of inheritance. Moreover, many disease processes are influenced by the genetic background of the affected individual. There is often a complex interaction between environmental factors and genetic predisposition. Many of these disorders (e.g., hypertension, coronary artery disease, asthma, diabetes) represent major public health problems and the elucidation of their pathogenesis remains an extraordinary challenge. Other genetically determined forms of disease include the syndromes caused by chromosomal aberrations and inherited cancer syndromes, as well as nonhereditary somatic cell DNA defects that occur in cancer. DNA AND THE TRANSMISSION OF GENETIC INFORMATION The double-stranded macromolecule DNA contains the genetic information. It consists of two complementary strands that wrap around one another to form a double helix. Each strand is a linear arrangement of four different bases: adenine (A), thymine (T), cytosine (C), and guanine (G). The order of the bases forms the DNA sequence. The two DNA strands are held together by hydrogen bonds formed between the complementary bases G and C, or A and T. Because the nucleotide chains within the doublestranded DNA are strictly complementary, each can serve as template for the formation of a new strand. This semiconservative form of replication ensures that each dividing cell receives an identical copy of DNA (see Chapter 6). In the coding sequence of a gene, each set of three DNA bases forms a codon, the genetic code for the incorporation of one of the 20 amino acids found in proteins (Fig. 4-1). A series of codons thus defines which amino acids will be synthesized into a given polypeptide. The protein-coding instructions found in the DNA

CHROMOSOMES, GENES, AND DNA A detailed description of molecular genetic techniques is beyond the scope of this chapter, but it is useful to present some of the key concepts and techniques that are essential for the analysis of the genome and genetic diseases (see Chapters 1 and 2). Aspects of gene manipulation are also covered in chapters on transgenic models (see Chapter 10) and gene therapy (see Chapter 18). SIZE AND ORGANIZATION OF THE HUMAN GENOME The genome is encoded by a coiled polymer of deoxyribonucleic acid (DNA) that is separated into several segments, the chromosomes (see Chapter 1). In humans, every somatic cell is diploid and contains two sets of 23 chromosomes. A child receives one copy of every chromosome from its mother, and one copy from its father. The 22 autosomal chromosomes are present in two copies, and are similar in females and males. The sex chromosomes, X and Y, are dissimilar. Only a small segment of the short arm of both the X and Y chromosomes, the pseudoautosomal region (PAR), shares homology. Females have two X chromosomes (XX), while males are heterogametic (XY). The DNA of a single or haploid set of chromosomes consists of approximately 3 × 109 bp (3000 Mb), with the smallest chromosome containing ~50 Mb (chromosome 21) and the largest ~263 Mb (chromosome 1). A gene consists of regulatory DNA sequences, coding sequences (exons), and intervening noncoding sequences (introns). The sequences of a gene that encode a protein are located in exons, which are spliced together to form the messenger ribonucleic acid (mRNA). Based on the number of expressed mRNAs, it has been estimated that the human genome contains about 70,000 to 100,000 genes. The size and structure of genes is highly variable. Some small genes consist of a few hundred basepairs, while the largest gene known to date, the X-chromosomal gene encoding the protein dystrophin, contains about 2 × 106 bp with more than 100 introns. Mutations in this gene lead to two distinct clinical conditions, Duchenne Muscular Dystrophy (DMD) and Becker Muscular Dystrophy (BMD). Some chromosomal regions, particularly near the centromeres, do not contain genes, while others display a high gene density. Occasionally, genes are grouped into clusters in which several copies of genes with similar function are located near one another. Examples of gene clusters are the growth hormone genes, or the genes encoding α and β globin. In other

From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

43

44

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Figure 4-1 Genetic code. A codon for an amino acid consists of three nucleotides (triplet). One amino acid may be encoded by several different codons (degeneracy of the genetic code). The code includes sequences for the start codon and the stop codon.

instances, functionally related genes are located on different chromosomes. Expressed genes encompass only about 20% of the total genome. A large amount of genomic DNA is not formed of singlecopy sequences, but consists of repetitive sequences of various types. Short interspersed repetitive elements (SINE; 100–500 bp) and long interspersed repetitive elements (LINE; 6000–7000 bp) are spread throughout the genome. Another form of repetitive sequences, microsatellites, consist of 10–50 copies of simple sequence repetitions. Their functional role is not well defined, but they provide an important tool for establishing genetic maps and performing linkage studies (see below). MITOSIS AND MEIOSIS For tissue development and growth, eukaryotic cells progress through cycles of cell division and differentiation (see Chapter 6). Each cell division (mitosis) results in two genetically identical diploid (2n chromosome sets) daughter cells that are derived from the precursor cell. The first step in mitosis is the duplication of each parental chromosome, yielding two pairs of sister chromatids (2n → 4n). The chromosomes then condense and the nuclear membrane disappears. Subsequently, the chromosomes are arranged along the equatorial plate at metaphase. The two identical sister chromatids, held together at the centromere, are attached to the mitotic spindles. They then divide and migrate to opposite poles of the cell. After formation of a nuclear membrane around the two separated sets of chromatids, the cell divides and two diploid daughter cells are formed. The gametes, oocytes or sperm, contain a single copy of each chromosome, a haploid (1n) chromosome set. When a sperm fertilizes an egg, the two haploid sets are combined, resulting in a new zygote containing two copies of each chromosome (diploid or 2n). The generation of the haploid gametes from diploid germ cells (oogonia, spermatogonia) occurs through meiosis, which consists of two cell divisions. During the first mitotic cell division, there is an exchange between homologous chromosomes generating

Figure 4-2 Crossing-over and recombination. During chiasma formation, either of the two chromatids on one chromosome pairs with one of the chromatids of the homologous chromosome. Genetic recombination through crossing-over results in recombinant and nonrecombinant chromosome segments in the gametes. Such recombination events occur frequently and, together with the random segregation of maternal and paternal chromosomes, generates genetic diversity in gametes.

genetically unique gametes. After formation of the two sister chromatids (2n → 4n), homologous chromosomes pair and form chiasmata (Fig. 4-2). Either of the two chromatids can pair with the homologous chromosome and undergo a process called crossingover, in which segments of the maternal homolog recombine with the paternal homolog to form hybrid chromosomes in the place of the original ones. Such recombination events occur frequently, and it appears that at least one chiasma occurs on each chromosome arm. After these exchanges have been completed, the chromosomes segregate randomly. Because there are 23 chromosomes, there are 223 (>8 million) possible combinations of chromosomes. Together with the exchanges that occur during recombination, chromosomal segregation generates great genetic diversity in gametes. After the first meiotic division, which gives rise to two daughter cells (2n), the two chromatids of each chromosome are separated during the second meiotic division and this results in four gametes with one chromatid (1n). GENOTYPE AND PHENOTYPE A segment of DNA that is inherited in a Mendelian fashion (e.g., a gene) is called a locus. The genetic information at a given gene locus is defined as a genotype. Differences in the DNA sequence, and thus the genotype, are defined as alleles. If two alleles are present at a given locus, there are three possible genotypes. The genotype can be identical (homozygous) for one or the other allele, or it can consist of the combination of the two alleles (heterozygous). If there are more than two alleles, the number of possible genotypes increases accordingly. The phenotype defines the visible or measurable characteristics of an individual. The term wild type describes the normal gene, including different allelic variants. A mutant genotype may or may not change the phenotype. If the phenotype of a mutant allele can be recognized in the heterozygous state, it is dominant. If alleles can only be recognized in the homozygous state, they are recessive. If both alleles can be recognized in the heterozygous state, they are codominant (e.g., the alleles A and B in the ABO blood system).

CHAPTER 4 / TRANSMISSION OF GENETIC DISEASE

The widespread nature of genetic diversity is illustrated by subtle phenotypic variations between, and within, different ethnic groups. Biochemical polymorphisms, not readily perceptible by phenotype, can also be detected in many normal proteins that exist in various forms in a population. These variants are explained by the existence of multiple alleles at the level of the gene. The first such polymorphism to be clearly defined was the ABO blood group system. Molecular genetic analyses have revealed a surprisingly great degree of genetic variation. About 1 in 200 bp in the human genome is polymorphic. They are particularly frequent in noncoding regions of the genome and include single-basepair changes and variations in repetitive sequences. A polymorphism may change the protein sequence when it is located in coding regions and the substitution changes the amino acid specified by a codon. Polymorphisms are inherited according to Mendelian laws, a feature that is used in linkage studies or forensic applications. Polymorphisms may have no impact on phenotype, or they may contribute to population diversity without being associated with disease. Alternatively, they may contribute to the susceptibility to, or expression of, certain diseases. A high degree of genetic diversity increases the ability of a population to adapt to changing environmental conditions, and it decreases the risk of recessive diseases. GENETIC HETEROGENEITY Genetic heterogeneity refers to a similar phenotypic alteration that results from defects in different genes. For example, genetic heterogeneity occurs in Maturity-Onset Diabetes of the Young (MODY), which can result from defects in the glucokinase gene, as well as the genes encoding HNF-1α (Hepatocyte nuclear factor-1α) or HNF-4α (see Chapter 47). Allelic or intragenic heterogeneity refers to alternate genotypes at the same locus that result in the same or similar phenotype. In cystic fibrosis, more than 600 mutations have been identified in the gene encoding the cystic fibrosis transmembrane conductance regulator protein (see Chapter 37). In allelic heterogeneity, linkage studies will identify the same locus. In contrast, genetic heterogeneity results in a similar phenotype, but linkage analyses identify distinct loci in different families.

TECHNIQUES USED IN GENETICS RESTRICTION ENDONUCLEASES Restriction endonucleases are bacterial enzymes that recognize and cleave DNA at specific nucleotide sequences, called restriction sites (see Chapter 2). The most common restriction sites are sequences of 4–6 bp. Many of these sequences are palindromic (twofold axis of symmetry). A restriction enzyme recognizing a four-base sequence will find a target sequence every 44 (256) bases, a six-base cutter every 46 (4036) bases in a randomly composed DNA fragment. Cleavage can result in blunt ends of DNA or short single-stranded ends known as cohesive or sticky ends. The ability to manipulate DNA with restriction enzymes is of fundamental importance and includes applications like Southern analysis (see below) and the cloning of DNA fragments. If a sequence variation leads to loss or introduction of restriction sites, it is called a restriction fragment length polymorphism (RFLP). These polymorphic markers can be used for linkage studies. PCR The polymerase chain reaction (PCR), introduced in 1985, has profoundly changed the way DNA analyses are performed, and it has become a cornerstone of molecular biology and

45

Figure 4-3 Polymerase chain reaction. The polymerase chain reaction (PCR) generates multiple copies of a DNA segment. The double stranded (ds) template DNA is denatured, specific synthetic oligonucleotides primers of about 20 bp are annealed on each side of the segment of interest, and the complementary strand is then synthesized by a heat-stable DNA polymerase. After each cycle, the dsDNA from the preceding cycle serves as template. The amount of DNA is therefore doubled after every cycle, resulting in an exponential increase.

genetic analysis. PCR allows the amplification of a defined DNA segment several millionfold. In the first step, the double stranded (ds) template DNA is denatured by heating (Fig. 4-3). Second, specific synthetic oligonucleotides primers of about 20 bp are annealed on either side of the segment of interest. The complementary strand located between the primers is then synthesized by a DNA polymerase. This procedure is usually repeated for about 30 cycles using automated heating blocks. After each cycle, all the dsDNA present from the preceding cycle serves as template. The DNA is therefore doubled after every cycle, resulting in an exponential increase in the amount of replicated DNA. Reverse transcriptase-PCR (RT-PCR) is another important strategy for molecular analysis. The starting material in this case is mRNA. The mRNA is reverse-transcribed into cDNA by the enzyme reverse transcriptase (first-strand synthesis). Subsequently, cDNA segment(s) can be amplified by PCR as described above. DNA SEQUENCING Several protocols have been developed for the determination of DNA nucleotide sequence. In both traditional methods (Maxam-Gilbert or Sanger), the template DNA is used to generate fragments of DNA, which differ in size. Using high-resolution gel electrophoresis, these DNA molecules are separated at single-base resolution, allowing the sequence of the template DNA to be deduced. Maxam-Gilbert sequencing takes advantage of chemicals that cleave DNA at specific bases, thus resulting in fragments of different lengths. In the Sanger sequenc-

46

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Figure 4-4 Automated DNA sequencing using fluorescently labeled dideoxynucleotides. Among the many methods in use for DNA sequencing, the chain termination methodology is widespread, and automated procedures are now used with increasing frequency. Automated procedures use fluorescently labeled dideoxynucleotides (or primers) and direct computer analysis of the sequence. A sequencing primer is bound to the template DNA. Dideoxynucleotides carrying different fluorescent labels are included in these reactions to terminate DNA polymerization when they are incorporated instead of the normal nucleotide. Electrophoresis of these reactions allows separation of the elongation products, which are directly analyzed by a computer.

ing method, DNA chains of varying length are generated by using dideoxynucleotides to terminate extension of the sequence by a DNA polymerase (chain termination method). By running four reactions in parallel (one for each dideoxynucleotide) and separating the products by electrophoresis on polyacrylamide gels, it is possible to determine the sequence of the template. Automated procedures are now used with increasing frequency to analyze DNA sequences. These procedures are usually based on the chain termination method and employ fluorescently labeled dideoxynucleotides or primers followed by direct computer analysis of the sequence. A typical example is shown in Fig. 4-4. These methods are, however, still relatively labor- and cost-intensive. For example, it has been estimated that sequencing the whole human genome by such an approach would result in costs of at least $3 billion at $1–$2 per base, and that it would require 30,000 work-years. Currently, efforts are underway to develop new DNA sequencing technologies that are faster, more sensitive, and more costeffective. Methodologies that are being explored include detection of fluorescently labeled bases in flow cytometry, direct reading of the base sequence on a DNA strand with the use of scanning, tunneling, or atomic force microscopies, mass spectrometric analy-

sis of DNA sequence, and sequence analysis with DNA chips consisting of large arrays of oligonucleotides to which the DNA can be hybridized. CYTOGENETICS AND FLUORESCENT in situ HYBRIDIZATION (FISH) Chromosomes can be stained and visualized under a light microscope. They display a characteristic pattern of light and dark bands reflecting regional variations in DNA composition. Differences in size and banding patterns allows one to distinguish the 22 autosomes and sex chromosomes to determine the karyotype of an individual. This type of cytogenetic analysis reveals the presence of major chromosomal abnormalities, including missing or extra copies of a chromosome, gross deletions, insertions, or translocations. Fluorescent in situ hybridization (FISH) combines conventional cytogenetics with DNA hybridization. Chromosomes are prepared from nucleated cells and hybridized with fluorochromelabeled DNA probes specific for various loci. Numerical and structural aberrations can then be detected by analysis of the fluorescence pattern. FISH is gradually replacing conventional cytogenetic analyses. The high resolution of FISH is useful for diagnostic purposes, and it is also helpful in the construction of chromosomal maps. Using interphase chromosomes, such probes can improve map resolution to about 100,000 bp. GENE MAPPING Given the size and complexity of the human genome, the identification and location of genes is not trivial. Maps of the entire genome have now been developed to facilitate the localization of a gene of interest. Different types of maps (genetic or physical) describe the order of markers and the distances between them on each chromosome. One of the primary goals of the Human Genome Project is to increase the resolution of these maps until the exact location of every gene is known (see Chapter 5). A genetic map determines the relative position of a gene or locus based on the recombination frequencies relative to other loci on the same chromosome. It is expressed in recombination units or centiMorgans (cM). The genetic map is up to 40% longer in chromosomes derived from females because of a higher frequency of recombination during the formation of oocytes. Any polymorphic sequence whose inheritance pattern can be followed is useful for mapping purposes. Examples of useful markers include RFLPs and microsatellite repeats. The genetic map is then constructed by assessing how frequently two markers are inherited together by linkage studies. The chromosomal location of a gene responsible for an inherited disease can be determined by tracking the inheritance of DNA markers (see below). Physical maps indicate the position of loci in absolute values. The various types of physical maps differ in their degree of resolution. A cytogenetic or chromosomal map determines the position of genetic loci relative to characteristic chromosomal bands observed under light microscopy. cDNA maps are of particular interest because they allow analyses of expressed genomic regions. After synthesis of cDNA from mRNA using reverse transcriptase, the origin of the cDNA can be mapped to particular chromosomal regions by hybridization. Sequence-tagged sites (STS) are short DNA segments with known locations on the genetic map (see Chapter 5). They can be used to array DNA fragments that have been cloned into yeast artificial chromosomes (YACs). The presence or absence of a given STS in a cloned fragment of DNA can then be compared in different YACs to identify overlapping clones. This leads to the characterization of contiguous DNA

CHAPTER 4 / TRANSMISSION OF GENETIC DISEASE

47

Figure 4-5 STS content map. The presence or absence of sequence tagged sites, short DNA segments of known sequence, can be determined in various clones of yeast artificial chromosomes (YAC). The comparison of STS sequences in different YAC clones allows their order to be determined.

sequences, commonly referred to as contigs (Fig. 4-5). The highest resolution physical map will provide the complete DNA sequence of each chromosome in the human genome. LINKAGE Genetic linkage provides the basis for the development of genetic maps, and it is used for the detection of new genes by positional cloning (see below). Linkage is also used diagnostically to predict transmission of a disease gene. A prerequisite for linkage analyses are various types of sequence polymorphisms, like RFLPs, variable number of tandem repeats (VNTR), or microsatellites, which are used to distinguish the parental origin of alleles. Gene loci in the parental chromosomes may undergo recombination or remain nonrecombinant. The observed frequency of recombination between two loci is a function of their distance and is expressed in centiMorgans (cM). If the recombination frequency between two loci is 1%, the two loci are said to be 1 cM apart (1 cM corresponds to about 1 Mb of DNA). Because recombination frequency increases as a function of genetic distance, the closer together two loci are, the higher the likelihood that they will be inherited together (genetic linkage). A set of closely linked markers that are inherited together define a haplotype. DNA loci are identified by a specific nomenclature. For example, the locus D7S525 refers to human chromosome 7, segment 525, and is found on the short arm of chromosome 7. In order to identify a chromosomal locus that segregates with a disease, it is necessary to determine the genotype of DNA samples from one or several pedigrees. Subsequently, one can determine whether certain marker alleles cosegregate with the disease. When a series of genetic markers are used to establish linkage to a disease phenotype in large informative kindreds, markers that are closest to the disease gene are less likely to undergo recombination events and will attain a higher linkage score. The data resulting from the

Figure 4-6 Linkage analysis in multiple endocrine neoplasia type 1 (MEN-1). (A) Schematic representation of human chromosome 11 depicting the position of the MEN 1 gene at band q13. In any individual, there is a paternal and a maternal allele for MEN 1. Hypothetical microsatellite markers A and B are shown near the MEN 1 locus. The A marker is closer to the MEN-1 gene. Based on the number of repeats in the microsatellites, genotypes are defined such that the paternal allele is “3-4” and the maternal allele is “2-2.” Based on the pedigree (or DNA sequencing if available), the “3-4” genotype is shown to be linked to the MEN 1 gene. (B) Pedigree of a family with MEN 1. Alleles of family members are denoted by the genotypes based on microsatellites at A and B. Within the affected family, the disease is carried on the “3-4” allele (bold). Those affected or carrying the disease are shaded. Note that the male in generation II, who is not part of the original family, also possesses the “3-4” allele, but is not affected, indicating that a specific genotype is linked only within a family, not in the general population.

characterization of multiple loci are then analyzed by computer programs. Linkage is usually expressed as a Lod (logarithm of odds) score, which is a ratio of the probability that the disease and marker loci are linked rather than unlinked. Lod scores are expressed as the logarithm to the base 10 such that positive numbers favor linkage and negative scores support nonlinkage. Lod scores of +3 (1000:1) are generally accepted as supporting linkage, whereas a score of –2 is consistent with the absence of linkage. An example of the use of linkage analysis is shown in Fig. 4-6. In this case, the gene for the autosomal dominant disorder, multiple endocrine neoplasia type 1 (MEN-1), is known to be located on chromosome 11q13. Hypothetical polymorphic microsatellite markers are close to the MEN-1 gene, which encodes a protein, menin (currently of unknown function). In the pedigree shown, the affected grandfather in generation I carries alleles 3 and 4 on the

48

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

chromosome with the mutated MEN-1 gene and alleles 2 and 2 on his other chromosome 11. Consistent with the linkage of the 3/4 genotype to the MEN-1 locus, his son in generation 2 is affected, whereas his daughter (who inherits the 2/2 genotype from her father) is unaffected. In the third generation, transmission of the 3/4 genotype indicates risk of developing MEN-1, assuming no genetic recombination between the 3/4 alleles and the MEN-1 gene. If a specific mutation in the MEN-1 gene is identified within a family, it is possible to track transmission of the mutation itself, thereby eliminating uncertainty caused by recombination. CLONING OF GENES DNA may be amplified by cloning DNA fragments into suitable vectors that can be propagated in host cells. Vectors include DNA molecules derived from viruses, bacteria, or yeast that replicate independent of the genome of the host cell. After digestion of both the vector and the DNA to be inserted with restriction enzymes, they can be joined using DNA ligase. The vector carrying the fragment is then transferred into host cells (bacteria, yeast, or eukaryotic cells) where it is replicated. These manipulations allow the generation of large amounts of DNA for further experimental studies (see Chapter 2). This type of manipulation can be performed with a single DNA fragment to be amplified, or one can create libraries in which all fragments derived from genomic DNA are represented in the final collection of colonies. In principle, a genomic library represents multiple cloned fragments of the entire genome. Genomic libraries can also be generated from single chromosomes after selection with fluorescence-activated cell sorters (FACS). cDNA libraries can also be created, using the mRNA of a selected tissue, and several modifications allow enrichment for certain mRNAs (e.g., poly[A] or size-selected), thus increasing the chance to find clones of interest. After the preparation of libraries, the next step consists of isolating the recombinant clone containing the genomic or coding sequence of a gene. If there is some information on the protein encoded by this gene, microsequencing of a partial amino acid sequence allows synthesis of oligonucleotides based on the genetic code. Because of the degeneracy in the genetic code, this strategy employs a group of oligonucleotides. The oligonucleotides are radiolabeled and hybridized to clones within the library. Clones with a positive signal are isolated and propagated. Sequencing the DNA fragment inserted in the vector will provide information on at least a part of the gene of interest. It is also possible to clone genes without having prior information on the protein. This strategy is referred to as positional cloning or reverse genetics (Fig. 4-7). This approach was introduced in the late 1980s has been successful for the identification of many genes, including those causing cystic fibrosis, Duchenne muscular dystrophy, polycystic kidney disease, and the MEN syndromes. The first step consists of establishing genetic linkage between a disease phenotype and DNA markers. This allows determination of the chromosomal region where the candidate gene is located. The identification of markers residing close to a gene provides the starting point for chromosome walking or jumping, which can be used to move progressively closer to the gene of interest. These methods allow one to screen sequences between the markers for the presence of functional genes. If one or several candidate genes are identified, they can be cloned and sequenced. The final step involves the demonstration that the isolated gene harbors a mutation that segregates with the disease.

Figure 4-7 Positional cloning. A schematic outline is presented for the typical steps involved in the identification of a disease gene by positional cloning.

MUTATIONS In a broad sense, a mutation can be defined as any change in the primary nucleotide sequence of DNA, regardless of its functional consequences. A mutation in germ cells will lead to its presence in every cell of an organism, and it will be transmitted to the offspring. Some mutations may be lethal, while others are less deleterious, and some confer an evolutionary advantage. Somatic mutations, limited to a clone of cells in a given tissue, play an important role in the development of neoplasms. With the exception of triplet repeats which can expand (see below), mutations usually remain stable. Structurally, mutations are very diverse. They can involve the entire genome as in triploidy, as well as gross numerical or structural alterations in chromosomes (Table 4-1). Large deletions may affect a portion of a gene, an entire gene, or if several genes are involved, they may lead to a contiguous gene syndrome. Unequal crossing-over between homologous genes can result in fusion gene mutations, well-illustrated by color blindness. Mutations can also involve changes in only a few or a single basepair. Mutations involving single nucleotides are referred to as point mutations. Analogous to gross chromosomal changes, point mutations may consist of substitutions, deletions, or insertions. Substitutions may be silent or change the respective codon. If the change occurs in a coding region, it is called a missense mutation. Substitutions are called transitions if a purine is replaced by another purine base (A↔G) or a pyrimidine is replaced by another pyrimidine (C↔T). Changes from a purine to a pyrimidine or vice versa are referred to as transversions. Deletions or insertions cause a shift of the translational reading frame, typically resulting in a premature stop (nonsense mutation). Mutations in intronic sequences may eliminate or create splice donor or splice acceptor sites, resulting in an abnormally spliced mRNA from the mutated gene. Mutations may

CHAPTER 4 / TRANSMISSION OF GENETIC DISEASE

49

Table 4-1 Examples of Different Types of Mutations Type of mutations Genome Abnormal chromosome set Chromosome Abnormal number of autosomal chromosomes Abnormal number of sex chromosomes Translocation Deletion Duplication Inversion Gene Deletion Duplication, insertion Fusion Inversion Triplet expansion Missense point mutation Nonsense point mutation Frameshift Splicing mutation

Example Triploidy, tetraploidy Trisomy 21, 18, 13 Klinefelter syndrome, Turner syndrome Acute myeloid leukemia t(9;22)(q34:q11) “Philadelphia chromosome” Cri du chat syndrome 5pDosage-sensitive sex reversal (Xp dup) Ret oncogene inv (10)(q11.2q21) in papillary thyroid carcinoma (PTC) Duchenne muscular dystrophy, Thalassemia Charcot-Marie-Tooth Type I Glucocorticoid-remediable aldosteronism Hemophilia A Fragile X syndrome, Huntington disease Cystic fibrosis Cystic fibrosis Cystic fibrosis β Globin

also be found in the regulatory sequences of genes and can cause reduced or absent transcription of genes. In some cancer syndromes, there is an inherited predisposition to tumor formation (see Chapter 7). In retinoblastoma, the tumor develops when both copies of the retinoblastoma gene become defective through two somatic events (sporadic retinoblastoma) or through a somatic loss of the normal allele in an individual with an hereditary defect in the other allele (hereditary retinoblastoma) (see Chapter 105). This “two-hit model,” initially proposed by Knudson, applies to several other inherited cancer syndromes. The defective allele is inherited in a Mendelian fashion and follows a dominant pattern, but tumorigenesis results from a recessive loss of the tumor suppressor gene in an affected tissue. In other instances, the development of cancer typically requires defects in multiple genes, a process termed multistep carcinogenesis (see Chapter 7). Testing for DNA mutations has several advantages in comparison to analyses at the protein level. DNA is easy to isolate and, in contrast to proteins, it is not differentially regulated in various tissues. DNA can be isolated from many sources, including white blood cells, tissue samples, exfoliated cells, and hair roots. A variety of methods have been developed for mutational analyses (Table 4-2). Large mutations, deletions, insertions, rearrangements, or expansions of triplet repeats can be detected by Southern blotting or PCR. In Southern analysis, high molecular DNA is digested with restriction enzymes. This results in multiple DNA fragments that can be separated by gel electrophoresis. After transfer of DNA to a membrane, the DNA can be denatured and hybridized with radioactive probes that detect a particular sequence among the countless other fragments. Differences in the hybridization patterns obtained by Southern blotting can indicate deletions or insertions in the genomic DNA (Fig. 4-8). For example, Southern blotting can be used in the diagnosis of the α-thalassemia variant found in Southeast Asia (see Chapter 20). In some cases of this autosomal recessive disorder, there is a large deletion of both α-globin genes from the 30 kb gene cluster on chromosome 1. Because the deletion

encompasses the ζ-globin gene, it leads to hydrops fetalis and intrauterine death. RT-PCR can be useful for detecting absent or reduced levels of expression of an mRNA resulting from a mutated allele. Screening for point mutations can be performed by numerous methods (Table 4-2). They are based on the recognition of mismatches between nucleic acid duplexes, electrophoretic separation of single or double-stranded DNA, or sequencing of DNA fragments amplified by PCR. DNA sequencing can be performed directly on PCR products or by using fragments cloned into plasmid vectors. The sensitivity of these methods varies between 80 and 100%. Protein truncation tests (PTT) can be used to detect mutations resulting in premature termination of a polypeptide during its synthesis. The isolated cDNA is transcribed and translated in vitro and the proteins are analyzed by gel electrophoresis. Comparison with the wild-type protein allows detection of truncated mutants. This approach is particularly helpful for the detection of loss of function mutations in large genes.

GENETIC DISEASES CATEGORIES OF GENETIC DISEASE Taken together, genetic diseases form a substantial group of human disease. More than one-third of all pediatric hospital admissions are because of disorders that are caused, at least in part, by genetic factors. About 6–8% consist of single-gene defects, 0.4–2.5% are from chromosomal disorders, and the remaining group are genetically influenced. Overall, 3–5% of diseases in the general population are estimated to be genetically determined. Accurate estimates are difficult because different genetic defects can lead to the same phenotype (genetic heterogeneity), and a large number of genetic diseases are relatively rare. These difficulties should be kept in mind when assessing genetic risk. If one includes disorders with a strong polygenic predisposition, such as Type 2 diabetes mellitus, hypertension, or hyperlipidemias, then the role of genetics in common disorders is even greater. Every classification of genetic diseases harbors some degree of oversimplification because of overlaps between the different cat-

50

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Table 4-2 Methods Used for the Detection of Mutations Method Cytogenetic analysis FISH (fluorescent in situ hybridization) Southern blot PCR (polymerase chain reaction) Sequencing SSCP (single-strand conformational polymorphism) DGGE (denaturing gradient gel electrophoresis) OSH (oligonucleotide specific hybridization) RNase cleavage RFLP (restriction fragment length polymorphism) RT-PCR (reverse transcriptase-PCR) PTT (protein truncation tests)

Principle

Type of mutation detected

Numerical and structural analysis of chromosomes Hybridization to chromosomes with fluorescently labeled probes Hybridization with DNA probe after digestion of high molecular DNA Amplification of DNA segment Direct sequencing of PCR products Sequencing of DNA segments cloned into plasmids PCR of DNA segment: mutations result in conformational change and altered mobility PCR, altered mobility of amplified segment harboring a mutation Hybridization of oligonucleotides ± altered sequence to amplified genomic DNA Cleavage of mismatch between mutated and wild-type sequence Detection of altered restriction pattern of genomic DNA (Southern blot) or PCR products Reverse transcription, amplification of DNA segment → absence or reduction of mRNA Transcription/translation of cDNA isolated from tissue sample

egories. Genetic diseases can be classified generally into the following major categories: 1. Chromosomal disorders with numerical or structural abnormalities in one or several chromosomes. 2. Mendelian or monogenic disorders, characterized by a single mutant gene inherited according to Mendelian rules. 3. Multifactorial diseases or complex disease traits, defined by the interaction of multiple genes and one or multiple environmental factors. 4. Nonclassical forms of genetic disease, a heterogeneous group that includes disorders influenced by genomic imprinting, or caused by uniparental disomy, trinucleotide repeats, and various forms of mosaicism. 5. Mitochondrial disorders resulting from mutations in the mitochondrial genome. Because mitochondrial DNA is transmitted through the maternal line, the pattern of inheritance differs from Mendelian disorders. 6. Mutations arising in differentiated somatic cells, which are of particular importance in the development of neoplasms. Although they are not inherited, they may occur in individuals with an inherited predisposition. CHROMOSOMAL DISORDERS Chromosomal or cytogenetic disorders are defined by numerical or structural aberrations in chromosomes. They are frequent causes of abortions, developmental disorders, malformations, and malignancies. In humans, numerical chromosome aberrations occur in approximately 1 of

Numerical or structural abnormalities in chromosomes Large deletions, insertions, rearrangements, expansions of triplet repeats, amplifications Expansion of triplet repeats, variable number of tandem repeats (VNTR), gene rearrangements, translocations Point mutations, small deletions, and insertions Point mutations, small deletions, and insertions Point mutations, small deletions, and insertions Point mutations, small deletions, and insertions Point mutations, small deletions, and insertions Point mutations, small deletions, and insertions Indirect evidence for presence of mutation, cDNA cloning for sequencing Mutations leading to premature truncations

400 neonates. Triploidy and tetraploidy refer to circumstances in which each chromosome is present threefold or fourfold, respectively, instead of the usual diploid set. Triploidy is one of the most frequent chromosomal aberrations in humans and is found in about 15% of spontaneous abortions and severe malformations. The additional set of chromosomes may be of paternal or maternal origin. Aneuploidy describes a deviation from the normal chromosome number in a single pair of chromosomes. In a trisomy, three copies of a chromosome are present, and in a monosomy, one of the chromosomes is missing. Trisomies and monosomies are caused by nondisjunction in the first or second meiotic division and may be of paternal or maternal origin. Some of the most frequent abnormalities of chromosome number are listed in Table 4-3. Structural defects in chromosomes arise from translocations, deletions, inversions, insertions, isochromosomes, dicentric chromosomes, and ring chromosomes. If no chromosomal material has been lost or gained, the rearrangement is referred to as balanced. If it is unbalanced, there is either a loss or gain of chromosomal DNA. A large number of congenital syndromes caused by structural changes in chromosomes can be recognized based on their characteristic clinical phenotype. Furthermore, chromosomal translocations play an important role in certain malignancies, such as leukemias and lymphomas (see Chapters 26 and 27). MENDELIAN OR MONOGENIC DISORDERS Because most traits examined by Mendel were caused by single genes, monogenic human diseases are frequently referred to as Mendelian disorders. Information about many of these genetic disorders

CHAPTER 4 / TRANSMISSION OF GENETIC DISEASE

Figure 4-8 Southern blot. A Southern blot can be used to detect alterations in gene structure (e.g., deletions, insertions, variable number tandem repeat [VNTR], or restriction fragment length polymorphism [RFLP]). Genomic DNA is digested with one or several restriction enzymes. The digestion products are separated on an agarose gel and transferred to a membrane. Hybridization of the immobilized DNA with a radiolabeled probe allows detection of specific fragments by autoradiography. As illustrated, an RFLP can lead to differences in the length of the detected fragments. Such polymorphisms are used in linkage studies to assess whether a disease cosegregates with a genetic marker.

can be found in a large, continuously evolving compendium, Mendelian Inheritance in Man (MIM), initiated by V. A. McKusick. It is also accessible on-line (OMIM: On-line Mendelian Inheritance in Man, http://www3.ncbi.nlm.nih.gov/Omim). The Mendelian laws predict the transmission of alleles within a family, and these are depicted graphically in family trees or pedigrees. Standard symbols used for describing pedigrees are shown in Fig. 4-9. As an example, the segregation of genotypes in the offspring of parents with two distinct alleles, one dominant and one recessive, is illustrated in Fig. 4-10. Some relatively common Mendelian disorders are listed in Table 4-4, and the characteristic modes of inheritance are shown in Fig. 4-11. A dominant allele, A, and a recessive allele, a, can display three Mendelian modes of inheritance: autosomal dominant, autosomal recessive, and Xlinked. The mode of inheritance for a phenotypic trait or disease

51

is determined by pedigree analysis. In autosomal dominant disorders, individuals can be affected in each generation. The disease does not occur in the offspring of unaffected individuals. Males and females are affected with equal frequency (Fig. 4-11A). The child of an affected individual has a 50% risk of inheriting the mutated gene. In an autosomal recessive disease, both parents of an affected individual are obligate heterozygotes (Fig. 4-11B). The affected individual, who can be of either sex, can be a homozygote or compound heterozygote (distinct mutations in each copy of the gene) for a gene defect. Normally, heterozygous carriers of a defective gene are clinically normal. There is a 25% chance that an offspring will be affected. Many autosomal recessive diseases are rare and often occur in cases of parental consanguinity. X-chromosomal inheritance refers to the transmission of genes located on the X chromosome (Fig. 4-11C). A daughter always inherits her father’s X chromosome together with one of the two maternal X chromosomes. A son inherits the Y chromosome from his father and one of the maternal X chromosomes. Thus, there is no father-to-son transmission in X-linked inheritance, and all daughters of an affected male are obligate carriers of the mutant allele. Since males have only one X chromosome, they are hemizygous for a mutant allele and are therefore more likely to develop the mutant phenotype, regardless of whether the mutation is dominant or recessive. A female may be either heterozygous or homozygous for a mutant allele, which may be dominant or recessive. In addition, the expression of X-chromosomal genes is influenced by X chromosome inactivation (see below). The frequency of Mendelian disorders varies substantially in different populations. For example, the autosomal recessive disorder sickle-cell anemia occurs with the highest frequency in populations originating from West Africa, β thalassemia is frequently found in Asians, and cystic fibrosis occurs most often in individuals of European descent. In many of the more prevalent disorders, a selective advantage has been proposed for heterozygous gene carriers and may provide an explanation for the high frequency of the mutated allele. In disorders with a low prevalence, but a relatively high frequency within a defined population, a founder effect is more likely. PHENOTYPIC VARIABILITY IN MENDELIAN DISORDERS The phenotype associated with certain mutations can vary in the degree of penetrance. For example, in hypertrophic obstructive cardiomyopathy (HOCM), an autosomal dominant disorder characterized by obstruction of left ventricular outflow resulting from ventricular hypertrophy, some mutations in the myosin heavychain β gene exhibit 100% penetrance in adults. In other words, all carriers of the mutation will develop the disease (see Chapter 13). Other mutations in the same gene display a penetrance of about 50%, such that an individual carrying the mutant genotype may not manifest the disease phenotype. In this situation, the mutant gene is called nonpenetrant. Although a carrier of the mutant gene may not develop the disease, it will still be transmitted to the next generation where it may be penetrant or nonpenetrant. In addition, the HOCM phenotype may be caused by molecular alterations in other genes, a phenomenon referred to as genetic heterogeneity (see above). Determination of penetrance may be dependent on the sensitivity of the methodology used for the assessment of the phenotype. Penetrance can also be dependent on the presence or absence of environmental factors: If there is no exposure to disease-causing agents, some genetic defects may not become apparent. This is illustrated by hyperlipidemias, hemochromatosis, or porphyrias, all disorders that are significantly modulated by diet.

52

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Table 4-3 Frequent Chromosome Abnormalities Disorder Abnormal no. of chromosome sets Triploidy Tetraploidy Abnormal no. of autosomes Trisomy 21 Trisomy 18 Trisomy 13 Abnormal no. of sex chromosomes Klinefelter syndrome XYY-syndrome Turner syndrome Triple-X syndrome

Chromosomal genotype

Frequency

69XX, 69XY 92XX, 92XY

Frequent in miscarriage Frequent in miscarriage 1/600 1/5000 1/15,000

47XXY 47XYY 45X0, 45X/46X0, 45X/46XY (mosaicism) XXX

1/1000 males 1/1000 1/10,000 females 1/1000

Figure 4-10 Segregation of alleles. Segregation of genotypes in the offspring of parents with a dominant (A) and a recessive (a) allele. The distribution or segregation of the parental genotypes to their children depends on the combination of the alleles in the parents. The Mendelian laws predict the combinations and frequencies in the offspring (see text).

Figure 4-9 Symbols used in pedigree analyses.

Variability in clinical expressivity characterizes the spectrum of phenotypic changes among different individuals with a given genotype. It includes differences in the type and severity of symptoms, as well as the age of onset of a disease. A classical example is the multiple endocrine neoplasia syndrome type 1 (MEN-1), an autosomal dominant syndrome characterized by tumors of the parathyroid glands, the anterior pituitary, and the endocrine pancreas. By age 50, almost 100% of the gene carriers develop parathyroid tumors (see Chapter 51). In contrast, there is a great variability in the development of other manifestations of MEN-1 among siblings carrying the mutated gene. AUTOSOMAL DOMINANT DISORDERS Diseases inherited in an autosomal dominant manner result from one mutant allele and a normal allele on the other chromosome. Unless it is a new germline mutation, each affected child has an affected parent. The probability that an offspring will be affected is 50% because the alleles segregate at meiosis. Children with a normal genotype

will have only unaffected offspring (Fig. 4-11A). Both males and females can be affected, since the defective gene is on one of the 22 autosomes. Autosomal dominant disorders can, however, manifest themselves in a sex-limited pattern as exemplified by familial malelimited precocious puberty, which is caused by activating mutations in the gene encoding the LH-receptor (Chapter 57). The clinical manifestations of autosomal dominant disorders may be variable because of differences in penetrance or expressivity (see above). In certain cases, a mutant gene may be nonpenetrant. These variations can lead to difficulties in the recognition that a disorder is inherited. Approximately 1 in 100,000 newborns will harbor a new mutation at a given locus, and in the case of a dominant allele, this may result in phenotypic changes. De novo germline mutations are thought to occur more frequently during later cell divisions in gametogenesis, which explains why siblings are rarely affected. In some cases, however, the mutation may occur early in gametogenesis, resulting in gonadal mosaicism, which can lead to multiple affected sibs, despite the fact that somatic cells of the parents do not harbor the mutation. Gonadal mosaicism is, for example, found in Duchenne muscular dystrophy and osteogenesis imperfecta.

CHAPTER 4 / TRANSMISSION OF GENETIC DISEASE

53

Table 4-4 Selected Monogenic Disorders Autosomal dominant disorders Familial hyperlipidemia Familial hypercholesterolemia Huntington disease Polycystic kidney disease Spherocytosis Marfan syndrome Neurofibromatosis Hereditary nonpolyposis colon cancer Polyposis of the colon Familial breast cancer Willebrand disease Hypertrophic obstructive cardiomyopathy (HOCM) Myotonic dystrophy Otosclerosis Autosomal recessive disorders Cystic fibrosis Sickle-cell anemia β-Thalassemia α1-antitrypsin deficiency Congenital adrenal hyperplasia Phenylketonuria Hemochromatosis Tay-Sachs disease X-linked disorders Color blindness Hemophilia A Hemophilia B Glucose-6-phosphate dehydrogenase deficiency Duchenne muscular dystrophy Becker muscular dystrophy Fragile X syndrome X-linked ichthyosis

New germline mutations occur more frequently in fathers of advanced age. The average age of fathers with new germline mutations causing Marfan syndrome is 37 years, whereas fathers who transmit the disease by inheritance have an average age of 30 years. The ability to reproduce may be affected by autosomal dominant disorders, and the frequency of the disease will then reflect the occurrence of new mutations during gametogenesis. A mutant dominant allele may result in an abnormal phenotype because of several pathogenic mechanisms. Haploinsufficiency refers to the fact that half of the normal gene product may not be sufficient to result in a normal phenotype. This applies, for example, to rate-limiting enzymes in heme synthesis causing porphyrias (see Chapter 42). An increase in dosage of a gene product may also result in disease. In Charcot-Marie-Tooth type 1a, a duplication of the PMP22 gene results in high levels of expression of peripheral myelin protein 22, and this dosage effect underlies the demyelinating neuropathy (see Chapter 101). A mutant protein may also interfere with the function of the normal protein and act as a dominant negative. This is illustrated by mutations in the thyroid hormone receptor β in the syndrome of resistance to thyroid hormones (see Chapter 50). AUTOSOMAL RECESSIVE DISORDERS Autosomal recessive disorders are clinically manifest only if the patient is a homozygote or compound heterozygote (distinct mutations in the gene from the two different chromosomes) for a defect in a single

Figure 4-11 Mendelian inheritance. Pairs of alleles can display three Mendelian modes of inheritance: autosomal dominant, autosomal recessive, and X-chromosomal. The mode of inheritance for a given phenotypic trait or disease is determined by pedigree analysis. (A) In autosomal dominant disorders, affected individuals are found in successive generations. The disease does not occur in the offspring of unaffected individuals. Both males and females can be affected. The offspring of an affected individual have a 50% risk of inheriting the mutated gene. New mutations in one allele may lead to disease. (B) In an autosomal recessive disease, both parents of an affected individual are usually heterozygotes for a defective allele. The affected individual, who can be of either sex, is a homozygote or compound heterozygote. Generally, heterozygous carriers of a defective allele are clinically normal. There is a 25% chance that an offspring will be affected (homozygous recessive). Many autosomal recessive diseases occur in cases of parental consanguinity. (C) Characteristic features of X-linked inheritance are that there is no father-to-son transmission, and all daughters of an affected male are obligate carriers of the mutant allele. Males are hemizygous for the mutant allele, and they are more likely to develop the mutant phenotype, regardless of whether the mutation is dominant or recessive. A female may be either heterozygous or homozygous for the mutant allele, which may be dominant or recessive. De novo mutations in males may result in disease.

gene. Heterozygous carriers of a defective allele are clinically normal. Although many recessive disorders appear to be truly recessive, it should be emphasized that heterozygotes may display subtle differences in phenotype that only become apparent with more precise testing or in the context of certain environmental influences. For example, in sickle-cell anemia, heterozygotes

54

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

are normally asymptomatic. In situations of dehydration or diminished oxygen pressure, they can occasionally experience sicklecell crises. In most cases of a recessive disorder, an affected individual is the offspring of two heterozygotes. In this situation, there is a 25% chance of a normal genotype, a 50% probability of a heterozygous state, and a 25% risk of disease. If a heterozygote mates with a homozygote, the probability of disease increases to 50% for each child and the pedigree analysis may mimic an autosomal dominant mode of inheritance. If a mutant recessive allele is rare in a population, affected individuals are often the offspring of consanguineous parents. In contrast to autosomal dominant disorders, new mutations for recessive alleles rarely manifest themselves, because they result in an asymptomatic carrier. Males and females are affected with the same frequency unless the mutation has a sexspecific effect (e.g., steroid 5α-reductase type II deficiency, which affects only males; see Chapter 59). The clinical expression of autosomal recessive disorders is usually more uniform than in autosomal dominant disorders. Most mutated alleles lead to a complete or partial loss of function. They frequently involve enzymes in metabolic pathways, receptors, or components of signaling cascades. Recessive disorders are generally rare because of the fact that homozygotes may have reduced biological fitness. The high frequency of certain autosomal recessive disorders such as sickle-cell anemia, cystic fibrosis, and thalassemia has been proposed to reflect a biological advantage for heterozygotes. X-LINKED DISORDERS The characteristic pattern of inheritance of X-linked disorders is shown in Fig. 4-11C. In X-linked inheritance, there is no father-to-son transmission, and all daughters of an affected male are obligate carriers of the mutant allele. The risk of developing a disease resulting from a mutant X-chromosomal gene is thus different in males and females. While males hemizygous for the mutant allele are always affected, the heterozygous females may be asymptomatic (X-linked recessive) or affected (X-linked dominant). New mutations can result in affected males or heterozygous females. As described below, X chromosome inactivation may influence the expression of an X-linked disease in a heterozygous female. Some X-linked disorders are listed in Table 4-4. In some cases (e.g., incontinentia pigmenti), the mutated allele is lethal in the hemizygous males, and affected women have an increased frequency of spontaneous abortions of male fetuses. Y-LINKED DISORDERS Only a few genes are known on the Y chromosome. One of these, the sex-region determining Y factor (SRY) or testis determining factor (TDF), is crucial for the development of the normal male phenotype. Since the SRY region is close to the pseudoautosomal region of the sex chromosomes, crossing-over occasionally occurs between the X and Y chromosomes. Translocations between the X and Y chromosome can subsequently result in XY females with the Y chromosome lacking the SRY gene or XX males harboring the SRY factor on one of the X chromosomes (see Chapter 58). Point mutations in the SRY gene may also result in individuals with an XY genotype and an incomplete female phenotype. Most of these mutations occur de novo. Men with oligospermia/azoospermia frequently have microdeletions on the long arm of the Y chromosome that involve the azoospermia factor (AZF). MULTIFACTORIAL DISEASES OR COMPLEX DISEASE TRAITS In many diseases that appear to have some genetic com-

ponent, there is no clear evidence for classical Mendelian inheritance. However, many common diseases such as asthma (see Chapters 35 and 36), rheumatoid arthritis (see Chapter 34), diabetes mellitus (see Chapter 47), schizophrenia (see Chapter 109), hypertension (see Chapter 16), and many others exhibit familial predisposition. In many of these cases, interactions between genetic factors and the environment may be involved in the pathogenesis of the disease. Several lines of evidence point towards polygenic (multiple genes) inheritance. A role for genetics is supported if a disorder occurs at a higher frequency in certain ethnic groups. Since environmental factors may have an important influence, it is important to assess whether the disease frequency is modulated if members of such a population move to a different environment. A genetic component in a multifactorial disease is also suggested if the disorder occurs at a higher frequency in the immediate relatives of an index case in comparison to the general population. Finally, the concordance in monozygotic and dizygotic twins may give important clues for the presence or absence of complex traits. If the concordance rate is high in monozygotic twins in comparison to dizygotic twins, genetic factors are likely to be important. X-CHROMOSOME INACTIVATION, GENOMIC IMPRINTING, AND UNIPARENTAL DISOMY Although females contain two X chromosomes, the expression of many X-chromosomal genes is not greater in females than in males. This is because of random inactivation of either the paternal or the maternal X chromosome in each somatic cell of a female during early embryogenesis, a phenomenon also referred to as lyonization. X-inactivation is caused by differential methylation of cytosine nucleotides. Methylation of the inactive copy of the X chromosome results in formation of the X chromatin, which is visible as a Barr body. Xinactivation is not reversible, and the same chromosome is inactivated in daughter cells. Methylation is not only involved in X-chromosome inactivation, but also plays an important role in the regulation of gene expression. Active genes often display diminished or absent methylation, whereas inactive genes are hypermethylated. Methylation occurs at the 5' position in the cytosine of cytosine-guanine (CpG) dinucleotides. Unmethylated or hypomethylated CpG (cytosine-phosphate-guanine) clusters often indicate the beginning of a structural gene that is being expressed, and they are sometimes used in positional cloning for the identification of genes. The mechanism by which methylation leads to inactivation of gene expression is not well-understood. It could alter chromatin structure or interfere with the binding of transcription factors to regulatory DNA sequences. Alternatively, methylation might be an epiphenomenon that occurs following the inactivation of a gene by other mechanisms. The complex control of gene expression at the transcriptional level is discussed in Chapter 3. According to Mendelian principles, the parental origin of a mutant gene is irrelevant for the expression of a phenotype. As described above in the context of X-chromosome inactivation, there are some important exceptions to this rule. Inactivation of genes or chromosomal regions on one of the two chromosomes also occurs on autosomes. This phenomenon, referred to as genomic imprinting, leads to preferential expression of an allele, depending on its parental origin, resulting in transmission of disease in a manner that is dependent on the sex of the transmitting parent. The fact that identical regions can differ in their functional activity depending on whether they are of maternal or paternal

CHAPTER 4 / TRANSMISSION OF GENETIC DISEASE

origin is also illustrated by zygotes with two maternal or two paternal sets of chromosomes. Cells with two copies of maternal chromosomes develop into the inner cell mass of the embryo, whereas cells with two sets of paternal chromosomes almost exclusively form the extraembryonal cells of the trophoectoderm. These observations indicate that some genes are differentially expressed from paternal and maternal chromosomes. The principle of genomic imprinting has important implications for a subset of human disorders. Classical examples are the Prader-Willi-Syndrome and the Angelman Syndrome (Chapter 117). The Prader-Willi-Syndrome, characterized by diminished fetal activity, obesity, hypotonia, mental retardation, short stature, and hypogonadotropic hypogonadism, is caused in most cases by deletions on the short arm of chromosome 15. The deletions in the Prader-Willi-Syndrome occur exclusively on the paternally derived chromosome. In contrast, patients with Angelman syndrome (which is characterized by mental retardation, seizures, ataxia, and hypotonia) have deletions at the same site of chromosome 15, but they occur only on the maternally derived chromosome (Fig. 4-12A). The two syndromes may also result from uniparental disomy (Fig. 4-12B). In this case, the syndromes are not caused by deletions on chromosome 15, but by inheritance of either two maternal chromosomes (resulting in Prader-Willi syndrome) or of two paternal chromosomes (causing Angelman syndrome). Genomic imprinting, or uniparental disomy, is involved in the pathogenesis of several other disorders and malignancies. Hydatiform mole contains a normal number of diploid chromosomes, but they are all of paternal origin. The opposite occurs in ovarian teratomata with 46 chromosomes of maternal origin. Overexpression of the imprinted gene for insulin-like growth factor II (IGF-II) is involved in the pathogenesis of the cancer-predisposing Beckwith-Wiedemann syndrome (BWS) (see Chapter 116). These children show somatic overgrowth with organomegaly and hemihypertrophy, and they have an increased risk of developing embryonal malignancies such as Wilm's tumor (see Chapter 71). Normally, only the paternally derived copy of the IGF-II gene is active. Imprinting of the IGF-II gene is regulated by H19, an RNA transcript that is not translated into protein. Disruption or lack of methylation of H19 leads to a relaxation of IGF-II imprinting which subsequently results in overgrowth. TRINUCLEOTIDE REPEATS Trinucleotide repeats are found in several genes and their number can vary among healthy individuals (see Chapter 100). For example, the number of CAG repeats in the first exon of the androgen receptor (AR) gene is lowest in African Americans, intermediate in Caucasians, and highest in Asians (Chapter 60). Of note, the frequency of prostatic cancer is inversely proportional to the length of the repeats. An increase in the number of repeats above a certain critical threshold is associated with several diseases (Table 4-5). The repeats can be located within the coding region as in Huntington disease (see Chapter 97) or the X-linked form of spinal and bulbar muscular atrophy (SBMA, Kennedy syndrome), which is caused by an expansion in the polyglutamine tract encoded by CAG repeats in the first exon of the androgen receptor. In other instances (e.g., in myotonic dystrophy and in the fragile X syndrome), the repeats are not located in coding regions of genes. If an expansion is present, the DNA fragment is unstable and generally tends to expand further with additional cell divisions. The length of the expansion and the age of onset, or the severity of a disease, often correlate. Since the repeat length tends to

55

Figure 4-12 Genomic imprinting and uniparental disomy. Model for Prader-Willi syndrome and Angelman syndrome. The two distinct disorders are caused by loss of function of two closely adjacent loci on chromosome 15, which undergoes genomic imprinting. (A) Normal situation. The maternal PWS locus and the paternal AS locus are inactivated. (B) PWS and AS caused by deletions of the paternal and the maternal alleles, respectively, of chromosome 15q11-13. Southern blots with the marker D15S11 illustrate the absence of a positive hybridization for the paternal (PWS) or maternal (AS) alleles. (C) Uniparental disomy, the inheritance of two maternal copies (PWS) or two paternal copies (AS) can lead to the same phenotypes. In this case, a Southern blot with the probe D15S11 shows an increased intensity for one of the maternal (PWS) or paternal alleles (AS).

increase from generation to generation, the manifestations of the disease are observed at an earlier age, a phenomenon called anticipation. In addition, the repeat number may vary in some disorders within an individual in a tissue-specific manner. In myotonic dystrophy, the CTG repeat can be tenfold higher in muscular tissue compared to lymphocytes (Chapter 94). The mechanism by which the expansion of such repeats occurs is still debated. Moreover, the molecular mechanisms leading to the respective diseases are poorly defined.

56

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Table 4-5 Selected Disorders Caused by Nucleotide Repeats Disease

Locus

Repeat

Triplet length

Gene product

MIM no.

Spinobulbar muscular atrophy (SBMA)

X

CAG

Androgen receptor

313200

Fragile X-syndrome FRAXA

Xq27.3

CGG

FMR-1 protein

309550

Fragile X-syndrome FRAXE

Xq28

GCC

FMR-2 protein

309548

Myotonic dystrophy

19q13.3

CTG

Myotonin protein kinase

160900

Huntington disease

4p16.3

CAG

Huntingtin

143100

Spinocerebellar atrophy type 1

6p21.3-21.2

CAG

Ataxin 1

164400

Dentatorubral pallidoluysian atrophy

12p

CAG

Atrophin

125370

Machado Joseph

14q24.3-32

CAG

MJDI

109150

Friedreich ataxia

9q13

GAA

Normal: 11-34 Disease: 40-62 Normal: 6–50 Disease: 200–300 Normal: 6–25 Disease: >200 Normal: 5–30 Disease: 200–1000 Normal: 11–34 Disease: 37–121 Normal: 19–36 Disease: 43–81 Normal: 7–23 Disease: 49–75 Normal: 13–36 Disease: 68–79 Normal: 7–22 Disease: 200–900

Frataxin

229300

MOSAICISM Mosaicism refers to the presence of two or more cell lines in an individual that differ in their genotype. Cell mosaicism can result from a mutation that occurs during embryonic, fetal, or extrauterine development. The developmental stage at which the defect arises will determine whether germ cells or only somatic cells are involved. Chromosomal mosaicism results from nondisjunction at an early embryonic mitotic division, leading to the persistence of more than one cell line. Somatic mosaicism is characterized by a patchy distribution of somatic cells containing a mutation. For example, the McCuneAlbright syndrome is caused by activating mutations in the stimulatory G protein, Gsα, that occur early in development. The clinical phenotype varies depending on the tissue distribution of the mutation. The manifestations can include ovarian cysts that secrete sex steroids and cause precocious puberty, polyostotic fibrous dysplasia, café au lait skin pigmentation, pituitary adenomas, and hypersecreting autonomous thyroid nodules. Germ cell mosaicism or gonadal mosaicism should be suspected when parents seem to be normal on genetic testing, but have several affected offspring with a dominant or X-linked disorder. In these instances, the mutation may have occurred early in gametogenesis. MITOCHONDRIAL INHERITANCE Several diseases arising from mutations in the mitochondrial genome are known in humans (see Chapter 103). Each mitochondrion has several copies of a circular chromosome. This DNA predominantly encodes proteins that are components of the respiratory chain. They are transmitted through the maternal line because sperm do not contribute cytoplasmic components to the zygote. The disease will therefore not be transmitted from an affected man to his children, but all the children from an affected mother will be affected. In contrast to the nuclear chromosomes, the mitochondrial chromosome is present in numerous copies in the cell, and variable numbers of mitochondria are present in cells of different tissues. Different offspring may inherit various ratios of mutant and wildtype mitochondrial genomes (heteroplasmy) that lead to phenotypic variability. During cell replication, the genotype may also shift in the direction of the wild type or mutant chromosomes. Examples of mitochondrial diseases are listed in Table 4-6.

SOMATIC MUTATIONS Mutations that occur during embryogenesis, or later in development, are referred to as somatic mutations. Because these mutations do not involve the germline, they are not transmitted to the offspring. Somatic mutations have their most important role in various forms of neoplasia (see Chapter 7). Undoubtedly, many mutations occur that are silent, because they fail to alter the expression or function of genes. Another large group of somatic mutations may be deleterious to cellular function, but they result in apoptosis or programmed cell death, and are therefore not discovered (see Chapter 6). Rarely, mutations will enhance cell proliferation or prolong cell survival, and this group is associated with the development of tumors. These mutations can be identified because they confer a selective survival advantage, resulting in clonal expansion of cell population harboring the mutation. Activating mutations have been described in many oncogenes. Some of these mutations activate growth factor receptors (e.g., ret), whereas others stimulate growth factor signaling pathways (e.g., ras). Another group of somatic mutations inactivate tumor suppressor genes such as retinoblastoma (Rb). Somatic “second hits” in tumor suppressor genes occur in several inherited cancer syndromes in which the “first hit” has already been transmitted in the germline. Examples of these syndromes include retinoblastoma (Rb), Li-Fraumeni (p53), and multiple endocrine neoplasia type I (MEN-I). Many models of tumorigenesis invoke the concept of multistep carcinogenesis. In this scenario, an individual may inherit certain mutations or a predisposition to errors in DNA repair. However, these features alone are not usually sufficient for neoplasia in the absence of additional, somatic mutations. Multistep carcinogenesis proposes an accumulation of mutations, some of which may involve the activation of oncogenes, others inactivate tumor suppressor genes, and another group may impair apoptosis (e.g., p53). The characterization of these types of somatic mutations is used increasingly to classify tumors for the purpose of prognosis and may eventually provide new strategies for treatment.

SCREENING FOR GENETIC DISORDERS The detection of genetic disorders by means of DNA testing is playing an increasing role in prenatal diagnosis, newborn, and

57

CHAPTER 4 / TRANSMISSION OF GENETIC DISEASE

Table 4-6 Selected Mitochondrial Diseases Disease/syndrome

MIM no.

MELAS syndrome: mitochondrial myopathy with encephalopathy, lactic acidosis, and stroke Leber optic atrophy: hereditary optical neuropathy Kearns-Sayre syndrome (KSS): ophthalmoplegia, retinal pigment degeneration, cardiomyopathy MERRF syndrome: myoclonic epilepsy and ragged red fibers Maternally inherited myopathy and cardiomyopathy (MMC) Neurogenic muscular weakness with ataxia and retinitis pigmentosa (NARP) Progressive external ophthalmoplegia (CEOP) Pearson syndrome (PEAR): bone marrow and pancreatic failure Autosomal dominant inherited mitochondrial myopathy with mitochondrial deletion

540000 535000 530000 545030 590050 551500 258470 557000 157640

population screening, as well as predictive testing (see Chapter 9). Although these tests are extremely helpful in some instances, they are associated with difficult ethical questions, some of which are briefly addressed at the end of this chapter. Chorionic villus sampling (CVS) and ultrasound-guided umbilical vein puncture allow analyses of fetal DNA during the first trimester, and amniocentesis is used to test for genetic diseases during the second trimester. Generally accepted indications to perform such analyses include the presence of a known familial genetic disease and advanced maternal age. Disorders for which prenatal diagnosis is routinely available include the thalassemias, hemophilias A and B, cystic fibrosis, and chromosomal abnormalities that can be identified by fluorescent in situ hybridization (FISH). Determination of the sex has major implications in X-linked disorders (e.g., Duchenne muscular dystrophy or hemophilia). Other applications of prenatal DNA testing include the diagnosis of congenital infections and Rhesus blood group incompatibility. Newborn screening for frequent diseases is well-established for disorders that can be altered by early treatment, such as congenital hypothyroidism, phenylketonuria, and galactosemia. In some countries, these newborn screening programs have been expanded, and the application of PCR to newborn blood spots opens the door for further testing, as exemplified by screening for cystic fibrosis, sicklecell disease, or congenital adrenal hyperplasia. Issues that have to be addressed before accepting new screening programs include costeffectiveness, adequate counseling, and therapeutic consequences. Population screening strategies are also becoming more practical because PCR allows testing of a large number of samples relatively rapidly and economically. Population screening can be performed using different strategies. In selective screening programs, DNA testing is performed by mutational analysis or linkage studies in individuals at risk for a genetic disorder known to be present in a family (e.g., Tay-Sachs disease). Mass screening programs require tests of high sensitivity and specificity in order to be cost-effective. In epidemiological surveys, this type of screening program allows determination of the prevalence of certain disorders. Prerequisites for their success are that the disorders be potentially serious, that they can be influenced at a presymptomatic stage by changes in behavior, diet and/or pharmaceutical manipulations, and that the screening does not result in any harm or discrimination (“primum nil nocere”). Screening for the autosomal recessive neurodegenerative storage disease, Tay-Sachs, in Jewish populations resulted in a reduction of the number of affected individuals. In contrast, the screening for sickle-cell trait and disease in African Americans led to unanticipated problems of discrimination by health insurers and employers. Mass screening

programs harbor further potential problems. Although screening for the most common genetic alteration in cystic fibrosis (∆F508 mutation), with a frequency of ~70% in Northern Europe, is feasible and seems to be effective, it is important to keep in mind that the disease can be caused by more than 600 other mutations. The search for these less common mutations would substantially increase the costs and workload, but result in little impact on the effectiveness of the screening program as a whole. Population screening is also performed as a part of occupational screening programs that aim to detect increased risk for certain activities (e.g., α1-antitrypsin deficiency and smoke or dust exposure). Detection and exclusion of individuals with increased risk should not replace or diminish efforts to increase the safety of the working environment.

ETHICAL CONSIDERATIONS Ethical issues surrounding genetics and genetic testing continue to be a topic of considerable debate. The sequencing of the human genome, the identification of its genes, and the association of genetic defects with disease raise many issues concerning the implications for the individual and mankind. The recent cloning of mammals underlines the relevance of these issues. Ethical discussions are influenced by education, cultural traditions, religion, attitudes toward human values, and also political structures and historical context. Progress and the advantages of genetic medicine have to be balanced against their potential risks. Although the magnitude of benefits is difficult to predict, it is essential that genetic testing does not result in harm for individuals or groups within a society. Modern genetics has demonstrated that the concept of eugenics as a strategy for the elimination of human disease cannot be effective. The fact that this approach has been misused in the past must not be forgotten. Testing for disease predisposition may lead to a profound change in the way we screen and treat disease, and it opens the possibility for early recognition. However, it also entails a risk of discrimination and loss of privacy, health benefits, and employment. In many countries, lawmakers are therefore establishing legislation that will prevent health plans from discriminating against people on the basis of their genetic inheritance, or using genetic screening to deny coverage or to establish premiums. Prenatal diagnosis of a disorder may lead to ethical dilemmas for parents (see Chapter 9). The ethical implications of terminating a pregnancy continue to be controversial in many societies. In all these situations, adequate counseling is therefore paramount. Other difficult situations may arise if DNA testing results in detection of nonpaternity, which occurs in 3–5% of randomly stud-

58

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

ied children in many cultures. Particular problems may arise when healthy children are tested. Testing may be acceptable in the context of a disorder for which early intervention will improve prognosis, but it is more problematic when this is not the case. This leads to the question of when or whether tests for these types of inherited disorders should be performed. For example, in Huntington disease, which cannot currently be influenced by any form of treatment, individuals are not tested before they are able to give informed consent (see Chapter 97). The complexity of these issues will probably increase further as multifactorial disorders are characterized more thoroughly at the genetic level. The hope that molecular medicine will contribute significantly to diminish the burden of disease in the long term seems, however, realistic. For molecular medicine to be successful, it is essential that the ethical principles of social justice, equality of treatment, confidentiality of test results, and the absence of discrimination based on genetic screening are fully respected.

SELECTED REFERENCES Alberts B, Bray D, Lewis J, Ruff M, Roberts K, Watson JD. Molecular Biology of the Cell, 3rd ed. New York: Garland, 1994. Beaudet AL. Genetics and disease. In: Fauci AS, Braunwald E, Isselbacher KJ, et al., eds. Harrison’s Principles of Internal Medicine, 14th ed. New York: McGraw-Hill, 1997; pp. 365–395.

Brock DJH. Molecular Genetics for the Clinician, 1st ed. Cambridge, U.K.: Cambridge University Press, 1993. Collins FS. Positional cloning moves from perditional to traditional. Nat Genet 1995;9:347–350. Dracopoli NC, Haines JL, Korf BR, et al., eds. Current Protocols in Human Genetics. New York: Wiley, 1994. Gelehrter TD, Collins FS. Principles of Medical Genetics. Baltimore, MD: Williams & Wilkins, 1990. Lewin B. Genes V. Oxford, UK: Oxford University Press, 1994. McKusick VA. Mendelian Inheritance in Man: A Catalog of Human Genes and Genetic Disorders, 11th ed. Baltimore, MD: Johns Hopkins University Press, 1994. Ott J. Human Genetic Linkage, 2nd ed. Baltimore, MD: Johns Hopkins University Press, 1992. Passarge E. Color Atlas of Genetics. New York: Thieme, 1995. Pembrey ME. Genetic factors in disease. In: Weatherall DJ, Ledingham JGG, Warrell DA, eds. Oxford Textbook of Medicine, Oxford, UK: Oxford University Press, 1996; pp. 100–138. Reilly PR, Boshar MF, Holtzman SH. Ethical issues in genetic research: disclosure and informed consent. Nat Genet 1997;15:16–20. Rimoin DL, Conner JM, Pyeritz RE, Emery AEH. Emery and Rimoin’s Principles and Practice of Medical Genetics, 3rd ed. New York: Churchill Livingstone, 1996. Scriver CR, Beaudet CR, Sly WS, Valle D, eds. The Metabolic and Molecular Bases of Inherited Disease, 7th ed. New York: McGraw-Hill, 1995. Thompson MW, McInnes RR, Willard HF, eds. Genetics in Medicine, 5th ed. Philadelphia, PA: WB Saunders, 1991. Weatherall DJ. The New Genetics in Clinical Practice, 3rd ed. Oxford, U.K.: Oxford University Press, 1992.

CHAPTER 5 / THE HUMAN GENOME PROJECT

5

59

The Human Genome Project J. LARRY JAMESON

INTRODUCTION

approximately 50,000–80,000 genes. It is estimated that only about 5–10% of the human genome encodes protein-coding regions of genes. The remainder of the genome comprises regulatory regions, introns, and repetitive sequences. Thus, only a minority of the genome is expressed in the form of mRNA, which is subsequently translated into protein. It is for this reason that analyses of expressed sequence tags (ESTs), short fragments of cDNA sequence from a particular tissue, provide a powerful approach for identifying gene sequences that encode proteins. The total length of DNA is about 3 billion bp, nearly 1000-fold greater than the Escherichia coli genome. At the outset of the HGP, most molecular biologists might be able to sequence and analyze a 30,000-kb locus over the course of a year. Thus, in the absence of technological advances, 100 years would be required for 1000 such individuals to complete DNA sequencing. For this reason, initial emphasis was placed on mapping and on the development of new sequencing and computer database technologies. An analogy that has been used to illustrate the complexity of the human genome is to compare it with the amount of information in an encyclopedia that is written in the four-letter code of DNA (A, G, T, and C). Thus, the chromosomes might be analogous to volumes of the encyclopedia, with genes corresponding to paragraphs. This analogy emphasizes the challenge of identifying single nucleotide changes that might cause a disorder such as cystic fibrosis or sickle-cell anemia. In this case, the mutation would consist of a single letter change on one of the pages of the encyclopedia. In addition to identifying disease-causing genes, a critical first step is to create a table of contents and an index for the book, which allows one to find the relevant genes. This is the goal of genetic and physical mapping stages of the genome project.

The Human Genome Project (HGP) was conceived in the mid1980s as an ambitious effort to characterize the human genome, ultimately culminating with a complete DNA sequence by the year 2005. The accomplishment of this goal would locate the ~80,000 genes and provide the DNA sequence (~3 × 109 bp) for the entire genome at an estimated cost of $3 billion over 15 years. The project has evolved as an international effort, driven forward by numerous scientific groups and funding agencies. Exchange of information and biological materials has been facilitated by the Human Genome Organization (HUGO). In the United States, the genome project was officially launched in 1990 by the National Institutes of Health (NIH) and the Department of Energy (DOE). The goals, as initially conceived were (1) creation of genetic maps, (2) development of physical maps, and (3) determination of complete sequence of human DNA.

INITIATION OF THE GENOME PROJECT AS “BIG SCIENCE” From its inception, the goals of the HGP seemed daunting, and some considered it unrealistic, if not foolhardy. In addition to issues of cost, there was much debate concerning the goals of the genome project, its organization and timetable, and whether it would raise insurmountable ethical issues. There was particular concern that resources committed to the genome project would detract from traditional investigator-initiated research. On the other hand, proponents emphasized that analyses of the genome were inevitable and that, in the absence of a systematic approach, genetic studies would be inefficient and even more costly. In terms of its potential impact, analogies were drawn to the space program and to attempts to unravel the periodic table of elements. Most anticipated that the genome project would spin off new technologies in parallel with completion of its scientific goals. Moreover, it was expected that progress in the genome project would provide even more opportunities for hypothesis-driven research and for applications to clinical medicine. As described below, the genome project appears to be living up to the expectations of its proponents.

GENETIC MAPS The complexity of the human genome emphasizes the importance of first establishing a genetic and a physical map. In this manner, the locations of genes can be described. In addition, it allows genetic linkage to be determined. That is, genes that are physically close to one another on a chromosome will be transmitted together, except when a recombination event has occurred during meiosis to exchange homologous regions of DNA. The genetic and physical maps differ in the following manner. A genetic map is measured in centiMorgans (cM) and is based on distances between genes that are estimated by recombination frequency. The physical map is the actual distance (in bp of DNA sequence) between genes.

SCOPE OF THE PROJECT A few facts help to appreciate the scope of the human genome project. The 23 pairs of human chromosomes are thought to encode From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

59

60

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Figure 5-1 Detection of allelic variants using microsatellite markers. An example of a microsatellite consisting of a triplet nucleotide repeat (CAG) is shown. PCR primers on either side of the repeat results in PCR products of varying length, depending on the number of triplet repeats. In this example, the father has alleles of 110 and 113 and the mother has alleles of 107 and 116. The children each inherited the 116 allele from their mother and distinct alleles from their father. In this manner, microsatellites allow determination of chromosomal transmission within a family.

The unit of genetic distance, centiMorgans, is named after Thomas H. Morgan, a geneticist who provided early evidence for genetic linkage and established the concept that recombination frequency varies as a function of the distance between two genetic loci. By definition, 1 cM corresponds to a recombination frequency of 1%. For example, consider two polymorphic markers (A, B) on chromosome 11 that are linked to the gene causing multiple endocrine neoplasia type 1 (MEN-1). If sufficient pedigrees are examined, it is possible to determine which marker is closer to the disease locus by determining the frequency with which each marker undergoes recombination relative to the disease phenotype. If marker A undergoes recombination at a rate of 0.2, whereas marker B exhibits a lower recombination frequency (0.1), it can be concluded that marker B lies between marker A and the MEN-1 gene as long as there is data that neither marker is on the other side of MEN-1. There are several reasons why there might be discrepancies between a genetic and a physical map. For example, some genomic regions may exhibit relatively high recombination frequencies. This would result in a greater genetic distance relative to a physical map. Recombination appears to be about twice as common during meiosis in females. Thus, most of genetic distances are “sex-averaged” distances. Rapid progress in the creation of a genetic map reflects synergistic interactions among investigators from many countries. In addition, there have been important technical advances, particularly involving the use of highly polymorphic microsatellite markers. Microsatellites consist of di-, tri-, or tetranucleotide repeats. Within any given microsatellite marker, the number of repeats is highly variable in the population (Fig. 5-1). This makes it probable (~70% chance) that an individual will be heterozygous for the marker. In addition, their spouse is also likely to be heterozygous, or may even have a distinct set of repeats, greatly enhancing the power of linkage studies (Fig. 5-2). In the example shown, the array of microsatellite markers makes it possible to determine genetic transmission within a family with great certainty. A large array of these markers have now been identified and have been ordered throughout the genome.

Figure 5-2 Genotype analysis using microsatellite markers. The pedigree illustrates autosomal recessive transmission. An array of microsatellite markers are shown along the long arm of one of the chromosomes. Note that some markers are dinucleotide repeats whereas others are trinucleotide repeats. The parents are heterozygous for most of the markers, which have been selected based on their high degree of heterozygosity within the population. Thus, when several markers are combined, it is relatively easy to document chromosomal transmission within the pedigree. One can assign letters (A–D) that correspond to the various chromosomes, reflecting the transmission of linked alleles. In this case, a putative disease gene is carried by each parent. The mutation resides on chromosome B in the father and chromosome D in the mother. Thus, a son with the BD genotype is affected, a daughter with genotype AC is unaffected, and the children with genotypes BC and AD are each carriers even though they inherited the disease gene from different parents. An example of recombination is illustrated in the affected son. In this case, the chromosome labeled B* received the 128 (black shading) microsatellite marker, which must be derived from the father’s chromosome A. Because this son is affected, this crossover event indicates that the disease-causing gene resides centromeric to this microsatellite.

The Genethon human linkage map was updated in 1996. At that point, the map consisted of more than 5000 microsatellite repeat polymorphisms with a mean heterozygosity of 70%. Genotyping with the microsatellite markers was performed using the CEPH (Centre d’Etudes du Polymorphisme Humaine) families. The estimated length of the map is 3699 cM, with an average interval size of about 1.6 cM. In combination with other genetic maps and markers, most regions of the genome can now be analyzed at the cM level. In the future, it may be possible to develop an array of single nucleotide polymorphisms that could be used in conjunction with DNA chip technology. In addition to increasing the density of markers, chip technology would be amenable to begin high throughput screening of genetic loci.

CONSTRUCTION OF PHYSICAL MAPS AND CLONING OF THE ENTIRE GENOME In physical terms, 1 cM is roughly 1 million bp (1 Mb). The physical map reflects the arrangement and distances between genes and exists at several levels of resolution. A low-resolution map would reveal which chromosome carries a particular gene. The use of such techniques as fluorescence in situ hybridization (FISH) allows determination of the location of a particular gene on a chromosome. By using different fluorescent tags, it is possible to “paint” chromosomes and demonstrate the relative locations of genes, but still at a relatively low resolution. A higher level of

CHAPTER 5 / THE HUMAN GENOME PROJECT

61

Table 5-1 Size of DNA Inserts in Various Vectors Vector Yeast artificial chromosome (YAC) Bacterial artificial chromosome (BAC) Cosmid λ Phage Plasmid

DNA insert, kb 1000 250 35 15 10

resolution can be achieved after cloning large DNA fragments and estimating distances using empiric methods such as gels to determine the sizes of an array of overlapping fragments, or DNA sequencing to determine the actual length between genes. As noted below, a parallel goal of the genome project is to insert large segments of the human genome into cloning vectors such as yeast artificial chromosomes (YACs) (Table 5-1). Additional mapping often involves the use of BACs (bacterial artificial chromosomes) and traditional cloning vectors such as lambda phage. Although YACs allow insertion of large fragments of DNA, BACs are gaining favor because of less frequent recombination. The isolation and physical separation of chromosomes has also allowed the preparation of chromosome-specific libraries. The cloning of large segments of genomic DNA into YACS and cosmids greatly facilitates high-resolution gene mapping, particularly using “contigs,” in which an ordered array of overlapping clones can be established (Fig. 5-3). An example of this approach on chromosome 21 is described further in Chapter 117. The goal of achieving a high-resolution physical map of the human genome has now been achieved and greater than 95% of the genome has been cloned into overlapping fragments. A major advance has been the use of sequence tagged sites (STSs) as a standard unit for physical mapping. The STSs serve as landmarks that allow overlapping cloned fragments to be arranged in the same order in which they occur in the genome. The STSs consist of 200–500 bp, which can be retrieved from computer databases as opposed to having to use stored clones. Thus, using PCR, investigators can amplify an STS to gain ready access to map locations. The Whitehead group has now described the locations of more than 15,000 STS markers, which were screened against about 30,000 YAC clones from the CEPH-Genethon libraries. As of 1995, this map had an average distance of about 200 kb between markers. Ideally STSs would be available at intervals between 50 and 100 kb, which would allow the rapid localization of most disease genes. This will require at least 30,000 STSs. The feasibility of this approach has been enhanced by automation using a “genomatron” that performs 150,000 PCR reactions per run.

DNA SEQUENCING A long-term goal of the genome project is to obtain DNA sequences for the entire human genome as well as model organisms, such as Saccharomyces cerevisiae (now completed, 12.4 million bp), Caenorhabditis elegans, and the mouse. At present, only a few percent of the human genome have been sequenced. The amount of DNA sequencing required is daunting for several reasons. Currently available technology is relatively expensive (0.20–0.30 dollars/base) and slow. Sequence accuracy and information management must be assured in parallel with technical advances. A current goal is to achieve 99.99% (1 error in 10,000 bp) accuracy. The issue of polymorphic variants in

Figure 5-3 A physical genomic map with an ordered array of “contigs.” A genetic map is shown with polymorphic markers A and B, shown on either side of a disease gene. Below the genetic map, a physical map consisting of overlapping clones is shown. The group of clones are referred to as a contig and can be ordered based on shared sequences, usually a sequence-tagged site (STS).

human sequence will only be addressed in the long term. There are probably polymorphic sequence variants approximately 1 out of every 1000 bp. Although this variation is low (99.9% identical), it suggests as many as 3 million sequence differences between any two unrelated individuals. Despite these challenges, there have been remarkable advances in sequencing capabilities. A decade ago, most DNA sequencing was performed manually. Radioactive sequences had to be developed onto film and the nucleotide sequence was entered into databases manually. Automated DNA sequencing has made a major impact on throughput, particularly because the fluorescently labeled sequence can be read directly into computer systems. A goal is to develop sequencing capability of approximately 50–80 Mb per year. This will require the combined use of robotics and automation in conjunction with informatics. It is presently unclear whether the DNA sequence of the entire genome will be completed by the year 2005. However, several aspects of the genome project have progressed at an accelerated rate because of unanticipated technological advances. At present, most sequencing efforts remain based on gels that resolve DNA sequence at a single nucleotide level. There are efforts to develop new sequencing technologies, including the use of DNA chips. This technology, based on the use of arrays of oligonucleotides that are applied to DNA chips, is particularly promising for detecting variations and mutations in known DNA sequences.

MAJOR ADVANCES AND FUTURE CHALLENGES Considering its controversial inception, the human genome project has progressed remarkably well. The genome project has already catalyzed many advances in human genetics. The development of a high-resolution genetic and physical map makes it possible to more rapidly identify disease-causing genes by positional cloning. These efforts in turn provide additional sequence and physical map information about specific regions of the genome. As depicted in Fig. 5-4, the number of disorders identified by positional cloning has increased markedly in recent years; this trend can be expected to continue. The identification of the cystic fibrosis gene in 1989 represented a landmark example of positional cloning. In 1996, examples of positional cloning included Friedrich’s ataxia, long QT syndrome, basal cell nevus syndrome, hemochromatosis, a form of Maturity Onset Diabetes of the Young (MODY), and Treacher Collins syndrome, among many others. The genome project has stimulated interest in the private sector in the applications of genetics. Characterization of expressed sequence tags (ESTs) has greatly enhanced the database of known

62

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Figure 5-4 Number of genes that cause inherited disorders identified by positional cloning.

expressed genes. In the future, we can expect these databases to be merged with physical maps such that expressed genes can be superimposed onto physical maps. We can also expect increasing interest in changes in pattern of gene expression in disease states, including cancer. Implicit in the human genome project is the idea that identifying disease-causing genes can lead to improvements in diagnosis, prognosis, and treatment. It is estimated that most individuals harbor several serious recessive genes. Because the frequency of mutations in most of these genes is low, it is unlikely that both members of a couple would carry mutations in the same gene, unless there is consanguinity or a restricted gene pool. However, the polymorphic variation in DNA also confers risk to a large number of less serious susceptibility genes. Most polygenic disorder reflect the cumulative risk of several different susceptibility genes. The ability to test for disease genes, whether serious mutations with relative high gene frequencies or various susceptibility genes, is already a reality and can be expected to expand rapidly with improvements in technology. The diagnosis of most disorders is currently made on clinical grounds in conjunction with radiographic or laboratory tests. Nevertheless, there is often some degree of uncertainty, resulting in a “differential diagnosis.” The ability to add genetic tests will not always remove uncertainty, but it adds a powerful new dimension. For example, a patient with medullary thyroid cancer and a germline mutation in the ret oncogene can be diagnosed with multiple endocrine neoplasia type II. Molecular genetics also improves the categorization of disease. For example, Duchenne’s and Becker’s muscular dystrophies are now known to involve the same genetic locus (dystrophin), but the mutations differ in their impact on the encoded protein. Similarly, the “classic” and “adult-onset” forms of congenital adrenal hyperplasia are now known to represent mutations in the 21-hydroxylase gene that vary in severity. In general, the diagnostic capability of molecular diagnostics is much greater than the prognostic information that results. In part, this reflects still-limited genotype– phenotype correlations. In addition, genetic background and environmental influences can greatly modify the course of many genetic diseases. For example, individuals with lipoprotein abnormalities can have a highly variable clinical course, depending on other genes that act on lipid metabolism as well as environmental influences such as diet composition and weight gain. An ultimate goal is to use the results of the genome project to improve the therapy of diseases. Success in this area might result from several venues. Perhaps the first advance will come from

screening and prevention. Thus, patients known to carry a risk factor gene could be screened selectively for certain diseases or treated prospectively. For example, individuals known to carry genes that predispose to colon cancer would undergo earlier and more intensive screening. Individuals with genes that predispose to non–insulin-dependent diabetes mellitus could receive counseling about diet and weight gain, and may be treated early to prevent metabolic decompensation. Gene discovery will provide targets for screening for new drugs. The availability of recombinant hormones such as growth hormone, insulin, and erythropoietin has already had a major impact on clinical medicine. The potential role of gene therapy has received much attention. For certain enzyme deficiency disorders, such as adenosine deaminase deficiency, the technique has great promise. Its role for many other diseases, such as cystic fibrosis or sickle-cell anemia, remains to be established. In addition to classic genetic disorders, gene therapy may have an adjunctive role in the management of atherosclerotic lesions or the treatment of cancers. Currently, major issues that face gene therapy include the need to target therapy to specific tissues (e.g., cancers), the need to regulate gene expression (e.g., insulin in diabetes), and the need to sustain long-term expression (e.g., cystic fibrosis). The genome project raises many ethical issues, which have been addressed in parallel with the other major goals of the project. Advances in such areas as the predisposition to cancer, atherosclerosis, or degenerative neurologic diseases such as Huntington’s or Alzheimer’s disease raise important questions concerning genetic counseling, privacy, and genetic discrimination in the workplace or by insurance companies. In addition, the scientific advances of the genome project require ongoing public and professional education, and there is the potential for unrealistic fears or expectations. The US federal government has established the “Ethical, Legal, and Social Implications” (ELSI) working group to help address ethical issues that arise from the genome project. Analogous groups exist in several other countries. Many issues raised by the genome project are familiar in principle to medical practitioners. For example, a patient with increased LDL cholesterol, high blood pressure, or a strong family history of early myocardial infarction, is known to be at increased risk of coronary heart disease. Likewise, patients with phenylketonuria, cystic fibrosis, or sickle-cell anemia are often identified as having a genetic disease early in life. These precedents can be helpful for adapting policies that relate to genetic information. The genome project is, nevertheless, accelerating the rate of new information. In addition, the new information is greatly expanding the repertoire of diseases that can be characterized at the genetic level. One confounding aspect of the rapid expansion of information is that the ability to make clinical predictions often lags behind the genetic advances. For example, when genes that predispose the breast cancer, such as BRCA1, are described, there is tremendous public interest in the potential to predict disease; but many years of clinical research are required to rigorously establish genotype and phenotype correlations. In the future, increased education will be required to adequately address many of the issues related to advances in molecular medicine. Whether related to informed consent, participation in research, or the management of a genetic disorder that affects an individual or their families, there is great need for more information on the fundamental principles of genetics. The lay press has glorified many aspects of genetics and, in such

CHAPTER 5 / THE HUMAN GENOME PROJECT

cases as gene therapy, expectations have been raised to unrealistic levels. The pervasive nature of molecular medicine also makes it imperative for physicians and other health care professionals to become more informed about genetics and to provide advice and counseling in conjunction with trained genetic counselors (see Chapter 9).

SELECTED REFERENCES Collins F, Galas D. A new five-year plan for the U.S. Human Genome Project. Science 1993;262:43–46. Haseltine WA. Discovering genes for new medicines. Sci Am 1997; 276:92–97.

63

Hudson KL, Rothenberg KH, Andrews LB, Kahn MJ, Collins FS. Genetic discrimination and health insurance: an urgent need for reform. Science 1995;270:391–393. Jordan E, Collins FS. A march of genetic maps. Nature 1996;380: 111,112. Schuler GD, Boguski MS, Stewart EA, Stein LD, Gyapay G, Rice K, et al. A gene map of the human genome. Science 1996;274: 540–546. Venter JC, Smith HO, Hood L. A new strategy for genome sequencing. Nature 1996;381:364–366. Watson JD. The human genome project: past, present, and future. Science 1990;248:44–49. Weissenbach J. Landing on the genome. Science 1996;274:479.

CHAPTER 6 / THE CELL CYCLE

6

65

The Cell Cycle LYNDA Q. NGUYEN AND J. LARRY JAMESON

INTRODUCTION

stages: G1, S, and G2 (Fig. 6-2). In addition, G0 refers to a resting state when cells have exited from the active cell cycle and have not committed to reenter the process of cell division. During G1, the effects of extracellular nutrients, mitogens, and growth factors induce the transcription of genes that are necessary for DNA synthesis, and the cell transitions into a committed state that will ultimately lead to the replication of DNA during the S (synthetic) phase. During G2, there is additional cellular growth and repair of DNA replication errors before the cell enters mitosis (M phase). After mitosis, cytokinesis, the actual process of cell division, occurs. If conditions are not appropriate for another round of cell division, G0 can be entered for an indefinite period of time before the cell makes the transition back into G1. In a rapidly proliferating somatic cell, the entire cycle requires about 18–24 h. G1 occupies the longest time and may require up to 12 h. DNA replication during S phase may last from 6 to 8 h, depending on the size of the genome and the number of replication origins that are initiated simultaneously. G2 requires approximately 3–4 h, and mitosis can occur within an hour.

The cell cycle consists of a set of highly ordered events that result in the duplication and division of a cell. This process requires the synthesis of a new copy of DNA, segregation of chromosomes, mitosis, and apportionment of the cellular contents. Multiple extracellular signals control entry and exit from the cell cycle to coordinate normal cell growth and to avoid uncontrolled cell proliferation. Various steps in the progression of the cell cycle are regulated rigorously to allow surveillance of the cycle and to avoid errors in DNA replication. This chapter reviews the molecular basis for control of the cell cycle. Much progress has come from studies of yeast, clams, and other model systems. These mechanisms are largely conserved in mammalian systems, which are the main focus of this chapter.

GENERAL MECHANISMS Early studies using light microscopy allowed some of the events the cell cycle, such as mitosis, to be observed. When DNA was radiolabeled with 3H-thymidine, it was possible to visualize the segregation of chromosomes. From these experiments, it was learned that newly replicated DNA becomes condensed and aligned at the mitotic spindle, and that sister chromatids segregate to opposite poles of the cell before the actual division of the cell. There are few visible changes during the interval between cell divisions (interphase) except for the increase in cell volume and mass. For this reason, molecular and genetic approaches have been essential for understanding the processes that control the cell cycle. Cellular replication can be divided into several distinct phases. Mitosis (M phase) refers to the process of chromosomal segregation and actual cell division. Mitosis results in the division of one cell into two identical daughter cells, each bearing a complete diploid (2n) complement of chromosomes. Mitosis is divided into prophase, metaphase, anaphase, and telophase. These stages, along with several additional subdivisions, are based on characteristic features associated with chromosomal segregation, mitotic spindle formation, and cell division (Fig. 6-1). The cell cycle may also lead to meiosis, producing gametes with a haploid (1n) number of chromosomes. Although the mechanisms controlling mitosis and meiosis have some similarities, this chapter will focus on the mitotic cycle of the somatic cell. The interval between cell divisions is referred to as interphase, which is classically divided into three

THE CYCLINS AND THEIR CYCLIN-DEPENDENT KINASES The cyclin-dependent kinases (CDKs) coordinate specific transitions that occur at defined times during the cycle. Originally described by Hartwell in his studies of the budding yeast Saccharomyces cerevisiae, the genes encoding these proteins were named cell-division cycle or cdc genes. Mutations in the cdc genes caused an arrest at specific points in the cell division cycle. One of the first cdc genes to be isolated was cdc2, which was found by Nurse during studies of the fission yeast, Schizosaccharomyces pombe. Mutations in cdc2 caused arrest at two points in the cell cycle; either before “start,” where the yeast cell becomes committed to DNA replication, or just before mitosis. The cdc2 gene encodes a 287 amino acid protein with high homology to protein kinases. As opposed to the yeast S. cerevisiae, which contains a single cell cycle CDK (cdc28), and S. pombe (cdc2), mammalian cells have several: Cdc2, CDK2, CDK3, CDK4, CDK6, and CDK7, which act at different transitions in the cell cycle. Activation of CDK activity requires their association with another group of proteins known as the cyclins (cyclin D, H, E, A, and B). Cyclins were initially identified by Hunt in developing sea urchin and clam eggs. The levels of the various cyclins fluctuate during the cell cycle in a characteristic manner (Fig. 6-3). Cyclin D is induced during G1 and remains elevated throughout the cell

From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

65

66

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Figure 6-1 Stages of mitosis. For simplicity, only the nucleus is shown, and only two chromosomes are illustrated. Although mitosis is a continuous process, it is useful to divide it into several distinct stages. Prophase refers to the initial stage of mitosis during which chromosomes begin to condense, centrioles appear and initiate the formation of the mitotic spindle, and the nucleus begins to disintegrate. Metaphase is characterized by the alignment of chromosomes along the equitorial plate of the cell. The chromosomes are highly contracted at this stage and are attached to the microtubules via kinetochores. Anaphase begins with the separation of the two chromatids of each chromosome. Telophase is characterized by reformation of the nuclear envelope around the dividing nuclei and by the decondensation of chromatids as the cells are restored to the interphase state. Several of these stages of mitosis that can be subdivided into intermediate events are not shown.

Figure 6-2 Overview of the cell cycle. A schematic representation of various changes in cyclin/CDK activity during the cell cycle is depicted. Growth factors induce the synthesis of D-type cyclins, thereby causing an association with CDK4 and CDK6 in early G1. Additional phosphorylation of CDK4/CDK6 by CAK is required for CDK activation. Phosphorylation of Rb by cyclin D/CDK4/CDK6 is required to drive the cell past the restriction point, where mitogenic stimulation is no longer needed, and signals the entry into S phase. Cyclin E/CDK2 kinase activity is subsequently needed to maintain Rb in its hyperphosphorylated state. Although cyclin A is synthesized in late G1, it associates with CDK2 throughout S phase and later with cdc2 in late S phase and early G2. Cyclin B complexes with cdc2 during G2; this complex was originally known as MPF. CDK inhibitors either block cyclin/CDK assembly and/or inhibit CDK activities, serving to arrest cells at specific points until conditions are optimal to reenter the cell cycle. Cells emerging from quiescence in G0 or from mitosis may reenter G1 and continue with the division process.

cycle. Cyclin E is produced transiently in late G1 and declines before S phase. Cyclins A and B are also known as the mitotic cyclins. Cyclin A is synthesized in late G1 and throughout S phase. It activates CDK2 during S phase and cdc2 during G2. Cyclin B is produced during G2 and also activates cdc2. Both cyclins A and B are degraded during M phase. All cyclins contain a characteristic

100-amino acid “cyclin box” which is located in the aminoterminal region. Proteolysis of the mitotic cyclins is conferred by a 9-amino acid motif referred to as the destruction box (D box), which is also located at the amino-terminal region of the proteins. The D box is necessary for ubiquitination and subsequent proteolytic degradation. Proteolysis of cyclins A and B triggers exit

CHAPTER 6 / THE CELL CYCLE

Figure 6-3 Changes in cyclins during the cell cycle. The levels of different cyclins are depicted schematically. Early in G1, growth factors and mitogens induce the transcription of the D-type cyclins (D1, D2, D3) that accumulate and persist through most of the cell cycle. After G1, cyclin D is translocated from the nucleus to the cytoplasm. Cyclin associates with several CDKs (CDK 2, CDK 4, CDK 6). Cyclin E accumulates in late G1, associates with CDK2, and is destroyed rapidly as cells enter S phase. The mitotic cyclins, cyclin A and cyclin B, increase in S phase and G2, and are degraded by the anaphase promoting complex (APC) before the end of mitosis. Cyclin A associates with CDK 2 during S phase and with cdc 2 during G2. cdc2cyclin B complexes are restricted to M phase.

from mitosis. As described below, proteolysis also occurs at other phases of the cell cycle to enable specific transitions (e.g., proteolysis of CDK inhibitors during G1; proteolysis of inhibitors to allow transition from metaphase to anaphase). The cyclin-CDK complexes are subject to input from multiple pathways. The kinase activity of the CDKs is regulated by phosphorylation (both positively and negatively) and by dephosphorylation. The cyclin/CDK complexes are also negatively controlled by several groups of inhibitors (Fig. 6-2). Proteins such as p21WAF1/CIP1 act by directly inhibiting CDK kinase activity, whereas proteins such as Suc1 modify the specificity or accessibility of CDKs for regulatory proteins. Cyclins also contribute to CDK substrate specificity. The phosphorylation of substrates (e.g., Rb) by specific cyclin/ CDK complexes induces the transcription of genes, whose products are needed to drive the cell through each transition of the cell cycle. It can be speculated that each of these regulatory proteins, in addition to the cyclins and their associated kinases, are potential targets for dysregulation of cell cycle control.

RESTRICTION POINT CONTROL AND G1 TO S PROGRESSION Mitogenic signals, such as growth factors, induce cells to enter and progress through G1. The D-type cyclins (D1, D2, and D3) serve as growth factor sensors and are transcribed in response to growth factor stimulation (Fig. 6-3). Cyclin synthesis, and interaction with their CDK partners (CDK4 and CDK6), also depends

67

on mitogenic stimulation. As long as nutrient and growth factor availability is sustained, cyclin synthesis and the assembly of cyclin D/CDK4 and cyclin D/CDK6 complexes continue to occur. After a cell reaches the restriction point (R) in late G1, it is committed to DNA replication and subsequently enters S phase, where mitogenic stimulation is no longer required for cell cycle progression. Cyclin D/CDK4/CDK6 complexes are activated through phosphorylation by a CDK-activating enzyme known as CAK, which is now recognized to be a complex between cyclin H and CDK7. The kinase activities of these cyclin/CDK complexes are crucial in the activation of substrates that play a vital role in the G1 to S phase transition. A primary target of the cyclin/CDK complex is the protein product of the retinoblastoma (Rb) tumor suppressor gene. Phosphorylation of Rb results in the dissociation of E2F transcription factors and is necessary for passage through the restriction point (see below). A family of inhibitors known as the Inhibitors of CDK4 (INK4) proteins block the activities of the cyclin D/CDK4/CDK6 complexes. These inhibitors, which include p16Ink4, p15Ink4B, p18Ink4C, and p19Ink4D, can induce G1 arrest (Fig. 6-2). Inactivation of these inhibitors allow uncontrolled cell proliferation and increased genomic instability, and p15 and p16 are targets for mutations in certain malignancies (see Chapters 7 and 85). Inactivation of the Rb gene shortens the G1 phase, reduces cell size, and decreases the dependence on mitogenic stimulation. Disruption of the Rb pathway occurs in several forms of cancer (see Chapters 7 and 105). In cultured cells that lack Rb, the ectopic expression of INK4 proteins no longer arrests cells, confirming that Rb is an important target for these inhibitors. In late G1, the synthesis of cyclin E is upregulated, presumably through E2F, since the cyclin E gene has been shown to contain an E2F-response element. Cyclin E/CDK2 activity may therefore be regulated by a positive feedback mechanism. Entry into S phase is highly dependent on the proteolytic degradation of cyclin E by ubiquitin-dependent pathways. Phosphorylation by its catalytic partner, CDK2, targets cyclin E for destruction. As cyclin E/CDK2 activity decreases, cyclin A synthesis is induced. Accumulation of the cyclin A/CDK2 complexes signals entry into S phase. Cyclin D/CDK2 and cyclin A/CDK2 complexes may interact with the DNA replication origins. Cyclin A/CDK2 also binds E2F/DP-1 dimers, phosphorylating a regulatory residue on DP-1, thereby preventing DNA binding of the E2F/DP-1 complex. Inactivation of E2F helps to ensure cell cycle progression into S phase and prevents reversion back to G1. The cyclin D/CDK4/CDK6, cyclin E/CDK2, and cyclin A/CDK2 kinase activities are inhibited by another group of proteins that includes p21WAF1/CIP1, p27KIP1, and p57KIP2 (Fig. 6-2). These CDK inhibitors (CDKIs) also impair CAK activity. The p27 gene product may be one the most important proteins involved in restriction point control; p27 represses the activity of cyclin E/CDK2 and cyclin A/CDK2 until entry into S phase. Depriving proliferating fibroblasts of serum mitogens increases p27, thereby causing an immediate arrest at G1. The tumor suppressor gene, p53, also induces p27. In addition, p15 and p27 may provide a pathway for TGF-β-mediated growth suppression. The p15 gene is activated in response to TGF-β, a growth factor that suppresses the proliferation of several cell types, including colon and prostate epithelial cells; p15 binds cyclin D/CDK4 and cyclin D/CDK6 complexes to displace p27, allowing it to bind cyclin E/CDK2 and block cell cycle progression.

68

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

S PHASE Once a cell is committed to the initiation of DNA replication, the process continues until duplication of the entire genome is complete. In the human genome, this entails the accurate duplication of approximately 3 billion bp of DNA. An enormous level of replication fidelity is required to accurately synthesize this amount of DNA. In addition to a highly regulated and well-conserved replication system, the existence of a proofreading mechanism (3' to 5' exonuclease) that detects misincorporated nucleotides, and a mismatch-repair system, helps to assure replication fidelity. In eukaryotes, an intricate network of DNA repair systems can delay the cell cycle to allow polymerase errors to be recognized and repaired. These systems also repair DNA damage that is induced by mutagens, such as oxidation products, spontaneous depurination or depyrimidination, chemical adducts, and the effects of ultraviolet (UV) radiation and ionizing radiation. Defects in the DNA repair process predispose to several forms of cancer. The cancer-prone disorder, xeroderma pigmentosum (XP), results in deficient repair of UV-induced and oxidationinduced lesions (see Chapter 81). DNA repair defects and hypersensitivity to DNA damage are also associated with several other autosomal recessive syndromes, including Bloom’s syndrome (BS), Fanconi’s anemia (FA) and ataxia telangiectasia (AT). The search for the human homologs of yeast mismatch-repair genes allowed the identification of MSH2. Soon thereafter, defects in this gene were shown to cause hereditary nonpolyposis colon cancer (HNPCC). In addition, MSH2-deficient mice have been shown to be highly susceptible to lymphoid tumors. Several other genes, including MLH1 and PMS2, whose products are also involved in mismatch repair, are additional causes of HNPCC. Another form of DNA repair, called transcription-coupled repair, occurs during the transcription of DNA into RNA. Long strands of DNA that are unwound during transcription must be mended to reform the double strands. Patients with Cockayne syndrome (CS) exhibit defective mechanisms in transcription-coupled repair. One of the important roles of the p53 tumor suppressor occurs during times of damage to DNA. DNA strand breaks induced by UV or ionizing radiation induce production of the p53 protein; this in turn stimulates the expression of genes (e.g., p21) that cause cell cycle arrest. p21 prevents DNA synthesis by binding proliferating cell nuclear antigen (PCNA), a subunit of the DNA polymerase δ enzyme complex involved in both DNA replication and repair. p53 also induces GADD45 and cyclin G, both of which contribute to G1 cell cycle arrest, allowing time for DNA repair. Germline mutations in p53 cause Li-Fraumeni syndrome (LFS), which is associated with a high incidence of multiple cancers, including breast, ovary, and brain tumors (see Chapter 7). Inactivating p53 mutations are also among the most common somatic mutations in malignancies, emphasizing its importance in the maintenance of normal cell growth.

MECHANISMS CONTROLLING DNA REPLICATION Important insight into mechanisms controlling DNA replication initially came from a series of cell fusion experiments. When a cell in G1 was fused to a cell in S phase, the nucleus of the G1 cell began to replicate its DNA prematurely, indicating that the G1 nucleus is competent for replication and that the S phase cell contains an activator. However, when G2 cells were fused with G1 cells, the G2 nuclei failed to reinitiate DNA replication, whereas

the G1 nuclei replicate normally, indicating that G2 nuclei cannot rereplicate DNA until passage through mitosis. The same held true when G2 nuclei were fused with cells in S phase. Several conclusions were drawn from the above experiments. First, only chromosomes from the G1 phase of the cell cycle are competent for DNA replication. Second, cells in S phase contain an activator that initiates DNA replication from G1 chromosomes. Third, G2 cells do not contain repressors and their nuclei must progress through mitosis before DNA replication can occur. The control of DNA replication relies on the ordered assembly of specific proteins at the origins of replication. These proteins are necessary to form a competent, prereplicative chromosomal state. In eukaryotes, the origins of replication are determined by cisacting sequences called replicator elements, and trans-acting proteins (initiator proteins) bind to the replicators. Because of the large amount of DNA that must be duplicated in eukaryotic genomes, multiple origins of replication are formed to accommodate the constraints of size and time. In the eukaryotic genome, approximately 103 to 105 replication initiation events occur during each cell cycle. Recent discoveries in yeast have further elucidated this process. In S. cerevisiae, a multisubunit complex known as the origin recognition complex (ORC) has been shown to bind the initiator elements A and B. Consisting of six proteins, the ORC serves as a docking site for protein–protein interactions that regulate the initiation of replication. Other proteins, including cdc6p from budding yeast, and cdc18+ from fission yeast, are also needed to form the prereplication complex. The proper assembly of this multisubunit complex defines the prereplicative state of the chromosomes in G1. Studies using Xenopus egg extracts have provided evidence for activators of DNA replication. G1 nuclei from human HeLa cells are able to initiate DNA replication in the presence of egg extracts, but G2 nuclei are not. When the G2 nuclei are permeabilized and then repaired, DNA replication occurs. This finding led to the idea that G1 nuclei are competent, because factors sequestered in the cytoplasm are able to interact with chromatin when the nuclear envelope is disassembled during mitosis. However, once initiation of DNA replication occurs, the factor is destroyed, thereby preventing rereplication of the DNA. Using yeast genetics, the mini-chromosome maintenance (MCM) family of proteins have been identified as factors that confer the replication-competent state of chromatin in G1. The MCM proteins include MCM2, MCM3, MCM4/cdc54, MCM5/ cdc46, and MCM7/cdc47, and mutations in their genes cause a high rate of mini-chromosome loss. MCM proteins that are bound to chromatin are degraded as S phase progresses. The formation of a prereplication complex therefore consists of the ORC, proteins such as Cdc6p, and the MCM complex. The activity of cyclins and their associated CDKs is important for activating DNA replication as well as blocking the reassembly of the prereplication complex once DNA synthesis is complete. During S phase, this requires the association between cyclin A and CDK 2, and in M phase, cyclin A/cdc2 and cyclin B/cdc2. The mechanisms involved in the actual process of DNA replication are reasonably well-understood. As described above, DNA synthesis is initiated from multiple independent replicons. Because DNA is double-stranded, it is necessary to copy both strands simultaneously (Fig. 6-4). This can occur in either a unidirectional

CHAPTER 6 / THE CELL CYCLE

69

G2 TO M PHASE TRANSITION During G2, there are checkpoints to assure that DNA has been faithfully copied before entry into mitosis. DNA repair systems can be activated to correct errors that may have occurred during the replication process. In addition, cellular proteins involved in the assembly of the mitotic spindle and cell division are produced. The cyclin/CDK complex involved in M phase was first identified through use of the Xenopus oocyte assay. When meiotic oocytes arrested in G2 are injected with M-phase cytoplasm from a mature unfertilized egg, the recipient oocyte is driven into M phase. These experiments led to the conclusion that a maturation promoting factor (MPF) in the cytoplasm of the egg caused the oocyte to complete cell division. MPF is now known to be a complex between the mitotic cyclin B and cdc2. Cyclin B, which is synthesized in late S phase, complexes with unphosphorylated cdc2. In S. cerevisiae, an additional level of regulation is provided by the competing actions of the tyrosine kinase Wee1 and the phosphatase cdc25. Phosphorylation by Wee1 on tyrosine 15 inactivates cdc2 kinase activity, whereas dephosphorylation of the same residue by cdc25 reactivates the complex. It is likely that mammalian homologs of these proteins also exist. Targets for MPF are not well-characterized, but both mitogen activated kinase (MAPK) and MPM2 kinase are potential substrates.

PROTEOLYSIS AND THE M TO G1 TRANSITION

Figure 6-4 Replication of DNA. A replication fork is illustrated. After separation and unwinding of the DNA double helix, DNA replication proceeds in the 5' to 3' direction on opposite strands. The leading strand is synthesized continuously towards the replication fork. Synthesis of the lagging strand is initiated discontinuously as the fork unwinds, resulting in Okazaki fragments that are subsequently ligated together. At the end of chromosomes, telomeres exist to allow repair of the ends of DNA. The enzyme, telomerase, extends the 3' end of the parental strand, providing an elongated template.

(one replication fork) or bidirectional (two replication forks) manner. An RNA fragment that is complementary to the singlestranded region exposed by the replication fork serves as a primer for DNA synthesis, which occurs in a 5' to 3' direction. On the “leading” strand, DNA synthesis occurs continuously, whereas it is discontinuous on the opposite “lagging” strand, which is made available as the fork unwinds. The discontinuous Okazaki fragments on the lagging strand are subsequently ligated together. A unique circumstance arises for DNA replication at the ends of the chromosomes. Since DNA polymerase requires a labile RNA primer to initiate synthesis, some bases at the 3' end of each template are not copied. Thus, the DNA strands would become progressively shorter with each cycle of replication in the absence of a mechanism to repair the ends. The identification of repeated GC-rich fragments at the ends of chromosomes, referred to as telomeres, provides a compensating mechanism for this process. Telomeres act as a template for the enzyme, telomerase, which extends the repeats (in the 3' direction) with each cycle of division. The presence of telomerase is required for ongoing cell division, and the loss of telomerase activity may represent a mechanism of cellular senescence.

Protein degradation is required for the preparatory events that occur before DNA replication and for entry into and exit from mitosis. Two major ubiquitin-dependent pathways mediate proteolysis at distinct points in the cell cycle (Fig. 6-5). The first pathway, involving a protein called cdc34, initiates DNA replication by degrading the CDK inhibitor SIC1. Using temperaturesensitive cdc34 mutants in budding yeast, it was discovered that cdc34 encodes a ubiquitin-conjugating enzyme, E2. Following activation by the E1 ubiquitin-activating enzyme, ubiquitin is transferred to the target protein, SIC1, by cdc34 (E2) through a transesterification reaction. Subsequent multiubiquitination by the ubiquitin-ligating enzyme E3 targets SIC1 to the 26S proteasome for degradation; cdc34 also participates in the destruction of the G1 cyclins, CLN2 and CLN3, in yeast. Ubiquitin-dependent degradation of a CDK-inhibitor represents a crucial event needed for entry into S phase. The second pathway involves a multiprotein complex called the anaphase-promoting complex (APC). This complex mediates the entry and exit from mitosis by targeting anaphase inhibitors (PDS1 and CUT2) and mitotic cyclins A and B for destruction, also via the ubiquitin-proteasome pathway (Fig. 6-5). Although APC appears to be a large E3, it acts primarily by bringing together ubiquitin-charged E2s and D-box-containing substrates. As discussed previously, the D-box is a 9-amino acid motif in the aminoterminal domain of mitotic cyclins, which is necessary for degradation via the ubiquitin pathway. This degradation represents an irreversible step that assures exit from mitosis.

CELL-CYCLE CHECKPOINTS The empirical definition of a cell-cycle checkpoint was provided by Hartwell and Weinert as an event B that is dependent on the completion of a prior event A. A checkpoint occurs if a lossof-function mutation relieves the dependence. In biochemical terms, a checkpoint represents a surveillance system that allows the detection of an incomplete previous step in the cell cycle or

70

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Figure 6-5 Role of proteolysis in cell cycle control. Proteolysis plays a critical role in several transitions during the cell cycle (see review by Kirschner and colleagues for details). Degradation of the mitotic cyclins (cyclin A and B) is mediated by the APC and is required for exit from mitosis. During G1, the cdc34 pathway degrades inhibitors of G1 CDKs (e.g., CDK2, CDK4), allowing passage into S phase. Inactivation of the APC allows the accumulation of mitotic cyclins that are involved in S-phase as well as the transition into mitosis. The APC also mediates proteolysis of anaphase inhibitors, permitting a transition from metaphase to anaphase.

damage to the genome or mitotic spindle. When cell-cycle checkpoint pathways are functioning correctly, cells arrest at specific phases, depending on when damage is sensed, to allow time for repair and to activate genes involved in the repair process. Checkpoint regulatory pathways have been elucidated primarily based on analyses of cdc mutants in budding and fission yeast. Checkpoints exist for DNA damage, DNA replication blocks, and improper spindle assembly. The DNA damage checkpoint can be implemented at three times during the cell cycle: G1/S transition, progression through S phase, and at the G2/M transition. Four genes in S. cerevisiae, RAD9, RAD17, RAD24, and MEC3, serve as sensors of single-strand DNA, and they are required for G1 and G2 arrest in response to DNA damage. Three genes act at the replication checkpoint in budding yeast: POL2, DPB11, and RFC5. They are all candidate sensors of DNA replication. DPB11 is required for replication arrest in response to hydroxyurea, an inhibitor of ribonucleotide reductase. DNA damage occurring during S phase is sensed by POL2, whereas DNA damage incurred during G1 and G2 is sensed by RAD9. RCF5 is a component of the replication factor C that binds to gapped DNA (e.g., during lagging-strand DNA synthesis). RCF5 recruits PCNA and DNA polymerase δ for repair. Two proteins, MEC1, a member

of the phosphoinositide kinase family, and RAD53, a protein kinase, transduce the DNA damage signals to the repair systems. The sensors of spindle assembly are less well characterized. However, several yeast genes—MAD1, MAD2, and MAD3 (mitotic arrest defective), and BUB1, BUB2, and BUB3 (budding uninhibited by benzimidazole)—have been implicated in the spindle assembly checkpoint. Similar checkpoint control pathways exist in mammalian cells, although fewer of the proteins have been identified. Three proteins controlling the DNA damage checkpoint include ATM (mutated in ataxia telangiectasia), the tumor suppressor p53, and the CDK inhibitor p21. p53 is a transcription factor that is induced in response to DNA damage, although how it is activated remains unknown. As noted above, p53-mediated G1 arrest results from activation of the p21 gene. p21 directly inhibits the activity of CDKs needed for entry into S phase, and p21 knockout mice exhibit a partial failure to arrest at G1. Inactivation of p53 (e.g., by SV40 large t-antigen) may extend the life of cells that are of late passage, thereby inplicating p53 in cellular senescence. Cells from p53 nullizygous mice are able to escape cellular senescence and produce aneuploid immortalized cell lines. p53 appears to be at least partially regulated by ATM, as cells lacking ATM show reduced and delayed activation of the p53 gene in response to DNA damage. However, independent regulatory mechanisms must exist, as ATM mutant cells can still undergo p53-dependent apoptosis. Apoptosis can be considered a type of cell-cycle checkpoint. In the intact organism, destruction of damaged cells by apoptosis provides an important control against the accumulation of genetic damage. The p53 gene is one of the most frequent mutations in human cancers. The ATM gene is a target of mutations as well; cells from patients with the cancer-prone syndrome ataxia telangiectasia exhibit defects in all three DNA damage-induced checkpoints. Cellular senescence may also be considered as a cell-cycle checkpoint. This phenomenon is especially evident in culture, where normal cells undergo a limited number of divisions before reaching senescence. As noted above, shortened chromosomal telomeres and suppression of telomerase activity may be a mechanism for cell senescence. The re-expression of telomerase activity often occurs in tumorigenesis, resulting in the development of immortal cells.

CONTROL OF CELLULAR PROLIFERATION AND DIFFERENTIATION A multitude of factors influence the ability of a cell to proceed through the cell cycle, to advance into a more differentiated state, or to undergo apoptosis. Factors such as cell-to-cell contact, viral infection, and mitogenic or inhibitory signals (or their withdrawal) initiate intracellular cascades that communicate signals to the molecular machinery in the nucleus. Many of these pathways, including their respective receptors and corresponding intracellular signaling molecules (second messengers and protein kinases), have been identified, and some have been implicated in tumorigenesis (see Chapters 7 and 46). Although remarkable progress has been made in our understanding of the complex signaling pathways for growth factors, the mechanisms by which they induce cell division remain unclear. Terminally differentiated cells do not reenter the cell cycle. However, many growth factors induce both cell proliferation and differentiation, presumably by acting on progenitor cells that have

CHAPTER 6 / THE CELL CYCLE

not progressed beyond an irreversible stage of differentiation. For example, insulin stimulates the division of many cells, but it also induces differentiation (e.g., 3T3-L1 fibroblasts into differentiated adipocytes). In the case of hematopoietic cells, various cytokines stimulate the division of progenitor cells, but also move cells along a differentiation pathway. One explanation for these dual effects of growth factors involves their ability to activate parallel signaling cascades that ultimately have distinct cellular effects. It is also important to view the effects of a growth factor in the context of other extracellular stimuli that may act synergistically or antagonistically. Thus, the convergence of signaling pathways from different growth factors can elicit selective cellular responses. The pattern and timing of growth factor exposure is also important in determining cellular responses. For example, brief exposure to nerve growth factor (NGF) may induce cell division, whereas prolonged exposure favors neuronal differentiation. Extracellular signals exert their strongest influence on the cell cycle during a narrow window of time in G1. Signals from growth factors such as epidermal growth factor (EGF) and insulin-like growth factor-1 (IGF-1) are integrated along with those from inhibitory factors such as transforming growth factor-β (TGF-β). An example of growth factor signaling is illustrated by EGF, which is a potent stimulator of cell proliferation. EGF binds to its membrane-bound tyrosine kinase receptor, resulting in receptor dimerization and autophosphorylation of specific tyrosine residues. These phosphorylated tyrosines induce the recruitment of Grb2 through its SH2 domain and the Sos protein through its SH3 domain. This complex activates the Ras signaling pathway (see Chapter 46). GTP-bound Ras stimulates the serine-threonine kinase Raf, which phosphorylates the mitogen-activated protein kinase (MAPKK or MEK). MAPKK phosphorylates the mitogenactivated protein kinase MAPK (or extracellular signal-related kinase; ERK). MAPK acts in the nucleus to phosphorylate and regulate the activity of oncogene products involved in cell cycle progression and growth, including c-fos, c-jun, and c-myc. In addition, cyclin D is upregulated in response to mitogens, leading to the activation of cyclin-dependent kinase activity. Interactions mediated by cadherins provide an additional means of cell–cell as well as intracellular signaling. Cadherins are transmembrane Ca2+-dependent adhesion receptors that play important roles in cell recognition and sorting during development and in the maintenance of solid tissue. Cadherins form complexes with cytoplasmic proteins known as catenins (α and β catenin). The association of cadherins with catenins is important for cytoskeletal alterations associated with cell recognition and morphogenesis. The intracellular domain of several cadherins (e.g., E cadherin) interacts with cytoskeletal-related proteins, such as globular actin, α actinin, vinculin, and ankyrin. The cadherins can also associate with signaling molecules such as the protein tyrosine kinases, Src, and Yes. β catenin is a substrate for Src, and its phosphorylation perturbs cadherin-mediated cell–cell adhesion. Intracellular signaling by cadherins can influence the availability of free β catenin, which serves as a substrate in the Wnt signaling pathway that regulates early development and cell fate in Drosophila. Defects or downregulation in cadherin or β-catenin function has been implicated in tumorigenesis (e.g., APC [adenomatous polyposis coli], DCC [deleted in colorectal cancer]), as loss of cell adhesion is associated with increased invasiveness of tumor cells. Other proteins activated by less direct mechanisms also contribute to the control of cell-cycle progression. Several growth

71

Figure 6-6 Control of Rb activity during G1 to S progression. (A) In its hypophosphorylated form, Rb is capable of binding E2F, preventing E2F-mediated transcriptional activation. (B) Phosphorylation of Rb by the active cyclin/CDK complexes during mid-to-late G1 causes the release of E2F from Rb, allowing E2F to heterodimerize with members of the DP family of transcription factors. As long as Rb is maintained in its hyperphosphorylated state, E2F/DP-mediated transcription of genes required for G1 to S phase progression continues. Factors favoring mitogenesis can regulate transcriptional activity by modulating Rb phosphorylation. In turn, growth inhibitory factors acting through CDKIs can reverse the hyperphosphorylated state of Rb.

factors activate phospholipase C, which hydrolyzes phosphatidylinositol-4, 5-bis-phosphate (PIP2) into two second messengers: diacylglycerol (DAG) and inositol 1, 4, 5-tris-phosphate (IP3). DAG is able to activate protein kinase C, thereby stimulating gene transcription and cell proliferation. IP3 releases calcium ions from intracellular stores and calmodulin, a calcium-binding protein, can also regulate the activities of other kinases in the cell. Accumulating evidence suggests that Rb is a central target for the convergence of many of these signaling pathways (Fig. 6-6). Cells emerging from mitosis require prolonged and continuous growth factor stimulation until the restriction point is reached during mid to late G1. After the restriction point, serum is no longer required, and the cell can progress through S phase and the remainder of the cell cycle. Regulation of the activity of Rb occurs largely as a result of phosphorylation. In its hypophosphorylated form, Rb associates with the E2F family of transcription factors (E2F1–E2F5). E2F-1 through E2F-3 associate with Rb (p105), whereas E2F-4 and E2F-5 associate preferentially with the Rb-related proteins, p107 and p130. The association with Rb can occur while the E2Fs are bound to the promoters of various genes and may account for the ability of hypophosphorylated Rb to actively repress gene tran-

72

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

scription. Cyclin D/CDK4/CDK6 mediate the phosphorylation of Rb, and cyclin E/CDK2 is required to maintain Rb in a hyperphosphorylated state. When Rb is hyperphosphorylated, E2Fs dissociate and heterodimerize with the DP family of transcription factors (DP-1, DP-2, DP-3). These heterodimeric transcription factors bind to sequences in the regulatory regions of genes important in the control of cell growth, including c-myc, B-myb, cdc2, DHFR, thymidine kinase, and E2F-1 itself. Growth inhibitory factors prevent Rb phosphorylation through modulation of the CDKs that phosphorylate Rb. These signals block Rb phosphorylation by activating various CDK inhibitor proteins. Three signals have been identified that block Rb phosphorylation in this manner: TGF-β, cyclic AMP (cAMP), and contact inhibition. As described previously, TGF-β prevents Rb phosphorylation by inducing the expression of p15, which competes with D cyclins for binding to CDK4/CDK6. In addition, TGF-β has been shown to decrease levels of CDK4 in certain cell types. cAMP activates the protein kinase A (PKA) signaling cascade, which can inhibit Rb phosphorylation. cAMP also results in the phosphorylation of transcription factors, such as the cAMP response element-binding protein (CREB). Phosphorylated CREB recruits the coactivator CBP (CREB-binding protein), which links CREB to the basal transcription machinery, promoting the induction of gene transcription (see Chapter 3). In some cell types, activation of cAMP-responsive genes in combination with the inhibition of Rb phosphorylation leads to cellular differentiation. It is important to note that a variety of other Rb-binding proteins have been identified. Among these are Elf-1, MyoD, PU.1, ATF-2, and c-Abl, although the effector functions of these proteins are largely unknown. Hypophosphorylated Rb has been shown to bind the catalytic domain of the nuclear tyrosine kinase c-Abl. This interaction blocks the kinase function of the proto-oncogene; phosphorylation of Rb releases c-Abl and allows it to phosphorylate nuclear substrates, some of which may be involved in cell growth.

APOPTOSIS Apoptosis, or programmed cell death, is a normal cellular process that occurs in response to many different stimuli, including UV and ionizing radiation, chemotherapy, hypoxia, viral agents (E1A and E7), proto-oncogene expression (E2F-1 and c-myc), and growth factor or hormone withdrawal. As opposed to cellular senescence, apoptosis is an energy-dependent process that results in distinct morphological changes in the cell. Nuclei become condensed and fragmented, DNA is degraded, and cells shrink and are eventually degraded by a lysosome-mediated pathway. Apoptosis plays a critical role in development and homeostatic processes, including development of the nervous system and in the lymphoid selection process of the immune system. The apoptotic response can vary depending on the cell type and the nature of the stimulus. A number of gene products involved in cell cycle control mediate this response, including Rb and c-myc, although most work has centered around the actions of p53. p53mediated apoptosis occurs primarily in response to DNA damage induced by UV radiation, hypoxia, or viral infection. A growing body of evidence points toward a cooperative interaction between the p53 and Rb/E2F pathways. The adenovirus E1A and the human papilloma virus (HPV) E7 proteins stabilize and activate p53, thereby inducing p53-dependent apoptosis. In response to p53 actions, viral proteins such as the HPV E6 protein and E1A can bind the tumor suppressor Rb and inactivate it. This relieves

Rb-mediated repression of E2F-DP-1 complexes. If p53 is absent, free E2F can promote uncontrolled progression through the cell cycle. In fact, overexpression of E2F results in failure of cells to arrest in G1 or to exhibit apoptosis. Similar effects have been seen with the loss of Rb. The apoptotic response appears to be induced by separate p53 transcription-dependent and transcription-independent pathways. The bax and insulin-like growth factor-binding protein 3 (IGF-BP3) genes are both regulated by p53. The bax gene encodes a protein with homology to the survival factor Bcl-2. Heterodimerization with Bax inactivates Bcl-2, thereby promoting cell death. p53 regulation of IGF-BP3 can block the mitogenic response of cells to IGF-1 by inhibiting IGF-1 interaction with its receptor. Although programmed cell death can occur by pathways independent of p53 (e.g., glucocorticoids), the p53-mediated pathways prevail in most circumstances. In situations in which DNA damage has occurred, or when the supply of survival factors is low, the p53-dependent apoptotic response becomes engaged. This response is important to preclude the development of tumors, and it also appears to play a role in effective responses to cancer therapy, including chemotherapy and radiation treatments.

SELECTED REFERENCES Baserga R. Oncogenes and the strategy of growth factors. Cell 1994;79: 927–930. Chiu CP, Harley CB. Replicative senescence and cell immortality: the role of telomeres and telomerase. Proc Soc Exp Biol Med 1997; 214:99–106. Collins K, Jacks T, Pavletich NP. The cell cycle and cancer. Proc Natl Acad Sci USA 1997;94: 2776–2778. Elledge S. Cell cycle checkpoints: preventing an identity crisis. Science 1996;274:1664–1672. Evans T, Rosenthal ET, Youngblom J, Distel D, Hunt T. Cyclin: a protein specified by maternal mRNA in sea urchin eggs that is destroyed at each cleavage division. Cell 1983;33:389–396. Hartwell LH. Cell division from a genetic perspective. J Cell Biol 1978;77:627–637. Hartwell LH, Weinert T. Checkpoints: controls that ensure the order of cell cycle events. Science 1989;246:629–634. Holzman D. Mismatch repair genes matched to several new roles in cancer. J Natl Can Inst 1996;88:950,951. Hunter T. Oncoprotein networks. Cell 1997;88:333–346. Kastan MB. Molecular biology of cancer: the cell cycle. In: DeVita VT, Hellman S, Rosenberg, SA, eds. Principles and Practice of Oncology. Philadelphia, PA: Lippincott-Raven, Philadelphia, 1997; pp. 121–134. King RW, Deshaies RJ, Peters J-M, Kirschner MW. How proteolysis drives the cell cycle. Science 1996;274:1652–1659. Ko LJ, Prives C. p53: puzzle and paradigm. Gen Dev 1996;10:1054–1072. Levine AJ. p53, the cellular gatekeeper for growth and division. Cell 1997;88:323–331. Lewin B. Genes V. Oxford, UK: Oxford University Press, 1994. McIntosh JR, Koonce MP. Mitosis. Science 1989;246:622–628. Nasmyth K. Viewpoint: putting the cell cycle in order. Science 1996;274: 1643–1645. Nurse P, Bisset Y. Gene required in G1 for commitment to cell cycle and in G2 for control of mitosis in fission yeast. Nature 1981;292:558–560. Paulovich AG, Toczyski DP, Hartwell LH. When checkpoints fail. Cell 1997;88:315–321. Pines J. Cyclins and their associated cyclin-dependent kinases in the human cell cycle. Signaling from the plasma membrane to the nucleus. Biochem Soc Trans 1993;21:921–925. Sherr CJ. Cancer cell cycles. Science. 1996;274:1672–1677. Stillman B. Cell cycle control of DNA replication. Science. 1996;274: 1659–1664. Weinberg RA. The retinoblastoma protein and cell cycle control. Cell 1996;81:323–330.

CHAPTER 7 / ONCOGENES AND TUMOR SUPPRESSOR GENES

7

73

Oncogenes and Tumor Suppressor Genes J. LARRY JAMESON

DEFINITION AND DISCOVERY OF ONCOGENES AND TUMOR SUPPRESSOR GENES

of normal cellular genes, which are referred to as proto-oncogenes. The ability of retroviruses to “reverse transcribe” mRNA accounts for their ability to capture cellular gene products. Analyses of such viral oncogenes has helped to identify many critical cellular genes including src, ras, raf, kit, jun, fos, ets, and others. DNA tumor viruses have also played an important role in our current understanding of neoplasia. In this case, the viruses produce proteins that target key cellular regulatory proteins, such as retinoblastoma (Rb) and p53. For example, the SV40 large T antigen associates with, and inactivates, Rb. Remarkably, other viral products including adenovirus E1A and papilloma virus protein E7 also target Rb. Although these viral proteins are structurally unrelated, they each appear to associate with the so-called pocket of Rb to prevent interactions with cellular transcription factors like E2Fs, which are involved in control of S-phase genes (Fig. 7-2). These cellular targets of DNA tumor viruses are now known to be tumor suppressor genes. Historically, the presence of tumor suppressor genes was suspected, based on several lines of evidence. The malignant phenotype of certain tumors could be suppressed by fusion with normal cells, implying the presence of a suppressor gene in the normal cell. Chromosomal losses in the hybrids eventually resulted in reversion back to the malignant phenotype. Ultimately, it was possible to observe partial alterations in tumorigenicity by the introduction of single chromosomes into the malignant cells. For example, insertion of chromosome 11 (now known to harbor the WT-1 gene) was sufficient to suppress tumorigenicity in a Wilm’s tumor cell line. These laboratory experiments were paralleled by observations in humans that certain hereditary forms of cancer are transmitted in an autosomal dominant manner (see below). Based on the agedependent appearance of retinoblastoma, Knudson postulated a “two-hit” model for the disorder (Fig. 7-3). In this model, the first “hit” or mutation is inherited in the germline and the second mutation is acquired as a somatic event involving either the remaining normal allele or another gene. As described below, the second hit typically involves the remaining normal allele. This model, which has proven to explain a number of hereditary cancer syndromes, implied a loss of function, and was in a sense, recessive. The identification and characterization of the Rb gene has converted this statistical model into concrete mechanisms (see Chapter 105). Cytogenetic studies in patients with inherited forms of retinoblastoma revealed deletions of chromosome 13q14. Additional studies showed that tumors from patients with retinoblas-

Cancer is caused by an accumulation of genetic alterations that confer a survival advantage to the neoplastic cell. These genetic changes can affect multiple facets of cellular function, including an increased rate of cell proliferation, resistance to apoptosis, altered tissue invasiveness, production of growth and angiogenic factors, and the ability to escape immune surveillance. Different cancers reflect these features to varying degrees, depending on the nature of their cellular functions and genetic changes. The genetic basis for cancer is reflected in the clonal nature of neoplastic cells. As depicted in Fig. 7-1, cell proliferation can involve either polyclonal or monoclonal expansion of the cell population. Polyclonal growth or hyperplasia reflects the response to an extrinsic growth factor or to an intrinsic genetic alteration that is shared by all (e.g, MEN-2, germline ret mutation) or some (e.g., McCune-Albright, postzygotic somatic Gsα mutation) of the cells. Hyperplastic cells may subsequently acquire one or more somatic mutations and develop clonal derivatives. Monoclonal growth reflects the acquisition of somatic mutations that confer a survival advantage. Multistep models of tumorigenesis (see below) postulate that multiple different mutations are acquired over time. Thus, an initial clone would give rise to additional clonal variants as they acquire distinct mutations, some of which may foster tumor growth and expansion by favoring tissue invasion or metastasis. The recognition that cancer is a genetic disease has led to an intensive effort to characterize genetic alterations in tumors and to understand how these genes function in the context of the normal and neoplastic cell. The idea that cancer might be caused by genetic changes has long been suspected by clinicians who noted that certain cancers tend to “run in families.” Several landmark advances in virology, cytogenetics, and molecular biology have converted this intuitive assessment into a firm scientific foundation. In the early 1900s, Rous demonstrated transmission of sarcoma in chickens by using cell-free filtrates that contained a virus now known to be a retrovirus, Rous sarcoma virus (RSV). Identification of the transforming principle in the virus revealed that it harbored an altered form a normal cellular gene, src. The viral gene product was referred to as an oncogene, derived from the Greek, onkos, which refers to a mass or tumor. Subsequent studies have shown that many viral oncogenes correspond to altered versions From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

73

74

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Figure 7-1 Clonal expansion of cancer cells. Growth responses can be divided into hyperplasic (polyclonal) or monoclonal expansion of cells. Hyperplasia reflects the actions of growth factors that increase the proliferation of a population of cells. In contrast, monoclonal expansion reflects increased proliferation from a single ancestral cell, reflecting a growth advantage conferred by a somatic mutation. Monoclonal and polyclonal tumors can be distinguished based on patterns of X-chromosomal inactivation (in females) or based on the presence of a characteristic somatic mutation in a monoclonal tumor. Hyperplasia may predispose cells to develop somatic mutations, allowing a monoclonal population to emerge. Additional mutations in monoclonal tumors allows new clonal populations with increased rates of growth or invasiveness to emerge. Adapted with permission. (Jameson JL. Principles of hormone action. In: Weatherall DJ, Ledingham JGG, Warrell DA, eds. Oxford Textbook of Medicine, 3rd ed. Oxford, U.K.: Oxford Medical, 1996; pp.1553–1573.)

toma frequently showed loss of heterozygosity (LOH) in this region of chromosome 13, and that the remaining abnormal allele was transmitted from the affected parent. These experiments confirmed the loss of function hypothesis and established the importance of using loss of heterozygosity as a means for detecting tumor suppressor genes. The cloning of the Rb gene allowed demonstration of germline mutations in patients who did not have deletions. In addition, it was shown that patients with sporadic forms of retinoblastoma had acquired two somatic hits in which each allele received an independent mutation. The Rb protein plays a pivotal role in control of the cell cycle. As depicted in Fig. 7-2, Rb helps to govern the restriction point between G1 and S phase (see Chapter 6). Hypophosphorylated Rb binds transcription factors, including E2Fs. Upon Rb phosphorylation by Cyclin-CDK complexes, E2Fs (particularly E2Fs 1–3) dissociate from Rb to stimulate the transcription of genes that are involved in DNA synthesis. The loss of both Rb genes is therefore predicted to enhance cell proliferation. Because Rb is expressed ubiquitously, it is unclear why only certain tumors (retinoblastomas, osteosarcomas, soft tissue sarcomas, melanomas) develop with increased frequency in patients with germline Rb mutations. Cytogenetics have played an important role in the characterization of genetic alterations, particularly in hematologic malignancies (see Chapters 26 and 27). Characteristic chromosomal rearrangements suggested the presence of an oncogene. For example, in chronic myelogenous leukemia (CML), the description of the Philadelphia chromosome t(9;22)(q34;q11) was recognized early as a characteristic cytogenetic abnormality. This

Figure 7-2 The pivotal role of the Rb tumor suppressor gene in control of the cell cycle. The cell cycle is depicted schematically with its characteristic stages (G0, M, G1, G2, S). G0 represents a quiescent state in which cells exit or enter the cell cycle. A key restriction point is noted in G1. After passage through this point, cell are committed to progress into the S phase. An additional checkpoint is noted in G2 before cells initiate mitosis. The Rb gene product plays a key role in progression through the restriction point. In its hypophosphorylated state, Rb binds E2F transcription factors, making them inaccessible for transcription of S-phase genes. Increasing amounts of cyclin D1 in association with cdk4 or cdk6 result in hyperphosphorylation of Rb. Cyclin E, in association with cdk2, also contributes to Rb phosphorylation. Phosphorylated Rb no longer associates with E2Fs, freeing them to interact with target genes (e.g., DNA polymerase, thymidine kinase, dihydrofolate reductase) that are involved in S phase. p16 is capable of blocking the actions of Cyclin D1/cdk 4,6, and this may account for its role as a tumor suppressor in some malignancies, as the absence of p16 would favor unrestricted progression through the cell cycle. A possible role for p53 is also shown. Cell injury can stimulate p53 levels leading to increased p21, a probable inhibitor of the actions of Cyclins D and E. Thus, the absence of p53 might prevent restriction of cell division in damaged cells. Viral gene products (E1a, large TAg, E7) bind Rb and dissociate E2Fs, therefore having similar effects as Rb hyperphosphorylation.

translocation is now known to fuse the bcr gene on chromosome 22 with the c-abl gene on chromosome 9. The fusion protein combines the tyrosine kinase activity of c-abl with abnormal regulation and cellular localization conferred by bcr. Another striking example involves a translocation t(8;14)(q24;q32) of the IgH chain gene with c-myc in Burkitt’s lymphoma. In this case, the regulatory regions of the immunoglobulin gene result in the overexpression of c-myc in lymphoid cells, causing increased cell proliferation. A number of such translocations and chromosomal inversions have now been described (Table 7-1). In general, they involve either abnormal expression of a cellular kinase or transcription factor or, the production of a fusion protein with altered functional properties.

CLASSES AND FUNCTIONS OF ONCOGENES GROWTH FACTORS Oncogenes can be categorized according to their cellular functions (Fig. 7-4). A few oncogenes function as growth factors (Table 7-2). For example, v-sis is structurally related to β chain of platelet-derived growth factor (PDGF), and int-2 corresponds to members of the fibroblast growth factor (FGF) family. Although these oncogenes do not play major roles

75

CHAPTER 7 / ONCOGENES AND TUMOR SUPPRESSOR GENES

Table 7-1 Examples of Oncogene Rearrangements and Fusions Type of rearrangement

Rearranged oncogene(s)

Gene activation

IgH-c-MYC IgH-Cyclin D1 TCR-β-TAL-2 TCR-α-HOX11 PTH-Cyclin D1 ELE1-RET

Disorder Burkitt’s lymphoma Mantle cell lymphoma T-cell acute lymphoblastic leukemia T-cell acute lymphoblastic leukemia Parathyroid adenomas Papillary thyroid cancer

Gene fusions BCR-ABL AML-1-CBFα FUS-CHOP EWS-FLI-1 PML-RARα

Figure 7-3 Two-hit model for retinoblastoma inactivation. The principles involved in the “two-hit” model are illustrated schematically. The Rb gene is located on chromosome 13q14. The Maternal (M) and Paternal (P) chromosomes are denoted as A or B. The first hit involves a germline mutation (depicted by an X) on chromosome MB. Subsequently, several different types of somatic second hits can occur. A second point mutation can inactivate the normal allele inherited from the father (PA). Deletions of part of chromosome 13, internal deletions, or translocations can also lead to inactivation of the normal Rb gene. Nondisjunction with loss of the normal chromosome 13 can cause loss of the normal Rb gene. The germline mutation causes a predisposition to retinoblastoma, as any inactivating second hit will be selected because of clonal proliferation. An analogous series of events can occur in sporadic tumors. In this case, both the first and second hits occur somatically.

in human cancer, they illustrate the importance of growth factors and their signaling pathways for the growth of tumors. A more prominent functional role for growth factors stems from their overproduction as a consequence of the actions of other oncogenes. Many tumors produce growth factors, such as insulinlike growth factor-1 (IGF-1), IGF-2, epidermal growth factor (EGF), and tumor growth factor-α (TGF-α), which stimulate cell growth in an autocrine manner. The mechanisms that lead to growth factor overexpression are relatively obscure, but account in part for the ability of transformed cells to grow in the presence of reduced serum and for the ability of conditioned media from transformed cells to foster the growth of other cell lines. As depicted in Fig. 7-2, growth factors act in the cell cycle to enhance the entry from the quiescent state (G0) and they stimulate progression to the restriction point in G1. Other growth factors, such as TGF-β, can oppose entry into the cell cycle, in part by increasing levels of cyclin dependent (cdk) inhibitors. Thus, the balance of growth factors and the activity of their receptors can have profound effects on cell proliferation. Growth factors also influence

Chronic myelogenous leukemia Acute myelogenous leukemia Myxoid liposarcomas Ewing’s sarcoma Promyelocytic leukemia

Figure 7-4 Sites of cellular actions of oncogenes and tumor suppressor genes. Oncogenes act at multiple sites of cellular function. (1) The extracellular environment can influence cell function by the overproduction of various growth factors. Not shown are important extracellular effects of angiogenic factors and proteolytic enzymes. (2) At the level of the cell membrane, adhesion molecules such as DCC and APC influence tumor invasiveness. Membrane receptors can be constitutively activated (e.g., Ret, TSH-R) or the receptors can be amplified (e.g., ErbB-2). (3) Ras is a prototypical example of signal transduction molecule that stimulates proliferation when activated. (4) Nuclear proteins include a number of targets for oncogenesis. Proteins such as AP-1 (Jun/Fos) and Ets are targets for several of the transducing signal transduction cascades, including Src, Ras, and ERKs. Some nuclear proteins, such as Myc, are amplified or overexpressed (Cyclin D1) in certain tumors. p53 plays a role in controlling the cellular response to injury (inhibits G1/S) and helps to direct apoptosis. Several DNA repair enzymes cause familial adenomatous polyposis when mutated. Adapted with permission. (Jameson JL. Principles of hormone action. In: Weatherall DJ, Ledingham JGG, Warrell DA, eds. Oxford Textbook of Medicine, 3rd ed. Oxford, U.K.: Oxford Medical, Oxford, 1996; pp.1553–1573.)

the initiation of apoptosis (programmed cell death). Withdrawal of growth factors (or other nutrients) favors apoptosis, and their presence helps to prevent it.

76

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Table 7-2 Functional Classes of Oncogenes and Tumor Suppressor Genes Class Growth factors

Oncogene sis int-2

Receptor tyrosine kinases

erbB-2 ret

Non-receptor tyrosine kinases

abl src

Cell surface proteins

DCC APC

GTPases

ras

Protein kinases

Gsα mos raf

Transcription factors

jun myc

Cell-cycle factors

Rb p53

DNA repair

ATM

hMSH2

Programmed cell death

Bcl-2

Function Homologous to the β chain of platelet-derived growth factor (PDGF) Integration site for MMTV; fibroblast growth factor-3,4 (FGF-3,4) genes Orphan receptor related to the epidermal growth factor receptor (EGF-R) Activated form of tyrosine kinase; binds glial derived nerve growth factor Translocates to bcr; interacts with cell cycle proteins Membrane associated tyrosine kinase; activates mitogenic signaling cascades Deleted in colorectal cancer gene involved in cell adhesion Adenomatous polyposis gene that interacts with β-catenin and E-cadherin Activates signaling cascades including MAPKs/ERKs Activates adenylyl cyclase A MEK-like enzyme Serine kinase in the ras pathway B-Zip protein that complexes with fos to form AP-1 bHLH protein that regulates cell cycle factors Retinoblastoma undergoes phosphorylation to regulate cell cycle progression Transcription factor that regulates cell-cycle genes and apoptosis Ataxia-telangiectasia gene involved in cell-cycle checkpoints and p53 directed apoptosis Human mutS homolog involved in DNA mis-match repair Blocks apoptosis; interacts with other cell proteins such as Bax to control apoptosis

The functional effects of growth factors on tumor development are emphasized by several transgenic models. Disruption of the inhibin α-subunit gene results in ovarian granulosa cell tumors with nearly complete penetrance (see Chapter 10). Inhibin normally combines with a β-subunit of activin and antagonizes the action of the activin homodimer. Activin is a member of the TGF-β family of growth factors that exerts pleomorphic growth effects in

Figure 7-5 The cycle of ras activation and effect of mutations. Ras is a guanine-nucleotide protein. When GTP is bound, Ras is active, and it is normally inactivated by hydrolysis of GTP to GDP, resulting in an inactive state. SOS (son of sevenless), also called guanine-nucleotide releasing protein (GNRP), regulates Ras activity and responds to input from the SH2 domain proteins, GRB-2 and SHC. The catalytic activity of Ras is also modulated by GAP (GTPase activating protein). The NF-1 gene (causes Von Recklinghausen’s neurofibromatosis) is homologous to gap. Mutations in ras prevent GTP hydrolysis, resulting in constitutive activation. Ras participates in several signaling pathways, including activation of Raf, MEK, and ERKs, which stimulate mitogenesis.

various tissues, including the ovary. In this model, the loss of inhibin is thought to result in unopposed actions of activin, which induce ovarian and adrenal tumors. Crossbreeding the inhibin-deficient mice with p53-deficient mice greatly accelerates the development of tumors. This model illustrates the profound effect of unbalanced growth factor action and its ability to act in a tissueselective manner. Another informative model involves disruption of the gene encoding the IGF-1 receptor. Fibroblast cells derived from these mice are resistant to transformation by SV40 large T antigen and/or activated ras. Reintroduction of the receptor confers the ability to undergo transformation, emphasizing the role of the IGF-1 pathway in cell proliferation and transformation. Growth factors act by initiating a variety of signaling cascades that involve other oncogenes. These signal transduction pathways are discussed further in Chapter 46. EGF, for example, activates a tyrosine kinase receptor. Ligand-induced homodimerization of the EGF-receptor results in autophosphorylation that allows the recruitment of SH2 domain-containing adaptor proteins such as Grb-2. Grb-2, in turn, associates with the guanine-nucleotide exchange protein, son of sevenless (SOS), through an SH3 domain to enhance the activation of Ras (Fig. 7-5). Ras plays a pivotal role in growth factor signaling. It can also be activated by mitogenic Giα-coupled seven transmembrane receptors, probably through the release of βγ-subunits and by activation of src kinase. Activation of focal adhesion kinase (FAK) by integrins can also lead to

CHAPTER 7 / ONCOGENES AND TUMOR SUPPRESSOR GENES

Ras activation. Ras initiates a mitogen-activated kinase (MAPK) cascade that includes Raf, MEK, and ERKs (extracellular regulated kinases). These kinases act on numerous cellular targets including the cytoskeleton and activation of transcription factors such as AP-1 (c-Jun/c-Fos) and Ets-1. c-jun and c-fos, which are themselves early response genes, form the AP-1 complex that activates a variety of genes involved in cell growth control. RECEPTOR TYROSINE KINASES Growth factor receptors not only serve to transduce the signals of extracellular growth factors, but they are also important targets of mutations or altered expression that contribute to tumorigenesis. The tyrosine kinase class of receptors includes oncogenes such as ret (a receptor tyrosine kinase that binds glial derived nerve growth factor), trk (nerve growth factor class), kit (steel receptor), fms (colony stimulating 1 receptor), erbB (epidermal growth factor, EGF, receptor), and erbB-2/neu (EGF-related receptor). As noted above, this group of receptors are thought to act primarily by signaling through the ras-MAPK cascade. In some cases, such as erbB and erbB-2/neu, receptor amplification leads to marked overexpression. When amplification involves deletion of the ligand binding domain, constitutive activation can result. Amplification of these receptors is particularly common in head and neck cancers, esophageal and gastric cancers, and in breast cancer. Point mutations can also activate tyrosine kinase receptors. The ret receptor is a prototype for activation by point mutations that cause constitutive activation of the receptor. Mutations in ret cause multiple endocrine neoplasia type 2 (MEN-2) (see Chapter 54). The MEN-2 mutations are clustered in a group of cysteines that are thought to form disulfide bonds in the extracellular ligand-binding domain of the receptor. MEN-2 is transmitted in an autosomal dominant manner. Individuals who inherit a mutant ret receptor exhibit hyperplasia of the calcitonin-producing medullary thyroid cells at a very early age. Subsequently, they have a high incidence of developing medullary thyroid cancer. Presumably, the development of medullary thyroid cancer involves a second hit, although the target(s) of subsequent mutations are not known. In practical terms, it is now recommended that patients who inherit ret mutations undergo prophylactic thyroidectomy and receive screening for additional features of MEN-2 (pheochromocytoma and hyperparathyroidism). It is notable that the action of the ret oncogene is conceptually distinct from that described above for Rb, even though both are transmitted in an autosomal dominant manner. As a tumor suppressor gene, Rb requires inactivation of both alleles, whereas the mutant form of ret initiates the process of tumorigenesis directly. Interestingly, mutations in the tyrosine kinase domain of ret cause a phenotypically distinct disorder, MEN-2b, which differs by including mucosal neuromas. Ret also plays a role in papillary cancer of the thyroid. In this case, chromosomal inversions (inv[10]) bring the tyrosine kinase domain of ret under the control of one of several fusion genes (e.g., ELE-1). This recombination leads to overexpression of the activated form of ret in the thyroid cell (where it is not normally expressed). Further expanding the role of ret in human disease is the observation that inactivating mutations cause Hirschprung’s disease. SIGNAL TRANSDUCTION The role of signal transduction molecules as oncogenes is exemplified by ras. There are three ras genes (N-ras, H-ras, K-ras) that were initially identified as transforming rodent sarcoma viruses. Mutant forms of ras were later identified as the transforming principle in DNA derived from sev-

77

eral types of human tumors. Ras is now recognized as one of the most common genetic alterations in human cancers. Ras mutations occur in 90% of certain cancers (e.g., pancreatic) and in 10–30% many other types of cancer including leukemias, melanomas, thyroid cancer, colon cancer, and lung cancer. Although the ras gene is amplified in some tumors, more commonly it is activated by point mutations. The biochemical basis of ras mutations is now well understood in terms of protein structure and activity. Ras activity is controlled by the binding and hydrolysis of GTP. The cycle of Ras activation and inactivation is depicted in Fig. 7-5. When GTP is bound, Ras is active, and it is inactivated by hydrolysis of GTP to GDP. Extensive structural, mutagenesis, and biochemical analyses have revealed that ras mutations act by preventing GTP hydrolysis, retaining the protein in an active state. The locations of ras mutations are clustered in amino acids 12, 13, and 61, each of which has been shown to play a critical role in nucleotide binding and/or hydrolysis. The inability to inactivate Ras leads to stimulation of the MAP kinase signal transduction cascade and uncontrolled cell proliferation. Ras catalysis of GTP is regulated by SOS, also called guaninenucleotide releasing protein (GNRP), which responds to input from the SH2 domain-containing proteins, GRB-2 and SHC. In this manner, Ras activity can be activated by numerous growth factor pathways (Fig. 7-6). The catalytic activity of Ras is also modulated by GAP (GTPase activating protein). The NF-1 gene (which causes Von Recklinghausen’s neurofibromatosis type 1) is homologous to GAP. Inactivating mutations of both NF-1 alleles are predicted to result in diminished GTPase activity and increased activity of Ras or other related proteins. Thus, NF-1 functions like a tumor suppressor gene. Ras is tethered to the cytoplasmic side of the cell membrane by farnesylation. Mutations that eliminate the farnesylation site in the carboxyterminus of Ras eliminate its signaling function, presumably because of a requirement for subcellular localization to contact other components of the signal transduction cascade. Efforts to inhibit Ras farnesylation represent a promising therapeutic strategy for blocking the actions of activated forms of ras. The guanine nucleotide-binding G proteins are structurally related to Ras. Moreover, homologous mutations have been described in Gsα, Giα2, and Giα3. In the case of Gsα, the mutations occur in codons 201 and 227, and also prevent GTP hydrolysis, resulting in constitutive activation of the adenylyl cyclase pathway. Gsα mutations that develop postzygotically cause McCune-Albright syndrome, which is characterized by polyostotic fibrous dysplasia, café au lait spots, hyperfunctioning ovarian cysts, and autonomous function in other glands, including the thyroid and pituitary glands (see Chapters 46 and 59). The development of these lesions reveals tissues in which the cAMP pathway is mitogenic. Somatic mutations in Gsα occur in about 30% of GH-secreting pituitary tumors (see Chapter 48) and can cause autonomously functioning thyroid nodules. It is notable that mutations have also been identified at other steps in the cAMP signaling pathway. Activating mutations have been described in the thyroid-stimulating hormone (TSH) receptor. These mutations, which appear to cause hormone-independent coupling to Gsα, cause congenital hyperthyroidism and nodular goiter when transmitted in the germline or autonomous nodules when acquired as somatic mutations (see Chapter 50).

78

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

products that result from translocations, inversions, or retroviral insertions (Table 7-1). These rearrangements can activate transcription factors in inappropriate tissues, at the wrong stage of cellular development, or at increased levels (see Chapter 3). In some cases (e.g., PML-RARα), the transcription factor prevents normal cellular differentiation and allows cells to retain responsiveness to proliferative signals. In other cases (e.g., IgH-Myc), overexpression of Myc favors cell proliferation by enhancing entry into the cell cycle. A number of the signal transduction pathways discussed above converge on the transcription factor proto-oncogenes. c-Jun and c-Fos form the transcription factor, AP-1, which is a member of the B-Zip family of transcription factors that dimerize via a leucine zipper motif. Reflecting its dimeric structure, AP-1 recognizes variations of a palindromic DNA element, TGAGTCA, that resides in the regulatory regions of target genes. Numerous genes contain AP-1 sites, although the genes involved in v-Jun-mediated transformation remain unknown. Jun is one of the transcription factor targets for the Ras signaling pathway. Amino-terminal phosphorylation of Jun by Jun N-terminal kinases (JNKs) and stress-activated protein kinases (SAPKs) enhances its ability to transactivate genes.

CLASSES AND FUNCTIONS OF TUMOR SUPPRESSOR GENES

Figure 7-6 Growth factor signaling pathways. Representative growth factor receptors are shown including a cytokine receptor, the insulinlike growth factor-1 (IGF-1) receptor, a tyrosine kinase receptor (e.g., EGF), and a G-protein-coupled receptor (different members can signal via Giα or Gsα). Each of these receptors can activate an array of signaling pathways, not all of which are shown, and the central role of ras is emphasized. Cytokine receptor signaling may also interface with the Ras cascade, but most prominently activates the STAT (signal transducers and activators of transcription) proteins, which, during phosphorylation by Janus kinase (JAK), translocate to the nucleus to activate transcription of target genes. Tyrosine kinases induce the binding of an SH2 domain protein Grb2, which binds to guanine nucleotide exchange factor, SOS. This pathway can also be activated by the IGF-I receptor via insulin related substrate (IRS). Gsα-coupled receptors act primarily through the cAMP-PKA (protein kinase A) pathway, which is mitogenic in some cell types. Giα-coupled receptors can activate the Ras pathway through βγ-subunits, likely through src kinase. Ras activates several downstream kinases in a cell type specific manner. The preferential activation of a particular mitogen activated protein kinase (MAPK) module, such as activation of the extracellular-regulated kinase (ERK) rather than the stress-activated protein kinase (SAPK) pathway provides a mechanism for specific responses. There are many targets of these kinases, including a variety of transcription factors that are involved in cellular growth responses.

TRANSCRIPTION FACTORS Several of the oncogenic transcription factors were initially identified based on their association with transforming viruses (e.g., Jun, Fos, Ets, Myb, Rel, Erb-A). Others were recognized because of their presence as fusion

Tumor suppressor genes are defined as genes that sustain loss of function in the development or progression of neoplasms. Thus, tumor suppressor genes are characterized by inactivating mutations, whether inherited in the germline or acquired by somatic mutation. Although the initial tumor suppressor genes like Rb are involved in cell-cycle control, the class of tumor suppressors has broadened to include genes with a variety of cellular functions. As described below, the products of tumor suppressor genes include proteins involved in transcriptional regulation (e.g., WT-1), cellcycle progression (e.g., p53, p16), RNA elongation, (e.g., VHL), signaling pathways (e.g., NF-1, DPC-4, Patched), and DNA repair (e.g., MSH-2). As a group, tumor suppressor genes function at key junctures in the control of cell proliferation, differentiation, apoptosis, and response to genetic damage. CELL-SURFACE PROTEINS Cell-surface proteins other than receptors are now recognized as important sites for oncogenesis. These factors function like tumor-suppressor genes in the sense that loss of function (by mutation or deletion) contributes to tumorigenesis. Examples in this class include DCC (deleted in colorectal cancer) and adenomatous polyposis coli (APC). DCC was recognized because of localization to a region of chromosome 18q that was frequently deleted in colorectal cancer (~70% of cases). The DCC gene encodes a large protein that is involved in cell adhesion. Loss of DCC also occurs in other types of tumors including brain, breast, and pancreatic cancers. The APC gene causes familial adenomatous polyposis (FAP) with a prevalence of about 1 in 7000 individuals. It was localized to chromosome 5q21 based on cytogenetic and linkage studies. The APC gene encodes a large protein that contains β-catenin binding sites that allow linkage to the cytoskeleton and E-cadherin. In this manner, APC probably plays a role in linking the extracellular matrix to intracellular signaling pathways, and it may play a role in apoptosis. Germline mutations in APC cause the development of multiple benign colonic polyps. Most mutations result in truncation of the protein and, because of the size of the gene,

CHAPTER 7 / ONCOGENES AND TUMOR SUPPRESSOR GENES

mutations can be screened using a protein truncation test. Tumorigenesis caused by APC mutations follows the two-hit model that charaterizes classic tumor suppressor genes. In FAP, neoplasia is initiated on inactivation of the remaining normal allele. APC mutations are also very common in sporadic colon cancers, which occasionally demonstrate inactivation of both APC alleles. CELL-CYCLE CONTROL PROTEINS Not unexpectedly, a number of tumor suppressor genes play a role in control of the cell cycle. Rb was the first of these to be recognized. As shown in Fig. 7-2, Rb acts primarily at the restriction point in G1 of the cell cycle. Early in G1, Rb is hypophosphorylated and binds transcription factors E2F 1–3. E2Fs dimerize with a family of DP proteins (DP 1–3). The E2F/DP complexes bind to a consensus sequence, TTTCGCGC, which is present in the promoters of many genes, including thymidine kinase, dihydrofolate reductase, c-myc, and E2F itself. When Rb is bound to E2F/DP complexes, it can function as a repressor. On the other hand, dissociated E2Fs are transcriptional activators. Overexpression of E2F appears to be sufficient to push cells through the restriction point in G1. Phosphorylation of Rb by cyclin D/cdk4,6 complexes results in dissociation of E2Fs from Rb. Several pathways can act to prevent Rb phosphorylation and inhibit E2F dissociation from Rb. In the setting of cell damage, p53 levels are increased, and induce the cdk inhibitor, p21, which inhibits the actions of cyclin E/cdk2, another activator of Rb. p16 can also inhibit cyclin D1/cdk4. In addition to Rb mutations in retinoblastoma, somatic mutations in Rb are relatively common in small cell lung cancer, sarcomas, and bladder cancer (see Chapters 40 and 105). Deletion of p16 has been observed in small cell lung cancers in which the Rb gene is expressed. The absence of p16 would be expected to diminish inhibition of Rb-mediated progression through G1. Expression of the papilloma virus E7 protein in cervical cancers functionally inactivates Rb. Overexpression of cyclin D1 occurs by amplification in some breast, esophageal, and squamous cell carcinomas. In addition, cyclin D1 is ectopically expressed as a result of translocations in some parathyroid tumors (PTH-cyclin D1) (see Chapter 51) and B-cell lymphomas (IgH-cyclin D1). Mutations of p53 and LOH at 17p occurs commonly in colorectal, lung, breast, and bladder cancers. Germline mutations of p53 are found in Li-Fraumeni syndrome, which is characterized by a greatly increased risk for multiple tumors including breast cancer, soft-tissue sarcomas, osteosarcomas, brain tumors, and leukemias. Although p53 is one of the most prevalent mutations in advanced stages of tumorigenesis, its cellular functions are only partly understood. p53 is capable of acting as a transcription factor and induces the expression of inhibitors (e.g., p21) of the G1/S transition. p53 also appears to govern entry into the apoptosis pathway (by stimulating bax) in response to genotoxic damage. Thus, the absence of p53 not only favors progression through the cell cycle, but it also allows cells with damaged DNA to avoid undergoing apoptosis. TRANSCRIPTION FACTORS The Wilms’ tumor (WT-1) suppressor gene causes renal tumors in children (see Chapter 71). The WT-1 locus on chromosome 11p13 was identified based on a contiguous gene syndrome, WAGR (Wilm’s, aniridia, genitourinary disorders, retardation). The WT-1 gene encodes a transcription factor that contains zinc finger domains and is expressed in the urogenital ridge and in the developing kidney. WT-1 appears to suppress gene expression, including the IGF-II gene, which is also overexpressed in Beckwith-Wiedeman syndrome (see Chapter

79

116). Mutations in WT-1 are found in only a minority of cases of Wilm’s tumor, suggesting that other loci are also involved in the pathogenesis of this tumor. Von Hippel-Lindau (VHL) syndrome includes retinal angiomas, predisposition to renal cancer, pheochromocytomas, hemangioblastomas of the central nervous system, and cysts in several tissues. The VHL gene was identified on chromosome 3p based on linkage. Tumors in patients with germline mutations of VHL show mutations or loss of the remaining normal VHL gene. Somatic mutations in both alleles of VHL are also found in the majority of clear cell renal cancers (see Chapter 71). VHL encodes a protein that binds to elongin and is involved in transcriptional elongation. The normal function of VHL appears to be inhibition of elongin activity. The BRCA 1 (breast cancer predisposition) and BRCA 2 genes are located on chromosomes 17q21 and 13q, respectively (see Chapter 63). The proteins are weakly homologous and likely function as transcription factors. Germline mutations predispose to early breast and ovarian cancers (especially BRCA 1). Although the BRCA 1 and 2 genes were identified in large pedigrees with early breast cancer, these genes appear to account for a relatively small fraction of increased breast cancer risk, even when there is an affected first-degree relative. Their functional roles and their importance for genetic testing and counseling remain to be determined. DNA REPAIR PROTEINS The role of DNA repair processes is most clearly illustrated in hereditary nonpolyposis colorectal cancer (HNPCC). Based primarily on studies by Lynch and colleagues, this form of hereditary colon cancer was strongly suspected to be caused by predisposing “cancer genes.” HNPCC is now known to be caused mutations in one of several different DNA repair enzymes (hMSH2, hMLH1, hPMS2). Initially, linkage analyses indicated association with chromosomes 2p16 or 3p21. However, unlike the situation with Rb, MEN-1, and other heritable tumor suppressor genes, tumor samples did not always exhibit loss of heterozygosity in the putative normal allele at these loci. Rather, new microsatellites (highly polymorphic repeat sequences, see Chapters 4 and 5) were detected, and these aberrations were present throughout the patient’s genome, suggesting microsatellite instability. This type of genomic instability is similar to that found in microorganisms with mutations in mismatch repair (proofreading) enzymes (e.g., mutS, mutL), spurring a search for human homologs of these genes. Germline mutations have now been identified in the human mismatch repair enzyme genes (hMSH2, hMLH1, hPMS2) in kindreds with HNPCC. Of these, mutations in hMSH2 and hMLH1 appear to account for the majority of HNPCC cases. Although HNPCC is relatively uncommon (2–4% of colorectal cancer), sporadic mutations in DNA repair enzymes may result in genomic instability in as many as 15% of colorectal cancers. The TGF-β receptor, which mediates apoptosis of colon cells, has been hypothesized as a target of inactivation by microsatellite instability. The mutated in ataxia telangiectasia (ATM) gene represents another example of a DNA repair enzyme defect. Heterozygotes who carry this gene appear to be at increased risk for breast cancer. The defective ATM gene product appears to cause impaired p53 responses to DNA damage. Thus, ATM may play an important role in cell cycle checkpoints for DNA damage. Xeroderma pigmentosa represents another example of defects in DNA repair. In this case, a mutation in one of several different genes involved

80

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

in nucleotide excision and repair leads to a high incidence of skin tumors, particularly in response to UV radiation (see Chapter 81). APOPTOSIS-RELATED PROTEINS Although cancer is frequently conceptualized in terms of cell proliferation, the role of abnormalities in cell death, or apoptosis, is now becoming more fully appreciated. Most malignancies demonstrate abnormalites in programmed cell death. In addition to contributing to the immortality of the replicating cell, alterations in the apoptotic pathway also favor the accumulation of additional genetic alterations that would normally induce cell death. Responses of tumors to certain types of chemotherapy or irradiation also appear to correlate with the ability of tumors to undergo apoptosis. Bcl-2 was initially discovered based on its translocation in follicular lymphomas. In this case, the recombination of the immunoglobulin locus on chromosome 14 with the Bcl-2 gene on chromosome 18 causes persistent overexpression of Bcl-2 in lymphoid cells. It was shown that the clonal expansion of lymphoid cells reflected increased survival rather than enhanced proliferation. Bcl-2 is a member of a family of proteins that regulate the apoptotic pathway. While Bcl-2 protects against cell death, others such as Bax, favor cell death. Bcl-2 and Bax can form homodimers or heterodimers, and the balance of Bcl-2 and Bax expression regulates the predisposition to cell death. As noted above, p53 also plays an important role in apoptosis. Since p53 mutations are found in up to 50% of human cancers, its role in cell-cycle checkpoints and in apoptosis may represent one of the most common lesions in neoplastic cells. The exact function of p53 in the apoptotic pathway has not been fully defined. Thymocytes from mice deficient in p53 fail to undergo apoptosis in response to DNA damage. However, glucocorticoids can still induce apoptosis in the absence of p53. Thus, p53 is not absolutely required for all pathways of cell death. It has been suggested that p53 may induce apoptosis primarily in the presence of DNA damage or when survival factors are limiting. It is notable that p53 can increase Bax expression, providing one mechanism for shifting the cell into an apoptotic pathway.

INHERITED CANCER SYNDROMES Several inherited cancer syndromes have already been described as examples of the functional roles of oncogenes (MEN-2), tumor suppressor genes (Rb), or DNA repair enzymes (HNPCC). Although many of these syndromes cause relatively rare disorders, they have been enormously important for the identification of cancer-related genes and for establishing principles of cellular abnormalities in neoplasia. In addition, many of the genes that cause inherited forms of cancer are frequent targets for somatic mutations in sporadic tumors. Thus, these syndromes often identify candidate oncogenes or tumor-suppressor genes. Some of the inherited cancer syndromes are listed in Table 7-3. Most of these disorders are discussed in greater detail in specific specialty sections of the book. In general, inherited forms of cancer are caused by a limited number of mechanisms. The classical two-hit model for neoplasia, in which one copy of the mutant gene is inherited and the second normal copy is subsequently lost (see Fig. 7-3), is illustrated by disorders such as retinoblastoma (Rb), Li-Fraumeni (p53), MEN-1 (MENEN), and Wilm’s tumor (WT-1). MEN-2 (Ret), and some of the signaling pathway disorders such as neurofibromatosis type 1 likely involve inherited stimulation of cell growth followed by the accumulation of additional mutations that lead to malignancy. Another group of disorders involve

Table 7-3 Inherited Cancer Syndromes Syndrome/cancer Retinoblastoma Li-Fraumeni Familial adenomatous polyposis Lynch’s syndrome (HNPCC) Von Hippel-Lindau

Oncogene Rb p53 APC

Cell cycle control Cell cycle; apoptosis Cell-cell interactions

MSH2, MLH1, PMS1,2 VHL

DNA repair

Von Recklinghausen’s NF-1 Neurofibromatosis NF-2 type 2 Wilms’ tumor WT-1 Ataxia-telangiectasia Nevoid basal cell cancer syndrome Multiple endocrine neoplasa type I Multiple endocrine neoplasa type II Xeroderma pigmentosa Bloom’s syndrome Breast cancer

Function

Transcriptional elongation GAP-like Ezrin-like

ATM Patched

Transcriptional repression DNA repair Transmembrane protein

MENIN

Tumor suppressor

Ret

Tyrosine kinase

XP A-G

DNA repair

BLM BRCA1, BRCA2

DNA helicase Tumor suppressors

abnormalities in DNA repair, predisposing to a high frequency of genetic alterations at numerous loci. This group includes genes that cause HNPCC, xeroderma pigmentosa, Bloom’s syndrome, and ataxia-telangiectasia.

MULTISTEP THEORY OF NEOPLASIA Pathologists have long recognized a spectrum of histologic abnormalities that correlate with different degrees of tumorigenicity. These observations have led to various grades of malignancy based on the degree of abnormal cellular morphology or to stages of invasiveness such as dysplasia, carcinoma in situ, locally invasive tumors, or metastatic lesions that have spread beyond their original site of growth. There is now strong evidence that these morphological and clinical characteristics of tumors can be explained in molecular terms. The process of “multistep carcinogenesis” has been studied in many different models. It is sometimes divided into stages of initiation, promotion, and progression. These categories are undoubtedly oversimplistic, and it is likely that different tumor types involve distinct pathways for progression. In addition to developing altered pathways for proliferation and escape from apoptosis, tumors require angiogenesis and escape from immune surveillance in order to continue clonal expansion. Early models for cellular transformation demonstrated cooperation among different classes of oncogenes such as ras and myc. However, analogous events do not hold true for all malignancies. Many hematologic malignancies harbor distinct translocations that are not found in other tumor types. In some of these cases, however, it remains true that progression to more aggressive tumors involves the acquisition of additional genetic abnormalities. In fact, surveys of most

CHAPTER 7 / ONCOGENES AND TUMOR SUPPRESSOR GENES

Figure 7-7 Example of multistep carcinogenesis involving thyroid carcinoma. There is evidence for multistep carcinogenesis for most cancers. In the case of thyroid cancer, various mutations affect the differentiated phenotype of the cancer as well as the degree of invasiveness. The normal thyroid follicular cell can be converted to an autonomously functioning “hot” nodule by two different types of mutations. Mutations in the TSH-receptor can activate Gsα and the cAMP signaling pathway, which is mitogenic in the thyroid. Gsα is a guanine nucleotide-binding protein that is structurally related to Ras. Gsα mutations that inhibit GTP hydrolysis also cause constitutive activation of the cAMP pathway. Follicular adenomas exhibit ras mutations (~30%) and, in the progression to follicular cancers, there is an increasing prevalence of loss of heterozygosity (LOH) at several distinct loci. Papillary cancers have a high frequency (20–40%) of a unique mutation caused by an inversion on chromosome 10 that brings the ret tyrosine kinase receptor under the control of several different genes. Ras mutations are found in papillary cancer with similar prevalence to follicular neoplasms. Mutations in p53 are relatively common in anaplastic cancer, most of which are thought to arise from follicular cancers. Adapted with permission. (Jameson JL. Applications of molecular biology in endocrinology. In: Degroot LJ, ed. Endocrinology, 3rd ed. Philadelphia, PA: WB Saunders, pp. 119-147.)

hematologic malignancies for genetic alterations reveal numerous examples of gene amplification or loss of heterozygosity. The concept of multistep carcinogenesis has been particularly well-studied for colon cancer by Vogelstein and colleagues. In colon cancer, it appears that seven or more genetic events (APC, ras, DCC, mismatch repair, TGF-b-R, p53, and others) may be required before an invasive carcinoma or metastases develop. Based on this list of genetic alterations, it is apparent that colon cancers can contain nearly every type of oncogene, tumorsuppressor gene, and DNA repair defect. Patients with germline APC mutations may first develop a second hit in this gene and subsequently develop additional growth promoting mutations such as mutations in ras or p53. However, the order in which these mutations occurs is variable in different patients. Another example of multistep carcinogenesis is depicted for thyroid cancer in Fig. 7-7. Thyroid cancer is appealing as a model because several mutations give rise to distinct tumor phenotypes and because some of the growth promoting factors have been wellcharacterized. Proliferation of the thyroid follicular cell is largely dependent on thyroid-stimulating hormone (TSH) and IGF-1 (see Chapter 50). TSH stimulates growth through a Gsα-coupled receptor that acts via the cAMP pathway, whereas IGF stimulates a tyrosine kinase receptor that activates IRS-1, PI-3 kinase, and the

81

Ras pathway. These pathways function synergistically to induce cellular proliferation. Activating mutations in the TSH-R, or in Gsα, cause constitutive activation of the cAMP pathway. These mutations cause clonal proliferation of a highly differentiated (benign) nodule that functions autonomously and secretes excess thyroid hormone. Although these mutations cause cellular proliferation, this mutation pathway is rarely observed in malignant thyroid cancers, which are not well-differentiated and rarely produce excess hormone. A group of inversions on chromosome 10 bring the ret oncogene under the control of various promoters. These fusion genes were initially termed papillary thyroid cancer (PTC) because they are essentially unique to this histologic variant. Approximately 20–30% of papillary thyroid cancers contain PTC rearrangements. Ras mutations (K-ras, N-ras, H-ras) have been found in all types of thyroid neoplasms, including benign macrofollicular adenomas, papillary cancers, and follicular cancers. As in most other cancers, the more malignant thyroid tumors show numerous examples of loss of heterozygosity, presumably involving tumor-suppressor gene loci. LOH is particularly common in follicular thyroid cancer, which has a greater tendency to metastasize. Anaplastic thyroid cancers have a very poor prognosis and show a marked increase in p53 mutations. Although it is unclear whether tumors show a gradual progression from ras mutations to loss of tumor suppressor loci, to p53 mutations, this seems likely, as anaplastic lesions are often noted to develop within a pre-existing follicular cancer. The loss of p53 may represent a pivotal step in which there is a defect in normal apoptosis pathways and a more rapid accumulation of additional genetic abnormalities. Superimposed on these genetic abnormalities, it is well-established that thyroid cancers respond to normal growth factors such as TSH, leading clinicians to suppress TSH with thyroid hormone as one means of therapy. This feature of the thyroid tumors emphasizes the overlap between normal and abnormal influences on cell growth. Thus, in the model of thyroid cancer, there are examples of mutations that are associated with specific tumor phenotypes (e.g., TSH-R, PTC), but other mutations are less specific (e.g., ras) for distinct tumor types. IMPLICATIONS FOR DIAGNOSIS AND TREATMENT Ideally, one would hope to “genotype” a tumor to establish a specific diagnosis and to aid in prognosis. In some cases, such as hematologic malignancies, specific translocations can be used for diagnostic purposes. Characteristic genetic changes can also be used to detect minimal residual disease using PCR or other sensitive techniques. However, for most malignancies, the array of genetic alterations is too large and the pattern of mutations is too variable to provide a diagnostic or prognostic code. It is possible in the future that the growing body of correlations between clinical course and genetic alterations will allow such analyses, at least for some tumors. In the short term, the greatest advances are likely to involve assessment of genetic risk based on inheritance of predisposing cancer genes. Thus, in kindreds with autosomal dominant MEN-2, it is now possible to determine early in life whether or not an individual carries the mutant ret gene. In conjunction with genetic counseling and clinical assessment, this has allowed careful monitoring for characteristic tumors (medullary thyroid cancer, pheochromocytoma, parathyroid adenomas) and, in some cases, prophylactic thyroidectomy to prevent thyroid cancer. In addition, unaffected family members can be spared the anxiety associated with repeated testing for possible malignancy. By analogy, earlier

82

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

and more careful screening procedures can be anticipated in affected individuals with inherited cancer syndromes such as FAP and HNPCC. In addition to issues of medical intervention, the assessment and detection of genetic risk carries an obligation for genetic counseling and raises numerous ethical issues involving health insurance, effects on the individual’s sense of self, and interpersonal relationships. For many disorders, there is an opportunity for prevention as well as enhanced surveillance. In patients with increased risk of colon cancer, changes in diet to reduce red meat and to increase fiber content are reasonable, and nonsteroidal anti-inflammatory agents may be protective. Patients with xeroderma pigmentosa have increased sensitivity to UV light and need to avoid or block UV exposure. In the long term, it is hoped that an improved understanding of the molecular basis of neoplasia will provide new avenues for therapy. The field is still very new and, given the usual timeline for drug discovery, it is not surprising that there are relatively few successes at this stage. The recognition that ras and p53 are commonly involved in many types of tumors has led to aggressive efforts to find agents that can alter their functions. In the case of Ras, it has been possible to inhibit farnesylation, which is required for the protein to function. Efforts are underway to search for drugs that might mimic some of the actions of p53. The recognition that the retinoic acid receptor was involved in a translocation in promyelocytic leukemia led to trials of its ligand, all trans-retinoic acid, which has proven to differentiate the tumor cells. Glucocorticoids induce apoptosis in many lymphoid malignancies, and antagonists of estrogens have been useful in hormonally responsive breast cancers. Further studies of the mechanisms of oncogene action and apoptosis may provide additional therapeutic strategies. Finally, gene therapy holds some promise for cancers. Because of the large number of abnormalities in most cancer cells and because of rapid clonal selection, it is unlikely to be practical to replace or to correct mutant genes. However, strategies to target toxic genes to malignant cells hold some promise. The primary challenge (as with many chemotherapeutic approaches) is to achieve specificity for the tumor and to target a very high fraction of the tumor cells. Several strategies are underway to augment immunologic responses (using cytokines) or to alter the drug sen-

sitivity of tumors. Attempts are also being made to genetically modify (e.g., with multidrug resistance genes) normal bone marrow constituents to make them more resistant to chemotherapy.

SELECTED REFERENCES Bale AE, Li FP. Principles of cancer management: cancer genetics. In: DeVita VT, Hillman S, Rosenberg SA, eds. Cancer: Principles and Practice of Oncology, 5th ed. Philadelphia, PA: Lippincott-Raven, 1997; pp. 285–294. Baserga R. Oncogenes and the strategy of growth factors. Cell 1994; 79:927–930. Cooper GM. Oncogenes, 2nd ed. Boston, MA: Jones and Bartlett, Boston, 1995. Fearon ER, Vogelstein B. Tumor suppressor and DNA repair gene defects in human cancer. In: Holland J, Frei E, eds. Cancer Medicine, 4th ed. Baltimore, MD: Williams and Wilkins, 1997; pp. 97–118. Haber D, Harlow E. Tumor-suppressor genes: evolving definitions in the genomic age. Nat Genet 1997;16:320–322. Hunter T. Oncoprotein networks. Cell 1997;88:333–346. Kastan MB. Molecular biology of cancer: the cell cycle. In: DeVita VT, Hillman S, Rosenberg SA, eds. Cancer: Principles and Practice of Oncology, 5th ed. Philadelphia, PA: Lippincott-Raven, 1997; pp. 121–134. Kinzler KW, Vogelstein B. Lessons from hereditary colorectal cancer. Cell 1996;87:159–170. Korsmeyer SJ. Regulators of cell death. Trends Genet 1995;11: 101–105. Levine AJ. p53, the cellular gatekeeper for growth and division. Cell 1997;88:323–331. Oltvai ZN, Korsmeyer SJ. Checkpoints of dueling dimers foil death wishes. Cell 1994;79:189–192. Perkins AS, Stern DF. Molecular biology of cancer: oncogenes. In: DeVita VT, Hillman S, Rosenberg SA, eds. Cancer: Principles and Practice of Oncology, 5th ed. Philadelphia, PA: Lippincott-Raven, 1997; pp. 79–102. Rabbitts TH. Chromosomal translocations in human cancer. Nature 1994;372:143–149. Roth JA, Cristiano RJ. Gene therapy for cancer: what have we done and where are we going? J Natl Cancer Inst 1997;89:21–39. Schichman SA, Croce CH. Oncogenes. In: Holland J, Frei E, eds. Cancer Medicine, 4th ed. Baltimore, MD: Williams and Wilkins, 1997; pp. 85–96. Sherr CJ. Cancer cell cycles. Science 1996;274:1672–1677. Weinberg RA. E2F and cell proliferation: a world turned upside down. Cell 1996;85:457–459.

CHAPTER 8 / MOLECULAR DIAGNOSTIC TESTING

8

83

Molecular Diagnostic Testing C. SUE RICHARDS AND PATRICIA A. WARD availability of gene sequence information and probes specific for the gene are essential for technology transfer from research to diagnostic laboratory. For the purpose of this limited discussion we will present examples of well-characterized mutation types represented in several common genetic diseases.

BACKGROUND Molecular diagnostic testing is a rapidly evolving field that currently includes analysis for single-gene disorders, multifactorial disorders, some forms of cancer, infectious disease, microbial epidemiology, and personal identification. The era of molecular medicine began in the late 1970s with the cloning of the β-globin gene and identification of the point mutation responsible for sicklecell anemia. At the same time, population-based carrier testing, using other technologies, for common genetic diseases such as Tay-Sachs and sickle-cell disease and prenatal diagnosis for chromosomal abnormalities were becoming widely available. The diagnostic capability of molecular technology has expanded tremendously, and there are now tests available for more than 300 genetic diseases being performed in greater than 200 diagnostic laboratories, academic and commercial, throughout the country (information provided by Helix; Seattle, Washington). The completion of the Human Genome Project will facilitate the identification of all disease genes, making a greater number of molecular diagnostic tests possible. While improved technological tools are essential for molecular diagnostics to move forward, the basic strategies for designing and implementing testing, which we describe in this chapter, will be long-lasting. To illustrate these principles, we will focus on the application of molecular diagnostic testing for single-gene disorders using DNA technology.

DELETION AND DUPLICATION MUTATIONS IN DUCHENNE MUSCULAR DYSTROPHY Duchenne muscular dystrophy (DMD) is an X-linked recessive neuromuscular disorder affecting approximately 1 in 3500 males with one-third of isolated cases resulting from a new mutation in the dystrophin gene. The dystrophin gene, located on Xp2l, is the largest gene identified to date, having greater than two megabases encompassing 79 exons. Two-thirds of patients have structural mutations with approximately 60% having intragenic deletion mutations and 6% duplication mutations. In addition, these mutations occur more frequently in certain regions of the gene reported as “hot spots for recombination.” This knowledge, gene sequence information, and the full-length cDNA dystrophin probe provide the tools needed to design a comprehensive and efficient strategy for molecular analysis of DMD. MULTIPLEX PCR DNA from affected males can be analyzed for the presence or absence of specified exons in the dystrophin gene, utilizing multiplex PCR reactions. Our laboratory examines 32 of the 79 dystrophin exons in three multiplex amplification reactions. These exons have been shown to account for over 98% of detectable deletions in the dystrophin gene. In addition, careful design of multiplexing strategy allows for establishment of deletion endpoints, which may be useful for phenotype–genotype correlations. Results from the multiplex reactions are analyzed for the presence or absence of specified exons, using agarose gel electrophoresis and ethidium bromide detection of DNA fragments (Fig. 8-1A). A positive result—i.e., the absence of a single PCR product or several gene contiguous PCR products from this multiplex—is virtually diagnostic of Duchenne or Becker muscular dystrophy. In contrast, a negative result—i.e., the presence of all exonic PCR products—neither confirms nor refutes the diagnosis. Dystrophin gene duplications can be problematic to detect by PCR analysis as band intensity can be modified by PCR conditions. Various approaches to overcome this technical challenge have included fluorescencelabeled primers and automated gene analyzer equipment to visualize fragment intensities compared to normal male and female control DNAs.

METHODS AND MUTATIONS DNA testing becomes feasible when the disease gene responsible is mapped or identified. If the gene has been mapped but not identified or if the mutations in a particular gene have not been characterized, then linkage analysis may be possible. If the gene has been cloned and mutations characterized, then it may be possible to perform direct mutation analysis. Direct mutation analysis is the method of choice, since it is more accurate and offers confirmation of diagnosis, in addition to precise carrier and fetal risk prediction. In contrast, linkage analysis cannot be used to confirm a diagnosis and provides a less accurate statistical prediction of carrier or fetal risk based on analysis of DNA markers surrounding a gene.

DIRECT MUTATION DETECTION Both the technology and the strategy used for mutation detection is determined by the nature of the specific mutation. The From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

83

84

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Figure 8-1 Dystrophin DNA analysis for deletion and duplication mutations. (A) Multiplex PCR analysis. For the two multiplex PCR panels shown, lanes represent: (a) a patient deleted in exons 46 through 48, and (b) a normal male control. Arrows point to deleted exons. (B) Southern analysis of HindIII fragments using cDNA probe 47-4B. Patient samples are in lanes b–k; male control (a), male patients (b–f); female patients (g–k); female control (l). Patient b is deleted for exons 44–47; patients c and d, deleted for exons 45–47; patient e, deleted for exon 45 only; patient f, no deletion detected. Carrier analysis of female relatives (lanes j and k) of affected male in lane c identifies the familial deletion (see Fig. 8-2).

SOUTHERN DELETION ANALYSIS For male probands with a negative result by multiplex PCR or in the absence of endpoint resolution, Southern analysis should be done to examine the entire dystrophin gene for deletions and duplications. In our laboratory we use eight cDNA probes that cover the entire dystrophin gene to probe HindIII-digested patient DNAs to examine for the presence or absence of HindIII-fragments and their intensity as compared to a known male control DNA (Fig. 8-1B). Carrier detection in females can also be performed using Southern and dosage analysis for deletions and duplications (Fig. 8-2). Dosage analysis is also used for identification of duplication mutations in male probands. Ideally, a male proband is analyzed first to determine whether an identifiable mutation is present. If a mutation is identified in the affected male, then females at risk are examined for the presence or absence of the specific mutation using scanning densitometry to determine gene dosage as compared to a known female control DNA. Whereas two copies of each dystrophin exon are normal for females, one copy indicates a gene deletion in a female, and three copies indicate a duplication. In the event that an

Figure 8-2 Dystrophin gene dosage analysis by scanning densitometry. Scans of Southern dystrophin analysis in Fig. 8-1B, control female (lane 1), and deletion carrier female (lane k). Arrows indicate deleted exons at approximately half gene dosage as compared to the normal female control.

affected male is unavailable for DNA analysis, females can undergo DNA testing for carrier determination. However, in the absence of a male proband, female mutation analysis using DNA dosage analysis is less reliable. Recently, deletion and duplication analysis by fluorescence in situ hybridization (FISH) has proven useful for diagnosis of female carriers of DMD when the familial deletion or duplication has been previously identified by DNA analysis. In the family shown in Fig. 8-3, the proband II-1 with a suspected diagnosis of DMD is analyzed by multiplex PCR and found to have a deletion of exons 45 through 50, confirming the diagnosis. Thus, molecular analysis of a blood specimen has avoided the invasive and costly procedure of muscle biopsy to confirm the diagnosis. The mother, I-2, who prior to DNA analysis has a 66% carrier risk, is analyzed by Southern deletion analysis and densitometry and found to carry the same deletion in one of her dystrophin genes. Similar analysis of the daughter, II-2, who has a prior carrier risk of 33%, does not identify this deletion mutation in either dystrophin gene. Thus her carrier risk is the same as a female from the general population, less than 1 in 2000. The male fetus, II-3, has a prior risk of 33% to be affected with DMD and is analyzed using multiplex PCR and does not have the familial

CHAPTER 8 / MOLECULAR DIAGNOSTIC TESTING

85

Figure 8-4 CF mutation analysis for ∆F508. Autoradiogram from exon 10 PCR products from appropriate controls and patient DNAs spotted on membranes and hybridized to the normal (F508) and mutant (∆F508) radiolabeled ASOs. For example, patient specimens in positions GI, E1, and E2 represent an affected child and both parents, respectively. Homozygote G1 shows hybridization only to ∆F508, while heterozygotes E1 and E2 hybridize to both F508 and ∆F508.

Figure 8-3 detection.

DMD risk modification following direct mutation

deletion mutation. Additional comparative studies using highly polymorphic DNA markers are performed on the maternal and fetal samples in order to rule out maternal cell contamination and demonstrate inheritance of one maternal allele. These results indicate that this fetus is unaffected with DMD. Accurate prenatal diagnosis is available for future pregnancies of I-2 and for her other female relatives at risk.

POINT MUTATION ANALYSIS IN CYSTIC FIBROSIS Cystic fibrosis (CF) is a common autosomal recessive disorder that, in US Caucasians of Northern European background, has an incidence of approximately 1 in 2000 and a carrier frequency of 1 in 25. With the cloning of the cystic fibrosis transmembrane regulator gene (CFTR) in 1989 and the subsequent identification of a number of common mutations, a DNA-based laboratory test was quickly put in place to provide more precise carrier detection and prenatal diagnosis for CF families. In addition, molecular analysis is helpful to confirm a CF diagnosis for some patients and is indicated for abnormal fetal ultrasound suggesting echogenic bowel. Recent guidelines recommend that potential gamete donors also have CF carrier screening performed, as appropriate to the donor’s ethnic background. Although debate in the genetics community has inhibited general population screening protocols, the general trend has been for progressive genetics centers to offer such testing. While hundreds of mutations have been identified, a detection rate of approximately 90% in US Caucasians can be achieved with the analysis of only 20–30 mutations. Carrier frequency and detection rate will vary, depending on ethnicity and geographical location of a specific population. Although various mutation detection methods exist, we will present the method used in our laboratory, allele-specific oligonucleotide (ASO) hybridization. This test uses multiplex PCR to analyze 10 exons and one intron in the CFTR gene for the presence or absence of 30 CF mutations. A partial test result illustrating the most common CF mutation, ∆F508, is shown in Fig. 8-4. This

Figure 8-5

CF modification following mutation analysis.

analysis can be partially automated using robotics, multiplexing, and pooling strategies. In addition, this strategy is applied in our laboratory to other genetic diseases, including Tay-Sachs, Gaucher, Canavan, Fanconi anemia, achondroplasia, sickle-cell and hemoglobin SC disease, and α-1-antitrypsin deficiency. The case shown in Fig. 8-5 illustrates the impact of molecular analysis on CF risk modification. All individuals are of Northern European Caucasian descent. Individuals II-2 and II-4 had a brother affected with CF who died prior to CF testing and the two carrier parents, I-1 and I-2, are also deceased. Thus, the prior carrier risk for II-2 and II-4 is two-thirds and their fetuses are both at 1 in 150 risk to be affected. CF testing reveals that II-2 is a CF carrier with one copy of a ∆F508 mutation. Individual II-1, the spouse of individual II-2, has a negative family history for CF; thus

86

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

his prior carrier risk is 1 in 25. His CF mutation analysis is negative. His modified carrier risk is 1 in 241. Thus, the modified CF risk for fetus III-1 is now 1 in 964 and prenatal diagnosis is not recommended. In contrast, individual II-4 does not have a ∆F508 or any of the other 29 CF mutations tested. The carrier risk for II-4 can be modified using Bayesian analysis. Based on detection of a ∆F508 mutation in II-2, the prior risk for II-4 is 1 in 2, since only one parental mutation has been identified. Based on a ~90% detection rate using this panel of 30 CF mutations, the prior risk of 50% can be conditionally modified using a 10% chance that a mutation will not be detected. Thus, II-4s modified carrier risk is 1 in 11. The reproductive partner, individual II-5, who has a negative family history of CF and a prior carrier risk of 1 in 25, is found by molecular analysis to carry a W1282X mutation, and thus, is a definite CF carrier. As a result, fetus III-2 is at a 1-in-44 risk to be affected with CF. This presents a complex prenatal diagnosis situation. If prenatal diagnosis is performed and the results are negative for W1282X, the fetus is unaffected with CF, but carrier risk is ambiguous at a 1-in-22. However, if the fetus inherits the W1282X mutation, then the fetus is a definite carrier and at 1 in 22 risk to be affected with CF.

TRINUCLEOTIDE REPEAT MUTATION AND METHYLATION ANALYSIS IN FRAGILE X MENTAL RETARDATION SYNDROME Fragile X syndrome was the first disorder in which the molecular basis was explained by the expansion of a trinucleotide repeat. Fragile X is the most common cause of inherited mental retardation, occurring in approximately 1 in 1500 males and 1 in 2500 females in all ethnic groups studied. Prior to the cloning of the FMR-1 gene, cytogenetic analysis was used for diagnostic confirmation. However, cytogenetic analysis proved unreliable, particularly for carrier detection and prenatal diagnosis. With the cloning of the FMR-1 gene and identification of a trinucleotide repeat expansion, the phenomenon of anticipation and reduced penetrance in fragile X families, the Sherman Paradox, was explained, and molecular analysis became the method of choice for fragile X testing. A CGG repeat occurs within the 5' untranslated region of the FMR-1 gene, which in normal individuals varies in size up to approximately 40 repeats and is inherited stably over generations. However, in affected individuals this CGG repeat region expands to greater than 200 repeats and becomes hypermethylated. In unaffected carriers, an intermediate size repeat region, termed as premutation, is seen ranging from approximately 60 to 200 repeats, but hypermethylation is not present. Premutation alleles are unstable during female meiosis and, depending on size (>70 repeats), have a significant likelihood of expanding to a full mutation in the next generation. However, unstable premutation alleles found in normal transmitting males generally do not expand to a full mutation in transmission to their carrier daughters. An indeterminate range between 40 and 60 repeats presents a challenge in counseling families regarding the potential risk of fragile X occurring in their future generations. Alleles in this range can be either normal stable alleles or unstable premutations, but have never been shown to be transmitted as full mutations. Fragile X testing is indicated for individuals with mental retardation or developmental delay of unknown cause, or individuals with a positive family history of fragile X. While general population carrier screening for fragile X carrier detection has been done primarily on an investigational basis, this application may become more widely accepted in the future.

Figure 8-6 Fragile X PCR analysis. Radiolabeled PCR products from patient DNA generated from primers flanking CGG repeat are separated by electrophoresis and analyzed by comparison to a standardized sizing ladder to determine allele size. An amplification control is included to rule out test failure in males with large expansions that fail to amplify by PCR. Normal alleles range up to 42 repeats; inconclusive, 43–59 repeats; premutation, 60 to approximately 200 repeats.

Figure 8-7 Fragile X Southern analysis. EcoRI- and BSSHIIrestricted DNAs from patients (b–k) and normal control (female, l and male, m) hybridized to radiolabeled pE5.1 probe isolated from the CGG repeat region of FMR-1. The normal EcoRI fragment size detected by this probe is 5.2 kb. In normal males additional digestion using the methylation-sensitive enzyme, BSSHII, results in a 2.8-kb fragment only, representing an active X chromosome. Normal females have one 5.2-kb fragment, due to normal X inactivation, and one 2.8kb fragment. Lanes b and e show affected males with full hypermethylated mutations; lane c and h, mosaic males having a full mutation and a premutation, respectively; lane d, a female with two premutations, the mother of affected male in lane e; lane f, a female with a full mutation; lane g, a female with a small premutation (68 repeats); lanes i, j, k, females having no detectable expansion.

Fragile X analysis is performed in our diagnostic laboratory using simultaneous PCR analysis to determine the size of normal and small premutation alleles (Fig. 8-6), and Southern analysis to examine methylation status and to estimate the size of large premutation and full mutation alleles (Fig. 8-7).

CHAPTER 8 / MOLECULAR DIAGNOSTIC TESTING

Figure 8-9

87

DMD Linkage.

The family shown in Fig. 8-8 illustrates the impact of direct mutation detection for fragile X for diagnostic clarification as well as carrier and fetal risk prediction. Individual II-1 reports to her obstetrician that she has a paternal uncle with mental retardation of unknown cause. DNA analysis of the uncle, I-1, reveals an expansion of approximately 500 CGG repeats, thus confirming his diagnosis as fragile X. Subsequent PCR analysis of her father, I-2, who is of normal intelligence, identifies a premutation allele of 80 repeats, thus indicating that he is a normal transmitting male. His daughter, II-1, is found to have one normal allele of 30 repeats and one premutation allele of 150 repeats, confirming her carrier status. Thus, her male fetus, III-1, is at a 1-in-2 risk to be affected with fragile X. PCR analysis of fetal DNA indicates a null allele, i.e., failure of amplification across the CGG region, suggesting the presence of a full mutation. Southern analysis of this fetus reveals a CGG expansion of >700 repeats with hypermethylation, thus predicting that this fetus is affected with fragile X mental retardation syndrome.

within the dystrophin gene can be used to rapidly accomplish this analysis. However, due to the large size and high mutation frequency within the dystrophin gene, intragenic recombinations do occur and must be accounted for in the risk analysis. We analyze DNA markers at the 5', central, and 3' regions of the dystrophin gene. Those females whose two-dystrophin marker patterns (haplotypes) can be distinguished are termed “informative.” If all markers are informative, a 1% chance of unrecognized recombination remains. If 5' and central, 3' and central, or only central markers are informative, a 5% chance for unrecognized recombination exists. If only 5' or 3' markers are informative, a 10% chance of recombination is used in the risk calculation. In this example shown in Fig. 8-9, the affected male, II-1, does not have an identified deletion or duplication mutation. Linkage analysis of the mother is informative at the 5' and central sites, but not at the 3' site. The sister of the affected and the male fetus has inherited the opposite maternal haplotype. The paternal haplotype of II-2 can be deduced from the data without the analysis of her father’s DNA. Based on the new mutation rate in DMD, I-2’s carrier risk is 67% and linkage analysis does not modify her risk. The carrier risk for individual II-2 based on family history is 33%; but, using linkage data and taking into account two possible recombination events with a 5% recombination frequency for each event, this risk can be modified to 6.4%. Similarly, the risk that the male fetus, II-3, is affected can also be modified from 33 to 6.4%, based on linkage analysis.

LINKAGE ANALYSIS

FUTURE DIRECTIONS

When direct mutation detection is not possible, either because the gene has not been cloned or a particular family does not have an identifiable mutation, then linkage analysis may be possible, provided that key family members are available and willing to be studied and that the necessary molecular tools are available for the analysis. For example, in the Duchenne muscular dystrophy case depicted in Fig. 8-3, if molecular analysis did not detect a mutation in the affected male, then linkage analysis in this family would be possible and could provide risk modification for the sister and the fetus. PCR analysis of highly polymorphic intronic DNA markers

With the expansion of molecular diagnostics will come improved health care. To date, the major impact of molecular analysis has been reproductive decision making regarding single gene disorders, which occur in a small percentage of the population. Future trends in genetic testing will include more predictive tests for susceptibility to common diseases, including cancer, cardiovascular disease, diabetes, and psychiatric disorders. When offering genetic testing, particularly for those disorders with adult onset, the molecular laboratory must be aware of the legal and ethical implications. Comprehensive protocols attentive to

Figure 8-8

Fragile X mutation analysis.

88

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

national guidelines must be implemented that should include genetic counseling, medical and (in some instances), psychological evaluations, and informed consent. Predictive testing is currently being performed for several adult onset diseases having trinucleotide repeat expansion mutations, including Kennedy disease, myotonic dystrophy, Huntington’s disease, spinocerebellar ataxia type 1, dentatorubral-pallidoluysian atrophy, and Machado Joseph/spinocerebellar ataxia type 3. In addition, cancer susceptibility genes, including those responsible for some inherited breast and ovarian cancer and colorectal cancer, have been identified, and tests have been developed for individuals at risk. While susceptibility testing is currently performed under research protocols, it is likely that testing will become more widely available. Predictive testing, while having positive implications for the individual and family, has the potential risk of insurance or employment discrimination and other negative implications. Such legal, social, and ethical issues specific to genetic testing are currently being debated and addressed at a national level. A key component to the future success of this discipline is the evolution of technology development to address volume and cost issues. Laboratory testing, which is now routinely done in highly specialized genetic testing laboratories in both academic and industrial settings, may eventually be done in hospital chemistry laboratories and physician offices. Gene and technology patenting issues further complicate and potentially limit the widespread application of genetic testing. However, the major challenge to ensure widespread access to genetic testing is education of health care professionals, consumers, and insurance carriers.

SELECTED REFERENCES Broholm J, Cassiman JJ, Craufurd D, et al. Guidelines for the molecular genetics predictive test in Huntington’s disease. Neurology 1994; 44:1533–1536. Chamberlain JS, Gibbs RA, Ranier JE, et al. Deletion screening of the Duchenne muscular dystrophy locus via multiplex DNA amplification. Nucleic Acid Res 1988;16:11,141–11,156. Clemens PR, Fenwick RG, Chamberlain JS, et al. Carrier detection and prenatal diagnosis in Duchenne and Becker muscular dystrophy families, using dinucleotide repeat polymorphisms. Am J Hum Genet 1991;49:951–960. DeMarchi JM, Caskey CT, Richards CS. Population-specific screening by mutation analysis for diseases frequent in Ashkenazi Jews. Hum Mut 1996;8:116–125. DeMarchi JM, Richards CS, Fenwick RG, Pace R, Beaudet AL. A robotics-assisted procedure for large scale cystic fibrosis mutation analysis. Hum Mut 1994;4:281–290.

Fu YH, Kuhl DPA, Pizzuti A, et al. Variation of the CGG repeat at the fragile X site results in genetic instability: Resolution of the Sherman Paradox. Cell 1991;67:1047–1058. Fu YH, Pizzuti A, Fenwick RG, et al. An unstable triplet repeat in a gene related to myotonic muscular dystrophy. Science 1992;255: 1256–1258. Grompe M. The rapid detection of unknown mutations in nucleic acids. Nat Genet 1993;5:111–117. Hejtmancik JF, Ward P, Tantravahi U et al. Genetic analysis: a practical approach to linkage, pedigrees, and Bayesian risk. In: Rowland LP, Wood DS, Schon EA, DiMauso S, eds. Molecular Genetics of Brain, Nerve and Muscle. New York: Oxford University Press, 1989; pp. 191–207. Kan YW, Dozy AM. Antenatal diagnosis of sickle-cell anemia by D.N.A. analysis of amniotic-fluid cells. Lancet October 28, 1978:910,911. Kawaguchi Y, Okamoto T, Taniwaki M, et al. CAG expansions in a novel gene for Machado-Joseph disease at chromosome 14q32.1. Nat Genet 1994;8:221–228. Kerem B, Rommens JM, Buchanan JA, et al. Identification of the cystic fibrosis gene: Genetic analysis. Science 1989;245:1073–1080. Koenig M, Hoffman EP, Bertelson CJ, et al. Complete cloning of the Duchenne muscular dystrophy (DMD) cDNA and preliminary genomic organization of the DMD gene in normal and affected individuals. Cell 1987;50:509–517. Korf B. Molecular Medicine Molecular Diagnosis (first of two parts). New Engl J Med 1995;332:1218–1220. Korf B. Molecular Medicine Molecular Diagnosis (second of two parts). New Engl J Med 1995;332:1499–1502. MacDonald ME, Ambrose CM, Duyao MP et al. A novel gene containing a trinucleotide repeat that is expanded and unstable on Huntington’s disease chromosomes. Cell 1993;72:971–983. Miki Y, Swensen J, Shattuck-Eidens D, et al. A strong candidate for the 17q-linked breast and ovarian cancer susceptibility gene BRCA1. Science 1994;266:66–71. Orr HT, Chung MY, Banfi S, et al. Expansion of an unstable trinucleotide CAG repeat in spinocerebellar ataxia type 1. Nat Genet 1993;4:221–226. Shiang R, Thompson LM, Zhu YZ, et al. Mutations in the transmembrane domain of FGFR3 cause the most common genetic form of dwarfism, Achondroplasia. Cell 1994;78:335–342. Tsui LC. Mutations and sequence variations detected in the cystic fibrosis transmembrane conductance regulator (CFTR) gene: a report from the Cystic Fibrosis Genetic Analysis Consortium. Hum Mut 1992;1:197–203. Verkerk AJMH, Pieretti M, Sutcliffe JS, et al. Identification of a gene (FMR-1) containing a CGG repeat coincident with a breakpoint cluster region exhibiting length variation in Fragile X syndrome. Cell 1991;65:905–914. Warren ST, Nelson DL. Advances in molecular analysis of fragile X syndrome. JAMA 1994;271:536–542. Zielenski J, Rozmahel R, Bozon D, et al. Genomic DNA sequence of the cystic fibrosis transmembrane conductance regulator (CFTR) gene. Genomics 1991;10:214–228.

CHAPTER 9 / GENETIC COUNSELING

9

89

Genetic Counseling BETH A. FINE

INTRODUCTION

birth defects, or disease-susceptibility, and investigate the problem in the family. Genetic counselors analyze inheritance patterns and risk of occurrence/recurrence, interpret and communicate information about the disorder, and explore and discuss available options with the family. Genetic counselors aim to communicate complex genetic and medical information to individuals and families in a manner that makes sense within the context of the patients’ lives. A patient’s educational level, culture, religion and psychosocial issues inform the way that genetic counselors frame risks and discuss the implications of genetic risk with families. Therefore, the information must be presented in a culturally sensitive manner in order to facilitate informed decision-making. Genetic counselors often work with patients and families in crisis, at the time of a diagnosis, or during loss of a pregnancy or family member. They provide supportive counseling to families, serve as patient advocates, and refer individuals and families to community resources/services. Some genetic counselors also serve as educators and consultants for other health-care professionals and for the general public. Genetic counselors work in the public-health setting, in academic and commercial laboratories as liaisons to clinicians, and are involved in clinical research and public policy.

The practice of genetic counseling has evolved in the last quarter century into a communication process between an appropriately trained health professional and a patient/client, couple, and/ or family, regarding the occurrence or risk of occurrence of a genetic condition and relevant implications. In response to advances in molecular genetics and diagnostic technologies, a new health care professional, the master’s-level genetic counselor, emerged in the early 1970s to participate in a multidisciplinary team approach to meeting the informational, psychosocial, and medical needs of individuals and families affected with or at risk for genetic conditions. In this chapter, a historical perspective on the profession and practice of genetic counseling in the United States will be followed by a discussion of the principles and types of service delivery in the settings in which genetic counseling is conducted. Next, the philosophical underpinnings of genetic counseling along with a review of the psychosocial, ethical, religious, and ethnocultural contexts in which patients experience genetic counseling and testing will be reviewed. The chapter will end with a description of geneticcounselor education, training, and credentialing. Since more nongenetics health professionals are participating and will continue to participate in aspects of genetic-counseling activities, the aim of this chapter is to introduce the components of genetic counseling together with the relevant issues faced by professionals and patients and their families as the Human Genome Project progresses. Resources for referrals to genetic counselors are included in Table 9-1.

THE GENETIC-COUNSELING PROCESS In 1974, an ad hoc committee of the American Society of Human Genetics defined genetic counseling as a communication process that deals with the human problems associated with the occurrence or risk of occurrence of a genetic disorder in a family. This process involves an attempt by one or more appropriately trained persons to help the individual or family to:

THE PROFESSIONAL GENETIC COUNSELOR Genetic counselors are health professionals with specialized graduate degrees and experience in medical genetics and counseling, often with undergraduate education in biology, genetics, psychology, public health, social work, or nursing. Approximately 1500 genetic counselors are members of the National Society of Genetic Counselors (NSGC); most practice in university or community hospitals. Genetic counselors work as members of a health-care team, providing information and support to individuals and families who have members with birth defects or genetic disorders, and to families who may be at risk for a variety of inherited conditions. They identify individuals and families at risk for genetic conditions,

1. Comprehend the medical facts, including the diagnosis, probable course of the disorder, and the available management. 2. Appreciate the way heredity contributes to the disorder and the risk of recurrence in specified relatives. 3. Understand the options for dealing with the risk of recurrence. 4. Choose the course of action that seems appropriate to them in view of their risk and the family goals and act in accordance with that decision. 5. Make the best possible adjustment to the disorder in an affected family member and/or to the risk of recurrence of that disorder (Fraser, 1974). Historically, the practice of genetic counseling, defined as providing information about recurrence risks for particular conditions within a family, began in the early part of this century. Genetic

From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

89

90

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Table 9-1 Resources for Referrals to Genetic Counselors Clinical Genetics Web Sites http://www.faseb.org/genetics http://www.kumc.edu/gec/geneinfo.htm http://members.aol.com/nsgcweb/nsgchome/htm National Cancer Institute Familial Cancer Risk Counseling and Genetic Counseling Information Genetic Counselor Directory http://cancer net.nci.nih.gov/www.prot/genetic/genesrch.shtml Genetic Counseling Resources National Society of Genetic Counselors 233 Canterbury Drive Wallingford PA 19086-6617 610-872-7608 E-mail: [email protected] American Board of Genetic Counseling 9650 Rockville Pike Bethesda MD 20814-3998 301-571-1825 American College of Medical Genetics 9650 Rockville Pike Bethesda MD 20814-3998 301-530-7127 American Society of Human Genetics 9650 Rockville Pike Bethesda MD 20814-3998 301-571-1825 Alliance of Genetic Support Groups 35 Wisconsin Circle, Suite 440 Chevy Chase MD 20815-7015 1-800-336-GENE

counseling was conducted by biologists with research interests in particular conditions. The goal of providing risk information in the absence of diagnostic or preventive measures reflected the eugenic philosophy of the time. In fact, the first genetic-counseling clinic, the Eugenics Records Office, was established in the early 1900s in Cold Spring Harbor, NY. The only options available to these families were to “take their chances” and initiate a pregnancy or to refrain from having biological children. In the 1940s, heredity clinics staffed by physicians who were interested in genetic diseases were established. Genetic counseling at that time was conducted by physicians and PhD geneticists. They primarily used a medical/preventive model that involved providing “facts”—risk figures, natural history, and treatment information— so that couples and families could make informed reproductive decisions. Geneticists assumed that these decisions were based on the patients’ perception of the risk, while studies revealed the importance of the perception of the burden of the disease. In essence, the genetic counseling was primarily an education process. With the advent of carrier testing and prenatal diagnosis in the late 1960s and early 1970s, families’ range of choices broadened. Geneticists recognized that information alone regarding genetic conditions and testing procedures was not enough to facilitate decision making and to truly serve the patients and families.

Dr. Melissa Richter, a visionary, initiated the development of a master’s degree program in genetic counseling at Sarah Lawrence College to prepare a new professional to address the needs of families at genetic risk. These new professionals would also serve as “physician-extenders,” providing the genetic information in a psychosocial context using skills the physicians did not possess. Thus, the team approach to genetics services began. As genetic counseling and related services moved from a more preventive focus to a more client-centered approach, some practitioners adhered to the decision-making model in which information is presented within the psychological and cultural context, so that patients can come to an appropriate decision regarding genetic testing and reproductive options. Genetic counselors came to espouse a nondirective approach to genetic counseling about reproductive issues to demonstrate respect for the patients’ autonomy and to move away from the eugenic underpinnings of earlier genetic counseling. Today, it is recognized that many psychosocial and medical factors influence the genetic-counseling process. The information communicated in most genetic-counseling sessions is emotionladen, often leading to a crisis. Therefore, many contemporary genetic counselors adhere to the psychotherapeutic genetic counseling model that involves psychosocial assessment, provision of information, psychological support and counseling, and referrals to other mental-health professionals as needed. It is often true that the genetic counselor functions as a liaison to the other specialists dealing with a patient and the family, providing a forum for sharing feelings, having them validated, and for processing information and resources. Genetic counselors work in pediatric or obstetric settings as well as in a wide array of specialty clinics, focusing on a particular disorder or group of disorders. While the contexts may differ, the process remains the same. Genetic counseling involves developing a relationship and rapport with a patient or family, usually in a short time period. In many instances, there is only one encounter between a patient and a genetic counselor, although many individuals affected with genetic conditions are followed on a long-term basis. The goal of genetic counseling is to make genetic and medical information relevant to the patient or family accessible and useful so that they can optimize their health care and lifestyle decisions. The genetic counselor begins each session by asking the patient what his or her goals are for the session and the relationship, e.g., their questions about diagnosis, testing, options for dealing with risks, therapeutic modalities, support, access to resources. The genetic counselor then places the patient’s expectations in context for what is possible within the limitations of scientific knowledge and available technology. Whether the genetic-counseling process occurs in one clinic visit or on a periodic basis over many years, the following components are necessary for the goals that the patient and professional wish to meet. INFORMATION GATHERING Geneticists and genetic counselors have been compared to detectives with regard to their work in making a diagnosis or discovering an underlying etiology for a condition that runs in a family. The sleuth work begins with gathering a great deal of information from the patient, his or her physician, and other family members. Whereas an accurate diagnosis is the most important component of quality genetic counseling, several types of histories must be obtained to arrive at that diagnosis. First, the genetic counselor will elicit a detailed family history, constructing a pedigree or a family tree. The purpose of the pedi-

CHAPTER 9 / GENETIC COUNSELING

gree is to create a graphic record of the family history for use in the medical record. The pedigree may lead to an accurate diagnosis and precise risk assessment for a particular condition in family members. Approaches to testing or treatment may be determined at least partially by information gathered for a pedigree. There are standard symbols used for pedigree notation that have been approved by the National Society of Genetic Counselors and the American Society of Human Genetics (Bennett et al., 1996). The pedigree should include at least three generations, including the proband (the affected person about whom the consultation is sought) or consultand (an unaffected individual seeking risk information), their children, siblings, parents, grandparents and aunts, uncles and cousins. Since most patients do not have complete medical information on all family members, it is important to encourage them to inquire about health histories from their relatives. They should inquire about symptoms or traits that run in families, as they can provide clues to a diagnosis. For example, if several relatives in more than one generation experienced sudden death from aortic dissection, one might consider the diagnosis of Marfan syndrome. The genetic counselor will pose questions about any medical, developmental, or physical abnormalities that may run in the family. In addition, questions will be asked about pregnancy losses, stillbirths, environmental exposures during pregnancy, birthdates or ages of individuals, ages and causes of death, and about ethnicity and possible consanguinity. When necessary, medical records may be requested to confirm diagnoses. This type of questioning also helps build a rapport with the patient, so that difficult or emotional issues can be addressed more easily. Genetic conditions are considered taboo or shameful in some families, so that the patient must feel safe and comfortable in discussing such matters. The questioning may be focused on features of specific conditions when the patient seeks genetic counseling for a particular disorder. For preconceptual counseling or prenatal diagnosis because a woman is over 35 years of age, the family history should be more general. However, if a family history of a particular condition is discovered, genetic counseling can address the risks for that condition if the patient is interested. The pedigree is useful as a communication tool in genetic counseling. When the patient or family is involved in constructing the diagram, it often jars their memory about other family members. It can be used to illustrate modes of inheritance and how risk figures are calculated. Finally, when genetic testing in a family is conducted, the pedigree is useful in determining which family members need to be tested and in interpreting their test results. In general, the pedigree is the genetic counselor’s primary tool for conducting risk assessment and counseling. Second, if there is an affected individual in the family, the genetic counselor must gather information about the medical history. This should include symptoms prior to diagnosis, age at diagnosis, confirmatory tests, course of the disorder, and current treatment. Confirmation of a diagnosis is necessary for accurate genetic counseling. Third, a developmental history is important, particularly in pediatric genetic counseling, since a large proportion of children with inherited conditions have developmental disabilities. Knowledge of specific disabilities is essential for accurate diagnosis in some cases. In other situations, a geneticist may make recommendations for special education or other therapies based on the specific condition or constellation of problems.

91

Finally, the genetic counselor gathers information about patient and/or family concerns, fears and questions regarding the genetic counseling process, the condition, or possible testing or treatment. It is also important to understand the family dynamics and support systems that often influence the care for the affected individual. In addition, the genetic counselor must assess the level of baseline knowledge and understanding of genetics, the genetic disorder, and the medical components of the condition. This information will guide the genetic counselor in presenting the information being offered to the patient and family. DIAGNOSIS An accurate diagnosis is central to providing appropriate and effective genetic services. A genetic diagnosis can be made in a variety of ways, depending on the age of the patient and the nature of the symptoms or characteristics leading to the referral to a genetics center. The physical examination provides a great deal of information, particularly in the case of dysmorphic features. The identification of physical findings, which include a constellation of characteristics and symptoms that represent a known syndrome, is important in providing answers about prognosis, natural history, and possible therapies or management strategies. In addition, a genetic diagnosis may lead to prevention of adverse symptoms or amelioration of a particular condition. For example, a healthy baby born with macrosomia, macroglossia, characteristic facies, and creases in the earlobes may carry the diagnosis of Beckwith-Wiedemann syndrome. While the features described here are not medical problems, infants with this syndrome are at increased risk for hypoglycemia, which, if left untreated, can lead to mental retardation. In addition, these children are at increased risk for Wilms’ tumor. Thus, by making the diagnosis, mental retardation can be prevented by monitoring blood sugar levels, and screening for early diagnosis of the kidney malignancy is possible. Many genetic diagnoses are made in the laboratory. Cytogenetic analysis performed on a blood sample from a child or adult, or on amniocytes or chorionic villi obtained by prenatal diagnosis leads to the preparation of a karyotype that reveals aneuploidy or chromosomal rearrangements. Chromosome analysis is the most frequently ordered genetic test, since the majority of women who have prenatal diagnosis are age 35 or older and are therefore at increased risk for chromosomal abnormalities such as Down syndrome. Biochemical assays are available to diagnose many inborn errors of metabolism in individuals and in fetuses. These conditions are primarily autosomal recessive conditions in which the gene product has been identified. For example, an assay to measure the level of α-iduronidase in a child with coarse facial features, hepatosplenomegaly, and developmental delay may lead to the diagnosis of Hurler syndrome. Biochemical assays are also used for carrier detection for rare recessive disorders. In most cases, carrier testing is offered for individuals with a specific ethnic origin who are at increased risk for carrying a gene for a specific condition. For example, carrier testing for Tay Sachs disease among Ashkenazi Jews involves measurement of the level of hexosaminidase A in a blood sample. The accuracy, specificity, and sensitivity of such tests vary; genetic counselors remain current on the standard procedures of this type of testing. Increasing numbers of genes for inherited conditions have been mapped and sequenced over the last several years. With the completion of the Human Genome Project, the genes for all genetic conditions will be mapped and sequenced. While the goal of the

92

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Human Genome Project is to understand how genes work under normal circumstances, many of these discoveries have resulted in the availability of DNA-based genetic tests. The large number and varied frequencies of mutations in each disease gene result in a range of accuracy for each test. The complexity of the interpretation of DNA test results necessitates genetic counseling that involves an informed consent process and psychosocial counseling before, during, and after testing. Again, as technology changes rapidly, genetic counselors are obligated to remain current on standard clinical tests and to identify appropriate research studies for their patients and families. Imaging techniques, such as ultrasonography, particularly in prenatal diagnosis, and X-rays for diagnosis of skeletal dysplasias and other syndromes, are often useful in diagnosing specific conditions. The use of ultrasonography in pregnancy to identify fetuses at increased risk for chromosomal abnormality or syndromes that involve structural abnormalities is increasing. Again, genetic counseling to explain the benefits and limitations of ultrasound diagnosis and issues surrounding uncertainty and loss is important to maximize patients’ ability to cope successfully. RISK ASSESSMENT Many patients seek genetic counseling with the explicit question, “What is the chance that this condition will occur in one of our family members?” or “What is the chance that this will happen again?” Geneticists use a variety of sources to attempt to answer one or both of these questions. While many patients and clinicians believe that arriving at a specific, accurate number to quantify the risk is the information desired, studies have shown that reproductive decisions are more likely to be based on a patient’s perception of the burden of the disease, rather than the magnitude of the risk of occurrence or recurrence. Also, the risk must be considered within the context of the patients’ lives, including an understanding of how the individual processes numerical values such as percents or ratios. Risk assessment followed by communication is a prerequisite to any type of decision making regarding reproduction or taking genetic tests. Recurrence or occurrence risks may be calculated by pedigree analysis in which a known mode of inheritance is identified. If two parents are found to be carriers of a recessive gene, then the risk for each future offspring is 25%. This information will affect different couples’ reproductive decisions in different ways. Empiric risks are often used when disorders are not caused by single genes, but are polygenic or multifactorial in nature. Empiric risks are derived from studying disease frequencies in a given population. If the data is gathered in an unbiased manner, using appropriate study design, the risk figures provide a sound basis for counseling. However, one must critically evaluate whether such studies are applicable to an individual patient because of differences in ethnic groups, factors that modify gene expression, and whether new developments led to reclassification of disorders that were once thought of as multifactorial. Calculations based on genetic testing results can also lead to risk assessment that is specific to a particular family. For example, a woman, with a brother affected with cystic fibrosis (CF), and her husband, who has no family history of CF, asked about their risk of having a child with CF. This risk, prior to any testing, is 1/120. The affected brother is found to have two copies of the common dF508 mutation; DNA analysis for the woman revealed that she carries this mutation. Her husband, of Northern European descent, tested negative for 30 common mutations. Since this test detects about 90% of all mutations, his risk of being a carrier is about

1/240. Thus, after testing, their risk to have an affected child is 1/1920, much lower than the a priori risk. Again, changes in the detection rates for many DNA tests occur often; genetic counselors must remain aware of new standards in testing and counseling. PROVISION OF INFORMATION Genetic counseilng is truly a communication process, designed to answer the concerns of patients and their health care providers about diagnosis, prognosis, and treatment/management options. Genetic counseling can be thought of as a conversation between genetic counselor and the patient and his or her family members. In meeting the needs of patients, genetic counselors often discuss the risk of recurrence or occurrence of a disorder in the patient and/or other relatives. The genetic counselor assesses the patient’s emotional and cognitive status so as to present the information in an appropriate manner. A significant component of genetic counseling involves a discussion of genetic testing as an option for patients. While not all genetic-counseling sessions involve a discussion of genetic testing, genetic counseling should be part of any genetic testing process. Obtaining informed consent for testing includes reviewing the risks, benefits, limitations, costs, ramifications, and accuracy of any genetic test. The genetic counselor must then facilitate the decision-making process regarding testing, based on the patient’s response to the informed consent process. If a patient elects to undergo genetic testing, either in the form of prenatal diagnosis of chromosomal disorders by amniocentesis, or in the form of a blood test on an individual, plans are made for the disclosure of results either by telephone or in person. Once a patient consents to testing, the genetic counselor must disclose test results when they become available. Usually, a genetic counselor will review the possible results and the patient’s reaction to each outcome prior to testing. A plan for communicating results either by telephone, mail, or in person is arranged. Finally, the genetic counselor facilitates an exploration of possible options following particular results. Communication, exploration of options, and support for the patient’s decision follow the provision of results. Genetic counselors can continue to be resources to the family and referring professionals in such circumstances. It is standard practice to provide written documentation of the session(s) to the patient and referring physician. This letter or report may also include diagrams or figures that illustrate the information shared with the family. Written communication serves as a record the patient and physician can refer to at a later time; it can also be used to assist in explaining the information with other family members. PSYCHOLOGICAL ASSESSMENT AND COUNSELING Since most genetic-counseling relationships are short-term, the genetic counselor must assess the psychological state of the patient and family rather quickly and tailor the plan for counseling and testing to meet the patient’s needs. A new diagnosis or possible diagnosis often presents a crisis to the patient; the genetic counselor must determine how to intervene in the time of crisis, providing the proper amount and type of information, or planning for a more comprehensive approach at a later date. After establishing a rapport with the patient and family, the genetic counselor elicits the patient’s perceptions about the genetic condition, testing, and the genetic-counseling process. The genetic counselor identifies the anxieties, fears, concerns, and hopes of the patient, accomplishing this by assessing verbal and nonverbal cues, and speaking with the patient, family members, and, occasionally, the referring physician.

CHAPTER 9 / GENETIC COUNSELING

The diagnosis of a genetic condition, or the realization that a health problem exists, leads to many emotions and reactions for the patient. By providing anticipatory guidance for the possible outcomes of genetic counseling and testing, the genetic counselor assists the patient by exploring issues of loss, grief, and survivor guilt. The genetic counselor can also assist the patient in examining the meaning of the diagnosis or possible outcomes of testing in the context of religious, spiritual, ethical, and cultural beliefs. When a patient faces difficulty in coping with the diagnosis and its implications, genetic counselors will make appropriate referrals to mental-health professionals or to any type of resource that will provide support financially, medically, or emotionally. SUPPORT FOR DECISION-MAKING AND ONGOING PATIENT SUPPORT In many genetic-counseling relationships, one goal of the session or sessions is to assist a patient in informed decision making regarding reproductive options, pursuing genetic testing, or acting on a test result. The genetic counselor aims to facilitate decision-making by providing a forum for discussion of the patient’s reactions to any and all possible outcomes for each choice. A supportive environment in which a patient or family does not feel judged by the genetic counselor for any option they choose is essential to successful genetic counseling. Once a decision is made, the genetic counselor will facilitate arrangements for the patient to act on the choice. For example, if an asymptomatic son of a man who died of hereditary nonpolyposis colon cancer elects to have predictive testing, the genetic counselor will make arrangements with an appropriate laboratory and arrange for follow-up counseling either in person or on the telephone. In many cases, the genetic counselor can help meet the support needs of patients by making referrals to peer-support volunteers or support groups on a local or national level. Often, patients who learn they have a family history of a genetic condition feel alone, not knowing anyone else with the disorder. While professionals can be supportive, many patients seek the true empathy that only an individual in the same situation can provide. The Alliance of Genetic Support Groups is an excellent resource for providing contacts for this type of support (see Table 9-1). Genetic counselors refer patients to a variety of medical specialists who can provide long-term care or treatment for patients with genetic conditions affecting a variety of systems. Ideally, a multidisciplinary continuity clinic can provide integrated, comprehensive care for such patients. Genetic counselors often serve on these clinic teams; others are aware of such clinics on a local, regional, and national basis. For example, a patient diagnosed with Marfan syndrome would benefit from attending a clinic staffed by specialists in cardiology, genetics, ophthalmology, and medicine. Many institutions have a Craniofacial Clinic at which children are followed and treated by otolaryngologists, orthodontists, plastic surgeons, speech pathologists, and audiologists. When patients are eligible for research protocols for diagnosis, treatment, and management, genetic counselors are excellent resources for referral to such programs. The experience of loss and related grief occurs with the diagnosis of many genetic conditions, whether or not the disorder leads to the death of an individual or a pregnancy loss. Parents of newborns diagnosed with a genetic condition often grieve the loss of the expected healthy child they had planned for and dreamed about. A woman who learns she has the Huntington disease gene grieves the loss of a healthy future. Genetic counselors maintain relationships by contacting patients by telephone or mail at the time of the

93

anniversary of a diagnosis or at the time a birth would have occurred if a pregnancy loss had not taken place. Plans for follow-up visits are also often made. In some cases, the genetic counselor provides support when seeing a patient or family when there are subsequent pregnancies or when new information on a disorder or test becomes available.

GENETIC TRIAGE: THE GENETIC COUNSELOR/ PRIMARY CARE PROVIDER PARTNERSHIP Since increasing numbers and types of disorders are being identified as genetic conditions, primary care providers are being faced with questions from patients regarding etiology, risk assessment, and management of genetic conditions. Most practicing physicians, physician assistants, and nurses have had little genetics education during their training or in postgraduate courses and therefore are unable to provide comprehensive genetic services. An important role for primary care providers is to identify patients at risk for genetic conditions and to make appropriate referrals to a genetics center. After eliciting a genetic family history, the primary care provider can consult with a genetics professional for risk assessment and identification of patients who may benefit from genetic counseling and/or testing. The primary care provider can prepare patients for the genetic-counseling process by providing anticipatory guidance, explaining what to expect from the genetics consultation. Genetic counselors maintain ongoing communication with the referring health professional and the patient, informing them of new developments when appropriate. Genetics professionals and the primary care provider will establish a cooperative plan for follow-up, medical management, and support for the patient as needed. For example, a 57-year-old, Ashkenazi Jewish woman, who was diagnosed with breast cancer at age 41, came for genetic-susceptibility testing, learning that she has the common BRCA1 mutation, 185delAG. Her unaffected daughter, age 32, came for risk assessment, genetic counseling, and testing. She was found to have the same mutation, indicating that she has an increased lifetime risk of developing breast cancer. The genetic counselor referred the young woman to her internist with recommendations for surveillance by mammography and clinical breast exam as determined by her physician. The team approach is one used by geneticists and genetic counselors since the beginning of the profession. Genetics professionals usually function as consultants to those who provide ongoing care to the patient.

PROFESSIONAL VALUES OF GENETIC COUNSELORS The practice of genetic counseling grew out of the medical model, whereby information was provided to patients, followed by advice from the clinician about reproductive choices. This prescriptive advice reflected the clinician’s perception of the genetic condition and the burden the disorder placed on an individual and his family. This view has been associated eugenic policies and practices of the early part of this century. However, the eugenic Nazi atrocities led clinical genetics in America away from directive, prescriptive practice. The birth of the genetic-counseling profession was concomitant with the women’s movement, the self-help movement, and the reproductive-rights movement. Thus, genetic counselors developed a technique of nondirectiveness with regard to reproductive decision making and prenatal-diagnosis counseling that reflects a respect for patient autonomy. It became obvious that in a situation that involved reproductive decisions,

94

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

such as deciding whether or not to continue a pregnancy after a fetal diagnosis of Down syndrome, there was no one right choice for all couples. Therefore, genetic counselors strive to provide information, support, and counseling in a context that is unique for each family. By demonstrating empathy, respect, and unconditional positive regard for their patients, genetic counselors can guide patients and families in understanding their own values, in making decisions that are appropriate for them, and in helping them cope with their grief. Nondirectiveness does not imply that genetic counselors give information, refuse to answer the question posed by many patients, “What would you do in this situation?” and remove themselves from the relationship when difficult decisions need to be made. The role of the genetic counselor is to elicit beliefs, values, and feelings from patients so that they can understand for themselves which decision is best for them. When there is clearly a better option for a couple, genetic counselors are often faced with a dilemma: a conflict of the bioethical principles of respect for autonomy and beneficence. These types of dilemmas are difficult, if not impossible, to resolve since the principles are not assigned varying levels of importance. Thus, when the National Society of Genetic Counselors (NSGC) considered the format for its Code of Ethics, it elected to draft the Code as an “ethic of care,” since this approach emphasizes the interdependence of individuals and best reflects the values, principles, and beliefs of genetic counselors. The relationships are: genetic counselors themselves; genetic counselors and their clients; genetic counselors and their colleagues, and genetic counselors and society. The NSGC Code of Ethics provides guidance and a framework for approaching dilemmas that genetic counselors face in practice. It includes the value of respect for client autonomy as an important focus. Genetic counselors, therefore, function within the health care system as patient advocates who aim to empower patients and families to find solutions to problems and ways of coping that are optimal for them.

GENETIC-COUNSELOR EDUCATION AND TRAINING In response to the development of amniocentesis for prenatal diagnosis and carrier screening for several recessive conditions, the first master’s-level program in genetic counseling was established at Sarah Lawrence College in 1969. By 1977, there were six programs that collectively graduated about 50 genetic counselors per year. These programs had unique curricula designed by faculty at each university. In 1979 a conference was convened to explore the training, role, and function of the genetic counselor. This was the first organized effort to standardize curriculum and training. The recommendations from this meeting have provided the backbone of education for genetic counselors. Some curricular modifications have resulted from the development of new diagnostic technologies in the laboratory, in the clinic, and from genetic counseling research. As the role of the genetic counselor expanded into specialty areas, the curricula changed to address the educational and practice needs of the students. For example, as questions from health professionals and pregnant women regarding the risks of exposures in pregnancy to the fetus arose with increasing frequency, genetic counselors (with their knowledge of birth defects,

embryology, epidemiological literature, and methods for presenting risks in a meaningful way) led the way in developing and staffing Teratogen Information Services. Similarly, as academic clinical molecular genetics laboratories and commercial biotechnology companies developed, their directors recognized the need to have a “customer service representative” for referring physicians and patients. Again, genetic counselors have filled many of these positions because of their skill in bridging the gap between the laboratory and the clinic. Throughout the 1980s, leaders in the field recommended course topics and approaches to clinical training, yet there were no requirements or guidelines for programs to adhere to. Today, 23 master’s-degree-granting programs are accredited by the American Board of Genetic Counseling (ABGC). Approximately 130 new graduates enter the field each year. Graduates of accredited programs must demonstrate proficiency in a wide range of appropriately supervised cases in the form of a log book and letters of recommendation to be eligible to take the ABGC certification examination.* Each program comprises didactic courses, clinical rotations, academic departmental activities such as case conferences, journal clubs, rounds, and clinical research (optional). Within this framework, each program is unique and reflects the approaches of the faculty and the nature and culture of the patient populations. Each program can guide students in developing these skills in any way it chooses: in the classroom, the clinic, or both. Instructional curricular content must include the following topics: 1. Principles and applications of human genetics and related sciences. 2. Principles and practice of clinical/medical genetics. 3. Methods of genetic testing. 4. Theory and application of interviewing and counseling. 5. Social, ethical, and legal issues. 6. Health care delivery systems/public health. 7. Teaching skills. 8. Research methods. Curricula and program accreditation are based on each program’s ability to support the development of 27 practice-based competencies in genetic counseling drafted by a consensus development team comprised of program directors and the ABGC Board of Directors. These competencies define what an entry-level genetic counselor should be able to do upon graduation from an accredited program. The 27 competencies fall into four skill categories: communication, critical thinking, interpersonal counseling and psychosocial assessment, and professional ethics and values. The competencies are published in the September 1996 issue of the Journal of Genetic Counseling. Graduate programs undergo an accreditation process every 6 years. A written self-study document is reviewed by the ABGC Accreditation Committee and recommendations are made regarding areas to explore during the 2-day site visit. Administrative/ institutional support, curricular content and design, clinical training, and student supports are several areas that are explored as part of the accreditation process. Programs are required to submit an annual report to document that the program is remaining in compliance with ABGC requirements.

*The 2-part national certification examination includes a general genetics examination taken by all geneticists (doctoral and masters level) and a separate specialty examination in genetic counseling.

CHAPTER 9 / GENETIC COUNSELING

The interdisciplinary academic training of genetic counselors is essential if patients’ needs are to be met successfully. The integration of course work in genetics, psychology, communication, ethics, and medicine, with supervised clinical experience, is essential for effective practice in the clinic and in society.

SELECTED REFERENCES Alliance of Genetic Support Groups. Directory of National Genetic Voluntary Organizations and Related Resources. Chevy Chase, MD, 1995. Andrews LB, Fullarton JE, Holtzman NA, Motulsky AG. Issues in genetic counseling. In: Assessing Genetic Risks: Implications for Health and Social Policy. National Academy Press, Washington, DC: National Academy, 1995; pp. 146–184. Bartels DM, LeRoy BS, Caplan AL. Prescribing Our Future: Ethical Challenges in Genetic Counseling, Hawthorne, NY: Aldine de Gruyter, 1993. Benkendorf JL, Callanan NI, Grobstein R, Schmerler S, FitzGerald KT. An explication of the National Society of Genetic Counselors Code of Ethics. J Genet Counseling 1992;1:31–40. Bennett RL, Steinhaus KA, Uhrich SB, et al. Recommendations for standardized human pedigree nomenclature. J Genet Counseling 1995;4: 267–280.

95

Fine BA, Baker DL, Fiddler MB, et al. Practice-based competencies for accreditation of and training in graduate programs in genetic counseling. J Genet Counseling 1996;5:113–122. Fraser FC. Genetic counseling. Am J Hum Genet 1974;26:636–659. Harper PS. Practical Genetic Counseling, 4th ed. Oxford, UK: Butterworth Heinemann, 1993. Kelly PT. Dealing with Dilemma: A Manual For Genetic Counselors. New York: Springer-Verlag, 1977. Kessler S. Genetic Counseling: Psychological Dimensons. New York: Academic, 1979. National Society of Genetic Counselors. National Society of Genetic Counselors Code of Ethics. J Genet Counseling 1992;1:41–44. Robinson A, Linden MG. Clinical Genetics Handbook, Cambridge, UK: Blackwell Scientific, 1993. Thompson MW, McInnes RR, Willard HF. Thompson & Thompson Genetics in Medicine, 5th ed. Philadelphia, PA: WB Saunders, 1991. Walker AP. Genetic counseling. In: Rimoin DL, Connor JM, Pyeritz RE, eds. Emery and Rimoin’s Principles and Practice of Medical Genetics, 3rd ed. New York: Churchill Livingstone, 1997; pp. 595–618. Walker AP, Scott JA, Biesecker BB, Conover B, Blake W, Djurdijinovic L. Report of the 1989 Asilomar meeting on education in genetic counseling. Am J Hum Genet 1990;46:1223–1230.

CHAPTER 10 / TRANSGENIC MICE AS MODELS OF DISEASE

10

97

Transgenic Mice as Models of Disease T. RAJENDRA KUMAR AND MARTIN M. MATZUK oped years ago, involves the infection of mouse embryos at different developing stages by retroviruses. Because of many technical constraints, this method presents problems for the routine production of transgenic mice. The second method, which has been the most widely used procedure, involves the direct microinjection of foreign DNA into the pronuclei of one-cell fertilized mouse embryos (Fig. 10-1). This results in the chromosomal integration of one or more copies of the injected foreign DNA (“transgene”) at a random site, and all cells of the embryo including the extraembryonic tissues will carry the injected DNA, since the injected transgenes usually integrate into the genome prior to embryo cleavage. The injected embryos are then transferred back to foster pseudopregnant mothers, and the offspring eventually are screened for the introduced transgene by standard protocols involving either Southern blot or polymerase chain reaction. The offspring generated are termed founder mice, and lines are eventually generated from these founder mice and maintained for successive generations by breeding to determine the transmission of the stably integrated transgene to the progeny. Because a wealth of information is available on different promoter or regulatory sequences of many genes, it is possible to over- or underexpress a given gene in a cellor tissue-specific manner. Specific promoters or regulatory sequences can be used to selectively express reporter genes, toxin genes, oncogenes, or antisense constructs, as well as wild-type and mutant genes, in transgenic mice. Promoters can be constitutive, inducible, or even conditionally regulated, depending on the situation. In instances in which cell transfection studies are not possible, transgenic mouse technology can be exploited as a powerful expression system. Promoters containing hormone-responsive elements, enhancers, silencers, and locus-controlling regions can be finely mapped and functionally dissected. By targeting coding sequences of commercially and pharmacologically important proteins using mammary gland-specific promoters, bulk purification of these products is also possible. The third method for generating transgenic mice is based on a gene-targeting approach in embryonic stem (ES) cells (Fig. 10-2). ES cells are derived from the cells of the inner cell mass of the blastocyst. The ES cells are pluripotent and can contribute to all tissues of the embryo. ES cells can be genetically manipulated in vitro and so-called “gene-targeting” vectors can be used to manipulate DNA at specific loci. The locus to be manipulated can be deleted, and this deletion can span a region as small as a base pair to as large as several megabases. Different types of vectors (insertional or replacement) can be designed, depending on the

BACKGROUND Efforts to manipulate the genome have been the constant pursuit of geneticists since the end of the 19th century. Methods to improve the quality of the species have been practiced and perfected by plant breeders. Induction of random mutations by UV-radiation and consequent screening for interesting phenotypes in bacteriophage or fruit flies have set forth a trend to identify the genetic basis of structural and functional malformations in these organisms. The advent of chromosomal mapping and gene cloning techniques and the availability of breeding data in many animal species have made it possible to selectively manipulate the genomes of species such as mice, rats, pigs, and cattle. This technology, called “transgenic animal” technology, has already revolutionized our current understanding of how organisms develop and how several physiological processes are regulated. In addition, transgenic models have increased our understanding of the genetic basis for many human diseases, including cancer. Although gene manipulation is theoretically possible in many species, the mouse has become the obvious choice for several reasons. Mice are relatively inexpensive to maintain and easy to breed, and an exhaustive store of information is already available on chromosomal mapping and linkage analysis of many genes in the mouse. In addition, micromanipulation of mouse embryos is technically easier and more feasible compared to that of other species. The manner in which transgenic technology has been used until today and will be used in the future has implications for basic and clinical research. Using essentially the same principles, but with specific modifications in the approach, it is now possible to create mice with precise targeted genetic lesions or generate mice that express any gene product at a designated time and place. As a result, numerous strains of mice have been created to address diverse issues such as early development, organogenesis, growth and differentiation, and the molecular basis of many diseases. In the following sections, we will briefly discuss the principles of transgenic technology and illustrate the efficacy and power of this novel approach by drawing some important examples in several physiological systems.

PRINCIPLES OF TRANSGENIC TECHNOLOGY Foreign DNA can be introduced into the chromosomal DNA of mice essentially in three ways. The first method, originally develFrom: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

97

98

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

driven by a cell/tissue-specific promoter. The Cre protein can excise the gene sequences between the lox sites and thus result in a tissue-specific gene deletion. This method has many advantages in situations where the regular mutation is embryonic lethal or the mice eventually develop multiple defects in several tissues. In this manner, the specific function of a gene product in a given cell type can be studied by loss of function in that cell without affecting its expression in other cell types of the mouse. Alternatively, it is also possible to “knock-in” or replace a different new gene into a locus and disrupt the actual gene at that locus. In the following sections, we will describe the successful application of transgenic technology to study whole-animal biology and the development of mouse models for many human diseases. We will illustrate the functional analysis of each physiological system with a few examples.

APPLICATIONS OF TRANSGENIC TECHNOLOGY

Figure 10-1

Production of transgenic mice.

situation. Typically, replacement vectors are used that have positive and negative selectable markers, and sequences homologous to the target gene flank the positive selectable marker. The targeting vector DNA is then introduced into ES cells in culture, and a simultaneous double-selection procedure is used to enrich for the correct homologous recombination event. The negative selection eliminates many of the cells with random integrations. Cell clones that survive the double selection are expanded, and the DNA from these clones is analyzed by the polymerase chain reaction or Southern blot analysis to identify clones that contain the correctly targeted recombinant allele. ES cells carrying the desired mutation are microinjected into 3.5-day-old blastocysts and implanted back into uteri of the foster mice. Most ES cell lines are derived from mice that have an agouti coat color, and the blastocysts are usually collected from mice that have either a white or black coat color. Therefore, when the genetically manipulated ES cells mix with the inner cell mass cells of the injected blastocysts, the offspring are chimeric because of the patchy appearance of these distinct coat colors. The chimeric mice are subsequently mated to wild-type mice to obtain germline transmission of the mutant allele. The heterozygous mice (i.e., mice carrying one copy of the mutant allele) are bred, if viable and fertile, to obtain homozygous null mice (i.e., mice with two mutant alleles). If the mutation is not lethal in embryos, then theoretically 25% of the offspring will be homozygous for the mutation, following a typical Mendelian inheritance pattern. Recently, it has also been possible to generate “tissue-specific” knockout mice. The mutation in the gene of interest is achieved as described above in ES cells but with insertion of phage sequences called lox sites that flank (f) the region to be deleted. These “floxmice” are then mated to another line of transgenic mice that harbor a transgene containing the phage Cre protein coding sequences,

CELL-CYCLE CONTROL The process of cell multiplication involves a complex series of regulatory steps. Each of these steps is either activated, inhibited, or blocked by several protein kinases, phosphatases, cyclin proteins, growth factors, proto-oncogenes, or tumor suppressor proteins. Gain of function and/or loss of function mutations in the genes encoding these regulatory proteins often will lead to important insights into the origin and/or development of many tumors. p53 is an important tumor-suppressor protein. It is ubiquitously expressed in mouse and human tissues. Under normal physiological situations, p53 is phosphorylated by cdc2 kinase in a cellcycle–dependent manner. Thus, activation leads to the binding to and inhibition of a yet unidentified cell-replication protein and prevents the cells from entering the S phase. Mutations and allelic loss of the p53 gene have been associated with many human tumors. These genetic lesions are frequently observed in spontaneous human cancers. Mutations in p53 are associated with the inherited cancer-susceptibility Li-Fraumeni syndrome in humans. To study the function of p53 in an animal model, mice carrying mutations in one or both copies of the p53 gene have been generated. Surprisingly, p53-deficient mice develop normally. Therefore, p53 does not play an essential role in cell-cycle control during embryonic development. However, these mutant mice develop tumors at a very early age with 100% penetrance. The tumors are often derived from multiple cell lineages, suggesting that the loss of growth control is not cell-type-restricted. Mice heterozygous for the mutant p53 allele also develop tumors, although it usually takes more time to develop these tumors. The analysis of tumors from the heterozygous mice demonstrates that the tumor cells usually lose the remaining wild-type p53 allele. Heterozygous p53 mutant mice are more susceptible to tumors when challenged with carcinogens, which supports a role of p53 in DNA damageresponse pathways. Because p53-deficient mice are highly susceptible to a wide spectrum of tumors as diverse as sarcomas, lymphomas, choriocarcinomas, and gonadoblastomas, these mice are invaluable for testing suspected carcinogens and cancer therapeutic agents. Several proteins are known to interact with p53 under different conditions in a given cell. One of the proto-oncogene products called Mdm2 (named after a gene located on a mouse doubleminute chromosome present in 3T3 fibroblast cells) forms a complex with the p53 protein and inhibits the p53-mediated transregulation of gene expression. The human Mdm2 gene is amplified in 30–40% of sarcomas and is overexpressed in many

CHAPTER 10 / TRANSGENIC MICE AS MODELS OF DISEASE

Figure 10-2

99

Production of knockout mice.

leukemic cells. To understand the biological role of Mdm2 during development, null mice have been created. This mutation leads to lethality of the mice between embryonic day 4.5–7.5. In normal mice, both p53 and Mdm2 are ubiquitously expressed on embryonic days 6.0–6.5, and at this point in normal mouse development, there is a sudden increase in cell-cycle rate. To further resolve the issue of whether the embryonic lethality in Mdm2-deficient mice is a result of alterations in Mdm2-mediated gene expression or a loss in the ability of Mdm2 to downregulate p53 activity, doublehomozygous mutants for both Mdm2 and p53 are generated by intercrossing the corresponding compound heterozygotes. Interestingly, absence of p53 rescues the embryonic lethality in Mdm2deficient mice. These results indicate that Mdm2 functions normally in development primarily as a downregulator of p53 activity. Mdm2/p53 double-mutant mice develop normally, are fertile, and they appear phenotypically similar to mice deficient in only p53. The retinoblastoma susceptibility gene (Rb) is another tumor suppressor gene associated with a wide variety of tumor types, including retinoblastomas, sarcomas, and breast, prostate, and lung carcinomas. The inheritance of a mutant form of the Rb gene by an individual is associated with a 90% incidence of multifocal, bilateral retinoblastomas. Loss of heterozygosity has been observed at the Rb locus in these tumors, confirming that both copies of the Rb gene must be mutated for the development of retinoblastoma. Rb protein has been implicated to play key role(s) in cell-cycle control, and the gene is ubiquitously expressed in different cell types. However, homozygous mice with mutations in both Rb genes develop to mid-gestation but exhibit defects in the hematopoietic system and central and peripheral nervous systems leading to embryonic lethality. In contrast, heterozygous mice, like humans, exhibit a dramatic predisposition to tumor development with 100% penetrance by 18 months of age. However, the mice do not develop retinoblastoma, indicating that there are species differences in the susceptibility of differentiated cell types to loss of function muta-

tions in Rb. Instead, these heterozygous mice develop multifocal tumors of the intermediate lobe of the pituitary gland and demonstrate elevated serum levels of α-melanocyte-stimulating hormone. It is now feasible to consider using mice with specific genetic mutations as an assay system for the identification of tumorsuppressor genes because the effects of such targeted mutations in either known or unknown genes can be directly tested in vivo. For example, using this principle, inhibin has been shown to be the first secreted tumor-suppressor protein with specificity restricted only to the gonads and adrenals. Inhibins are heterodimers (α:β), which belong to the transforming growth factor (TGF)-β superfamily. They share the β-subunits (βA- or βB-subunits) with activins. Inhibins were originally discovered as Sertoli and granulosa cell products that had FSH-suppressing activity. These peptides are also expressed in other tissues outside the reproductive tract. Mice deficient in the α-subunit of inhibin are viable, indicating that inhibin is not essential for embryonic development. However, between 5 and 20 weeks, both male and female mice deficient in inhibin develop hemorrhagic and bilateral gonadal sex cordstromal tumors with nearly 100% penetrance. As the tumors progress, they secrete high levels of activin, leading to a humancancer cachexia-like wasting syndrome in the mice with hepatocellular necrosis and weight loss leading to eventual death. In contrast, gonadectomized, inhibin-deficient mice survive beyond 20 weeks but develop adrenal cortical tumors and also die of a similar wasting syndrome. These knockout mouse studies have provided an excellent opportunity to examine similar human gonadal tumors for mutations in the inhibin α-subunit gene, which was not previously envisioned. DEVELOPMENTAL BIOLOGY Highly programmed molecular events triggered by the process of fertilization between the mammalian sperm and egg lead to distinct layers of cells and pattern formation in the embryo. The changing topography of the body axis during normal development eventually results in organogenesis. Mutations affecting critical events associated with early

100

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

embryogenesis may result in severe deformities of the developing embryo at distinct steps or even embryonic lethality. Several human mutations affecting the fetus have been described, and many interesting lines of mutant mice have been generated with the aim of dissecting the critical molecular events during early embryogenesis. Numerous genetic manipulation studies on fruit flies, zebrafish, and C. elegans have paved the way to identify homeobox genes and other related pattern-regulating genes. Many of these genes encode transcription factors involved in organogenesis, body axis determination, and distinct layer formation. The corresponding mouse homologs have been identified, and mice with mutations in these genes have been created and characterized. In some cases, these critical factors have been ectopically expressed to follow the cell fate and consequences of this aberrant expression. Cleft palate is a commonly occurring craniofacial developmental disorder in humans. It has been possible to identify some of the critical factors during craniofacial development by generating different strains of mutant mice. For example, the majority of the mice that are deficient in activin βA, a member of TGF-β family, develop cleft palate, tooth defects, and lack of whiskers. Some of the mice that do not have cleft secondary palate exhibit either incomplete or complete failure of hard-palate formation. As a result, these mice cannot nurse and die perinatally. Similarly, mice deficient in follistatin, an activin-binding protein, also exhibit cleft palate and have hard-palate defects. In addition, mice that are deficient in a transcription factor MSX-1 display similar phenotypic characteristics. Thus, using transgenic mice, one can verify the in vitro observations in an in vivo context and also draw conclusions on the involvement of several factors in a common pathway controlling a physiological process. Several lines of transgenic mice have also been created to study embryonic formation of limbs, notochord, different brain regions, musculoskeletal structures, and body-axis. Mice with mutations in the presumptive “mesoderm-inducers” have also been generated. It has been possible to dissect out the early developmental events in the formation of virtually every organ system using transgenic mice. A few examples are given below. The Wnt-1 (int-1) proto-oncogene, which encodes a putative signaling molecule, is expressed exclusively in the developing central nervous system (CNS) and adult testes. To test the presumptive function of Wnt-1 in mammalian neural-tube organization and patterning role in CNS development, Wnt-1 null mice have been generated. These mice develop to term but die within 24 h after birth as a result of abnormal brain development. Most of the neural tube and nonneural tissues appear normal. However, as early as embryonic day 14.5, a substantial portion of the midbrain fails to develop. This includes a large contiguous domain, comprising approximately the caudal two-thirds of the midbrain, the normal midbrain-metencephalic junction, and rostral metencephalon. Thus, for the first time, using an in vivo approach, it has been established that Wnt-1 acts as a determinant for the development of a specific region of the CNS. Defining and understanding the nature of vertebrate organizers in molecular terms have been among the fundamental issues in embryology. Toward this end, several putative DNA transcription factors that are expressed in the organizer region have been identified in Xenopus, chick, and mouse. The Lim 1 transcription factor is the mouse homolog of xlim-1 in Xenopus, which is expressed in the node, the developing kidney, and portions of the CNS. Lim-

1-deficient mice lack head structures anterior to the optic vesicle but develop the remaining body axis normally. The mutant embryos lack forebrain and midbrain because of the absence of an organized node, head process, and prechordal mesoderm by embryonic day 7.5. Some mutant embryos develop a partial anterior secondary axis as a result of a disrupted early organizing region. By embryonic day 10.5, most of the embryos die but a few “headless” pups (4/1000) are delivered stillborn. These mutant mice also lack kidneys and gonads. Thus, using knockout mouse technology, the requirement of Lim-1 in head-organizer formation has been clearly confirmed. Combinatorial interactions between products of clusters of Hox genes direct regional development in the embryo. Hox genes encode transcriptional regulators that contain sequence-specific DNA binding motifs called homeodomains. Mutations in most of these genes often result in homeotic transformations. In the mouse, four clusters of Hox genes exist that are homologous to the Drosophila homeotic complex. Both gain-of-function and loss-offunction mutations in these genes have been created in transgenic mice. These experiments have demonstrated that Hox genes act as key selector genes and have led to the “Hox code”-hypothesis, in which several Hox genes control the specification of regional identity (i.e., they control the formation of the anteroposterior body axis and secondary axis of the limb). By either overexpression or deletion of a particular set(s) of Hox genes, it has been possible to specify and demarcate the boundaries of expression of each gene during development and determine the consequences of such a perturbation. One of the models developed to establish the specific role of a Hox gene in a specific cluster has been to introduce Hoxb-4 mutations into the mouse germline. Hox genes are organized in tightly linked clusters with poorly understood and complex mechanisms of transcriptional regulation. Because the conventional “replacement” gene-targeting strategy introduces exogenous sequences that may affect transcription of the Hox gene cluster and disrupt normal regulation at several important genetic loci, genetic subtle mutations have been engineered into the Hoxb-4 locus. Hoxb-4 is a member of the Hoxb cluster on mouse chromosome 11. Hoxb-4 is first expressed at embryonic day 8.5 from the posterior end of the prospective spinal cord and eventually the anterior limit at the boundary between rhombomeres 6 and 7 in the hindbrain. In addition, Hoxb-4 is widely expressed in tissues including spinal ganglia, the nodose ganglion of cranial nerve X, prevertebrae up to the second cervical vertebra or axis (C2), and in mesodermal components of lung, kidney, and gut. Mutant mice that harbor a disrupted first exon of Hoxb-4 show two obvious skeletal changes—a partial homeotic transformation of C2 from axis to atlas and a defective morphogenesis of the sternum. These phenotypes are variable in penetrance and expressivity depending on the genetic background. Mice generated from ES cells that are engineered to have a premature stop codon in the second exon, by an insertion of a oligonucleotide using the “hit-and-run” method, also show partial homeotic transformation of axis to atlas; these mice, however, do not show abnormalities in sternum. Whereas only a portion of the first-category mice survive to adulthood, mice that carry the second type of mutation are viable, healthy, and fertile. These experiments illustrate that ES cell technology can be effectively used to introduce subtle mutations into the mouse genome, and this can be used as an assay system to study their effects on functions of complex clusters of Hox genes. Selected

CHAPTER 10 / TRANSGENIC MICE AS MODELS OF DISEASE

genetic crosses between individual mutant mice belonging to individual clusters give important clues to the combinatorial interactions that operate among these key regulatory molecules. MUSCLE BIOLOGY A distinct set of transcription factors called myogenic basic helix-loop-helix (HLH) factors have been implicated in the regulation of skeletal-muscle differentiation. These include MyoD, Myogenin, Myf-5, and Myf-6/MRF-4. Each of these factors, when ectopically expressed in nonmuscle cells, can activate the skeletal-muscle differentiation program. Each factor can induce other factors, including a number of skeletal muscle-specific genes. Null mutations in each of these factors have been introduced into the germline of mice. Further, mice with multiple genetic lesions in these factors have also been obtained by breeding. It has been possible to formulate a hierarchical network of interactions involved in the program of muscle differentiation based on the functional analyses on these mutant mice. Mice deficient in MyoD are viable and show no skeletal-muscle abnormalities. The inactivation of this gene causes a twofold increase in Myf5 mRNA expression, whereas myogenin and MRF4 levels are not altered. Myf-5-deficient mice also develop normal skeletal muscle, but die at birth because of the absence of the distal parts of the ribs, which leads to breathing problems. Double-mutant mice that lack both MyoD and Myf5 produce no detectable skeletal-muscle-specific markers (skeletal-actin, troponin T, and so on) and appear to lack skeletal myoblasts, suggesting that both of these factors function similarly in myoblast formation. Mice lacking myogenin have normal numbers of skeletal myoblasts at birth, but show a severe reduction of skeletal-muscle fibers. Both MyoD and Myf5 are expressed in undifferentiated myogenic cells, suggesting that they act earlier in development than myogenin. MRF4deficient mice show only a slight reduction in a subset of muscle-specific genes. Consistent with the in vitro observations that MRF4 downregulates the expression of myogenin, myogenin levels are elevated in adult skeletal muscle from MRF4-deficient mice. Surprisingly, MRF4-null mice exhibit multiple rib abnormalities, possibly through an indirect mechanism affecting the rib primordia. Duchenne muscular dystrophy (DMD) is an X-linked recessive disease in humans caused by defective expression of dystrophin. Progressive skeletal-muscle weakness and wheelchair confinement are the most obvious symptoms of DMD. Most patients die of respiratory failure resulting from progressive atrophy of the diaphragm. There is a naturally occurring mutant mouse strain called the mdx mouse that has been identified. This mouse has a point mutation that eliminates the expression of the 427-kDa muscle and brain isoforms of dystrophin. These mdx mice do not display muscle weakness or impaired movement, but the diaphragm muscles show progressive myofibre degeneration and fibrosis similar to the human condition. Transgenic mouse technology has demonstrated the feasibility of gene therapy for DMD. High-level and tissue-specific expression of a full-length murine dystrophin was achieved in transgenic mice by using the regulatory regions of the mouse muscle creatinine kinase gene, and the transgene was introduced into the mdx genetic background by crossbreeding. The transgene positive mdx mice show complete correction of the dystrophic pathology in skeletal muscles and the diaphragm as assessed by both functional and immunohistochemical tests. It is now feasible to generate viral vectors that encode the full-length human dystrophin to treat and/or cure human patients with DMD.

101

HEMATOLOGY/IMMUNOLOGY The most thoroughly studied aspects of mammalian physiology using transgenic technology are perhaps the immune and hematopoietic systems. Both by conventional microinjection techniques and using ES cell technology, mutant mice have been generated that either overexpress or lack cell-specific transcription factors, immune cell-specific light chains, the recombination-associated proteins, receptors for antibodies, MHC antigens, and immune cell-specific growth factors (lymphokines/ cytokines). Because many human disorders involve aberrations in hematopoietic/immune systems, these mice serve as excellent models to understand and treat these human disorders. Mice that lack an important component of these systems often are more prone to many microbial diseases and are often sensitive to immediate and/or delayed hypersensitivity reactions. These transgenic mice can be effectively used to screen for the efficacy of various anti-inflammatory drugs and other chemotherapeutic products and are also useful models in transplantation biology. β2-microglobulin is a 12-kDa polypeptide and is associated with the heavy chain of the polymorphic MHC class I proteins encoded by the H2-K, H2-L/D, and Qa/Tla loci. One of the first mouse models generated using ES cell technology involved the targeted disruption of the β2-microglobulin gene. Mice deficient in β2-M microglobulin were generated to precisely understand its biological function. These mice are normal and fertile but have no mature CD4–CD8+ T cells in either the thymus or peripheral lymphoid organs and are defective in CD4–CD8+ T-cell-mediated cytotoxicity. Although it was widely believed that γδ T cells interact with class-I molecules, this subset of cells develops normally in these mutant mice. TGF-βs are homodimeric peptides that control a wide variety of cellular processes such as cell proliferation, differentiation, embryonic development, extracellular matrix formation, bone development, wound healing and hematopoiesis, and immune and inflammatory cell response. Three distinct TGF-β genes (TGF-b1, -b2, and -b3) have been identified in mammals. The three mammalian TGF-βs are structurally highly similar (75% identity at the amino acid level) and are often coexpressed and colocalized in a spatiotemporal manner. Many additional functions are also attributed to TGF-βs based on in vitro observations. Obviously, to delineate the individual functions of these peptides, gene-targeting approaches have begun to be employed. Approximately 50% of TGF-β1 null mice show no gross developmental abnormalities. By 3–4 weeks of age, however, these viable mutant mice succumb to a wasting syndrome and die. All mutant mice exhibit a marked but variable degree of mixed inflammatory cell infiltration and tissue necrosis in multiple organs. Many of these lesions resemble those found in human autoimmune disorders, graft vs host diseases, and some viral diseases, and suggest that the TGF-β1– deficient mice are an important model to study these diseases. Gene-targeting approaches have also recently helped to understand the molecular basis of human inflammatory bowel disease (IBD). These studies have led to the unexpected discovery that mice with deletions in genes encoding specific cytokines and T-cell receptor subunits develop chronic intestinal inflammation. In humans, IBD manifests as a chronic, noninfectious, inflammation limited to the large bowel, known as ulcerative colitis, or as a granulomatous inflammation anywhere along the gastrointestinal tract, known as Crohn’s disease. Several lines of evidence suggest that human patients with IBD are hyperresponsive to nor-

102

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

mal gut constituents. Mutant mice that are deficient in T-cell receptor (TCR)-α, TCR-β, both TCR-β/δ, or class II major histocompatibility complex all develop chronic intestinal inflammation. This intestinal disorder resembles ulcerative colitis but without ulceration or bleeding. These mice have normal number of B cells but lack class II-MHC-restricted CD4+ αβ T cells. It was hypothesized that, in the absence of a tolerance to dietary or microbial antigens, an autoimmune attack on the intestinal epithelium results in the inflammatory response. This suggests that the αβ T-cell-mediated suppression of B cells does not occur in these mice. Interleukins are cytokines of the immune system. Interleukin-2 (IL-2) is produced by activated T cells. It is a key regulator of immune and inflammatory responses. T-cell proliferation in vitro, differentiation of B cells, activation of macrophages, and NK cells are all dependent on IL-2 signaling. IL-2–deficient mice at 4 weeks of age develop normal thymus glands, with unaltered thymocyte and peripheral T-cell subset composition. The in vitro responses to T-cell mitogens on isolated T cells from these mice are reduced. The differentiation of B cells is affected in these mice, with a drastic increase of serum levels of immunoglobulin (Ig) G1 and IgE. IL-2–deficient mice beyond 6 weeks of age develop an IBD similar to ulcerative colitis in humans. These mice develop ulcerations and bloody diarrhea, and the mucosa and submucosal tissue of the large bowel shows pronounced thickening. Crypt abscesses are found, and the epithelial layer shows loss of goblet cells. Similar to IL-2–deficient mice, mice with a gene deletion in IL-10 also develop mucosal inflammation. However, the inflammation in these mice is restricted to the villi. The crypt is associated with pseudopolyps and villous atrophy. These mice show elevated levels of interferon-γ. The intestinal disease in these mice may result from the absence of a suppressive effect of IL-10 on macrophage production of inflammatory cytokines. However, these knockout mouse models are not identical to patients with IBD, because IBD patients have T cells and produce IL-2 and IL-10. However, the mouse models elucidate that T-cell abnormalities could be the cause of many forms of intestinal inflammation. These mice may also prove useful in testing the involvement of other genetic and nongenetic factors in the etiology of human IBD. PULMONARY BIOLOGY Transgenic mouse technology has successfully been used in four major aspects of pulmonary biology research. First, the molecular mechanisms controlling epithelial-cell gene expression have been determined. Second, the cellular interactions that contribute to lung morphogenesis and maturation have been elucidated. Third, models that may be clinically relevant for developing new strategies for therapy have been developed. Fourth, functional analysis of genes by targeted expression or mutation and their consequences in lung growth and development has been undertaken. The respiratory-cell epithelium produces distinct gene products, including surfactant proteins A, B, and C (SP-A, B, C), clara cell secretory protein (CCSP) or uteroglobulin. The most exhaustively studied genes are SP-C and CCSP. Expression studies using transgenic mice have identified that a 3.7-kb human SP-C promoter contained the necessary elements to drive a chloramiphenicol acetyltransferase reporter gene correctly to bronchiolar and alveolar respiratory epithelial cells. The same promoter has been used to express diphtheria toxin-A chain in transgenic mice. These mice die immediately after birth as a result of cytotoxic lesions of the respiratory epithelium induced by the diphtheria toxin expression. Similarly, to study lung epithelial-mesenchymal interactions in

vivo, transgenic mice have been created that express human TGF-α under the control of the 3.7-kb SP-C promoter sequences. These mice develop marked pulmonary fibrosis, exhibit increased collagen deposition, and demonstrate disruption of elastin fibers. Two other mutant strains of mice have been created with lungspecific promoter sequences. In one case, a dominant negative FGF-receptor mutant is expressed in transgenic mice, and in the second case, SV40 large T-antigen expression is directed from the 5' flanking sequences of the uteroglobin gene. While the mice expressing the former transgene die of respiratory failure similar to the 3.7-kb SP-C diphtheria toxin-A-bearing mice, the latter mice die because of the development of pulmonary adenocarcinoma. These models are highly useful for the study of molecular pathogenesis and therapy for adenocarcinoma of the lung in humans. Transgenic mice have also been created to study the human disease cystic fibrosis. Cystic fibrosis is the most common fatal autosomal-recessive disorder, affecting about 1 in 2500 newborns. The disease is characterized by defective chloride transport in epithelial cells leading to excess mucus secretion. Meconium ileus, or intestinal blockage, is also a diagnostic feature of the disease. Associated defects include pancreatic insufficiency, malabsorption in the gut, chronic opportunistic lung infections, and reproductive tract defects. To develop mouse models, both insertional mutagenesis, which disrupts exon 10 of the CF gene, and a replacement-type vector to delete the same exon, have been employed. The cystic fibrosis transmembrane regulator (CFTR) mutant mice demonstrate many similarities in their pathology to human patients. The majority of mutant mice die from intestinal perforation and peritonitis following intestinal blockage. This is primarily caused by dehydrated secretions in the crypts of the intestine. The pancreatic effects seen in humans are not recapitulated in mice, however. Pathological changes in the respiratory tract of mutant mice mimic those in humans-goblet cell number increases, gland ducts in the nasal and proximal trachea dilate, and destructive changes in the upper-airway epithelia occur. Interestingly, there are no reproductive defects in male mice that survive to the adult stage, but female mice show infertility. Mice deficient in N-myc proto-oncogene exhibit markedly slow growth rates in embryonic lungs, resulting in hypoplastic lungs at birth. Targeted disruption of a novel homeobox gene, Gsh-4, similarly results in neonatal respiratory failure in the mutant mice. The expression of Gsh-4 mRNA is confined only to the CNS, suggesting that the mice die as a result of a respiratory-center failure in the brain at birth. Both of these mouse models have provided previously unanticipated results demonstrating the power and the advantage of transgenic mouse technology to establish novel functions of known gene products. CARDIOVASCULAR BIOLOGY Hypertension and atherosclerosis are the most common human disorders. Many transgenic animal models have been developed to study cardiovascular function. Cardiovascular function is dependent on a multitude of hormones, regulatory peptides, and cell-signaling pathways, in addition to the genetic background. Only a few candidate genes have been identified thus far, and many quantitative trait analyses are being performed to identify loci that control cholesterol levels in mice. Antidiuretic hormone (ADH) is a nona-peptide synthesized in the hypothalamus as a precursor and later transported to and stored in the posterior pituitary. The major endocrine function of ADH is to increase reabsorption of water in collecting ducts of the kidney. ADH also increases blood pressure by constricting the arterioles.

CHAPTER 10 / TRANSGENIC MICE AS MODELS OF DISEASE

Absence of ADH results in diabetic insipidus, a common human disease with problems of water retention. Surprisingly, transgenic mice that chronically express a preproarginine vasopression, by using a metallothionein promoter, have a mild nephrogenic diabetes insipidus. Atrial natriuretic peptide (ANP) is released from the heart in response to cardiac atrial stretch. This peptide is also involved in sodium balance. It is highly concentrated in the preoptic and median eminence areas within the CNS. A 10-fold overexpression of ANP by using a transthyretin promoter in transgenic mice resulted in lowered blood pressure without affecting the electrolyte balance. ANP has been implicated to mediate its natriuretic and vasorelaxant effects through the guanylyl cyclase-A (GC-A) receptor. Approximately half of all humans with essential hypertension are resistant to salt. However, a gene defect for this phenotype has not been described. Recently, GC-A gene knockout mice have been generated. As a result of this mutation, these mice have chronic elevations in blood pressure when placed on a normal diet, and even in response to either minimal or high-salt diets, the blood pressure remains elevated and unchanged. Aldosterone and ANP concentrations are not affected in these mice. Therefore, mice with GC-A mutations can be useful models to study some salt-resistant forms of essential hypertension in humans. The renin-angiotensin system is a major regulator of blood pressure and sodium and volume homeostasis. Transgenic mice that overexpress the rat or human renin or angiotensin genes develop high blood-pressure levels. Recently, knockout mice have been generated to systematically address the issues of hypertension in relation to the angiotensin system. Mice carrying one, two, three, or four functional copies of the murine angiotensinogen gene at its normal chromosomal location have been created by gene targeting. These mice demonstrate plasma angiotensinogen levels corresponding to the number of gene copies they carry (i.e., mice with four copies show 145% elevated levels compared to one-copy animals). Accordingly, the blood pressures also show significant and linear increases in each of these mice. These elegant experiments have established a direct causal relationship between angiotensinogen genotypes and blood pressure. Atherosclerosis and other lipid-associated disorders are very common in humans. There are four major types of plasma lipoproteins—chylomicrons, very low-density lipoproteins (VLDL), intermediate density, and low-density lipoproteins (IDL and LDL), and high-density lipoproteins (HDL). These are synthesized either in liver or intestine. The lipid-free protein components of plasma lipoproteins are termed apolipoproteins. These are represented by five major types (A, B, C, D, and E), and some of them also can be categorized further into subtypes, for example, A-I, II, IV and C-I, II, III. The coordinated network of interactions between these complexes and the corresponding receptors regulates cholesterol homeostasis in blood. To mimic human atherosclerosis in mice, several candidate genes have been either overexpressed or knocked-out in transgenic mice. Apolipoproteins A-I (apo A-I) is the major protein associated with HDL. It is synthesized in the liver and small intestine. Apo A-I serves as a cofactor for the enzyme lecithin-cholesterol acetyltransferase that catalyzes cholesterol ester formation. Human individuals deficient in apo A-I because of various types of mutations are more prone to developing atherosclerosis. The human apo AI transgene has been genetically transferred into the atherosclerosis-susceptible inbred mouse strain C57BL/6. This

103

results in elevated levels of apo A-1 and HDL and correction of atherosclerosis. Mice that are deficient in apo A-1 have low levels of HDL cholesterol, but interestingly, do not develop atherosclerosis when supplemented with an atherogenic diet. This suggests that low levels of HDL may only predispose, but low HDL does not by itself, accelerate atherosclerosis. Transgenic mice that overexpress apo E do not show atherosclerotic plaques, even when fed a high-fat diet. In contrast, apo E-deficient mice accumulate VLDL and remnant particles in plasma. Young mice lacking apo E that are fed a low-fat diet develop atherosclerosis similar to that seen in humans. Transgenic mice that overexpress dominant apo E mutants have also been generated. These types of mutations are associated with a dominant form of type III hyperlipoproteinemia in humans. Mice overexpressing these mutant forms of apo E develop hypertriglyceridemia and hypercholesterolemia. Finally, the “familial hypercholesterolemic mouse” has been developed by knocking-out the LDL-receptor gene. These mice have elevated cholesterol levels and demonstrate delayed plasma clearance of VLDL and LDL. When placed on an atherogenic diet for prolonged time, the mice develop extensive fibroproliferative atherosclerosis and thus are a useful model to study human disease. In contrast, high levels of LDL-receptor expression in transgenic mice results in eight times faster clearance of blood LDL than normal. This has exciting possibilities of genetically correcting some human familial hypercholesterolemia syndromes caused by LDL-receptor defects. Mutant mice with gene deletions in the endothelial receptor tyrosine kinase tek die in utero. These mice have defects in the integrity of endothelium and show abnormalities in heart development. These in vivo studies have identified that the tek signaling pathway plays a critical role in the differentiation, proliferation, and survival of endothelial cells in the mouse embryo. Vascularsystem defects are the major phenotypic characteristics in mutant mice that lack ras GTPase-activating protein. The ras family consists of small guanine-nucleotide-binding proteins that are known to participate in normal- and oncogenic-signaling pathways. Four GTPase-activating proteins (GAPs) specific for ras proteins have been identified. These function as negative modulators of ras activity. One of the proteins, p120-ras GAP, interacts with various proteins, including tyrosine kinases, through its src-homology 2 (SH2)-binding domains. Mutant mice that lack this important biological signaling protein die by embryonic day 10.5. These embryos have a reduced number of somites and show a failure in reorganization of endothelial cells into a vascular network. The mutant embryos also exhibit a thinning of the dorsal aorta and aberrant ventral branches. In later stage mutants, local ruptures in the embryonic vasculature result in leakage of blood into the body cavity. At this stage, the pericardium is distended. These mutant mice have provided important insights into the understanding of the molecular mechanisms involving the interactions between SH signaling proteins and downstream targets of tyrosine kinases. RENAL BIOLOGY Kidney development requires the coordinated differentiation of two distinct tissues, the ductal epithelium and the nephrogenic mesenchyme. Both of these tissues are derivatives of the early embryonic-intermediate mesoderm. The ductal epithelium gives rise to the genital tracts, ureters, and collecting duct system. The mesenchymal components undergo epithelial transformation to form nephrons, the functional unit of the kidney. The nephron consists of the renal corpuscle and proximal and distal tubules, which are derived from the metanephric blastema.

104

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Mutations in genes expressed during kidney development have been created using transgenic mice technology. The phenotypes of these mutant mice, in many cases, resemble many of the well-known human renal disorders. The most striking phenotype of the Wilms’ tumor (WT)-1 gene knockout mouse is the absence of kidneys. WT-1 encodes a protein with a proline/glutamine rich N-terminal domain and with four Cys-His zinc fingers at the C-terminal domain. Germline mutations in WT-1 gene are associated with human urogenital malformations and a childhood kidney tumor called Wilms’ tumor. Mutant mice that are deficient in WT-1 protein die at embryonic day 11.5. At this stage, the metanephric blastema cells undergo cell death, and the embryos show a failure in ureteric bud growth and an absence of the inductive events, which lead to the formation of a definitive kidney. Whereas human individuals with the WAGR (Wilms’ tumor, aniridia, genitourinary, and mental retardation) syndrome, with retention of one functional copy of the WT-1 gene, are predisposed to develop kidney tumors, WT-1 heterozygous mutant mice are normal. This situation is reminiscent of the absence of retinoblastomas in Rb-heterozygous mice as described above. Abnormal kidney development is also seen in mutant mice that lack either the ligand, platelet-derived growth factor (PDGF) B or its cognate receptor PDGFβ-R. In both cases, the kidney glomerular tufts do not form because of the absence of mesangial cells, and there are also cardiovascular defects. Failure of kidney mesenchymal-epithelial interactions is also obvious in bone morphogenetic protein-7 (BMP-7) mutant mice. BMP-7 is a member of the TGF-β superfamily. BMP-7 is expressed in the kidney at the time of the earliest inductive event. BMP-7–deficient mice die within 48 h of birth as a result of small dysgenic kidneys with hydroureters. The kidneys show a severe reduction in the number of glomeruli, lack identifiable metanephric mesenchyme, and do not show any evidence of glomerular formation in the cortical region. At embryonic day 14.5, all mesenchymal derivatives are absent in the mutant embryos. The expression of the early markers in kidney development, for example, WT-1 and Pax-2, are either reduced or completely absent in BMP-7–deficient embryos. These studies have thus identified BMP-7 as an early glomerular inducer. The BMP-7 mutant mice may also be a useful animal model to dissect the genetic pathway for several human genetic diseases accompanied by impairment of glomerular formation. Recently, a prostaglandin synthase-2 gene knockout mouse model has been created that accurately mimics the human congenital renal disease called oligomeganephronia. This condition is characterized by small kidneys with reduced number of nephrons. The glomeruli and tubules are hypertrophied, and the kidneys eventually develop focal segmental glomerular sclerosis. These mice will be important models for studying human oligomeganephronia. Transgenic mice have been generated using glomeruli-specific promoters driving the expression of SV40 T antigen. These mice develop tumors of the glomeruli, and cell lines derived from these tumors are useful for studying renal-cell biology in vitro. Transgenic mice that express a constitutively activated allele of the human c-erbB2 growth factor receptor gene under the regulatory control of a MMTV-LTR promoter develop hyperplastic lesions in kidney, in addition to defects in other organs. In the kidney, multifocal, severe hyperplasia of renal tubules appear that may lead to microcyst formation. Within the tubules, the epithelial cells show atypical proliferation into the intraluminal compartment. The morphology of these cells also shows dramatic changes, with increased cell volume and enlarged nuclei.

Renal cysts also appear in transgenic mice that chronically overexpress TGF-α under the control of a metallotheionein (MT) promoter. TGF-α is a member of the epidermal growth factor (EGF) family of proteins. Like EGF, TGF-α binds to the EGF receptor and elicits its cellular function. In human autosomaldominant polycystic kidney disease, EGF and its stimulation have been implicated to cause this renal pathology, and thus the MT–TGF-α transgenic mice may be mimicking the human pathology. NEUROBIOLOGY One of the challenging areas of research in present-day modern biology is to understand the development of the mammalian nervous system. Many naturally occurring mutant mice provide excellent models to study human disorders associated with this complex system. The identification of many neuronal cell-specific genes that encode receptors, neurotransmitters, growth factors, surface antigens, and structural proteins has resulted in generation of a vast array of transgenic and knockout mice strains. These aid in understanding many neurological diseases like Prion disease, Alzheimer’s disease, poliomyelitis, many myelination disorders, leukodystrophy, Amylotropic lateral sclerosis, peripheral neuropathy, Down’s syndrome, Parkinsonism, and many others. Transgenic mice that express viral oncogenes in specific neural-cell types are excellent sources of generating immortalized cell lines. In addition, the development of mouse models has made it possible to probe into the early developmental aspects of the nervous system. Generation of transgenic and knockout mice for neurotrophins and their receptors has helped to obtain a clearer picture of the factors necessary for survival of many types of neurons. Transgenic and mutant mice are also currently being extensively used to explore the genetic basis of complex behavioral patterns and cognitive functions such as learning and memory. Numerous mouse models are already available for understanding the development and function of special sense organs; for example, the eye, ear, and skin. Transgenic-mouse models that mimic many human neurodegenerative disorders have been developed. We will give examples of mouse models for neurodegenerative disorders in the following section. Amyotrophic lateral sclerosis (ALS) in humans is a degenerative disease of motor neurons. It is characterized by accumulation of neurofilaments in the perikarya and proximal axons. Neurofilaments (NF) belong to the intermediate filament family, are approximately 10 nm in diameter, and show polymorphism depending on the cell type. Transgenic mice that overexpress the human NF-heavy-subunit gene (NF-H) show neuropathological symptoms close to those seen in human patients. These include neuronal and axonal swellings of α motoneurons and dorsal root ganglion cells, distal axonopathy, and secondary muscular atrophy. The L5 ventral roots from NF-H transgenic mice at old age reveal massive degeneration of large axons derived from spinal motor neurons. Metabolic labeling experiments and ultrastructural analyses on these mice have provided valuable insights into the mechanism of neurodegeneration resulting from disorganized neurofilaments. Most importantly, there are dramatic defects in axonal transport of neurofilaments, tubulin, and actin. Ultrastructural analysis on these mice indicate significant reduction of mitochondria in the degenerating axons. Because this affects the energy metabolism, neuropathy may result in these mice. Similar phenotypic characteristics are also apparent in some human familial ALS cases induced by mutations in the superoxide dismutase gene SOD-1 and in transgenic mice that express a mutant form of

CHAPTER 10 / TRANSGENIC MICE AS MODELS OF DISEASE

human SOD-1. The NF-H transgenic-mouse model, therefore, provides an important assay system to test drugs that are capable of downregulating neurofilament expression and consequently help to devise therapeutic approaches to treat human ALS. Huntington’s disease (HD) is a dominant neurodegenerative disorder. Typical symptoms of the disease include chorea, psychiatric alterations, and intellectual decline. The disease results from the presence of an expanded stretch of CAG trinucleotides (37–40 U) in one copy of the gene encoding huntingtin. It is a ~350-kDa cytoplasmic protein with unknown biological function, found in fetal and adult peripheral tissues and nervous system. Because the CAG repeat, when translated, elongates the N-terminus of huntingtin protein by a polyglutamine segment, it may reduce the normal activity of the protein. However, humans with one copy of the normal gene inactivated because of translocations do not develop HD. Similarly, a loss of function or a gain of function mutation in huntingtin (Hdh) leading to HD may also exist. To assess these possibilities, the mouse Hdh gene has been inactivated by gene targeting. Heterozygous mice are viable, fertile, and display no phenotypic abnormalities, similar to individuals with translocations. The homozygous Hdh mouse embryos implant normally, exhibit abnormal gastrulation at embryonic day 7.5, and resorb by day 8.5. These results suggest that Hdh is critical for embryonic development in mice, before the emergence of neural tube, presumably between embryonic days 8.0 and 8.5. Because the Hdh inactivation does not lead to the development of Hd neuropathology, this suggests that the human disease involves a gain of function. These experiments now open up the possibility of generating an accurate model of HD by producing transgenic mice that carry the Hdh with an expanded polyglutamine segment. A similar strategy has recently been adopted by generating a mouse model for another neurodegenerative disorder, called spinocerebellar ataxia-1 (SCA-1). SCA-1 is an autosomal-dominant inherited adult-onset human neurologic disorder characterized by loss of Purkinje cells in the cerebellar cortex and neurodegeneration within the brain stem and spinocerebellar tracts. Both the human and mouse SCA1 genes have been isolated and genetically mapped. Whereas the human gene normally has 6–40 CAG repeat units, the mouse homolog has only two CAGs. The human gene encodes a protein called ataxin-1 with 792–826 amino acids, depending on the number of glutamine residues. It is expressed in a variety of both neuronal and nonneuronal cell types and can be localized to nucleus and cytoplasm, respectively. In contrast, in cerebellar Purkinje cells, ataxin-1 has both nuclear and cytoplasmic localization. To study the intergenerational stability of trinucleotide repeats in mice, the human SCA-1 gene with normal or an expanded CAG tract containing 30 or 82 repeats, respectively, has been introduced into the mouse germline under the regulatory control of Purkinje cell-specific gene Pcp2 sequences. Several lines of heterozygous mice carrying the human transgene have been obtained. Some lines are bred to homozygosity. Transgenic mice that carry the normal human SCA1 allele are neurologically normal even after 1 year of age, indicating intact Purkinje cell function. In contrast, as early as 8–10 weeks of age, transgenic mice that carry the expanded repeats show slightly reduced cage activity, a gentle swaying of the head while walking, and early signs of general incoordination. By 16–30 weeks, these mice become clearly ataxic when walking. The onset of the neurological abnormalities and ataxia phenotype vary in independent 82-repeat-transgenic lines, without any correlation to the mRNA

105

levels of the transgene. Those lines of mice that show earliest onset of ataxia tend to have the most severe neuropathologic disease. Typical phenotypic characteristics include significant loss of Purkinje cell population, with Bergmann glial proliferation, and shrinkage and gliosis of the molecular layer. These mice also show numerous ectopic Purkinje cells in the molecular layer, sometimes in the granular layer, and abnormal dendritic arrays. Heterozygous 82-repeat-transgenic mice do not show ataxia symptoms but display only mild cerebellar pathology. These mice do not show loss of Purkinje cells or gliosis, but exhibit only occasional Purkinje cells in both the molecular and granular layers. In humans, CAG repeats in the expanded range are subject to intergenerational instability, but those in the unexpanded range are stable. Eightytwo CAG repeat-containing SCA-1 transgenic mice show no variation in repeat length with parent-to-offspring transmission. These results suggest that the mechanisms of DNA repair, replication, and the higher order structure of chromatin may be important factors in determining the CAG repeat instability in humans. Irrespective of the molecular mechanisms that control CAG repeat instability, this in vivo transgenic approach has successfully resulted in the generation of a mouse model that phenocopies the human SCA-1 disorder. These mice may also be amenable for further manipulation to devise therapeutic strategies and for targeting SCA-1 transgene expression to different tissues within the CNS. In addition, loss of function mutations can be engineered into the murine SCA-1 locus to explore its biological function during normal development of the mouse. Po is a cell-adhesion protein that belongs to the immunoglobulin (Ig) superfamily of recognition molecules. It contains only one variable-like Ig-related extracellular domain. Members of this family have been implicated in cell migration, axon outgrowth, fasciculation, and myelination. The other mammalian members of this neural-cell-adhesion molecule superfamily include neuralcell-adhesion molecule (N-CAM) proteins, L1, and MAG. Each of these molecules can exhibit both homophilic or heterophilic interactions, and their signal transduction pathway includes regulation of cytoskeletal and second-messenger systems. Po is uniquely expressed by myelinating Schwann cells of the peripheral nervous system during the first 3 weeks after birth and accounts for almost 60% of the protein in the peripheral myelin sheath. These characteristic features have been taken into consideration in generating the Po-deficient mouse model. The mutant mice exhibit lack of normal motor coordination, tremors, and occasional convulsions. The peripheral nerve axons show severe hypomyelination and degeneration. Many of the molecules involved in myelination are abnormally regulated in Schwann cells of the mutant mice. These results suggest that Po is essential for the normal spiraling, compaction, and maintenance of the peripheral myelin sheath and for the continued integrity of associated axons. In addition, these Po mutant mice may help to decipher the molecular mechanisms of genetically transmitted peripheral neuropathies in humans. Although the above-mentioned mouse models mimic human neurodegenerative disorders, the initial efforts to develop a useful transgenic mouse model for the fatal Alzheimer’s disease have not been successful. However, several attempts have been made to develop such a mouse model. Alzheimer’s disease (AD) is an irreversible neurodegenerative disorder that affects the human CNS. The disease is characterized by massive deposition of fibrillary aggregates, intracellularly as neurofibrillary tangles, extracellularly as amyloid plaques. Patients with Down’s syndrome

106

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

also show similar pathological symptoms. Principally, the plaques are deposited in parenchyma of amygdala, hippocampus, and neocortex. The human β-amyloid protein (Aβ) is derived from the amyloid precursor protein (APP), encoded by at least 18 exons located on chromosome 21. Four distinct transcripts of APP premRNAs are generated by alternate splicing. The rationale for developing an APP overexpressing mouse model is based on the fact that AD may be initiated or accelerated by high levels of APP. In addition, high levels of APP have been observed in Down’s syndrome patients. Unfortunately, most of the transgenic mice that harbor several copies of human APP cDNA do not show any pathological symptoms of AD. Only in one case, in which the expression of a human APP cDNA is driven by a mouse metallothionein promoter, have behavioral changes been noticed. Because in many instances, the introduction of foreign cDNA does not result in high levels of transgene expression, a different strategy has been used to derive APP transgenic mice. A yeast artificial chromosome (YAC) containing 400-kb human APP gene and approximately 250 kb of flanking sequences has been transfected into mouse ES cells. These cells have been used to generate transgenic mice that retain the APP-YAC transgene with stable integration and transmission. These mice show high levels of human APP mRNA and protein in brain and peripheral tissues. In particular, human APP is expressed in cell bodies of neurons in the hippocampus, neocortex, and cerebellum in a 3-month-old APP YAC transgenic mouse. Surprisingly, this mouse also does not show any AD-type pathology. While transgenic mice that express human APP cDNAs or gene (through YAC) do not show any AD symptoms, recently two other mouse models have been generated to understand the AD pathology and to explore the normal function of APP in mouse. The first model is a transgenic-mouse model in which the mouse Aβ expression is restricted to only neuronal cells by using 1.8 kb of 5' flanking sequences from the mouse neurofilament-light (NF-L) gene. This promoter is transcriptionally active throughout adult life. Several independent lines overexpressing this NF-L–transgene have been generated. Although transgene expression is also seen in peripheral tissues, only brains from these mice show pathological changes. Similarly, in humans, even though Aβ deposits are seen in skin and intestine without much clinical consequence, the CNS is the main target of the APP toxicity. At least three independent lines of transgenic mice show extensive neuronal degeneration. In situ hybridization analysis suggests that apoptosis is the predominant cause for this degeneration. These mice also show reactive gliosis. Over 50% of the transgenic mice die by 12 months of age. However, the mice do not show neurofibrillary tangles and senile plaques characteristic of human AD. It may be possible that factors required for plaque formation in humans may be absent in mice. Nevertheless, this model is important in understanding the APP toxicity and the mechanisms of apoptosis that lead to neuronal-cell death and degeneration. Lastly, to understand the physiological function of APP, the mouse APP gene was disrupted by insertion of a premature stop codon in exon 2 in ES cells, and mutant mice have been generated from these ES cells. Homozygous mutant mice continue to express the APP mRNA in brain and other tissues, although at 5- to 10-fold lower levels than wild-type mice. This could possibly result from skipping of the disrupted exon during splicing. A shortened β-APP–specific protein has also been detected in low levels in different regions of brain. These mutant mice show severe

impairment of spatial learning and exploratory behavior. Most strikingly, these mice show an increased incidence of agenesis of the corpus callosum. Although the original goal of generating a true APP null mouse has not been produced here, this model does help to understand the normal physiological function of APP during mouse development. ENDOCRINOLOGY Mouse models for human hypothalamic-pituitary disorders, thyroid disorders, pancreas disorders, and adrenal disorders have, in general, been useful for analysis of the endocrine function in mammals. Transgenic mice that overexpress either growth hormone (GH) or growth hormone-releasing hormone (GHRH) are excellent models for both acromegaly and hyperprolactinemia, common endocrine disorders in humans. MT-GH transgenic mice are historically one of the earliest mouse models developed and were used to correct a genetic defect in the dwarf little mouse. The little mouse, a naturally occurring mouse strain, shows a deficiency of GH, which causes its short stature. Several MT-GH fusion genes have been injected to generate transgenic mice that overexpress growth-hormone mRNA in several tissues and consequently elevated serum-GH levels. These mice show accelerated growth rates. When the transgene is introduced into the “little” background, the phenotype is rescued, and fertility is also restored. Using cell ablation techniques in transgenic mice, the development of somatotropes, lactotropes, and thyrotropes can be suppressed. These experiments have clarified the lineage relationships and identification of common progenitor cells for these types of anterior pituitary cells. A similar strategy has also been helpful to study proopiomelanocortin (POMC) producing corticotropes and melanotropes. Transgenic mice that express the SV-40 T antigen oncogene, driven by a human α-glycoprotein hormone or a rat POMC promoter sequences, develop anterior and intermediate lobe of pituitary tumors, respectively, and have provided novel cell lines to study the function of these cell types. Transgenic mouse models for analyzing the etiology of human type-I diabetes have been useful in the validation of the current hypothesis on the human disorder. Essentially different types of transgenes have been targeted to β cells of the pancreas to selectively destroy these cells. Transgenic mice in which the autoimmune response is evoked in pancreatic β cells by induction of MHC antigens develop diabetes accompanied by a selective loss of these cells. To mimic the typical autoimmune response in β cells, transgenic mice have been created that express a viral protein in pancreatic islet cells. Because the virus shares some antigenic epitopes common to mouse proteins, the immune response generated against the viral proteins also destroys the β cells. As a result, these mice develop diabetes. Although this model is useful in understanding the pathogenesis of human typeI diabetes, it differs from the human disease in two aspects. First, the presence of virus is required to elicit the response in mice, whereas in humans, the autoimmunity is maintained by endogenous β-cell-specific proteins. Second, the β cells are destroyed at a rapid rate in the mouse model unlike in the human situation, where the disease process is rather prolonged. Finally, transgenic mice that express a rat-insulin promoter-SV40 oncogene sequences develop heritable pancreatic β-cell tumors. They have β-cell hyperplasia and elevated serum insulin levels. The α and δ cells of the islets of Langerhans are disordered, and their number is drastically reduced in the pancreatic tumor tissue. These mice develop highly vascularized solid pancreatic tumors that cause

CHAPTER 10 / TRANSGENIC MICE AS MODELS OF DISEASE

premature death of the mice around 9–12 weeks of age. This animal model has become invaluable in studying the biogenesis of islet cells, cell–cell interactions within the endocrine pancreas, and understanding of the pathobiology of human insulinomas. The three pituitary glycoprotein hormones—namely, luteinizing hormone (LH), follicle stimulating hormone (FSH), and thyroid stimulating hormone (TSH)—are heterodimeric proteins. They share a common α-subunit that is noncovalently linked to a hormone-specific β-subunit. Whereas LH and FSH control gonadal functions, TSH regulates thyroid function. Mice deficient in the glycoprotein hormone α-subunit (α-GSU) have been generated by gene targeting in ES cells. These mice are viable, but exhibit growth insufficiency and are hypothyroid and hypogonadal. The thyroids are hypoplastic and contain small disorganized follicles. The animals are infertile and show lack of some secondary sexual characteristics, although the sex organs differentiate normally. Within the pituitary, the thyrotropes exhibit hypertrophy and intensely stain for TSH-β subunit antibody. The gonadotrope cell population is normal, whereas both somatotropes (GH expressing cells) and lactotropes (prolactin-expressing cells) are reduced and absent, respectively. Absence of α-GSU does not affect the distribution and appearance of gonadotropin-releasinghormone (GnRH) neurons that migrate to different regions of the hypothalamus. Experiments with these mice have thus established the physiological functions of these important trophic hormones in mammalian growth and differentiation. REPRODUCTIVE BIOLOGY The hypothalamic-pituitarygonadal axis, commonly known as the reproductive axis, is regulated by a complex network of interactions between neuropeptides, trophic hormones, gonadal steroids, and peptides and growth factors. Distinct genes encoding several critical factors are activated during the establishment of the primitive reproductive axis, functional maturation of the gonads, and maintenance of their differentiated function. Disruption of any key step may lead to aberrant reproductive function and may even cause sterility as demonstrated in the analysis of several transgenic and knockout mice. Still, many mouse models are currently being generated to understand human infertility and associated disorders to test and design novel strategies for contraception. These many transgenic models have been extensively summarized elsewhere, and we will only discuss a few models related to sex determination. Mammalian sex differentiation is dependent on interactions between several factors. The early molecular events that dictate the choice between the pathways of male and female sex-organ development in mammalian early embryo are poorly understood, including those in humans. Several genetic lesions have been characterized in humans that affect this complex pathway. Successful application of the transgenic-mouse technology in recent years has given important clues toward identifying functions of regulatory molecules that control the sexual-differentiation process. Most important among these are transcription factors including WT-1 (described in an earlier section) and steroidogenic factor-1 (SF-1), SRY, and Mullerian inhibiting substance (MIS), a member of the TGF-β superfamily. Gene deletions of WT-1 or SF-1 by homologous recombination in ES cells and generation of knockout mice show drastic phenotypes. Absence of either WT-1 or SF-1 results in failure of development of the gonads. Transgenic mice that express SRY, a testis-determining gene, demonstrate testis formation in XX females. The function of MIS in mammalian reproductive development has been elucidated both by gain-of-

107

function transgenic mice that chronically overexpress MIS and by an MIS knockout mouse model. Absence of MIS in males prevents regression of the Mullerian duct, and therefore male mice develop uteri and testicular Leydig cell hyperplasia. METABOLIC DISORDERS Metabolic defects most often result from an accumulation of a substrate because of absence or reduced activity of an enzyme, often causing formation of a toxic side product or absence of an important product for a subsequent enzyme reaction. Such defects can cause life-threatening problems during childhood or adulthood. There are many transgenic-mouse models available to investigate the pathogenesis and to evaluate therapeutic approaches for such human diseases. Some of the transgenic mice that are useful models in understanding these human metabolic disorders are discussed below. The Lesch-Nyhan syndrome in human males is a rare neurological and behavioral disorder. The disorder is caused by an inherited deficiency in the level of activity of the X-chromosome–linked purine salvage enzyme hypoxanthine-guanosine phosphoribosyl transferase (HPRT). Hemizygous mutant male mice have been generated that carry a mutation in the HPRT gene. Interestingly, these mice are viable and normal. No gross neurological symptoms have been apparent in male mice. Even though these mutant mice do not show any metabolic, neurological, or behavioral symptoms characteristic of the human syndrome, the basic principles of gene-targeting technique were originally developed during the creation of this animal model. Urate oxidase, or uricase, is a purine metabolic enzyme. The enzyme catalyzes the conversion of uric acid to allantoin in most mammals. In humans and other primates during evolution, deleterious mutations in the urate oxidase gene resulted in the loss of this enzyme. Therefore, the sparingly soluble uric acid is the end product of purine metabolism in humans, instead of a more soluble allantoin. Because humans lack urate oxidase, they are predisposed to metabolic disorders, such as gouty arthritis and renal stone formation, as a result of elevated uric acid levels. To understand the biological function of urate oxidase in lower mammals and to develop an animal model for human hyperuricemia, knockout mice have been created by gene-targeting techniques. Approximately 65% of mice deficient in urate oxidase die at 3–4 weeks of age. Both serum and urinary uric acid levels are elevated 10- and 5-fold higher, respectively, in mutant mice compared to wild-type control mice. Consistent with the biochemical action of the enzyme, mutant mice show a decrease in urinary allantoin content. Urate oxidase-deficient mice show small cortical cysts and white yellow deposits of uric acid in both kidney cortex and medulla as early as 6 days after birth. With the progression of the deposition, the mice show kidney tubular degeneration leading to severe hydronephrosis at 5 weeks. This eventually results in glomerular atrophy and occurrence of chronic inflammation within the kidney interstitium, similar to acute hyperuricemia nephropathy in humans. Similar to human patients, urate oxidasedeficient mice, when treated with allopurinol, an inhibitor of xanthine oxidase, show dose-dependent reduction in serum uric acid levels. Although urate oxidase is absent in humans, most humans do not develop hyperuricemia at a young age and do not show increased mortality, unless in conjunction with other environmental and genetic factors. However, generation and analysis of urate oxidase knockout mice resolved the long-standing debate on the biological function of urate oxidase in mice and other lower mammals that retain it.

108

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

DERMATOPATHOLOGY The application of transgenic technology to skin research has resulted in many advances in our understanding of skin development and function. The skin is a complex organ system. It consists of two different layers, epidermis and dermis, derived from ectoderm and mesoderm, respectively. Cells of the outermost layer of the epidermis, called periderm, continually undergo keratinization and are replenished by cells arising from the basal layer that lies beneath. The fully developed epidermis consists of five distinct layers—stratum corneum, stratum lucidum, stratum granulosum, stratum spinosum, and stratum germinativum. The epithelial cells of the epidermis are called keratinocytes. They produce important proteins that regulate the epidermal–dermal interactions. The epidermis has been the major target for several gene-manipulation studies. The epidermis exerts its protective function through an extensive cytoskeletal network. The major structural components of this network are called keratins. Keratins are members of the intermediate filament (IF) superfamily. They are α-helical proteins that can assemble into 10-nm filaments. Based on amino acid sequence, keratins can be subdivided into two groups—the acidic, type I, keratins and the basic, type II, keratins. Type I keratins include K9–K20 and are in the molecular-weight range of 64–40 kDa, whereas type II keratins include K1–K8 and weigh 67–53 kDa. Heteropolymeric combinations of type I and type II can result in the formation of 10-nm filaments. Proliferating basal cells express K5 and K14 mRNAs, and they differentiate in the first suprabasal layer, inducing the expression of K1 and K10. During the wound-healing process and in cultured epidermal cells, K6 and K16 are induced. Many of the human and mouse keratin genes and other epidermal-specific genes have been isolated and characterized in detail. This structural analysis has led to the design of many transgenic and knockout mice to understand cutaneous development and functions including hair development and skin cancer. Epidermolytic hyperkeratosis (EHK) in humans is a dominantly inherited disorder of keratin, characterized by hyperplasia of the stratum corneum. Mutations in highly conserved regions of K1 and K10 genes have been identified in patients with EHK. Transgenic mice expressing a mutant form of K1 have been generated. Most of the homozygous mice with two copies of the mutant transgene die within 24–36 h after birth resulting from progression of skin erosions. Similar to EHK patients, the spinous layer shows many clear cells that have clumping of keratin filaments around the nucleus. Mice that survive to the juvenile and adult stage do not show suprabasal epidermal splitting (blistering) in their back skin, but their footpad epithelium, which lacks hair follicles, show cell lysis and splitting in the spinous layer, similar to human EHK disorder. Hyperkeratosis is also seen in mice deficient in follistatin, an activin-binding protein. Both granular and stratum corneum layers show increased thickness in these follistatin-mutant mice. Psoriasis is a hyperproliferative inflammatory skin disorder in humans. Approximately 2% of the world’s population is affected by this skin disorder. Transgenic mice that express either human integrin β1-subunit or heterodimers of β1- and α2- or β1- and α5-subunits in the suprabasal layers of epidermis exhibit a phenotype that resembles the human disorder. Integrins are extracellular matrix receptor proteins; three types are expressed by epidermal keratinocytes. They share a common β-subunit. The α2β1 complex is a collagen receptor, and the α3β1 is a laminin receptor; both of these are expressed in normal epidermis. The α5β1 is the keratinocyte fibronectin receptor and is upregulated during wound-

healing, in psoriasis, and in cultured keratinocytes. Normally, integrin expression is confined to the keratinocytes in the basal layer of the epidermis. Since their expression is often seen in suprabasal layers during wound-healing and in psoriasis, integrin subunits are targeted to this layer in transgenic mice, using a suprabasal layer cell specific promoter called involucrin. Several lines of mice have been obtained that express each of the integrin subunits, α2, α5, and β1, separately. Subsequently, β1 transgene-bearing mice have been mated to α2 or α5 mice to generate mice that express the heterodimers. A majority of the α5, β1, and α5β1 transgenic mice have defects in eyelid closure, hair, and whisker development. The eyelids noticeably fail to fuse prior to birth, and often an inflammatory exudate in the cornea and eyelid is apparent. The hair of the coat does not display a uniform orientation, with disorganized follicles, and the whiskers are short and curly. These phenotypic features are characteristic of those seen in many other mutant mice in which the growth factor signaling is disrupted. For example, activin/inhibin βB knockout mice display eyelid-closure defects. Similarly, TGF-α and EGF receptor-deficient mice and BMP-2, BMP-4 overexpression transgenic mice exhibit abnormal skin development. In mice deficient in activin βA-subunit, the whisker follicles differentiate in a delayed manner, with abnormal hair papilla and irregularly positioned cells at the base of the hair bulb. These observations are all consistent with the fact that growth factors regulate integrin levels in keratinocytes and may explain the consequences of perturbances in integrin expression during embryonic development. The β1 alone or β1α2 or β1α5 integrin– transgenic mice also show the typical features of psoriasis (i.e., increased keratinocyte proliferation, abnormal differentiation, increased dermal mitoses, capillary dilation, and an influx of CD4+ and CD8+ T-lymphocytes). The abnormal keratinocyte development and inflammation in these mice result in flaking of the epidermis on the chin, behind the ears, reddening of skin and paws, and pustule formation on dorsal-skin surface. Consistent with the belief that psoriasis in humans is influenced by several genetic and environmental factors, all transgenic mice that express the integrin subunits do not show the symptoms. Thus, the transgenic technology has provided important answers to the understanding of the etiology of psoriasis. These mice have a potential as important models in the treatment of psoriasis using immunosuppressive agents. Recently, gene-targeting approaches have been used to create knockout mice to mimic the human disease xeroderma pigmentosum (XP). XP is a rare autosomal disease in humans, resulting from a defective nucleotide excision repair process. It is characterized by hypersensitivity of the skin to sunlight and a very high risk of skin-cancer incidence in body parts exposed to sunlight. There are eight complementation groups in XP (A–G and a variant form). Mice deficient in XPC are viable and fertile and do not exhibit an increased susceptibility to spontaneous tumor generation. However, following 10 d of UV exposure, the ears of the mutant mice exhibit marked atrophy. By 20 weeks of exposure, they develop skin cancers of different types. Most of these belong to the squamous-cell carcinoma category. These mice also show epidermal hyperplasia with focal areas of hyperkeratosis and varying degrees of dysplasia, acantholysis, and/or dyskeratosis. The UV treatment of the mutant mice also results in keratitis and corneal ulceration of the eye. The spectrum of these phenotypic features are characteristic of human XPC. Similar to XPC-deficient mice, mice deficient in XPA also are viable and fertile and are highly susceptible to chemical and UV-induced skin carcinogenesis. These mice also develop squa-

CHAPTER 10 / TRANSGENIC MICE AS MODELS OF DISEASE

mous-cell carcinoma and eye defects. These knockout mouse models are useful in two ways. First, they can be used to understand XP disorders, and cells from these mice can be used to study the molecular mechanisms of nucleotide excision. Second, sensitive short-term in vivo carcinogenicity of genotoxic agents can be tested using these excision repair enzyme-deficient mice.

CONCLUSIONS AND FUTURE DIRECTIONS Transgenic animal technology has become highly popular in recent years. The primary reasons are because this is an in vivo manipulative approach, and the specificity with which a target gene can be expressed makes it an extremely powerful approach. Virtually every physiological system has been and will continue to be investigated from a developmental and functional perspective (i.e., both physiological and pathological). In many instances, the clinical data and the naturally occurring mutant strains of mice complement the efforts on gene-manipulation research. What can we envision in future with this technology? Animal species such as goat and cattle can routinely be used as bioreactors to produce pharmacological products in high quantities. The functional redundancies or specificities between close members of a superfamily can be resolved by breeding and generating mice with combinations of mutations in distinct or unrelated pathways. Large-scale deletions that mimick many human chromosomal haploidy situations can be created in mice, and thus sets of genes on a specific chromosome can be identified structurally and functionally. The “cre-lox” system will be used in many cases to generate tissue-specific knockout mice. Many mouse models will be generated to address the genetic basis for the complex issues of behavior, learning, and memory. The potential possibilities of generating ES cells from species other than mouse and rat can be anticipated. The rapid pace with which the genome-mapping data are being accumulated in parallel will set a continued pace to identify and define the biological functions of many newly discovered genes. The final goal is to generate cheap, convenient, and easily manipulative animal models that exactly reflect many humandisease conditions. Both human health and health-care management will thus benefit richly from this fascinating technology.

SELECTED REFERENCES Aguzzi A, Brandner S, Sure U, Ruedi D, Isenmann S. Transgenic and knockout mice: Models of neurological disease. Brain Path 1994;4:3–20. Bradley A, Hasty P, Davis A, Ramirez-Solis R. Modifying the mouse: design and desire. Bio/Technology 1992;10:534–539. Burright EN, Clark HB, Servadio A, et al. SCA1 transgenic mice: a model for neurodegeneration caused by an expanded CAG trinucleotide repeat. Cell 1995;82:937–948. Capecchi MR. Targeted gene replacement. Sci Amer 1994;270:52–59. Cox GA, Cole NM, Matsumura K, et al. Overexpression of dystrophin in transgenic mice eliminates dystrophic symptoms without toxicity. Nature 1993;364:725–729. Giese KP, Martini R, Lemke G, Soriano P, Schachner M. Mouse Po gene disruption leads to hypomyelination, abnormal expression of recognition molecules, and degeneration of myelin and axons. Cell 1992;71: 565–576. Gilbert SF. Developmental Biology. Sunderland, MA: Sinauer Associates, 1994. Glasser SW, Korfhagen TR, Wert SE, Whitsett JA. Transgenic models for study of pulmonary development and disease. Am J Physiol 1994;267: L489–L497. Grandaliano G, Choudhury G G, Abboud HE. Transgenic animal models as a tool in the diagnosis of kidney diseases. Semin Nephrol 1995;15:43–49.

109

Greenhalgh DA, Roop DR. Dissecting molecular carcinogenesis: development of transgenic mouse models by epidermal gene targeting. Adv Cancer Res 1994;64:247–296. Gu H, Marth JD, Orban PC, Mossmann H, Rajewsky K. Deletion of a DNA polymerase beta gene segment in T cells using cell type-specific gene targeting. Science 1994;103–106. Hammer RE, Palmiter RD, Brinster RL. Partial correction of murine hereditary growth disorder by germ-line incorporation of a new gene. Nature 1984;311:65–67. Hanahan D. Heritable formation of pancreatic β-cell tumours in transgenic mice expressing recombinant insulin/simian virus 40 oncogenes. Nature 1985;315:115–122. Henkemeyer M, Rossi DJ, Holmyrad DP, et al. Vascular system defects and neuronal apoptosis in mice lacking Ras GTPase-activating protein. Nature 1995;377:695–701. Jones SN, Roe AE, Donehower LA, Bradley A. Rescue of embryonic lethality in Mdm2-deficient mice by absence of p53. Nature 1995; 378:206–208. Kendall SK, Samuelson LC, Saunders TL, Wood RI, Camper SA. Targeted disruption of the pituitary glycoprotein hormone alpha-subunit produces hypogonadal and hypothyroid mice. Gen Devel 1995; 9:2007–2019. Kreidberg JA, Sariola H, Loring JM, et al. WT-1 is required for early kidney development. Cell 1993;74:679–691. Kuehn MR, Bradley A, Robertson EJ, Evans MJ. A potential animal model for Lesch-Nyhan syndrome through introduction of HPRT mutations into mice. Nature 1987;326:295–298. Kumar TR, Donehower LA, Bradley A, Matzuk MM. Transgenic mouse models for tumor-suppressor genes. J Int Med 1995;238: 233–238. Lamb BT. Making models for Alzheimer’s disease. Nat Genet 1995;9:4–6. Luo G, Hofman C, Bronckers ALJJ, Sohocki M, Bradley A, Karsenty G. BMP-7 is an inducer of nephrogenesis, and is also required for eye development and skeletal patterning. Gen & Devel 1995;9: 2808–2820. Luo X, Ikeda Y, Parker KL. A cell-specific nuclear receptor is essential for adrenal and gonadal development and sexual differentiation. Cell 1994;77:481–490. Matzuk MM, Finegold MJ, Su J-GJ, Hseuh AJW, Bradley A. α-Inhibin is a tumor-suppressor gene with gonadal specificity in mice. Nature 1992;360:313–319. Matzuk MM. Functional analysis of mammalian members of the transforming growth factor-β superfamily. Trends Endocrinol Metab 1995;6:120–127. Matzuk MM, Kumar TR, Vassalli A, Bickenbach JR, Roop DR, Bradley A. Functional analysis of activins during mammalian development. Nature 1995;374:354–356. McMahon AP, Bradley A. The Wnt-1 (int-1) proto-oncogene is required for development of a large region of the mouse brain. Cell 1990;62: 1073–1085. Molkentin JD, Black BL, Martin JF, Olson EN. Cooperative activation of muscle gene expression by MEF 2 and myogenic bHLH proteins. Cell 1995;83:1125–1136. Morham SG, Langenbach R, Loftin CD, et al. Prostaglandin synthase 2 gene disruption causes severe renal pathology in the mouse. Cell 1995;83:473–482. Nishimori K, Matzuk MM. Transgenic mice in the analysis of reproductive development and function. Reviews of Reproduction, 1996;1:203–212. Paigen B, Plump AS, Rubin EM. The mouse as a model for human cardiovascular disease and hyperlipidemia. Curr Opin Lipidol 1994;5: 258–264. Palmiter RD, Brinster RL. Germ-line transmission of mice. Annu Rev Genet 1986;20:465–499. Ramirez-Solis R, Liu P, Bradley A. Chromosome engineering in mice. Nature 1995;378:720–724. Ramirez-Solis R, Zheng H, Whiting J, Krumlauf R, Bradley A. Hoxb-4 (Hox-2.6) mutant mice show homeotic transformation of a cervical vertebra and defects in the closure of the sternal rudiments. Cell 1993;73:279–294.

110

SECTION I / INTRODUCTION TO MOLECULAR MEDICINE

Sands AT, Abuin A, Sanchez A, Conti CJ, Bradley A. High susceptibility to ultraviolet-induced carcinogenesis in mice lacking XPC. Nature 1995;377:162–165. Shawlot W, Behringer RR. Requirement for Lim-1 in head-organizer function. Nature 1995;374:425–430. Smithies O, Kim H-S. Targeted gene duplication and disruption for analyzing quantitative genetic traits in mice. Proc Natl Acad Sci USA 1994;91:3612–3615. Snouwaert JN, Brigman KK, Latour AM, et al. An animal model for cystic fibrosis made by gene targeting. Science 1992;257:1083– 1088. Stewart TA. Models of human endocrine disorders in transgenic rodents. Nature 1993;4:136–141.

Strober W, Ehrhardt RO. Chronic intestinal inflammation: an unexpected outcome in cytokine or T cell receptor mutant mice. Cell 1993;75: 203–205. Taverne J. Transgenic mice in the study of cytokine function. Int J Exp Path 1993;74:525–546. Wagner J, Thiele F, Ganten D. Transgenic animals as models for human disease. Clin Exp Hyperten 1995;17:593–605. Wu X, Wakamiya M, Vaishnav S, et al. Hyperuricemia and urate nephropathy in urate oxidase-deficient mice. Proc Natl Acad Sci USA 1994;91:742–746. Zijlstra M, Bix M, Simister NE, Loring JM, Raulet DH, Jaenisch R. Beta 2-microglobulin deficient mice lack CD4–8+ cytotoxic T cells. Nature 1990;344:742–746.

111

CHAPTER 11 / MOLECULAR CARDIOLOGY

CARDIOLOGY SECTION EDITOR:

ANTHONY ROSENZWEIG

II

112

SECTION II / CARDIOLOGY

CHAPTER 11 / MOLECULAR CARDIOLOGY

113

11 Molecular Cardiology An Overview

ANTHONY ROSENZWEIG Not too long ago a textbook section entitled “Molecular Cardiology” would have seemed curiously paradoxical and impractical. Molecular biology appeared to have little to do with either clinical cardiology or even its scientific foundations. Classical large animal physiology experiments provided insights that were readily extrapolated from the laboratory to the bedside. In contrast, the clinical relevance of molecular biology seemed remote. This perception is changing because of the substantial progress made in applying the tools of molecular biology to cardiovascular biology and medicine. An imperfect measure of these changes is reflected in the number of papers published in these fields over the past 30 years. Acknowledging that the number of papers published in all fields has burgeoned, some insight may be gained from comparing the relative changes in publications related to cardiology, molecular biology, or the intersection of these disciplines (Fig. 11-1). As one might expect, the number of publications in molecular biology has increased over 30-fold during this period, which has witnessed the genesis of many important technologies under this general umbrella. The number of papers in cardiology has also increased over this period by approximately fivefold. However, the number of papers over the past 30 years within what might be termed “molecular cardiology” has increased nearly 75-fold, perhaps reflecting the growing realization of valuable connections between these two fields. Molecular biology provides a set of powerful tools that can be brought to bear on clinical and scientific problems. These tools are rapidly elucidating the molecular basis of many clinical phenomena. One example highlighted by Dr. Michel in Chapter 15 is the identification and molecular cloning of the enzymes responsible for production of endothelium-dependent relaxing factor (EDRF), now recognized as nitric oxide (NO). Absence or dysfunction of endothelial cells results in less NO production. This is recognized in the catheterization laboratory as a paradoxical vasoconstrictive response to acetylcholine that dilates normal vessels through an endothelium-dependent mechanism. Such an abnormality is seen both in atherosclerotic vessels and patients with risk factors for atherosclerosis even in the absence of angiographically demonstrable lesions. Identification of the nitric oxide synthase (NOS) genes responsible for NO production has permitted a demonstration of their role in vascular tone homeostasis. Such studies form the essential foundation for future analysis of the role of NOS and From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

other regulators of vascular tone in a variety of cardiovascular conditions, including hypertension and atherosclerosis. Molecular technologies that are particularly relevant to the clinical arena include molecular genetics, generation of genetic animal models, and somatic gene transfer. The exponential increase in available genetic markers has dramatically hastened analysis of monogenic disorders. Genetic analysis of such disorders is of immediate interest to clinicians because of its the potential to provide both diagnostic and prognostic insights. These themes are echoed repeatedly throughout this section in chapters on congenital heart disease, cardiomyopathies, arrhythmia, and hypertension. In Chapter 12, Dr. Chin reviews the growing list of genetic defects linked to congenital heart disease. Furthermore, he illustrates the importance of animal models for understanding the molecular basis of cardiac development, especially in relationship to congenital heart disease. Although standard genetic models are limited for congenital heart disease, there is great potential in the study of mutant zebrafish. Clinicians are occasionally skeptical of such studies both because the animals studied seem so far from clinical experience and because even human monogenic conditions represent only a small subset of a broad clinical spectrum of sporadic and polygenic disorders. However, both genetic animal models and mendelian conditions provide a rare window of opportunity for observations that can have much broader relevance. In addition, identification of disease-causing mutations can illuminate unanticipated molecular functions or interactions. Admittedly most of the cardiovascular diseases encountered routinely by clinicians result in fact from a complex interaction between multiple genetic loci and superimposed environmental influences. Using hypertension as an example, Drs. Koike and Jacob (Chapter 16) illustrate evolving approaches to understanding the genetic bases of polygenic disorders. Surprisingly, little is known about the genetic basis of most coronary artery disease, despite the prevalence of the phenotype. It is likely that approaches similar to those described by Drs. Koike and Jacob will shed light on the genetic bases of this problem as well. Somatic gene transfer is of interest not only as a potential experimental tool for the development of local overexpression models, but also as a vehicle for somatic gene therapy. Gene therapy within the cardiovascular system has been broadly defined to include not only replacement of a defective gene but also overexpression of a beneficial gene product. This approach to gene therapy within the cardiovascular system differs from gene

113

114

SECTION II / CARDIOLOGY

Figure 11-1 Relative number of papers published 1966–1995. The MEDLINE database was searched using BRS Colleague to identify papers published between January 1966 and December 1995. Papers were identified as relating broadly to cardiology (contain the word “heart” or the root “cardi-”), molecular biology (contain the wordroot “gene-” or “molecul-”), or molecular cardiology (falling into both categories). For each group, the number of papers was normalized to the initial reference period (1966–1970), which is arbitrarily assigned the value 1. The number of papers in subsequent 5-year periods is displayed relative to this initial reference period. 䊏, cardiology; 䉬, molecular, 䊉, molecular cardiology.

replacement therapy in other disciplines in several important ways. Access to the target tissue is often more difficult than in other systems, and may require specific catheter delivery systems. In addition, the target cells for gene transfer within the cardiovascular system (endothelial cells, vascular smooth muscle cells, or cardiomyocytes) replicate either slowly or not at all and are notoriously resistant to gene transfer techniques. These limitations have sparked interest in viral vectors to achieve acceptable efficiencies in vivo. On the positive side, gene therapy for several potential cardiovascular applications (such as restenosis) has the advantage that transient expression of beneficial gene products may be sufficient to mediate significant clinical benefit. Drs. Vassalli and Dichek discusses the status of cardiovascular gene therapy in Chapter 18. The potential power of molecular approaches to clinical cardiology is perhaps best illustrated in the evolving story of hypertrophic cardiomyopathy, recounted by Dr. Seidman et al. in Chapter 13. This clinically important, often familial condition has been recognized for approximately 30 years. However, the fundamental etiologies remained obscure. The Seidman laboratory first established genetic linkage for this condition in several large affected families. They then went on to identify disease-causing mutations in multiple genes, including cardiac myosin, troponin T, and tropomyosin. This work has had important implications for our understanding of hypertrophic cardiomyopathy and basic muscle biology. The clinical prognosis of patients with this condition is dramatically affected by the specific mutation they carry. As one might expect, more conservative mutations, in general, carry a better prognosis than more radical alterations. In addition, these genes all participate in forming the sarcomere, and elucidat-

Figure 11-2 Absolute number of papers published 1966–1995. The MEDLINE database was searched as described for Fig. 11-1. Here the absolute number of papers published in each 5-year period is shown graphically for each of the three categories as defined previously. 䊏, cardiology; 䉬, molecular; 䊉, molecular cardiology.

ing the genetic abnormalities responsible for the clinical phenotype will likely continue to provide insights into their molecular functions and interactions. The genetic analysis of the long QT syndrome (LQTS) discussed by Dr. London in Chapter 17 is a similar evolving success story, which also illustrates the potential synergism between basic and clinically focused research. The LQTS is an uncommon but life-threatening condition. Afflicted individuals suffer high rates of syncope and sudden cardiac death. The condition is clinically and genetically heterogeneous. Genetic pedigree analysis has implicated three ion channel genes in the condition. Simultaneously, basic investigation has substantially advanced our understanding of the contribution of these channels to the cardiac action potential. We are just beginning to realize the substantial promise of synergism between these fields. For example, one recent report suggests it may become feasible to tailor pharmacological therapy to the individual LQTS genotype. Specifically, patients in whom LQTS is due to delayed inactivation of the sodium channel benefit from mexiletene, which decreases inward sodium current. In contrast, in patients in whom LQTS is due to reduced function of a potassium channel, no benefit is seen with mexiletene. It should be noted that this particular potassium channel was initially cloned as a homolog of a mutant drosophila gene not then known to be relevant to cardiovascular disease. Further evidence that the dividends of basic research are far-reaching and difficult to predict. The ability to manipulate putative disease-causing genes in the germline of mice allows such genetic analysis to be brought fullcircle, and provides an invaluable animal model for analysis of pathophysiology and potential therapeutic interventions. Altering cardiac expression of ion channels in mice through germline manipulation has improved our understanding of the molecular basis of the cardiac action potential and offers promise of animal models of arrhythmia in general as well as LQTS in particular. This may well become the paradigm for future investigation in molecular medicine: genetic analysis of human conditions providing the basis for development of genetically manipulated animal models. Those models, in turn, provide a unique opportunity to

CHAPTER 11 / MOLECULAR CARDIOLOGY

return to the fundamental physiologic focus with which cardiovascular investigation began. The dog lab of the past may be supervened by the mouse lab of the future to allow analysis of physiology and pathophysiology with full advantage of the power of modern genetics. Moreover, such models will likely become the optimum testing ground for pharmacological or genetic therapies. It is clear that clinical cardiology has only begun to feel the impact of these advances in molecular biology. Although, as pointed

115

out earlier, there has been a dramatic increase in the relative number of papers published in “molecular cardiology,” these papers still constitute an exceedingly small minority of the absolute number of papers published in either cardiology or molecular biology as a whole (Fig. 11-2). The transformation has only begun. It is likely that the most dramatic changes lie ahead. The chapters in this section provide a glimpse of a revolution very much in progress, but whose repercussions have yet to fully impact clinical cardiology.

CHAPTER 12 / CONGENITAL HEART DISEASE

12

117

Congenital Heart Disease ALVIN J. CHIN

BACKGROUND Sufficient progress has occurred in identifying, characterizing, and surgically repairing physiologically important congenital cardiac defects (which occur at the rate of 20,000/yr in the United States) that 85% of these infants can now expect to survive to adulthood. Since only the most severe structural malformations thwart current management schemes, prevention and early in utero detection are the remaining strategies to further reduce the impact of heart malformation on the health of infants and children; however, any prevention approach would depend largely on knowing the percentage of congenital heart defects that are genetic. For example, if the vast majority of congenital heart anomalies are due to intrauterine insults (infectious agents, toxins, mechanical stress, and so on) to the embryo during the first 30 days of gestation, then a prevention strategy would be ineffective; since most pregnancies are already monitored by ultrasound at least once for accurate dating, perhaps adding a “cardiac surveillance” portion to the imaging protocol would suffice to increase the chance of prenatal identification. However, evidence has begun to accumulate for the hypothesis that the majority of congenital heart malformations are due to gene alterations. Therefore, the scientific challenge of the next several decades will be to unravel the sequence of molecular decisions that result in the construction of the heart and blood vessels from the first embryonic tissue layers. Although more than 50 genes have already been reported to be involved in cardiovascular morphogenesis (Table 12-1), the way they function to form the heart and great vessels correctly (in space and time) remains obscure. Both forward and reverse genetic approaches are being tried, and a variety of organisms are being scrutinized since the underlying mechanisms of patterning appear to be widely shared among vertebrates. Murine cardiac development is difficult to study in vivo without new imaging techniques because of the relatively inaccessible embryo. Furthermore, a systematic screen of the mouse genome for embryonic lethal mutations affecting organogenesis is impractical. Other vertebrates such as the chick Gallus gallus and the African clawed frog Xenopus laevis have long generation times. The teleost fish Danio (formerly Brachydanio) rerio (zebrafish) has emerged as a model system because of its prolific egg production, rapid (90-d) generation time, optically transparent embryo, and extremely rapid heart development (48 h). Despite the advantages of the zebrafish system for mutational analysis, there are some aspects of cardiovascular construction, namely From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

septation and pulmonary artery formation, that must be studied in higher vertebrates. This overview will discuss the anatomical malformations that make up the vast majority of cases of human congenital heart disease but will specifically exclude heritable cardiomyopathies (Chapter 13), long QT syndrome (Chapter 17), and Marfan syndrome, and will only briefly mention diGeorge syndrome, since this is also covered in Chapter 120.

CLINICAL FEATURES The majority of human newborns with hemodynamically significant heart defects present in the first 2 weeks of life with one of four physiological arrangements: obstruction to pulmonary arterial circulation, obstruction to systemic arterial circulation, inadequate mixing between pulmonary and systemic circulations, or pulmonary venous obstruction. Because of the two-connectedpumps-working-in-parallel construction of fetal cardiac circulation (Fig. 12-1), these four types of hemodynamic aberration are well-tolerated for a 40-week gestation. An obstruction to either the pulmonary or aortic flow is reliably compensated for by blood flow shifting to the contralateral side of the heart. Transposition of the great arteries with intact ventricular septum does not have antenatal hemodynamic consequences since it merely constitutes a different variety of two-connected-pumps-working-in-parallel configuration. Finally, anomalous pulmonary venous connection with pulmonary venous obstruction does not substantially alter prenatal hemodynamics because very little blood flow is normally allowed into the lungs in utero. The four physiological derangements are unmasked when the circulation is acutely changed at birth to a two-unconnected pumpsworking-in-series arrangement, with lungs but without placenta (Fig. 12-1). The rapidity of presentation is critically dependent on the time course of foramen ovale closure or ductal closure or both. Obstruction to pulmonary arterial circulation (Fig. 12-2A,B) manifests as cyanosis. Obstruction to systemic arterial circulation manifests as “low output syndrome.” Inadequate mixing between the pulmonary and systemic circulations also appears as cyanosis. Pulmonary venous obstruction (Fig. 12-2C) initially manifests as tachypnea out of proportion to arterial desaturation. Although these four hemodynamic subsets account for virtually all cases of heart defect appearing early, severe cases of the rare malformation “absent pulmonary valve syndrome,” whose cardinal feature is ventilatory failure due to airway impingement by adjacent markedly dilated pulmonary arteries, may also show up on the first day of life.

117

118

SECTION II / CARDIOLOGY

Table 12-1 Genes Involved in Cardiovascular Morphogenesis Temporal and spatial expression of mRNA/phenotype of knockout homozygote/overexpression (mRNA injection) data Gene family

Drosophila

Transcription factors NK tinman: no heart or gut muscle precursors

Xenopus XNkx2.3 in pharyngeal endoderm and cardiac mesoderm @ stage 16 XNkx2.5(similar to XNkx2.3)

Hox

Sox GATA

XGATA-5 in ventral cardiac mesoderm @stage 18, by stage 34, only in endocardium MADS

Dmef2 in all cardial cells @ 5.5 h; lies directly downstream of tinman

bHLH

twist activates tinman HLH Myc

MFH-1 Pax

Xtwist in cephalic neural crest @ stage 16

Mus

Nkx2.5 in myocardial [email protected] d, in [email protected] d, also in sinus venosus @9.0 d, [email protected] d, but not in aorta; Nkx2.5 [email protected] d shows no looping, cushions, trabeculae carnae a-2 in aorta @ 9.75 d, in vasculature @11.5 d; knockout does not have heart phenotype a-3 in atrium and aortic [email protected] d; a-3 knockout has defective aortic and pulmonary valves, thin aortic wall, atrial and left ventricular hypertrophy, dilatation of systemic veins Sox-4 knockout die @ 14 d from dysplasia of both semilunar valves; there is also a defect in septation of the outflow of the heart GATA-4 in caudal end of heart tube @8.0 d, in endocardium and myocardium (stopping at distal outflow tract) @ 9.0 d, in endocardial cushions and myocardium @10 d; knockout dies by 9.5 d, showing cardia bifida and absence of the pericardial coelom, defects rescuable by wild-type endoderm GATA-5 in precardiac mesoderm @7.0 d, in atrium and ventricle @ 9.5 d, in only atrial endocardium @ 12.5 d GATA-6 in precardiac mesoderm and primitive streak @ 7.0 d, in atrium and ventricle @ 9.5 d, throughout atrial and ventricular endocardium and myocardium @ 12.5 d Mef2a in heart @ 8.5 d Mef2c in heart [email protected] d, in sinus venosus @8.0 d, in atrium and ventricle @ 8.5 d; knockout dies by 10.5 d Mef2d in heart @ 8.5 d No looping occurs; the right ventricle and sinus venosus fail to form dHAND; knockout dies @ 10.5 d, showing hypoplastic right ventricle and outflow tract but a dilated aortic sac eHAND in outflow tract and left ventricle @ 8.0 d M-twist in first branchial arch @ 8.0 d, in atrioventricular canal endocardial cushion @ 10.0 d; knockout does not have heart phenotype Id in ventral, aortae, aortic arch, and dorsal aortae @9.0 d; in atrioventricular cushions and outflow tract ridges @ 10.0 d 1/3 of c-myc knockouts show enlarged heart and pericardial effusion; die by 10.5 d N-Myc in heart @ 9.5 d in only compact layer of ventricular myocardium @10.5 d; knockout shows reduction in compact layer of ventricular myocardium Knockout shows type B interruption of left aortic arch in 10/12 and perinatal death Homozygous splotch2H show truncus arteriosus @ 13.5 d approximately 50% of the time

119

CHAPTER 12 / CONGENITAL HEART DISEASE

Table 12-1 (continued) Gene family Transcription factors Zfh-1

Evi Secreted proteins TGF-β

Drosophila In dorsal vessel cardioblasts @ stages 11–15; mutant manifests breaks in anterior region of heart and kink in heart shape

Vg-1 DVR

dpp required for activation of dorsal mesoderm genes

Wnt

wgts shifted to nonpermissive temperature @3–5h: ↓#pericardial and cardial cells

VEGF

Xenopus

In truncal and conal ridges and valves @12.5 d

Injection of mRNA into right-dorsal blastomere randomizes heart looping Xbmp4 overexpression ventralizes mesoderm; inhibition dorsalizes mesoderm

Neuregulin Nodal Extracellular proteins Tenascin Fibronectin

Laminin Receptors Receptor for PDGF

Xnr-1 in left lateral plate mesoderm @ stage 19 odd Oz in cardioblasts @9.5 h; encodes tenascin like protein

Heart is “broken” (dissociation of pericardial cells)

Mus

β1 in yolk-sac blood islands and heart mesoderm @ 7.0 d, in endocardium @8.0 d, only in endocardial cushions (both atrioventricular canal and outflow tract) @9.5 d. Knockout does not have heart phenotype (due to transplacental passage of maternal TGF-β1) unless mother is TGF-β1 null. In the latter case, knockout shows poorly formed ventricular lumina and disorganized valves and ventricular myocardium; half die @ 10.5 d due to defective yolk sac vasculogenesis, endothelial differentiation, and hematopoiesis. β2 in sinus venosus @ 8.0 d, in outflow tract and aortic sac @ 8.5 d, in atrioventricular canal and proximal outflow tract myocardium @ 9.5 d β3 in pericardium @ 9.0 d

bmp2 in atrial myocardium @9.5 d (low levels in truncus and dorsal aortae), in myocardium of atrioventricular canal (not truncus) @10.5 d bmp4 in myocardium @8.5 d, more localized to atrioventricular canal region @ 9.0 d, more localized to truncus (aorta) @10.5 d; knockout: heart phenotype varies, depending on genetic background, dies between 7.5 and 10.5 d

In myocardium @ 8.5 d; heterozygotes die @ 11.0 d showing no blood vessels In endocardium @ 10.5 d; knockout shows absent ventricular trabeculae and distorted endocardial cushions Mouse nodal is in left side of node @ 4-somite stage and in left lateral plate mesoderm @ 5-somite stage Tenascin-X is in epicardial layer of sinus venosus and atrium @ 12.0 d In mesoderm during primitive streak stage, in endocardium @ 5-somite stage; knockout may not develop a heart, depending on genetic background (heart primordia not fused @ 8.5 d) Knockout does not have heart phenotype

In dorsal aortae @9.5 d; in ventricular trabeculae, cushions, and pericardium @ 11.5 d; in valves @ 15 d; Patch mutation (deletion for PDGFRα) homozygotes show truncus arteriosus, and >50% die by 9.5 d (continued)

120

SECTION II / CARDIOLOGY

Table 12-1 (continued) Genes Involved in Cardiovascular Morphogenesis Temporal and spatial expression of mRNA/phenotype of knockout homozygote/overexpression (mRNA injection) data Gene family

Drosophila

Xenopus

Receptors for neuregulin Receptors for TGF-β–like ligands

VEGF receptors

Integrins

Ig superfamily N-CAM PECAM-1 VCAM-1

Retinoic acid receptors RXR

RAR

Injection of truncated activin receptor type IIB mRNA into left-dorsal blastomere (16-cell stage) randomizes heart looping

Mus ErbB4 knockout dies @ 10.5 d and shows absent ventricular trabeculae. ErbB2 knockout dies @ 10.5 d and shows absent ventricular trabeculae Activin receptor type II B knockout dies perinatally, showing randomized heart position (and doubleoutlet right ventricle, right atrial appendage isomerism, and other anolmalies) flk-1 in heart mesoderm @ 7.0 d, in yolk-sac blood islands @ 8.0 d, in vasculature and endocardium @ 8.5 d; knockout shows no yolk sac blood islands, endocardium, or blood vessels; dies by 9.5 d flt-1 knockout forms endothelial cells but cannot assemble blood vessels, shows disorganized endocardium, and dies by 9.0 d tie-1 in vasculature and endocardium @ 8.5 d; knockout shows edema (followed by hemorrhage) due to leaky blood vessels; dies @ birth tie-2 (also called tek) in yolk sac blood islands @ 8.0 d, in vasculature and endocardium @ 8.5 d; knockout shows malformation of vascular network including coronary vessels; dies by 10.5 d showing disorganized dorsal aorta and heart trabeculae α1 α4 knockout shows no epicardium or coronary vessels and dies @ 11 g/dL, these children grow and develop normally until early puberty, when they begin to show signs of progressive hepatic, cardiac, and endocrine disturbances, including liver failure, diabetes, hypoparathyroidism, and delayed or absent secondary sexual development. These changes are owing largely to tissue siderosis from the progressive accumulation of iron derived from transfusions. Unless iron overload is controlled by regular chelation therapy, death results in the second or third decade, from acute or intractable congestive cardiac failure. Individuals with β-thalassemia minor are typically asymptomatic. Splenomegaly is rare. Diagnosis Untransfused hemoglobin levels in β-thalassemia major can be as low as 2–3 g/dL. The red cells show severe hypochromia, marked anisopoikilocytosis, target cell formation, and basophilic stippling. Poorly hemoglobinized nucleated red cells are frequently found in the peripheral blood and may reach very high levels after splenectomy. Despite the severe anemia, the reticulocyte count is usually not very high because of the massive destruction of erythroid precursor cells in the bone marrow (i.e., ineffective erythropoiesis). The bone marrow shows marked erythroid hyperplasia, characterized by poorly hemoglobinized normoblasts. Ragged inclusions (α-chain aggregates) in the normoblasts are revealed under phase microscopy or after supravital staining (e.g., methyl violet). Increased iron deposition is also seen in the bone marrow; the majority of the iron granules are randomly distributed. Biochemical evidence of hemolysis and progressive iron loading are observed. Other biochemical changes may include evidence of diabetes and endocrine dysfunction, such as parathyroid insufficiency. The findings on Hb electrophoresis vary with the β-thalassemia genotype and is informative only in the previously untransfused patient. In homozygous βo-thalassemia, Hb A is completely absent, and the hemoglobin consists of F and A2 only. In β+-thalassemia (homozygous or compound heterozygous), a variable amount of Hb A is present. The Hb F is usually elevated and vary from 10 to 90% of the total hemoglobin. The Hb A2 level is of no diagnostic value. In vitro globin chain biosynthesis of peripheral blood reticulocytes or bone marrow shows globin chain imbalance with a marked excess of α- over β- and γ-chain production. In βo-thalassemia, there is a complete absence of β-chain synthesis. The heterozygous state for β-thalassemia (βo or β+) is remarkably uniform hematologically. If present, anemia is mild, and the diagnosis is based on a low MCV and MCH, accompanied by an increased proportion of Hb A2, from 3.5 to 5.5%, with the exception of a subgroup that has a normal level of Hb A2. Globin chain biosynthesis shows α chains in excess of about twofold. Normal A2 heterozygous β-thalassemia may be difficult to distinguish hematologically from heterozygous α-thalassemia, since both cases are characterized by hypochromic microcytic red cells and a normal Hb A2 level. The distinction is made by globin chain biosynthesis and DNA analysis. Type I normal A2 β-thalassemia is “silent,” in that the red cell indices are almost normal, and the phenotype is caused by very mild mutations that cause only a minimal deficit in β-chain production (e.g., C-T mutation in position –101 of the β gene). Other cases (type 2 normal Hb A 2 β-thalassemia) are a result of the coinheritance of δ-thalassemia in cis or in trans to the β-thalassemia gene. DNA analysis frequently shows the β-globin cluster to be intact and point mutations within the β gene or its immediate flanking regions.

181

Genetic Basis of Disease The β-thalassemias are considered to be autosomal recessive disorders, because individuals who have inherited one abnormal β gene (carrier) are asymptomatic and the inheritance of two abnormal β-globin genes is required to produce a clinically detectable phenotype. Molecular analysis of the β-thalassemia genes has demonstrated a striking heterogeneity. Although >150 β-thalassemia alleles have been characterized, population studies indicate that probably only 20 β-thalassemia alleles account for >80% of the β-thalassemia mutations in the whole world. This is because in each of the high-frequency areas, only a few (four to six) mutations are common, with a varying number of rare ones; each of these populations has its own unique group of mutations. This is particularly relevant to prenatal diagnosis, because direct detection of these mutations by DNA analysis becomes feasible. The vast majority of β-thalassemia mutations are single-base substitutions, small insertions or deletions of one to two bases involving the critical sequences that control the various stages of gene expression (Fig. 20-2A). Approximately half of these mutations completely inactivate the β gene and cause a phenotype of β o -thalassemia. Mutations that allow the production of some β globin lead to the phenotype of β+-thalassemia. Mutations affecting the conserved sequences in the 5' promoter (i.e., TATA box, proximal CACCC and distal CACCC box), typically cause a 70–80% reduction in promoter activity and are often very mild. Mutations affecting the polyadenylation signal (AATAAA) at the 3' end also generally result in a mild β+-thalassemia phenotype. Studies show that the different inphase termination mutants exhibit a “positional” effect. Frameshifts and nonsense mutations that result in premature termination early in the sequence (in exon 1 and 2) are associated with minimal amounts of mutant β mRNA. In such cases, no β chain is produced from the mutant allele, and only half the normal β globin is present, resulting in a phenotype of typical heterozygous β-thalassemia. In contrast, mutations that produce inphase terminations later in the β sequence—in exon 3— are associated with substantial amounts of mutant β-mRNA. Such mutations, even when present in a single copy, result in a moderately severe anemia and are said to be “dominantly inherited.” Small amounts of truncated β-variant chains have been isolated in one case (heterozygous β codon 121). However, these truncated β chains are nonfunctional and not able to form viable tetramers, resulting in ineffective erythropoiesis and clinical disease, even in the heterozygous state. β-thalassemia is rarely caused by deletions (Fig. 20-2B). Of these, only the 619-bp deletion at the 3' end of the β gene is common, but even that is restricted to the Sind populations of India and Pakistan, where it constitutes approximately 30% of the β-thalassemia alleles. The other deletions, although extremely rare, are of particular phenotypic interest, because they are associated with an unusually high level of Hb A2 in heterozygotes. The mechanism underlying the markedly elevated levels of Hb A2 and the variable increases in Hb F in heterozygotes for these deletions is related to the removal of the 5' promoter region of the β-globin gene that removes competition for the upstream β-LCR, leading to an increased interaction of the LCR with the γ and δ genes in cis, thus enhancing their expression. Even more rare are three upstream deletions that remove all or part of the β-LCR, but leave the β gene itself intact, yet cause β-thalassemia. Because of the vast number of different β-thalassemia mutations, many patients with thalassemia major are compound heterozygotes for two different molecular lesions.

182

SECTION III / HEMATOLOGY

Figure 20-2 (A) Point mutations causing β-thalassemia. The β-globin gene is represented by 3 exons (gray) interrupted by 2 introns with the 5' and 3' untranslated regions (UTRs, striped boxes). The vertical lines represent the sites of the different mutations that can be found in the UTRs, exons and introns. (B) Deletions causing β-thalassemia. Two groups of deletions are shown: (1) group of upstream deletions which remove extensive regions (30–100 kb) of the 5' end of the β cluster but leave the β gene intact; (2) deletions that remove part or all of the β gene. These range from 290 bp to >45 kb in size. The 619-bp deletion (number 4) removes the 3' end of the β gene but leaves the 5' end intact, whereas the other deletions remove, in common, the 5' promoter region of the β gene, including sequences from positions –125 to +78 (relative to the β mRNA cap site).

Pathophysiology The molecular defects in β-thalassemias result in absent or reduced β-chain production, whereas α-chain synthesis proceeds at a normal rate. This imbalance in globin synthesis in β-thalassemia gives rise to excess α chains (Fig. 20-3A), which are extremely unstable and precipitate in the red cell precursors, forming inclusion bodies. These inclusions interfere with the red cell maturation and are responsible for the intramedullary destruction of the erythroid precursors, hence, the ineffective erythropoiesis that characterizes all β-thalassemias. The anemia of β-thalassemia results from a combination of underproduction of

hemoglobin and ineffective erythropoiesis and is ultimately related to the degree of imbalance between the α- and non–α-globin chains. HB E/β-THALASSEMIA Throughout Southeast Asia and the Indian subcontinent, it is not uncommon to encounter individuals who are compound heterozygotes for Hb E and β-thalassemia. Because Hb E is also insufficiently synthesized, the compound heterozygous state results in a clinical picture closely resembling homozygous β-thalassemia, ranging from severe anemia and transfusion dependency to thalassemia intermedia. The diagnosis of Hb E/β-thalassemia is confirmed by finding Hb E and F on hemoglo-

CHAPTER 20 / HEMOGLOBIN STRUCTURE AND SYNTHESIS DISORDERS

Figure 20-3 Diagrammatic representation of the pathophysiology of (A) β-thalassemia; (B) α-thalassemia.

bin electrophoresis and by demonstrating Hb E trait in one parent and β-thalassemia trait in the other. THE δβ-, γδβ-THALASSEMIAS AND HPFH SYNDROMES This group of disorders is characterized by a reduced synthesis of β- and δ-globin chains and a variable increase in γ-chain production. They are remarkably heterogeneous at the molecular level. In some cases, they result from extensive deletions of the β-gene cluster, removing the β- and δ- or β-, δ- and γ-globin genes, whereas in others, they result from point mutations in the γ-globin promoters (HPFH). Occasionally, δβ-thalassemia results from unequal crossing over between the δ- and β-genes producing a δβ-fusion gene. The δβ-fusion chains combine with α chains to form Hb Lepore (α2(δβ)2). In a subgroup of HPFH, heterocellular HPFH, the β cluster—including the γ-globin genes—is intact. The distinction between HPFH and δβ-thalassemia is subtle and made on both clinical and hematological grounds.

183

HPFH is characterized in heterozygotes by levels of Hb F of up to 30% with normal red cell indices and near normal hematocrit levels, whereas heterozygotes for δβ-thalassemia tend to have elevated levels of Hb F that are lower (up to 20%) with mild anemia and hypochromic microcytic red cells. The importance of this group of conditions lie in their interaction with β-thalassemia. Compound heterozygotes for HPFH and β-thalassemia tend to have a mild disease compared to β-thalassemia homozygotes, whereas compound heterozygotes for δβ- and β-thalassemia exhibit a wide spectrum in disease severity, ranging from mild anemia to thalassemia major. INTERMEDIATE FORMS OF β-THALASSEMIA Thalassemia intermedia is an ill-defined clinical term used to describe patients with phenotypes that are more severe than the asymptomatic thalassemia trait but milder than the transfusion-dependent thalassemia major. The criteria on which the diagnosis is based is that patients present later in life relative to thalassemia major and that they are capable of maintaining a reasonable level of hemoglobin (≥ 6 g/dL) without transfusion. At the severe end of the spectrum, patients present between the ages of 2 and 6 years; although they are just capable of surviving without blood transfusion, it is clear that growth and development are retarded. Many will show the skeletal and facial changes and progressive splenomegaly, as seen in untreated thalassemia major. As they become older, they develop iron-overload because of increased gastrointestinal absorption of iron. At the other end of the spectrum, patients are completely asymptomatic until adulthood and are transfusion-independent with hemoglobin levels of 10–12 g/dL. Such patients are diagnosed either during episodes of infection, when they become anemic, or by a chance hematological examination. There is usually some degree of splenomegaly. Table 20-1 lists some of the molecular interactions associated with the phenotype of thalassemia intermedia. Because of the extreme variability of these disorders, these patients should be regularly followed from early childhood and the growth charts and iron accumulation carefully monitored. THE α-THALASSEMIAS The geographical distribution of α-thalassemia is very similar to that of β-thalassemia. Although the α-thalassemias are more common, they pose less of a public health problem because the severe homozygous states cause death in utero, and the milder forms that survive into adulthood do not cause a major disability. Clinical Features The clinical disorders resulting from α-thalassemia range from death in utero (Hemoglobin Bart’s hydrops syndrome) to a completely silent carrier state. Hb Bart’s hydrops is a frequent cause of stillbirths in Southeast Asia. Affected infants are usually stillborn with gross pallor, generalized edema, and massive hepatosplenomegaly. The placenta is enlarged and friable, frequently causing obstetric difficulties. The intermediate form of α-thalassemia is Hb H disease, commonly seen in the Mediterranean, Middle East, and Southeast Asia. A spectrum of disease severity is also encountered in Hb H disease. Generally, these individuals have a moderately severe anemia and splenomegaly but are usually transfusion-independent, except during episodes of hemolysis associated with infection. The skeletal deformities and growth retardation characteristic of β-thalassemia are not common. Diagnosis Diagnosis is based on the hemoglobin level, red cell indices, examination of the peripheral blood smear, and hemoglobin analysis. Infants with Hb Bart’s hydrops are severely ane-

184

SECTION III / HEMATOLOGY

Table 20-1 Molecular Basis of β-Thalassemia Intermedia 1. Homozygous or compound heterozygous state for β-thalassemia. a. Inheritance of mild β+-thalassemia alleles. • e.g., βIVS1-6 T-C, β promoter mutations. b. Coinheritance of α-thalassemia. • Effect more evident in β+ thalassemia. c. β Thalassemia with elevated γ-chain production. • Polymorphism at position –158 Gγ gene. (Xmn I-Gγ site). • γ Promoter mutations. • Heterocellular HPFH. X-linked, 6q-linked. 2. Compound heterozygotes for β-thalassemia and deletion forms of HPFH or δβ-thalassemia. 3. Compound heterozygotes for β-thalassemia and β-chain variants (e.g., Hb E/β-thalassemia). 4. Heterozygotes for β-thalassemia. a. Coinheritance of extra α-globin genes. (ααα/αα or ααα/ααα) b. Dominantly inherited forms of β-thalassemia (including some thalassemic hemoglobinopathies).

mic with hemoglobin levels of 6–8 g/dL. The blood film shows severe thalassemic changes with numerous hypochromic nucleated red cells. There is no Hb A or F, the hemoglobin consists mainly of Hb Bart’s (γ4 tetramers), with small amounts of embryonic hemoglobin and Hb H (β4 tetramers). Biosynthetic studies confirm the complete absence of α chains. Patients with Hb H disease run an hemoglobin level of 7–10 g/dL, with moderate reticulocytosis. Again, typical thalassemic changes are seen in the blood film. On incubation of the red cells with brilliant cresyl blue, numerous inclusion bodies are generated by precipitation of the Hb H, forming typical “golf balls.” Hemoglobin analysis shows 5–40% Hb H, with the major component being Hb A and a normal or reduced level of Hb A2. Sometimes, there is also a small amount of Hb Bart’s. Carriers for α-thalassemia may be slightly anemic with hypochromic microcytic red cells or “silent,” with minimal hematologic changes. The Hb electrophoretic pattern is normal, and globin biosynthetic studies show a deficit of α-chain production. Diagnosis of α-thalassemia is confirmed by DNA analysis, which commonly reveals deletions of the α-gene cluster, removing one or both α genes. Less commonly, the α genes are present and DNA sequence analysis reveals point changes within the α2 gene or its immediate flanking regions. Genetic Basis of Disease Normal individuals (αα/αα) have two α-globin genes, α2 and α1, on each chromosome 16. The α-thalassemia syndromes result from underproduction of α chains and are frequently caused by deletions that remove one of the linked α genes (–α3.7/αα or –α4.2/αα). Extensive deletions that remove both the linked α2 and α1 genes tend to be geographically isolated, and so are often referred to by their geographical origin for example, ––SEA/αα. More rarely, upstream deletions that remove the HS –40 region but leave the α genes themselves intact can also cause α-thalassemia. Loss of one functioning α gene (αα/–α) is almost completely silent with normal or only slightly hypochromic red cells. Loss of two α genes (—/αα or –α/–α) produces a mild hypochromic microcytic anemia. Individuals who have inherited only one α gene (—/–α) have Hb H disease, whereas inheritance of no α genes (—/—) gives rise to a lethal, intrauterine hemolytic anemia termed the Hb Barts hydrops fetalis syndrome. Deficiency of α chains

gives rise to an excess of γ chains (in fetal life) or β chains (in adult life) that form γ4-tetramers (Hb Barts) and β4-tetramers (Hb H). The presence of Hb Barts or Hb H is thus diagnostic of α-thalassemia. Anemia results from a combination of the underproduction of hemoglobin and hemolysis owing to the intracellular precipitation of Hb Barts and Hb H, and is ultimately directly related to the degree of α-chain deficiency. Less commonly, α-thalassemia results from point mutations involving the critical sequences that control the various stages of gene expression, as encountered in β-thalassemia. With the exception of one phenotypically mild α thalassemia mutant, all of these mutations affect the dominant α2-globin gene. In general, the nondeletion α-thalassemia variants (αTα/αα) give rise to a more severe reduction in α-chain synthesis than the single α-gene deletions (–α/αα), the homozygous state (αTα/αTα) for such variants often result in Hb H disease. A common nondeletion α-thalassemia variant in Southeast Asia is Hb Constant Spring (Hb CS), which is caused by a single base substitution (TAA → CAA) in the α2 globin termination codon. This results in read through of the 3' untranslated sequence until another inphase termination codon is encountered 31 codons later. Homozygotes (αCSα/αCSα) or compound heterozygotes (αCSα/—) for Hb CS have a less severe form of Hb H disease. Very low levels of this elongated α globin chain (5–8% of the total hemoglobin in homozygotes) are found; the defective αCS chain production is a consequence of the instability of the αCS mRNA. Three other variants (Hb Icaria, Hb Seal Rock, and Hb Koya Dora) involving different base substitutions in the α2 termination codon have been identified. Pathophysiology There is a fundamental difference in the pathophysiology of the α- and β-thalassemias (Fig. 20-3). Because γ4 and β4 tetramers are soluble, they do not precipitate to a significant degree in the bone marrow (i.e., erythropoiesis is more effective than in β-thalassemia). However, these β 4 tetramers do precipitate as the red cells age, forming inclusion bodies. Hemolysis occurs because of the red cell membrane damage and obstruction in the spleen. α-THALASSEMIA WITH MENTAL RETARDATION (ATR) SYNDROMES There are two distinct syndromes of α-thalassemia and mental retardation. One group, ATR-16, results from

CHAPTER 20 / HEMOGLOBIN STRUCTURE AND SYNTHESIS DISORDERS

185

Table 20-2 Clinical Disorders Caused by Structural Hemoglobin Variants 1. Sickle syndromes causing hemolysis and tissue damage. Hb S and interaction of Hb S with other Hb variants. (Hbs S/C, S/O-Arab, and S/D-Punjab) and β-thalassemia (S/β-thal). 2. Chronic hemolysis—unstable hemoglobin variants (congenital Heinz body anemia, CHBA). e.g., Hb Köln, Hb Bristol. 3. Congenital polycythemia—high oxygen affinity Hb variants. 4. Congenital cyanosis—Low oxygen affinity Hb variants. M hemoglobins. 5. Hypochromic microcytic anemia (thalassemic hemoglobinopathy). e.g., Hb E—β structural variant. Hb Constant Spring—α structural variant. Hyperunstable hemoglobin variants. 6. Drug-induced hemolysis (e.g., Hb Zurich).

extensive deletions of 1–2 Mb of the tip of chromosome 16, which remove the α-globin gene cluster. The mental retardation is associated with variable dysmorphic features and is thought to result from loss of a variable number of genes at the tip of chromosome 16p. ATR-X is characterized by more severe mental retardation, characteristic dysmorphic facies, and genital abnormalities. In these cases, the α-globin gene cluster is intact; the underlying mutation is a trans-acting abnormality encoded in the XH2 gene on the X chromosome. ACQUIRED HB H DISEASE Hb H disease is occasionally seen in individuals with a variety of hematological disorders within the myelodysplastic syndromes. These individuals are predominantly elderly males and hematologically normal before the onset of the disease. The α-globin gene cluster is intact, and the acquired α-thalassemia is likely to be caused by reduced α-gene transcription, but the nature of the specific defect remains completely unknown. MANAGEMENT The thalassemias are a major health problem in many populations; because there is no definitive treatment, major efforts are concentrated on prevention. Preventive programs in the past were based on education, population screening, heterozygote detection, and genetic counseling, but they were not entirely effective. Most countries now combine this approach with screening programs at antenatal clinics. When heterozygous mothers are detected, their partners are tested; if their partners are also carriers, the couples are offered prenatal diagnosis and selective termination of pregnancy. Regular transfusion and iron chelation remain the cornerstones of treatment for severe β-thalassemia and are primarily palliative. Currently, the most useful chelating agent is desferrioxamine (Desferal). Unfortunately, Desferal is not only expensive, but it also has to be administered parenterally via an infusion pump. It is not surprising that noncompliance becomes a problem, particularly during adolescence. Considerable effort is directed towards the development of a safe and effective oral iron chelator. The most promising of these is deferiprone (L1); results of recent clinical trials suggest that although deferiprone is initially effective, there is a loss of efficacy in the long term. Another alternative that is currently being developed is a new formulation of depot desferrioxamine which can be given as a single bolus dose subcutaneously once a week. The development of hypersplenism as shown by the falling platelet and white-cell counts, and increasing blood transfusion requirements is an indication for splenectomy.

Bone marrow transplantation is the only form of treatment that can cure the severe forms of β-thalassemia, but it is dependent on the availability of an HLA-compatible related donor. Young wellchelated patients without evidence of liver disease are the best candidates, having 80% chance of cure. With increasing knowledge of cell and molecular biology, a long-term aim would be the replacement of the defective β gene with its normal counterpart. Although, significant progress has been made in addressing the problems of gene transfer and expression, major biological problems remain. It has long been known that the severity of disease in β-thalassemia can be ameliorated by coinheritance of genetic factors that increase Hb F production. Thus, activation of the normal γ genes in patients with β-thalassemia represent a potentially important approach of therapy for this disorder. Several compounds, including 5-azacytidine, hydroxyurea, butyrate and butyrate analogs, have been tried, but, to date, none could achieve the therapeutic levels of Hb F needed without inducing significant toxic side effects.

STRUCTURAL HEMOGLOBIN VARIANTS More than 600 structurally different hemoglobin variants have been described. Many of them are harmless and have been discovered in population surveys using electrophoretic analyses of human hemoglobin. Because only variants that alter the charge of the hemoglobin molecule are detectable in routine electrophoresis, this number is probably an underestimate. Diseases resulting from structural abnormalities of hemoglobin are shown in Table 20-2. In this section, only those abnormal hemoglobins of clinical importance are described. SICKLE-CELL DISEASE Background Sickle-cell anemia was first described by Herrick from Chicago in 1910, in a West Indian student. Peculiar elongated and sickle-shaped red blood cells were observed in the peripheral blood films and suggested the term sickle cell anemia. In 1949, Pauling et al. demonstrated that this sickling phenomenon was related to an abnormal hemoglobin present in all patients with sickle-cell anemia; this hemoglobin had an abnormally slow electrophoretic migration. Subsequently, in 1956, the sickle hemoglobin was chemically characterized by Ingram and was shown to differ from normal adult hemoglobin (Hb A, α2β2) by the single substitution of glutamic acid to valine at position six in the β-

186

SECTION III / HEMATOLOGY

Table 20-3 The Major Sickling Disorders

β Genotype

α Genotype

Hb Electrophoresis

Sickle-cell trait Sickle-cell trait Sickle-cell Anemia SC disease SO-Arab disease SD-Punjab disease Sβ + thal

β A /β S β A /β S β S /β S β S /β C β S /β OArab β S /β D-Punjab β S /β Th

αα/αα α–/αα or α–/α– αα/αα αα/αα αα/αα αα/αα αα/αα

Sβ o thal

β S /β Th

αα/αα

S HPFH

β S /β c

αα/αα

Hb S approx 40%; Hb A 60% Hb S 20–35%; Hb A 65–80 Hb S 80–100%; Hb F 0–20% Hb S 50%; Hb C 50% Hb S, HbO Arab a Hb S 50%; Hb D Punjab 50% b Hb S 50–80%; Hb F 0–20% Hb A 10–30%; Hb A 2 3–6% Hb S 75–100%; Hb F 0–20% Hb A 2 3–6% Hb S 70–80%; Hb F 20–30%; Hb A 2 decreased

aHbC, HbO-Arab, and HbE are not separated on bQuantitation based on agar gel electrophoresis. cDeletion of β-globin cluster.

routine alkaline electrophoresis.

subunit. Since then, molecular characterization has shown that, apart from the homozygous state for the βs gene, the syndrome of sickle-cell disease can also arise from the compound heterozygous state for HbS and other structural variants such as Hbs C and D and β-thalassemia (Table 20-3). The sickling disorders occur predominantly in black African populations, but they are also prevalent throughout the Mediterranean, Middle East and parts of India. It appears that heterozygotes for the βs gene are protected from the severe effects of Plasmodium falciparum, thus explaining the high gene frequencies in those malarious regions. The βs gene in these diverse population groups is caused by the same molecular defect (β codon 6 GAG to GTG). Recent investigations using β haplotypes constructed from the linked-groupings of DNA sequence polymorphisms in the β globin cluster have provided some insights into the origins and migration of the βs gene. The βs gene occurs on four different β haplotypes in Africa, known as the Senegal, Benin, Central African Republic (or Bantu), and the Cameroon types. In addition, it is associated with a different β haplotype in Saudi Arabian and Asian Indian sickle patients. The evidence suggests multiple independent origins of the βs mutation, although gene conversion on regionally specific β haplotypes cannot be excluded. Clinical Features Sickle-cell trait is a benign condition and generally does not cause any clinical disability. However, under certain extreme conditions such as severe pneumonia, flying in unpressurized aircraft, and exercise at high altitude, vaso-occlusive episodes can occur. The clinical manifestations of sickle cell disease range from a chronic hemolytic anemia interspersed with painful crises to a completely asymptomatic state, detected only by chance on routine hematological examination. Most patients fall between these two extremes and are relatively asymptomatic, except for the occasional clinical crisis. Sickle-cell anemia normally presents in infancy with attacks of painful dactylitis, the so-called “hand-foot syndrome,” with swelling of the fingers and feet. At this stage, the infant is usually anemic with mild jaundice and the spleen palpable. Splenomegaly usually resolves owing to repeated infarctions of the spleen, the “autosplenectomy” manifested by typical postsplenectomy changes in the peripheral blood film. It is unusual to feel the spleen after the first decade of life. Typically, these children have a

chronic hemolytic anemia. The hemoglobin level varies between 6 and 8 g/dL, with a reticulocyte count of 10–20%, slight elevation of the serum bilirubin level, and increased urinary urobilinogen. Examination of the peripheral blood smear shows polychromasia and poikilocytosis, with a variable number of sickled erythrocytes. The chronic hemolysis of sickle-cell anemia is punctuated by acute exacerbations of the illness termed sickling crises traditionally classified as vaso-occlusive, sequestration, aplastic, or hemolytic. The most common are the vaso-occlusive crises, characterized by acute painful episodes caused by blockage of small vessels with sickled erythrocytes and tissue infarction. Commonly, the patient experiences a rapid onset of deep throbbing bone pain in the lumbosacral spine and limb bones, usually without physical findings but sometimes accompanied by local tenderness, warmth, and swelling. Marrow aspirated from areas of bone tenderness have revealed infarction of the marrow tissue. Occasionally, abdominal pain is the major symptom and can pose a difficult problem in differential diagnosis. The abdominal crisis is accompanied by distension and rigidity with loss of bowel sounds—findings typical of an acute surgical abdomen. Vaso-occlusion in the lung, acute chest syndrome, is characterized by acute dyspnea and pleuritic pain and often accompanied by a significant fall in the hemocrit that may reflect sequestration of the sickled cells in the pulmonary vessels. The distinction from pulmonary infection is often very difficult, particularly as infection and infarction usually coexist. Acute chest syndrome is the most common cause of death after 2 years of age; the patient with acute chest syndrome is extremely vulnerable because hypoxia has a profound effect on sickling. Acute vaso-occlusion in the central nervous system (CNS) usually presents in childhood, either as fits, transient neurological symptoms resembling ischemia attacks, or with a fully developed stroke. Recurrent attacks are common, 70% of patients experience a recurrence within 3 years, and many are left with permanent motor and intellectual disabilities. Priapism is a distressing problem resulting from vaso-occlusion of the outflow vessels from the corpora cavernosa by sickled erythrocytes. This complication may present as multiple shortlived episodes (“stuttering” priapism) that may progress to “severe prolonged” priapism, lasting several days and leading to permanent sexual dysfunction.

CHAPTER 20 / HEMOGLOBIN STRUCTURE AND SYNTHESIS DISORDERS

187

Figure 20-4 Pathophysiology of sickle-cell disease. Initially, intracellular polymerization is reversible, but repeated cycles of polymerization-depolymerization in the circulating red cell results in membrane damage, and the cell becomes irreversibly sickled. This results in two main effects; first, sickled erythrocytes are mechanically fragile with a short intravascular life-span resulting in a chronic hemolytic anemia. Second, sickled erythrocytes are not so flexible, particularly in the microvasculature; they adhere abnormally to the vascular endothelium and form aggregates leading to vascular stasis, local hypoxia, and further sickling, complete vaso-occlusion and tissue infarction.

Temporary marrow aplasia can have a profound effect with reticulocytopenia and a very sudden drop in hematocrit. These aplastic crises appear to result from intercurrent infections, particularly owing to parvovirus B19 and often occur in epidemics, frequently involving more than one sibling in the same family. Sequestration crises are the most serious of the acute crises and commonly involve the spleen in the first 2 years of life. Acute splenic sequestration is characterized by sudden rapid massive enlargement of the spleen that becomes engorged with sickled erythrocytes. As the crisis progresses, a large proportion of the circulating red cell mass may be trapped in the spleen, leading to profound anemia and death. Splenic sequestration shows a tendency to recur in the same individual. A similar type of sequestration may occur in the liver in adult life, causing a dramatic fall in the hematocrit. Patients with sickle-cell disease are particularly susceptible to infection because of Streptococcus pneumonia, Salmonella, Escherichia coli, and Hemophilus influenza. Osteomyelitis is common and results from infection of bone infarcts. Pneumococcal pneumonia and overwhelming septicemia are particularly important causes of death in infancy and childhood because of hyposplenism. Repeated vaso-occlusive events ultimately result in end-organ damage and almost any organ can be affected. The vertebral bodies and femoral heads are particularly prone to infarction. A vascular necrosis of the femoral head may lead to total disability, frequently requiring a total hip prosthesis. Virtually every patient with sickle-cell anemia has some form of renal impairment. Sickling of the erythrocytes is enhanced in the hypertonic, hypoxic, and acidotic environment of the renal medulla, leading to progressive infarction of the medullary papillae. There is progressive inability to concentrate urine, polyuria, nocturia, and enuresis, which is common in children. Eventually, the glomerular damage

causes chronic renal failure, particularly in patients over 40 years of age. Because of chronic hemolysis, gallstones are very common and are seen in one-third of SS patients by 10 years of age. However, it is difficult to assess its clinical significance, because only a minority develop clear-cut cholecystitis. Recurrent chronic leg ulceration is common and can be a major handicap. The lesions normally occur just above the medial malleoli and seem to be more common in those patients with severe anemia. Proliferative retinopathy leading to progressive visual loss is an important ocular complication, although this is more common in Hb SC disease. Diagnosis Diagnosis is established by the combination of several tests: hemoglobin electrophoresis on cellulose acetate (pH 8.6) and agar (pH 6.2) and positive sickling or solubility tests. Diagnostic difficulties can often be resolved by family studies. Genetic Basis and Molecular Pathophysiology of Disease Sickle hemoglobin (Hb S, α2β2S) is produced by a single base substitution (GAG→GTG resulting in Glu to Val) in codon 6 of the β-subunit of hemoglobin. Fundamental to the sickle-cell pathophysiology is the insolubility of deoxy-Hb S in the red cells (Fig. 20-4). This leads to intracellular polymerization, which alters successively the cytoplasmic viscosity, the topography, and microrheology of the membranes, the ion fluxes, and the cell–cell interactions in their adherence to the vascular endothelium. The rate of polymerization varies exponentially with the intracellular hemoglobin concentration and composition (percentage of Hb S and non-S hemoglobin), as well as oxygen saturation. Oxygen binding to HbS has a dramatic effect on polymerization. Sickling is influenced by a wide variety of inherited and acquired factors, both intracellular and extracellular. Abnormal endothelial adherence of the sickled erythrocytes is also influenced by cellular constituents such as von Willebrand factor, cytokines, and vascular tone.

188

SECTION III / HEMATOLOGY

Management Currently, there is no specific treatment that is useful in sickle cell disease. Management consists of continuous general medical care and treatment of complications as they arise. In high-risk populations, neonatal screening programs should be established to identify babies with sickle-cell anemia as soon as possible. These babies should be given prophylactic penicillin, followed by polyvalent pneumococcal vaccine between the ages of 6 and 12 months. Folate supplements should be given during pregnancy and in patients with severe anemia. Transfusions are not usually required except in special circumstances. Episodes of infection should be treated early and sudden exposure to cold and high altitudes avoided. All but the mildest of crises should be managed in hospital. The patient should be examined for any underlying infection, kept warm and adequately hydrated, either orally or intravenously, and given appropriate antibiotics. Prompt and adequate relief of pain is of prime importance. In selected patients, exchange transfusion may be effective in preventing recurrent painful crises. Most patients with sickle-cell anemia tolerate the relatively low levels of hemoglobin quite well, and blood transfusion is usually not required. However, a blood transfusion is indicated when there is a sharp fall in the hemoglobin because bone marrow failure (aplastic crisis) or increased hemolysis. A sequestration crisis requires very close surveillance, and urgent transfusion is usually indicated because of the prompt development of profound anemia. “Acute chest syndrome” and the “brain syndrome” should be treated by partial exchange transfusion to maintain a level of sickle hemoglobin of 120; 42

6.8–7.2

HS HS

Band 3

911 (102)

1,200,000

EPB3; 17q12–q21

17; 20

4.7

HS

Protein 4.1 Protein 4.2

588 (66) 691 (77)

~ 200,000 ~ 200,000

EL1; 1p33–p34.2 ELP42; 15q15–q21

> 250; > 23 20; 13

5.6 2.4

HE HSa

aAtypical form of HS. bThe inheritance pattern

8

Inheritance patternb (autosomal) Dominant Recessive Dominant or recessive Dominant Dominant or recessive Dominant or recessive Dominant Recessive

depends on complex factors in addition to the mere nature of the mutations.

Figure 21-2 A series of changes in the SDS-PAGE profile in HS and HE/HPP. (䉴), primary decrease or absence; (䉯), secondary decrease; (*), duplication (+ decrease). A and B: HS associated with ANK1 gene mutations. A: one allele (heterozygosity) carries an early frameshift mutation. It abolishes one haploid set of ankyrin and leads to the secondary reduction of spectrin and protein 4.2 (Morlé et al., unpublished results). B: one allele (heterozygosity) carries a nonsense mutation near the 3' end of the gene. It yields a truncated ankyrin molecule (Hayette et al., unpublished results). C and D: HS associated with EPB3 gene mutations. C: compound heterozygosity for two abnormal alleles generates a sharp reduction of band 3 and, secondarily, a noticeable decrease of protein 4.2 (Alloisio et al., unpublished results). D: compound heterozygosity for two other abnormal alleles yields a total absence of protein 4.2; one allele is of the same type as in C (abolishing one haploid set of band 3) and the other allele disrupts the binding site of band 3 cytoplasmic domain for protein 4.2 (Kanzaki et al., unpublished results). E: HS associated with a frameshift mutation in the ELP42 gene (homozygosity). F: HE associated with an exon skipping near the 3' end of SPTB transcript (heterozygosity), yielding a truncated β-spectrin molecule. G: HE associated with the abolition of the downstream initiation codon (homozygosity) of the EL1 gene. (Electrophoreses done by Ms. M.-A. Schreiner.)

terminal region) for a complementary site in the β chain (Nterminal region). As a consequence, the nucleation process, i.e., the onset of dimerization, becomes impossible, and all exon 46lacking spectrin α chains are discarded. However, this remains innocuous, even in the homozygous state, because a large excess of α chains is normally synthesized. Half of the αLELY products—those bearing exon 46—are sufficient to meet the needs of the erythrocyte. On the contrary, in the αHE/αLELY situation, there is a twofold excess of αHE chains compared to αLELY chain having retained exon 46. An increased proportion of αHEβ dimers are assembled. Subsequently, those dimers prove unable to self-associate, hence the aggravation of the disease. Of course, some αHE alleles, coined αHE-LELY alleles, carry the LELY mutations in cis of a HE mutation. As one would anticipate, the expression of HE is attenuated.

4.1(-) Hereditary Elliptocytosis Among Caucasians, approximately 30% of HE results from defects involving protein 4.1, usually in the heterozygous state and rarely in the homozygous state (Fig. 21-2G). Protein 4.1 is a key element of the junctional complex (Fig. 21-1). Heterozygous 4.1(–) HE is symptomless and discloses numerous, smooth and well-elongated elliptocytes. Homozygous 4.1(–)HE achieves a picture of HPP. Owing to the size of EL1 gene (>250 kb) and of its introns, and also because of the extraordinary complexity of the alternative splicing characterizing protein 4.1 transcript, only two mutations have been elucidated at the gene level so far. One of them cancels the downstream initiation codon for translation— the only initiation codon used in the erythroid cells—whereas the other disrupts the binding site of protein 4.1 for spectrin and actin.

194

SECTION III / HEMATOLOGY

Figure 21-3 Location of mutations in band 3 (cDNA level) (recorded 8/31/95). The cytoplasmic domain (positions 1 to 403) and the membrane domain (404–881) with its transmembrane segments (circled numbers) are individualized. 䊉, single base mutations; 䊊, polymorphisms; x, frameshift or nonsense mutations; 䉭, mutations abolishing the binding of protein 4.2.

Figure 21-4 Clustering of mutations responsible for HE/HPP at/or in the vicinity of the self-association region (recorded 8/31/95). Circled numbers designate the homologous segments at the point where they begin. a: intronic mutations that yield skipping of SPTB gene exons 30 or 31. Inset: correspondence between the homologous (or repeating) segments (linear) and the conformation units (three-dimensional). Each conformational unit contains one helix with a color, h3, belonging to a homologous segment, and two white helices, h1 and h2, belonging to the preceding homologous segment. (Courtesy of Dr. P. Maillet.)

CHAPTER 21 / RED CELL DISORDERS

In homozygous 4.1(–)HE, the total absence of protein 4.1 causes a sharp reduction of transmembrane glycophorin C and the total absence of a third protein, called p55 (not shown). This reflects a triangular interaction between these proteins (a vertical interaction; see Fig. 21-1). On the other hand, the Leach phenotype, a rare blood group resulting from the primary absence of glycophorin C, is associated with elliptocytosis. Protein 4.1 is slightly diminished (hence the elliptocytosis) and p55 is, again, entirely missing.

MOLECULAR PATHOPHYSIOLOGY Few studies have tried to bridge the abnormalities at the molecular and cellular levels. In a simple fashion, spherocytes stem from discocytes by “spleen conditioning,” i.e., the removal of microvesicles (50–80 nm) that yield a reduction of the membrane surface area. It is intriguing that mutations in at least five distinct proteins result in the same cellular phenotype. Elliptocytes do not recover the discocyte shape following shear stress in large vessels. This lack of elastic deformability generates increased mechanical fragility. In severe cases, the cells shatter against the walls of major arteries, achieving pyropoikilocytosis. Again, it is remarkable that mutations in spectrin and protein 4.1 have a similar cellular phenotype.

TREATMENT Molecular genetics has allowed reclassification HS and HE/HPP and provide a rational basis of diagnosis. Treatment is based on occasional or regular blood transfusions. Splenectomy is beneficial; it is nearly curative in HS. Identification of the causative mutation at the DNA level allows prenatal diagnosis when a mutation, that is severe or lethal in the homozygous state, occurs in consanguineous couples (Alloisio et al., unpublished data). The efficiency of prenatal and postnatal managements do not justify gene therapy.

FUTURE PROSPECTS Beside the practical interest in describing mutations, many of the latter offer precious natural models. They allow far-reaching studies on gene expression and structure–function relationships within proteins.

SELECTED REFERENCES Alloisio N, Morlé L, Maréchal J, et al. SpαV/41: a common spectrin polymorphism at the αIV-αV domain junction. Relevance to the expression level of hereditary elliptocytosis due to α-spectrin variants located in trans. J Clin Invest 1991;87:2169–2177. Alloisio N, Texier P, Vallier A, et al. Modulation of clinical expression and band 3 deficiency in hereditary spherocytosis. Blood 1997;90:414–420. Amin KM, Scarpa AL, Winkelmann JC, Curtis PJ, Forget BG. The exonintron organization of the human erythroid β-spectrin gene. Genomics 1993;18:118–125. Conboy JG. Structure, function, and molecular genetics of erythroid membrane skeletal protein 4.1 in normal and abnormal red blood cells. Semin Hematol 1993;30:58–73. Dalla Venezia N, Gilsanz F, Alloisio N, Ducluzeau MT, Benz EJ Jr, Delaunay J. Homozygous 4.1(-) hereditary elliptocytosis associated with a point mutation in the downstream initiation codon of protein 4.1 gene. J Clin Invest 1992;90:1713–1717. Delaunay J, Dhermy D. Mutations involving the spectrin heterodimer contact site: clinical expression and alterations in specific functions. Semin Hematol 1993;30:21–33.

195

Eber SW, Gonzalez JM, Lux ML, et al. Ankyrin-1 mutations are a major cause of dominant and recessive hereditary spherocytosis. 1996;13:214–218. Gallagher PG, Ferriera JD. Molecular basis of erythrocyte membrane disorders. Curr Opin Hematol 1997;4:128–135. Gallagher PG, Kotula L, Wang Y, et al. Molecular basis and haplotyping of the alphaII domain polymorphisms of spectrin: application to the study of hereditary elliptocytosis and pyropoikilocytosis. Am J Hum Genet 1996;59:351–359. Gallagher PG, Tse WT, Scarpa AL, Lux SE, Forget BG. Structure and organization of the human ankyrin-1 gene. Basis for complexity of pre-mRNA processing. J Biol Chem 1997;272:19,220–19,228. Hassoun H, Vassiliadis JN, Murray J, et al. Characterization of the underlying molecular defect in hereditary spherocytosis associated with spectrin deficiency. Blood 1997;90:398–406. Hayette S, Dhermy D, dos Santos ME, et al. A deletional frameshift mutation in protein 4.2 gene (allele 4.2 Lisboa) associated with hereditary hemolytic anemia. Blood 1995;85:250–256. Jarolim P, Murray JL, Rubin HL, et al. Characterization of 13 novel band 3 gene defects in hereditary spherocytosis with band 3 deficiency. Blood 1996;88:4366–4374. Jenkins PB, Abou-Alfa GK, Dhermy D, et al. A nonsense mutation in the erythrocyte band 3 gene associated with decreased mRNA accumulation in a kindred with dominant hereditary spherocytosis. J Clin Invest 1996;97:373–380. Kotula L, Laury-Kleintrop LD, Showe L, et al. The exon-intron organization of the human erythrocyte α-spectrin gene. Genomics 1991;9: 131–140. Lorenzo F, Dalla Venezia N, Morlé L, et al. Protein 4.1 deficiency associated with an altered binding to the spectrin-actin complex of the red cell membrane skeleton. J Clin Invest 1994;94:1651–1656. Lux SE, Palek J. Disorders of the red cell membrane. In: Handin RI, Lux SE, Stossel TP, eds. Blood, Principles and Practice of Hematology. Philadelphia: JB Lippincott, 1995; pp. 1701–1818. Peters LL, Lux SE. Ankyrins: structure and function in normal cells and hereditary spherocytes. Semin Hematol 1993;30:85–118. Peters LL, Shivdasani RA, Liu SC, et al. Anion exchanger 1 (band 3) is required to prevent erythrocyte membrane surface loss but not to form the membrane skeleton. Cell 1996;86:917–927. Randon J, Miraglia del Giudice E, Bozon M, et al. Frequent de novo mutations of the ANK1 gene mimic a recessive mode of transmission in hereditary spherocytosis: three new ANK1 variants: ankyrins Bari, Napoli II and Anzio. Br J Haematol 1997;96:500–506. Sahr KE, Laurila P, Kotula L, et al. The complete cDNA and polypeptide sequences of human erythroid α-spectrin. J Biol Chem 1990;265: 4434–4443. Sahr KE, Taylor WT, Daniels BP, Lubin HL, Jarolim P. The structure and organization of the human erythroid anion exchanger (AE1) gene. Genomics 1994;24:491–501. Schofield AE, Martin PG, Spillett D, Tanner MJA. The structure of the human red blood cell anion exchanger (EPB3, AE1, Band 3) gene. Blood 1994;84:2000–2012. Tanner MJA. Molecular and cellular biology of the erythrocyte anion exchanger (AE1). Semin Hematol 1993;30:34–57. Tanner MJ. Physiology: The acid test for band 3. Nature 1996;382: 209,210. Wilmotte R, Maréchal J, Morlé L, et al. Low expression allele αLELY of red cell spectrin is associated with mutations in exon 40 (αV/41 polymorphism) and intron 45 and with partial skipping of exon 46. J Clin Invest 1993;91:2091–2096. Winkelmann JC, Chang JG, Tse WT, Scarpa AL, Marchesi VT, Forget BG. Full-length sequence of the cDNA for human erythroid β-spectrin. J Biol Chem 1990;265:11,827–11,832. Winkelmann JC, Forget BG. Erythroid and nonerythroid spectrins. Blood 1993;81:3173–3185. Yawata Y. Red cell membrane protein band 4.2: phenotypic, genetic, and electron microscopic aspects. Biochim Biophys Acta 1994;1204: 131–148.

CHAPTER 22 / RED CELL ENZYMOPATHIES

22

197

Red Cell Enzymopathies LUCIO LUZZATTO AND ROSARIO NOTARO

INTRODUCTION The mature red cell is the product of a developmental pathway that brings the phenomenon of differentiation to an extreme. An orderly sequence of events produces synchronous changes, whereby the gradual accumulation of a huge amount of hemoglobin in the cytoplasm (to a final level of 340 g/L, i.e., about 5 mM) goes hand in hand with the gradual loss of cellular organelles and of biosynthetic abilities. In the end, the erythroid cell undergoes a process that has features of apoptosis, including nuclear pycnosis and actual loss of the nucleus. However, the final result is more altruistic than suicidal; the cytoplasmic body, instead of disintegrating, is now able to provide oxygen to all cells in the human organism for some remaining 120 days of the red cell life-span. This unique behavior has important consequences, not only with respect to the physiology of red cells (which cannot be covered here), but also with respect to their pathophysiology. Specifically, the chemical machinery of intermediary metabolism is drastically curtailed in mature red cells (Fig. 22-1); for instance, because cytochrome-mediated oxidative phosphorylation has been lost with the loss of mitochondria, there is no backup to anaerobic glycolysis for the production of adenosine triphosphate (ATP). Also, because the capacity of making protein has been lost with the loss of ribosomes, the existing metabolic machinery is always at risk: if any component deteriorates, it cannot be replaced, as in most other cells. Another consequence of the relative simplicity of red cells is that they have a very limited range of ways to manifest distress under conditions of hardship: in essence, any sort of metabolic failure will eventually lead either to membrane damage or to failure of the cation pump. In the former, the cell will suffer direct mechanical breakdown; in the latter, it will undergo osmotic lysis. In either case, the life-span of the red cell is reduced, which is the definition of a hemolytic disorder. If the rate of red cell destruction exceeds the capacity of the bone marrow to produce more red cells, the hemolytic disorder will manifest as hemolytic anemia. Inherited abnormalities of red cell enzymes (enzymopathies for short) are a distinct set of genetic disorders. Most of the enzymes involved are housekeeping enzymes that, by definition, are present in all cells. Therefore, it is of interest first of all to consider the possible consequences of the deficiency of any of them. On one hand, we might expect that a severe reduction in the activity of any housekeeping enzyme would have generalized clinical manifestations, because in principle, all organs would be From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

affected. On the other hand, in many cells there might be “metabolic redundancy,” whereby other enzymes provide the missing function in a surrogate capacity; or, if the deficiency is not total, an increase in the synthesis of the enzyme involved might provide the minimum of activity compatible with cell viability. Because both of these compensatory mechanisms are not available to red cells, they would be more severely affected than other cells. In practice, we observe a clinical picture somewhere within this spectrum for each individual enzymopathy. In the majority of cases the main problem is hemolytic anemia; in some cases, hemolytic anemia is associated with other clinical manifestations; and in others the latter—particularly neurological damage—may dominate the clinical picture. In this chapter, we shall concentrate on those enzymopathies affecting red cell metabolism of which the molecular basis has been elucidated. We will not cover conditions in which an enzyme abnormality is expressed also in red cells, but the main clinical manifestations are elsewhere (e.g., the porphyrias, acatalasemia, galactosemia, the Lesch-Nyhan syndrome). For the sake of brevity, some enzymopathies will be discussed as a group, with details of individual enzymes given in additional notes and within the illustrative material.

ENZYMOPATHIES OF GLYCOLYSIS PREVALENCE All of these defects are rare to very rare (Table 22-1, pp. 200–201). CLINICAL FEATURES All of these defects cause hemolytic anemia with varying degrees of severity (the detailed description of the clinical features and of the pathophysiology of hemolytic anemia is not within the scope of this book). If the anemia is severe, it will usually present very early in life, sometimes with severe neonatal jaundice that may require exchange transfusion; if the anemia is less severe, it may present later in life, or it may even remain asymptomatic and be detected incidentally when a blood count is done for unrelated reasons. The spleen is often enlarged. When other systemic manifestations occur, they involve either the neuromuscular system, or the central nervous system (CNS)— sometimes entailing severe mental retardation—or both. DIAGNOSIS The diagnosis of hemolytic anemia is usually not difficult, based on the triad of normo-macrocytic anemia, reticulocytosis, and hyperbilirubinemia. Enzymopathies should be considered in the differential diagnosis of any chronic Coombsnegative hemolytic anemia. Once hemoglobinopathies have been ruled out, the field is narrowed down to membrane abnormalities and enzymopathies. Red cell morphology is usually conspicuously

197

198

SECTION III / HEMATOLOGY

Table 22-1, the majority of mutations so far identified in the genes encoding glycolytic enzymes are of the missense type, causing single amino acid replacements. This is important, because in all these cases, the activity of the respective enzyme is reduced, but it is not completely lost. Thus, the residual activity can still support a decreased but nonzero flow through the glycolytic pathway: this explains how red cells survive in circulation, even though their life-span is reduced. As for the molecular basis for the reduction in enzyme activity, we must consider two basic mechanisms: 1. In the majority of cases, loss of activity is probably owing to a decreased stability of the protein. In such cases, we would predict that other cells might be much less affected than red cells, because the former can compensate for decreased stability through increased synthesis of the enzyme. 2. In some cases, the amino acid replacement may affect the active center of the enzyme. This, in turn, may affect either substrate binding (Km) or the catalytic rate of the enzyme (kcat), or both: in this case, other cells in which the rate of glycolysis is critical will be affected, as well as red cells.

Figure 22-1 Intermediary metabolism in red cells. The diagram shows the glycolytic pathway and related reactions (not the complete metabolic machinery of the red cells). Enzymes are enclosed in rounded boxes. Abbreviations are as in Table 1. Additional abbreviations: DPG, diphosphoglycerate; GSH, reduced Glutathione; GSSG, Glutathione; HbFe2+, hemoglobin; HbFe3+, methemoglobin; γGluCys, γ-Glutamylcysteine.

abnormal in membrane disorders, but conspicuously normal in glycolytic enzymopathies. A useful test is the so-called autohemolysis test, in which red cells are simply incubated at 37°C under defined conditions, and the percent lysis measured after 48 hours. With membrane abnormalities, an abnormal hemolysis takes place, but this can be limited or abolished by the addition of sufficient concentrations of glucose; with glycolytic enzyme abnormalities, this “glucose correction” is only partially or totally lacking. This test has been criticized because it is unspecific: however, in our hands, it can be used as a reliable screening for enzymopathies. If the test is abnormal, the final diagnosis must be done by assaying individual glycolytic enzymes by the appropriate spectrophotometric method. For the sake of economy, it makes sense to carry out these rather laborious tests in order of frequency of the various enzymopathies (i.e., first PK, then GPI, and so on; Table 22-1). If a particular molecular abnormality is already known in the family, of course one can test for that directly on the patient’s DNA, bypassing the need for enzyme assays. GENETIC BASIS Because these conditions are all recessive, there are usually no affected members in the previous generations. However, inquiries for consanguinity should always be made, and there may be of course an affected sibling. MOLECULAR PATHOPHYSIOLOGY At the biochemical level, all glycolytic enzymopathies have something in common, because their main consequence is to reduce the capacity of the red cell to produce chemical energy in the form of ATP. As seen in

MANAGEMENT Currently, there is no specific treatment for these conditions. Patients with moderate anemia may require occasional blood transfusion when they experience exacerbations of the anemia because of increased rate of hemolysis, or because of decreased red cell production secondary to infection (the most extreme example being “aplastic crisis” from parvovirus infection). Very rarely, chronic anemia may be sufficiently severe to require regular blood transfusion treatment with associated iron chelation. In some patients, splenectomy has been beneficial. In severe cases, bone marrow transplantation would be a rational form of treatment for patients who have a suitable donor, provided there are no systemic manifestations other than hemolytic anemia, and provided it is carried out before there is organ damage (e.g., from iron overload). ADDITIONAL NOTES ON INDIVIDUAL GLYCOLYTIC ENZYMOPATHIES HEXOKINASE (HK) Several isoenzymes of HK are known. The main isoenzyme in normal red cells is type 1. In the very rare cases of HK deficiency, this defect is only expressed in red cells, presumably because the other isoenzymes contribute sufficient HK activity in other cells. It is possible that even in red cells HK2 (normally present as a minor species) might help in damage limitation. GLUCOSE 6-PHOSPHATE ISOMERASE (GPI) Most mutations entail single amino acid replacements, compatible with some residual enzyme activity (Fig. 22-2, p. 202). The splicing defects near exon 16 and the stop codon mutation in exon 18 might still result in the production of some enzymatically active protein. The nonsense mutation in exon 4 might be lethal in the homozygous state, but it has been found, together with a missense mutation, in two brothers with HA who were genetic compounds for these GPI mutations. PHOSPHOFRUCTOKINASE (PFK) In many metabolic pathways, the first reaction is the one most exquisitely regulated (Fig. 22-3, p. 202). In the case of glycolysis in red cells it appears that PFK, the second kinase of the pathway, is more subject to regulation, particularly by ATP and adenosine monophosphate (AMP), which affect both kcat and KmF6P. An interesting point of potential physiological significance is that this step is downstream of the pentose phosphate shunt; thus, the two pathways can be

CHAPTER 22 / RED CELL ENZYMOPATHIES

regulated independently. Five isoenzymes encoded by two genes are present in red cells. In PFK deficiency, all the mutations have been in the M subunit (Table 22-1), which explains the basis for the myopathy (glycogenosis type VII). In all these cases, we presume that the contribution of the L4 species of PFK will be unaffected, and hybrid tetramers may be partially active. This may explain why the HA associated with PFK deficiency is rather mild. Conversely, it is possible that L subunit mutations may exist, but they may not cause HA as long as the M subunit is normal. ALDOLASE Aldolase A, present in red cells, is a different enzyme from that which, when deficient in liver, causes fructose intolerance. Very few families with this red cell enzymopathy have been reported, and in only two was the molecular defect identified. In one family it was a missense mutation, causing an amino acid replacement of glycine for aspartic acid at position 128, a change that was shown to cause marked instability of the enzyme. Two members of the family were homozygous for this mutation and had enzyme levels of about 5% of normal, showing convincingly that aldolase deficiency can cause HA. TRIOSEPHOSPHATE ISOMERASE (TPI) This is a very rare but one of the most serious red cell enzymopathies, because HA is usually associated with severe involvement of the CNS (Fig. 22-4, p. 203). One might reasonably surmise that this is owing to the fact that a low level of TPI compromises some crucial metabolic step in the CNS; however, there is a single report in the literature of two brothers with TPI deficiency, one of whom died with severe mental retardation, whereas the other lives at the age of 22 with only hemolytic anemia. This intriguing observation would be consistent with the CNS manifestations requiring the coexistence of a mutation in some other unlinked gene. Another extraordinary finding has been that of the same missense mutations (105 Glu→Asp) in 17 unrelated families from distant parts of the world. In addition, this mutation is in strong linkage disequilibrium with several polymorphic sites. DIPHOSPHOGLYCERATE MUTASE (DPGM) DPGM is the only enzyme in the table that is not housekeeping; on the contrary, it is probably erythroid-specific. It is important in controlling the level of DPG, which is a regulator of the hemoglobin–oxygen dissociation curve. With DPGM deficiency, the DPG level is reduced, with consequent decreased delivery of oxygen to tissues, functional hypoxia, and polycythemia. PHOSPHOGLYCERATE KINASE (PGK) Like most other enzymes in the table, the X-linked gene encoding PGK is expressed in virtually all cells: a notable exception are sperm cells, in which a closely related autosomal PGK gene is expressed (Fig. 22-5, p. 203). Apart from HA, patients with PGK deficiency often have either CNS involvement, a myopathy, or both. Attempts have been made to correlate various combinations of clinical manifestations with the extent of enzyme instability produced in red cells and in muscle cells by individual mutations. However, the correlation does not always hold. PYRUVATE KINASE (PK) PK deficiency is by far the most common glycolytic enzymopathy (Fig. 22-6, p. 204), and it is also that for which the largest number of mutant alleles has been reported (Table 22-1). In a recent survey of 30 patients (60 alleles), 19 different mutations were observed. A single mutant, 1529A (corresponding to 510 Arg→Gln in Fig. 22-6), accounted for 42% of the total. This mutation had been already found in several families, and it is in strong linkage disequilibrium with a polymorphic site elsewhere in the gene, consistent with a single ancient origin.

199

Although this might be an example of founder effect, it is also compatible with some selective advantage of this allele in heterozygotes, possibly related to a slightly increased DPG level in these subjects. Homozygotes are at a disadvantage because of HA, but this is often characteristically well-tolerated, as a markedly increased DPG favors oxygen supply to the tissues. One patient with PK deficiency is known to have successfully completed a marathon run (EC Gordon-Smith, personal communication). PK deficiency is also the only glycolytic enzymopathy for which a significant proportion of mutant alleles are of the “null” variety, by virtue of having nonsense or frameshift mutations. In most cases, residual activity can be provided by the other allele in genetic compounds. However, several patients have been identified who were homozygous for the frameshift mutation called Gipsy. In these patients, some residual activity may be contributed by the M isoenzyme, present at low abundance in red cells. A mouse model of PK deficiency has been described that exhibits splenomegaly and nonspherocytic hemolytic anemia. This strain may be useful as an experimental model of PK deficiency.

GLUCOSE 6-PHOSPHATE DEHYDROGENASE (G6PD) DEFICIENCY EPIDEMIOLOGY G6PD deficiency is distributed worldwide. Areas of high prevalence are found in Africa, Southern Europe, the Middle East, South East Asia, and Oceania. In the Americas and in parts of Northern Europe, G6PD deficiency is also quite prevalent as a result of migrations that have taken place in relatively recent historical times. The overall geographic distribution of G6PD deficiency and its heterogeneity, together with clinical field studies and in vitro culture experiments, strongly support the view that this common genetic trait has been selected by Plasmodium falciparum malaria, by virtue of the fact that it confers a relative resistance against this highly lethal infection. CLINICAL FEATURES Three types of clinical presentations are well characterized: acute hemolytic anemia, neonatal jaundice, and chronic nonspherocytic hemolytic anemia. 1. The vast majority of G6PD deficient people are asymptomatic most of the time, but they are at risk of developing acute hemolytic anemia (AHA), which may be triggered by drugs, infections, or fava beans. During a hemolytic attack, the hemoglobin may fall very little, or it may plummet to values as low as 40 g/L or less; however, even if severe, the anemia is usually self-limited and tends to resolve spontaneously. Depending on the proportion of red cells that have been destroyed (reflected in the severity of the anemia), the hemoglobin level may be back to normal in 3–6 weeks. 2. The risk of developing neonatal jaundice (NNJ) is much greater in G6PD deficient than in G6PD normal newborns. NNJ related to G6PD deficiency—unlike “classical” Rhesus-related NNJ—is very rarely present at birth; the clinical onset is usually between day 2 and day 3. The severity varies enormously from being subclinical, to overlapping with “physiological jaundice,” to imposing the threat of kernicterus, if not treated. Prematurity, infection, and environmental factors (e.g., the use of naphthalene— camphor balls—used in baby bedding and clothing) are known aggravating factors.

Enzyme Hexokinase (HK) Glucose 6-phosphate isomerase (GPI) Phosphofructokinase (PFK)h

1 M L A

1 B 1 (α) Rm B 1

Prevalence of enzyme deficiency

Main clinical features associated with enzyme deficiencyc

Very rare Rare Very rare

HA HA, NM, CNS HA, myopathy

Very rare Very rare Very rare Very rare Very rare

HA, myopathy HA, CNS, NM, HA Polycythemia HA, CNS, NM

Very rare Rare Common Rare Very rare Very rare Very rare Very rarer

HA HA HA pseudocyanosis, CNS HA, CNS HA HA, CNS ?r

Benefit from splenectomyd

Chromosomal localization

Number of exons

Partial Partial

10q22 (q11?) 19q13.1 1cen-q32 21q22.3 16q22–24 12p13 12p13.31–p13.1 7q31-q34 Xq13 10q25.3 1pter-p36.13 1q21 Xq28 22q13.31–qter 9q34.1 6p12 20q11.2 3q11–q12

18 18 24 22 12 7 9 3 11

None Partial Partial None Partial

12 13 9 7 12

Number of known mutationsf Deletion–insertion Enzyme Hexokinase (HK) Glucose 6-phosphate isomerase (GPI) Phosphofructokinase (PFK)h Aldolase Triosephosphate isomerase (TPI) Glyceraldehyde 3-phosphate dehydrogenase (GAPD)k Diphosphoglycerate mutase (DPGM) Phosphoglycerate kinase (PGK) Monophosphoglycerate mutase (PGAM-B) Enolasek Pyruvate kinase (PK) Glucose 6-phosphate dehydrogenase (G6PD) Cytochrome-b5 reductase Adenylate kinase (AK) γ-Glutamylcysteine synthetase (GLCLC)k,q Glutathione synthetase (GSS) Glutathione peroxidase (GSH-Px)

Number of amino acidse 917 558 780 784 364 249 335 259 417 254 434 574m 515 276o 194 637 474 201

5'UTRj

Missense 1 19 7

1

2 9

Nonsense

65 115n 7 2 13

With frameshift

Affecting splicing

Total

1–0

2 6

4g 23 15

3–0 2 1 2

1 8 1

In frame

1–0 1–0 1–0

7 1 2 1

2i 13 2

2l 11

2–2 6–0 2–0

7–3

9 1 2

96 123 13 3p

1–0

1–0

1

16

SECTION III / HEMATOLOGY

Aldolase Triosephosphate isomerase (TPI) Glyceraldehyde 3-phosphate dehydrogenase (GAPD)k Diphosphoglycerate mutase (DPGM) Phosphoglycerate kinase (PGK) Monophosphoglycerate mutase (PGAM-B) Enolasek Pyruvate kinase (PK) Glucose 6-phosphate dehydrogenase (G6PD) Cytochrome-b5 reductase Adenylate kinase (AK) γ-Glutamylcysteine synthetase (GLCLC)k,q Glutathione synthetase (GSS) Glutathione peroxidase (GSH-Px)

Isoenzymeb characteristic of red cells

200

Table 22-1 Synopsis of Red Cell Enzymopathiesa

CHAPTER 22 / RED CELL ENZYMOPATHIES

aWe have listed all enzymes in the intermediary metabolism of red cells for which—to the best of our knowledge—the corresponding cDNA/gene has been cloned. There are other enzymes the deficiency of which may be associated with HA, but for which no molecular information is yet available: for instance pyrimidine 5'-nucleotidase. bNo entry in this column means that there are no known isoenzymes; therefore, it is assumed that the same enzyme type is present in all tissues. cThe following abbreviations have been used. CNS, central nervous system involvement; HA, hemolytic anemia; NM, neuromuscular manifestations. dData available only on some patients. eIncluding N-terminal methionine, which is cleaved off in most or all cases. fEach individual molecular change, if observed in more than one patient, has been counted only once. gTwo HK mutations were found in the same patient: 529 Leu→Ser and ∆162-193. Two large deletions were found in another patient: 䉭 exons 3–14, ∆ exons 5–8. hPFK in normal red cells consists of a mixture of the five tetrameric species that can be formed from random association of the M (muscle) and L (liver) highly homologous subunits (i.e., M , 4 M3L, M2L2, ML3, L4). iThe only known aldolase mutations are 128 Asp→Gly and 206 Glu→Lys. j5'UTR, 5' Untranslated region. kSince no mutations have yet reported, there is no formal proof that HA associated with this enzyme deficiency is due to mutation of the corresponding gene. lThe only two DPGM mutations known were found in the same patient. They were: 89 Arg→Cys and ∆ C205 (or 206). The deletion causes a frameshift resulting in an abnormal protein of 46 amino acids in which only the first 19 N-terminal amino acids are correct. mThe red cell form of PK called R is produced by the gene encoding the L (liver) subunit. Because a different promoter is used (see Fig. 22-5) the size of liver PK is 543 amino acids. nThe 115 missense mutations include two variants with normal activity, A and São Borja. Seven variants have 2 missense mutations each: these include G6PD Santamaria, G6PD Mount Sinai and the three variants that are called G6PD A—both of which have the mutation of G6PD A plus another mutation; G6PD Honiara and G6PD Bangkok. G6PD Vancouver variant has 3 different missense mutations (see Fig. 22-6). oThe cytoplasmic form of this enzyme, present in red cells, differs from the microsomal form present in other cells because, as a result of an alternative splicing pathway, it lacks the first 25 Nterminal amino acids. Therefore, in other cells, the size of the enzyme is 301 amino acids. pThe known AK mutations are: 128 Arg→Trp, 164 Tyr→Lys, and 107 Arg→Stop. qγ-glutamylcysteine synthetase consist of two subunits, a catalytic subunit and a regulatory subunit; the data concerning the catalytic subunit are shown here. rThere is no clear evidence that inherited deficiency of glutathione peroxidase exists.

201

202

SECTION III / HEMATOLOGY

Figure 22-2 Diagram of mutations in the GPI gene. In this and in the following figures, the central vertical line represents the genomic structure, with exons shown as thick blocks and introns as thin lines. Exons are numbered. Long introns are signalled by a diagonal double line. 123Additional symbols are as follows: 䊏, trans123 lated exon region; 123, untranslated exon region; 䉭, deletion; Ins, insertion; ∫, splicing error. The GPI genomic gene spans approximately 40 kb. A question mark (?) next to a diagonal double line indicates that the exact length of the intron is unknown. aThis variant has a mutation in the acceptor site that causes an aberrant splicing of exon 16. bThis variant has a deletion of the last 2 nucleotides of exon 16 and of the first 2 nucleotides of intron 16, with possible abnormal splicing.

3. Chronic nonspherocytic hemolytic anemia (CNSHA). In contrast to the large majority of G6PD deficient subjects who have minimal and subclinical hemolysis in the steady state, a small minority of G6PD deficient individuals have chronic anemia of variable severity. This rare condition is rather similar to CNSHA associated with glycolytic enzymopathies (see above) and, again, it is of variable severity. However, it is characteristically exacerbated by the same agents that can cause acute hemolytic anemia in people with the ordinary type of G6PD deficiency. DIAGNOSIS When there is anemia, it is usually normocytic and normochromic, and it may be from moderate to extremely severe. In AHA, the anemia results from intravascular hemolysis; and hence, it is associated with hemoglobinemia, hemoglobinuria, and low or absent plasma haptoglobin. The blood film shows anisocytosis, polychromasia, and other features associated with acute

Figure 22-3 Diagram of known mutations in the PFK-M gene. The PFK-M genomic gene spans approximately 30 kb. aThis mutation at the donor site of intron 5 causes exon 5 to be skipped (in frame) and thus a protein lacking 26 amino acids is produced. bEach one of these two PFK mutations has been found recurrently in several Ashkenazi families. cThis mutation at the acceptor site of intron 6 causes the production (thanks to the presence of two cryptic splice sites in exon 7) of two abnormal mRNA species, one lacking 5-bp and the other lacking 12-bp: the latter is more abundant and in frame, thus producing a protein lacking 4 amino acids. dThe mutation in the donor site causes splicing to a cryptic site in the intron, resulting in the retention of some intronic sequence and a premature stop. eThe mutation at the donor site of intron 15 causes splicing to a cryptic splicing site in exon 15 (in frame) and thus a protein lacking 25 amino acids is produced. f This mutation at the donor site of intron 19 causes exon 19 to be skipped (in frame), thus, a protein lacking 55 amino acids is produced. g This single basepair deletion causes a frameshift with a stop codon 47 nucleotides downstream and a truncated protein of 683 amino acids, with the last 16 C-terminal amino acids being incorrect. (For other symbols, see legend to Fig. 22-2.)

hemolysis, including spherocytes. In severe cases, the poikilocytosis is very marked, with presence of bizarre forms, numerous red cells that appear to have an uneven distribution of hemoglobin within them (hemighosts), and red cells that appear to have had parts of them bitten away (“bite cells,” or “blister cells”). Supravital staining with methyl violet, if done promptly, reveals the presence of Heinz bodies, consisting of precipitates of denatured hemoglobin. In CNSHA, the morphology is less characteristic. The final diagnosis must rely on the direct demonstration of decreased activity of G6PD in red cells. In most cases, one of the commercially available “spot tests” is adequate, but the gold standard is, as for all other enzymes, a spectrophotometric assay.

CHAPTER 22 / RED CELL ENZYMOPATHIES

Figure 22-4 Diagram of mutations in the TPI gene. The TPI genomic gene spans 3.2 kb. aSequencing of this variant reveals of two single base change in the 5' untranslated region. It is not clear whether and how this causes enzyme deficiency. bThe deletion of these two nucleotides causes a frameshift. (For other symbols, see legend to Fig. 22-2.)

GENETIC BASIS G6PD is a homodimeric molecule, and its single subunit is encoded by an X-linked gene (Table 22-1). Therefore, in areas with high prevalence of G6PD deficiency, male hemizygotes and female heterozygotes are common, but female homozygotes are rare. As a result of the phenomenon of X-chromosome inactivation in somatic cells, female heterozygotes are genetic mosaics, in whom approximately one-half of the red cells are normal and approximately one-half are G6PD deficient; however, sometimes the ratio is greatly imbalanced. Clinical manifestations in heterozygotes are milder than in hemizygotes and in homozygotes, roughly in proportion to the fraction of red cells that are G6PD deficient. MOLECULAR PATHOPHYSIOLOGY AHA associated with G6PD deficiency results from the action of an exogenous factor on intrinsically abnormal red cells. Although the sequence of events ending in hemolysis is not completely understood, it is clear that the exogenous agent taxes the capacity of the red cell to detoxify oxygen radicals, which is impaired by the short supply of NADPH in G6PD deficient red cells. Hence, the notion of oxidative hemolysis. AHA is seen with variants of G6PD that retain some 10% of the normal enzyme activity, or even less, if the kinetic parameters are favorable. By contrast, with some other variants the steady-state level of G6PD is so low that, even in the absence of any oxidant challenge, it becomes limiting for red cell survival. This is the case in the patients with CNSHA, who may have a red

203

Figure 22-5 Diagram of known mutations in the PGK gene. The PGK genomic gene spans approximately 23 kb. The methionine initiator is counted as residue zero. , hemolytic anemia; , myopathy; *, CNS involvement. aThis mutation in the donor site of intron 3 causes splicing to a cryptic site within intron 3 with the consequent insertion of 10 extra amino acids in the protein. bThis mutation in exon 7 causes either a missense mutation and the alteration of the consensus donor splice sequence site with alternative splicing to a cryptic site in intron 7 and a premature stop. cThe variant PGK-II, found at a polymorphic frequency in New Guinea, has normal activity. (For other symbols, see legend to Fig. 22-2.)

cell life-span of between 10 and 50 days. Very numerous point mutations in the G6PD gene causing CNSHA have been identified (see Fig. 22-7). Although we cannot explain the reason for a severe clinical phenotype in every case, a cluster of mutations causing CNSHA in exons 10 and 11 corresponds closely to the region of the molecule where the two subunits interface. It is not surprising that amino acid replacements in this region will interfere with dimer formation or will cause marked instability of the dimer. MANAGEMENT The most common manifestations of G6PD deficiency, NNJ and AHA, are largely preventable or controllable by screening, surveillance, and avoidance of triggering factors, particularly fava beans, by G6PD deficient subjects. When a patient presents with AHA, and once the cause is diagnosed, no specific treatment may be needed if the episode is mild. However, at the other end of the spectrum—especially in children—AHA may be a medical emergency requiring an immediate blood transfusion. The management of NNJ does not differ from that of NNJ owing to causes other than G6PD deficiency. To prevent neurological damage, it may have to include phototherapy and/or exchange blood transfusion. The management of CNSHA is simi-

204

SECTION III / HEMATOLOGY

different mutations have been identified: three patients were homozygotes, six were compound heterozygotes, and in three only a missense mutation was found in heterozygosity (Fig. 22-8).

CYTOCHROME B5 REDUCTASE DEFICIENCY A small amount of methemoglobin is formed in red cells continuously as a by-product of the cyclic oxygenation and deoxygenation of hemoglobin. It is estimated that methemoglobin would accumulate in blood at the rate of about 2% per day, unless it was continuously reduced. Red cells have both a NADPH-dependent and a NADH-dependent mechanism for the reduction of methemoglobin, but several lines of evidence indicate that the latter is by far more important physiologically. Cytochrome-b5 reductase (formerly called diaphorase) is the currently accepted name for the enzyme that reduces methemoglobin using NADH as the electron donor. Patients with cytochrome-b5 reductase deficiency can be also described as having methemoglobinemia. However, because this is caused by an enzyme defect rather than to a hemoglobin abnormality (Hb M), the former term is preferable. PREVALENCE Cytochrome-b5 reductase deficiency is relatively common in Alaskan Eskimos, but it is a rare condition elsewhere.

Figure 22-6 Diagram of mutations in the PK gene. The PK genomic gene spans 8.6 kb. The same gene encodes the R (red cell) and L (liver) form of PK. In the erythroid cell transcription starts from the upstream promoter (Pr );123 in the liver transcription starts from the downstream promoter (Pl 123 ). Thus, exon I of erythroid PK mRNA is 123 missing in liver PK mRNA. aThe consequences of this mutation at the acceptor site of intron 2 on RNA processing are not known. b This deletion removes the entire exon 11 and parts of intron 10 and intron 11 (total 1149 bp); it also causes a frameshift. (For other symbols, see legend to Fig. 22-2.) For reasons of space not all mutations can be shown.

lar to that of CNSHA caused by glycolytic enzymopathies, but, in addition, it is important to avoid exposure to potentially hemolytic drugs. Again, although there is no evidence of selective red cell destruction in the spleen (e.g., as seen instead in hereditary spherocytosis), splenectomy has proven beneficial in severe cases.

GLUTATHIONE SYNTHETASE DEFICIENCY Glutathione (GSH) is a ubiquitous tripeptide with a range of important biological functions. In red cells, the main function of GSH is protection against oxidative damage. GSH is produced from glutamic acid, cysteine, and glycine through the action first of γ-glutamylcysteine synthetase and then glutathione synthetase (GSS) (see Fig. 22-1). GSS deficiency is a rare autosomal recessive disorder resulting in very low levels of GSH in red cells. Clinically, GSS deficiency may present in two different clinical forms: a mild form that causes only a compensated hemolytic anemia, and a severe form that causes also 5-oxyprolinuria and neurological manifestations. Twelve unrelated patients with GSS deficiency have been studied for mutations in the GSS gene and 16

CLINICAL FEATURES The main finding is cyanosis, because since this is a result of the presence of methemoglobin rather than deoxyhemoglobin, it is sometimes referred to as pseudocyanosis, but with the naked eye, the two are not easily distinguished. There may be also a mild erythrocytosis, resulting from the erythropoietin-drive associated with the reduced oxygenation of tissues. A subset of patients with cytochrome-b5 reductase deficiency (i.e., designated as having type II disease) may have also mental retardation, which can be severe, and other neurological signs including microcephaly, opisthotonus, and athetoids movements. DIAGNOSIS Usually a patient presenting with cyanosis is suspected to have congenital heart disease or acquired cardiac or pulmonary disease; the suspicion may be reinforced by the coexistence of erythrocytosis. When the appropriate investigations are negative, a test for methemoglobin will explain the (pseudo)cyanosis. The differentiation of methemoglobinemia caused by cytochrome-b5 reductase deficiency from that caused by Hb M is based on electrophoretic and spectral studies of hemoglobin. An enzyme assay or a therapeutic trial of methylene blue will clinch the diagnosis. GENETIC BASIS Clinically affected individuals are homozygotes or genetic compounds; therefore, methemoglobinemia secondary to cytochrome-b5 reductase deficiency exhibits an autosomal recessive pattern of inheritance. This is in contrast to methemoglobinemia associated with Hb M, in which the pattern of inheritance is autosomal dominant. MOLECULAR PATHOPHYSIOLOGY Several point mutations have been identified in the cytochrome-b5 reductase gene in patients with congenital methemoglobinemia (Fig. 22-9). Not surprisingly, patients with type II disease have mutations that are different from those found in patients with type I disease: indeed, they include amino acid deletions and splicing mutations. Presumably, the resulting abnormal molecules are more unstable, or compromised in catalytic efficiency; therefore, the level of enzyme activity becomes severely reduced, not only in red cells, but also in nerve cells. In the latter, the enzyme is thought to be involved in lipid synthesis.

CHAPTER 22 / RED CELL ENZYMOPATHIES

205

Figure 22-7 Diagram of mutations in the G6PD gene. The G6PD genomic gene spans approximately 18 kb. , variant with normal enzymatic activity; , variant that cause acute hemolytic anemia; , variant that cause chronic nonspherocytic hemolytic anemia (CNSHA). a This variant has, in addition to the mutation shown, also the mutation (454 Arg→Cys) of G6PD Andalus. bThese variants have, in addition to the mutation shown, also the mutation (126 Asn→Asp) of G6PD A. cThis variant has been reported to have three different mutations; two are unique (106 Ser→Lys and 182 Arg→Trp), whereas one is the mutation (198 Arg→Cys) of G6PD Coimbra. dThe deletion of the last two nucleotides of intron 10 destroys the acceptor site with unknown effect on the processing of the transcript. (For other symbols, see legend to Fig. 22-2.) For reasons of space not all mutations can be shown.

MANAGEMENT Most patients with type I disease suffer little disability. However, if they are fair-skinned, their discoloration may create a cosmetic problem: this responds very well to the administration of methylene blue or ascorbic acid. The former activates the NADPH-dependent pathway of methemoglobin reduction; therefore, it is not effective if the patient happens to be also G6PD deficient. Unfortunately, no treatment is known for the neurological manifestations of type II disease. Knowledge of the molecular lesion makes it possible to offer prenatal diagnosis to a couples at risk for type II disease.

FUTURE DIRECTIONS Red cells have been for decades a model system in the study of intermediary metabolism, and red cell enzymes have been a classical tool of biochemical genetics and of population genetics. With the cloning of most of the genes encoding red cell enzymes, it has been possible to validate many of the items predicted in earlier times: for instance, the prediction that most of the individuals affected with rare autosomal enzymopathies—barring inbreeding—would be genetic compounds rather than true homozygotes. From the point of view of genotype–phenotype correlations, the

206

SECTION III / HEMATOLOGY

Figure 22-8 Diagram of mutations in the GSS gene. The GSS genomic gene spans approximately 23 kb. aThis single base deletion causes a frameshift. bThe mutation in the donor site results in aberrant splicing, whereby exon 4 is skipped. cThis varient has normal activity. (For other symbols, see legend to Fig. 22-2.)

analysis of at what point mutations cause what degree of change in stability or in catalytic properties of each enzyme has only just begun. The universe of patients with enzymopathies constitutes a living spontaneous mutagenesis laboratory in which structure– function relationships are, so to speak, displayed in vivo. From the practical point of view, the molecular analysis has not yet had any impact on management, which is still unsatisfactory, but it does provide the option of prenatal diagnosis in cases where previous family history has led to identifying a specific enzymopathy. Because chronic hemolytic anemia is the main clinical problem associated with these enzymopathies (in the majority of cases), they are an ideal target for correction by gene transfer into hematopoietic stem cells. Indeed, since the deficiency of most of these enzymes is likely to be a disadvantage even before erythroid cell differentiation reaches the end point of the mature red cell, it is conceivable that phenotypic correction might be self-selecting, thus abrogating the need for bone marrow ablation in future gene therapy protocols.

ACKNOWLEDGMENTS We thank our colleagues F Alfinito, L Baronciani, S Miwa, L Pastore, A Rovira, JL Vives-Corrons, TJ Vulliamy, and A Zanella, for kindly communicating information on mutations not yet published. We are also very grateful to Professor A Afolayan and Professor B Rotoli for their support and critical reading of the manuscript.

Figure 22-9 Diagram of known mutations in the cytochrome b5 reductase gene. The genomic gene spans approximately 31 kb. , region of exon 1 and 2 encoding for transmembrane domain (25 aa), missing in the erythrocyte cytoplasmic form of the enzyme. The methionine initiator is counted as residue zero. There are two types of cytochrome-b5 reductase deficiency. In type I, the enzyme deficiency is limited to the red cell cytoplasmic form, and its main clinical manifestation is pseudocyanosis. In type II, the enzyme deficiency involves both the cytoplasmic and the microsomal forms of the enzyme, and all tissues are affected. In addition to pseudocyanosis, there is mental retardation and other neurological signs. It is likely that type I mutations affect mainly the stability of the enzyme, whereas type II mutations either produce no viable protein or compromise drastically its catalytic activity. aThis variant, found at a polymorphic frequency in African American, has normal activity. bThe mutation at the donor site of intron 5 causes the skipping of exon 5 and produces in red cells a truncated protein of 102 amino acids. cThe consequences of this mutation at the acceptor site of intron 8 on the RNA are not exactly known, but the protein product is undetectable by immunological analysis. (For other symbols, see legend to Fig. 22-2.)

SELECTED REFERENCES Baronciani L, Bianchi P, Zanella A. Hematologically important mutations: Red cell pyruvate kinase. Blood Cells Mol Dis 1996;22:259–264. Baronciani L, Beutler E. Molecular study of pyruvate kinase deficient patients with hereditary nonspherocytic hemolytic anemia. J Clin Invest 1995;95:1702–1709. Beutler E. Hemolytic anemia in disorders of red cell metabolism. New York, Plenum Medical Book Company, 1978. Beutler E, West C, Britton HA, et al. Glucosephosphate Isomerase (GPI) Deficiency Mutations Associated with Hereditary Nonspherocytic Hemolytic Anemia (HNSHA). Blood Cells Mol Dis 1997;23:402–409. Bianchi M, Magnani M. Hexokinase mutations that produce nonspherocytic hemolytic anemia. Blood Cells, Mol Dis 1995;21:2–8.

CHAPTER 22 / RED CELL ENZYMOPATHIES

Cappellini MD, Martinez di Montemuros F, De Bellis G, Debernardi S, Dotti C, Fiorelli G. Multiple G6PD mutations are associated with a clinical and biochemical phenotype similar to that of G6PD Mediterranean. Blood 1996;87:3953–3958. Cohen-Solal M, Valentin C, Plassa F, et al. Identification of new mutations in two phosphoglycerate kinase (PGK) variants expressing different clinical syndromes: PGK Creteil and PGK Amiens. Blood 1994;84:898–903. Dacie JV. Haemolytic anaemias. In: The Hereditary Haemolytic Anaemias, vol I, 3rd ed. London: Churchill Livingstone, 1985. Dacie JV, Lewis SM. Practical Haematology, 8th ed. London: Churchill Livingstone, 1995. Dahl N, Pigg M, Ristoff E, et al. Missense mutations in the human glutathione synthetase gene result in severe metabolic acidosis, 5oxoprolinuria, hemolytic anemia and neurological dysfunction. Hum Mol Genet 1997;6:1147–1152. Harris RW. The red cell—production, metabolism, destruction: normal and abnormal. Cambridge: Harvard University Press, 1963. Hollan S, Fujii H, Hirono A, et al. Hereditary triosephosphate isomerase (TPI) deficiency: two severely affected brothers, one with and one without neurological symptoms. Hum Genet 1993;92:486–490. Kanno H, Fujii H, Wei DC, et al. Frame shift mutation, exon skipping, and a two-codon deletion caused by splice site mutations account for pyruvate kinase deficiency. Blood 1997;89:4213–4218. Kishi H, Mukai T, Hirono A, Fujii H, Miwa S, Hori K. Human aldolase A deficiency associated with a hemolytic anemia: thermolabile aldolase due to a single base mutation. Proc Natl Acad Sci USA 1987;84: 8623–8627. Kreuder J, Borkhardt A, Repp R, et al. Inherited metabolic myopathy and hemolysis due to a mutation in aldolase A. N Engl J Med 1996;334:1100–1104. Lemarchandel V, Joulin V, Valentin C, et al. Compound heterozygosity in a complete erythrocyte bisphosphoglycerate mutase deficiency. Blood 1992;10:2643–2649. Lukens JN. Hereditary hemolytic anemias associated with abnormalities of erythrocyte anaerobic glycolysis and nucleotide metabolism. In: Lee GR, Bittell TC, Foerster J, Athens JW, Lukens JN, eds. Wintrobe’s Clinical Hematology, 9th ed. Philadelphia: Lea & Febiger, 1993; pp. 990–1005. Luzzatto L, Mehta A. Glucose 6-phosphate dehydrogenase deficiency. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The Metabolic and Molecular Bases of Inherited Disease, 7th ed. New York: McGrawHill, 1995; pp. 3367–3398.

207

Manabe J, Arya R, Sumimoto H, et al. Two novel mutations in the reduced nicotinamide adenine dinucleotide (NADH)-cytochrome b5 reductase gene of a patient with generalized type, hereditary methemoglobinemia. Blood 1996;88:3208–3215. Matsuura S, Igarashi M, Tanizawa Y, et al. Human adenylate kinase deficiency associated with hemolytic anemia: a single base substitution affecting solubility and catalytic activity of the cytosolic adenylate kinase. J Biol Chem 1989;264:10,148–10,155. McCaneu JR, Thomas K. Human testis-specific PGK gene lacks introns and possesses characteristics of a processed gene. Nature 1987;326:501–504. Miwa S, Fujii H. Molecular basis of erythroenzymopathies associated with hereditary hemolytic anemia: tabulation of mutant enzymes. Am J Hematol 1996;51:122–132. Naylor CE, Rowland P, Basak AK, et al. Glucose 6-phosphate dehydrogenase mutations causing enzyme deficiency in a model of the tertiary structure of the human enzyme. Blood 1996;87:2974–2982. Qualtieri A, Pedace V, Bisconte G, et al. Severe erythrocyte adenilate kinase deficiency due to homozygous A→C substitution at codon 164 of human AK1 gene associated with chronic haemolytic anaemia. Br J Haematol 1997;99:770–776. Raben N, Sherman JB. Mutations in muscle phosphofructokinase gene. Hum Mutat 1995;6:1–6. Schneider A, Cohen-Solal M. Hematologically important mutations: Triosephosphate isomerase. Blood Cells Mol Dis 1996;22:82–84. Schneider A, Westwood B, Yim C, et al. The 1591C mutation in triosephosphate isomerase deficiency, tightly linked polymorphism and a common haplotype in all known families. Blood Cells Mol Dis 1996;22:115–125. Sierra-Rivera E, Summar ML, Dasouki M, Krishnamani MRS, Phillips JA, Freeman ML. Assignment of the gene (GLCLC) that encodes the heavy subunit of gamma-glutamylcysteine synthetase to human chromosome 6. Cytogenetics and Cell Genetics 1995; 70: 278–279. Vieira LM, Kaplan JC, Kanh A, Laroux A. Four new mutations in the NADH-cytochrome b5 reductase gene from patients with recessive congenital methemoglobinemia type II. Blood 1995;85: 2254–2262. Vulliamy T, Luzzatto L, Hirono A, Beutler, E. Hematologically important mutations: glucose-6-phosphate dehydrogenase. Blood Cells Mol Dis 1997;23:302–313. Yoshida A. Molecular Abnormalities of phosphoglycerate kinase. Blood Cells Mol Dis 1996;22:265–267.

CHAPTER 23 / COAGULATION DISORDERS

23

209

Coagulation Disorders MARTINA DALY, ANNE GOODEVE, PETER WINSHIP, AND IAN PEAKE

INTRODUCTION Inherited disorders of the coagulation system have been recognized for many years, and their study formed the basis of laboratory experiments from which grew the concept of clotting factors and the coagulation cascade with its subsequent refinements. The genes encoding these factors have now also been identified and cloned (Table 23-1), and, in most cases, a series of causative mutations have been detected. The gene for von Willebrand factor (VWF) has also been identified and cloned. Although not a procoagulant factor, VWF plays an integral part in the primary hemostatic process, and its deficiency results in von Willebrands disease (VWD), the most common bleeding disorder. This chapter will concentrate on the genetic basis of the three most common disorders, hemophilia A (factor VIII deficiency), hemophilia B (factor IX deficiency), and VWD. Reference to rarer defects in other factors will also be made.

THE GENETIC BASIS OF HEMOPHILIA A Hemophilia A is a recessively inherited X-linked disorder of coagulation factor VIII (FVIII) with an incidence of 1 in 5000 in the male population. The gene for FVIII was identified and cloned in 1984 and is found on the long arm of the X chromosome at Xq28. Human FVIII is synthesized primarily in hepatocytes as an approximately 300-kDa single-chain polypeptide, having a repeated domain structure A1-A2-B-A3-C1-C2. Activation of FVIII occurs by thrombin cleavage at sites within the B-domain, giving rise to a heterodimer comprised of a constant sized light chain (A3-C1-C2) and a heavy chain (A1-A2). FVIII circulates in complex with VWF, an association that protects it from proteolytic degradation. Thrombin activation releases FVIII from VWF and allows it to function as a cofactor for FIXa in the conversion of FX to its activated form (FXa). FACTOR VIII GENE STRUCTURE The factor VIII gene is 186 kb in length and contains 26 exons. The FVIII mRNA is approximately 9 kb long and contains 7053 nucleotides of coding sequence. The mature factor VIII protein contains 2332 amino acids (Fig. 23-1). Of particular interest is intron 22 of the gene that contains a CpG island from which two adjacent genes, F8A and F8B are encoded in opposite orientations. This was the first example in humans of genes being found wholly contained within another gene. The F8A gene is in the opposite orientation to the FVIII From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

gene, is intronless, and produces a mRNA transcript of 1.8 kb found in many tissues. F8B is transcribed in the same direction as FVIII. Its first exon, encoding eight amino acids, is within intron 22 and, in the mRNA, is spliced to exons 23–26 of the FVIII gene. The function of the F8A and F8B genes and their predicted protein products are not known. F8A lies within a 9.5-kb region of DNA (int22h) repeated two additional times approximately 400 kb telomeric to FVIII. The three homologous regions (int22h 1, 2 and 3) are identical in over 99.9% of their sequence, and each of the three F8A genes is transcribed. DEFECTS IN THE FACTOR VIII GENE Following the cloning of the FVIII gene in 1984, it became possible to begin analysis for mutations resulting in hemophilia A. Southern blotting analysis initially revealed a few patients with total or partial deletions or point mutations at specific restriction enzyme sites (e.g., C to T transitions at CpG sites identified by TaqI digestion—see below). The more sensitive techniques now applied to mutation detection are all based on amplification of DNA by the polymerase chain reaction (PCR), followed by mismatch detection methods (e.g., SSCP, DGGE, and so on) and DNA sequencing to characterize the mutations found. These techniques can be applied to genomic DNA or cDNA, produced by reverse transcription and PCR amplification (RT-PCR) from FVIII mRNA. The current hemophilia A database demonstrates the variety of mutations responsible for FVIII defects and is summarized in Table 23-2. The mutations reported have not been located by a systematic search/analysis of the gene and thus cannot be taken as a true representation of the relative incidence of the different mutations within the hemophilia A population. The table demonstrates that mutations can occur throughout the entire coding region of the FVIII gene, although no promoter or polyadenylation signal mutations have been identified to date. Twenty-five percent of single basepair substitutions occur at CpG dinucleotides; multiple mutations at these sites are frequent in the database and reinforce its “mutation hotspot” nature. Despite this, most of these mutations are “private,” and are found in only one or a few unrelated families. In contrast to the above, a novel type of mutation is now known to occur in the FVIII gene and accounts for nearly 50% of cases of severe hemophilia A. This DNA inversion is mediated by the three copies of int22h one located in intron 22 of FVIII, the other two 400 kb 5' and telomeric to the gene (Fig. 23-1). Homologous recombination appears to occur between the intron 22 copy (int22h1) and either the distal (telomeric, int22h-3) or more proximal (int22h-2) copy, and this process occurs most readily at male meiosis when the X chromosome is largely unpaired. Recombina-

209

210

SECTION III / HEMATOLOGY

Table 23-1 Blood Coagulation Factor Genes Clotting factor

Gene, kb

Fibrinogen (I)

α5.4 β8.2 γ8.4 20.2 ~80 13 186 34 25 23 12 A160 B28 178

Prothrombin (II) Factor V Factor VII Factor VIII Factor IX Factor X Factor XI Factor XII Factor XIII von Willebrand Factor (VWF)

Location 4q23–q23 11p 1q21–q25 13q34 Xq28 Xq27.1 13q34–qter 4q32–q35 5q33–qter A 6p24–25 B 1q31–32;1 12p12–pter

mRNA, kb

Inheritance pattern

α2.2 β~2.5 γ~2.5 2 7 ~2.5 8.8 1.8 1.5 2 2-4 A4 B>2 9

Autosomal dominant or recessive Autosomal recessive Autosomal recessive Autosomal recessive X-linked recessive X-linked recessive Autosomal recessive Autosomal recessive Autosomal recessive Autosomal recessive Autosomal dominant or recessive

Figure 23-1 The factor VIII gene. (A) Structural domains of factor VIII. Triplicated A domains of 330 amino acids, the unique B domain of 980 amino acids and duplicated C domain of 150 amino acids, are shown. (B) Exon (vertical bar) and intron (horizontal line) structure of factor VIII gene. Intron 22 is shown expanded, the locations of F8A and F8B transcripts are shown and the three homologous regions int22h-1 within factor VIII, and -2 and -3 400 kb 5' and telomeric to factor VIII are indicated. (C) An intrachromosomal crossover event, mediated by homologous recombination between int22h-1 and -3 is shown. This results in the common distal form of factor VIII gene inversion shown in (D). (D) Following an inversion event, introns 1–22 of the factor VIII gene now lie 400 kb 5' and telomeric to exons 23–26 of the gene, in the opposite transcriptional orientation.

tion results in FVIII exons 1–22, in addition to adjacent DNA becoming inverted and relocated some 400 kb from its normal position, with exons 23–26 remaining in their original location. Therefore, intact FVIII transcript and protein cannot be produced.

Crossover with the most distal int22h sequence results in a type 1 inversion. This is the most common inversional event, accounting for 82% of all inversions and 35% of all patients with severe disease. Crossover involving the proximal int22h results in a type 2

211

CHAPTER 23 / COAGULATION DISORDERS

Table 23-2 Summary of 1998 Database for Mutations in the FVIII and FIX Genes in Patients with Hemophilia A and Ba Hemophilia Aa,b

Hemophilia Bc

536 304 48 80 155 27 48

1713 652 132 29d 389 55 58

Patient entries Unique molecular events Short deletions or insertions Large deletions Different amino acid substitutions STOP mutations Percentage involving CpG transitions

aFrom Kemball-Cook et al. Nucleic Acids Res 1998;26:216–219. http://europium.mrc.rpms.ac.uk bDoes not include the FVIII gene inversion responsible for up to 50% of all severe hemophilia A. cFrom Giannelli et al. Nucleic Acids Res 1998;26:265–268. http://www.umds.ac.uk/molgen/

haemBdatabase.htm. dNot included in Giannelli et al. (1998) database.

Table 23-3 Useful DNA Polymorphisms Within or Flanking the Human Factor VIII, Factor IX and von Willebrand Factor Genes Gene

Type of polymorphism

Location

Detection methodd

Heterozygosity (Caucasian)a

5' Int7 Int13 Int18 Int19 Int22 Int22 Int22 3' 3' 5' 5' Int1 Int3 Int3 Int4 Int4 Codon 148 Codon 192 3'

Probe PCR PCR PCR PCR Probe PCR PCR Probe Probe PCR PCR PCR PCR PCR PCR PCR PCR PCR PCR

0.40 0.33 0.80 (10 alleles) 0.39 0.38 0.01b 0.49 0.55 (6 alleles) 0.25 0.43 0.44 0.04 0.36 0.11 0.41 0.45 0.32 0.44 0.01c 0.48

Restriction enzyme

F VIII

RFLP TaqI (G/A) — VNTR CA repeat — RFLP BclI RFLP HindIII RFLP MspI RFLP XbaI VNTR CA repeat — RFLP BglI RFLP MspI F IX RFLP MseI RFLP BamHI DEL/INSe — RFLP BamHI RFLP XmnI RFLP TaqI RFLP MspI RFLP MnlI (A/G) — RFLP HhaI VWF RFLP: Over 30 have been reported. VNTR (AGAT repeats): two regions in intron 40. Multiallelic. aHeterozygosity rates can vary in different ethnic groups. b0.13 in Asians. c0.11 in Asians. dProbe (Southern blot) method when PCR-based technique eDimorphic with two major forms differing by 50 bp.

has not been reported.

inversion and occurs less frequently (15%). The number of copies of int22h outside the FVIII gene appears to be polymorphic, with some individuals having extra complete or partial copies that can result in rarer types of inversion (3A, 3B, or other variants). Inversions can be readily detected by Southern blotting following DNA digestion with BclI and using a probe for part of the int22h sequence. Overall, this inversion is the causative mutation in 43% of patients with severe disease and accounts for 20% of all cases of hemophilia A. GENETIC SCREENING FOR HEMOPHILIA A In families affected by hemophilia A, female relatives of patients wish to know their carrier status. This can be achieved either by tracking the defective FVIII gene through the family by using polymorphic

genetic markers or by detection of the causative mutation. The latter procedure is the most technically demanding and expensive, unless the inversion mutation is present, but is the ideal approach. Since mutation detection is not available in many laboratories, the approach most frequently adopted for noninversion families is that of gene tracking, using DNA polymorphism analysis. A number of intragenic restriction fragment length polymorphisms (RFLP) in the FVIII gene have been described. However, the CA dinucleotide repeat polymorphisms located in introns 13 and 22 are the most useful (see Table 23-3). Heterozygosity rates in females from different ethnic backgrounds vary from 30 to 90%. The suggested strategy for hemophilia A analysis is thus to first determine whether the FVIII gene inversion is present. If not, then

212

SECTION III / HEMATOLOGY

dinucleotide repeats should be examined; if these are not informative, try RFLPs. Approximately 85% of families examined will be informative by using one of these intragenic polymorphisms. The probability of error caused by crossover at meiosis between a factor VIII gene mutation, and the intragenic polymorphism used to track its inheritance is 90% of the population, with the MS and MZ phenotypes the next most common. The MS, MZ, and SS phenotypes, which are associated with modest deficiencies of α1-AT (about half of the normal serum concentration), do not present an increased risk of emphysema, although there has been an increased frequency of Pi MZ individuals in some COPD populations. Pi MS individuals may have an increased frequency of airway hyperreactivity. Because individuals with Pi SZ, who have an average of 37% of normal α1-AT serum concentration, rarely develop emphysema, serum levels >35% of normal are thought to be enough to provide protection. As mentioned above, cigaret smoke can oxidize a methionine residue in the reactive center of α1-AT, inactivating its capacity as a proteinase inhibitor. The potential consequences of this reaction were demonstrated in a dog model in which animals treated with chloramine-T, an agent that profoundly depresses α1-AT functional activity, developed pulmonary emphysema. However, the initial studies in smokers that demonstrated oxidatively inactivated α1-AT in the bronchoalveolar lavage fluid have not been corroborated. Several α1-AT phenotypes are associated with very low serum concentrations of α1-AT. Of these, the Pi Z phenotype is by far the most common, accounting for >95% of such individuals. Pi Z individuals have about 15% of the normal serum concentration of α1-AT. The prevalence of the Pi Z phenotype in the United States is about one in 3000 people. The Z allele is rare in Asians and African-Americans.

345

The small number of other individuals with marked deficiency of α1-AT have Pi SZ, Pi null-null, or Pi null-Z phenotypes. Most Pi Z individuals eventually become symptomatic with COPD as a result of emphysema, but there is considerable variation and some individuals reach advanced age with minimal symptoms. Silverman and colleagues confirmed the wide variability in pulmonary function among Pi Z subjects and found evidence for familial factors that segregated with deterioration in pulmonary function. Smoking has a marked effect on the age at which shortness of breath appears. On the average, Pi Z smokers have symptoms by age 40, about 15 years earlier than Pi Z nonsmokers. The Pi Z phenotype is caused by a point mutation involving a single nucleotide at codon 342 that results in coding for lysine instead of glutamic acid. This amino acid substitution alters the charge attraction between the amino acids at positions 342 and 290 present in the normal form of α1-AT and prevents the formation of a fold in the molecule. This change in tertiary structure promotes dimerization of α1-AT molecules. It is the dimerization that leads to the aggregation of α1-AT in the endoplasmic reticulum, which impedes secretion of the protein from the cell and results in the low levels of α1-AT in plasma and other body fluids. The Z form of α1-AT also has a rate of association with neutrophil elastase that is significantly slower than the association rate of normal α1-AT with neutrophil elastase. Thus, the Pi Z phenotype leads to both a deficiency of α1-AT protein and a form of α1-AT that is less effective than normal α1-AT as an inhibitor of neutrophil elastase. The S variant of α1-AT, which involves a single nucleotide alteration leading to substitution of glutamic acid264 with a valine, does not accumulate in the liver. This protein is less stable, presumably because of loss of a salt bridge between the glutamic acid in position 264 and the lysine in position 387. The Pi null phenotype arises because either the α1-AT gene is missing or there is a mutation in the α1-AT gene that results in premature termination of the gene’s transcription.

REPAIR OF LUNG ELASTIN The seminal observations about α1-AT deficiency and production of emphysema in animals with elastolytic enzymes (and only with elastolytic enzymes) led to the concept that destruction of elastin in the lung parenchyma is key to the development of emphysema. Elastin is the principal component of elastic fibers. Elastic fibers, which possess rubber-like reversible extensibility, come under tension and provide elastic recoil throughout the respiratory cycle. In the lung parenchyma, elastic fibers loop around alveolar ducts, form rings at the mouths of the alveoli, and penetrate as wisps into the alveolar septae, where they are concentrated at bends and junctions. Elastin is secreted as a soluble protein of 60–70 kDa called tropoelastin (Fig. 38-4). Tropoelastin molecules, encoded by a gene on chromosome 7 in the human, are deposited into the extracellular space and aligned on a “scaffold” of microfibrils that consist of a number of proteins, including fibrillins, microfibrilassociated proteins, and latent transforming growth factor-β (TGF-β) binding proteins. In the extracellular space, lysyl oxidase modifies most of the lysine residues in tropoelastin monomers, causing them to crosslink and form elastin, a highly insoluble, rubber-like polymer. The lysine-derived crosslinks in elastin are known as desmosines. Desmosines are unique to elastin, and therefore can be used to quantify elastin in tissues and as markers of elastin degradation in body fluids.

346

SECTION V / PULMONOLOGY

Figure 38-4 The synthesis of tropoelastin and assembly of the elastic fiber. Under the influence of extracellular and intracellular factors, tropoelastin pre-mRNA is transcribed within the nucleus of the elastogenic cell. Differential splicing of tropoelastin pre-mRNA leads to different tropoelastin mRNAs and tropoelastin isoforms. After tropoelastin is secreted from the cell it associates with microfibrils adjacent to the cell surface. Uncertainty exists as to whether there is a carrier protein that facilitates the secretion of tropoelasin. Microfibrils are thought to be a scaffold on which tropoelastin monomers align. On the microfibril, most of the lysines in tropoelastin monomers are modified by lysyl oxidase to form covalent crosslinks (desmosines) between the monomers. The resultant polymer is elastin. (Provided by William C. Parks, PhD).

Elastin is resistant to many proteinases, most notably the collagenases that cleave interstitial collagens, but there are a number of enzymes that may come in contact with the lung that can degrade elastin (Table 38-1). Under normal conditions, elastin synthesis in the lung begins in the late neonatal period, peaks during early postnatal development, continues to a much lesser degree through adolescence paralleling lung growth, and stops in adult life. There is some evidence that the tropoelastin gene always remains transcriptionally active but rapid mRNA degradation prevents expression of the protein. Multiple cell types are responsible for elastin synthesis in the lungs and associated structures. Lung elastin normally lasts a human lifespan, and there is virtually no elastin synthesis in the normal adult lung. Although destruction of lung elastin appears to be necessary for the development of emphysema caused by smoking, it remains unknown precisely how the breakdown of elastin translates into the deformity recognized as emphysema. Elastin depletion appears to be restricted to the sites of emphysema, rather than being a global deficiency of the lung that contains regions of emphysema. It has also been difficult to determine the capacity of the lung parenchyma to undergo repair after proteolytic injury. It is not known if normal elastic fibers can be properly formed in the lung after the period of growth and development. After an intratracheal injection of human neutrophil elastase into an experimental animal, there is acute depletion of elastin followed by a burst of extracellular matrix synthesis, so that, during the course of a few weeks, the elastin content of the lungs returns to normal, although the lungs develop emphysema. However, the newly synthesized elastic fibers appear disorganized, similar to the elastic fibers in human emphysema. Recent studies of lung growth after pneumonectomy in adult rats indicate that elastin synthesis can be reinitiated in the adult

lung and deposited at the sites at which elastin is normally produced during lung development. Nothing is known about the turnover of other extracellular matrix components in human lungs affected by COPD.

CONCLUDING COMMENTS Intrapulmonary proteolytic activity that degrades lung elastin is generally accepted as the predominant mechanism for emphysema in smokers. Although elastin degradation may be central to emphysema, the biology of emphysema is clearly complex and still poorly understood. Moreover, degradation of extracellular matrix components besides elastin, particularly collagens, may also be essential. To probe the roles of the many factors involved in the development of emphysema and the lung’s response to the remodeling called emphysema, researchers are now taking advantage of transgenic mice technology. From these studies, it should be possible to assess more precisely than ever before what is happening in the lung injury that culminates in the development of emphysema.

SELECTED REFERENCES American Thoracic Society Statement: Standards for the diagnosis and care of patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1995;152:S77–S120. Bartecchi C, MacKenzie T, Schrier R. The human costs of tobacco use. N Engl J Med 1994;330:907–912; 975–980. Cantor J. Emphysema, lung disease and retinoic acid. Nat Med 1997;3:817. Chapman HA Jr, Munger JS, Shi G-P. The role of thiol proteases in tissue injury and remodeling. Am J Respir Crit Care Med 1994;150:S155– S160. Cox DW. α 1 -Antitrypsin deficiency. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The Metabolic and Molecular Bases of Inherited Disease, 7th ed. New York: McGraw-Hill, 1995; pp. 4125–4158.

CHAPTER 38 / PULMONARY EMPHYSEMA

D’Armiento J, Dalal SS, Okada Y, Berg RA, Chada K. Collagenase expression in the lungs of transgenic mice causes pulmonary emphysema. Cell 1992;71:955–961. Hautamaki RD, Kobayashi DK, Senior RM, Shapiro SD. Macrophage elastase is required for the development of emphysema induced by cigarette smoke in mice. Science 1997, in press. Joslin G, Fallon RJ, Bullock J, Adams SP, Perlmutter DH. The SEC receptor recognizes a pentapeptide neodomain of alpha 1-antitrypsinprotease complexes. J Biol Chem 1991;266:11,282–11,288. Mecham RP, Davis EC. Elastic fiber structure and assembly. In: Yurchenko P, Birk D, Mecham R, eds. Extracellular Matrix Assembly and Structure. San Diego: Academic, 1994; pp. 281–314. Owen CA, Campbell MA, Sannes PL, Boukedes SS, Campbell EJ. Cell surface-bound cathepsin G on human neutrophils: a novel non-oxidative mechanism by which neutrophils focus and preserve catalytic activity of serine proteinases. J Cell Biol 1995;131:776–789. Perlmutter DH. Alpha-1-antitrypsin deficiency: biochemistry and clinical manifestations. Ann Med 1996;28:385–394.

347

Sanford AJ, Weir TD, Pare PD. Genetic risk factors for chronic obstructive pulmonary disease. Eur Respir J 1997;10:1380–1391. Shapiro SD. The pathogenesis of emphysema: elastase:antielastase hypothesis 30 years later. Proc Assoc Am Physicians 1995;3:346–352. Shapiro SD, Endicott SK, Province MA, Pierce JA, Campbell EJ. Marked longevity of human parenchymal elastic fibers deduced from prevalence of D-aspartate and nuclear weapons-related radiocarbon. J Clin Invest 1991;87:1828–1834. Silverman EK, Pierce JA, Province MA, Rao DC, Campbell EJ. Variability of pulmonary function in alpha-1-antitrypsin deficiency: clinical correlates. Ann Intern Med 1989;111:982–991. Snider GI. Emphysema: the first two centuries—and beyond: a historical overview with suggestions for future research: parts 1 and 2. Am Rev Respir Dis 1992;146:1334–1344; 1615–1622. Tetley TD. Matrix metalloproteinases: a role in emphysema? Thorax 1997;52:495. Wright JL. Emphysema: concepts under change—a pathologist’s perspective. Mod Pathol 1995;8:873–880.

CHAPTER 39 / SURFACTANT DEFICIENCY

39

349

Surfactant Deficiency JEFFREY A. WHITSETT AND TIMOTHY E. WEAVER

BACKGROUND Pulmonary surfactant is a complex mixture of phospholipids and associated proteins that reduces surface tension at the air– liquid interface at the alveolar surfaces. In the absence of surfactant, collapsing forces of ~70 dyne/cm2, generated by unequal intramolecular forces among water molecules at the alveolar surfaces, cause atelectasis, cyanosis, and associated respiratory distress. Pulmonary surfactant is a phospholipid–protein complex produced by type II epithelial cells lining the alveolus of the lung. Surfactant is secreted into the alveolar space, forming an array of macromolecular forms that produce a monolayer or multilayer of phospholipids at the air–liquid interface, reducing collapsing forces and maintaining lung volumes at end-expiration. The critical role of pulmonary surfactant in lung function was first discerned from the studies of respiratory distress syndrome (RDS) in premature infants. Subsequent biochemical, biophysical, and molecular studies have elucidated important features of the pulmonary surfactant system that have led to the widespread application of surfactant replacement therapy for RDS and other clinical abnormalities of surfactant. The present review will discuss the structure, function, and regulation of pulmonary surfactant, focusing attention to the role of surfactant proteins in surfactant function and respiratory distress associated with mutations in the surfactant protein-B gene. Readers are referred to several texts and reviews that are listed in the references.

SURFACTANT DEFICIENCY AND RESPIRATORY DISTRESS Surfactant deficiency is a life-threatening disorder, the pathophysiology of which is most clearly represented by the respiratory failure associated with infantile respiratory distress (IRDS) that accompanies premature birth. Pulmonary surfactant is composed primarily of phospholipids (PL) that are particularly enriched in dipalmitoylphosphatidylcholine (DPPC). Surfactant phospholipids are packaged in lamellar bodies of type II epithelial cells and secreted into the airspace by exocytotic processes stimulated by stretch, catecholamines, and purinoceptor agonists. After secretion, lamellar bodies form tubular myelin, a highly organized phospholipid–protein mixture that is the most abundant form of extracellular pulmonary surfactant (Fig. 39-1). Surfactant phospholipids move rapidly from tubular myelin to the gas–liquid interface, forming a monolayer in multilayers of phospholipids, From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

thereby creating an interface between alveolar gas and the fluid subphase overlying alveolar cells. The dense packing of phospholipids at this interface is maintained by the parallel packing of the acyl chains of the phospholipids, and further strengthened by interactions of the choline and glycerol head groups of the phospholipid molecules. This dense packing of phospholipids excludes interactions between gas and liquid molecules, reducing surface tension in the alveolus. The reduction of surface tension in the alveolus maintains lung volumes at end-expiration, stabilizing the alveolar spaces from collapse. Although surfactant proteins represent less than 10% of the mass of pulmonary surfactant, they appear to play a critical role in the routing, storage, secretion, and activity of pulmonary surfactant. Three abundant proteins, surfactant proteins A, B, and C, mediate the packaging of surfactant lipids within the type II cell, contribute to tubular myelin formation, and enhance the spreading and stability of surfactant phospholipids in the alveolar space. Deficiencies in these various components of surfactant contribute to lung dysfunction in a variety of human disorders. Surfactant proteins and lipids are subject to a variety of regulatory stimuli and their expression is precisely controlled during perinatal development, the concentrations of surfactant lipid and protein synthesis increasing markedly with advancing gestational age. In infants, the paucity of surfactant lipids and proteins related to prematurity is associated with a lack of biochemical and morphologic differentiation of the distal respiratory epithelium. Surfactant deficiency is also associated with respiratory failure in adult respiratory distress syndrome (ARDS), a disorder often associated with aspiration, infection, trauma, or shock in older patients. Advances in our understanding of the structure and function of surfactant have recently led to the widespread clinical use of exogenous surfactant replacement for prevention and therapy of IRDS in premature infants. Surfactant replacement is also being studied in the care of infants with other pulmonary disorders, including meconium aspiration, and is under investigation for therapy of ARDS and other lung disorders in older patients.

STRUCTURE AND FUNCTION OF SURFACTANT PROTEINS The recognition that mixtures of surfactant phospholipids alone do not form a fully active surfactant led to the identification of surfactant proteins A, B, and C and the recognition that each plays a unique role in the organization, function, and catabolism of pulmonary surfactant.

349

350

SECTION V / PULMONOLOGY

SURFACTANT PROTEIN A (SP-A) Gene: SPA-Structure and Regulation of Expression SP-A belongs to a family of proteins, referred to as calcium-dependent lectins (collectins), encoded by a cluster of genes on the long arm of human chromosome 10. There are two SP-A genes in the human that differ by approximately 1% in the coding region. Although both genes are transcribed, the functional significance of two genes remain unclear, particularly in view of the fact that other species (mouse, rat, and rabbit) contain only a single gene. Transcription from the SP-A locus is regulated both temporally and spatially. SP-A expression is restricted to alveolar type II epithelial cells and nonciliated bronchiolar cells (Clara cells) of the respiratory tree. Synthesis of SP-A mRNA is detectable as early as the second trimester of pregnancy, but secretion of SP-A into the amniotic fluid is not generally detectable until approximately 30 weeks’ gestation. Accumulation of SP-A in amniotic fluid during late gestation correlates strongly with surfactant function and lung maturity. The developmental, cell-specific, and basal regulation of SP-A expression is complex. Binding of the nuclear transcription protein TTF-1, thyroid transcription factor-1, to the SP-A promoter is required but is not sufficient to direct appropriate temporospatial SP-A gene transcription. The identification of transcription factors that modulate expression of SP-A independent of or by interacting with TTF-1 is an active area of research that will provide new insight into the regulation of SP-A expression during normal growth and development and in response to lung injury. Structure and Biosynthesis Translation of the human SP-A mRNA gives rise to a preprotein of 248 amino acids. Cotranslational cleavage of the 20-amino acid signal peptide results in an SP-A monomer comprising four discrete functional domains, including (1) a 7-residue amino-terminal domain involved in association of multiple SP-A subunits, (2) a 72-residue collagenlike domain, (3) a 130-residue C-terminal carbohydrate-recognition domain (lectin-binding domain), and (4) a short neck region that links the lectinbinding and collagenlike domains (Fig. 39-2). The cotranslational addition of asparagine-linked carbohydrate and subsequent processing of the oligosaccharide tree(s), including sialation and sulfation, results in multiple forms of monomeric SP-A ranging in size from 28,000 to 36,000 Dalton. Assembly of mature, oligomeric SP-A is initiated in the endoplasmic reticulum and involves the interaction of the collagenlike domains of three SP-A monomers to form a triple helix; subsequent formation of sulfhydryl bonds between the amino-terminal domains of adjacent trimeric subunits leads to the formation of the mature SP-A molecule comprising 18 monomers. Secretion studies in isolated Type II epithelial cells, fetal lung explants, and rabbit lung in vivo suggest that most of the newly synthesized SP-A is secreted independently of surfactant phospholipids. Functional Domains SP-A avidly binds to surfactant phospholipids, particularly DPPC, a property thought to be mediated by the C-terminal domain of the protein. The role of SP-A in organizing newly secreted surfactant phospholipids in the airway was deduced from in vitro studies demonstrating that SP-A promotes the calcium-dependent aggregation of lipid vesicles and is also required for the formation of tubular myelin. Binding of SP-A to a receptor on the surface of the type II epithelial cell results in an inhibition of secretagog-stimulated secretion of phospholipid consistent with a role for SP-A in modulation of surfactant secretion. SP-A has also been shown to facilitate the uptake of phospholipid by isolated type II cells. The effect of SP-A on structure,

Figure 39-1 Life cycle of surfactant. Newly synthesized surfactant lipids and proteins are stored in the form of lamellar bodies within the type II epithelial cell. Lamellar bodies are secreted into the alveolus and unravel into tubular myelin, a process facilitated by SP-A, SP-B, and calcium. Adsorption of phospholipids to the air–liquid interface and formation of the surface-active monolayer is facilitated by surfactant proteins B and C. Repeated compression and expansion of the monolayer results in the formation of small lipid vesicles, which are taken up by the type II cell and incorporated into lamellar bodies for reuse; a minor proportion of surfactant is degraded in lysosomes of type II cells or macrophages.

secretion, and uptake of surfactant phospholipid by the type II cell is consistent with the key role for this protein in regulation of surfactant homeostasis. This hypothesis, however, requires careful reassessment in view of the apparently normal lung function in mice lacking SP-A protein (SP-A knockout mice): genetic ablation of the mouse SP-A gene in transgenic mice markedly disrupted the formation of tubular myelin in homozygous SP-A(–/–) mice, but did not alter lung structure, phospholipid content, or function in vivo. Apart from surfactant-related functions, there is growing evidence that SP-A may play an important role in host defense of the airways. SP-A has been shown to stimulate migration of macrophages and to increase production of oxygen radicals involved in the destruction of pathogens internalized by macrophages. SP-A is an opsonin, enhancing phagocytosis of viruses and bacteria, and stimulates phagocytosis of bacteria opsonized by complement or immunoglobulins. Consistent with the broad binding specificity of SP-A, the carbohydrate moieties, the neck region, and the lectin-binding domain have all been implicated in the binding of SP-A to various pathogens. Further insight into the role of SP-A in nonimmune lung defense is certain to come from the study of SP-A knockout mice. Indeed, preliminary evidence with these SP-A(–/–) mice suggests their susceptibility to lung infection. SURFACTANT PROTEIN B (SP-B) SP-B Gene: Structure and Regulation of Expression SP-B is encoded by a single gene on human chromosome 2. Transcription from the SP-B locus

CHAPTER 39 / SURFACTANT DEFICIENCY

351

Figure 39-2 Structure of SP-A. SP-A is a 32- to 36-kDa polypeptide that forms trimers, which further assemble into hexamers and larger multimers. The molecules contains a globular lectin domain (C-terminal) and a rigid collagenous domain (N-terminal).

is regulated both developmentally and in a cell-specific manner. As for SP-A, expression of SP-B is restricted to Clara cells and type II cells of the respiratory epithelium. Cis-acting elements regulating cell-specific expression have been mapped to the proximal region of the SP-B promoter (–80 to –110) and shown to bind TTF-1 and HNF-3 (hepatocyte nuclear factor-3); in addition, TTF-1 binds at distal sites (–459 to –331) in the promoter resulting in significant enhancement of transcriptional activity. Phosphorylation of TTF-1 by cAMP-dependent protein kinase significantly increases transcription of SP-B consistent with a broad role for cAMP in the regulation of surfactant protein expression. Regulation of developmental expression by binding of specific nuclear proteins to sites within the SP-B promoter has not been systematically evaluated although it is likely that both TTF-1 and HNF-3 are involved in the temporal and spatial determination of SP-B gene expression. SP-B mRNA is detected as early as 12 weeks of gestation but significant amounts of protein are not detected in amniotic fluid before 31 weeks; secretion of SP-B increases significantly to term and is strongly correlated with surfactant function and lung maturity in infants. Overall, the developmental, cell-specific, and basal transcription of the SP-B gene is strongly influenced by TTF-1 and HNF-3; however, the widespread expression of these transcription factors in tissue derived from the embryonic foregut axis indicates that other nuclear proteins may also contribute to the unique pattern of SP-B expression. Structure and Biosynthesis Human SP-B is synthesized as a preproprotein of 381 amino acids; however, the active peptide associated with surfactant phospholipids is only 79 residues, indicating that extensive proteolytic processing occurs in type II epithelial cells (Fig. 39-3). In the alveolus, SP-B is closely associated with surfactant phospholipids, contributing to tubular myelin formation. Translocation of newly synthesized preproprotein into the endoplasmic reticulum is accompanied by cleavage of a 23-amino acid

signal peptide and the addition of asparagine-linked oligosaccharide to the COOH-terminal propeptide. Although glycosylation of the proprotein occurs in all species examined, the functional significance of this modification is not apparent; the mature peptide is not modified by oligosaccharide addition and glycosylation of the proprotein is not required for intracellular trafficking of SP-B. In contrast, the 177-residue NH2-terminal propeptide is absolutely required for trafficking of SP-B. The propeptide acts as an intramolecular chaperone to prevent nonspecific interactions of the extremely hydrophobic mature peptide with cellular membranes during transport through the secretory pathway; further, the NH2terminal propeptide is essential for directing the mature peptide to the lamellar body. Both the NH2- and COOH-terminal propeptides are proteolytically cleaved before incorporation of the mature peptide into the lamellar body. The function of the COOH-terminal propeptide, the metabolic fate of the cleaved propeptides, and the identity of the endoprotease(s) involved in their cleavage are currently not known. Functional Domains Mutations in the SP-B gene leading to a complete absence of mature peptide are neonatal lethal in human infants and in SP-B-deficient mice created by gene targeting. Fetal lung development proceeds normally in the absence of SP-B but inflation of the newborn lung is compromised resulting in the rapid onset of respiratory distress. It is likely that many of the histopathologic changes described in SP-B-deficient infants are related to intensive postnatal therapy. Several attempts to rescue SP-Bdeficient infants with surfactant preparations containing SP-B have not been successful, suggesting that replacement by the mature peptide alone is insufficient. Consistent with this conclusion, SP-B deficiency is accompanied by aberrant processing of the SP-C proprotein to a diagnostic fragment (Mr = 12,000) detectable in amniotic fluid or alveolar lavage; in addition, mature SP-C peptide is significantly reduced, likely exacerbating RDS in these infants. These findings, coupled with the observation that lamellar body

352

SECTION V / PULMONOLOGY

Figure 39-3 Structure of SP-B. Active SP-B, a 79-amino acid polypeptide, is produced by proteolytic processing of a 381-amino acid precursor. SP-B associates with the surfaces of phospholipid membranes, contributing to the adsorption and stability of the phospholipids.

formation is abnormal in SP-B knockout mice, suggest that the SP-B proprotein plays a critical role in assembly of the bioactive pulmonary surfactant complex. The mature SP-B peptide is very hydrophobic and avidly associates with surfactant phospholipids. There are four regions within the peptide that have the potential to form amphipathic helices, and one or more of these domains may facilitate its interaction with phospholipids. The presence of a large number of cationic residues in SP-B likely accounts for the affinity of this peptide for phosphatidylglycerol, an abundant surfactant phospholipid containing a negatively charged head group. SP-B has been shown to accelerate the transition of newly secreted surfactant to the surface-active film and facilitate respreading of the surface film after dynamic compression and expansion. SP-B has also been shown to stimulate the uptake of phospholipids by type II cells in vitro, suggesting that the peptide may play a role in clearance of surfactant. SURFACTANT PROTEIN C (SP-C) SP-C Gene: Structure and Regulation of Expression SP-C is encoded by a single gene on human chromosome 8. As for surfactant proteins A and B, transcription of the SP-C locus is regulated both developmentally and in a cell-specific manner; however, unlike the other surfactant proteins, expression of SP-C is restricted to type II epithelial cells. Although TTF-1 plays an important role in the regulation of basal and lung-specific transcription of the SP-C gene, other as yet unidentified transcription factor(s) are clearly required to maintain type II cell-specific expression. Elements of the SP-C promoter contained within 3.7 kb of the 5' flanking sequence of the gene have been shown to confer appropriate lung-specific and developmental expression on a reporter gene in transgenic mice. SP-C mRNA is detected at low levels early in the second trimester and increases significantly during the third trimester coincident with mRNAs for SP-A and SP-B. Assessment of SP-C protein levels is complicated by the lack of a monospecific antiserum directed against the active airway peptide; however, immunocytochemical studies with antisera directed against the SP-C proprotein indicate that the time course of SP-C precursor synthesis during lung development approximates that for SP-A and SP-B and therefore parallels overall lung development and functional status.

Structure and Biosynthesis The SP-C gene encodes a proprotein of 197 amino acids (Fig. 39-4); however, differential splicing of the primary mRNA transcript leads to several different SP-C mRNAs, at least one of which results in a reduction in the size of the proprotein to 191 amino acids. It is not known if the multiple SP-C mRNAs are all translated and whether this apparent redundancy is functionally significant. Unlike SP-A and SP-B, the SP-C proprotein is not glycosylated and does not contain an NH2-terminal signal peptide. The hydrophobic mature peptide (amino acids 24–58) mediates translocation of the NH 2-terminal propeptide into the lumen of the endoplasmic reticulum resulting in a transmembrane protein. As for the SP-B proprotein, the mature 35-amino acid SP-C peptide is generated by proteolytic cleavage of the NH 2 - and COOH-terminal propeptides, an event that likely occurs in the multivesicular body of type II cells. The identity of the protease(s) involved in propeptide cleavage and the fate and function of the liberated propeptides are not known. It is also not clear how the mature SPC peptide, which is presumably oriented within the interior of the vesicle membrane, is assembled with surfactant phospholipids for secretion into the airspace. Functional Domains Mature SP-C is an extraordinarily hydrophobic peptide consisting of 77% nonpolar amino acids. The most hydrophobic portion of the peptide is the 23-residue, valinerich C-terminal region that forms a rigid α-helix capable of spanning the membrane bilayer. The relatively hydrophilic NH2terminal portion of the peptide contains several positively charged residues that likely anchor SP-C to one side of the membrane by interacting with the negatively charged phospholipid head groups; in addition, the covalent attachment of palmitic acid to cysteines at positions 5 and 6 ensures that the NH2-terminus of SP-C is firmly anchored to the membrane. Although the function of SP-C is not entirely clear, the membrane-spanning helical region of the peptide is required for rapid adsorption of surfactant phospholipids to the air–liquid interface. The role of SP-C in the surfaceactive monolayer remains much less clear. Recent studies in vitro suggest that SP-C may play in important role in the clearance of alveolar surfactant lipids by the type II epithelial cell.

CHAPTER 39 / SURFACTANT DEFICIENCY

353

Figure 39-4 Structure of SP-C. SP-C is produced by proteolytic processing of a 197-amino acid precursor. The active SP-C peptide of 33–35 amino acids inserts deeply into lipid monolayers and bilayers, enhancing the surface activity of the phospholipids.

LIFE CYCLE OF PULMONARY SURFACTANT Pulmonary surfactant phospholipids and proteins are synthesized by type II epithelial cells lining the alveolar surface of the lung. Phospholipids are synthesized de novo and are re-used at high rates by the type II epithelial cell in a highly regulated, conservative manner. Phospholipids and proteins are synthesized and trafficked to membranous structures of the endoplasmic reticulum, Golgi, and multivesicular bodies, and ultimately packaged in lamellar bodies, the storage form of phospholipids and proteins in the type II cell. Surfactant, in the form of lamellar bodies, is secreted into the airway by exocytosis. The presence of calcium, SP-A, SP-B, and phospholipids produces tubular myelin, the most abundant form of alveolar surfactant. Monolayers or sheets of monolayers are likely formed by the actions of surfactant proteins B and C and phospholipids. The membranes spread over the alveolar surface, reducing surface tension at the air–liquid interface. The dynamic compression of surfactant during the respiratory cycle generates a variety of structural forms containing surfactant lipids and proteins, and it is likely that some of these forms determine whether the particles are catabolized or taken back up by type II cells for reutilization. Inactive particles, in the form of small, less dense vesicles, that are relatively depleted in surfactant proteins are likely catabolic forms that are taken up by the respiratory epithelium. Surfactant proteins B and A serve to maintain the large aggregates (tubular myelin) that are highly surface active. Differences in the abundance or type of surfactant proteins associated with the various lipid particles may influence the rates of uptake and recycling. Although surfactant proteins and lipids are taken up by the type II epithelial cell and recycled at high rates, a subfraction of surfactant lipids and proteins are taken up and degraded by

alveolar macrophages. Although only approximately 15% of phospholipids are catabolized by alveolar macrophages in vivo, this process may be a critical determinant of the steady state concentrations of surfactant, which are maintained at precise levels in the alveolar spaces of the adult animal. Thus, the life cycle of pulmonary surfactant represents a balance between synthesis, secretion, formation of distinct structural forms, reuptake, and catabolism. Each of these processes is influenced at various regulatory levels.

HEREDITARY SURFACTANT PROTEIN B DEFICIENCY The critical role of surfactant protein B (SP-B) in lung function was emphasized by the finding that full-term infants lacking SP-B develop lethal respiratory distress immediately after birth. Hereditary SP-B deficiency was first recognized in 1993 in a kinship of three full-term infants dying of respiratory failure at birth. Affected infants develop respiratory distress with grunting, retraction, cyanosis, pulmonary hypertension, and radiographic findings consistent with diffuse atelectasis after birth, similar to the findings in premature infants with IRDS. Initial chest radiograms show reticular-granular infiltrates that progress to severe atelectasis. Infants require intubation and intensive care immediately after birth. Therapy with oxygen and positive-pressure ventilation fails in spite of vigorous and continued respiratory management. Although the number of treated infants is not extensive to date, the infants do not have a consistent or prolonged response to exogenous surfactant replacement. Where available, SP-B-deficient infants are often treated with extracorporeal membrane oxygenation (ECMO) before the diagnosis is entertained. Hereditary SP-B deficiency has been consistently fatal and infants usually succumb in the first months of life in spite of continued intensive respiratory

354

SECTION V / PULMONOLOGY

support. Several infants with hereditary SP-B deficiency have undergone lung transplantation. DIAGNOSIS OF HEREDITARY SP-B DEFICIENCY The diagnosis of SP-B deficiency is made on the basis of family history. All infants diagnosed to date have two defective SP-B genes inherited as autosomal recessive genes. The infants develop respiratory distress in the first 24 h and are generally critically ill with increasing respiratory distress in the absence of prematurity, infection, or other underlying causes of respiratory failure. Tracheal aspirates from infants lack the active SP-B peptide and contain a diagnostic fragment of proSP-C (of approximately 12 kDa), which accumulates in the air spaces. Amniotic fluid from these infants contains the 12-kDa proSP-C fragment, a paucity of phospholipids, a decreased lecithin:sphinyomyelin (L:S) ratio, and the absence of the SP-B active peptide and phosphatidylglycerol (PG). Lung biopsy samples generally reveal diffuse lung disease with accumulation of proteinaceous material in the air spaces of affected infants. This material is distinct from that detected in adults with alveolar proteinosis, the former being devoid of the SP-B and enriched in the proSP-C fragment. Electron microscopic examination of lung tissue from SP-B-deficient infants demonstrated the lack of lamellar bodies and tubular myelin, as well as abnormal and injured epithelial cells likely related to prolonged intensive care and oxygen therapy. The definitive diagnosis is made by genetic analysis of the SP-B alleles from the infants and family members. MOLECULAR DIAGNOSIS OF CONGENITAL SP-B DEFICIENCY The most common genetic defect associated with SP-B deficiency results from a frameshift mutation in exon 4 (121 ins 2), which has been detected in approximately 75% of the affected infants analyzed to date. At present, 38 infants with this disease have been diagnosed since the recognition of the disorder in 1993. Missense and nonsense mutations have also been identified, with the infants generally presenting as compound heterozygotes in concert with the common exon 4 mutation. Definitive diagnosis is made by sequence analysis of the SP-B alleles. PATHOPHYSIOLOGY OF HEREDITARY SP-B DEFICIENCY IN SP-B GENE-TARGETED MICE The murine SP-B gene was disrupted by homologous recombination in embryonic stem cells used to make SP-B gene-targeted (knockout) mice. Like affected infants, SP-B-deficient mice succumb from respiratory distress postnatally, but are otherwise unaffected. Abnormalities in these mice are confined to the lungs, which are atelectatic but have developed normally. Type II cells of the SP-B(–/–) mice lack lamellar bodies and contain large atypical multivesicular bodies. Tubular myelin is lacking in the air spaces, and there is decreased content of the active SP-C peptide and a complete lack of both proSP-B and SP-B in the gene-targeted mice. Insertional mutagenesis in exon 4 of the mouse gene and the missense mutation in exon 4 of the human gene both result in an unstable messenger RNA and the lack of synthesis of proSP-B. Mutations in other regions of the human gene are associated with expression of abnormal forms of SP-B and proSP-B, which are detected intracellularly and extracellularly. ProSP-C, normally routed through the secretory pathway of the type II epithelial cell in concert with proSP-B, is aberrantly processed, resulting in the secretion and accumulation of proSP-C of Mr = 12 kDa in the air spaces of SP-B(–/–) in humans and mice. Thus, in addition to the lack of SP-B, SP-B(–/–) mice and patients are deficient in SP-C active peptide and accumulate high concentrations of proSP-C fragments in the air spaces that likely further impair surfactant function. The aberrant packaging,

storage, and organization of surfactant lipids in the SP-B deficiency also are likely to contribute to surfactant dysfunction, leading to respiratory failure in hereditary SP-B deficiency. HETEROZYGOUS SP-B DEFICIENCY SP-B deficiency is generally inherited as an autosomal recessive gene. Although clinical lung disease has not been associated with heterozygous family members of SP-B-deficient infants, these families have not been intensively studied to date. Heterozygous SP-B gene-targeted mice develop normally for more than a year in the laboratory without apparent abnormalities. However, SP-B mRNA, proSP-B, and the mature SP-B peptide are reduced by approximately 50% in SP-B(+/–) offspring. Careful analysis of lung function in the SP-Bdeficient heterozygous mice revealed air trapping and decreased lung compliance, consistent with abnormalities of small airway function. Whether SP-B(+/–) humans have abnormalities of lung function or are susceptible to lung dysfunction is unknown at present. The small airway dysfunction and air trapping seen in the SP-B(+/–) mice suggests that phenotypic alterations in SP-B deficiency might include asthma and other forms of small airway disease. DISTINCTION OF HEREDITARY SP-B DEFICIENCY FROM PULMONARY ALVEOLAR PROTEINOSIS (PAP) Although associated with the accumulation of surfactant lipids in the air space, the hereditary form of SP-B deficiency is distinct from infants or adults with pulmonary alveolar proteinosis in which both surfactant lipids and proteins accumulate in the alveolar space. Adult forms of alveolar proteinosis are generally idiopathic and are associated with malignancy, immune disorders, and exposure to inhaled particles and are not associated with the acute respiratory failure seen in SP-B-deficient infants. Alveolar proteinosis appears to be caused by disruption in surfactant homeostasis that leads to decreased clearance of surfactant by alveolar macrophages or by the respiratory epithelium. Genetic ablation of granulocyte-macrophage colony-stimulating factor (GM-CSF) or its receptor causes severe alveolar proteinosis in transgenic knockout mice. Unlike hereditary SP-B deficiency, surfactant from these animals or patients with PAP is highly surfaceactive, contains normally processed surfactant proteins, and lacks the proSP-C fragment, the latter being diagnostic for hereditary SP-B deficiency. The analysis of the GM-CSF of knockout mice has provided novel insight into the role of GM-CSF in lung homeostasis consistent with a critical role for GM-CSF in autocrineparacrine signaling within the lung. These findings support a role for GM-CSF in the modulation of surfactant homeostasis in the postnatal lung and suggest the potential role of GM-CSF and its receptor subunits in the pathogenesis of alveolar proteinosis in humans. SECONDARY CAUSES OF SP-B DEFICIENCY Hereditary SP-B deficiency must be distinguished from other causes of surfactant or surfactant protein B deficiency in newborn infants. Secondary causes of SP-B deficiency may be related to infection, prematurity, or to other as yet unknown factors. SP-B content was below the level of detectability in lung lavage fluid from some infants with severe respiratory failure and with no discernible abnormalities in the SP-B genes. SP-B deficiency has also been observed in the lungs of infants dying of congenital lung disease associated with abnormal lung morphogenesis, such as congenital acinar hypoplasia. Lungs from these infants are also deficient in SP-A and SP-C, and the pathogenesis of this disorder may be related to the lack of synthesis of regulatory factors that influence the expression of surfactant proteins in lung epithelial cell differentiation or on pulmonary organogenesis per se.

CHAPTER 39 / SURFACTANT DEFICIENCY

FUTURE DIRECTIONS FOR DIAGNOSIS AND THERAPY OF HEREDITARY SP-B DEFICIENCY Because of the severe disruption of surfactant homeostasis of hereditary SP-B deficiency and the severity of respiratory distress, these infants generally succumb within the first months of life despite extensive ventilatory support and intensive care. SP-B(–/–) infants have not responded to exogenous surfactant replacement even when initiated relatively early in the course of the disease. Genetic testing of the fetus can be used to identify infants with hereditary SP-B deficiency before birth. Both protein and DNA analyses have been used to identify siblings of previously affected kindred with SP-B deficiency. Such information can be useful in genetic counseling of these families. Several infants with SP-B deficiency have received lung transplants for correction of SP-B deficiency and two of these infants have now survived more than a year. Long-term outcome of neonatal lung transplantation is unknown, however, and this procedure should be considered experimental. SP-B mRNA can be transferred to respiratory epithelial cells for the correction of SP-B deficiency in the laboratory, and adenoviral vectors capable of efficiently transferring SP-B mRNA to type II epithelial cells have been used in animal models. However, correction of hereditary SP-B deficiency in infants will require relatively high levels of SP-B gene expression within type II epithelial cells. At present, the inefficiency of gene transfer vectors and immune responses for the use of such vectors provide formidable barriers to successful gene therapy of SP-B deficiency in the lung.

355

Nevertheless, hereditary lung diseases, such as SP-B deficiency, cystic fibrosis, and α1-antitrypsin deficiency, may someday be amenable to somatic cell gene transfer.

SELECTED REFERENCES Hohlfeld J, Fabel H, Hamm H. The role of pulmonary surfactant in obstructive airways disease. Eur Respir J 1997;10:482–491. Johansson J, Curstedt T. Molecular structures and interactions of pulmonary surfactant components. Eur J Biochem 1997;224:675–693. Khoor A, Stahlman MT, Gray ME, Whitsett JA. Temporal-spatial distribution of SP-B and SP-C proteins and mRNAs in development respiratory epithelium of human lung. J Histochem Cytochem 1994;42:1187–1199. Kuroki Y, Voelker DR. Pulmonary surfactant proteins. J Biol Chem 1994;269:25,943–25,946. Nogee LM. Surfactant protein-B deficiency. Chest 1997;111(Suppl): 129S–135S. Nogee LM, Garnier G, Dietz HC, et al. A mutation in the surfactant protein B gene responsible for fatal neonatal respiratory disease in multiple kindreds. J Clin Invest 1994;93:1860–1863. Robertson B, Taeusch HW. Surfactant therapy for lung disease. In: L’Enfant C, Lung Biology in Health and Disease. New York: Marcel Dekker, 1995. Rooney SA, Young SL, Mendelson CR. Molecular and cellular processing of lung surfactant. FASEB J 1994;8:957–967. van Golde LMJ. Potential role of surfactant proteins A and D in innate lung defense against pathogens. Biol Neonate 1995;67:2–17. Whitsett JA, Nogee LM, Weaver TE, Horowitz AD. Human surfactant protein B: structure, function, regulation, and genetic disease. Physiol Rev 1995;75:749–757.

CHAPTER 40 / LUNG CANCER

40

357

Lung Cancer The Role of Tumor Suppressor Genes STEVEN JAY WEINTRAUB

BACKGROUND The leading cause of cancer deaths in the United States is carcinoma of the lung. Despite the innumerable advances in the treatment of other types of cancer, lung carcinomas remain nearly uniformly fatal. In the past few years, however, there have been many new findings regarding the series of events that occur on a molecular level during lung carcinogenesis. These findings should prove to be of value not only in understanding the evolution of lung cancer, but also in its diagnosis and treatment. The different types of lung cancer have been divided into two categories. Small cell lung cancer (SCLC) is thought to be of neuroendocrine origin. Chemotherapy extends survival of patients with this type of lung cancer by several months. Squamous cell carcinoma, adenocarcinoma, and large cell carcinoma are all thought to arise from a bronchoalveolar progenitor cell. This group of carcinomas has been designated non–small cell lung cancer (NSCLC). Current chemotherapeutic regimens have no effect on survival of patients that are afflicted with NSCLC. Early cytogenetic studies of both SCLC and NSCLC demonstrated that several chromosomal abnormalities occurred consistently in lung cancer tissue and cell lines derived from these tumors, suggesting that a specific set of mutations occurs in each type of lung cancer. Indeed, it is now known that several genes are frequently expressed aberrantly in lung cancer. Of these, the retinoblastoma gene and the p53 gene, both of which are tumor suppressors, are the best studied. Therefore, the function of the retinoblastoma protein (Rb) and p53 protein (p53) in normal cell growth and differentiation and the mechanism by which alteration of their activity promotes lung carcinogenesis will be the focus of this chapter.

THE RETINOBLASTOMA PROTEIN–E2F PATHWAY Studies of the neoplastic process initially led to the description of oncogenes, which are genes whose expression induces cellular transformation. However, a consistent finding among human retinoblastomas was the deletion of chromosomal band 13q14, suggesting the existence of a gene whose inactivation is necessary for progression to retinoblastoma (see Chapters 7 and 105). Thus the existence of antioncogenes or tumor suppressor genes was postulated. Accordingly, a gene, now known as the retinoblastoma gene From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

and found at this locus, is expressed in all normal retinal cells, as well as all normal adult tissue, but is either absent or mutated in all retinoblastomas. Confirmation that Rb is a tumor suppressor was provided by experiments in which wild-type Rb was artificially introduced into retinoblastoma cells. Restoration of Rb in these cells resulted in growth inhibition and loss of tumorigenicity, indicating that inactivation of Rb is necessary for tumorigenesis. In addition to retinoblastomas, loss of Rb function has been associated with osteosarcomas and bladder, breast, prostate, cervical, and lung cancers. Among patients with bladder cancer, those with Rb(–) carcinomas have more invasive tumors and their prognosis is significantly worse than those with Rb(+) bladder cancers. In cancer of the cervix, disruption of Rb function must be necessary for tumor progression because almost all cervical cancers fall into one of two categories: either they express the human papilloma virus protein E7, a protein that is known to bind to and inactivate Rb, or they produce only a mutant, inactive form of Rb. Similarly, loss of functional Rb appears to be important in lung cancer since it is mutated in over 90% of SCLCs and 20–30% of NSCLCs. Underscoring the role of disruption of Rb function in the pathogenesis of lung cancer is the finding that relatives of retinoblastoma patients who have germline mutations in the Rb gene are at markedly increased risk of developing SCLC. Rb has two domains that are strongly conserved across species and among two other cellular proteins, p107 and p130, and the region that encompasses these domains has become known as the pocket (Fig. 40-1). Accordingly Rb, p107, and p130 are collectively known as the pocket proteins. The integrity of the pocket is critical for the interaction of these proteins with several other cellular proteins, including the transcription factor E2F (see below). An intriguing early finding was that most mutations of Rb in tumors that express a stable protein involve the pocket, suggesting that a function of the pocket has an essential role in the tumor suppressing activity of Rb (i.e., disruption of an Rb pocket function is important in tumor pathogenesis). Additionally, several DNA tumor virus oncoproteins bind to the Rb pocket and displace cellular pocket-binding proteins. Any mutation in these oncoproteins that interferes with their capacity to bind Rb and displace these proteins disrupts their oncogenicity, suggesting that they are oncogenic because they block the interaction of the Rb pocket with other cellular proteins. It is now known that Rb plays a role in regulation of progression through the normal cell cycle (see Chapter 6). Rb does so at least

357

358

SECTION V / PULMONOLOGY

Figure 40-1 Schematic of the pocket proteins. The pockets of Rb, p107, and p130 each consist of two regions, domain A and domain B, that contain amino acid sequences that are conserved both among these three proteins and across species and a nonconserved spacer between these two conserved regions. Most of the cellular functions of the pocket proteins that have been identified to date are dependent on the integrity of the pocket domain. The white areas represent the regions that are conserved between Rb, p107, and p130 and among each of these proteins across different species.

in part through the interaction of the pocket with a family of transcription factors known collectively as E2F. Binding sites for E2F are found in the promoters of several genes whose expression is necessary for progression through the cell cycle, including thymidylate synthase, dihydrofolate reductase, and DNA polymerase-α. Rb is phosphorylated in a cell cycle-dependent fashion. It is in its “hypophosphorylated” state at the end of mitosis and remains so during most of G0/G1, but as the cell enters late G1 it becomes “hyperphosphorylated.” The phosphorylation cycle controls the interaction of Rb with E2F: when it is in its hypophosphorylated form, it binds to E2F; however, when it is hyperphosphorylated in late G1, it dissociates from E2F leaving free E2F behind on the promoter (Fig. 40-2). Only the hypophosphorylated form of Rb, the form that binds E2F, inhibits growth. The complex of hypophosphorylated Rb and E2F is a transcriptional repressor that inhibits the activity of trans-activating factors on the promoter, and the “free” E2F that is left on the promoter upon hyperphosphorylation and dissociation of Rb from E2F is a transcriptional activator. The complex of the Rb-related proteins p107 and E2F are also transcriptional repressors. p107and p130 also bind to E2F in a cell cycle-dependent fashion, but their pattern of binding differs from that of Rb. So it is envisioned that cell cycle progression is regulated by the Rb–E2F pathway as follows: several genes whose expression is necessary for cell cycle progression contain E2F sites and, at specific stages of the cell cycle, free E2F activates these genes; however, when Rb or one of the other pocket proteins bind to E2F, the complex of the pocket protein and E2F inhibits the activity of these same genes. In this fashion Rb and related proteins control the expression of these genes in a cell cycle-specific fashion. If Rb is inactivated through mutation, some of these genes will be constitutively expressed, removing an important constraint on cellular proliferation. It is likely that this is a critical step in the pathogenesis of lung cancer as well as other tumors. The mechanism by which the Rb–E2F complex represses transcription has recently been elucidated. It was found that the Rb pocket has intrinsic transcriptional repressor activity. This was demonstrated through the use of a chimeric protein in which the pocket was fused to a heterologous DNA binding domain, that of the yeast transcription factor Gal4. The activity of the resultant protein was assessed using an artificial transcriptional reporter construct that contains the Gal4 DNA binding sequence (Fig. 40-3).

Figure 40-2 Rb is phosphorylated in a cell cycle-dependent fashion and its interaction with E2F is controlled by its phosphorylation state. Rb becomes hyperphosphorylated in late G1 and remains so until late in mitosis; during anaphase Rb becomes hypophosphorylated. Only the hypophosphorylated form binds E2F. E2F is a family of transcription factors that activate genes whose expression is necessary for progression through the cell cycle. When E2F is complexed with Rb (or one of the other pocket proteins), however, the complex is a transcriptional repressor that inhibits expression of these same genes. Rb is thought to regulate cellular proliferation through its interaction with E2F and the resultant transcriptional repression of these genes.

This facilitated the targeting of Rb to a promoter in an E2F-independent fashion. When the activity of a Gal4-Rb-pocket fusion protein was assessed in this fashion, it was found to be as effective a transcriptional repressor as wild-type Rb when it is tethered to a promoter by E2F. When control proteins or other regions of Rb were fused to Gal4, the resulting chimeric proteins were inactive in this assay. Thus, it is likely that the repressor activity of the Rb–E2F complex is a function of the Rb pocket, and the role of E2F in this complex is to target the repressor activity of Rb to the appropriate promoters. Further, it was shown that this repressor activity is a function of the ability of Rb to disrupt critical interactions of several proteins that activate transcription. Even though Rb has been shown to interact with several cellular proteins, there is evidence, at least in one cell line, that E2F is the most important mediator of Rb activity. In these cells, overexpression of a form of E2F that lacks Rb, p107, and p130 binding activity, but still encodes a trans-activation function will disrupt the growth-inhibitory effect of Rb. When it is overexpressed, this protein will bind to E2F sites and displace cellular proteins that are bound to these sites, such as the Rb–E2F complex, and, like free E2F, it will trans-activate promoters that contain E2F sites. Unlike E2F, because it does not bind pocket proteins, it will not mediate transcriptional repression by Rb (or any other effect of the pocket proteins) and it will not affect the interaction of the pocket proteins with other proteins in the cell. The net result of overexpression of this protein then is disruption of the effect of Rb and Rb-related proteins on promoters that contain E2F sites with-

CHAPTER 40 / LUNG CANCER

359

SCLC cells express wild-type Rb, it is likely to be functionally inactive because of the lack of p16. These findings suggest that inactivation of Rb, whether by mutation or constitutive hyperphosphorylation, is a necessary step in the pathogenesis of lung cancer; in fact, it has recently been postulated that the Rb–E2F pathway may be dysregulated in all human malignancies.

THE p53 CHECKPOINT FOR DNA DAMAGE

Figure 40-3 The transcriptional repressor activity of Rb is a function of its pocket domain. Chimeric proteins in which the different regions of Rb are fused to the DNA-binding domain of the yeast transcription factor Gal4 are targeted to promoters that contain a Gal4binding site. This permits an assessment of the activity of the different domains of Rb. When the Rb pocket domain is targeted to a promoter in this fashion it represses the activity of enhancers on the promoter, thereby blocking expression of the gene. The only region of Rb that has repressor activity in this assay is the pocket, suggesting that the repressor activity of Rb is a function of its pocket domain.

out a direct effect on any other function of the Rb family of proteins. The finding that this protein disrupts the growth inhibitory effect of Rb indicates that the interaction of Rb with Rb binding proteins other than E2F is not sufficient for growth suppression. Indeed, it has been suggested that E2F may be the only critical effector of Rb activity. Although Rb is not inactivated by mutation or deletion in all tumors, it may be functionally inactive in several tumor types that express wild-type Rb. As outlined above, the activity of Rb is controlled by its phosphorylation state, hypophosphorylated Rb being the active growth-inhibitory form and hyperphosphorylated Rb the inactive form. It is becoming apparent that the most important regulators of Rb phosphorylation are the D cyclins and their associated cyclin-dependent kinases (CDKs). The D cyclins activate certain CDKs, which in turn phosphorylate Rb. It has been found that in some NSCLCs that express wild-type Rb, cyclin D1 is overexpressed. This could lead to constitutive hyperphosphorylation and, therefore, inactivation of the Rb. So even though these cells express wild-type Rb, they are likely to be functionally Rb negative. Additionally, a newly discovered class of proteins has been found that inhibit the activity of cyclin-CDKs. The most relevant one to this discussion appears to be p16. This protein binds to cyclin D–CDK complexes and inhibits phosphorylation of Rb, maintaining Rb in its active growth-suppressive state. Disruption of p16 function would result in inappropriate hyperphosphorylation and, therefore, inactivation of Rb. Accordingly, p16 has been found to be mutated in several different types of cancer and it has been shown that replacement of wild-type p16 into p16(–) cancer cell lines results in inhibition of their growth and suppression of tumorigenicity. It is therefore of note that in a study of 55 SCLC cell lines it was found that whereas the 48 that lacked wild-type Rb expressed p16, 6 of the cell lines expressed wild-type Rb but lacked p16. In the second set of cells, lack of the p16 inhibitor will lead to increased kinase activity of cyclin D–CDK complexes, which in turn should result in increased phosphorylation and functional inactivation of Rb. So although these

p53 plays a critical role in controlling cellular proliferation by either arresting cells in G1 or by inducing apoptosis under the appropriate conditions (see Chapter 7). p53 functions in a pathway that serves as a checkpoint for DNA damage in that various agents that are known to cause mutagenesis of DNA, such as ionizing radiation, will induce either p53-dependent cell cycle arrest or apoptosis. Both of these effects are known to be dependent on p53 activity because mutation of p53 disrupts them and reintroduction of p53 into p53(–) cells restores them. Whether a cell undergoes p53-dependent arrest or apoptosis appears to be cell type-specific. It is likely that the cell cycle arrest that occurs with DNA damage allows the cell to pause and repair its DNA whereas apoptosis serves to limit proliferation of cells with genomic damage. Evidence for these hypotheses is provided by the finding that cells that lack p53 are one million times more likely to contain abnormally amplified DNA than cells that express p53. Additionally, mice that have homozygous p53 mutations develop normally through birth but invariably develop tumors by 6–9 months. This is probably because of the fact that cells that lack the p53-dependent checkpoint for DNA damage continue to proliferate while accumulating mutations—if the appropriate set of mutations occurs in these cells they will undergo malignant transformation. The p53 gene is the most frequently mutated gene among cancers of all types. In regard to lung cancer, p53 is mutated in about 90% of SCLCs and 60% of NSCLCs. Additionally, although wildtype p53 is expressed in some lung cancers, it may be inactivated by mechanisms other than mutation. In a recent study, it was found that several NSCLCs that express wild-type p53 overexpress a second protein, mdm2, that binds to and inactivates p53. Overexpression of mdm2 was found to occur only in cells that expressed wild-type p53. These findings suggest that in certain lung tumors the p53 pathway is inactivated by overexpression of mdm2 instead of mutation of p53. This is analogous to the situation with Rb outlined above. Even though Rb is not inactivated by mutation in all SCLCs, it may be functionally inactivated by inappropriate phosphorylation if cyclin D1 is overexpressed or if CDK inhibitor p16 activity is absent. After exposure of cells to ionizing radiation there is an increase in cellular p53 levels that is coincident with p53-dependent cell cycle arrest. Like Rb, p53 is a transcription factor, except that p53 is a trans-activating protein. Underscoring the role of its transactivating function in cell cycle arrest and, therefore tumor suppression, is the finding that most of the mutant forms of p53 found in tumors have lost this activity. This suggested that wild-type p53 inhibits cellular proliferation by activating the expression of genes that block progression through the cell cycle (Fig. 40-4). It is now known that in nontransformed cells the increase in the cellular level of p53 that occurs with ionizing radiation increases expression of at least two proteins that arrest cellular proliferation, p21 and GADD45. p21, like p16, inhibits the kinase activity of cyclin– CDK complexes. Increased cellular levels of p21 will block phosphorylation of Rb by inhibiting the activity of the Rb-specific cyclin–CDK complexes, thereby maintaining Rb (and possibly

360

SECTION V / PULMONOLOGY

SELECTED REFERENCES

Figure 40-4 The ability of p53 to stimulate transcription is essential for tumor suppression. The critical role that the transcriptional activation function of p53 has in tumor suppression was first suggested by the finding that transforming mutants of p53 do not activate gene expression.

Figure 40-5 Schematic of the mechanism by which p53 mediates ionizing radiation-induced cell cycle arrest. The promoters of the genes for p21 and GADD45 contain binding sites for p53. Ionizing radiation increases p53 in the cell resulting in activation of transcription of the p21 and GADD45 genes. These proteins arrest cell growth.

Rb-related proteins) in its hypophosphorylated growth-suppressive form—resulting in a G1 cell cycle arrest. GADD45, the other protein whose expression is induced by p53, is thought to arrest cellular proliferation by inhibiting DNA synthesis, probably through an interaction with proliferating cell nuclear antigen (PCNA) (Fig. 40-5).

CONCLUSIONS A series of mutations resulting in aberrant production of a specific set of proteins culminates in cellular transformation. It is estimated that 10–20 mutations occur during the pathogenesis of lung cancer. Several genes have been identified that are consistently mutated in lung cancers, but the best studied of these are the Rb and p53 genes. Both of these genes encode tumor suppressors. An understanding of the role of these and other tumor suppressors in maintenance of normal cellular physiology and the significance of their inactivation in cancer pathogenesis has advanced significantly during the past decade. The retinoblastoma protein is important for maintaining a controlled progression through the cell cycle by regulating the activity of several growth-promoting genes. p53 limits the transmission of potentially harmful genetic alterations that occur with DNA damage to future generations of cells either by arresting the cell cycle to allow for DNA repair or by inducing apoptosis. Therefore, the inactivation of Rb or p53 disrupts constraints that inhibit the transformation of normal lung tissue to lung carcinoma. The understanding of these processes should allow for more efficient treatment of lung cancer in the future, possibly by the restoration of their function in malignant cells.

Albelda SM. Gene therapy for lung cancer and mesothelioma. Chest 1997;111(Suppl):144S–149S. Antelman D, Machemer T, Huyghe BG, Shepard HM, Maneval D, Johnson DE. Inhibition of tumor cell proliferation in vitro and in vivo by exogenous p110RB, the retinoblastoma tumor suppressor protein. Oncogene 1995;10:697–704. Benedict WF. Altered RB expression is a prognostic clinical marker involved in human bladder tumorigenesis. J Cell Biochem 1992; 161(Suppl):69S–71S. Bookstein R, Lee WH. Molecular genetics of the retinoblastoma suppressor gene. Crit Rev Oncog 1991;2:211–227. Buchkovich K, Duffy LA, Harlow E. The retinoblastoma protein is phosphorylated during specific phases of the cell cycle. Cell 1989;58:1097–1105. Cai DW, Mukhopadhyay T, Liu Y, Fujiwara T, Roth JA. Stable expression of the wild-type p53 gene in human lung cancer cells after retrovirus-mediated gene transfer. Hum Gene Ther 1993;4:617–624. Carbone DP, Minna JD. In vivo gene therapy of human lung cancer using wild-type p53 delivered by retrovirus. J Natl Cancer Inst 1994; 86:1437,1438. Chen PL, Scully P, Shew JY, Wang JY, Lee WH. Phosphorylation of the retinoblastoma gene product is modulated during the cell cycle and cellular differentiation. Cell 1989;58:1193–1198. Cordon-Cardo C, Wartinger D, Petrylak D, et al. Altered expression of the retinoblastoma gene product: prognostic indicator in bladder cancer. J Natl Cancer Inst 1992;84:1251–1256. Donehower LA, Harvey M, Slagle BL, et al. Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumours. Nature 1992;356:215–221. Dulic V, Kaufmann WK, Wilson SJ, et al. p53-dependent inhibition of cyclin-dependent kinase activities in human fibroblasts during radiation-induced G1 arrest. Cell 1994;76:1013–1023. el-Deiry WS, Tokino T, Velculescu VE, et al. WAF1, a potential mediator of p53 tumor suppression. Cell 1993;75:817–825. Ewen ME. The cell cycle and the retinoblastoma protein family. Cancer Metastasis Rev 1994;13:45–66. Fields S, Jang SK. Presence of a potent transcription activating sequence in the p53 protein. Science 1990;249:1046–1049. Fujiwara T, Grimm EA, Mukhopadhyay T, Cai DW, Owen-Schaub LB, Roth JA. A retroviral wild-type p53 expression vector penetrates human lung cancer spheroids and inhibits growth by inducing apoptosis. Cancer Res 1993;53:4129–4133. Harbour JW, Lai SL, Whang-Peng J, Gazdar AF, Minna JD, Kaye FJ. Abnormalities in structure and expression of the human retinoblastoma gene in SCLC. Science 1988;241:353–357. Harper JW, Adami GR, Wei N, Keyomarsi K, Elledge SJ. The p21 Cdkinteracting protein Cip1 is a potent inhibitor of G1 cyclin-dependent kinases. Cell 1993;75:805–816. Hu QJ, Dyson N, Harlow E. The regions of the retinoblastoma protein needed for binding to adenovirus E1A or SV40 large T antigen are common sites for mutations. EMBO J 1990;9:1147–1155. Huang HJ, Yee JK, Shew JY, et al. Suppression of the neoplastic phenotype by replacement of the RB gene in human cancer cells. Science 1988;242:1563–1566. Kaelin WJ, Pallas DC, DeCaprio JA, Kaye FJ, Livingston DM. Identification of cellular proteins that can interact specifically with the T/E1A-binding region of the retinoblastoma gene product. Cell 1991;64:521–532. Kastan MB, Canman CE, Leonard CJ. p53, cell cycle control and apoptosis: implications for cancer. Cancer Metastasis Rev 1995;14:3–15. Kastan MB, Onyekwere O, Sidransky D, Vogelstein B, Craig RW. Participation of p53 protein in the cellular response to DNA damage. Cancer Res 1991;51:6304–6311. Kastan MB, Zhan Q, el-Deiry WS, et al. A mammalian cell cycle checkpoint pathway utilizing p53 and GADD45 is defective in ataxiatelangiectasia. Cell 1992;71:587–597. Koh J, Enders GH, Dynlacht BD, Harlow E. Tumour-derived p16 alleles encoding proteins defective in cell-cycle inhibition. Nature 1995;375: 506–510.

CHAPTER 40 / LUNG CANCER

Kuerbitz SJ, Plunkett BS, Walsh WV, Kastan MB. Wild-type p53 is a cell cycle checkpoint determinant following irradiation. Proc Natl Acad Sci USA 1992;89:7491–7495. Livingstone LR, White A, Sprouse J, Livanos E, Jacks T, Tlsty TD. Altered cell cycle arrest and gene amplification potential accompany loss of wild-type p53. Cell 1992;70:923–935. Marchetti A, Buttitta F, Pellegrini S, et al. Mdm2 gene amplification and overexpression in non–small cell lung carcinomas with accumulation of the p53 protein in the absence of p53 gene mutations. Diagn Mol Pathol 1995;4:93–97. Minna JD. The molecular biology of lung cancer pathogenesis. Chest 1993;103:449S–456S. Momand J, Zambetti GP, Olson DC, George D, Levine AJ. The mdm-2 oncogene product forms a complex with the p53 protein and inhibits p53-mediated transactivation. Cell 1992;69:1237–1245. Nevins JR. E2F: a link between the Rb tumor suppressor protein and viral oncoproteins. Science 1992;258:424–429. Nevins JR. Transcriptional regulation. A closer look at E2F. Nature 1992;358:375,376. Otterson GA, Kratzke RA, Coxon A, Kim YW, Kaye FJ. Absence of p16INK4 protein is restricted to the subset of lung cancer lines that retains wildtype RB. Oncogene 1994;9:3375–3378. Qian Y, Luckey C, Horton L, Esser M, Templeton DJ. Biological function of the retinoblastoma protein requires distinct domains for hyperphosphorylation and transcription factor binding. Mol Cell Biol 1992;12:5363–5372. Qin XQ, Livingston DM, Ewen M, Sellers WR, Arany Z, Kaelin WG. The transcription factor E2F-1 is a downstream target of RB action. Mol Cell Biol 1995;15:742–755. Raycroft L, Wu HY, Lozano G. Transcriptional activation by wild-type but not transforming mutants of the p53 anti-oncogene. Science 1990;249:1049–1051. Roth JA, Nguyen D, Lawrence DD, et al. Retrovirus-mediated wild-type p53 gene transfers to tumors of patients with lung cancer. Nat Med 1996;2:985–991. Sanders BM, Jay M, Draper GJ, Roberts EM. Non-ocular cancer in relatives of retinoblastoma patients. Br J Cancer 1989;60:358–365. Schauer IE, Siriwardana S, Langan TA, Sclafani RA. Cyclin D1 overexpression vs. retinoblastoma inactivation: implications for growth control evasion in non–small cell and small cell lung cancer. Proc Natl Acad Sci USA 1994;91:7827–7831. Scheffner M, Munger K, Byrne JC, Howley PM. The state of the p53 and retinoblastoma genes in human cervical carcinoma cell lines. Proc Natl Acad Sci USA 1991;88:5523–5527.

361

Serrano M, Hannon GJ, Beach D. A new regulatory motif in cell-cycle control causing specific inhibition of cyclin D/CDK4. Nature 1993;366:704–707. Shirodkar S, Ewen M, DeCaprio JA, Morgan J, Livingston DM, Chittenden T. The transcription factor E2F interacts with the retinoblastoma product and a p107-cyclin A complex in a cell cycle-regulated manner. Cell 1992;68:157–166. Slebos RJ, Lee MH, Plunkett BS, et al. p53-dependent G1 arrest involves pRB-related proteins and is disrupted by the human papillomavirus 16 E7 oncoprotein. Proc Natl Acad Sci USA 1994;91:5320–5324. Smith ML, Chen IT, Zhan Q, et al. Interaction of the p53-regulated protein Gadd45 with proliferating cell nuclear antigen. Science 1994;266: 1376–1380. Swafford DS, Middleton SK, Palmisano WA, et al. Frequent aberrant methylation of p16INK4a in primary rat lung tumors. Mol Cell Biol 1997;17:1366–1374. Weinberg RA. The retinoblastoma protein and cell cycle control. Cell 1995;81:323–330. Weinberg RA. Tumor suppressor genes. Science 1991;254: 1138–1146. Weintraub SJ, Chow K, Luo RX, Zhang SH, He S, Dean DC. Mechanism of active transcriptional repression by the retinoblastoma protein. Nature 1995;375:812–815. Weintraub SJ, Prater CA, Dean DC. Retinoblastoma protein switches the E2F site from positive to negative element. Nature 1992;358: 259–261. Whang PJ, Knutsen T, Gazdar A, et al. Nonrandom structural and numerical chromosome changes in non–small-cell lung cancer. Genes Chromosom Cancer 1991;3:168–188. Whyte P, Williamson NM, Harlow E. Cellular targets for transformation by the adenovirus E1A proteins. Cell 1989;56:67–75. Wiest JS, Franklin WA, Otstot JT, et al. Identification of a novel region of homozygous deletion on chromosome 9p in squamous cell carcinoma of the lung: the location of a putative tumor suppressor gene. Cancer Res 1997;57:1–6. Yin Y, Tainsky MA, Bischoff FZ, Strong LC, Wahl GM. Wild-type p53 restores cell cycle control and inhibits gene amplification in cells with mutant p53 alleles. Cell 1992;70:937–948. Yonish-Rouach E, Resnitzky D, Lotem J, Sachs L, Kimchi A, Oren M. Wild-type p53 induces apoptosis of myeloid leukemia cells that is inhibited by interleukin-6. Nature 1991;352:345–347.

363

CHAPTER 41 / HEPATOLOGY

GASTROENTEROLOGY VI SECTION EDITOR:

JAMES C. REYNOLDS

364

SECTION VI / GASTROENTEROLOGY

CHAPTER 41 / HEPATOLOGY

365

41 Hepatology PIET C. DE GROEN AND NICHOLAS F. LARUSSO INTRODUCTION During the past century, hepatology has gradually progressed from a discipline focused on simply describing the signs and symptoms associated with a specific liver disease, via clinical observations, biochemical abnormalities, and organ pathology, to a science that involves describing and understanding abnormal cell function. Rapid progress in the techniques of molecular biology in the last three decades has added an entirely new dimension to the concept of disease; it has led to the development of molecular hepatology as an emerging area of interest that attempts to generate new knowledge of the molecular mechanisms responsible for normal and abnormal liver function. Recently, several genes responsible for inherited liver diseases have been either localized to a specific part of a chromosome or cloned and characterized. Some of these genes have been found by applying traditional molecular techniques (e.g., the gene was cloned using antibodies to a normal or abnormal protein or cDNA probes were used to screen CDNA libraries [α-1 antitrypsin deficiency]). Other genes have been found using linkage analysis and chromosome jumping or walking (e.g., cystic fibrosis, Wilson’s disease, hemochromatosis) or cytogenetic deletions and rearrangements (e.g., Alagille syndrome). Relevant bacterial and viral DNA/RNA compositions have also been identified using a variety of techniques, such as virus purification (Hepatitis B) or cDNA cloning (Hepatitis C and G). At present, we are beginning to understand why certain genes are expressed only in liver. Hepatocyte-specific promoters have been described and the search for specific transcription factors is ongoing. In vitro as well as in vivo transfection techniques are being developed, and the first human trials of gene therapy in both inherited as well as acquired (e.g., cancer) liver diseases have been started. At least five viruses are now known to preferentially infect the liver, and new therapeutic modalities aimed at selectively interfering with viral metabolism are being developed. Because of limited space and the fact that substantial components of what might be covered in a chapter on molecular hepatology will be covered in other chapters, our approach will be a very focused one. In this chapter, we briefly describe liver development, bile formation (including the enterohepatic cycle of bile acids), and the molecular mechanisms leading to cirrhosis. Next, From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

we discuss five inherited (i.e., Crigler-Najjar, Gilbert’s, DubinJohnson, Alagille, and the progressive familial intrahepatic cholestasis [PFIC] syndromes) and two important immune-mediated (i.e., primary biliary cirrhosis [PBC] and primary sclerosing cholangitis [PSC]) cholestatic diseases; finally, we review selected developments in autoimmune hepatitis, polycystic liver disease, amyloidosis, and alcohol- and drug-induced (i.e., acetaminophen, aspirin, and NSAIDS) liver disease. The classic inherited disorders of the liver are reviewed in Chapter 42 by Ruiz and Wu, who address the molecular basis of Wilson’s disease hemochromatosis, α-1-antitrypsin deficiency, cystic fibrosis, the porphyrias, galactosemias, and fructose deficiency. In Chapter 43, Gupta and Shafritz explain in detail the molecular aspects of hepatitis caused by viruses, with particular attention to hepatitis B, C, and D.

MOLECULAR ASPECTS OF LIVER FUNCTION AND BILE FORMATION During embryogenesis, the liver develops as an extension of cells from the hepatic diverticulum, the caudal portion developing into the gallbladder and cystic duct, and the cranial portion forming the common bile duct and liver bud. The endogenous liver cell population is composed of hepatocytes and intrahepatic bile duct epithelial cells, or cholangiocytes, which are of epithelial cell origin, and a variety of mesenchymal cells, including Kupffer cells, vascular and sinusoidal endothelial cells, fat-storing cells, and pit cells. Within the liver, these cells are organized into histologic and functional units forming the classic hepatic lobules. The hepatocyte has three important, unique features: first, it is the sole producer of a large number of serum proteins, carbohydrates, and lipids; second, it metabolizes endogenous and foreign substances and excretes the modified products into the biliary system. Third, it synthesizes primary bile acids. All features require hepatocyte-specific promoters and transcription factors for hepatocyte-specific expression of secretary proteins, receptors, or enzymes. Several hepatocyte-specific promoters are now known and include, among others, the albumin, transthyretin, and phosphoenol pyruvate carboxykinase promoters. These specific promoters have been identified by a standard molecular strategy. First, the coding region of the genes was cloned (cDNA) and, subsequently, their genomic structure identified. Next, the upstream untranslated genomic sequence was analyzed for DNA sequences (i.e., cis-acting elements) able to drive expression of reporter genes only in hepatocyte-derived cell lines, such as HepG2

365

366

SECTION VI / GASTROENTEROLOGY

Figure 41-1 Albumin promoter. Transcription factors binding to regulatory elements in the albumin gene are depicted in this schematic, simplified diagram. Parts of the gene (double-stranded DNA) representing the enhancer and promoter regions are shown as a double horizontal line; the distance between the enhancer and the promoter is approximately 10-kb pairs. The arrow to the right of the TATAA box indicates the position and orientation of the transcription start of the albumin mRNA. Transcription factor binding sites are shown as ovals or circles, with the names of several transcription factors that bind to these sites shown above and with the names of representative tissues where these factors have been found shown below. The high degree of tissue specificity of this gene is conferred by the combination of the enhancer as well as promoter regulatory elements. NF-1, nuclear factor 1 or CCAAT transcription factor (CTP); LAP, liver-enriched transcriptional activator protein [also named nuclear factor interleukin-6 (NF-IL6), C/EBPβ, and cysteine-rich protein 2 (CRP2)]; NF-Y, nuclear factor Y. (Adapted from: Zaret KS. Control of hepatocyte differentiation by liver-enriched transcription factors. In: Tavoloni N, Berk PD, eds. Hepatic Transport and Bile Secretion: Physiology and Pathophysiology. New York: Raven, 1993; pp. 135–143.

cells. Finally, the unique DNA sequences able to drive expression were analyzed for binding sites for proteins or transcription factors able to drive hepatocyte-specific transcription. Examples of these factors include CCAAT/enhancer-binding protein (C/EBP) and hepatocyte nuclear factors 1, 3, and 4 (HNF-1,3,4) (Fig. 41-1). Other well-known, unique hepatocyte proteins include the asialoglycoprotein receptor, bile acid formation regulating enzymes, and several transport proteins in the hepatocyte canalicular membrane. Bile acids are organic acids derived from cholesterol. As natural ionic detergents, they are of vital importance in absorption, transport, and secretion of lipids. The primary bile acids in humans are cholic and chenodeoxycholic acid. Both are excreted predominantly in conjugated form, linked either to glycine or taurine. Dehydroxylation of these bile acids by intestinal bacteria produces the more hydrophobic, secondary bile acids, deoxycholic acid and lithocholic acid. In other mammalian species, alternative hydroxylations can occur. For instance, bears generate ursodeoxycholic acid, the 7β-epimer of chenodeoxycholic acid. After release into the intestine, the majority of bile acids is absorbed and returned to the liver via the portal venous system. This cycling of bile acids between liver and intestine is called the enterohepatic circulation. A small fraction of secondary bile acids is further metabolized in the liver to tertiary, conjugated forms, which are not reabsorbed after excretion. Most enzymes required for cholesterol and bile acid synthesis have been extensively studied, and metabolic defects for several enzymes have been described. For example, deficiency in 26-hydroxylase results in a syndrome of premature arteriosclerosis, cataracts, and dementia, known as cerebrotendinous xanthomatosis. A defect in the final step of the bile acid synthesis pathway, that is, side-chain cleavage in peroxisomes, causes the cerebrohepatorenal syndrome of Zellweger, characterized by severe hypotonia, profound psychomotor retardation, and a characteristic facial appearance. Recently, proteins required for bile acid conjugation and transport have been cloned, either by searching for proteins able to bind to labeled bile acids or by expression cloning for proteins able to transport bile acids. For

example, an hepatic sodium/bile acid cotransporter, an ileal bile acid binding protein, an enzyme capable of conjugating bile acids with both glycine and taurine, and the protein responsible for most, if not all, of the sulfation of bile acids in human liver have been cloned. Secretion of bile acids is coupled to secretion of phosphatidylcholine and cholesterol and thereby provides a major pathway for cholesterol excretion in bile. Deficiency of bile acids results in reduced cholesterol solubility in bile and is thought to contribute to formation of gallstones. On the other hand, expansion of the bile acid pool with ursodeoxycholic acid causes a reduction of biliary cholesterol excretion and has been found to be of therapeutic value in patients with cholesterol gallstones. The exact molecular mechanisms leading to cholesterol gallstone formation are unknown; however, bile with a high cholesterol saturation index strongly predisposes to cholesterol stone formation. Epidemiological studies have shown a familial risk of gallstones and suggest that probably the main cause of this risk is genetic. This concept is supported by the recent description of the lith1 gene associated with development of gallstones in mice. Until recently, the bile duct epithelial cell, or cholangiocyte, was considered a rather passive cell, that formed a three-dimensional network of interconnecting ducts that served as simple conduits for delivering bile to the intestine. New data show that cholangiocytes actively participate in secretory and absorptive processes. For instance, these cells express the cystic fibrosis transmembrane conductance regulator (CFTR), a regulatory chloride channel, and the chloride/bicarbonate exchanger at their apical domain; the water channel aquaporin 1, or CHIP28, likely at both apical and basolateral domains, and a variety of ion exchangers and ATPdriven pumps at the basolateral domain. Together with cytosolic carbonic anhydrase, these proteins form the basis of a secretory system that is activated, or inhibited by, changes in cytosolic cAMP concentrations. This whole system is finely regulated by peptide hormones. For example, secretin binds to the secretin receptor at the basolateral domain, activates adenyl cyclase, and increases cAMP concentrations, thereby inducing secretion of bicarbonate

CHAPTER 41 / HEPATOLOGY

367

Figure 41-2 The uridine diphosphate (UDP) glucuronosyltransferase 1 gene (UGT1) and bilirubin UDP-glucuronosyltransferase 1 mRNA. The more than 100-kb UGT1 gene is located on chromosome 2. Ten mRNAs are derived from the UGT1 gene, each encoding a different UDP-glucuronosyltransferase. The mRNAs differ only in their most 5' sequence. Each of the 5' exons (exon 1A to 1J) encodes the aminoterminus of one UGT1 isoform. Bilirubin UDP-glucuronosyltransferase 1 mRNA uses exon 1A as is shown. Exon 1A encodes for the substrate binding site, exons 3 and 4 for the uridine diphosphate glucuronic acid binding site and exon 5 for the membrane spanning region and endoplasmatic reticulum retention signal. A (TA) insertion in the TATAA box of the promoter preceding exon 1A has been associated with Gilbert’s syndrome; point and splice site mutations, deletions, insertions, and formation of stop codons with Crigler Najjar type I syndrome; and only point mutations with Crigler Najjar type II syndrome. Distances are not to scale. (Adapted from: Janssen PLM, Bosma P. Inherited unconjugated hyperbilirubinemias. AASLD postgraduate course 1995 syllabus, pp. 87–100.)

and water; in contrast, binding of somatostatin to its receptor has the opposite effect. Kupffer cells are macrophages lining the sinusoidal spaces between rows of hepatocytes. They have a variety of functions, including: 1. 2. 3. 4.

Endocytosis of viruses, bacteria, and endotoxins. Antigen processing. Secretion of prostaglandins, interferons, and cytokines. A direct cytotoxicity for parasites and tumor cells.

Pit cells, a population of large granular lymphocytes with natural killer cell activity, are thought to protect the liver from colonization by metastatic tumor cells and viruses. In addition, they may function in the regulation of hepatocyte proliferation. Endothelial cells lining the hepatic sinusoids act as sieves, allowing only free diffusion of particles smaller than 0.2 µm, such as chylomicron remnants, through their fenestrations into the space of Disse. They also are actively engaged in endocytosis and excrete prostaglandins and cytokines in response to various stimuli. Fat-storing cells, lipocytes, Ito cells, or stellate cells are mesenchymal cells situated in the space of Disse, between endothelial cells and hepatocytes. Under normal conditions, they store large quantities of vitamin A in the form of membrane bound droplets. Recent evidence suggests that fat-storing cells are able to transform into myofibroblast-like cells in response to hepatic injury. The factors involved in activation and transformation of these cells include TNF, TGF-β, PDGF, and, possibly, acetaldehyde. Of these, TGF-β probably is the most important. Secreted by fatstoring cells as well as other nonparenchymal cells in response to injury, it activates fat-storing cells and induces production of collagen and other extracellular matrix proteins. At the same time, synthesis of matrix-degrading proteolytic proteins is inhibited, thereby further disrupting the balance between collagen formation and degradation.

In addition to induction of collagen synthesis, a change in the relative amounts of collagen types occurs. Quiescent cells produce mainly collagen types III and IV, laminin and to a lesser extent collagen type I and fibronectin. Activated myofibroblastlike cells secrete collagens type I, III and IV, fibronectin, laminin and a variety of proteoglycans at a much higher level. In particular, collagen type I, fibronectin and proteoglycans containing chondroitin and dermatan sulfate are expressed at very high levels. Together, these changes result in marked hepatic fibrosis. With continued exposure to agents inducing the abnormal extracellular matrix formation, irreversible fibrosis develops. When this fibrotic process is associated with destruction of the classic hepatic lobule, a histology characteristic of cirrhosis develops: regenerating nodules of hepatocytes separated by dense strands of fibrous tissue.

CHOLESTATIC LIVER DISEASES: INHERITED AND IMMUNE-MEDIATED A single enzyme, UDP-glucuronosyltransferase 1, is responsible for most of the bilirubin glucuronidating activity in humans. Absence of UDP-glucuronosyltransferase 1 results in a syndrome known as Crigler-Najjar type I, whereas mutations causing severe deficiency of the enzyme result in Crigler-Najjar type II (Fig. 41-2). Both syndromes are inherited in a recessive pattern. Crigler-Najjar type I patients have severe unconjugated hyperbilirubinemia with plasma bilirubin levels fluctuating between 26 and 45 mg/dL. As a result, they develop severe neurologic deficits (kernicterus) and frequently die in their first year of life. Crigler-Najjar type II patients have less severe unconjugated hyperbilirubinemia (usually less than 20 mg/dL) and usually do not develop neurologic symptoms. In contrast to type I, the bile of these patients does contain bilirubin conjugates, and treatment with phenobarbital or glutethimide increases bilirubin metabolism resulting in markedly reduced plasma bilirubin concentrations.

368

SECTION VI / GASTROENTEROLOGY

Gilbert’s syndrome is characterized by a mild, chronic unconjugated hyperbilirubinemia in the absence of liver disease or overt hemolysis. The mode of inheritance is not clear, as both autosomal dominant as well as recessive patterns of inheritance have been described. The defect is a decreased hepatic glucuronidating activity to approximately 30% of normal, resulting in an increased relative concentration of bilirubin monoglucuronide in bile. The enzyme UDP-glucuronosyltransferase 1 itself is normal (i.e., normal sequence of coding region) in many patients thus far examined, but the 5' promoter region of the gene contains two extra bases (TA) in the TATAA element [A(TA7)TAA rather than normal A(TA6)TAA] (Fig. 41-2). The presence of the longer TATAA element results in the reduced expression of a reporter gene, encoding firefly luciferase, in a human hepatoma cell line and therefore appears to be a cause for reduced expression of bilirubin UDP-glucuronosyltransferase 1 in vivo. However, this explanation is not sufficient to explain the complete manifestation of the syndrome. Dubin-Johnson syndrome is a rare autosomal recessive disorder characterized by chronic conjugated hyperbilirubinemia. Patients have impaired hepatobiliary transport of nonbile salt organic anions. Recently the disease was shown to be caused by a mutation in the canalicular multispecific organic anion transporter, or multidrug-resistance-associated protein 2 (MRP2), on chromosome 10q24. This protein mediates adenosine triphosphate-dependent transport of a broad range of endogenous and xenobiotic compounds across the apical canalicular membrane of the hepatocyte. Alagille syndrome (also referred to as Alagille-Watson syndrome, syndromic bile duct paucity, or arteriohepatic dysplasia) is an autosomal dominant disorder characterized by abnormal development of liver, heart, skeleton, eye, face, and, less frequently, kidney. Major manifestations of this syndrome are neonatal jaundice and cholestasis in older children; histopathology shows paucity of intrahepatic bile ducts and cholestasis. Analyses of patients with cytogenetic deletions or rearrangements led to the identification of the JAG1 gene on chromosome 20 as the gene mutated or absent in this syndrome. JAG1 encodes a ligand for the Notch transmembrane receptor. This receptor, called Notch for the shape of the malformed wings of fruit flies with only one functional copy, plays a critical role during morphogenesis. The features of Alagille syndrome seem to develop due to the absence of adequate amounts of JAG1 protein during morphogenesis (so-called haploinsufficiency) rather then due to pertubations caused by a mutant protein. Progressive familial intrahepatic cholestasis (PFIC) is a heterogeneous group of autosomal recessive liver disorders, characterized by early onset of cholestasis that usually progresses to cirrhosis and liver failure before adulthood. Currently, PFIC is divided in three subcategories: PFIC type 1, or Byler disease, is thought to be caused by defective bile salt secretion. Two loci for Byler disease have been found and mapped to chromosome 18q21q22 and 2q24. The locus on chromosome 18 also maps the gene for benign recurrent intrahepatic cholestasis (BRIC), and it is possible that different mutations in a single gene lead to this variant of Byler disease and BRIC. A second type of PFIC is thought to be caused by defects in bile salt synthesis; multiple genetic loci may be responsible for this phenotype. Patients with this second type and patients with Byler disease have normal serum γ-glutamyltransferase levels. A third type of PFIC can be distinguished from

Figure 41-3 Histopathology of primary biliary cirrhosis. This liver specimen, obtained at the time of liver transplantation, shows end stage PBC. A regenerative nodule surrounded by inflammatory tissue and blood vessels can be detected; however, bile ducts and ductules are completely absent. (original magnification ×31.)

the other two types by high serum γ-glutamyltransferase activity and liver histology showing portal inflammation and ductular proliferation at an early stage. The genetic defect is due to mutations in the multidrug resistance P-glycoprotein MDR3. This protein is responsible for biliary phospholipid excretion at he canalicular membrane of the hepatocyte as was shown by knockout of the mouse homolog MDR2. As patients with MDR3 mutations, the MDR2 knockout mice develop severe liver disease, characterized by inflammation of the portal tracts, proliferation of the bile ducts, and fibrosis. Primary Biliary Cirrhosis (PBC) is a liver disease occurring primarily in women and characterized by cholestasis as a result of destruction of septal and interlobular bile ducts (Fig. 41-3) resulting in jaundice and pruritus, hyperlipidemia, fat and fat-soluble vitamin malabsorption, and metabolic bone disease. The liver histology progresses over years to decades from lymphocytic portal infiltrates with bile duct destruction and ductular hyperplasia, to end-stage biliary cirrhosis with complete absence of septal and interlobular bile ducts. At this stage, the usual complications of cirrhosis develop: portal hypertension, variceal hemorrhages, ascites, spontaneous bacterial peritonitis, hepatic encephalopathy, and hepatorenal syndrome. The molecular defect or abnormality responsible for development of PBC is not known, but an autoimmune basis seems most likely. This presumption is based on the presence of autoantibodies, a possible association with HLA-antigens, known association with other autoimmune diseases, a lymphocytic infiltrate in the early histopathology, and a partial response to a variety of immunefunction modifying drugs. The major autoantibodies seen in PBC are directed against the E2 subunit of the mitochondrial pyruvate dehydrogenase enzyme complex and the E2 subunit of the branched chain 2-oxo-acid dehydrogenase complex. It is unclear whether these antibodies are simply a result of the disease process or whether they are actually involved in bile duct destruction. There is an increased frequency of HLA class II antigen DR8 in patients with PBC; how this antigen increases susceptibility to PBC remains unexplained. Exogenous factors also have been implicated; for instance, antibodies as well as T-cell clones derived from patients with PBC crossreact with exogenous antigens such as an E. coli pyruvate decarboxylate enzyme complex-E2 peptide.

CHAPTER 41 / HEPATOLOGY

This suggests that molecular mimicry (i.e., the recognition by the immune system of endogenous proteins with structural similarity to microbial-derived proteins) at the T cell clonal level may be a possible cause of PBC. Totally effective medical treatment for PBC does not exist. A large number of immune-function modifying and anti-inflammatory drugs have been tried in randomized, controlled, and prospective studies. These drugs included prednisone, azathioprine, chlorambucil, D-penicillamine, cyclosporine, methotrexate, colchicine, and malotilate; none have substantially improved survival, symptoms, or biochemical parameters. However, use of ursodeoxycholic acid has been shown to improve liver tests as well as symptoms, and a recent meta-analysis suggests increased survival or time to liver transplantation; the mechanisms of this apparently beneficial effect are unknown. Finally, liver transplantation clearly has been shown to improve survival, with 1- and 5-year survivals reported as greater than 90 and 80%, respectively. Unlike PBC, primary sclerosing cholangitis (PSC) is seen predominantly in men. Destruction of bile ducts also forms the basis of this disease, but the inflammatory process is much less pronounced when compared to PBC. Extensive fibrosis of the portal tracts, especially involving the bile ducts (periductal fibrosis), is a histologic hallmark. Unlike PBC, all bile ducts can be affected; that is, the disease can extend from the common bile duct (extrahepatic) to the smallest septal bile ducts (intrahepatic). Involvement of the larger ducts results in a very characteristic radiologic appearance: When the bile ducts are filled with contrast, multiple strictures and localized areas of ectasia are seen (Fig. 41-4). Over years to decades, PSC progresses and end-stage disease is associated with similar complications as discussed for PBC. However, three additional complications are frequently seen. First, stasis of bile in the biliary tree as a result of strictures can lead to intraductal stone formation as well as to development of recurrent infection of the biliary tree (i.e., bacterial cholangitis). Second, in approximately 10% of patients, transformation of bile duct epithelial cells into bile duct cancer occurs. Third, chronic ulcerative colitis, a disease associated with PSC in approximately 70% of cases (see below), predisposes to the development of colon cancer. PSC, like PBC, probably is an autoimmune disease. PSC is also associated with other autoimmune diseases, particularly with chronic ulcerative colitis. As mentioned above, this association is seen in approximately 70% of patients and supports the presumed autoimmune basis of the disease. However, antibodies specific for PSC have not been identified, and an obvious, strong association with a particular HLA class I or II phenotype is not seen. Weak associations have been reported for HLA class II DR3, DR8, and DR52 antigens. All medical regimens tried thus far have been unsuccessful. At present, only liver transplantation offers patients with PSC significant improvement in survival as well as symptoms.

AUTOIMMUNE HEPATITIS Autoimmune hepatitis can be defined as an idiopathic autoimmune disease in which the hepatocyte is the predominant target of immune-mediated destruction. It can occur at any age but tends to follow a bimodal age distribution (10–30 and above 50 years), most frequently in women. Symptoms and signs are nonspecific and include general malaise, lethargy, and fatigue. The disease is characterized by negative viral serologies, hypergammaglobulinemia, and a number of circulating autoantibodies. Liver histol-

369

Figure 41-4 Cholangiogram of a patient with primary sclerosing cholangitis. This cholangiogram shows the typical features of primary sclerosing cholangitis: multiple strictures and dilatations throughout the intra- and extrahepatic bile ducts. Three metal clips were placed during cholecystectomy. (Source: LaRusso NF, Wiesner RH, Ludwig J, MacCarty RL. Primary sclerosing cholangitis. N Engl J Med 1984;310:899–903.)

ogy shows active hepatitis with mononuclear infiltration of portal tracts and surrounding liver lobule (i.e., piecemeal necrosis). Frequently, plasma cells and eosinophils are seen within the infiltrate, in addition to a mixture of helper and suppressor lymphocytes. In general, the disease progresses to cirrhosis. However, treatment with corticosteroids, alone or in combination with azathioprine, markedly delays disease progression and ameliorates symptoms. As for PBC and PSC, the pathobiology of autoimmune hepatitis is poorly understood. There is a strong linkage with the MHC class II antigens DR3 and DR4 on chromosome 6. In addition, autoimmune hepatitis is also associated with other autoimmune diseases, such as thyroiditis and Sjogren’s syndrome. Based on the type of autoantibodies present, several subgroups can be identified. In type I autoimmune hepatitis, autoantibodies are directed against non–organ- and non–species-specific antigens (e.g., antinuclear and antismooth muscle antibodies). These antibodies are unlikely to be involved in the pathogenesis. Type II autoimmune hepatitis is characterized by antibodies against specific cytochrome P450 isoenzymes (anti–liver-kidney microsomal antibodies; putative target cytochrome P450 2D6), which may be expressed on the cell membrane of hepatocytes. Immunogenic cytochrome P450 also can be induced by drug metabolism and haptenation, which indicates that environmental or medicinal xenobiotics may initiate autoimmune liver disease. Finally, a type

370

SECTION VI / GASTROENTEROLOGY

III autoimmune hepatitis has been proposed based on the presence of antisoluble liver antigen antibodies. Treatment goals are to decrease inflammation and prevent progression to cirrhosis. In general, immunosuppressant therapy is only used in symptomatic patients. As mentioned, the therapeutic efficacy of corticosteroids, with or without the addition of azathioprine, is well established. Survival with therapy is excellent. Once cirrhosis and its complications are present, only liver transplantation results in restoration of an acceptable quality of life.

POLYCYSTIC LIVER DISEASE Simple hepatic cysts are detected in 2–5% of patients undergoing ultrasound examinations of the abdomen. Cysts are seen more frequently in older patients and women. In 60–75%, cysts are solitary and, in general, a maximum of three cysts are detected in patients with multiple lesions. Polycystic liver disease, therefore, should be considered in patients who are found to have four or more hepatic cysts. The most common cause of polycystic liver disease is autosomal dominant polycystic kidney disease (ADPKD). At least three genetic loci are known to predispose to the disease and are located, respectively, on chromosome 16 (ADPKD1), chromosome 4 (ADPKD2), and a still unmapped location (ADPKD3) (see Chapter 70). In 1994, the gene from ADPKD1 was cloned and, based on DNA sequence analysis, it codes for a large transmembrane protein called polycystin-1. The function of this protein remains to be determined; a role in cell–cell or cell–matrix interactions is hypothesized. Preliminary evidence suggests that this protein is expressed during fetal development. In normal adults, it is present at very low levels in kidney and biliary epithelial cells; however, its expression is significantly increased in the epithelia lining both kidney and liver cysts of patients with ADPKD1, suggesting that failure to downregulate polycystin expression may be the molecular defect leading to polycystic disease. Recently, the ADPKD2 gene also was cloned; it too seems to encode an integral membrane protein, named polycystin-2, which is ~50% homologous to polycystin-1 at the amino acid level. Polycystin-2 also shows homology to the family of voltage-activated calcium channels. In addition to ADPKD, development of renal as well as hepatic cysts can be a result of other genetic defects, such as autosomal recessive polycystic kidney disease, obstruction of ureter or common bile duct, or a variety of drugs. Finally, multiple families have been described with polycystic liver disease in the absence of significant renal cystic disease; whether these families have a genotype different from all the above-mentioned is unknown. Liver cysts can range in size from microscopic to several inches in diameter and, therefore, the liver can be normal to massively enlarged (Fig. 41-5). Cysts usually are thin-walled, composed of a single layer of cuboidal epithelial cells, filled with fluid resembling the bile salt independent fraction of bile (i.e., approximate values for sodium, potassium, chloride, and bicarbonate of, respectively, 145, 4, 115, and 25 mEq/L). The cells lining cysts stain positively for biliary epithelial cell markers (i.e., cytokeratin 7 and 19) and secrete ions and water into the cyst in response to secretin, strongly supporting a biliary epithelial origin. Liver cysts are rare in children with ADPKD, but the frequency increases with age to approximately 70% in the seventh decade. Women are likely to have more and larger cysts at an earlier age than men, perhaps related to the use of estrogens and pregnancies. Most patients with polycystic liver disease do not have symptoms. If symptoms occur,

Figure 41-5 CT Scan of a patient with polycystic liver disease. Multiple, large hepatic cysts are present within the liver of this patient with ADPKD.

these reflect abdominal distension due to liver size, venous or biliary obstruction, hemorrhage into cysts, and torsion, rupture, or infection of cysts. Because the volume of the hepatic parenchyma remains normal, overall hepatic function is preserved. The primary goal of treatment is to reduce the volume of the cysts; this can be established by avoidance of factors associated with cyst development or expansion, such as estrogens and possibly alcohol. Cyst aspiration followed by installation of a sclerosing agent (e.g., 95% alcohol or tetracycline) is used for patients with symptoms a result of one or a few dominant cysts. Finally, surgical fenestration or fenestration combined with cyst resection is performed in patients with advanced disease, whereas liver transplantation is used in the rare patient without liver segments free of cysts.

AMYLOIDOSIS Amyloidosis is caused by deposition of amyloid substance in the extracellular matrix of various tissues. Frequently the liver is involved, although symptoms and signs because of liver involvement, predominantly hepatomegaly, are nonspecific and relatively infrequent. The amyloid substance is composed of fibrils, which are polymers of low molecular weight proteins in β-pleated sheet conformation. Systemic amyloidoses can be classified into four types as listed in Table 41-1. It is beyond the scope of this chapter to discuss all types in detail; instead, we will discuss the molecular basis of one of the best studied types of hereditary amyloidoses: familial amyloid polyneuropathy caused by point mutations in the transthyretin gene. Transthyretin is the product of a single copy gene on chromosome 18. This gene has four exons that code for a 127 amino acid mature protein and an 18 amino acid signal peptide. The mature protein functions as a plasma transport protein for both thyroxine and retinol-binding protein. Most of the protein is synthesized in the liver; the remainder is synthesized in the choroid plexuses of the brain and eye. The structure of transthyretin is dominated by eight β strands, which contain approximately 50% of the amino acids and form two four-stranded β sheets. This extensive β-sheet

CHAPTER 41 / HEPATOLOGY

371

Table 41-1 Systemic Amyloidosesa Type Immunoglobulin (AL)/primary Reactive (AA)/secondary Hereditary/familial amyloid polyneuropathies β2-microglobulin/dialysis

Protein Kappa or lambda light chains Serum amyloid A Transthyretin (prealbumin) Apolipoprotein A-1 Gelsolin β2-microglobulin

aSource: Reilly MM, King RH. Familial amyloid polyneuropathy. Brain Pathol 1993;3:165–176.

structure makes the transthyretin protein “amyloidogenic.” All 30 (approximately) point mutations predisposing to development of neuropathy are located in exons 2, 3, and 4; the most frequent one is a valine to methionine point mutation at amino acid 30. At the genome level, the frequent occurrence of this mutation can be explained by a “CpG hot spot” (i.e., the dinucleotide sequence CpG is more frequently the target sequence of mutations than would be expected on a random basis). Two additional point mutations in exon 2, one mutation in exon 3, and at least three mutations in exon 4 can lead to either hyperthyroxinemia or cardiomyopathy. Nearly all patients with a transthyretin mutation are heterozygotes; thus, both the normal as well as the mutated protein can be detected in blood. Because the liver is responsible for over 90% of transthyretin production, liver transplantation theoretically should correct the disease. Preliminary results after liver transplantation, performed in patients with familial amyloidosis polyneuropathy as a result of a point mutation converting the valine codon of amino acid 30 into a methionine codon, indeed show dramatic reductions in the abnormal transthyretin serum concentration and, in some cases, associated improvement in autonomic nervous system function.

ALCOHOL-INDUCED LIVER DISEASE Alcoholic liver disease occurs only in a subset of individuals who consume large amounts of alcohol (10–20%); the precise reason for this is not understood, but a difference in alcoholmetabolizing enzymes caused by genetic differences is suspected to be responsible for the individual variation in the severity of liver disease among heavy drinkers. Alcohol is predominantly metabolized in the liver. Two enzyme systems are of key importance; the first one consists of alcohol dehydrogenase (ADH) and aldehyde dehydrogenase (ALDH) (Fig. 41-6). ADH converts alcohol into acetaidehyde, which in turn is converted to acetate by ALDH. The second enzyme system is the cytochrome P450 2El or microsomal ethanol oxidizing system (MEOS). Acetaldehyde and increased production of reducing equivalents, such as NADH, are thought to be responsible for most of the hepatotoxicity of alcohol. At least seven different genomic loci are known to encode for over 20 different isoenzymes of ADH. All are expressed in liver and are thought to be derived from a common ancestral gene. They are grouped into five main classes based on the genetic and protein structure. Most are also expressed in other tissues such as the stomach, kidney, and lung. In women, the expression of ADH in the stomach is significantly lower than in men, and this, in part, may explain why women develop alcoholic hepatitis and subsequently progress to cirrhosis sooner, faster, and

Figure 41-6 Metabolic pathway of alcohol. Two enzymes, alcohol dehydrogenase (ADH) and Aldehyde dehydrogenase (ALDH), are responsible for most of the alcohol metabolizing capacity within the liver. Both enzymes require the presence of NAD+ to convert alcohol via the toxic intermediate product acetaldehyde into the nontoxic product acetate. The cytochrome P450 2El or microsomal ethanol oxidizing system (MEOS) pathway, in general, is not important with infrequent, low-dose alcohol use; however, it is markedly induced during chronic alcohol abuse.

after lower amounts of ethanol consumed than men. Two ALDHs are responsible for more than 90% of acetaldehyde metabolism: ALDH1 and ALDH2. Both are NAD+-dependent; ALDH1 is cytosolic whereas ALDH2 is mitochondrial. Many people with an Oriental background develop flushing after moderate alcohol use. This phenomenon is related to a decreased ALDH2 activity that results in relatively high acetaldehyde levels that, in turn, cause the flushing. This side effect of alcohol use may in part explain why alcohol abuse is less common in eastern Asia. Disulfiram is an inhibitor of ALDH; it produces rapid accumulation of acetaldehyde at levels 5–10 times higher than those found during metabolism of the same amount of alcohol alone. The unpleasant effects (flushing, headache, nausea, and vomiting) of high acetaldehyde concentrations caused by disulfiram, even after small amounts of alcohol, are used therapeutically in the treatment of alcoholism; it reminds motivated patients to abstain from alcohol. Acetaldehyde is known to condense with plasma proteins, forming stable adducts that can be recognized as foreign by the immune system. A prevalence of severe hypersensitivity reactions of 0.54% was found among a large population of non-Oriental individuals. The reactions were severe enough to deter these individuals from consuming all types of alcoholic beverages. Individuals presenting such reactions had significantly elevated levels of circulating anti-acetaldehyde-protein IgE antibodies. The exact mechanism whereby alcohol induces hepatitis and subsequently cirrhosis is not known. Cytokines and growth factors likely play a major role; interleukin-1, TNF-α, PDGF, EGF, and basic FGF all are known to stimulate Ito cell proliferation or enhance fibrogenesis. Other interleukins, in particular IL-6 and IL-8, also have been implicated in the pathogenesis of alcoholic liver disease. However, more research is needed to explain how alcohol induces production of these growth factors and cytokines.

DRUG-INDUCED LIVER DISEASES Nearly every drug can cause liver injury, because the liver is the central organ in most drug metabolic pathways. Some drugs cause a predictable, dose-dependent injury, whereas use of others only results in hepatotoxicity in patients with a genetic predisposition.

372

SECTION VI / GASTROENTEROLOGY

Figure 41-7 Metabolic pathway of acetaminophen. Under physiologic conditions acetaminophen is predominantly metabolized via phase 2 reactions (glucuronidation and sulfation). If the capacity of phase 2 reactions is exceeded, or if the cytochrome P450 complex is induced, increasingly phase 1 biotransformation occurs. In the presence of adequate amounts of reduced glutathione (GSH) the intermediate metabolite N-acetyl-p-benzoquinoneimine is converted into mercapturic acid; in the absence, covalent binding to cell proteins occurs. UDP, uridine diphosphate. (Adapted from: Lee WM. Drug-induced hepatotoxicity. N Engl J Med 1995;333:1118–1127.)

Two general mechanisms of injury are seen: first, direct disruption of the intracellular function or membrane integrity by the drug or its metabolites; and second, indirect damage from immunemediated mechanisms. Most drugs are hydrophobic compounds that require biochemical transformations for excretion, processes that transform the hydrophobic native drug into hydrophilic metabolites that are filtered by the glomerulus or excreted into bile. Biotransformation can be divided into two categories: phase 1 and phase 2 reactions. Phase 1 reactions include oxidation or demethylation and are mediated by the cytochrome P450 enzyme system. For instance, the result of a phase 1 reaction is generation of hydroxyl groups. In phase 2 reactions a large, water-soluble group is attached by glucuronidation or sulfation via ether or ester linkage to the phase 1 intermediate, such as an hydroxylated metabolite, or to the native drug. Enzymes involved in these reactions include the already discussed UDP-glucuronosyltransferases and several species of sulfotransferases. A third metabolic pathway for detoxifying many drugs involves a reduction reaction where an electrophilic compound is inactivated by glutathione-S-transferase–mediated glutathione coupling. Some drugs undergo all the processes (e.g., acetaminophen [Fig. 41-7]), others only one or two. The most common reaction leading to hepatotoxicity is the formation of covalent bonds between a metabolite of the native drug and cell proteins or DNA. Another mechanism likely to lead to cell death is lipid peroxidation. As already mentioned, the toxic effect of a specific drug can directly be related to the dose used (e.g., acetaminophen) or to a genetic variant in the enzymes required for its metabolism (e.g., autosomal recessive deficiency for cytochrome P450 2D6, which is required for metabolism of drugs such as debrisoquin, propranolol, and quinidine). How covalent binding of substrate or lipid peroxidation causes cell death remains unknown.

A commonly used classification of drug reactions is based on three end points: the histologic appearance, the cell type involved, and the clinical picture. Table 41-2 lists examples of toxic reactions and the causative agents. Detailed information regarding individual drugs is readily available and will not be discussed. A few widely used drugs and recently reported drug reactions deserve attention, however. Acetaminophen has already been mentioned. Its toxicity is enhanced by drugs, inducing the P450 system, such as barbiturates and alcohol, and by depletion of glutathione as seen in starvation and alcohol abuse (Fig. 41-8A). On the other hand, both cimetidine and N-acetylcysteine protect against acetaminophen toxicity, the former by inhibition of the P450 2E1 isoenzyme responsible for the conversion of acetaminophen into N-acetyl-p-benzoquinoneimine, and the latter by replenishing glutathione stores. Reye’s syndrome is defined as acute, noninflammatory encephalopathy in association with liver disease, without a clear explanation for both (Fig. 41-8B). Frequently, it follows a viral illness. Epidemiologic studies have shown a markedly increased risk for development of Reye’s syndrome after exposure to aspirin. The molecular mechanisms of aspirin-induced liver injury are poorly understood, but seem to involve inhibition of normal mitochondrial enzyme functions. In patients with hepatitis B, treatment with the nucleoside analog, fialuridine, induced a severe toxic reaction, characterized by hepatic failure, lactic acidosis, pancreatitis, neuropathy, and myopathy. The clinical picture of hepatic failure and the hepatic histology resembled the fulminant hepatic steatosis that occurs with Reye’s syndrome. Indeed, this toxic reaction also was a result of widespread mitochondrial damage. Similar toxic reactions may infrequently occur during treatment with other nucleoside analogs. Finally, recent evidence suggests that bacterial toxins, such as the emetic toxin of Bacillus cereus, also can inhibit hepatic mitochondrial fatty acid oxidation and cause fulminant hepatic failure.

CHAPTER 41 / HEPATOLOGY

373

Table 41-2 Types of Toxic Reactions and Examples of Provoking Agentsa Type of reaction

Examples of agents

Direct reaction

Acetaminophen, carbontetrachloride, mushrooms, phosphorus Isoniazid, disulfiram, propylthiouracil Halothane, isoflurane, ticrynafen Phenytoin, amoxicillin-clavulonic acid, sulfonamides Chlorpromazine, erythromycin estolate, estradiol, captopril, sulfonamides Diltiazem, quinidine, phenytoin, procainamide Nitrofurantoin, methyldopa, isoniazid, trazodone Amiodarone, perhexiline maleate, valproic acid Tetracyclines, aspirin, zidovudine, didanosine, fialuridine Methotrexate, vitamin A, methyldopa Cyclophosphamide, other chemotherapeutic agents, herbal teas Cocaine, sustained-release nicotinic acid, methylenedioxyamphetamine

Idiosyncratic reaction Toxic-allergic reaction Allergic hepatitis Cholestatic reaction Granulomatous reaction Chronic hepatitis Alcoholic hepatitis-like syndrome Microvesicular steatosis Fibrosis or cirrhosis alone Veno-occlusive disease Ischemic damage aSource:

WM Lee. Drug-induced hepatotoxicity. N Engl J Med 1995;333:1118–1127.

Nonsteroidal anti-inflammatory drugs (NSAIDS), frequently used anti-inflammatory, and analgesic agents are commonly implicated as causes of liver injury. One of the better studied drugs is diclofenac (Fig. 41-8C). In rare cases, it is associated with the development of fulminant hepatic necrosis. Both direct toxic effects of a diclofenac metabolite and hypersensitivity reactions have been suggested as possible molecular mechanisms of liver injury. In vitro experiments have shown that direct hepatocyte toxicity is increased with inhibition of the glucuronidation pathway and decreased by inhibition of cytochrome P450-dependent oxidative biotransformation. On the other hand, the covalent binding to hepatocellular proteins is greatly reduced when glucuronide formation was selectively blocked. Therefore, diclofenac acyl glucuronide formation is likely associated with covalent binding of a reactive metabolite to hepatocellular proteins, which may be toxicologically relevant for the expression of diclofenac hepatitis.

SUMMARY We now understand a growing number of the molecular mechanisms responsible for the signs and symptoms, the biochemical abnormalities, and the histopathology of many inherited as well as acquired liver diseases. At present, we are beginning to understand the regulation of gene transcription and translation within the various individual cell types of the liver. Among possible advances, this has resulted in the development of new classes of pharmacological agents and in the possibility of hepatocyte gene transfer. Although much has been learned, even more remains unknown: What are the molecular signals that determine overall liver size and induce the unique phenomenon of liver regeneration to its original size after drug- or virus-induced necrosis or surgical partial hepatectomy? How do different cell types within the liver communicate with each other? Why are genes transfected into

Figure 41-8 Histopathology of drug-induced liver disease. Liver histopathology of acetaminophen-, aspirin-, and diclofenac-hepatotoxicity. (A) Shows liver specimen with severe necrosis around the central veins (CV, zone 3 of the hepatic lobule) as the result of an acetaminophen overdose. The hepatic parenchyma around a large portal tract (PT, zone 1 of the hepatic lobule), showing a portal vein branch, an hepatic artery, and a bile duct, is well-preserved. (B) Shows the liver parenchyma of a patient with Reye’s syndrome after aspirin use. Within the hepatocytes there are abundant fat droplets; the fat has been dissolved and therefore the cells appear white in this illustration. (C) The liver specimen is from a patient who had received diclofenac for arthritis. While on the drug, she developed jaundice and therefore diclofenac was discontinued with resolution of symptoms. However, several months later, diclofenac was restarted with recurrence of jaundice. Only a few hepatocytes are seen in the part of the specimen shown in this picture (arrows). The close proximity of several bile ducts and the extensive inflammatory reaction suggest massive destruction of hepatocytes with collapse of the fibrous septa: submassive necrosis. The marked ductular proliferation is evidence of hepatic regeneration. (original magnifications: [A] ×31, [B] ×125, [C] ×62.

374

SECTION VI / GASTROENTEROLOGY

hepatocytes only transiently expressed in vivo? Clearly, the answers to many of these important questions will, in part, depend on further molecular studies of the diverse cell types forming the liver.

SELECTED REFERENCES HEPATOLOGY IN GENERAL Oxford Textbook of Clinical Hepatology. McIntyre N, Benhamou J-P, Bircher J, Rizzefto M, Rodes J, eds. Oxford: Oxford Medical Publ., 1992. Sherlock S, Dooley J. Diseases of the Liver and Biliary System. 9th ed. London: Blackwell Scientific Publ., 1993. Tung BY, Kowdley KV. Inherited liver diseases affecting the adult. Gastroenterologist 1996;4:245–261. Zakim D, Boyer TD. Hepatology, a Textbook of Liver Disease. 2nd ed. Philadelphia, PA: WB Saunders Company, 1990.

LIVER SPECIFIC GENE EXPRESSION Lai E, Darnell JE, Jr. Transcriptional control in hepatocytes: a window on development. TIBS 1991;16:427–430.

FIBROSIS AND CIRRHOSIS Friedman SL. The cellular basis of hepatic fibrosis. Mechanisms and treatment strategies. N Engl J Med 1993;328:1828–1835. Gressner AM. Activation of proteoglycan synthesis in injured liver—a brief review of molecular and cellular aspects. Eur J Clin Chem Clin Biochem 1994;32:225–237.

CHOLESTATIC LIVER DISEASES Alpini G, Phillips JO, LaRusso NF. The biology of biliary epithelia. In: Arias IM, Boyer JL, Fausto N, Jakoby WB, Schachter DA, Shafritz DA, eds. The Liver: Biology and Pathobiology, 3rd ed. Eds., New York: Raven, 1994; pp. 623–654. Bosma PJ, Chowdhury JR, Bakker C, Gantla S, de Boer A, Oostra BA, et al. The genetic basis of the reduced expression of bilirubin UDPglucuronosyltransferase 1 in Gilbert’s syndrome. N Engl J Med 1995;333:1171–1175. Ciotti M, Obaray R, Martin MG, Owens IS. Genetic defects at the UGT1 locus associated with Crigler-Najjar type I disease, including a prenatal diagnosis. Am J Med Genet 1997;68:173–178. de Vree JML, Jacquemin E, Sturm E, Cresteil D, Bosma PJ, Aten J, Deleuze J-F, Desrochers M, Burdelski M, Bernard O, Oude Elferink RPJ, Hadchouel M. Mutations in the MDR3 gene cause progressive familial intrahepatic cholestasis. Proc Natl Acad Sci USA 1998; 95:282–287. Green RM, Gollan JL. Crigler-Najjar disease type I: therapeutic approaches to genetic liver diseases into the next century. Gastroenterology 1997;112:649–651. Kaplan MM. Toward better treatment of primary sclerosing cholangitis. N Engl J Med 1997;336:719–721. Lee YM, Kaplan MM. Primary sclerosing cholangitis. N Engl J Med 1995;332:924–933. Leung PS, Chuang DT, Wynn RM, et al. Autoantibodies to BCOADC-E2 in patients with primary biliary cirrhosis recognize a conformational epitope. Hepatology 1995;22:505–513.

Li L, Krantz ID, Deng Y, Genin A, Banta AB, Collins CC, Qi M, Trask BJ, Kuo WL, Cochran J, Costa T, Pierpont ME, Rand EB, Piccoli DA, Hood L, Spinner NB. Alagille syndrome is caused by mutations in human Jagged1, which encodes a ligand for Notch1. Nature Genetics 1997;16:243–251. Oda T, Elkahloun AG, Pike BL, Okajima K, Krantz ID, Genin A, Piccoli DA, Meltzer PS, Spinner NB, Collins FS, Chandrasekharappa SC. Mutations in the human Jagged1 gene are responsible for Alagille syndrome. Nature Genetics 1997;16:235–242. Paulusma CC, Kool M, Bosma PJ, Scheffer GL, ter Borg F, Scheper RJ, Tytgat GN, Borst P, Baas F, Oude Elferink RP. A mutation in the human canalicular multispecific organic anion transporter gene causes the Dubin-Johnson syndrome. Hepatology 1997;25:1539–1542. Sato H, Adachi Y, Koiwai O. The genetic basis of Gilbert’s syndrome. Lancet 1996;347:557–558.

AUTOIMMUNE HEPATITIS Czaja AJ, Carpenter HA, Santrach PJ, Moore SB. Genetic predispositions for the immunological features of chronic active hepatitis. Hepatology 1993;18:816–822.

AMYLOIDOSIS Gertz MA, Kyle RA. Amyloidosis: prognosis and treatment. Seminars in Arthritis & Rheumatism 1994;24:124–138. Reilly MM, King RH. Familial amyloid polyneuropathy. Brain Pathol 1993;3:165–176.

POLYCYSTIC LIVER DISEASE Torres VE. Polycystic liver disease. In: Watson ML, Torres VE, eds. Polycystic Kidney Disease. Oxford: Oxford University Press, 1996; pp. 498–527. Mochizuki T, Wu G, Hayashi T, Xenophontos SL, Veldhuisen B, Saris JJ, Reynolds DM, Cai Y, Gabow PA, Pierides A, Kimberling WJ, Breuning MH, Deltas CC, Peters DJ, Somlo S. PKD2, a gene for polycystic kidney disease that encodes an integral membrane protein. Science 1996;272:1339–1342.

ALCOHOL- AND DRUG-INDUCED LIVER DISEASES Arnon R, Esposti SD, Zern MA. Molecular biological aspects of alcoholinduced liver disease. Alcohol Clin Exp Res 1995;19:247–256. Lee WM. Drug-induced hepatotoxicity. N Engl J Med 1995;333: 1118–1127. Lieber CS. Medical disorders of alcoholism. N Engl J Med 1995; 333:1058–1065. Mahler H, Pasi A, Kramer JM, Schulte P, Scoging AC, Bar W, Krahenbuhl S. Fulminant liver failure in association with the emetic toxin of Bacillus cereus. N Engl J Med 1997;336:1142–1148. Makin AJ, Wendon J, Williams R. A 7-year experience with severe acetaminophen-induced hepatotoxicity (1987–1993). Gastroenterology 1995;109:1907–1916. Mehendale HM, Roth RA, Gandolfi AJ, Klaunig JE, Lemasters JJ, Curtis LR. Novel mechanisms in chemically induced hepatotoxicity. FASEB J 1994;8:1285–1295. McKenzie R, Fried MW, Sallie R, Conjeevaram H, Di Besceglie AM, Park Y et al. Hepatic failure and lactic acidosis due to fialuridine (FIAU), an investigational nucleoside analogue for chronic hepatitis B. N Engl J Med 1995;333:1099–1105.

CHAPTER 42 / INHERITED LIVER DISEAESE

42

375

Inherited Liver Disease JUAN RUIZ AND GEORGE Y. WU

INTRODUCTION The liver is a major site of metabolism in humans, playing a central role in the metabolism of carbohydrate, protein, lipid, trace elements, and vitamins. In addition, detoxification and biliary excretion of various endogenous and exogenous compounds, especially lipophilic molecules, including most common drugs, and the degradation and elimination of a variety of hormones and hormonal metabolites, are distinct hepatic characteristics. Because of its involvement in so many metabolic pathways, it is not surprising that many inherited diseases of the liver are manifested as inborn errors of metabolism. Disease occurs as a result of lack of the required gene product or accumulation of a toxic metabolite. Molecular defects responsible for these disorders can be classified as single basepair substitutions, insertions, or deletions, and are exemplified in Fig. 42-1.

ABNORMAL METAL METABOLISM WILSON DISEASE Wilson disease (hepatolenticular degeneration) was described in 1912 by Kinnear-Wilson. The disease results from a disturbance in hepatic copper metabolism characterized by an inability to secrete copper into bile resulting in hepatic accumulation, progressive liver damage and, after hepatic overflow, accumulation in the brain and other organs. In addition, there is a defect in the incorporation of copper into ceruloplasmin. Copper is an important metal cofactor present in several enzymes (superoxide dismutase, tyrosinase, lysyl oxidase, dopamine-β-hydroxylase, peptide α-amidating enzyme, cytochrome oxidase, and ceruloplasmin). Although essential, only trace quantities are required and larger amounts are toxic. The normal body content is 70–100 mg and depends on the balance between intestinal absorption and biliary excretion (1–5 mg daily). The liver plays a major role in copper metabolism by the uptake of the recently absorbed metal that is excreted into bile and also incorporated into ceruloplasmin. Clinical Manifestations (Table 42-1) The disease has a worldwide prevalence of about 1–30,000 (1:200 for the gene) without ethnic preferences. There is a broad organ involvement. Approximately 40% of patients present with hepatic features, followed by neurological (34%) and endocrine, hematological (10%), or neuropsychiatric (12%). Hepatic manifestations tend to appear earlier than those of neurological origin (6–12 years for the hepatic From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

presentation vs 12 years or more for the neurological forms) and will be present always, irrespective of the initial presentation, if adequate treatment is not implemented. There may be a chronic insidious presentation, sometimes punctuated by transient icteric episodes or an acute onset with hemolysis and hepatic or renal failure as a result of an acute release of free nonceruloplasmin bound copper from the liver. Neuropsychiatric symptoms appear later than the hepatic manifestations and are associated with the presence of Kayser-Fleischer rings, which are a manifestation of copper deposition in the Descemet’s membrane. There is a strong correlation between the presence of Kayser-Fleischer rings in Wilson disease and copper deposition in the brain. Although the intellect is preserved, there may be a slow deterioration of the personality. Renal changes are related to copper deposition in tubular cells resulting in amino aciduria, glucosuria, uricosuria, hyperphosphaturia, hypercalciuria, renal tubular acidosis, and stone formation. Diagnosis It is most important to consider Wilson disease in the differential diagnosis of chronic liver disease in children and young adults and whenever liver disease is associated with neurological symptoms. Important parameters are: 1. Serum ceruloplasmin 250 µg/g dry weight. 3. Urinary copper excretion >100 µg/d. Diagnosis can be made by showing either parameter 1 and the presence of Kayser-Fleischer rings or both parameters 1 and 2. However, 50% of patients without neurological symptoms do not have Kayser-Fleischer rings at presentation, and 5–10% of patients have ceruloplasmin levels in the lower normal range. In those patients, the most important determination is liver copper content. If further discrimination is required, orally administered radiocopper (67Cu) incorporation into ceruloplasmin can be used. An absence of an increase in 67Cu serum levels after 24–48 h is diagnostically helpful. Linkage analysis using specific DNA markers can be applied to siblings. However, because of genetic heterogeneity, it is not useful for screening of potential heterozygotes in the general population. Confirmation in siblings should be sought by measuring serum copper, ceruloplasmin, and urinary copper. Genetic Basis of the Disease Wilson disease has an autosomal recessive inheritance. The gene (ATP7B), recently identified by positional cloning and the use of the closely related Menke’s

375

376

SECTION VI / GASTROENTEROLOGY

Figure 42-1 Different gene defects and its consequences. (A) Normal gene, messenger RNA, and encoded polypeptide. (B) Missense mutation leading to a single amino acid substitution: The results can be either a different isoform of the molecule (when the affected amino acid is not functionally essential) or a significant functional alteration of the product. (C) Missense mutation resulting in a premature stop codon and termination of the polypeptide: the final product is a truncated protein that has abnormal function or no function at all. (D) Insertion of a nucleotide producing a frameshift (modification of the amino acid sequence downstream from the mutation) and a premature stop codon (truncation of the polypeptide): The result is complete abolition of the protein activity (null phenotype). The same effect is produced by nucleotide deletions. (E) Scheme of the structure of a gene with the exons (boxes) and introns (lines). After transcription and splicing of the introns, the mRNA is produced. (F) Effects of mutations on the splice acceptor or donor sequences: Substitutions in exon/intron splice sites can result in improperly spliced message and abnormal gene product containing intronic sequences or missing exonic elements. In this case, exon 2 is removed from the mature mRNA. The black dot indicates the nucleotide mutation.

disease gene as a probe, is located on chromosome 13. It contains 23 exons and a 4.3-kb region that codes for a membrane-bound cation-transporting P-type adenosine triphosphatase. The family of P-type ATPases also includes the Menke’s Disease gene and several related bacterial and yeast genes. There is a 56% identity with the gene for Menke’s disease (ATP7A), an X-linked disorder characterized by a defective intestinal absorption of dietary copper. The identity is even higher (75–100%) at the amino acid level of functionally important domains. ATP7B is expressed mainly in the liver, kidney, and placenta. There is alternative splicing for some exons in the liver and the brain. The alternative isoforms may

have some functional significance, and the differential splicing may help regulate the amount of functional protein produced. The genetic defects in Wilson disease are heterogeneous. Using single-strand conformation polymorphism (SSCP) to identify the mutations, followed by sequencing, more than 30 different mutations have been described to date. Genotype analysis has revealed insertions and deletions causing frameshifts, stop codons (nonsense), altered splicing, and missense mutations. In particular, nucleotide substitutions have been described in the phosphorylation consensus sequence and in the hinge region of the ATPbinding site. Unlike Menke’s disease, where 20% of patients have

CHAPTER 42 / INHERITED LIVER DISEAESE

Table 42-1 Clinical Features of Wilson Disease Hepatic Neurological

Psychiatric Gastrointestinal Ophthalmological Hematological Renal Musculoskeletal Cardiovascular Endocrine

Acute hepatitis, fulminant hepatitis, chronic hepatitis, liver cirrhosis Akinesia, rigidity, muscle spasms, dystonia, dysarthria, dysphagia, resting and intention tremor, epilepsy, encephalopathy, peripheral neuropathy Cognitive impairment, behavioral changes, psychosis, affective disorders Cholelithiasis, pancreatitis Kayser-Fleischer rings, sunflower cataracts Hemolysis (acute, chronic), coagulopathy, intravascular coagulation, hypersplenism Renal tubular dysfunction, nephrolithiasis Osteoporosis, osteomalacia, osteoarthropathy Cardiomyopathy, arrhythmias Amenorrhea, miscarriages

large gene deletions, no large deletions have been found. Most patients studied are compound heterozygotes, having two different allelic mutations, a result of the large number of mutations. Molecular Pathophysiology of Disease It has been shown that copper transport across membranes is an ATP-dependent process. The P-type adenosine triphosphatase involved in Wilson disease contains several domains: six copper binding, nine transmembrane, transduction, channel, phosphorylation, phosphatase, and adenosine triphosphate-binding regions. A decrease in the number of active molecules or a decrease in its activity causes a defect in the secretion of copper into the bile because of a deficiency in the activity of the transporter system. The precise intracellular localization of the transport defect has not yet been elucidated. In addition, a reduction in the incorporation of copper into apo-ceruloplasmin to generate holo-ceruloplasmin, possibly a result of the same genetic defect, causes the accumulation of the metal in the hepatocytes. Damage in tissues where copper has accumulated is probably secondary to membrane lipid peroxidation, glutathione depletion, and polymerization of copperthionein, resulting in abnormal function of the mitochondria, lysosomes, microtubules, and nuclear DNA. Wilson disease is an excellent example of a metabolic disease caused by a single gene defect, with a correlation between genotype and phenotype based on the different defects in the mutated protein, e.g., domain modifications or complete absence of gene product. Specific types of mutations may predict the severity of the disease. For example, those mutations that destroy the function of the gene, stop codons truncating the protein or insertions, or deletions causing frameshifts, appear to correlate with an earlier onset of the disease, in one study approximately 7.2 years of age. Interestingly, the same study shows that in those mutations causing modifications in specific domains without affecting the whole molecule (missense mutations), the average age of onset of symptoms is higher, 16.8 years. In fact, some authors suggest than if the genetic defect is so substantial, e.g., large deletions leading to a null phenotype, the result would be a very early onset of the disease that could be misdiagnosed as some cases of Indian Childhood

377

Cirrhosis. Another important phenomenon is that of mutations affecting regions of the gene where a possible alternative splicing can occur. The removal of the mutation in some tissues would mitigate its deleterious effects and lead to a later onset of the disease. As an example, a severely detrimental mutation, 2010del7, located in exon 6, causing a frameshift, results in a later onset, neurological form of Wilson disease by an alternative splicing that removes the deletion. Ceruloplasmin carries 80% of the copper present in plasma and its function, although not well-understood, can be related to its ferroxidase, amine oxide, and antioxidant properties. The importance of the ferroxidase activity of ceruloplasmin is underlined by the deposition of iron and subsequent damage in the liver, pancreas, retina, and brain in patients with severe ceruloplasmin deficiency. Most patients with Wilson disease show a decrease in serum ceruloplasmin levels that correlate with the clinical manifestations of the disease. However, this decrease is not universal and no pathophysiological correlation has been established. Ceruloplasmin is synthesized in the liver in an inactive form, apo-ceruloplasmin, and is transformed to the active form, holoceruloplasmin, by the incorporation of up to six to seven atoms of copper per molecule. Both forms are different antigenically and holo-ceruloplasmin is the form usually detected in plasma in normal subjects. Normal amounts of ceruloplasmin in liver tissue from patients with Wilson Disease, normal ceruloplasmin RNA levels in Long Evans Cinnamon (LEC) rats, an animal model of Wilson Disease, together with the exclusive presence of apoceruloplasmin in serum, suggest a posttranslational defect in the incorporation of copper into ceruloplasmin rather than a defect in the ceruloplasmin gene itself. Management and Treatment Treatment is aimed at reducing the amount of copper in the body and to prevent the toxic effects of free copper in the tissues. D-penicillamine (1–2 g/d divided in two doses) is the most common treatment and must be given lifelong. D-penicillamine chelates copper, permitting urinary excretion. Additionally, it has detoxifying activity on intracellular copper. In some cases, neurological symptoms may worsen at the beginning of the treatment as copper is released from the liver. In cases of adverse effects to D-penicillamine, an alternative agent is trientine, another copper chelator. Other possible agents include tetrathiomolybdate, a chelator that is possibly more effective for neurologic disturbances, and zinc, which decreases intestinal copper absorption and stimulates metallothionein production, which in turn binds copper. Urinary copper determinations are used to monitor the treatment. Liver transplantation is the treatment of choice for patients with irreversible liver disease or fulminant hepatitis. Its role in cases of neurologic disease without hepatic insufficiency is not clear. Future Directions LEC rats possess a genetic defect characterized by a deletion of at least 900 bp at the 3' end of the rathomologous Wilson gene. No mRNA is detected in the liver or other tissues by Northern blots. The same clinical and biochemical features characteristic of Wilson disease have been demonstrated: excessive hepatic copper accumulation, defective holo-ceruloplasmin biosynthesis, and impaired biliary copper excretion. Gene therapy could be a potential option as a curative treatment of Wilson disease in the future and its feasibility can be tested in LEC rats. HEMOCHROMATOSIS The term hemochromatosis was first used by von Recklinghausen in 1889. The disease is caused by an inappropriately high intestinal absorption of iron that first ac-

378

SECTION VI / GASTROENTEROLOGY

cumulates in the parenchymal cells of the liver and later in other organs. Unlike most inborn errors of metabolism, the basic defect in hereditary hemochromatosis is not located primarily in the liver. However, its clinical manifestations are largely the result of liver disease and dysfunction. Iron is absorbed both as heme and nonheme forms (ferrous salt) by the intestinal mucosa. Once inside the enterocyte, it is coupled to transferrin and transported to tissues. Cells having transferrin receptors internalize the protein, and iron is released into the cytoplasm to be incorporated to heme and apoproteins. Because of its high toxicity, excess iron is stored in the body as ferritin or hemosiderin. Iron absorption is a poorly understood but extremely important process, because there is no physiological mechanism for excretion of excess iron. Absorption is responsible for regulation of normal iron balance. Losses because of shedding of mucosal cells along the gastrointestinal and genitourinary tracts and skin desquamation are relatively constant around 1 mg/d, and are compensated by a 10% absorption of the 10–20 mg of iron present in an average diet. Deregulation of the absorption in the presence of fixed losses lead to a constant increase in body iron stores (up to 40 g) with subsequent cell damage. Clinical Manifestations (Table 42-2) Although all races are affected, hemochromatosis is most common among Caucasoids of Celtic origin, with a prevalence of 1:300–400 and a carrier rate of 1:20. The disease is characterized by a latent period for many years while the iron stores increase in liver. This latent period depends greatly on the dietary iron intake as well as physiologic and pathologic blood loss. This explains the 9-to-1 ratio between affected men and women. The clinical picture is that of a multisystemic disease. Liver disease has mostly a benign and prolonged course with cirrhosis and liver failure present as late features. However, hepatocellular carcinoma can be the first manifestation of the disease. Hyperpigmentation of the skin is the result of an increase in melanin in the cells of the basal layer. Deposition of iron in the pancreas, heart, pituitary, and joints result in diabetes mellitus, cardiomegaly, arrythmias, heart failure, hypogonadism, hypoparathyroidism, hypothyroidism, chondrocalcinosis, demineralization, and chronic arthropathy. Diagnosis The introduction of routine serum iron studies for asymptomatic individuals as well as the use of HLA markers in families of affected individuals have improved the rate of early diagnosis and the time of diagnosis before the development of significant disease. This is important because treatment can restore life expectancy to normal if irreversible organ damage, e.g., cirrhosis, has not occurred. Diagnosis is made by serum iron studies, the most important being the transferrin saturation index. An index below 50% excludes the diagnosis, while one above 70% is suggestive of hemochromatosis. Serum ferritin concentration >200 µg/L in asymptomatic homozygotes and >900 µg/L in symptomatic patients is helpful. A high serum iron concentration >30 µmol/L alone is less specific. Confirmation of the diagnosis is made by liver biopsy, the definitive laboratory test, that shows increased iron deposition in hepatocytes (at the periportal areas in early stages) with paucity of iron elsewhere. This is in contrast to secondary iron accumulation in reticuloendothelial cells. Measurement of hepatic iron content allows the quantification of iron in the liver and the calculation of the hepatic iron index (tissue iron in mmol/g liver dry weight/age [years]). The index helps discriminate homozygotes, having indices >1.9, from heterozygotes and

Table 42-2 Clinical Features of Hereditary Hemochromatosis Features Weakness and lethargy Hepatomegaly Pigmentation (skin and mucosa) Diabetes mellitus Hypogonadotrophic hypogonadism Splenomegaly Loss body hair, dermal atrophy Anorexia, malaise Confusion, weight loss, vomiting Chronic arthropathy Hepatocellular carcinoma Abdominal pain Cardiomegaly, cardiomyopathy Vertigo, peripheral neuropathy Jaundice

Percentage 100 95 90 60 50 50 50 45 45 30 30 30 20 15 10

individuals with other iron-loading diseases with indices 29,000 cases. Subsequent large epidemics have occurred elsewhere in Asia, Africa, and Central America, with the epidemic in Xinjiang region of China in 1986–1988 affecting >100,000 people. The HEV is transmitted via the feco-oral route with a high attack rate. Pregnant women suffer from severe or fulminant hepatitis E with high mortality rates, approaching 20%. The virus is encountered sporadically among travelers to endemic areas in the West, including the United States. Infection with HEV has not been identified in animals other than man, although cynomolgus macaques and tamarins are susceptible and provide valuable animal models of HEV infection. The virus does not cause chronic disease. The HEV virions are spherical, nonenveloped particles with spikes and indentations on the surface (Fig. 43-1), which are characteristic of the calciviridae family, although HEV is thought to constitute a unique genus. The virus was first visualized by immunoelectron microscopy in stool from a patient and an experimentally infected cynomolgus macaque. The virus was subsequently cloned from RNA in infected bile and this utilized molecular hybridization screening of cDNA libraries. The HEV was found to have a 7.6-kb, single-stranded polyadenylated RNA genome. The viral genome contains three open reading frames with the ORF1 spanning 5079 bases in the 5' region and ORF2 spanning 1980 bases toward the 3' region with an additional 65 bases separating the ORF2 from the polyadenylated tail. The ORF3 spans 369 bases and overlaps the ORF1 by one nucleotide on its 5' end and ORF2 by 328 bases toward its 3' end. The ORF1 expresses nonstructural genes and contains a motif encoding an RNAdependent RNA polymerase, as well as two well-conserved motifs associated with helicases. The structural genes are in the 3' end of the genome and expressed via subgenomic transcripts. The virus expresses two polyadenylated mRNAs of 2- and 3.7-kb sizes. These mRNAs can be detected by PCR in the liver, serum, and fecal extracts of patients with HEV. The virus expresses type-specific epitopes, which are broadly reactive on immunotransblot analysis, in structural domains conferred by ORFs 2 and 3. There is evidence for significant regional variations in the ORF2 and 3 sequences in viral isolates obtained from different geographic parts of the world, although whether these differences are sufficient for molecular epidemiology studies needs further analysis. The hepatitis E antigen (HEAg) is recognized by immunostaining in the cytoplasm of infected hepatocytes and appears before the onset of clinically overt hepatitis. Serological tests have been developed by recombinant DNA technologies to assay for IgM and IgG class anti-HEV antibodies, although

394

SECTION VI / GASTROENTEROLOGY

diagnosis can also be made by detecting HEV nucleic acid sequences by PCR. HEPATITIS G VIRUS AND HEPATITIS GB VIRUSES The existence of patients with hepatitis and absence of markers for known hepatitis viruses generated ongoing interest in isolating additional viruses. In the early 1990s, three viruses were isolated from stored blood from a patient encountered in the 1950s with the initials GB. These viruses were designated GBV-A, GBV-B, and GBV-C, and assigned to the flaviviridae family. In 1995, a new RNA virus designated HGV was sequenced from the serum of a chronic HCV carrier by using an immunoscreening approach. The HGV was found to have an RNA genome of approximately 9.1 kb in size with an organization similar to flaviviridae, including HCV. The 5' end of the genome contains the structural genes and the 3' end the nonstructural genes. There are highly conserved sequence motifs for a helicase, two proteases, and an RNA-dependent RNA polymerase, although the polypeptides have not been extensively studied. Interestingly, sequence analysis showed that HGV was essentially identical to GBV-C (95 and 86% homology at amino acid and nucleotide levels, respectively). The global homologies between HGV and GBV-A (45%), GBV-B (25%), and HCV (26%) were somewhat less striking, although similarities in their genome organization and specific regions of the genome indicated close phylogenetic relationships. The 5' UTR region was found to be highly conserved in HGV, although sequences of viral isolates in different geographic regions show differences in the capsid protein. The HGV polyprotein expressed in a recombinant vaccinia virus system was found to produce multiple peptides using protease activities, similar to that observed with HCV. The HGV has been transmitted to chimpanzees, tamarins, and cynomolgus macaques with HGV RNA detected in serum by using reverse transcriptase-PCR. The HGV is prevalent in approximately 1–2% of volunteer blood donors in the United States, with increased prevalences in selected populations, such as polytransfused hemophiliacs (20%) and intravenous drug users (50%). However, although serum HGV RNA has been found for prolonged periods in patients with acute, fulminant, or chronic hepatitis, the significance of these findings is unclear because liver disease in HGV carriers can be blamed on other causes. By itself, HGV has not been found to cause liver disease. One view suggests that HGV is not actually pathogenic, and the issue of whether HGV and related HGBVs are even hepatitis viruses is presently under debate. The notion is that HGV is simply a harmless commensal virus.

MOLECULAR PATHOPHYSIOLOGY OF DISEASE Interactions between virus and host factors are most important in determining whether viral persistence and chronic liver disease will occur. The significant differences in the outcomes of infection with various hepatitis viruses are summarized in Table 43-4. Anicteric hepatitis is extremely common in all cases, specially hepatitis A and E. The HAV and HEV cause only acute hepatitis and are cleared by neutralizing antibodies with lifelong immunity. Similarly, HBV may also be cleared by anti-HBs, although the situation may rarely be complicated by the emergence of “vaccineescape” mutants (see below). On the other hand, HCV and HDV induce nonprotective antibody responses. When the host immune response is impaired, e.g., organ transplant recipients, chronic renal failure, chemotherapy, and so on, viral persistence is more

likely. Small viral inocula may be insufficient to elicit a vigorous immune response and also in this setting viruses may persist. The association of HBV and HCV with diseases involving circulating immune complexes (e.g., polyarteritis nodosa, membranoproliferative glomerulonephritis, aplastic anemia, cryoglobulinemias, and so on) is another example of how viral-host interactions may contribute to disease. Induction of chronic liver disease and disease exacerbations seen frequently in HBV and HCV infections are interlinked first and above all to the replicative states of the virus and the vigor of the host immune response. Both HBV and HCV are thought to cause chronic liver injury arising from activation of the cellular immune response. The injury is mediated by complex mechanisms involving both cytotoxic T lymphocytes (CTL) as well as secondary cytokine-mediated events. Additional mechanisms of cell injury involve expression of injurious viral proteins, although these need further characterization. Replicating and Nonreplicating States of Viruses These states of HBV are defined by the presence of replicative virus (serum HBV DNA+, HBeAg+/anti-HBe–, HBV DNA polymerase+, IgM anti-HBc+) or its absence (serum HBV DNA–, IgM anti-HBc–, HBeAg–/anti-HBe+). During HBV replication, hepatocytes display HBsAg and HBcAg and are surrounded by a mononuclear infiltrate containing activated lymphocytes. The HBV replication may be intermittent or persistent, at high levels or at low levels, and liver disease rapidly progresses with patients being infectious. During the nonreplicative phase, HBV DNA is negative and HBcAg is not produced. Active liver inflammation is absent and histologic evidence of HBV may be limited to “ground glass” hepatocytes or inactive cirrhosis. However, the states of replicative quiescence and activity are interchangeable. Such a change may occur in response to pharmacological interventions, such as use of glucocorticoids or chemotherapeutic agents, which increase viral replication, or spontaneously in response to unidentified stimuli. Spontaneous reactivation of HBV replication approaches an annual incidence of up to 10–35% and may lead to fulminant hepatitis, a rapid downhill course, or worsening of chronic liver disease. Nucleotide sequence analysis of HBV DNA in patients with reactivations has suggested that some episodes of reactivation represent reinfection with mutated HBV. During viral replicative states, serum HBV DNA may be found as partially double-stranded DNA, although full-length genomes and DNA:RNA hybrids may also be found. Maternal transmission is most significant for HBV, and the risk correlates with the viral replication status for both HBV and HCV. A characteristic feature of HBV is related to its integration in the host genome. In the replicative phase, HBV DNA in hepatocytes is mainly in free virion or replicating forms, and integration into host genome is not evident. In the nonreplicating phase, random integrations occur, replicative forms of free virion DNA are not found, and HBV is not secreted. In other nonpermissive infections, clusters of HBsAg-producing cells with the appearance of a focal clonal growth have been noted. In these cases, HBV has been identified in liver DNA integrated into the host genome. A mixed type of persistent infection may also occur with features of both replicating and nonreplicating infection in the same liver. In cases with impaired or blocked secretion of HBV, replicative intermediates and extrachromosomal forms may accumulate, facilitating HBV DNA integrations. It has been suggested that the region between DR1 and DR2 may facilitate HBV DNA integrations, but no consistently satisfactory mechanism has yet been identified. No specific cellular sites have

395

CHAPTER 43 / VIRAL HEPATITIS AND LIVER DISEASE

Table 43-4 Comparison of the Clinical Features of Viral Hepatitis Features Incubation (days) Fecal excretion Viremia Routes of transmission Oral Percutaneous Blood Transfusion Sexual Perinatal Age at onset

A

B

C

D

E

G

15–50 + Transient

28–160 – Prolonged

14–160 – Prolonged

20–120 – Prolonged

15–60 + Probably transient

14–30 – Prolonged

+ Rare Rare + – Children and adolescents

– + + – ± All ages

– + + – – All ages

+ – – Unknown Unknown Mostly children and adolescents

– Possible + Unknown Unknown Adult

Uncommon Uncommon +

Uncommon Uncommon –

May occur Frequent –

– – ?+

Rash or arthralgia Fever Waxing and waning hepatitis Biphasic hepatitis Cholestatic hepatitis Diagnosis of acute infection

Uncommon Frequent –

– + + + + All ages, neonatal in endemic areas May occur Uncommon –

Rare + Serum IgM anti-HAV

– – Serum IgM anti-HBc

– – Serum HCV RNA; IgM anti-HCV

+ – Serum IgM anti-HDV; HDV RNA

– + Serum anti-HEV; Serum or stool HEV RNA

– – Serum HGV RNA

Diagnosis of chronic infection

–

Serum IgM and IcG anti-HDV: HDV RNA Moderate to high

Serum HGV RNA

Rare

Serum HCV RNA; antiHCV Rare

–

Frequency of fulminant hepatitis Chronic carrier state Hepatocellular carcinoma Mortality during acute hepatitis

Serum HBsAg; serum HBV DNA Moderate

Very high in the pregnant

–

– –

+ +

+ +

+ ±

– –

+ –

Low, worse in >40 years of age

Low

Low

Low; very high during pregnancy

?None

Vaccine

Yes

Yes

No

May be high in superinfected chronic HBV carriers No

No

No

been found to support preferential HBV DNA integrations, although topoisomerase I, which is a nuclear enzyme, was shown to cleave open ciruclar HBV DNA at preferred integration positions and to potentially mediate DNA integrations. Nonetheless, more information is necessary to define how HBV DNA integrations occur and what their precise significance might be. During HBV replication, all viral proteins are expressed. The cellular immune response is thought to be induced most prominently by HBcAg expression. The preS1 and preS2 proteins are preferentially coexpressed with HBcAg during viral replication. Specific cellular immune responses against HBV antigens are wellrecognized. Circulating lymphocytes from chronic HBV carriers show cytotoxicity against autologous hepatocytes in vitro, which can be blocked by anti-HBs. The HBV proteins are not usually cytopathic, and liver injury is mediated by activated cytolytic T lymphocytes sensitized by HLA-restricted HBcAg expression. In contrast, HBeAg can occupy lymphocyte receptors used for binding HBcAg and provide immunological escape to the virus by

preventing clearance of infected hepatocytes despite HBcAg expression. The tolerance to HBV can indeed be broken in transgenic animals by administration of activated HLA-restricted CTL, which results in clearance of hepatocytes expressing HBV antigens, as well as decrease in viral gene expression because of cytokine release. In animals, extensive cell clearance has been associated with fulminant hepatitis, recreating the situation in patients with natural infection. The outcomes of liver disease in patients with HBV and HDV show variability, although it is unclear whether it is a result of viral factors, altered host immune response, or both. In susceptible hosts, HDV is highly infectious, and compared with HBV, its infectivity may be greater by multiple orders of magnitude. When HDV and HBV infections occur simultaneously (coinfection), HDV suppresses HBV replication and may cause a biphasic hepatitis, with the initial hepatic injury caused by HDV replication. Similarly, HDV suppresses HCV replication as well. On the other hand, when HDV infection occurs in chronic HBV carriers (superinfection),

396

SECTION VI / GASTROENTEROLOGY

the abundance of replicative HBV intermediates, as well as HBcAg, is reduced, and peripheral markers of HBV, including HBsAg, may disappear without actual virus clearance. In the setting of superinfection, even small HDV inocula that might not cause hepatitis in the setting of HBV and HDV coinfection, may be “rescued” and become pathogenic. Chronic HBV carriers are, therefore, at specially grave risk for acquiring HDV infection, when blood products negative for HBsAg could yet contain infectious HDV. Enhanced virulence of HDV during serial passages or because of unknown factors may cause severe or fulminant hepatitis. Although in vitro replication of HDV, as well as expression of HDAg in transgenic mice, does not necessarily lead to a cytopathic effect, when HDV is efficiently replicated, each cell may contain thousands of virion particles, and toxicity may potentially arise from hijacking of cellular resources. The exclusive use of RNA polymerase II by HDV may attenuate cellular gene expression. Also, evidence has been provided for sequence similarities between HDV RNA and the 7S signal recognition particle in cells for protein translocation, which may interfere with cellular protein synthesis. Finally, immune responses involving cytotoxic lymphocytes against HDV, similar to the host immune response against HBV, may be involved in cell clearance or cytokine-mediated cytotoxicity. In patients with HIV infection or immunosuppression, HDV replication is unchanged, suggesting that immunodeficiency does not influence HDV directly, which is in contrast with increased HBV replication in such settings. Studies of HCV in experimentally infected chimpanzees showed that HCV could be cleared rapidly, persist indefinitely with continuous replication, or replicate intermittently with variable periods of quiescence. During infection with HCV, the liver is markedly infiltrated with mononuclear cell aggregates containing B lymphocytes in the center surrounded by a network of reticular cells and another zone of various subpopulation of lymphocytes. The lymphocytes from chronic HCV carriers are cytotoxic in vitro to autologous hepatocytes, although this is not HLA-restricted. The activated lymphocytes recognize multiple viral antigens, including core, E1, and E2/NS proteins. In some, manifestations of autoimmune liver disease (liver-kidney-microsomal hepatitis) have been ascribed to molecular mimicry with HCV antigens. Similar to HBV, when HCV replication is attenuated, liver disease may be minimal. Of greater significance perhaps is the host immune response. “Healthy carriers” have been identified in whom chronic HCV infection and replication are established but have no liver disease. Disease progression may be arrested in the fortunate few responsive to interferon-α (IFN-α) with subduing or clearance of HCV replication (Fig. 43-6). Relationship with Hepatic Oncogenesis A number of epidemiologic studies clearly established relationships between chronic viral hepatitis and HCC. The relative risk for HCC in chronic HBV carriers may be increased by up to >200-fold. Similarly, hepadnavirus infection in woodchucks, and Peking ducks is associated with increased incidence of HCC. Observations in patients, as well as experimentally infected woodchucks indicate that simultaneous infection with HDV and HBV can accelerate the onset of HCC. The association between HCC and chronic HCV infection has also been confirmed by extensive epidemiological studies. For instance, in a large Japanese study, the relative risk for HCC in patients with chronic hepatitis was increased sevenfold in HBsAg+ patients and fourfold in patients with anti-HCV.

Figure 43-6 Representative responses in chronic HCV carriers treated with IFN-α. 䊉, serum ALT; 䊏, serum HCV RNA. (A) A patient who was found to respond to the drug during treatment but promptly relapsed with resumption of viral replication, as well as hepatic inflammation upon discontinuation of the drug. This is one of the most common responses to IFN-α treatment. (B) This patient showed less dramatic decrease in serum HCV RNA while on IFN-α, but the virus was cleared in a delayed manner. Most such patients clear virus temporarily, although in some sustained viral clearance may be associated with significant improvement in liver disease and decreased requirement for OLT.

The potential role of specific viral and cellular gene products in HCC has been subjected to extensive investigation. Whereas candidate genes have been identified for HBV, which integrates into the host genome, similar mechanisms have not been found for HCV, which does not integrate in the genome. The HBV DNA integrations were originally observed in HCC-derived cell lines and similar observations were subsequently made in patients. Most HBV DNA integrations occur during persistent infection in precancerous tissues. Although the majority of HBV integrations produce unique bands on DNA blot analysis, which indicates clonal expansion of cells containing integrated sequences, whether viral integrations contribute to cell transformation or clonal expansion has not been established. The integrated HBV DNA sequences vary from relatively simple to multiple rearrangements, deletions, inversions or duplications, and chromosomal translocations. Most integrations disrupt the viral genome organization, although some may contain greater than unit-length HBV DNA with apparently intact genomes. The HBV integrations have not

CHAPTER 43 / VIRAL HEPATITIS AND LIVER DISEASE

been shown to induce insertional mutagenesis, although random HBV integrations have been found in or adjacent to specific genes, including the erb A protooncogene and cyclin A gene. Other HBV integrations have been responsible for translocations between chromosomes 17q22 and 18q11, chromosomes 17 and 7, and chromosomes 9 and 5. However, HBV does not contain an oncogene, and transfection with HBV DNA in general lacks mutagenic potential. Evidence concerning whether HBV could augment or activate specific proto-oncogenes by itself has been inconclusive, although increased c-myc expression has been shown in HCC in woodchucks. Alternative mechanisms concerning the loss of tumor suppressor genes, such as p53, have not provided incriminating evidence related to HBV. Another way to examine the role of viral proteins in causing disease is by using transgenic mouse lines. In transgenic mice expressing only the small HBsAg molecule, no liver disease is observed. However, when large HBsAg is expressed in transgenic mice, the filamentous form of HBsAg is overproduced and entrapped in the cytoplasm. In the HBV transgenic mouse line designated 50-4, expression of large HBsAg leads to hepatic injury, ground glass hepatocytes, cellular degeneration, and Kupffer cell or hepatocyte hyperplasia, followed by aneuploidy, focal nodular hyperplasia or liver cancer. As ground glass hepatocytes are frequently observed in chronic HBV carriers, it is possible that similar pathophysiological mechanisms may be involved. The HBxAg has also been subjected to much study because this protein can transactivate cellular genes, which may be important in cell growth. However, evidence from three lines of transgenic mice expressing HBxAg showed that most animals did not develop HCC, although mild hepatitis, nuclear pleomorphism, focal necrosis, nodular hyperplasia, or increased mitotic activity were noted in some animals. In another transgenic line with more prolonged expression of HBxAg, there were progressive changes in the liver with eventual development of hepatic adenomas or carcinomas. Therefore, under certain circumstances, HBV gene products could interact with other cellular genes to induce oncogenesis. Whether HDV has an impact at the cellular and molecular level in hepatic oncogenesis requires more study. One possibility is that the HDV ribozyme may alter transcriptional regulation despite a >10-fold lower affinity for heterologous DNA sequences. In addition, HDAg is a basic phosphoprotein with histone-like DNA-binding capacity, and this could also potentially modulate the activity of regulatory genes in infected hepatocytes. Additional possibilities by which HBV, as well as HCV, could induce HCC must include activation of specific cell populations under selection pressure induced by ongoing cell necrosis and regeneration and the responsiveness of emerging cell clones to specific growth factors. The evidence for the involvement of growth factors in HCC has begun to emerge. The transforming growth factor-alpha (TGF-α) may help transform liver cells by increasing cell proliferation and turnover rates, and TGF-α transgenic mice show increased incidence of HCC. In addition, coexpression of TGF-α along with c-myc or SV40 TAg accelerates both the onset and growth rate of HCC. Therefore, the current evidence is inconclusive with regard to any single viral gene being oncogenic for the liver. It may be that HCC arises mainly from increased cell turnover and attempt of newly emerged cells to evade virus infection. Repeated cycles of liver injury and cell proliferation could amplify selection events with the help of hepatic growth factors.

397

VIRAL MUTATIONS AND MUTANT PROTEINS In general, RNA viruses are more prone to mutations because transcription with the RNA polymerase carries with it less fidelity. However, because HBV utilizes a reverse transcription step, it also shows significant mutations in its nucleotide sequence. Are there additional reasons for mutations to arise? Particularly, could mutations in viruses reflect evolutionary pressures or escape mechanisms? These tantalizing questions are unresolved, but viral mutations have been observed with greater frequency in patients with advanced liver disease, presumably related to active viral replication and robust immune responses, as well as in response to antiviral therapies, such as lamivudine treatment of chronic HBV carriers and IFN-α treatment of chronic HCV carriers. HAV MUTATIONS Based on differences in the nucleic acid sequence, seven genotypes of HAV have been recognized and designated genotypes I–VII. The genotypes show nucleic acid diversity of 15% or more. Only four HAV genotypes are associated with human disease, with the rest being pathogenic for monkeys. The genotypes I and III have been subtyped with additional nucleic acid sequence differences (IA and B; III A and B) approaching approximately 7.5%. The value of genotyping at present concerns molecular epidemiology studies with certain genotypes being more prevalent in some geographic areas. For instance, the genotype IA is most commonly isolated in the United States. However, despite the genotype variability, HAV constitutes a single serotype and development of neutralizing antibodies after natural infection confers lifelong immunity. The neutralizing IgG class antibodies bind to a common region on the surface of the virus. Although the amino acid sequences of VP1, VP2, and VP3 regions may also exhibit differences in various substrains, these changes do not provide an immunological “escape” for reinfection in an immune host. Analysis of immunogenic domains of the virus, including the use of recombinant DNA technology, has been subject to major efforts because of its significance in the development of effective vaccines. The HBV mutations identified involve the HBsAg, HBeAg, HBcAg, and possibly HBxAg, coding regions. The most commonly involved HBsAg region is responsible for conferring the “a” group-specific epitope, which is shared by all three small, middle, and large HBsAg molecules. The consequences of point mutations in HBsAg are substitutions in specific amino acids, such as at amino acid positions 122, 145, and others, which result in the loss of epitope conformation. As a result, the virus no longer can bind anti-HBs. The practical consequence is that in patients with anti-HBs, including after vaccination, reinfection with the mutated HBV could occur, along with chronic liver disease. A number of such cases termed “vaccine-escapes” have recently been identified with liver disease in previously immune patients. In addition, patients with mutations in the HBsAg region have been found to be younger and to present with more advanced disease, including cirrhosis, ascites, and HCC, albeit cirrhosis may be less active at such a presentation. Mutations in the precore HBV region were found in patients that were replicating HBV without producing HBeAg. Analysis of HBV DNA purified and amplified by PCR from sera of such HBV DNA+, HBeAg–/anti-HBe+ patients showed mutations in two codons of the precore region of the ORF C, one of which leads to generation of a TAG translational stop signal. The result is a mutant HBeAg, which is no longer secreted and may be indirectly associated with liver disease of greater severity because HBeAg would

398

SECTION VI / GASTROENTEROLOGY

no longer mask HBcAg from being recognized by activated CTL. In additional patients, precore mutant HBV has been identified in the setting of acute or fulminant hepatitis, although the evidence suggests that such mutant HBVs are not necessarily more likely to induce fulminant hepatitis in naive individuals. Nonetheless, patients with precore mutations may present more often with greater symptoms, jaundice, and hepatic inflammation. Mutations in the precore HBV regions have also been known to generate replication-incompetent viruses, which may function in a dominant negative fashion to suppress replication of the wild-type virus, and their potential use for gene therapy has recently been explored (see below). HCV MUTATIONS Genotyping by type-specific PCR; direct sequencing; restriction-fragment-length polymorphisms; DNA hybridization with type-specific probes or reverse hybridization with sub-type-specific probes; and serologic analysis with group-specific antibodies has identified multiple HCV variants. The substitution rate of HCV has been estimated at 0.9 × 10–3 per nucleotide per year. Proposals for genotype classifications are based on variations in the core region (four subtypes), NS3 region (two subtypes) and NS5 region (two subtypes). Six major genotypes are recognized with a series of subtypes. The original United States HCV isolate is designated HCV type I and is the commonest isolate (72%), followed by type II (14%), type III (6%), and type IV (1%). The type Ia HCV is most widespread in the United States, whereas type Ib is most prevalent worldwide. Mixed genotypes may also coexist (4%). The type I HCV is found in approximately 50% cases with posttransfusion HCV. Although patients with HCV types III and IV tend to be younger, viremia levels are in general similar in all genotypes, and no firm correlations have been established between genotype and disease activity. However, in Japan and Europe, type I HCV may be associated with greater viremia, compared with types 2 and 3. In addition, HCV types II and III may respond better to treatment with IFN-α as compared with type I HCV. Within individual patients, HCV may exist as a spectrum of closely related viral genomes, which is referred to as viral quasispecies. The development of HCV quasi-species appears to be related to increasing duration of carrier states, greater viremia, HCV type Ib, and a poor response to IFN-α treatment. The significance of viral quasi-species is unclear and this requires more analysis. Finally, the genotypic variability in HDV has resulted in tentative identification of three genotypes. The majority of HDV infection is with the genotype I (80% nucleotide sequence similarity among strains). The genotype II was found in an isolate from Japan (75% sequence similarity with genotype I). Three viral isolates from South America constituting genotype III have shown only 60–65% sequence similarity with the other two genotypes and have been responsible for fulminant hepatitis. The presence of HDV quasi-species within individual patients have also been noted. Its RNA genome is consistent with poor fidelity during replication and a mutation rate has been estimated of 3 × 10–2 to 3 × 10–3 substitutions/nucleotide/year.

THERAPIES The therapies are extremely limited. Patients with acute hepatitis are treated symptomatically because most cases resolve spontaneously. If fulminant hepatitis develops, intensive care management is necessary and orthotopic liver transplantation (OLT) may be lifesaving. The pharmacokinetics of drugs may be

altered in the presence of liver disease and this needs careful consideration. The availability of effective drugs to inhibit viral replication has been unsatisfactory. A variety of nucleoside analogs and other compounds have been ineffective. The largest experience has been with IFN-α, which inhibits the HBV DNA polymerase and also suppresses HCV replication. There are three kinds of interferons (α, β, and γ), which are endogenous, naturally occurring glycoproteins with antiviral, antiproliferative and immunomodulatory activities. IFN-α is derived from leukocytes, IFN-β from fibroblasts, and IFN-γ from T lymphocytes. Although IFN-α and -β are similar and share the same cell surface receptor, IFN-γ is a different molecule with a separate receptor. Most antiviral studies in HBV and HCV have been with IFN-α, as IFN-β is unstable. The mass of data concerning therapeutic trial of IFN-α in HBV and HCV is too large to discuss here. However, the drug benefits by decreasing viral replication, hepatic inflammation, and fibrosis. Patients with relatively low level of serum HBV DNA or serum HCV RNA, along with younger age, less advanced or prolonged disease, and possibly with HCV genotype other than type I are likely to benefit most. Despite the best circumstances, however, viral clearance is extremely rare, responses are most often temporary, side effects are multiple, and IFN-α therapy is expensive. Recently, lamivudine, a nucleoside analog, has shown promise in suppressing HBV replication, although rapid appearance of mutant strains in this situation is a limitation. Searches for alternative therapies have continued and include consideration of gene therapy approaches. The use of antisense gene sequences to suppress transcription offers a novel therapeutic approach, which is most attractive because of its potential specificity. In principle, stretches of nucleotides in a reversed orientation to the native mRNA sequences will hybridize and inactivate mRNAs. The suggested mechanisms in antisense gene regulation include perturbations in splicing at exon/intron junctions, mRNA transport from nucleus to cytoplasm, initiation factor-binding for translation, assembly of ribosomal subunits at start codons, and interference with ribosome travel along coding mRNA sequences. Studies with synthetic oligonucleotides have identified potential targets for HBV antisense sequences. Recently, HBsAg expression has been inhibited in vitro with the use of antisense oligonucleotides against HBsAg coding regions. The use of receptormediated endocytosis to deliver oligonucleotides specifically to the liver may potentially increase local drug delivery. In addition, antisense oligonucleotides against early preS or DR2 regions were effective in vivo, although continuous infusions and very high doses were required. However, despite these successes, extensive body distribution of oligonucleotides and low cellular uptake and degradation of oligonucleotides pose problems in in vivo applications. An alternative could be to express antisense sequences using cDNAs introduced into infected cells. This has not been extensively tested because unless dominant negative targets could be identified, such antisense strategies could potentially be defeated by the stochastic requirements of inactivating each and every mRNA transcribed. It may be possible to use cDNA sequences to covalently target and repress transcriptionally active DNA domains, but this requires more study. The use of dominant negative HBcAg mutants that markedly inhibit wild-type HBV replication has also been proposed. Using cell lines supporting viral replication, expression of the inhibitory HBcAg mutants with retroviral or adenoviral vectors almost completely suppresses HBV replication.

CHAPTER 43 / VIRAL HEPATITIS AND LIVER DISEASE

Efforts have also been ongoing to identify suitable genetic targets for treating HCV. The conserved 5' structural sequences are particularly attractive for this purpose. If expression of HCV core antigen could be suppressed, virion assembly and replication should be significantly affected. Studies with antisense oligonucleotides complementary to the HCV core antigen have shown marked inhibition of viral replication. Similarly, if ribosomal attachment to the IRES could be inhibited, polyprotein translation should be affected, with major effects on all aspects of the virus. Additional therapies could potentially be devised with better insights into mechanisms of viral persistence and molecular biology of specific viruses. The knowledge concerning the role of individual gene products is necessary for defining suitable genetic targets. The potential therapies could well involve novel antiviral compounds, such as the recently investigated BMS-200475, which is a proprietary guanosine analog. Whether immunological mechanisms could be harnessed for therapeutic purpose, such as by adoptive transfer or induction of antiviral CTL requires further work. Another potential mechanism could be to induce resistance in hepatocytes against viral infection. Similarly, intimate knowledge of the viral receptors would help develop novel drugs to block viral entry into cells. An understanding of viral transcriptional regulatory mechanisms would help develop novel drugs to block viral entry into cells. Also, it is clear that viruses are capable of infecting animals in a species-specific manner depending on the presence or absence of their receptors in cells. Although specific receptors responsible for internalizing hepatitis viruses into cells have not been fully characterized, further work could offer additional therapeutic targets. On the other hand, transplantation of cells or organs from another species resistant to viral hepatitis could offer means to treat patients in end-stage liver failure. Patients with chronic HCV are able to undergo OLT with meaningful survival despite reinfection of the transplanted liver, which is not the case in chronic HBV carriers. Starzl and colleagues proposed that transplantation of liver from baboons, which are not susceptible to HBV, will be successful if allograft rejection were prevented. However, in chronic HBV carriers that were subjected to OLT with baboon livers, survival was limited to only a few weeks, although the transplanted livers showed normal function. It was most likely that the baboon livers were rejected, and further work concerning xenograft tolerance is necessary if such an approach is to be seriously considered. VACCINES Prevention is the most effective strategy for reducing the burden of viral hepatitis. Success in culturing HAV and HBV has led to the development of highly successful vaccines. The early HBV vaccines were prepared from inactivated virus or HBsAg purified from pooled plasma of HBV carriers. Recombinant HBsAg vaccines contain smaller nonglycosylated HBsAg particles (17–25 nM) compared with the natural vaccine and are produced commercially by expressing HBV cDNAs in yeast. The recombinant HBV vaccines have been proven to be safe and highly effective (90–95%), although adequate levels of antiHBs require three doses during 6 months and additional boosters after several years. Similarly, effective vaccines have been developed for HAV (97% protection), although currently available vaccines use inactivated virus. Recombinant HAV vaccines based upon the P1 region or the simian HAV epitopes are under study. Despite significant efficacy, conventional vaccines are disadvantaged by the requirement for multiple doses, several months to adequate immunity, boosters to sustain antibody response, and

399

large amounts of antigens, which are expensive to produce. Recently, interest has developed in the use of genetic vaccines, whereby specific viral proteins are expressed by intramuscular injection of cDNAs. Such genetic vaccines against HBV and HCV have been shown in animals to efficiently express recombinant proteins and to rapidly generate high titer antibodies. The genetic vaccines have potential in converting nonresponders with no antibodies after HBV vaccination. In addition, repeated administration of genetic vaccines is being explored to induce antibody or CTL responses for clearing HBV in chronic carriers.

FUTURE DIRECTIONS AND CONCLUSIONS Further advances in our understanding of the molecular biology of hepatitis viruses is critical for pathophysiological and therapeutic insights. The availability of effective cell culture systems for several viruses, including HCV, HDV, and HEV will greatly facilitate analysis of molecular mechanisms and vaccine development. Recent progress in culturing HCV and HEV has been encouraging. Similarly, the availability of convenient animal models will be helpful and further allow testing of novel therapies in vivo. Transgenic mouse models will be helpful in analyzing the role of individual viral gene products. Information is necessary to fill gaps in our knowledge of viral life cycles, host-virus interactions, and viral pathogenesis. Parallel advances in gene transfer vectors and mechanisms regulating expression of introduced genes will facilitate gene therapy strategies while appropriate genetic targets are identified. The current pace of molecular discoveries suggests that the next decade will bring forth enormously exciting novel tools for unraveling the pathophysiology of viral hepatitis, as well as revolutionary ways to treat infected individuals.

SELECTED REFERENCES Alt M, Renz R, Hofschneider PH, Paumgartner G. Specific inhibition of hepatitis C viral gene expression by antisense phosphorothiote oligodeoxynucleotides. Hepatology 1995;22:707–717. Blumberg B, Alter HJ, Visnich S. A “new” antigen in leukemia sera. JAMA 1965;191:541–546. Bradley DW. Hepatitis E virus: a brief review of the biology, molecular virology, and immunology of a novel virus. J Hepatol 1995;22 (Suppl 1):140–145. Chisari FV. Hepatitis B virus transgenic mice: insights into the virus and the disease. Hepatology 1995;22:1316–1325. Dusheiko GM, Roberts JA. Treatment of chronic type B and C hepatitis with interferon alfa: an economic appraisal. Hepatology 1995;22: 1863–1873. Feinstone SM, Kapikian AZ, Purcell RH. Hepatitis A: detection by immune electron microscopy of a virus like antigen associated with acute illness. Science 1973;182:1026–1029. Goodarzi G, Gross SC, Tewari A, Watabe K. Antisense oligodeoxyribonucleotides inhibit the expression of the gene for hepatitis B virus surface antigen. J Gen Virol 1990;71:3021–3025. Gupta S, Govindarajan S, Cassidy WM, Valinluck B, Redeker AG. Acute delta hepatitis: serological diagnosis with particular reference to hepatitis delta virus RNA. Am J Gastroenterol 1991;86:1227–1231. Gupta S, Shafritz DA. Viral mechanisms in hepatic oncogenesis. In: Arias IM, Boyer J, Fausto N, Jacoby WR, Schachter D, Shafritz DA, eds. The Liver: Biology and Pathobiology. New York: Raven, 1994; pp. 1429–1453. Guptan RC, Thakur V, Sarin SK, Banerjee K, Khandekar P. Frequency and clinical profile of precore and surface hepatitis B mutants in AsianIndian patients with chronic liver disease. Am J Gastroenterol 1996;91:1312–1317. Houghton M, Weiner A, Han J, et al. Molecular biology of the hepatitis C viruses-implications for diagnosis, development and control of viral disease. Hepatology 1991;14:381–388.

400

SECTION VI / GASTROENTEROLOGY

Howard CR. The structure of hepatitis B envelope and molecular variants of hepatitis B virus. J Viral Hepat 1995;2:165–170. Ji W, St CW. Inhibition of hepatitis B virus by retroviral vectors expressing antisense RNA. J Viral Hepat 1997;4:167–173. Lai MM. The molecular biology of hepatitis delta virus. Annu Rev Biochem 1995;64:259–286. Linnen J, Wages J Jr, Zhang-Keck Z-Y, et al. Molecular cloning and disease association of hepatitis G virus: a transfusion-transmissible agent. Science 1996;271:505–508. Offensperger W, Offensperger S, Walter E, et al. In vivo inhibition of duck hepatitis B virus replication and gene expression by phosphorothioate modified antisense oligodeoxynucleotides. EMBO J 1993;12:1257–1262. Scaglioni P, Melegari M, Takahashi M, Roy Chowdhury J, Wands J. Use of dominant negative mutants of the hepadnaviral core protein as antiviral agents. Hepatology 1996;24:1010–1017.

Shafritz DA, Shouval D, Sherman ML, et al. Integration of hepatitis B virus DNA into the genome of the liver cells in chronic liver disease and hepatocellular carcinoma. New Engl J Med 1981;305: 1067–1073. Starzl T, Fung J, Tzakis A, et al. Baboon to man liver transplantation. Lancet 1993;341:65–71. van Doorn LJ. Molecular biology of the hepatitis C virus. J Med Virol 1994;43:345–356. Wu GY, Wu CH. Specific inhibition of hepatitis B viral gene expression in vitro by targeted antisense oligonucleotides. J Biol Chem 1992; 267:12,436–12,439. Zhang H, Chao SF, Ping LH, Grace K, Clarke B, Lemon SM. An infectious cDNA clone of a cytopathic hepatitis A virus: genomic regions associated with rapid replication and cytopathic effect. Virology 1995;212:686–697.

CHAPTER 44 / PANCREATIC EXOCRINE DYSFUNCTION

44

401

Pancreatic Exocrine Dysfunction DAVID WHITCOMB AND JONATHAN COHN

INTRODUCTION The pancreas is a retroperitoneal gland with both endocrine and exocrine functions. Pancreatic endocrine hormones are released by the islets into the circulation where they are essential for the regulation of intermediary metabolism (see Chapter 47). Pancreatic exocrine secretions (pancreatic juice), produced by the concerted action of pancreatic acinar and ductal cells, are delivered into the duodenum where they are essential for normal digestion. Different pancreatic enzymes promote the efficient digestion of complex carbohydrates, lipids, proteins, and other nutrients (Table 44-1). It is noteworthy that many of these digestive enzymes have the potential to digest components of the pancreas itself. Such autodigestion is normally prevented by several mechanisms. For example, many digestive enzymes exist in an inactive (proenzyme) form until they reach the duodenum. Other mechanisms either prevent activation of proenzymes inside the pancreas or inhibit the enzymes if they are inadvertently activated prematurely. Pancreatic exocrine dysfunction can present clinically either as an acute illness (acute pancreatitis) or as a chronic process (pancreatic insufficiency with or without chronic pancreatitis). Inherited causes of pancreatic exocrine disease include hereditary pancreatitis and cystic fibrosis (CF). In both conditions, the responsible gene is now known and information is rapidly emerging concerning the role of the gene’s protein product during normal pancreatic function and during the pathogenesis of pancreatic exocrine dysfunction.

ACUTE PANCREATITIS BACKGROUND Acute pancreatitis is defined clinically as pancreatic inflammation characterized by the sudden onset of abdominal pain in association with elevation in blood or urine pancreatic enzyme levels. In about 80% of the cases, acute pancreatitis is mild and recovery is expected. However, in the remaining cases, acute pancreatitis is severe and can produce potentially lifethreatening complications. The keys to clinical management are early recognition of severe acute pancreatitis, anticipation of complications, and prevention of recurrence. CLINICAL FEATURES Acute pancreatitis usually causes epigastric abdominal pain of acute onset. The pain tends to be constant and may radiate to the mid-back. Vomiting occurs in 80% of cases and provides little pain relief. Abdominal tenderness with guarding From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

occurs in half of cases but tends to be less severe than anticipated based on either the severity of reported pain or the degree of pancreatic inflammation (by abdominal CT or ultrasound), and this discrepancy results from the retroperitoneal location of the pancreas. DIAGNOSIS In a patient presenting with typical abdominal pain, a serum amylase or lipase level more then 2.5 times the upper limit of normal strongly suggests the diagnosis of acute pancreatitis. CT with intravenous contrast is the most sensitive method of assessing the degree of pancreatic injury, pancreatic necrosis, and local extrapancreatic complications but is usually unnecessary in mild pancreatitis. In most cases, acute pancreatitis is associated with an identifiable cause, with gallstones and alcohol ingestion being the most common factors (Table 44-2). When the cause of pancreatitis is identified early, this may suggest strategies to prevent ongoing pancreatic injury and to prevent recurrence. MOLECULAR PATHOPHYSIOLOGY A breakthrough in understanding the nature of acute pancreatitis came in 1896 when Chiari proposed that this condition represents autodigestion, rather than infection, of the pancreas. Autodigestion of the pancreas is a major threat because this organ normally produces enormous quantities of potentially dangerous digestive enzymes. To prevent autodigestion, three types of protective strategies are employed. In the first strategy, many enzymes (except amylase and lipase) are synthesized as proenzymes requiring enzymatic cleavage for activation. Activation normally occurs after the proenzymes are secreted by the pancreas into the intestinal lumen, where the intestinal brush border enzyme, enterokinase, cleaves trypsinogen activation peptide (TAP) from trypsinogen (the proenzyme) to generate trypsin (the active enzyme). Trypsin also activates trypsinogen and efficiently activates all other pancreatic digestive enzymes. Thus, the activation of trypsin in the duodenum is highly desirable as a key element in the rapid propagation of the digestive enzyme activation cascade. In the pancreas, however, widespread activation of trypsin may be disastrous. This is especially important in humans because human trypsinogen has a propensity to autoactivate. Additional protective strategies are therefore necessary to prevent autoactivation of the digestive enzyme cascade in the pancreas. For example, within the acinar cell, autoactivation of trypsinogen and other digestive enzymes is limited by maintenance of low intracellular calcium, by maintenance of a neutral pH, and by the subcellular isolation of zymogen granules. Furthermore, because of a constant background of trypsinogen autoactivation, a limited amount of pancreatic secretory trypsin inhibitor (PSTI) is synthe-

401

402

SECTION VI / GASTROENTEROLOGY

Table 44-1 Pancreatic Digestive Enzymes Proteinases Trypsin(ogen), Chymotrypsin(ogen), (Pro)carboxypeptidases, (Pro)elastases Kalikreinogen Pancreatic α-amylase Pancreatic lipase Other enzymes (Pro)colipase I, II (Pro)Phospholipase A2 Ribonuclease Deoxyribonuclease I

sized with trypsinogen. If trypsin activity in the acinar cells exceeds the inhibitory capacity of PSTI, then a trypsin self-destruct mechanism is initiated in which the peptide chain connecting the two globular domains of trypsin is cut at arganine II (R117) by trypsin and mesotrypsin. Cleavage at this site permanently inactivates trypsin and other digestive enzymes with this design because the two halves of the molecule separate, thereby eliminating the active site of the enzyme. Finally, if active digestive enzymes escape from the pancreatic acinar cells into the intracellular space or plasma, they are inactivated by hepatocyte-derived α2-macroglobulin and α1-protease inhibitors. Thus, multiple protective strategies exist to prevent pancreatic autodigestion and acute pancreatitis. The sequence of events initiating acute pancreatitis at the cellular level remains controversial, but several features are common to most models of experimental pancreatitis. Microscopically, zymogen granules accumulate at the apex of the acinar cells without being secreted. This functional block of secretion reflects the disruption of normal intracellular mechanisms, including the organization of microtubules and other proteins, which are necessary for exocytosis. The second feature is premature activation of digestive enzymes within the acinar cells. Although the intracellular site and mechanism responsible for digestive enzyme activation is uncertain, it seems clear that activation of these enzymes promotes “autodigestion” of pancreatic cells and initiates an inflammatory reaction. Once an episode of acute pancreatitis is triggered, a characteristic inflammatory reaction occurs that is independent of the initiating factors. When acinar cells are damaged, they begin leaking activated digestive enzymes and other cellular elements into the interstitial space. This triggers an inflammatory response with the release of proinflammatory mediators, including interleukin-1 (IL-1), IL-2, IL-6, IL-8, tumor necrosis factor-α (TNF-α), and platelet-activating factor (PAF) (see Chapter 30). If pancreatic injury is mild, then the inflammatory response is self-limited and the process contained. However, if pancreatic injury is severe, then a more intense inflammatory response ensues. This inflammatory response is responsible for many of the local and systemic complications of acute pancreatitis. For example, the acute phase reaction and hypermetabolic state in severe acute pancreatitis with fever, decreased peripheral vascular resistance, and increased cardiac output are driven by cytokines. The pulmonary complications of hypoxia and adult respiratory distress syndrome (ARDS) largely develop because of the release of phospholipase A2 from activated macrophages. Finally, the progression of pancreatic edema to

Table 44-2 Causes of Acute Pancreatitis Mechanical

Toxins

Hereditary Metabolic Trauma Infections

Vascular Idiopathic

Gallstones Biliary microlithiasis Post ERCP (multifactorial) Pancreatic cancer Anatomical predisposition Pancreas divisum Annular pancreas ? sphincter of Oddi dysfunction Chronic alcohol abuse (multifactorial) Scorpion bite Medications Anticholinesterases L-asparaginase Azathioprine Dideoxyinosine (DDI) Estrogen 6-Mercaptopurine Salicylates Thiazide-diuretics Valproic acid Vinca alkaloids Other (idiosyncratic reactions) Trypsinogen gene mutation (? others) Hyperlipidemia (triglycerides >1000) Hypercalcemia Blunt abdominal trauma Penetrating ulcer Postoperative/post-ERCP Mycoplasma pneumonia Virus (mumps, coxsackie, CMV, HSV, HIV) Bacteria Mycobacterium avium-intracellular Vasculitis (e.g., connective tissue diseases) Hypotension (Postoperative: e.g., CABG) Infarction Unknown

pancreatic necrosis may be driven, in part, by TNF-α. In animal models of acute pancreatitis, when proinflammatory cytokines are inhibited (i.e., PAF inhibitors) or the anti-inflammatory cytokine IL-10 is given, the severity of pancreatic injury is reduced, and survival improves. Although clinical intervention at the initiation of acute pancreatitis is impossible in most cases, the severity of acute pancreatitis may be reduced by early administration of agents that modify the immune response. Specific PAF inhibitors, for example, show great promise in this regard. Thus, acute pancreatitis could be considered to progress through three general phases: an acute injury from a variety of causes, a characteristic inflammatory response that depends on the severity of injury, followed by a healing phase. GENETIC PREDISPOSITION TO ACUTE PANCREATITIS Hereditary pancreatitis and familial hypertryglyceridemia predispose affected family members to acute pancreatitis. Hereditary pancreatitis represents a primary disorder of the pancreas, whereas patients with familial hypertryglyceridemia suffer from pancreatitis secondary to markedly elevated tryglycerides. Hereditary pancreatitis is a chronic, recurrent inflammatory disorder of the pancreas affecting family members in multiple generations. This disorder encompasses all features of acute pancreatitis and chronic pancreatitis. Thus, understanding the molecular basis

CHAPTER 44 / PANCREATIC EXOCRINE DYSFUNCTION

of pancreatic injury in hereditary pancreatitis may provide insight into the pathogenesis of other forms of acute and chronic pancreatitis. Hereditary pancreatitis is inherited as an autosomal dominant trait with 80% penetration and variable expression. Nearly 100 kindreds have been reported since the genetic nature of this disorder was recognized by Comfort and Steinberg in 1952. The cause of hereditary pancreatitis was determined after the gene was localized to chromosome VII and then identified by sequencing of candidate genes. This analysis showed that hereditary pancreatitis is caused by a mutation that was in a cationic trypsinogen gene, resulting in an arginine to histidine (R-H) substitution at residue 117. This R117H mutation was observed in all patients with hereditary pancreatitis and in all obligate carriers from five kindreds, but not in individuals who married into the families nor in 140 unrelated individuals. The effect of the R117H mutation on trypsinogen regulation has been studied through the X-ray crystal structure, molecular modeling, and protein digest analyses. Taken together, these data suggest that R117 occurs in a trypsin-sensitive site in the chain connecting the two globular domains of trypsinogen, and that the R117H mutation eliminates the final control of excessive trypsin activity in the pancreas by removing the natural cleavage site of the trypsin-like enzymes (Fig. 44-1). This finding explains why hereditary pancreatitis is an autosomal-dominant disorder. Even if only half of the trypsinogen molecules resist proteolytic degradation, this is sufficient to initiate the digestive enzyme activation cascade in the pancreas and therefore express the phenotype. This also explains why attacks of acute pancreatitis occur episodically rather than continually: Attacks occur when the rate of intrapancreatic trypsinogen activation overwhelms the protective effects of PSTI. The discovery of the hereditary pancreatitis gene has had important implications for understanding the pathogenesis of pancreatitis in general. First, it showed that intrapancreatic protease activation is sufficient to cause acute pancreatitis in humans. Second, it emphasizes the importance of trypsinogen autoactivation in humans and may explain why it has been difficult to study acute pancreatitis in animal models in which trypsinogen is more stable. Finally, it demonstrates a direct link between recurrent acute pancreatitis and chronic pancreatitis. MANAGEMENT AND TREATMENT Mild pancreatitis tends to be a self-limited process. Treatment is supportive and includes pancreatic rest (nothing by mouth), intravenous hydration, and analgesics. Patients usually recover within 3 days and can be discharged from the hospital when oral intake is adequate to prevent dehydration. By contrast, severe pancreatitis can result in life-threatening complications requiring aggressive management. Because these patients are at risk for sudden death, ICU admission is often indicated during the initial phase of severe acute pancreatitis. To predict which patients with acute pancreatitis are most likely to develop life-threatening complications, several prognostic schemes have been developed (see Table 44-3). These schemes remain useful for assisting in identification of the subset of patient with acute pancreatitis for whom intensive monitoring is most appropriate. Early recognition is key to the management of the most common life-threatening complications of severe acute pancreatitis. Cardiovascular collapse and respiratory failure each may occur rapidly and require aggressive support. Multiorgan failure may develop as nutritional stores are depleted, and early nutritional support should

403

Figure 44-1 Model of trypsin self-destruct mechanism preventing pancreatic autodigestion. (A) Autoactivation and enzymatic activation of trypsinogen generate trace amounts of active trypsin within pancreatic acinar cells. Active trypsin is inhibited by a limited supply of trypsin inhibitor (PSTI). If trypsin activity exceeds the inhibitory capacity of PSTI, then proenzymes, including mesotrypsin and enzyme Y, are activated. These enzymes feed-back to inactivate wildtype (wt) trypsinogen, trypsin, and other zymogens. (B) Activation of mutant (HP) trypsin in amounts that exceed the inhibitory capacity of PSTI results in activation of proenzymes. Since the Arg 117 cleavage site for trypsin-like enzymes is replaced by His in the HP mutant trypsin, trypsin continues to activate trypsinogen and other zymogens unabated, leading to autodigestion of the pancreas and pancreatitis.

Table 44-3 Prognostic Criteria in Acute Pancreatitis Ranson Criteriaa

Glasgow Criteriab

On admission Age >55 years Leukocyte count >16,000/mm3 Blood glucose >200 mg/dL Serum LDH >350 IU/L Serum GOT (SGOT/AST) >250 IU/dL At 48 h Hematocrit decreases >10% BUN rises >5 mg/dL Serum calcium 6 L On Admission Age >55 years WBC count >15,000/µL Glucose > 180 mg/dL BUN >45 mg/dL PO2 20 mg/dL over a 2-h period. Lactase deficiency is diagnosed by the lack of serum glucose elevation and, frequently, the onset of symptoms of malabsorption, with gas, diarrhea, bloating, and abdominal pain. If the serum glucose increases, yet symptoms also occur, one must consider occult diabetes mellitus. Alternatively, hydrogen breath testing a trial of dietary lactose avoidance may be attempted. After restriction of oral intake of milk and milk products, the patient is assessed for symptom resolution. Finally, small bowel biopsy allows for definitive diagnosis, but is unnecessary given the adequacy of the other two approaches. GENETIC BASIS AND MOLECULAR PATHOPHYSIOLOGY As determined by population and family studies, adultonset hypolactasia is an autosomal-recessive trait, and persistence of lactase activity is autosomal-dominant. The molecular basis for lactase persistence/nonpersistence has been the subject of intensive investigation. In the majority of studies, lactase-specific activity and protein levels correlate very well with mRNA concentrations, indicating transcriptional regulation of lactase expression. However, sequence analysis of the lactase cDNA and 1 kb of its promoter region has not revealed differences between sequences from persistent and nonpersistent patients. Yet, recent studies using known DNA marker polymorphisms in the exons of the lactase gene have suggested that cis-acting elements are responsible for the persistence/nonpersistence phenotype, based on differences in level of expression of lactase comparing one allele to the other. On the other hand, characterization of trans-acting factors that regulate lactase expression have revealed that levels of NF-LPH-1, an intestine-specific transactivator that binds to a sequence close to the TATA box, are high in newborn pigs with high lactase activity and low in adult pigs with low lactase activity. Still other data suggest that there is a second phenotype of lactase deficiency, in which posttranslational processing of lactasephlorizin is altered. Thus, the precise molecular mechanisms underlying lactase persistence/nonpersistence are still controversial. MANAGEMENT The treatment of symptomatic lactase deficiency consists of removal of lactose-containing nutrients from the diet and/or the use of oral enzyme replacement therapy. It is preferable to continue some milk and dairy food ingestion because of their rich calcium stores, especially in children and in women who are prone to osteoporosis. Lactase oral preparations are avail-

413

able in capsules or tablets and can be chewed. Replacement pills are taken with meals and titrated to diminish symptoms. Lactase may also be added directly to milk and ingested after an overnight incubation, or pretreated milk can be purchased. FUTURE DIRECTIONS The ready availability of oral lactase capsules and the ease of treatment of symptoms by dietary avoidance or supplementation makes more sophisticated approaches to treatment, such as gene therapy, unnecessary. The major research interest still lies in determining how epidemiologic observations are linked to genetics; i.e., what are the genetic alterations that led to lactase persistence in the milk-drinking populations of the world, and are there differences among different ethnic or racial groups? These are questions of great interest to population and evolutionary geneticists.

GLUCOSE–GALACTOSE MALABSORPTION CLINICAL FEATURES Glucose–galactose malabsorption is a rare congenital disorder in which enterocytic transport of glucose and galactose is absent, resulting from mutation of the gene encoding the Na+/glucose cotransporter SGLT-1. This disease is characterized by the onset of profuse watery diarrhea in newborns that responds to the removal of dietary sources of glucose and galactose. Diarrhea usually develops within 4 d of birth. The diarrhea is so severe that dehydration is frequent. As in other diseases of carbohydrate absorption, the presence of unabsorbed sugars leads to an increased osmotic load in the small intestine with increased luminal fluid. Laboratory findings include an acidic stool pH because of colonic bacterial metabolism of unabsorbed sugar. In addition, sugar is found in the urine, consistent with the expression of SGLT-1 in the kidney. Oral monosaccharide tolerance tests reveal that absorption of glucose and galactose, but not fructose, is impaired. GENETICS AND MOLECULAR PATHOPHYSIOLOGY Glucose galactose malabsorption is an autosomal recessive disease. The precise incidence of this disorder has not been determined, but it is clearly rare, as indicated by the few case reports in the literature. Frequently, there is a history of consanguineous marriage. Expression cloning of the cDNA that encodes SGLT-1, the Na/glucose cotransporter, by Wright’s group, provided a major breakthrough in understanding pathophysiology of this disorder. Specific mutations in SGLT-1 lead to loss of transporter function in humans, resulting in glucose and galactose malabsorption. Analysis of cDNA reverse-transcribed from RNA from intestinal biopsies of two affected sisters showed a single missense mutation, with a single base change from G to A at position 92, changing an aspartate to an asparagine. Their parents were heterozygous for this mutation. The autosomal recessive inheritance of this disorder was confirmed. The mutant SGLT1 mRNA was expressed in Xenopus laevis oocytes and compared to the normal SGLT1 mRNA for ability to transport a glucose analog, α-methyl d glucopyranoside. As anticipated, the mutant SGLT1 demonstrated no detectable transport activity. Subsequently, 30 other patients have been screened for mutations in SGLT1. The mutations included missense, nonsense, and frame shift mutations, and some occurred in splice sites at intronexon boundaries. Two areas were identified as “hot spots” for missense mutations, between residues 289–304 and 369–405. Analysis of the mutant proteins indicated that loss of transporter activity was a result of production of truncated or nonfunctional

414

SECTION VI / GASTROENTEROLOGY

protein, defective trafficking of the transporter, or defects in the transport cycle. The SGLT1 gene has been localized to chromosome 22 on the distal q arm. MANAGEMENT The treatment of glucose–galactose malabsorption is the immediate removal of lactose, glucose, and galactose-containing substances from the diet. Fructose is welltolerated, and a fructose-based formula can be substituted, with rapid resolution of symptoms. There are reports of increasing tolerance to glucose and galactose with age. Rechallenge with small amounts of these sugars may be possible. FUTURE DIRECTIONS The cloning of the SGLT-1 cDNA and the identification of naturally occurring mutations has already led to notable advances in our knowledge of the basic mechanisms of sugar transport. The regulation of expression of this and other cotransporters, their mode of insertion in the membrane, and structure–function correlation may be addressed in the research laboratory. With the identification of the SGLT1 gene, gene therapy may eventually be considered. However, in the near future, this is unlikely because of the relative ease of treatment by elimination diets, and the rarity of this disorder.

ABETALIPOPROTEINEMIA CLINICAL FEATURES Abetalipoproteinemia is a rare disorder, characterized by the absence of apolipoprotein (apo) B containing lipoproteins in plasma. Patients present in infancy with diarrhea resulting from severe fat malabsorption, poor weight gain, and acanthocytosis. Atypical retinitis pigmentosa and spinocerebellar degeneration become apparent at older ages. Ocular symptoms include decreased night and color vision, and neurologic signs of ataxia, spastic gait and dysmetria, loss of deep tendon reflexes, and decreased vibratory and proprioceptive sense appear in the second decade. Spontaneous hemorrhage characteristic of a bleeding diathesis may occur because of vitamin K deficiency. DIAGNOSIS The hallmark of this disorder is an abnormal plasma lipid profile, in association with signs of spinocerebellar degeneration and a pigmented retinopathy. Total cholesterol and triglyceride plasma levels are extremely low, and there is an absence of apo B-containing plasma lipoproteins, including chylomicrons, VLDLs, and LDLs. Most patients are anemic as a result of enhanced hemolysis of abnormal red blood cells (acanthocytes). Fat malabsorption leads to deficiency of fat soluble vitamins. In particular, vitamins E, K, and A are affected since they are transported in chylomicrons. Vitamin E deficiency is most severe, because its absorption from the intestine, transport to the liver, and circulation in the plasma are dependent on apo B-containing lipoproteins. Vitamin E deficiency is clinically important since it is thought to be responsible for the development of neurologic and retinal sequelae. Intestinal biopsy reveals lipidladen enterocytes but normal villi and is useful for distinguishing this disorder from other intestinal illnesses associated with fat malabsorption. GENETICS AND MOLECULAR PATHOPHYSIOLOGY Abetalipoproteinemia is a rare autosomal recessive disorder. Consanguinity is common in these families. An absence of plasma lipoproteins containing apo B led to the hypothesis that this disorder is a result of a defect in the apo B gene. However, linkage between apo B and abetalipoproteinemia in families has not been established by restriction fragment-length polymorphism analysis, and defects in the apo B gene have not been identified. Since assembly or secretion of apo B-containing lipoproteins were

most likely defective in this disorder, the microsomal triglyceride transport (MTP) protein was examined in affected individuals. This gene, located on chromosome 4q22-24, encodes a protein that catalyzes the transport of triglyceride, cholesteryl ester, and phospholipid between membranes. It is a heterodimer composed of a protein disulphide isomerase, and an 88-kDa subunit that is responsible for the lipid transport activity. This larger subunit shares homology with lipovitellin 1, a member of the lipid-binding lipovitellin complex that serves as a source for lipid in the developing embryo. Wetterau and colleagues showed that MTP activity, and particularly the 88-kDa subunit, normally present in microsomes from liver and intestine, was absent in intestinal biopsies from patients. Sequence analysis of cDNA and genomic DNA from patients with abetalipoproteinemia revealed homozygous frameshift and nonsense mutations in the gene encoding the 88kDa subunit, resulting in truncated proteins. Analysis of additional patients indicated frameshift and splice site mutations, most of which lead to the formation of a truncated protein. Transfection of Cos-1 cells with mutant MTP cDNAs lacking part of the carboxy-terminus show decreased lipid transfer activity, supporting the hypothesis that alterations of this protein are the primary cause of abetalipoproteinemia. MANAGEMENT A low-fat diet and supplementation with high doses of vitamin E as well as vitamins A and K are required. In particular, it is critical to avoid vitamin E deficiency, since this appears to be the primary cause of the severe and irreversible neurologic and ocular changes characteristic of this disease. FUTURE DIRECTIONS The cloning of the cDNA encoding the microsomal triglyceride transport protein raises the possibility of gene therapy in the future. Although many of the devastating neurologic and ocular sequelae may be avoided by high doses of vitamin E, there may be a role for gene replacement in refractory patients. Also, rapid, definitive diagnosis is crucial, so that prenatal screening of siblings may be useful. Finally, since patients with MTP deficiency have very low cholesterol and triglyceride levels, identification of selective inhibitors of its activity may prove to be a valuable addition to the treatment of hypercholesterolemia.

SELECTED REFERENCES Boll W, Wagner P, Mantei N. Structure of the chromosomal gene and cDNAs coding for lactase-phlorizin hydrolase in humans with adulttype hypolactasia or persistence of lactase. Am J Hum Genet 1991;48:889–902. Escher JC, de Koning ND, van Engen CGJ, et al. Molecular basis of lactase levels in adult humans. J Clin Invest 1992;89:480–483. Evans L, Grasset E, Heyman M, Dumontier AM, Beau JP, Desjexu JF. Congenital selective malabsorption of glucose and galactose. J Ped Gastroenterol and Nutr 1985;4:878–886. Goggins M, Kelleher D. Celiac disease and other nutrient related injuries to the gastrointestinal tract. Am J Gastroenterol 1994;89(Suppl): S2–S17. Harvey CB, Wang Y, Hughes LA, et al. Studies on the expression of intestinal lactase in different individuals. Gut 1995;36:28–33. Hediger MA, Coady MJ, Ikeda TS, Wright EM. Expression cloning and cDNA sequencing of the Na/glucose cotransporter. Nature 1987; 330:379–381. Heresbach D, Alizadeh M, Dabadie A, et al. Significance of interleukin1 beta and interleukin-1 receptor antagonist genetic polymorphism in inflammatory bowel diseases. Am J Gastroenterol 1997;92: 1164–1169. Hermiston ML, Gordon JI. Inflammatory bowel disease and adenomas in mice expressing a dominant negative N-cadherin. Science 1995; 270:1203–1207.

CHAPTER 45 / DYSFUNCTION OF THE BOWEL

Hugot JP, Laurent-Puig P, Gower-Rousseau C, Olson JM, Lee JC, et al. Mapping of a susceptibility locus for Crohn’s disease on chromosome 16. Nature 1996;379:821–823. Kagnoff MF, Harwood JI, Bugawan TL, Erlich HA. Structural analysis of the HLA-DR, -DQ, and -DP alleles on the celiac disease-associatedDR3 (Drw17) haplotype. Proc Natl Acad Sci USA 1989;86: 6274–6278. Kuhn R, Lohler J, Rennick D, Rajewsky K, Werner M. Interleukin-10deficient mice develop chronic enterocolitis Cell 1993;75:263–274. Lloyd M, Mevissen G, Fischer M, et al. Regulation of intestinal lactase in adult hypolactasia. J Clin Invest 1992;89:524–529. Marsh MN. Gluten, major histocompatibility complex, and the small intestine. A molecular and immunobiologic approach to the spectrum of gluten sensitivity (‘celiac sprue’). Gastroenterology 1992;102: 330–354. Martin MG, Turk E, Lostao P, Kerner C, Wright EM. Defects in Na/glucose cotransporter (SGLT1) trafficking and function cause glucose-galactose malabsorption. Nat Genet 1996;12:216–220. Rader DJ, Brewer HB. Abetalipoproteinemia: new insights into lipoprotein assembly and vitamin E metabolism from a rare genetic disease. JAMA 1993;270:865–869. Sadlack B, Merz H, Schorle H, Schimpl A, Feller AC, Horak I. Ulcerative colitis-like disease in mice with a disrupted interleukin-2 gene. Cell 1993;75:253–261.

415

Sharp D, Blinderman L, Combs KA, et al. Cloning and gene defects in microsomal triglyceride transfer protein associated with abetalipoproteinemia. Nature 1993;365:65–69. Sollid LM, Markussen G, Ek J, Gjerde H, Vartdal F, Thorsby E. Evidence for a primary association of celiac disease to a particular HLA DQ alpha/beta heterodimer. J Exp Med 1989;169:345–350. Stenson WF, Lobos E. Sulfasalazine inhibits the synthesis of chemotactic lipids by neutrophils. J Clin Invest 1982;69:494–497. Strober W, Ehrhardt RO. Chronic intestinal inflammation: an unexpected outcome in cytokine or T cell receptor mutant mice. Cell 1993;75: 203–205. Troelsen JT, Olsen J, Noren O, Sjostrom H. A novel intestinal trans-factor (NF-LPH-1) interacts with the lactase-phlorizin hydrolase promoter and co-varies with the enzymatic activity. J Biol Chem 1992; 267:20,407–20,411. Turk E, Zabel B, Mundlos S, Dyer J, Wright EM. Glucose/galactose malabsorption caused by a defect in the Na+/glucose cotransporter. Nature 1991;350:354–356. Wetterau JR, Aggerbeck LP, Bouma M, et al. Absence of microsomal triglyceride transfer protein in individuals with abetalipoproteinemia. Science 1992;258:999–1001. Yang Y, Harvey CB, Ratt WS, et al. The lactase persistence/nonpersistence polymorphism is controlled by a cis-acting element. Hum Molec Genet 1995;4:657–662.

CHAPTER 46 / MECHANISMS OF HORMONE ACTION

417

ENDOCRINOLOGY VII SECTION EDITORS:

MICHAEL J. MCPHAUL AND J. LARRY JAMESON

418

SECTION VII / ENDOCRINOLOGY

CHAPTER 46 / MECHANISMS OF HORMONE ACTION

46

419

Mechanisms of Hormone Action WILLIAM L. LOWE, JR., RICHARD G. PESTELL, LAIRD D. MADISON, AND J. LARRY JAMESON

INTRODUCTION Hormones provide a form of communication between different cells and from one organ to another. The dramatic effects of hormones on physiologic functions such as growth and metabolism have allowed relatively sophisticated studies to elucidate their sources of production, sites of action, and how they are controlled. Because hormones act by binding to specific receptors, studies that have defined their mechanisms of action have also provided important paradigms for how receptors activate intracellular signaling cascades that lead to altered cellular responses. Endocrine disorders ultimately involve abnormalities of hormone production or action. A relatively large number of these diseases have a well-defined genetic basis. There are now many examples of mutations that occur in genes that encode hormones, their receptors, second messenger signaling pathways, and the transcription factors that transduce hormone signals. The purpose of this chapter is to review basic principles of hormone signaling through their receptors. There has been a remarkable expansion of knowledge in this area in recent years. The structures and signal transduction networks for membrane receptors are finally coming into focus; nuclear receptors were cloned within the last decade and have revealed an unexpectedly complex family of proteins that serve as direct regulators of gene expression. The impact of this new information on our understanding of the pathophysiology of disease is already being felt in terms of the ability to predict specific receptors or pathways that may harbor mutations in different disorders. In this chapter, we will cite selected examples of such mutations, and the reader is referred to other chapters in this section for more detailed discussions of these disorders.

CLASSIFICATION OF HORMONES AND RECEPTORS Hormones can be divided generally into three major classes: (1) derivatives of amino acids (e.g., dopamine and catecholamines); (2) peptides and proteins (e.g., thyrotropin-releasing hormone or insulin); and (3) derivatives of steroids (e.g., estrogen or cortisol). For the most part, amino acid derivatives and peptide hormones interact with membrane receptors on the cell surface, whereas the steroid hormones act by crossing the plasma membrane to interact with intracellular receptors. From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

419

As described in detail below, membrane receptors can be divided into several distinct classes. After binding hormones, these receptors activate a complex array of second messenger signaling pathways that often involve cascades of kinases. Nuclear receptors, such as the glucocorticoid or progesterone receptors, bind hormones in the cytoplasm before translocation into the nucleus (Fig. 46-1). Other receptors in this family, such as the thyroid hormone receptor, bind hormone in the nucleus without a separate hormone-induced translocation step. In the nucleus, these receptors interact with DNA target sites to either stimulate or repress the expression of specific genes. Regardless of the class of receptors, certain principles apply to hormone–receptor interactions. Hormones bind to receptors with high affinity and specificity to allow appropriate physiologic responses. Most receptors bind hormones with affinity constants that approach the circulating levels of hormones (usually other Ubiquitous Other tissues > neural Neural, endocrine Neural, platelets Ubiquitous Ubiquitous Liver, lung, kidney Blood cells Ubiquitous Ubiquitous

Increases adenylyl cyclase and Ca2+ channel activity Increases adenylyl cyclase activity Increases cGMP phosphodiesterase activity Increases cGMP phosphodiesterase activity Unknown Decreases adenylyl cyclase and K+ channel activity Decreases adenylyl cyclase and K+ channel activity Decreases adenylyl cyclase and K+ channel activity Decreases Ca2+ channel activity Unknown Increases phospholipase C-β1 activity Increases phospholipase C-β1 activity Increases phospholipase C-β1 activity Increases phospholipase C-β1 activity Unknown Unknown

Table 46-2 Endocrine Disorders Associated with Mutations of Seven Transmembrane Domain Receptors and G Proteins Mutated protein Gain of function mutations Gαs Gαi2 LH receptor TSH receptor PTH receptor Calcium sensing receptor Loss of function mutations Gαs ACTH receptor Vasopressin receptor Calcium sensing receptor FSH receptor LH receptor GHRH receptor TSH receptor TRH

Disorder

OMIM no.

Type of mutation

McCune-Albright syndrome Acromegaly-somatotroph adenomas Hyperfunctioning thyroid adenomas Adrenal cortical and ovarian neoplasms Familial male precocious puberty Nonautoimmune hereditary hyperthyroidism Hyperfunctioning thyroid adenoma Jansen metaphyseal chondrodysplasia Hypoparathyroidism; hypocalcemia

174800

Somatic Somatic Somatic Somatic Autosomal dominant Autosomal dominant Somatic Autosomal dominant Autosomal dominant

Albright hereditary osteodystrophy Familial ACTH resistance Nephrogenic diabetes insipidus Familial hypocalciuric hypercalcemia Neonatal severe hyperparathyroidism Hypergonadotropic ovarian dysgenesis Male pseudohermaphroditism Dwarfism Congenital hypothyroidism Congenital hypothyroidism

300800 202200 304800 145980

139360 152790 275200 169468 145980

136435 152790 139191 275200 188545

Autosomal dominant Autosomal recessive X-linked Autosomal dominant Autosomal recessive Autosomal recessive Autosomal recessive Autosomal recessive Autosomal recessive Autosomal recessive

OMIM, Online Mendelian inheritance of man; LH, luteinizing hormone; TSH, thyroid stimulating hormone; PTH, parathyroid hormone; ACTH, adrenocorticotropic hormone; FSH, follicle-stimulating hormone; GHRH, growth hormone releasing hormone; TRH, thyrotropin releasing hormone.

responses to other receptors, as well as to the cytoplasmic and nuclear compartments. For the most part, these signaling pathways consist of enzyme cascades, although there are exceptions, including gases such as nitric oxide (NO), lipids, and transcription factors (such as the STAT proteins). In many cases, receptors activate several different signaling pathways. The strategy of using combinations of pathways provides an important mechanism for generating specific cellular responses. Another characteristic of signal transduction pathways is the potential for “crosstalk” in which one pathway may activate or inhibit another pathway. Again, this generates the potential for diverse responses. In the

section below, specific examples are provided for some of the better studied signaling pathways that play a key role in endocrine responses.

p21ras AND MITOGEN-ACTIVATED PROTEIN KINASES Because growth factors induce a variety of cellular responses, including proliferation, there has been great interest in delineating the pathways that mediate growth factor responses. Recently, there has been tremendous progress in the elucidation of numerous cytoplasmic and nuclear kinases involved in growth responses.

426

SECTION VII / ENDOCRINOLOGY

The number of different kinases is enormous, and they have been estimated to account for as many as 5–10% of transcribed genes. The family of mitogen-activated protein kinases (MAPKs) includes the extracellular signal-regulated kinases (ERKs), stressactivated or c-Jun N-terminal protein kinases (SAPKs on JNKs), and p38-kinases (Fig. 46-6). As noted above, tyrosine phosphorylation of growth factor receptors, such as EGF, creates a binding site for a variety of adaptor proteins. One adaptor, growth factor receptor-binding protein 2 (Grb-2), binds to tyrosine-phosphorylated proteins through its central SH2 domain. Grb-2 contains flanking SH3 domains that recruit proline-rich guanine nucleotide exchange factors, such as SOS (Drosophila son of sevenless). After growth factor treatment (e.g., EGF), the Grb-2–SOS complex associates with the tyrosine-phosphorylated receptor, which leads to the activation of p21ras and induction of downstream kinases. The signal transduction pathways downstream of growth factor receptors involve a complex interplay of parallel, diverging, and converging pathways. The Ras proteins link growth factor receptor activation to protein phosphorylation and gene regulation through cascades of protein kinases. Induction of MAPKs requires dual phosphorylation on threonine and tyrosine residues by MAPK kinases (MAPKK) or MEKs. Several different protein kinases are capable of functioning as MEK kinases (MEKKs), including Raf-1 (Fig. 46-6). In some cell types, the SAPKs and p38 kinases are activated by cellular stresses, including specific stimuli such as tumor necrosis factor-α (TNF-α), UV irradiation, or genotoxic alkylating chemicals. The mechanisms by which one particular MAPK module is activated in response to a particular stimulus may depend on which combinations of kinases are present as well as interactions among various signaling cascades. Transcription factors (e.g., c-Jun, c-Fos, ATF-2, Elk-1) are one of the targets of MAPK cascades and provide an important mechanism for altering patterns of gene expression in response to extracellular signals. Phosphorylation of the transcription factor can alter interactions with transcriptional partners, change DNA binding affinity for a particular DNA sequence, or affect the activation surfaces that interact with the basal transcription apparatus. As an example, c-Jun is phosphorylated by several different kinases, including ERKs and SAPKs. Amino-terminal phosphorylation of c-Jun by the SAPKs enhances transactivation function, whereas phosphorylation of c-Jun near its DNA binding domain enhances DNA binding. As with hormone regulatory systems, negative feedback is critical for dampening signal transduction systems once they have been activated. Because most of these pathways involve reversible protein phosphorylation, protein phosphatases provide a mechanism for temporally modulating specific signals. It has been estimated that there may be as many as 1000 distinct protein phosphatase genes. The physiologic importance of these phosphatases is exemplified by the ability of SV40 small t antigen to exert its cellular transforming effects by inhibiting protein phosphatase 2A (PP2A), with consequent induction of MEK and ERK activity in the cell. Various growth factors induce phosphatases differentially. This may explain, in part, why treatment with one growth factor can result in sustained activation of ERK, whereas another factor causes transient stimulation of the kinase.

CALCIUM SIGNALING Many cellular responses involve alterations in intracellular calcium as a mechanism for transmitting signals. There are two main pathways that initiate Ca2+ signaling. In nonexcitable cells,

Figure 46-6 Signal transduction cascades. Most cells express a wide variety of different growth factor receptors. Each of these receptors can activate an array of signaling pathways, not all of which are shown. Tyrosine kinases induce the binding of SH2 domain proteins, such as Grb2. Grb2 binds to guanine nucleotide exchange factors, such as son of sevenless (SOS). This pathway can also be activated by the insulin receptor via insulin-related substrate (IRS) or by cytokine receptors via Janus-activated kinases (JAK). Although Gsα-coupled receptors act primarily through the cAMP pathway, they can also activate mitogen-activated kinases (MAPK) through their β–γ-subunits. This figure emphasizes the observation that different receptors can activate the ras pathway along with one or more downstream kinase cascades. Ras activates several downstream kinases in a cell type-specific manner. The preferential activation of a particular MAPK module, such as activation of the extracellular regulated kinase (ERK) rather than the SAPK pathway, provides a mechanism for specific responses. There are many cellular targets of these kinases, including a variety of transcription factors that are involved in cellular growth responses.

the slow inositol pathway (inositol 1,4,5-trisphosphate [IP3]) predominates, initiated either by receptor tyrosine kinases or by G-protein-coupled receptors of the seven transmembrane domain class. Receptor tyrosine kinases stimulate phospholipase Cγ to produce IP3, which acts as a second messenger to trigger the release of Ca2+ from the endoplasmic reticulum. The G-protein-coupled receptors can increase intracellular Ca2+ by activating phospholipase Cβ. In excitable cells, voltage-dependent Ca2+ channels trigger more dramatic Ca2+ fluxes. Calcium binds to calmodulin (CaM), which serves as an intracellular sensor of calcium concentration. CaM binds to a number of enzymes (CaM kinases 1–5) protein phosphatases, and adenylate cyclases. The p21ras–MAPK signaling pathway is also activated by alterations in intracellular Ca2+.

CHAPTER 46 / MECHANISMS OF HORMONE ACTION

427

Figure 46-7 cAMP stimulation of transcriptional responses. G-protein-coupled receptors provide a major pathway for the generation of cAMP by adenylate cyclase. Increased cAMP levels within the cell cause dissociation of the tetrameric protein kinase A (PKA) holoenzyme leading to the release of the active catalytic subunit of PKA. The active catalytic subunit is translocated to the nucleus, where it phosphorylates transcription factor CREB, which binds as a homodimer to specific cAMP response elements (CREs) on target genes. The phosphorylation of CREB induces the binding of the coactivator, CREB-binding protein (CBP), which is also a substrate for phosphorylation. These transcription factors act by interacting with basal transcription factors such as TFIIB and TFIID. Several other kinases may phosphorylate CREB, including the calcium-regulated kinases.

Cyclic AMP response element binding protein (CREB) has been implicated as one of several transcription factors that can be activated by alterations in intracellular calcium. Ca2+-induced phosphorylation of CREB requires a novel p21ras-dependent 105-kDa kinase. This novel kinase leads to phosphorylation of CREB at Ser 133, and thereby enhances transactivation function (see below). Activation of these transcription factors by Ca2+ leads to the induction of temporally distinct immediate-early and delayed response genes.

cAMP-DEPENDENT SIGNALING cAMP-dependent signaling cascades are typically initiated by the binding of peptide hormones to their membrane receptors. However, intracellular cAMP concentrations are also modulated by processes intrinsic to the cell, such as the cell cycle. The pleiotropic array of cellular effects of cAMP are primarily mediated by protein kinase A, which acts on a number of cellular substrates, including enzymes, the cytoskeleton, and transcription factors. As depicted in Fig. 46-7, cAMP binds to a regulatory subunit of protein kinase A, leading to dissociation of an active catalytic subunit that phosphorylates specific substrates. In addition to receptors, ion channels, cytoskeletal proteins, and enzymes, a group of transcription factors are important targets of the protein kinase A pathway. These include the CREBs and activating transcription factors (ATFs) that are members of the B-Zip class of transcription factors. Posttranslational modification of CREB by phosphorylation at serine residues induces conformational changes that alter the affinity of CREB for coactivator proteins, such as CREB-binding protein (CBP) or the TATA boxbinding protein coactivator TAFII 110. CBP is thought to form a bridge between CREB and the basal transcription apparatus

(Fig. 46-7). CBP also interacts with other B-Zip proteins, such as c-Jun and c-Fos, other transcription factors, including c-Myb, as well as specific kinases. In addition to a clear role in cAMP signaling, CBP has also been implicated in mitogenic signaling. Consistent with its role in multiple cell signaling pathways, mutations of CBP cause Rubinstein-Tabi syndrome (OMIM 180849), which is associated with mental retardation, multiple congenital anomalies, and predisposition to malignancy. The B-Zip dimerization structure of the CREB and ATF proteins provides the basis for numerous combinations of different members of this family. The closely related cAMP response element modulator (CREM) gene product can act as a transdominant negative regulator of CREB transcriptional activity, either by forming inactive heterodimers with CREB or by binding to the CRE as an inactive complex. In addition to the canonical CRE, several other DNA regulatory elements and transcription factors are capable of stimulating gene transcription in response to cAMP. The transcription factor, activator protein 2 (AP-2), also functions in basal, phorbol ester, and cAMP-mediated transcriptional induction. The CAAT enhancer binding protein (C/EBP) induces transcription of the adipocyte protein 2 gene promoter (aP2 gene) and stearoyl acyl CoA desaturase (SCD-1) through a region that is also regulated by cAMP. cAMP-responsive regions of several genes appear to overlap with binding sites for nuclear receptors, some of which may also transduce cAMP effects.

CELL-CYCLE REGULATORY SIGNAL TRANSDUCTION PATHWAYS Phosphorylation plays an essential regulatory role in the cell cycle, and a large array of cyclin-dependent kinases (CDKs) have been identified (see Chapter 6). The regulatory subunits of the

428

SECTION VII / ENDOCRINOLOGY

CDKs, known as cyclins, form complexes with their catalytic partners, to function as heterodimeric holoenzymes that phosphorylate specific proteins, including the tumor suppressor protein pRB (see Chapter 105). Phosphorylation of pRB blocks its critical inhibitory function, allowing cell-cycle progression and differentiation to occur. Several proteins capable of binding cyclins and inhibiting CDK activity have been identified and are referred to as CDKIs. The CDKIs inhibit cell-cycle progression and inhibit tumor formation. Translocation of the cyclin D1 gene has been associated with certain cases of hyperparathyroidism (see Chapter 51), and actually led to the identification of cyclin D1 in humans. In these cases, a somatic inversion of chromosome 11 brings cyclin D1 under the control of the PTH promoter. Consequently, cyclin D1 is overexpressed in the parathyroid cell and results in cellular proliferation.

NUCLEAR RECEPTORS OVERVIEW OF NUCLEAR RECEPTOR ACTION The nuclear receptor superfamily consists of several hundred structurally related proteins that function by increasing or decreasing the rate of gene transcription (see Chapter 3). The nuclear receptors mediate the physiologic actions of small cell-permeable hormones, such as the sex steroid molecules (estrogen, testosterone, and progesterone), cortisol, aldosterone, and vitamin D, as well as retinoids that are derived from dietary vitamin A. There is also a large subfamily of nuclear receptors for which no ligand has been identified, or which may not require the binding of a ligand for functional activation. This group of receptors are referred to as orphan nuclear receptors. Nuclear receptors interact with specific DNA sequences, known as hormone response elements (HREs), that are typically located in the promoter regions of target genes. After the receptor binds to DNA, it is positioned to interact with other transcription factors to alter rates of gene transcription. STRUCTURE AND CLASSIFICATION OF NUCLEAR RECEPTORS Although nuclear receptors can be classified based on their ligands, structural similarities and DNA recognition sites provide a particularly useful classification system because it allows predictions to be made for receptor members that have not been characterized extensively. The DNA-binding domain (DBD) is the most highly conserved region of nuclear receptors (Fig. 46-8). The DBD consists of two centrally located zinc fingers, the structure of which has been well characterized using X-ray crystallography and nuclear magnetic resonance studies. The amino acid sequence at the carboxyterminal base of the first zinc finger is the primary determinant of DNA binding specificity, and provides a convenient means for classifying subfamilies of receptors that bind to related DNA elements. Other domains of nuclear receptors have also been clearly delineated. The ligand-binding domain (LBD) is located in the carboxyterminus of the receptor. This region of nuclear receptors has several overlapping functional domains, including regions required for dimerization and for transcriptional activation and repression. The structure of the ligand-binding domain has also been solved by X-ray crystallography, and several lines of evidence indicate that ligand binding induces significant conformational changes in the receptor. The amino-terminal region of nuclear receptors is the most variable and the least characterized. In some cases, it has been shown to contain additional transcription-activating domains. Nuclear receptors can generally be divided into two groups (Table 46-3). The classic steroid receptors share a similar DNA-

binding domain and DNA recognition sequence. This group includes receptors for glucocorticoids, mineralocorticoid, progesterone, and testosterone. Of note, even though the estrogen receptor (ER) binds to a steroid hormone, it shares more in common with the second group of nuclear receptors (see below). The steroid receptors often reside in the cytoplasm of the cell, complexed with proteins of the heat shock family. After ligand binds to the receptors, they disassociate from the heat shock proteins, translocate to the nucleus, and bind to hormone response elements in target genes. The steroid group of receptors bind to DNA only as homodimers, and their DNA binding sites reflect this twofold symmetry by being configured as palindromic pairs of monomeric “half-sites.” A second major group of nuclear receptors includes thyroid hormone receptor (TR), retinoic acid receptor (RAR), ER, and many others that share a similar DNA recognition sequence in the first zinc finger, and bind to a related DNA sequence, AGGTCA. This group of receptors also shows great flexibility in terms of dimerization. Many of these receptors can bind to DNA as monomers (particularly orphan receptors), but several bind to DNA as heterodimers as well as homodimers. The heterodimer partners typically include other members of this receptor subfamily. The retinoid X receptors (RXRs) serve as heterodimeric partners for several different nuclear receptors, including RARs, TRs, vitamin D receptor (VDR), and peroxisome proliferator-activating receptors (PPARs). The heterodimers typically bind to hormone response elements that are arranged as direct repeats rather than in a palindromic manner. The spacing between these half-sites, and variations in some of the contextual bases, provide specificity to different receptors, providing a partial explanation for how so many different receptors can share a common binding motif but elicit distinct cellular responses. Most members of this subfamily bind nonsteroid molecules and are located in the nucleus before ligand binding. They are able to bind to DNA in the absence of ligand, often with specific regulatory effects (e.g., gene silencing). Thus, the regulation of this group of receptors represents a complex interplay of receptor dimers, interactions with different DNA elements, and binding by specific ligands. A subject of great interest is whether orphan receptors function without ligands, or whether the ligand remains to be identified. The RXRs were initially classified as orphan receptors, but are now known to bind 9-cis retinoic acid. The PPARγ receptors have been found to bind a variety of eicosanoids, as well as thiazolidinediones that are used to treat non–insulin-dependent diabetes. Other orphan receptors may be regulated by posttranslational modifications such as phosphorylation. There has been rapid progress toward defining the mechanism by which nuclear receptors influence gene transcription. Several of the nuclear receptors have been found to physically interact with the basal transcription factor, TFIIB. Using in vitro transcription analysis, the nuclear receptors have been shown to facilitate the rate of assembly and stabilize basal transcription factors in the preinitiation complex that forms at the TATA box. An additional level of transcriptional control occurs through nuclear receptor interactions with “coactivator” or “corepressor” proteins, which in turn interact with proteins of the basal transcription factor complex. Several of these proteins have been identified on the basis of their ability to interact with nuclear receptors, but their role in transcriptional modulation is still being defined. Coactivators and corepressors may contribute to the development of specific diseases. Recently, a region of chromosome 20q that is often ampli-

CHAPTER 46 / MECHANISMS OF HORMONE ACTION

429

Figure 46-8 Nuclear hormone receptors. The nuclear receptors contain conserved modular functional domains. The central DNA-binding domain (C) is the most highly conserved region. Characteristic zinc finger domains are depicted. The carboxy-terminal region (E/F) contains several functional domains, including ligand binding, dimerization, and transcriptional activation region. A region (D) between the ligandbinding domain and the DNA-binding domain mediates nuclear localization and binds transcriptional corepressors. In most receptors, the amino-terminal domain (A/B) contains motifs involved in transcriptional activation. Table 46-3 Ligands and Antagonists for Nuclear Receptors Receptor

Ligand

Antagonists

DNA response element

P box

Glucocorticoid subfamily GR PR MR AR

Cortisol Progesterone Aldosterone, cortisol Dihydrotestosterone

RU-486 RU-486 Spironolactone Flutamide, cyproterone acetate

TGAACATGTTCT TGAACATGTTCT TGAACATGTTCT TGAACATGTTCT

GSCKV GSCKV GSCKV GSCKV

ER/TR Family ER

Estrogen

Tamoxifen, clomiphene, ICI 164, 384 None None Ro 41-5253 None None

AGGTCAnnnTGACCT

EGCKA

AGGTCAnnnnAGGTCA AGGTCAnnnAGGTCA AGGTCAnnnnnAGGTCA AGGTCAnAGGTCA AGGTCAnAGGTCA

EGCKG EGCKG EGCKG EGCKG EGCKG

TR VDR RAR RXR PPARγ

Triiodothyronine (T3) 1,25-Dihydroxyvitamin D All-trans retinoic acid 9-cis Retinoic acid Prostaglandin J2, thiazolidinedione

PR, progesterone receptor; GR, glucocorticoid receptor; MR, mineralocorticoid receptor; AR, androgen receptor; ER, estrogen receptor; TR, thyroid hormone receptor; VDR, vitamin D receptor; RAR, retinoic acid receptor; RXR, 9-cis retinoic acid receptor; PPAR, peroxisome proliferator activated receptor. The P-box refers to a five-amino acid sequence (in single letter code) at the base of the first zinc finger in the DNA-binding domain of the nuclear receptors. The half-sites refer to a consensus nucleotide sequence recognized by the DNA-binding domains of the receptor subfamilies.

fied in breast cancers was shown to encode a steroid receptor coactivator, AIB1, that enhances estrogen-dependent transcription and is overexpressed in ~65% of breast cancers. The CBP gene is also translocated in a subset of treatment-induced cases of acute myelogenous leukemias. Another important aspect of nuclear receptor action is their ability to crosstalk with other regulatory pathways. For example, growth factor pathways can alter the function of some nuclear receptors, such as the ER and progesterone receptor (PR). Certain nuclear receptors, including the glucocorticoid receptor (GR), TR, and RAR, interact with c-Jun or c-Fos, resulting in mutual antagonism. There is also evidence for nuclear receptor interactions with CBP, providing another means to inhibit the pathways (e.g., CREB, c-Jun, c-Fos) that converge on the CPB integrator protein. Finally, several nuclear receptors have been found to alter the conformation of DNA and associated chromatin, and in this manner may be able to alter the access of other transcription factors to the regulatory regions of genes. PHYSIOLOGIC EFFECTS OF NUCLEAR RECEPTORS Nuclear receptors serve an important role in development, cell differentiation, metabolism, and reproduction. Some of these func-

tions will be described briefly to provide examples of how nuclear receptors impact on physiology. The orphan receptor, steroidogenic factor-1 (SF-1) provides a good example of a receptor that plays a critical developmental role. SF-1 is selectively expressed in the adrenal gland and in reproductive tissues. Targeted disruption of SF-1 has been shown to prevent adrenal development and causes dysfunction at multiple levels of the reproductive axis. SF-1 appears to play dual role in cell viability and the direct regulation of a variety of steroidogenic enzyme genes. The glucocorticoid receptor mediates the widespread effects of cortisol on blood pressure, blood glucose levels and insulin action, neuropsychiatric status, diurnal variation in body temperature, and immune responses. The closely related mineralocorticoid receptor binds aldosterone and regulates the renal handling of potassium and sodium. The sex steroid receptors determine the phenotypic appearance of secondary sexual characteristics and control the production of oocytes and sperm. Testosterone itself binds relatively weakly to the androgen receptor, but it is converted to the more active metabolite, dehydrotestosterone, by the enzyme 5α-reductase,

430

SECTION VII / ENDOCRINOLOGY

Table 46-4 Disorders Caused by Mutations in Nuclear Receptors Receptor

Disorder

Androgen receptor

Androgen insensitivity

Androgen receptor Estrogen receptor Estrogen receptor Glucocorticoid receptor Thyroid hormone receptor β Vitamin D receptor

Prostate cancer Estrogen resistance Breast cancer Glucocorticoid resistance Resistance to thyroid hormone Vitamin D-resistant rickets

Retinoic acid receptor

Promyelocytic leukemia

OMIM no.

Mechanism

313700

Premature stop codon, DNA-binding domain mutation Point mutations Point mutations, deletions ? Postreceptor Point mutations Ligand-binding mutations DNA-binding mutant, ligand-binding mutant Fusion protein lacking RAR DNA-binding domain

133430 138040 188570 277440 180240

Inheritance X-linked Somatic Autosomal recessive Somatic Autosomal dominant Autosomal dominant Autosomal recessive Somatic translocation

OMIM, Mendelian inheritance of man; RAR, retinoic acid receptor.

which is expressed at high levels in many androgen-responsive tissues. Therefore, tissue specific variation in 5α-reductase expression can influence the extent of androgen responsiveness. The reproductive nuclear receptors are also notable because of the existence of clinically useful receptor antagonists (Table 46-3). RU-486 binds with high affinity to the progesterone receptor ligand-binding domain (it also binds the glucocorticoid receptor) in such a manner that its transcriptional activating properties are blocked. Similarly, tamoxifen binds to the estrogen receptor ligand-binding domain but fails to convert the receptor to a transcriptionally active form. The development of specific receptor antagonists for reproductive and other nuclear receptors is an active area of investigation, and promises to provide a plethora of useful new drugs and experimental tools. The nonsteroid nuclear receptors (e.g., thyroid hormone, retinoid, vitamin D, and the orphan receptors) mediate a variety of developmental and metabolic pathways. In several cases, multiple receptor isoforms provide an additional level of control. For example, there are two separate genes for TRs, three for all-trans RARs, and three for the retinoic 9-cis RXRs. In addition, many of these genes generate multiple receptor subtypes by virtue of alternate mRNA splicing or alternate promoter usage. The function of individual isoforms is not clear, but there is a complex pattern of overlapping tissue-specific expression. Receptor subtype expression is highly regulated during development, and there is striking tissue-specific expression of various nuclear receptor isoforms in adult organs. Therefore, like the peptide hormone receptors, tissue responsiveness to the hormone signal is regulated at its most fundamental level by selective expression of the nuclear receptor in responsive tissues. Nuclear receptors have been found to play a prominent role in human disease, many of which are described in detail in subsequent chapters (Table 46-4). The most well-described disorders are those of resistance to nuclear hormone action, caused by deletion or mutation of the receptors. Such hormone-resistance syndromes have been described for the estrogen, glucocorticoid, androgen, vitamin D, and thyroid hormone receptors. As an example, androgen resistance is inherited in an X-linked fashion, reflecting the X-chromosomal location of the androgen receptor. More than 50 complete or partially inactivating mutations of the androgen receptor have been described (see Chapter 56). In the fully resistant form (testicular feminization), 46,X,Y individuals have female secondary sexual features, but milder cases of resis-

tance can result in more subtle forms of feminization. Resistance to thyroid hormone is unusual because it is inherited in an autosomal dominant manner (see Chapter 50). In this disorder, the mutations are limited to the carboxy-terminal domain of the receptor, and result in loss of thyroid hormone binding or transactivation. However, because dimerization and DNA binding are preserved, the inactive mutant receptors can still bind to target genes and act as antagonists. This “dominant negative” property appears to explain the dominant pattern of inheritance. Another important example of nuclear receptor mutations causing human disease involves a translocation of the RAR receptor (t15; 17) in acute promyelocytic leukemia in which the RAR-α recombines with a putative tumor suppressor, PML. Treatment with high doses of retinoids causes the maturation of the leukemic cells and can result in clinical remission. Somatic mutations of other receptors (estrogen, androgen) are thought to play a role in hormonally responsive tumors, such as breast and prostate cancer. Recently, environmental toxins, such as DDT and halogenated aromatic hydrocarbons, have been found to interact with the nuclear receptors for androgens and estrogen. Finally, our knowledge of nuclear receptor biology will eventually allow a better understanding of conditions of nuclear hormone excess and deficiency states, such as Cushing’s disease, adrenal insufficiency, hyperthyroidism, and hypothyroidism.

SELECTED REFERENCES Anzick SL, Kononen J, Walker RL, et al. AIB1, a steroid receptor coactivator amplified in breast and ovarian cancer. Science 1997;277: 965–968. Boguski MS, McCormick F. Proteins regulating Ras and its relatives. Nature 1993;366:643–654. Chao MV. Neurotrophin receptors: a window into neuronal differentiation. Neuron 1992;9:583–593. Clapham DE. Calcium signaling. Cell 1995;80:259–268. Cobb MH, Goldsmith EJ. How MAP kinases are regulated. J Biol Chem 1995;270:14,843–14,846. Collu R, Tang J, Castagne J, et al. A novel mechanism for isolated central hypothyroidism: inactivating mutations in the thyrotropin-releasing hormone receptor gene. J Clin Endocrinol Metab 1997;82:1561–1565. Conklin BR, Bourne HR. Structural elements of G alpha subunits that interact with G beta gamma, receptors, and effectors. Cell 1993;73: 631–641. Gammeltoft S, Kahn CR. Hormone signaling via membrane receptors. In: DeGroot LJ, ed. Endocrinology. Philadelphia: WB Saunders, 1995; pp. 17–65. Ghosh A, Greenberg ME. Calcium signaling in neurons: molecular mechanisms and cellular consequences. Science 1995;268:239–247.

CHAPTER 46 / MECHANISMS OF HORMONE ACTION

Glass CK, Rose DW, Rosenfeld MG. Nuclear receptor coactivators. Curr Opin Cell Biol 1997;9:222–232. Habener JF. Cyclic AMP second messenger signaling pathway. In: DeGroot LJ, ed. Endocrinology. Philadelphia: WB Saunders, 1995; pp. 77–92. Hahn SA, Schutte M, Hoque AT, et al. DPC4, a candidate tumor suppressor gene at human chromosome 18q21.1. Science 1996;271: 350–353. Heldin CH. Dimerization of cell surface receptors in signal transduction. Cell 1995;80:213–223. Hunter T. Protein kinases and phosphatases: the yin and yang of protein phosphorylation and signaling. Cell 1995;80:225–236. Ihle JN. Cytokine receptor signalling. Nature 1995;377:591–594. Ihle JN. STATs: signal transducers and activators of transcription. Cell 1996;84:331–334. Insel PA. Seminars in medicine of the Beth Israel Hospital, Boston. Adrenergic receptors—evolving concepts and clinical implications. N Engl J Med 1996;334:580–585. Jameson JL. Applications of molecular biology in endocrinology. In: DeGroot LJ, ed. Endocrinology. Philadelphia: WB Saunders, 1995; pp. 119–147. Jameson JL. Principles of hormone action. In: Weatherall DJ, Ledingham JGG, Warrell DA, eds. Oxford Textbook of Medicine. Oxford: Oxford Medical Publishers, 1996; pp. 1553–1573. Josso N, di Clemente N. Serine/threonine kinase receptors and ligands. Curr Opin Genet Dev 1997;7:371–377. Karin M. The regulation of AP-1 activity by mitogen-activated protein kinases. J Biol Chem 1995;270:16,483–16,486. Kishimoto T, Taga T, Akira S. Cytokine signal transduction. Cell 1994;76:253–262. Malarkey K, Belham CM, Paul A, et al. The regulation of tyrosine kinase signalling pathways by growth factor and G-protein-coupled receptors. Biochem J 1995;309:361–375. Mangelsdorf DJ, Thummel C, Beato M, et al. The nuclear receptor superfamily: the second decade. Cell 1995;83:835–839. Marshall MS. Ras target proteins in eukaryotic cells. FASEB J 1995; 9:1311–1318.

431

Marshall CJ. Specificity of receptor tyrosine kinase signaling: transient versus sustained extracellular signal-regulated kinase activation. Cell 1995;80:179–185. Massague J. Receptors for the TGF-beta family. Cell 1992;69:1067–1070. Massague J, Weis-Garcia F. Serine/threonine kinase receptors: mediators of transforming growth factor beta family signals. Cancer Surv 1996;27:41–64. Mathews LS. Activin receptors and cellular signaling by the receptor serine kinase family. Endocr Rev 1994;15:310–325. McDonald NQ, Chao MV. Structural determinants of neurotrophin action. J Biol Chem 1995;270:19,669–19,672. Pestell RG, Jameson JL. Transcriptional regulation of endocrine genes by second messenger signalling pathways. In: Weintraub BD, ed. Molecular Endocrinology: Basic Concepts and Clinical Correlations. New York: Raven, 1995; pp. 59–76. Post GR, Brown JH. G protein-coupled receptors and signaling pathways regulating growth responses. FASEB J 1996;10:741–749. Saltiel AR. Diverse signaling pathways in the cellular actions of insulin. Am J Physiol 1996;270:E375–E385. Schlessinger J. How receptor tyrosine kinases activate Ras. Trends Biochem Sci 1993;18:273–275. Spiegel AM. Defects in G protein-coupled signal transduction in human disease. Annu Rev Physiol 1996;58:143–170. Spiegel AM, Weinstein LS, Shenker A. Abnormalities in G proteincoupled signal transduction pathways in human disease. J Clin Invest 1993;92:1119–1125. Tsai MJ, O’Malley BW. Molecular mechanisms of action of steroid/thyroid receptor superfamily members. Annu Rev Biochem 1994;63:451–486. van Biesen T, Luttrell LM, Hawes BE, Lefkowitz RJ. Mitogenic signaling via G protein-coupled receptors. Endocr Rev 1996;17:698–714. Van Sande J, Parma J, Tonacchera M, Swillens S, Dumont J, Vassart G. Somatic and germline mutations of the TSH receptor gene in thyroid diseases. J Clin Endocrinol Metab 1995;80:2577–2585. Wilks AF, Oates AC. The JAK/STAT pathway. Cancer Surv 1996;27: 139–163. Woodgett JR, Avruch J, Kyriakis J. The stress activated protein kinase pathway. Cancer Surv 1996;27:127–138.

CHAPTER 47 / DIABETES MELLITUS

47

433

Diabetes Mellitus WILLIAM L. LOWE, JR.

INTRODUCTION

GENETICS OF TYPE 1 DIABETES MELLITUS

Diabetes mellitus affects approximately 5% of the general population with its prevalence varying between ethnic groups and geographic regions. The majority of cases are accounted for by two different types of diabetes, type 1 and 2, which account for approximately 10 and 90% of cases of diabetes, respectively. Although these two disorders share a common phenotype, fasting and postprandial hyperglycemia, their etiology is distinct. Type 1 diabetes is characterized by pancreatic β-cell deficiency with a resulting absolute deficiency of insulin. The β-cell deficiency is most commonly secondary to autoimmune-mediated destruction. Type 2 diabetes, in contrast, is characterized by a deficiency of insulin action as a result of a combination of insulin resistance and β-cell dysfunction that is manifest as inadequate insulin secretion in the face of insulin resistance and hyperglycemia. The familial clustering of both type 1 and 2 diabetes has long suggested a genetic contribution to the origin of the diseases. In the case of type 1 diabetes, the concordance rate among monozygotic twins is 25–50%, whereas the relative risk to an individual with a first-degree relative with type 1 diabetes is approximately 5–10% compared with the 0.4% risk in the general population. Although these data are consistent with a genetic contribution to the origin of the disease, the lack of 100% concordance in monozygotic twins suggests that environmental factors also make a significant contribution to the pathogenesis of the disease. These environmental factors have not been clearly defined, but seasonal variation in the onset of type 1 diabetes and its association with preceding episodes of specific viral infections suggest that viruses may be one of the important environmental factors. Familial clustering of disease is much more apparent in type 2 diabetes. The concordance rate among monozygotic twins has been shown to be 50–95%, whereas approximately 40% of siblings and approximately 30% of offspring of affected individuals develop either type 2 diabetes or impaired glucose tolerance. Again, these data are consistent with a significant genetic contribution to the development of type 2 diabetes but suggest that environmental influences are also important in its etiology. With the advent of molecular genetics, significant progress has been made in defining the genetics of rare, monogenic forms of diabetes as well as more typical type 1 and 2 diabetes. This progress is reflected in the recent reclassification of diabetes that includes diagnostic categories of genetic defects in β-cell function and insulin action. This chapter will describe the recent advances in the genetics of both type 1 and 2 diabetes as well as monogenic forms of diabetes.

Type 1 diabetes often presents abruptly with marked hyperglycemia, polyuria, and ketoacidosis, which occur as a result of insulin deficiency. This dramatic presentation of the disease suggested initially that type 1 diabetes was the result of an acute event, but with an improved understanding of the pathogenesis of the disease, it is now clear that onset of the disease is a chronic process that can be divided into a series of stages. As will be discussed below, individuals have a genetic predisposition for diabetes. In certain of these individuals, an autoimmune process is initiated, presumably in response to some environmental exposure. There is then a period of active autoimmunity characterized by the presence of autoantibodies and progressive β-cell destruction but with maintenance of normal blood sugars and glucose tolerance. With sufficient β-cell destruction, impaired glucose tolerance, which is typically not clinically evident, and, subsequently, diabetes develop. Depending on the rate of decline of β-cell function, older patients may be presumed to have type 2 diabetes mellitus because of residual insulin secretion, albeit not sufficient insulin secretion to maintain euglycemia. As destruction of the residual β cells continues, a new stress, e.g., infection, often results in the acute presentation of diabetes. Ultimately, absolute insulin dependence develops because of total or near total β-cell destruction. The autoimmune-mediated destruction of pancreatic β cells is characterized by two features, autoantibodies and insulitis. Autoantibodies present in type 1 diabetes mellitus are directed against a variety of β-cell antigens, including insulin, glutamic acid decarboxylase (GAD65 and 67), membrane proteins that are homologous to tyrosine phosphatases (ICA512 and IA-2), and islet neuroendocrine ganglioside. Although the presence of two or more of these antibodies is predictive of progression to diabetes in relatives of affected individuals, their role in the immune-mediated destruction of the islets cells is still unclear. Insulitis is characterized by inflammatory infiltrates in the islets consisting primarily of CD8 cells but also of CD4 cells, B cells, macrophages, and natural killer cells. The cause of type 1 diabetes mellitus, and, thus, the trigger for the autoimmune process, is complex and involves both a genetic predisposition and environmental factors. The genetics of type 1 diabetes mellitus are now being defined. Type 1 diabetes does not follow a simple Mendelian pattern of inheritance. As demonstrated by studies in identical twins, there is not a 100% concordance between a susceptible genotype and disease presence. Presumably the penetrance of the disease genes is influenced by environmental factors that are, as yet, undefined. Moreover, studies in both mice and now humans clearly demonstrate that the disease is polygenic, suggesting that a sufficient

From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

433

434

SECTION VII / ENDOCRINOLOGY

complement of genes must be inherited to confer susceptibility to diabetes. Two approaches have been used to define genes that predispose to type 1 diabetes—identifying candidate genes by comparing the frequency of alleles of specific genes in diabetic and control populations (case-control or association studies) and genome scanning to identify chromosomal loci associated with disease susceptibility. Given the autoimmune origin of type 1 diabetes, the major histocompatibility locus (MHC) or human leukocyte antigen (HLA) region was examined initially as a candidate susceptibility locus. Association studies identified this region as a potential susceptibility locus, and this was confirmed in subsequent linkage studies. This locus is now referred to as IDDM1. The degree of family clustering of a disease is referred to as λs and can be estimated by comparing the disease risk in siblings of affected individuals to the prevalence of the disease in the general population. For type 1 diabetes, the λs is 15, whereas the λs for IDDM1 is estimated to be 2.6, suggesting that IDDM1 accounts for 35–40% of the familial inheritance of the disease. Progress has been made in defining how IDDM1 confers susceptibility to disease. The MHC is located on the short arm of chromosome 6 (6p21) and encodes proteins involved in the regulation of the immune process (see Chapter 31). The MHC consists of three major regions, A, B, and C, that encode class I genes, and the D region, which encodes class II genes. The class I molecules are highly polymorphic and present peptide fragments of foreign antigens to cytotoxic T lymphocytes. The class II molecules present foreign processed antigen to helper T cells and, thus, are involved in initiating the immune response. Class II molecules consist of an α and β chain that are encoded by different genes. There are two major classes of class II genes, the DR and DQ genes. The β but not the α chain in the DR molecules is polymorphic, whereas both the α and β chains are polymorphic in DQ molecules. Initially an association between class I alleles and type 1 diabetes was demonstrated, but it is now clear that class II molecules are more important in conferring risk. The original association with class I molecules was likely caused by the nonrandom association of class I alleles with the class D alleles (linkage disequilibrium). Moreover, linkage disequilibrium of the DQ and DR alleles has created extended haplotypes (the presence of several genes on the same chromosome). Some of these class II alleles predict susceptibility to type 1 diabetes, whereas others either confer protection or are neutral. It is likely that both DR and DQ alleles are important in regulating susceptibility to type 1 diabetes. Interestingly, all of the DQ β chain alleles encoding a gene product with an aspartic acid at position 57 are either protective or neutral. In contrast, alleles with alanine or valine at position 57 confer susceptibility, whereas the one allele with serine at position 57 confers weak susceptibility. Position 57 of the DQ β chain is not, however, the sole determinant of risk; susceptibility and protection are also conferred by other residues in the DQ β chain and by different alleles of the other chains. The biologic consequences of these and other amino acid differences are still being defined, but changes in peptide-binding sites of these class II molecules may alter the specificity of the immune response to foreign or self-antigens by affecting the binding affinity of different peptide antigens for the class II molecules. A second susceptibility locus that has been identified using a candidate gene approach is the insulin gene locus on chromosome 11p15.5. Linkage of this region, which is referred to as IDDM2, has also been demonstrated to type 1 diabetes. Among the polymorphisms in this region is the variable number of tandem repeats (VNTR) locus in the 5'-flanking region of the insulin gene. The

Table 47-1 Susceptibility Loci for Type 1 Diabetes Mellitus Susceptibility locus IDDM1a IDDM2a IDDM3 IDDM4a IDDM5a IDDM7 IDDM8a IDDM11 IDDM12a IDDM13

Chromosome 6p21 11p15 15q26 11q13 6q25 2q31 6q27 14q24.3–q31 2q33 2q34 7q

Linked markers HLA-DQB1, -DRB1 Insulin VNTR … FGF3, D11S1337 D6S476-ESR-D6S448 D2S152 D6S281 D14S67 CTLA-4 IGFBP-2, -5 GCK (glucokinase)

λs 2.60 1.29 … 1.07 1.16 1.13 1.42 … … … …

aLinkage

confirmed in independent studies using independent data sets. Adapted from Todd and Farrall (1996) and Cordell and Todd (1995).

VNTR consists of tandem repeats of a 14- to 15-bp core sequence. Three different alleles have been identified. The class I allele contains 34 to 45 repeats and is associated with a two- to fivefold increased risk of type 1 diabetes. The class III allele contains 141 to 209 repeats and has a dominant protective effect. The class II allele has approximately 80 repeats but is rare, and its association with disease susceptibility has not been defined. The λs for IDDM2 is 1.3, and it accounts for 10% of the familial inheritance of type 1 diabetes. IDDM2 is located within a 50-kbp region of chromosome 11 that contains the genes encoding tyrosine hydroxylase, insulin, and insulin-like growth factor-II (IGF-II). Multiple polymorphisms are present within this region, but an extensive series of genetic analyses mapped the IDDM2 susceptibility locus to the VNTR in the 5'flanking region of the insulin gene, suggesting that the variable number of repeats in some manner confers susceptibility to disease, perhaps because of effects on insulin gene expression. In vitro studies using the VNTR in reporter gene constructs have demonstrated that the VNTR modulates insulin gene expression, although conflicting data were obtained regarding the effect of the class III compared with class I allele on gene expression. In vivo studies have demonstrated that the class III allele is associated with decreased pancreatic insulin gene expression compared with transcription from the class I allele. A mechanism by which the VNTR could affect disease susceptibility via its effects on pancreatic expression of the insulin gene was not obvious. Subsequent studies, however, have examined effects of the VNTR on prenatal and postnatal expression of the insulin gene in the human thymus. Interestingly, the class III allele is associated with an approximately 2.5-fold increase in insulin gene expression in the thymus compared with the class I allele. Similarly, increased levels of proinsulin are present in thymi from class III/I heterozygotes compared with class I/I homozygotes. Increased expression of insulin in the thymus might facilitate tolerance to insulin by promoting deletion of insulin-specific T lymphocytes, and, thus, provide protection from type 1 diabetes. Because IDDM1 and IDDM2 account for only approximately 50% of the familial clustering of type 1 diabetes, additional genes must be involved. To identify these genes, genome scans have been performed using DNA from sibling pairs affected by type 1 diabetes. The aim of genome scans using affected sib pairs is to scan the entire genome with a collection of genetic markers, calculate the degree of allele sharing at each marker or locus, and identify those chromosomal loci in which allele sharing by the affected siblings occurs with greater frequency than expected,

CHAPTER 47 / DIABETES MELLITUS

435

Figure 47-1 Schematic model of the origin of type 2 diabetes mellitus. Type 2 diabetes is secondary to both insulin resistance and inadequate insulin secretion secondary to β-cell dysfunction. With sufficient β-cell function, euglycemia or impaired glucose tolerance is maintained in the presence of insulin resistance at the expense of hyperinsulinemia (compensated insulin resistance). With concomitant β-cell dysfunction, however, inadequate insulin secretion to compensate for the insulin resistance results in the onset of type 2 diabetes. Similarly, a primary defect in β-cell function may result in type 2 diabetes in the presence of some degree of insulin resistance. Both insulin resistance and β-cell dysfunction are influenced by genetic and environmental factors. Type 2 diabetes is multigenic, and its penetrance is secondary to the expression of several different genes, some of which are likely fixed and act independently of environmental factors. Other predisposing genes might be modifiable in that their expression or action is influenced by environmental factors. Interactions between genes are also likely to contribute to insulin resistance and β-cell dysfunction.

based on the frequency of the allele in the population. Increased sharing of alleles at a specific locus suggests that a gene(s) within that chromosomal locus contributes to the pathogenesis of the disease. Genome scans have now identified additional type 1 diabetes susceptibility loci (Table 47-1). Some of these loci have been confirmed by replication using additional families, whereas others await replication. None of these loci has an effect of the magnitude of IDDM1, but each may confer risk similar to that conferred by IDDM2. The genes at these loci that are responsible for conferring susceptibility to type 1 diabetes still await definition. As an increasing number of susceptibility loci are identified, an additional question that is raised is whether these different loci are acting epistatically. Epistasis, or interaction, between loci, indicates that the genotype at one locus affects the contribution of another locus. One implication of this is that loci that act epistatically may be acting through the same or related pathways. In contrast to epistasis is genetic heterogeneity, which indicates that two loci confer disease susceptibility independently, implying that the two loci confer risk via their effects on separate biologic pathways. This lack of interaction would suggest that an alteration in either pathway is associated with susceptibility to disease. IDDM1 and IDDM2 appear to act epistatically, whereas the action of IDDM1 and IDDM4 best fits a model of genetic heterogeneity. The potential interaction, or lack thereof, of the other susceptibility loci awaits further study. Definition of the genetic etiology of type 1 diabetes will have several clinical implications. Currently, the vast majority of affected individuals are identified subsequent to the onset of diabetes when nearly complete β-cell destruction has occurred. A relatively limited subset of individuals at risk for developing type 1 diabetes can be identified before the onset of disease based on a family history of type 1 diabetes and the presence of antibodies directed against islet cells. Not all such individuals develop type 1 diabetes, however, and the risk of developing diabetes in these individuals is defined by the degree of loss of first phase insulin

secretion during an intravenous glucose tolerance test. This loss of first phase insulin secretion occurs secondary to β-cell destruction and indicates activity of the autoimmune process. Currently, treatment of at risk individuals with insulin before the onset of altered glucose homeostasis is being tested as a means of preventing or delaying the onset of diabetes. On the basis of studies in animal models of diabetes, “vaccination” with insulin or a peptide that contains the epitope against which the majority of anti-insulin antibodies are directed is also being considered. Because a significant decrease in first phase insulin secretion indicates that marked β-cell destruction has already occurred, if susceptible individuals can be identified earlier, based on their complement of susceptibility genes, and screening can be extended beyond those with a positive family history, it may be possible to intervene earlier, before the onset of βcell destruction, and, thus, have a much greater impact on delaying or preventing the onset of type 1 diabetes. Second, as the pathways that are responsible for initiating the autoimmune process are identified, new therapies designed to interfere specifically with these pathways and prevent β-cell destruction can be developed. These therapies will likely be more efficacious in preventing the onset of diabetes than those that are currently available.

GENETICS OF TYPE 2 DIABETES MELLITUS Type 2 diabetes mellitus is a heterogeneous disorder that develops in response to both genetic and environmental factors (Fig. 47-1). In contrast to type 1 diabetes, type 2 diabetes is often diagnosed during routine screening by the detection of hyperglycemia or because of mild symptoms of hyperglycemia, e.g., polyuria. Much like type 1 diabetes, individuals with type 2 diabetes pass through a series of phases before the onset of diabetes. Initially, plasma insulin levels are increased because of insulin resistance, but euglycemia is maintained. In the second phase, postprandial hyperglycemia is present despite persistent hyperinsulinemia. Finally, insulin secretion declines in the face of persistent insulin resistance, which results in diabetes. Thus, affected individuals

436

SECTION VII / ENDOCRINOLOGY

demonstrate both insulin resistance, manifest as hyperinsulinemia and decreased insulin-stimulated glucose uptake into tissues, and abnormal β-cell function, manifest as altered glucose-induced insulin secretion. Obesity is an additional contributing factor to type 2 diabetes via its effects on insulin sensitivity. Obesity, in turn, is affected by both energy expenditure and intake. Energy expenditure is dependent to a large degree on resting metabolic rate, whereas energy intake is regulated by the central nervous system and behavioral regulation of eating and satiety. All of the above processes are in part heritable, but complex, and depend on a variety of different gene products, each of which may have the potential to contribute to the genetic predisposition to type 2 diabetes. Consistent with the heterogeneity of type 2 diabetes and the multiple contributing factors described above, mathematical modeling has suggested that type 2 diabetes is a polygenic disease. Consequently, onset of the disease likely requires the simultaneous presence of a subset of genes that affect the above processes. Because different subsets of genes are probably sufficient to confer susceptibility to type 2 diabetes, susceptibility genes likely vary between and, possibly, within populations. Moreover, environmental factors that have still not been fully defined contribute to the development of type 2 diabetes. Thus, disease susceptibility genes may be present in unaffected individuals because they lack a required complement of disease susceptibility genes or needed environmental factors to induce diabetes. This has and will continue to complicate attempts to define susceptibility genes for type 2 diabetes. An additional complication in defining the genetics of type 2 diabetes is that it is of late onset. Typically, parents, especially affected parents who may have succumbed to the complications of diabetes, are not available for study, and offspring are unlikely to be affected. Thus, family studies of type 2 diabetes are difficult. Finally, although it is clear that type 2 diabetes is polygenic and genetically heterogeneous, it is still unclear whether there are two or three major variants of the disease with additional minor variants or whether type 2 diabetes is truly heterogeneous with multiple variants, each of which accounts for a small percentage of affected individuals. Despite the above challenges, progress has been made in defining the genetics of late-onset type 2 diabetes. As with type 1 diabetes, the candidate gene approach and genome scanning have been used to identify genes that confer susceptibility to type 2 diabetes, but these studies have been complicated by all of the issues described above. Moreover, unlike IDDM1, which is a major susceptibility locus for type 1 diabetes, it is possible that there is not a major susceptibility gene for type 2 diabetes; rather, multiple genes that confer a limited degree of susceptibility may exist. GENETICALLY DEFINED FORMS OF TYPE 2 DIABETES Given the complexities of the genetics of type 2 diabetes, substantial effort has been applied using both candidate gene and positional cloning approaches to define the genetics of rare, but monogenic, forms of diabetes with the hope that this will provide insight into the genetics of type 2 diabetes. Mitochondrial Diabetes Mellitus The role of mutations in mitochondrial DNA as a cause of disease is becoming increasingly appreciated (see Chapter 103). Mitochondrial DNA is inherited maternally and encodes 13 polypeptide subunits involved in oxidative phosphorylation and the respiratory pathway, 22 transfer RNAs (tRNAs), and 2 ribosomal RNAs. Mitochondrial DNA is vulnerable to mutation because it is composed almost exclusively of coding sequences, lacks protection by histones, has inefficient repair mechanisms, and is exposed to reactive oxygen species produced during oxidative phosphorylation. Mitochondrial DNA undergoes mutation 5–10 times faster than nuclear DNA. The

possibility that mitochondrial DNA mutations might contribute to the pathogenesis of diabetes mellitus was suggested by the association of diabetes with several mitochondrial diseases and by the maternal inheritance of mitochondrial DNA, given the slight preference for maternal transmission of type 2 diabetes. Initially, a syndrome of diabetes mellitus and deafness caused by sensorineural hearing loss was identified in two pedigrees. In both, an A/G exchange at nucleotide 3243 in the mitochondrial tRNALeu(UUR) gene was noted. This particular mutation was located within a mitochondrial DNA-binding site for a protein that contributes to the termination of transcription at the boundary between the 16S ribosomal RNA and tRNALeu(UUR) genes. Thus, this mutation alters both tRNALeu(UUR) synthesis and mitochondrial protein synthesis in general. Interestingly, tRNALeu(UUR) is a hot spot for mutations; 10 disease-related mutations have been identified within this gene, 4 of which are associated with diabetes mellitus. Identification of these syndromes raised the question as to the role of mitochondrial DNA mutations in the pathogenesis of diabetes mellitus. A variety of studies have determined that of patients with type 1 and 2 diabetes mellitus, the A/G exchange in the tRNALeu(UUR) gene is present in 0.5–1.5% of patients. If only those patients with a family history of diabetes are considered, the prevalence of the mutation is two to five times higher. Characterization of affected individuals and their siblings demonstrated that 48% presented with diabetes and deafness, 21% had diabetes alone, 15% had deafness alone, 3% had deafness associated with some neurologic changes, and 13% had diabetes, deafness, and the findings of the MELAS syndrome (mitochondrial myopathy, encephalopathy, lactic acidosis, and strokelike episodes). The phenotypic variation is presumably caused by variable numbers of mitochondria containing normal compared with abnormal mitochondrial DNA (heteroplasmy) between tissues. Heteroplasmy results from unequal separation of mitochondrial populations during mitosis and differential accumulation of subsequent mitochondrial mutations. Although patients with mitochondrial DNA mutations were initially diagnosed as having either type 1 or 2 diabetes mellitus, there were differences between these patients and those with “typical” type 1 or 2 diabetes. Those diagnosed initially as having type 1 diabetes in general lacked anti–islet cell antibodies and did not typically have a history of diabetic ketoacidosis, whereas those initially diagnosed as having type 2 diabetes tended to be leaner and to be more likely to be treated with insulin compared with general populations with type 2 diabetes. Subsequent studies to define the etiology of diabetes in patients with mitochondrial mutations demonstrated decreased insulin secretion in response to an oral glucose tolerance test, whereas measures of glucose utilization demonstrated essentially normal insulin sensitivity. The above findings suggest that the primary defect in patients with mitochondrial mutations is in the pancreatic β cell. The pathway of glucose-induced insulin secretion is still being elucidated, but oxidative phosphorylation and ATP generation are thought to play an important role in insulin secretion (Fig. 47-2), so mutations in β-cell mitochondrial DNA would likely interfere with insulin secretion because of insufficient ATP generation. In patients with diabetes accompanied by hearing loss, a diagnosis of mitochondrial DNA mutations as the cause of the diabetes should be considered. Patients with these mutations should be advised that they will be much more likely to require insulin therapy because of the insulin deficiency that develops. Moreover, family members need to be carefully screened for both diabetes and evidence of sensorineural hearing. Frequently, the hearing

CHAPTER 47 / DIABETES MELLITUS

437

Figure 47-2 Schematic representation of glucose-stimulated insulin secretion by the pancreatic β cell. Glucose uptake is mediated by GLUT2, and glucose is phosphorylated by glucokinase to generate glucose-6-phosphate. Metabolism of glucose-6-phosphate via glycolysis yields pyruvate, which enters the tricarboxylic acid cycle in the mitochondria to generate ATP. The generation of ATP increases the ATP to ADP ratio in the cytoplasm, which inhibits activity of the ATP-sensitive K+ channels. The ATP-sensitive K+ channel is a complex of the sulfonylurea receptor (SUR) and an inwardly rectifying K+-channel protein (KIR6.2). Inhibition of these channels results in membrane depolarization and opening of voltage-dependent Ca2+ channels. The resulting increase in the intracellular concentration of Ca2+ stimulates insulin secretion. β-cell development and insulin production are regulated by a variety of factors, including several different transcription factors.

loss develops after the onset of diabetes, so those patients with diabetes need to be examined over time for hearing loss. Maturity Onset Diabetes of the Young (MODY) MODY is a subtype of diabetes that is monogenic and provides a model for studying the molecular genetics of type 2 diabetes. MODY is characterized by an early age of onset of diabetes and an autosomal dominant mode of inheritance that is based on the demonstration of a positive family history in three successive generations. Disease onset often occurs between 9 and 13 years of age and typically, although not universally, occurs before 25 years of age. Characterization of MODY patients from different kindreds revealed phenotypic differences between patients suggesting heterogeneity, which molecular genetic analyses confirmed. Linkage to polymorphisms in the glucokinase gene on chromosome 7p was demonstrated in some patients, whereas in other patients positional cloning techniques demonstrated linkage to markers on either chromosome 20q or 12q. Subsequent genetic analyses demonstrated that the kindred described as having MODY1 had a mutation in the hepatocyte nuclear factor (HNF)-4α gene on chromosome 20q, whereas kindreds with MODY3 had mutations in HNF-1α on chromosome 12q. Mutations in the glucokinase gene were shown to cause diabetes in kindreds with MODY2. Finally, a kindred was described recently in which early-onset type 2 diabetes (MODY4) cosegregated with an inactivating mutation in the insulin promoter factoral (IPF-1) gene. Mutations in these genes do not account for all cases of MODY, suggesting that additional MODY genes await description. The phenotypic differences between patients with the different forms of MODY provide some insight into the function of the responsible genes. Glucokinase is a key enzyme in glucose metabolism in β cells and hepatocytes and catalyzes the formation

of glucose-6-phosphate from glucose. Glucokinase participates in glucose sensing and links glucose to insulin secretion. In patients with glucokinase mutations, the severity of enzyme impairment is correlated to the degree of hyperglycemia. Because of the decreased sensitivity of the β cells to glucose in these patients, the dose–response curve for glucose-induced insulin secretion is shifted to the right, but increased insulin secretion is observed with increasing degrees of hyperglycemia. Whether altered glucose metabolism in hepatocytes contributes to the hyperglycemia is not known. Because of the underlying origin of the hyperglycemia, patients with MODY2 typically present early in life, usually before puberty, with mild fasting hyperglycemia that does not worsen over time. As noted, MODY1, MODY3, and MODY4 are caused by mutations in HNF-4α, HNF-1α, and IPF-1, respectively. HNF-4α is a transcription factor that is a member of the steroid–thyroid superfamily of nuclear receptors. Members of this receptor family are typically ligand-activated, but whether there is a ligand that activates HNF-4α is unknown. HNF-1α is a homeodomain transcription factor that functions either as a homodimer or as a heterodimer with the structurally related transcription factor HNF-1β. As their names imply, both of these factors are important for hepatic gene expression, but they are expressed in other tissues, including pancreatic islets. How mutations in these factors cause diabetes is unknown, although HNF-1α is known to be a weak trans-activator of the rat insulin I gene. HNF-4α is a positive regulator of HNF-1α expression and, thus, decreased HNF-1α production likely contributes to the cause of diabetes in patients with HNF-4α mutations. Little else is known about the function of HNF-1α and HNF-4α in pancreatic islets, but elucidation of their function will likely provide insight into islet cell development and function. IPF-1 is a transcription factor that regulates both pancreatic development and insulin

438

SECTION VII / ENDOCRINOLOGY

gene transcription. The other issue that has not been resolved is why MODY1, MODY3, and MODY4 are autosomal dominant diseases, implying that mutation of one allele results in disease. It is unclear whether the product of the mutant allele functions in a dominant negative fashion and inhibits activity of the product of the wild-type allele or whether a reduction in the amount of active transcription factor causes disease (haploinsufficiency). The phenotype of patients with MODY1 and MODY3 provides some insight into how mutations in HNF-1α and HNF-4α confer susceptibility to diabetes. The primary defect in patients with MODY1 and MODY3 is in the β cell. These patients have normal insulin sensitivity but demonstrate decreased insulin secretion in response to glucose. Unlike patients with MODY2, a progressive increase in insulin secretion with increasing hyperglycemia is not observed, suggesting a reduced capacity for insulin secretion. Similar defects, albeit not as severe, are present in individuals who have mutations in HNF-1α or HNF-4α but are not, as yet, diabetic. Thus, mutations in HNF-1α and HNF-4α appear to affect the regulation of insulin secretion or β-cell development. As noted, patients with MODY2 typically have mild fasting hyperglycemia and are much less susceptible to the microvascular and macrovascular complications of diabetes compared with patients with type 2 diabetes. Patients with MODY1 and MODY3, in contrast, exhibit the microvascular and macrovascular complications associated with diabetes, but do not have the phenotype commonly associated with late-onset type 2 diabetes, obesity, hypertension, and hypercholesterolemia. These observations raise the issue of the contribution of mutations in the MODY genes to the etiology of late-onset type 2 diabetes. Although approximately 5% of women with gestational diabetes mellitus and a first-degree relative with diabetes have mutations in the glucokinase gene, neither linkage studies nor mutational analyses suggest that mutations in glucokinase account for significant numbers of cases of late-onset type 2 diabetes. Mutations in HNF-4α appear to be rare, since only a single kindred with MODY1 has been described. Moreover, the mutation in HNF-4α associated with MODY1 was not present in a large population of Japanese with late-onset type 2 diabetes. Similarly, there is no evidence for linkage of the HNF-4α locus to type 2 diabetes. Further studies need to be done, but to date, there are no data to suggest that HNF-4α mutations account for a significant number of cases of late-onset type 2 diabetes. Mutations in HNF-1α appear to be more common. In studies of patients with early onset diabetes (3.0, the accepted threshold for linkage. Ten or more nonrelated nuclear families with affected pairs of siblings would be required to generate a similar score. Study of progeny of five consanguineous families of Saudi Arabian origin, using the homozygosity mapping approach, allowed placement of PHHI to a 5-cM interval on chromosome 11p14–15.1, between markers D11S1334 and D11S899 (Fig. 55-1).

513

514

SECTION VII / ENDOCRINOLOGY

Table 55-1 Differential Diagnosis of Hypoglycemia in Children Transient hypoglycemia Prematurity Small for gestational age Infants with systemic disease Infants of diabetic mothers Infants with erythroblastosis fetalis Persistent hypoglycemia Hyperinsulinemia Persistent hyperinsulinemic hypoglycemia of infancy β-Cell adenoma Beckwith-Wiedemann syndrome Leucine sensitivity Exogenous insulin abuse Sulfonylurea use Hormone deficiency Growth hormone Cortisol Glucagon Panhypopituitarism Inborn error of metabolism Glycogen storage diseases Fat metabolism disorders Amino acid metabolism disorders

In vitro studies of PHHI have suggested a defect of glucoseregulated insulin secretion in pancreatic islet β cells. Assignment of the PHHI locus to chromosome 11p excluded known genes involved in β-cell function, including the glucokinase, islet glucose transporter, and glucagon-like peptide-1 receptor loci, as candidates for PHHI. Current models of insulin secretion propose that, as a result of glucose metabolism, increases in the intracellular ATP/ADP ratio in pancreatic β cells inhibit ATP-sensitive potassium (KATP) channels leading to β-cell membrane depolarization, opening of voltage-gated calcium channels, and ultimately an increase in exocytosis of insulin (Fig. 55-2). Therefore, candidate genes for PHHI included those involved in the β-cell glucose sensing mechanism and insulin secretion. The high-affinity sulfonylurea receptor (SUR1), a subunit of the β-cell KATP channel, was considered as a candidate for the PHHI gene on the basis of its role as a modulator of insulin secretion. The human homolog of the SUR1 gene was localized to chromosome 11p15.1, the previously defined site of the PHHI locus, by fluorescent in situ hybridization, and this provided the impetus to begin mutational analysis. Because of the presence of two nucleotide-binding fold (NBF) consensus sequences, the SUR1 is classified as a member of the ATP-binding cassette superfamily. Contained within the NBF regions are highly conserved phosphate-binding loops, termed “Walker motifs,” that form intimate contact with the phosphates of bound nucleotide triphosphates. In other superfamily members, NBF-1 and NBF-2 have functional importance in the control of channel activity, through their interaction with cytosolic nucleotides. This class of molecules is involved in selective transport of substrate across the cell membrane against a concentration gradient using ATP hydrolysis as an energy source. Although each transporter is specific for a given substrate, the spectrum of substrates pumped by family members is diverse and includes amino acids, sugars, inorganic ions, polysaccharides, and peptides. The

Table 55-2 Characteristics of Persistent Hyperinsulinemic Hypoglycemia of Infancy Inappropriate elevation of serum insulin despite the presence of hypoglycemia Nonketotic hypoglycemia Increased glucose utilization rate (requires glucose infusion >15 mg · kg–1 · min–1 to maintain euglycemia) Glycemic response to glucagon

substrate for the SUR1 has not been elucidated. Other well-known members of this family include the multidrug resistance proteins, and the cystic fibrosis transmembrane conductance regulator (CFTR). Mutations in the CFTR have been shown to cause the autosomal recessive disorder cystic fibrosis, and the more frequent and severe disease alleles are located in the regions of the two NBFs. A search for mutations in the SUR1 gene in a group of Saudi Arabian individuals affected with PHHI revealed two separate point mutations in splice sites of the NBF-2 region (Fig. 55-3). The first of these mutations, G1400D(23)X, causes skipping of exon 35 of the SUR mRNA transcript with inclusion of a premature stop 24 codons later. The second mutation (3992–9 G→A) was identified in the 3' splice site sequence of exon 33, the first exon of the NBF-2 region. In the presence of this splice site mutation, three cryptic 3' splice sites were used in place of the wild-type splicing pattern, resulting in a 7-bp addition, a 20-bp deletion, or a 30-bp deletion to exon 33. Mutation of the NBF-1 region of SUR1, like disruption of the NBF-2 region, leads to the PHHI phenotype. Three mutations of the NBF-1 region of SUR1, including one point mutation that disrupts the Walker A motif (G716V), and two that alter RNA processing in the region (1672–20 A→G and 2552–1G→A), have been associated with familial PHHI. These additional mutations establish that both NBF regions of SUR1 are required for normal function, because mutation of either completely voids the ability of SUR1 to regulate insulin secretion and results in the PHHI phenotype. Two additional mutations have been described in SUR1 in those affected with PHHI. The first is an in-frame deletion within the NBF-2 region (∆F1388); the combination of the 3992–9 G→A and ∆F1388 mutations accounted for 88% of the PHHI chromosomes in the cohort of 25 Ashkenazi Jews studied. The second is a point mutation, again within the NBF2 region (G1479R) of SUR1, described in a single patient of unestablished ethnic origin; this mutation has been demonstrated to alter the sensitivity of SUR1 to MgADP. A summary of the published SUR1 mutations, depicting their relative locations within the NBF regions of the SUR1 gene, appears in Fig. 55-3. When expressed individually, SUR1 is unable to produce the β-cell ATP-sensitive potassium current IKATP. However, IKATP can be reconstituted by coexpression of the inward rectifier Kir6.2 with SUR1. Because the Kir6.2 locus is within 5 kb of the SUR1 gene on chromosome 11p15.1 and it is a necessary member of the β-cell KATP channel, we considered Kir6.2 as an additional candidate gene for PHHI. A homozygous point mutation (L147P), located in the conserved α-helical second transmembrane domain of Kir6.2, was identified in the genomic DNA of a severely affected child from a consanguineous Iranian family. No phenotypic dif-

CHAPTER 55 / HYPOGLYCEMIA WITH HYPERINSULINEMIA

515

Figure 55-1 Linkage of persistent hyperinsulinemic hypoglycemia of infancy (PHHI) to chromosome 11p15. (A) Linkage of PHHI to the interval between the chromosome 11 markers D11S1334 and D11S899 using the homozygosity gene mapping strategy. Shown are simplified pedigrees and genotypes in the region of homozygosity on chromosome 11 of members of two informative families affected with PHHI. Regions of homozygosity in affected children are boxed. In both families the parents are second cousins. Note that within each family the parents share the same disease allele haplotype, which presumably is derived from their common ancestor. (B) Summary of linkage data assigning PHHI to chromosome 11p15.1. Polymorphic repeat markers in the 11p15 region are detailed. The numbers between markers are approximate distances in centiMorgans. The vertical lines at the right represent the progressively narrowed regions to which PHHI was linked.

ference was discernible between this family and those with mutations of either the first or second NBF regions of SUR1. The identification of a mutation in Kir6.2 in those affected with PHHI demonstrates genetic heterogeneity for this disorder. In addition, demonstration that loss of function of either Kir6.2 or SUR1 leads to unregulated secretion of insulin and the PHHI phenotype supports the model that these two subunits cooperate in the formation of IKATP. The identified mutations do not represent the full spectrum of mutations in individuals affected with PHHI, because we are aware of many families that exhibit linkage to chromosome 11p15.1, yet demonstrate none of the mutations now described. By analogy to individuals affected with cystic fibrosis and adrenoleukodystrophy, both of which are caused by mutations in ATP-binding cassette superfamily members, the spectrum of mutations in familial PHHI may be very large. Except in those cases that may be attributed to founder effects, like the Saudi Arabian and Ashkenazi Jewish populations studied, the mutations found to date appear to be of low frequency in the general PHHI population. In

addition, the majority of those affected with PHHI exhibit sporadic, rather than familial, forms of the disease. Although sporadic forms of the disease are also associated with loss of KATP channel activity, it remains to be proven whether mutations of either SUR or Kir6.2 are responsible for these cases; it is possible that further genetic heterogeneity may be present and that additional genes with function related to KATP channel activity may be identified as responsible for sporadic forms of PHHI. Further comparison of phenotype and genotype status in those affected with PHHI is needed to ascertain whether a correlation may be made between specific mutations of KATP channel subunits and clinical course. Before prenatal diagnosis or genetic counseling may considered for clinical use for this disorder, further study to answer such questions is needed.

MOLECULAR PATHOPHYSIOLOGY OF DISEASE The resting membrane potential of pancreatic islet β cells is set by potassium channels. A link between the metabolic state of the cell and membrane electrical events is created by the inhibition of

516

SECTION VII / ENDOCRINOLOGY

All evaluated pancreatic specimens lacked KATP channel activity. Furthermore, the PHHI β cells were found to be spontaneously electrically active and to exhibit high basal cytosolic calcium concentrations as a result of calcium influx. Therefore, it can be conclusively stated that the pathogenesis of PHHI, as predicted by genetic evidence and supported by direct electrophysiologic measurements of patient pancreatic samples, is caused by abnormalities of the pancreatic islet KATP channel.

MANAGEMENT AND TREATMENT

Figure 55-2 A schematic drawing of events leading to insulin secretion by pancreatic islet β cells. Alterations in intracellular ATP level leads to closure of ATP-sensitive potassium (KATP) channels, which are composed at least of the sulfonylurea receptor (SUR) and the inward rectifier Kir6.2. Subsequent membrane depolarization, opening of voltage-gated calcium channels, and exocytosis of insulin follows. In the presence of mutations in either the SUR or Kir6.2, KATP channels are unresponsive to alterations in intracellular ATP levels, leading to inappropriate closure, membrane depolarization, elevated intracellular calcium levels, and, ultimately, unregulated exocytosis of insulin.

For individuals affected with PHHI, prompt recognition and treatment, with rapid resolution of hypoglycemia, is imperative to prevent permanent damage to the developing central nervous system. Treatment involves use of intravenous solutions of glucose and drugs, such as diazoxide and somatostatin analogs, which inhibit insulin secretion. These medical regimens are often not feasible for long-term use because of lack of treatment success, side effects, and practical problems with drug delivery. Near-complete (95%) pancreatectomy is frequently required for resolution of hypoglycemia. The need for additional treatment strategies remains apparent. Although patients who have undergone 95% pancreatectomy have demonstrated few ill effects during the short term, endocrine pancreatic insufficiency and growth failure is a significant problem affecting the majority of those examined for longer periods. The elucidation of the molecular basis for PHHI may allow the development of more effective therapy for this difficult management situation. One potentially promising recent addition to the therapeutic options for treatment of PHHI is the use of the calcium channel blocking agent verapamil, as proposed by Lindley et al. The trial use of this drug was based on the molecular pathophysiology of PHHI, which includes inappropriate activation of voltage-gated calcium channels and elevated intracellular calcium levels. In a single affected child, with persistent hypoglycemia despite treatment with glucagon, somatostatin, diazoxide, and 95% pancreatectomy, blood glucose levels and duration of fasting time increased after treatment with nifedipine. Studies of the resected pancreatic specimen from this child demonstrated that the excess spontaneous electrical activity present was reversibly blocked by verapamil. Further study of this treatment approach is needed to define the role for calcium channel blockade in the management of PHHI.

FUTURE DIRECTIONS Figure 55-3 Mutations of SUR1 associated with persistent hyperinsulinemic hypoglycemia of infancy (PHHI). The SUR1 gene is composed of 39 exons. The nucleotide-binding fold regions (NBF-1 and -2) are diagrammed here, with NBF-1 in the upper, and NBF-2 in the lower, portion. Exons are represented as boxes and individual exon numbers are included.

KATP channels by ATP. As intracellular ATP level increases, inhibition of KATP channels progressively increases, leading to cell depolarization. The model for the molecular basis of PHHI, based on the molecular genetic data, predicts that disruption of either SUR1 or Kir6.2 results in both inappropriate closure of KATP channels and membrane depolarization, and subsequently to excessive unregulated insulin secretion. Direct support for this model has been obtained by patch-clamp analysis of primary cultured human islet cells obtained from a group of five infants affected with PHHI.

Study of PHHI has demonstrated that this is a genetically heterogeneous disorder of pancreatic islet KATP channel dysfunction. Before PHHI may reasonably be considered as a candidate disease for gene replacement therapy, an understanding of the phenotype of specific mutations is required because with time, spontaneous abatement of clinical symptoms occurs in some cases. Further understanding of the molecular basis for PHHI, and the associated aberrant regulation of insulin secretion, may provide additional insight into effective therapeutic options for this and other disorders of pancreatic function and glucose metabolism.

SELECTED REFERENCES Aguilar-Bryan L, Nichols CG, Wechsler SW, et al. Cloning of the β-cell high-affinity sulfonylurea receptor: a regulator of insulin secretion. Science 1995;268:423–426. Ashcroft FM. Adenosine 5'-triphosphate-sensitive potassium channels. Annu Rev Neurosci 1988;11:97–118.

CHAPTER 55 / HYPOGLYCEMIA WITH HYPERINSULINEMIA

Aynsley-Green A, Polak JM, Bloom SR, et al. Nesidioblastosis of the pancreas: definition of the syndrome and the management of the severe neonatal hyperinsulinaemic hypoglycaemia. Arch Dis Child 1981;56:496–508. Glaser B, Chiu KC, Anker R, et al. Familial hyperinsulinism maps to chromosome 11p14-15.1, 30 cM centromeric to the insulin gene. Nature Genet 1994;7:185–188. Haymond MW. Hypoglycemia in infants and children. Endocrinol Metabol Clin North Am 1989;18:211–252. Higgins CF. ABC transporters: from microorganisms to man. Annu Rev Cell Biol 1992;8:67–113. Higgins CF. The ABC of channel regulation. Cell 1995;82:693–696. Inagaki N, Gonoi T, Clement JP, et al. Reconstitution of I-KATP: an inward rectifier subunit plus the sulfonylurea receptor. Science 1995;270:1166–1170. Kane C, Shepherd RM, Squires PE, et al. Loss of functional KATP channels in pancreatic β-cells causes persistent hyperinsulinemic hypoglycemia of infancy. Nat Med 1996;2:1344–1347. Kuvuvitis A, Deal C, Arbour L, Polychronakos C. An autosomal dominant form of familial persistent hyperinsulinemic hypoglycemia of infancy, not linked to the sulfonylurea receptor locus. J Clin Endocrinol Metab 1997;82:1192–1194. Lander ES, Botstein D. Homozygosity mapping: a way to map human recessive traits with the DNA of inbred children. Science 1987; 236:1567–1570. Leibowitz G, Glaser B, Higazi AA, Salameh M, Cerasi E, Landau H. Hyperinsulinemic hypoglycemia of infancy in clinical remission: high incidence of diabetes mellitus and persistent β-cell dysfunction at long-term follow-up. J Clin Endocrinol Metab 1995, 80:386–392.

517

Lindley KJ, Dunne MJ, Kane C, et al. Ionic control of β cell function in nesidioblastosis. A possible therapeutic role for calcium channel blockade. Arch Dis Child 1996;74:373–378. Nestorowicz A, Wilson BA, Schoor KP, et al. Mutations in the sulfonylurea receptor gene are associated with familial hyperinsulinism in Ashkenazi Jews. Hum Mol Genet 1996;5:1813–1822. Nichols CG, Shyng SL, Nestorowicz A, et al. Adenosine diphosphate as an intracellular regulator of insulin secretion. Science 1996;272: 1785–1787. Rorsman P. The pancreatic beta-cell as a fuel sensor: an electrophysiologist’s viewpoint. Diabetologia 1997;40:487–495. Shilyansky J, Fisher S, Cutz E, Perlman K Filler RM. Is 95% pancreatectomy the procedure of choice for treatment of persistent hyperinsulinemic hypoglycemia of the neonate? J Pediatr Surg 1997, 32:342–346. Thomas PM, Cote GJ, Hallman DM, Mathew PM. Homozygosity mapping, to chromosome 11p, of the gene for familial persistent hyperinsulinemic hypoglycemia of infancy. Am J Hum Genet 1995;56:416–421. Thomas PM, Cote GJ, Wohllk N, et al. Mutations in the sulfonylurea receptor gene in familial persistent hyperinsulinemic hypoglycemia of infancy. Science 1995;268:426–429. Thomas PM, Wohllk N, Huang E, et al. Inactivation of the first nucleotide-binding fold of the sulfonylurea receptor, and familial persistent hyperinsulinemic hypoglycemia of infancy. Am J Hum Genet 1996;59:510–518. Thomas P, Ye Y, Lightner E. Mutation of the pancreatic islet inward rectifier Kir6.2 also leads to familial persistent hyperinsulinemic hypoglycemia of infancy. Hum Mol Genet 1996;5:1809–1812.

CHAPTER 56 / REGULATION OF REPRODUCTION

56

519

Regulation of Reproduction MICHAEL J. MCPHAUL

INTRODUCTION Reproduction involves the integration of biologic processes as diverse as behavior and specialized cell divisions. This integration is accomplished by a complex web of interrelated hormonal and neural circuits that employ a variety of signaling molecules and pathways. The goal of this overview is to introduce the major effectors and pathways that participate in this regulation. A basic tenet of this summary is that these pathways can be viewed as a cascade comprised of several different levels. Although many differences exist in this cascade between men and women, the general features and overall hierarchical organization are similar.

HYPOTHALAMUS Early in embryogenesis, the cells destined to become gonadotropin-releasing hormone (GnRH)-secreting neurons migrate from the olfactory placode and became localized to specific regions of the preoptic and medial basal hypothalamus. Terminals of these neurons terminate on the long portal vessels, which carry the secreted GnRH to the gonadotropin-secreting cells of the anterior pituitary. Although unique in many ways, one of the important characteristics of these GnRH-secreting neurons is the rhythmicity of their secretory patterns. For this reason, the complex of GnRH-secreting neurons has been collectively termed the “pulse generator.” Episodic secretion of GnRH gives rise to measurable pulses of GnRH secretion approximately once each hour. Many influences that alter the secretion of gonadotropins can be traced to their effects on the pattern and amplitude of pulsatile GnRH release. This pattern of episodic secretion determines the activity of the other levels in the cascade below. GnRH is an ancient molecule and its structure has been highly conserved in vertebrate species, from teleost fishes to humans. The mature GnRH molecule is a decapeptide and is derived from a precursor 82 amino acids in length that is processed into the mature GnRH peptide by specific endoproteases (Fig. 56-1). Of interest, in contrast to the high degree of evolutionary conservation of the structure of the GnRH decapeptide itself, surrounding regions of the GnRH precursor are less conserved. Although adjacent segments of the prohormone (e.g., the carboxy-terminal fragment known as GAP) may serve a distinct function, the sequences encoding this polypeptide are much less conserved between species. This lesser degree of conservation suggests that From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

although it may play a distinct role in vertebrate physiology (one that has not yet been defined), its principal function may simply be its participation in the folding and processing of the mature hormone.

PITUITARY SECRETION OF LUTEINIZING HORMONE AND FOLLICLE-STIMULATING HORMONE The anterior pituitary is derived from the pharyngeal epithelium and comprises of five cell types that secrete six polypeptide hormones. Unlike the other cell types of the pituitary, gonadotrophs secrete both follicle-stimulating hormone (FSH) and luteinizing hormone (LH). This departure from the “one cell, one hormone” rule that characterizes the other cell types of the anterior pituitary poses an apparent paradox, because FSH and LH can have different secretory patterns. Although the mechanisms by which this is accomplished have not been completely defined, it seems likely that the documented morphologic and functional heterogeneity of gonadotrophs is sufficient to account for the differential regulation of these hormones. After delivery to the pituitary via the long portal vessels, GnRH binds to a specific receptor on the surface of the gonadotropinsecreting cells. This receptor is a member of a large gene family of membrane receptors that is predicted from its primary amino acid sequence to span the plasma membrane seven times and is most closely related to receptors from the same gene family that bind oxytocin and vasopressin (Fig. 56-2). In addition to those on the surface of the pituitary gonadotrophs, GnRH receptors have been detected in the gonads, prostate, and within selected regions of the brain (particularly GnRH-secreting cells of the hypothalamus). Although signal transduction is likely to involve a number of second messenger signaling pathways, the principal pathway appears to be via the G-protein-mediated stimulation of phospholipase C, rises in inositol 1,4,5-trisphosphate (IP3), and changes in intracellular calcium levels. Activation of the mitogen-activated protein kinase (MAPK) cascade by GnRH has also been demonstrated. An important feature of this step in the hypothalamic–pituitary– gonadal (HPG) axis is that the episodic secretion of GnRH by the GnRH-secreting neurons causes GnRH to be presented to the receptor in an intermittent fashion. Furthermore, it is clear that this pulsatile binding of GnRH to its receptor on gonadotrophs is a critical attribute of this signaling pathway. Binding of GnRH to the GnRH receptor in a manner that is not pulsatile will lead to the induction of a state of tachyphylaxis and result in the cessation of LH and FSH secretion. The principal effect of GnRH binding to its receptor on the surface of gonadotrophs is to stimulate the synthesis and release of

519

520

SECTION VII / ENDOCRINOLOGY

Figure 56-1 Reproductive hormones. The structures of the major classes of peptide involved in the process of reproduction are depicted. (Upper Panel) The positions of the pro- and GAP peptides are shown relative to the positions of the mature gonadotropin-releasing hormone (GnRH) molecule (shown as a dark bar). Luteinizing hormone (LH), follicle-stimulating hormone (FSH), and chorionic gonadotropin (CG) are heterodimeric glycoprotein hormones composed of a common α-subunit and a β-subunit that is unique for each hormone. (Middle panel) The structures of the hormones and subunits of inhibin and activin are presented. The two β-subunits and the single α-subunit are processed from larger precursors. As shown, combinations of these subunits give rise to a number of different heterodimeric (inhibins) and homodimeric combinations. (Lower panel) The structures of three of the major steroid hormones are shown. Each of the hormones is synthesized from cholesterol by the action of steroidogenic P450s and specific oxidoreductases.

LH and FSH. LH and FSH are heteromeric glycoprotein hormones that are structurally related and that likely evolved from a single ancestral precursor (Fig. 56-1). Both the common α-subunit and the unique β-subunits are processed from larger precursors that are cleaved by specific endoproteases to their mature forms. Glycosylation of both the α- and β-subunits occurs and has significant biologic effects, both in terms of the biologic activity and halflives of these molecules. Because of the pulsatile nature of GnRH secretion, serum levels of FSH and LH also vary in a pulsatile fashion and protocols for the measurement of serum gonadotropin levels should take this moment-to-moment variation into account.

Figure 56-2 Reproductive receptors. Schematic representation of three major classes of receptor that transduce signals in the reproductive axes. (Above) GnRH , LH, and FSH receptors belong to a large family of transmembrane receptors that are coupled to G-proteinsignaling pathways. The structures predicted for members of this large gene family include an extracellular domain that binds ligand, seven transmembrane-spanning segments, and cytoplasmic loops of varying lengths. The LH and FSH receptors are closely related to one another, whereas the GnRH receptor is more distant member of this gene family. (Middle) Activin receptors are composed of two nonidentical subunits that have been designated as type I and II (based on their molecular sizes and by analogy to the type I and II transforming growth factor-β receptors). Each comprises an extracellular domain, membrane-spanning segment, and cytoplasmic domain. The intracellular domain encodes a protein kinase that phosphorylates specific serine and threonine residues of target proteins. Signal transduction occurs only when ligand is bound to both type I and II subunits forming an active heteromeric complex. (Below) Structure of a prototypic member of the steroid hormone–thyroid hormone–retinoic acid family of nuclear receptors. Essential features include a DNA-binding domain that recognizes specific sequences within regulated genes, a hormone-binding domain that binds ligand with high affinity, and an amino-terminal segment that is of variable length, but that is important for the maximal regulation of responsive genes. This same general structure is applicable to additional members of this gene family (termed “orphan receptors”) for which a ligand has not been identified. Some orphan receptors (such as DAX and SF-1) have been shown to play critical roles in the development and function of the hypothalamic–pituitary–gonadal axis.

GONADS Although the patterns of hormone secretion may differ substantially between men and women at the level of the pituitary and hypothalamus, many of the components of this axis are similar.

CHAPTER 56 / REGULATION OF REPRODUCTION

This is not true of the gonads themselves. Both the testes and ovary are complex organs comprising a number of specialized cell types (discussed in Chapters 61 and 62). The focus of this section is to provide a partial description of the factors and receptors that have been shown to participate in the regulation of the gonadal axis. SITES OF ACTION OF LH AND FSH—TESTES LH is delivered to the testes via the blood and exerts its effects by binding to a high-affinity receptor expressed on the surface of Leydig cells. Characterization of cDNAs encoding this receptor predicts that the LH receptor is approximately 670–680 amino acids long (depending on the species) and is predicted to encode seven segments that span the plasma membrane. The activation of this G-protein-associated receptor has been correlated with increased intracellular levels of cAMP, which is viewed as an integral part of the second messenger pathway. Results from some investigators suggest that additional second messenger pathways, such as activation by phospholipase C, also contribute to gene regulation effected by LH receptor activation. In the testes, the FSH receptor is localized exclusively to the Sertoli cells. The isolation of cDNAs encoding the FSH receptor revealed it to be highly related to the LH receptor, both in terms of size, predicted amino acid composition, and structural organization. As with the LH receptor, it is also believed to be coupled to G-protein-dependent signaling pathways. Stimulation of adenylate cyclase appears to be a critical signaling pathway and modulation of intracellular calcium levels may also be important. Via these mechanisms, FSH is believed to modulate processes important for both Sertoli cell growth and differentiation and for Sertoli cell function in the regulation of spermatogenesis. SITES OF ACTIONS OF LH AND FSH—OVARY As in the testes, LH and FSH receptors are distributed in a cell-type-specific fashion in the ovary. The FSH receptor is localized exclusively to the surface of granulosa cells. Stimulation of this receptor in granulosa cells activates cyclic AMP-dependent signaling mechanisms. This activation promotes granulosa cell growth and differentiation and stimulates the gene encoding the aromatase P450, the ratelimiting step in estrogen biosynthesis. In contrast to the FSH receptor, the LH receptor is not confined to a single cell type within the ovary, but instead is expressed in both theca and granulosa cells. The levels of LH receptor detected in these cell types varies, depending on the developmental stage of the follicle. LH receptors are present only on the theca cells in immature follicles, but are detected on both theca and granulosa cells of mature antral follicles. In theca cells, LH is believed to stimulate the level of androgen synthesis via cyclic AMP-dependent pathways. STEROID HORMONE SYNTHESIS Testes The bulk of steroid hormone synthesis in the testes is confined to the Leydig cells, which in adult males synthesize approximately 5 mg of testosterone daily. As noted above (and depicted schematically in Fig. 56-3), substantial variations in serum testosterone occur moment to moment, owing to the pulsatility of LH secretion. Only small quantities of estradiol and 5α-dihydrotestosterone (5α-DHT) are secreted directly by the testes and the remainder of these hormones are derived from metabolic conversion in nongonadal tissues. Ovary Steroidogenesis in the ovary is complex and changes markedly during the follicular and luteal phases of the ovarian cycle. During early follicular development, enzymes early in the steroidogenic pathway (side-chain cleavage, 17-hydroxylase) are

521

Figure 56-3 Integration of signals in the male reproductive axis. The pulses of GnRH secreted by neurons of the GnRH pulse generator are delivered via the long portal vessels to the cells of the anterior hypophysis. This signal leads to the pulsatile secretion of LH and FSH into the blood (shown at the upper right for LH). LH acts at the level of the Leydig cells within the interstitium of the testes to stimulate the pulsatile secretion of testosterone. The secreted testosterone acts within the testes to stimulate spermatogenesis and systemically to effect a number of physiologically processes, including the feedback regulation of LH secretion. FSH binds to the membranes of Sertoli cells, where it acts to promote the process of spermatogenesis. Inhibin secretion by the Sertoli cell (particularly inhibin-B) acts as a component of the feedback loop that regulates the secretion of FSH.

localized to the thecal cells and aromatase is detected in the granulosa cells (“two compartment” model of ovarian steroidogenesis, see Chapter 62). Following ovulation, the cells of the corpus luteum continue to express high levels of side-chain cleavage P450 and 17-hydroxylase, but the levels of expression of aromatase drop. INHIBINS, ACTIVINS, AND FOLLISTATIN The development of assays to measure pituitary gonadotropin levels led to the identification of nonsteroidal activities of gonadal origin that could suppress FSH release in castrate animals. This activity, termed inhibin, was ultimately purified and found to be composed of heterodimeric molecules composed of α- and β-subunits. Subsequent cloning studies revealed the existence of genes encoding a single α-subunit and two distinct β-subunits (βA and βB), as shown in Fig. 56-1. During studies to characterize the inhibins, it was noted that selected fractions of follicular fluid exhibited a distinctive activity. When used in appropriate bioassays, it was found that instead of inhibiting FSH release, these fractions led to a stimulation of FSH release. Purification of this activity, collectively termed activins, revealed that the active proteins were dimeric molecules composed of combinations of the same β-subunits that participate in the formation of inhibin heterodimers. Owing to its dimeric structure, three different activin molecules are possible: activin A, activin AB, and activin B, as shown in Fig. 56-1. Studies by Ueno and coworkers established the existence of yet a third molecule capable of regulating FSH release in bioassays,

522

SECTION VII / ENDOCRINOLOGY

termed follistatin. Analysis of the structure of cDNAs encoding follistatin have demonstrated that this molecule is unrelated to either inhibin or activin. Current models suggest that a large portion of the biologic activity of this molecule is believed to derive from its ability to bind (and sequester) biologically active activins. Testes During embryonic development, the sites of α- and β-subunit expression vary, but become localized to the seminiferous tubule by the end of gestation. There is little question that Sertoli cells produce both inhibin and activin, although the mechanisms by which the production of these molecules by Sertoli cells are regulated is still unsettled. Available evidence suggests that the expression of inhibin subunits may vary in relation to the stage of development of the adjacent seminiferous tubule epithelium. Sertoli cells produce predominantly α-βB inhibin (inhibin-B), which appears to be regulated (directly or indirectly) by FSH, and likely by other paracrine factors as well. It is presumably through the production of inhibin-B that the testes effects a negative regulatory influence on FSH secretion by the pathway. The regulation of production of activin expression by Sertoli cells has been less completely defined. Ovary A number of studies have examined the identity of cell types that express α- and β-subunits in the ovary. Although some variation in the sites of subunit expression is evident in some studies, certain features are constant. First, it is clear that both α-subunit protein and α-subunit mRNA can be detected in the granulosa cells of growing follicles. Although patterns of β-subunit expression are more variable, it is detected in the thecal cells of dominant follicles in the human ovary. Nongonadal Sites of Expression A great surprise resulting from the development of assays for inhibin and activin was the recognition that many tissues outside the reproductive tract could be shown to express activin and inhibin subunits. The expression patterns reported for the different subunits suggests that activins are made in a variety of tissues and are likely to subserve a number of functions. The expression of inhibins, on the other hand, is more restricted and appears to play a central role in the modulation of FSH. Although the expression of follistatin often parallels the expression of inhibin and activin subunits, additional sites of expression have been described that imply functions beyond those related solely to the modulation of inhibin or activin. Mechanisms of Action The receptor that mediates the actions of activins is an area of active investigation. Specific proteins that bind activin have been identified and classified descriptively based on their physical properties. In crosslinking experiments, activin was shown to bind to two classes of membrane proteins (termed type I and type II receptors by analogy to the receptors for transforming growth factor [TGF]-β) having molecular weights of approximately ~50 and 70 kDa, respectively. The detailed characterization of the activin receptors awaited the isolation of cDNAs encoding these proteins. The first such cDNA was isolated by expression cloning methods based on its capacity to confer on transfected cells the capacity to bind 125Ilabeled activin with high affinity. Analysis of the cDNA sequence predicted a transmembrane protein that contained motifs that suggested it employed a serine and threonine kinase as its intracellular signaling mechanism. The size of this receptor protein revealed it to be a type II activin receptor (Act RII). Subsequent studies revealed the existence of additional related members of the same gene family.

When antibodies directed at the type II activin receptors (such as Act RII) were used to precipitate the protein complexes crosslinked with 125I-activin, the precipitates could be shown to contain proteins distinct from the type II receptors themselves. Screening strategies to isolate cDNAs encoding these proteins employed polymerase chain reaction (PCR)-based methods to identify cDNAs encoding novel transmembrane serine–threonine kinases. Subsequent crosslinking studies demonstrated that novel receptors isolated in this fashion (the activin type I receptors) could bind activin, but only in the presence of a type II activin receptor. The available evidence suggests that at least two distinct activin type I receptors exist. Thus, the combination of type I and type II receptors expressed in a given cell type is likely to afford an additional level at which differences in tissue responsiveness could be modulated. To date, receptors have not been described that specifically bind either inhibin or follistatin. In the case of inhibin, owing to its distinct biologic properties, it is likely that such specific receptors exist. In the case of follistatin, however, it is believed that many, if not all, of its biologic properties can be accounted for on the basis of its capacity to bind and sequester activin with high affinity and the existence of a specific follistatin receptor seems less likely. Insights From the Studies of Mice Carrying Targeted Disruptions of the Genes Encoding Inhibin, Follistatin, Activins, and the Activin Receptors The study of the physiologic roles of inhibins, activins, and follistatin have been complicated by the intricacies of the systems being studied. Investigators have attempted to unravel these intricate systems by deleting the functional genes in living animals. Such studies have provided a number of interesting insights. Inhibin Matzuk and coworkers disrupted the functional α-subunit gene that is a component of both the inhibin-A and inhibin-B genes (see Fig. 56-1). As would be expected on the basis of the assays used to define the inhibins as regulators of pituitary FSH production, animals homozygous for the disrupted allele displayed elevated levels of serum FSH (see Chapter 10). Although the male and female homozygous mice were phenotypically normal, fertility was not normal in crosses with wild-type mice. Examination of the gonads revealed that the testes and ovaries of the homozygotes developed gonadal stromal tumor cells in a very high proportion of animals. These findings suggest that in addition to its role in the regulation of FSH, that in the cells of the gonadal stroma, inhibin—either directly or indirectly—also acts as a tumor suppressor. The basis for this capacity to antagonize the development of gonadal tumors remains to be elucidated. However, it is notable that gonadotropins are required for the development of activin-dependent tumors as crossbreeding with gonadotropindeficient (hpg) mice prevents ovarian tumors. These findings indicate that gonadotropins are essential modifiers for tumor development in this model, perhaps reflecting their effects on follicle development. Activins Owing to the subunit composition of the activin molecules, testing the functional relevance of the different forms of activin via the generation of knockout mice was considerably more complex. The first information was from studies of mice in which the β-B subunit was disrupted. Mice homozygous for a β-B mutant allele demonstrated normal viability. The most marked abnormalities were abnormalities of eyelid development and female reproduction. Mesoderm formation and neurulation were normal.

CHAPTER 56 / REGULATION OF REPRODUCTION

Considerably different phenotypes were noted in mice in which the β-A subunit was disrupted. Although mice homozygous for the deleted β-A allele progressed to term normally, the mice died within 24 hours of birth. These mice also displayed several consistent phenotypic abnormalities, particularly defects in whisker and tooth development. A variable number displayed abnormalities of palatal development. When mice carrying the β-B and β-A mutant alleles were bred to produce compound homozygotes, it was observed that the phenotype of the resulting animals was additive. That is, the defects that were present were those observed in the animals carrying either mutant allele. This observation permitted two important conclusions. The first is that the molecules composed of the individual activin subunits (i.e., β-A and β-B homodimers) are not functionally compensating for one another. Second, these studies suggest that β-A/β-B heterodimers do not have a unique function during embryogenesis. Activin Receptor Type II The study of mice carrying targeted deletions of the activin receptor II provided a number of interesting contrasts to the results obtained in studies of mice carrying disruptions of the β-A or β-B genes. Male mice were delayed in reaching fertility and female mice were infertile. In male mice, these functional changes were traced to a decrease of seminiferous tubule diameter and total volume. In females, frequent follicular atresia was seen and corpora lutea were rare, consistent with abnormalities of the estrous cycle in the mutant mice. Investigations of mice homozygous for the deleted Act RII gene documented reduced FSH levels in both male and female animals, compared with wild-type animals, permitting the authors to conclude that the reproductive defects observed were a consequence of the reduced FSH levels caused by the absence of Act RII signaling. Although many of the phenotypic abnormalities were consistent with the findings of the activin subunit gene disruptions (particularly the effects on the pituitary and gonadal function), other major effects were noted that were specific for the Act RII disruption. In particular, mice carrying a disruption of this gene demonstrated defects of the first brachial arch (reminiscent of Pierre-Robin syndrome). These defects suggested that the Act RII receptor is involved in the transduction of signals that are involved in the formation of the first brachial arch. The nature of this signal (apparently distinct from the activins themselves since these features were not observed in the activin “knockouts” themselves) remains to be identified. Follistatin Follistatin was originally defined on the basis of its capacity to antagonize the actions of activins. The targeted disruption of the murine follistatin gene permitted an examination of the physiologic impact of this molecule. Mice with the homozygous deletion died rapidly and displayed a number of defects, some of which suggested an alteration of activin function and many of which were specific for the follistatin gene disruption. Whisker development was abnormal, suggesting that follistatin may be an important modulator of the action of activin β-A. The shiny, taut skin resembled the skin of transgenic animals in which TGF-β overexpression had been directed using the keratin promoter. In addition, a number of additional defects were observed in the development of the musculoskeletal system, including changes in the number of ribs and vertebrae, as well as defects in the development of the palate. These results suggested that in addition to the role played in the modulation of activin function, follistatin may well play additional roles in the modulation of the

523

Figure 56-4 Integration of signals in the female reproductive axis. The pulsatile secretion of GnRH stimulates the pulsatile release of LH and FSH in to the blood by the gonadotropes (shown at upper right). Unlike the male, the levels of FSH and LH vary dramatically at different points in the menstrual cycle. FSH binds to the FSH receptor (localized to the plasma membranes of granulosa cells), where it acts to stimulate the growth of granulosa cells of the dominant follicle, as well as the levels of aromatase that they express. The sustained increased estradiol levels that result in the late follicular phase trigger the midcycle surge of LH and FSH that is required for ovulation. In the early portion of the menstrual cycle, LH acts principally at the levels of the ovarian stromal cells and stimulates the secretion of androstenedione, the substrate for granulosa cell estrogen synthesis (twocompartment model of ovarian hormonogenisis). After ovulation, the vascularization of the corpus luteum makes increased quantities of cholesterol—in the form of low-density lipoprotein—available for steroidogenesis, accounting for the increased progesterone secretion characteristic of the corpus luteum. If fertilization and implantation do not occur, the function of the corpus luteum declines, and progesterone and estradiol levels fall. As indicated (lower right), the regulation of serum inhibin-A and inhibin-B levels differ markedly at different points in the menstrual cycle.

function of other TGF-β related proteins or that it may possess functions that are independent of its role in the modulation of activins.

INTEGRATION OF THE SIGNALS Each component of the hypothalamic–pituitary–gonadal axis is coordinated in a hierarchical fashion after puberty in men and women. In men, although levels of the gonadotropins and testosterone can be shown to vary minute by minute and to display some diurnal variation, large-scale changes do not occur (Fig. 56-3). In women, interplay between the ovary, hypothalamus, and pituitary results in the cyclic changes that characterize the menstrual cycle. A schematic representing the hormonal changes that occur during the menstrual cycle are depicted in Fig. 56-4.

524

SECTION VII / ENDOCRINOLOGY

MODULATION OF THE HYPOTHALAMIC– PITUITARY–GONADAL AXIS INHIBITION OF GONADAL FUNCTION Female Contraception The most common intervention affecting the function of the HPG axis is use of oral contraceptives. In the United States, these preparations include small doses of estrogen (ethinyl estradiol) and one of several progestational agents (norethindrone, norethindrone acetate, norethynodrel, norgestrel, ethynodiol diacetate, norgestimate, or desogestrel). The administration of these estrogen and progestogen combinations in low doses serves to inhibit the normal midcycle surge and to prevent ovulation. In addition to a large number of different combinations of these agents, contraceptives containing only progestational agents are available as oral agents and as depot preparations. As would be expected from their mechanisms of action, the inhibition of ovulation that is achieved using such agents is readily reversible and normal ovulatory cycles resume within 6 months in more than 95% of women after cessation of contraceptive use. Male Contraception The success achieved in the development of oral contraceptives for women stimulated interest in developing similar methods for male contraception. These efforts have employed the parenteral administration of various androgen preparations. Protocols using weekly injections of testosterone enanthate have been the most widely studied and have been found to reproducibly induce a state of azoospermia or oligospermia. Using such protocols, contraception rates have been estimated at 0.8 pregnancies per 100 person-years in those patients whose sperm counts fall into the azoospermic range. Although this rate is comparable to other reversible methods of contraception, the failure to induce azoospermia in as many as 50% of subjects in some studies is unacceptable and is a major impediment to the broader application of such methods. The development of effective, reversible hormonal forms of male contraception have focused on the identification of regimens (e.g., the combination of testosterone injections and oral progestogens) that will lead to a higher frequency of azoospermia. GnRH Analogs Attempts to deliver GnRH chronically to stimulate gonadal function led to the recognition that the pulsatile delivery of GnRH to the gonadotrophs was essential for normal gonadotroph function. As noted above, the continuous delivery of GnRH to the pituitary leads to the induction of a form of reversible secondary hypogonadism. The mechanism of this tachyphylaxis has been traced to a downregulation of GnRH receptors on the gonadotroph and an uncoupling of postreceptor pathways. These observations have led to the development and application of GnRH agonists to induce a state of reversible hypogonadism. These agents have been widely used to achieve a functionally castrate state by pharmacologic means. Such regimens are effective in both men and women and have been employed in the treatment of a wide range of conditions. STIMULATION OF GONADAL FUNCTION Female The techniques available to modulate the female reproductive tract and stimulate fertility have undergone explosive growth since the advent of successful in vitro fertilization (IVF) techniques in the late 1970s. Although the techniques and methodologies employed are quite varied, all rely on the use of agents to modulate the maturation of oocytes in a precise fashion, permitting the harvesting of ova for fertilization in vitro.

The simplest of such protocols employ the modulation of endogenous gonadotropin production using agents such as clomiphene citrate. This agent acts as an antiestrogen and acts to augment the secretion of endogenous gonadotropins. In some protocols, these effects are further augmented by the administration of exogenous gonadotropins. Although somewhat more cost-effective, overall pregnancy rates using such stimulation protocols are lower than those using more conventional IVF treatment regimens. In these latter protocols, GnRH agonists are used to downregulate the production of endogenous gonadotropins so that ovarian follicular development can be regulated under circumstances in which the effect of endogenous gonadotropins are negligible. When suppression of ovarian function is achieved using these protocols, ovarian stimulation is achieved using exogenous gonadotropins for 9–10 days until follicular development is achieved. When follicular maturity is reached (assessed by ultrasonography and estradiol measurements), human chorionic gonadotrophin (hCG) is administered (to simulate the LH surge of a normal ovarian cycle), and oocytes are retrieved by a laparoscopic procedure. Male In men, primary and secondary hypogonadism is treated most frequently with transdermal or injectable depot preparations of testosterone, owing to the rapid metabolism and inactivation of unsubstituted forms of testosterone and the hepatotoxicity associated with the oral administration of the more stable 17-substituted androgens. The systemic administration of testosterone—although providing hormone sufficient to support normal potency and secondary sexual characteristics—does not result in intratesticular concentrations of androgen that are capable of supporting normal spermatogenesis. For this reason, when fertility is desired by men with secondary forms of hypogonadism, different hormonal regimens must be employed that incorporate stimuli that mimic the trophic influences of gonadotropins on the testes. In men that have never undergone normal pubertal maturation (and as such have never established normal spermatogenesis), replacement regimens that supply stimuli that mimic the effects of both LH and FSH must be employed. By contrast, the administration of hCG may suffice to reestablish fertility in individuals that have developed an acquired form of secondary hypogonadism as adults.

SUMMARY Increasingly detailed information has been amassed regarding the molecules, the receptors and the mechanisms by which the function of the gonads are controlled. In some instances, this knowledge has permitted the definition of the pathophysiology causing specific disease entities. In other instances (e.g., polycystic ovarian disease), the use of this knowledge has not yet permitted a complete understanding of the disease entity to emerge.

ACKNOWLEDGMENTS This work was supported by NIH grants DK03892 and DK52678.

SELECTED REFERENCES Coerver KA, Woodruff TK, Finegold MJ, Mather J, Bradley A, Matzuk MM. Activin signaling through activin receptor type II causes cachexia-like symptoms in inhibin-deficient mice. Mol Endocrinol 1996;19:534–543. Cummings DE, Bremner WJ. Prospects for new hormonal male contraceptives. Endocrinol Metab Clin North Am 1994;23:893–922.

CHAPTER 56 / REGULATION OF REPRODUCTION

Danforth DR. Endocrine and paracrine control of oocyte development. Am J Obstet Gynecol 1995;172:747–752. Erickson GF, Danforth DR. Ovarian control of follicle development. Am J Obstet Gynecol 1995;172:736–747. Evans RM. The steroid and thyroid hormone receptor superfamily. Science 1988;240:889–895. Gaddy-Kurten D, Tsuchida K, Vale W. Activins and the receptor serine kinase superfamily. Recent Prog Horm Res 1995;50:109–129. Hillier SG, Whitelaw PF, Smyth CD. Follicular oestrogen synthesis: the ‘two-cell, two-gonadotrophin’ model revisited. Mol Cell Endocrinol 1994;100:51–54. Kumar TR, Wang Y, Lu N, Matzuk MM. Follicle stimulating hormone is required for ovarian follicle maturation but not male fertility. Nat Genet 1997;15:201–204. Kumar TR, Wang Y, Matzuk MM. Gonadotropins are essential modifier factors for gonadal tumor development in inhibin-deficient mice. Endocrinology 1996;137:4210–4216. Mangelsdorf DJ, Thummel C, Beato M, et al. The nuclear receptor superfamily: the second decade. Cell 1995;83:835–839. Matzuk MM, Kumar TR, Bradley A. Different phenotypes for mice deficient in either activins or activin receptor type II. Nature 1995;374:356–360.

525

Matzuk MM, Kumar TR, Vassalli A, et al. Functional analysis of activins during mammalian development. Nature 1995;374:354–356. Matzuk MM, Lu N, Vogel H, Sellheyer K, Roop DR, Bradley A. Multiple defects and perinatal death in mice deficient in follistatin. Nature 1995;374:360–363. Moley KH, Schreiber JR. Ovarian follicular growth, ovulation and atresia. Endocrine, paracrine and autocrine regulation. Adv Exp Med Biol 1995;377:103–119. Moore A, Krummen LA, Mather JP. Inhibins, activins, their binding proteins and receptors: interactions underlying paracrine activity in the testis. Mol Cell Endocrinol 1994;100:81–86. Saez JM. Leydig cells: endocrine, paracrine, and autocrine regulation. Endocr Rev 1994;15:574–626. Stojilkovic SS, Catt KJ. Expression and signal transduction pathways of gonadotropin-releasing hormone receptors. Recent Prog Horm Res 1995;50:109–129. Vassalli A, Matzuk MM, Gardner HA, Lee KF, Jaenisch R. Activin/ inhibin beta B subunit gene disruption leads to defects in eyelid development and female reproduction. Genes Dev 1994;8:414–427. Woodruff TK, Mather JP. Inhibin, activin and the female reproductive axis. Ann Rev Phys 1995;57:219–244.

CHAPTER 57 / SEX DETERMINATION AND DIFFERENTIATION

57

527

Disorders of Sex Determination and Differentiation CHARMIAN A. QUIGLEY

THE BASIS OF NORMAL SEX DETERMINATION AND DIFFERENTIATION Sex determination and differentiation are distinct, consecutive processes that follow the establishment of chromosomal sex at the time of fertilization. The term “sex determination” refers to the development of gonadal sex—a process that occurs at approximately 6–7 weeks’ gestation in the human male fetus, and around 10–11 weeks’ gestation in the female fetus. Male sex determination is synonymous with testis determination. This process (alternatively called primary sex differentiation) results from a coordinated series of exquisitely regulated, but as yet incompletely understood, events. As generally used, the term “sex differentiation,” refers to the processes downstream of gonadal development—the processes that are regulated by gonadal secretions or lack thereof (also called secondary sex differentiation). In essence, the sex chromosome complement endowed at fertilization determines gonad type and the latter determines the pattern of differentiation seen in the genital ducts and external genitalia. THE EMBRYOLOGY OF GONADAL AND GENITAL DEVELOPMENT In the first few weeks of gestation the primordial gonads develop from the condensation of a ridge of mesenchymal tissue located medial to the mesonephros (the forebear of the kidney), accompanied by thickening and proliferation of the coelomic epithelium, which penetrates the underlying mesenchyme. These primitive gonadal or genital ridges initially contain no germ cells. Between the 5th and 6th weeks of gestation (embryonic day 10.5–12 in the mouse) primordial germ cells migrate by ameboid movement from the endoderm of the yolk sac along the dorsal mesentery of the hindgut and into the indifferent gonad, where they invade the developing primary sex cords. Despite this, germ cells are not required for the processes of testis development to occur. At this stage the gonad comprises an outer cortex and inner medulla, and there is still no morphologic difference between the gonads of male and female fetuses. After the arrival of the germ cells, the gonad begins to differentiate: in the 46,XX fetus the cortex develops and the medulla regresses; in the 46,XY fetus the reverse occurs. The first discernible event in testis development is the appearance of primordial Sertoli cells, which differentiate from somatic cells of the coelomic epithelium. Soon after their appearFrom: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

ance in the gonadal ridge, Sertoli cells proliferate, aggregate around the primitive germ cells, and align into cordlike structures, which subsequently become the seminiferous tubules. The seclusion of germ cells within the tubules prevents meiosis and commits the germ cells to spermatogenic development. This organizational process appears to be regulated by the Sertoli cells themselves. About 1 week later (around 8 weeks), steroidogenic Leydig cells differentiate from primitive interstitial cells of mesonephric origin. This process too may be controlled by paracrine influences from Sertoli cells, possibly anti-Müllerian hormone (AMH). In contrast with the germ cell-independent development of the testes, the presence of germ cells is a prerequisite for normal ovarian differentiation. Without the germ cell seeding of the gonadal primordium, the tissue degenerates into a nonfunctional, mainly fibrous “streak.” Before week 10 of gestation the only feature that distinguishes an ovary is the absence of testicular features. Thereafter, ovarian structure becomes distinguishable, with the development of the primary medullary sex cords. Secondary cortical sex cords provide the supporting structure for the arriving germ cells. These break up into clusters, becoming the primordial follicles at about week 16. The primordial follicles contain the diploid oocytes, which remain quiescent until puberty. Follicular cells arise from the same cell lineage as Sertoli cells. Theca cells represent the ovarian counterpart of Leydig cells. There is a notable difference in the chronology of testicular versus ovarian development, the process of testis formation being completed by 8 weeks’ gestation, at which time the process of ovarian development has not yet begun. In fact, morphologic ovarian development is not completed until after most of the processes of phenotypic sex differentiation, described in the following section, have occurred. This, and the fact that phenotypic development is normal even in complete absence of the ovary, highlights the lack of involvement of the ovary in this process. The embryologic development of the testis and ovary are summarized in Fig. 57-1. The primitive internal genital tracts are also indistinguishable between the sexes until 7 weeks’ gestation (Fig. 57-2). The fetus is endowed with two sets of internal duct structures: the paramesonephric (Müllerian) ducts, and the mesonephric (Wolffian) ducts. From the 8th week of gestation hormonal secretions from the fetal testes induce masculinization of the internal genital structures. Initially the Wolffian ducts are stabilized by the action (mainly local) of testosterone and thereafter undergo differentiation into the epididymides, vasa deferentia, and seminal vesicles; these pro-

527

528

SECTION VII / ENDOCRINOLOGY

Figure 57-1 Embryologic development of the gonads. (A) The primordial germ cells are located in the wall of the yolk sac in early embryogenesis. (B) At about the 5th–6th week of human gestation, germ cells migrate by ameboid movement from the coelomic epithelium, along the dorsal mesentery of the hindgut, to the genital ridge, the site of the future gonad. (C) During formation of the indifferent (bipotential) gonad, the primordial germ cells infiltrate the primary sex cords, which subsequently become the seminiferous tubules of the testis, or the primordial follicles of the ovary. (D,E) Formation of the testis continues after arrival of the germs cells, although these are not required for the process to occur. Probably under the direction of the newly differentiated Sertoli cells, the sex cords differentiate as the seminiferous tubules, in which germ cells are embedded. Morphologic development of the testis is complete by 8 weeks’ gestation. (F,G) Formation of the ovary occurs later than that of the testis, at around week 10 of gestation, and does require the presence of germ cells. The primary sex cords develop into primordial follicles at about week 16 of gestation.

cesses occur between approximately 8 and 13 weeks’ gestation. Dihydrotestosterone does not appear to mediate these processes, since the enzyme required for its production is not expressed in Wolffian tissues at the time of their differentiation. In parallel to the masculinization of the internal genitalia represented by Wolffian development, a process of “defeminization” of the genital ducts occurs as the Müllerian ducts regress under the influence of locally acting AMH secreted by testicular Sertoli cells, at around 8–10 weeks. The Müllerian ducts are obliterated by the 11th week of gestation, the only remnant of their existence in the 46,XY fetus being the prostatic utricle. Absence of one testis results in retention of the Müllerian structures and only limited Wolffian development on that side, indicating that the effects of AMH and testosterone are mediated to some extent in a paracrine fashion.

In the absence of testicular secretions, as in the normal 46,XX fetus, the inverse set of genital tract developmental processes occurs. Without local androgen action, the Wolffian ducts regress between about weeks 9 and 13. Similarly, in the absence of AMH the Müllerian ducts are permitted to develop. Their upper portions form the Fallopian tubes. The lower sections fuse and differentiate as the uterus and upper part of the vagina. The lower portion of the vagina derives from the urogenital sinus, which remains patent in the absence of androgen action. Just as the gonads and internal genitalia are indistinguishable between the sexes for the first few weeks of life, so it is for the external genital primordia (Fig. 57-3). In the 4th week of gestation the external genitalia of both karyotypic sexes are represented simply by a midline protuberance—the genital tubercle. By week

CHAPTER 57 / SEX DETERMINATION AND DIFFERENTIATION

529

Figure 57-2 Embryologic development of the internal genitalia. (Indifferent Stage) The primitive internal genitalia at 7 weeks’ gestation, showing the presence of both paramesonephric or Müllerian (female) and mesonephric or Wolffian (male) duct systems. (Female Development) In the female fetus between weeks 9 and 13 of gestation there is regression of the Wolffian ducts (shown with broken lines) with differentiation and development of the Müllerian ducts (shaded). The upper portions of the Müllerian ducts form the Fallopian tubes; the lower portions fuse to form the uterus, cervix, and upper part of the vagina. The urogenital sinus remains patent in the absence of androgen action, forming the lower part of the vagina. (Male Development) Differentiation of the Wolffian ducts in the male fetus occurs under the influence of testosterone between 9 and 13 weeks of gestation. These develop as the epididymis, vas deferens, and seminal vesicles (shaded). By week 11, the Müllerian ducts are obliterated by a process of apoptosis that occurs under the direction of AMH secreted by Sertoli cells (broken lines).

6 (still indifferent) two medial folds, the urethral folds, flank the urogenital groove, and two larger folds, the labioscrotal folds, are present laterally. Under the influence of androgen action (primarily dihydrotestosterone), the genital tubercle elongates to form the body of the penis, and the urethral folds fuse ventrally from behind forward, to form the penile urethra. The labioscrotal swellings or folds grow toward each other, fusing in the midline to form the scrotum. Dihydrotestosterone (DHT) also induces the urogenital sinus to differentiate as the prostate, and inhibits the formation of the vesicovaginal septum. These processes are completed by week 12 of gestation. The testes migrate from their original lumbar location to the level of the internal inguinal ring above the scrotum, between 12 and 24 weeks’ gestation. Descent of the testes through the inguinal ring and into the scrotum begins around week 28, and in most infants is completed by term. In the absence of significant androgen action, such as in the normal female fetus, the genital tubercle elongates only slightly to

form the clitoris (Fig. 57-3). The urogenital sinus remains open and the vesicovaginal septum forms between the genital and urinary portions of the urogenital sinus, so that the urethra opens anteriorly and the vagina posteriorly. The vestibule of the urogenital sinus is bordered laterally by the urethral folds, which fail to fuse and instead develop as the labia minora. Further laterally, the labioscrotal swellings enlarge somewhat but also remain unfused, forming the labia majora. There is minor fusion posteriorly, forming the posterior commissure, and anteriorly, producing the mons pubis. These events occur from about week 7 to week 12 of gestation. The timetable of gonadal and genital differentiation is shown in Fig. 57-4. GENETIC AND HORMONAL CONTROL OF SEX DETERMINATION AND DIFFERENTIATION A complex interplay of genes, hormones, transcription factors, and receptors is required for normal sex determination and differentiation (Fig. 57-4; Table 57-1). The primary event governing the path down which morphologic sex differentiation proceeds is the development of

530

SECTION VII / ENDOCRINOLOGY

Figure 57-3 Differentiation of the external genitalia. (Top) At week 6 of gestation the external genitalia are undifferentiated, consisting of the midline genital tubercle, two medial urethral folds flanking the urogenital slit, and two larger labioscrotal swellings laterally. (Left) In the absence of androgen action the genital tubercle becomes the clitoris and the urogenital sinus remains patent. The vesicovaginal septum forms to separate the urethra from the vagina, located posteriorly. The urethral folds develop as the labia minora, whereas the labioscrotal folds enlarge slightly, remaining unfused, as the labia majora. (Right) In the male fetus in the presence of androgens (mainly DHT), the genital tubercle elongates to become the body of the penis, and the urethral folds fuse in the midline to form the penile urethra. The labioscrotal swellings fuse and enlarge to become the scrotum. These processes are completed approximately between weeks 8 and 12 of gestation.

the testis. In the 1930s the elegant experiments of Jost determined that maleness was a state imposed on the fetus that would otherwise develop as a female, because development of the ovary and female phenotype occur when the fetus is not exposed to the influences of specific “maleness-determining” genes. SEX AND TESTIS DETERMINATION The process of testis development appears to be controlled by a switchlike mechanism that involves the Y chromosomal SRY gene and its encoded protein, and, no doubt, other genes and proteins that either regulate or are regulated by SRY. However, for SRY to function, it must have a target—that is, there must first be development of the gonadal primordium. A candidate gene that may regulate the very earliest stages of gonadal development is the orphan nuclear receptor or transcription factor, steroidogenic factor 1 (SF-1). Although its exact role is still being elucidated, SF-1 appears to be a key factor in development of the reproductive tract, at least in rodents. Analysis of studies in mice with deletions for the murine homolog of the SF-1 gene reveals that SF-1 is required in both sexes for develop-

ment of the bipotential primordial gonads and the embryologic adrenal glands, and for development and function of pituitary gonadotrophs. SF-1 appears to be one of the earliest-acting factors in the sex determination cascade and likely regulates genes encoding factors acting further down the pathway of gonadal development, such as AMH, WT1, DAX-1, LHb, P450scc, and perhaps even SRY itself, either directly or indirectly, via intermediate steps. Colonization of the bisexual gonad by germ cells is one of the earliest events in gonadal development, occurring before any morphologic distinction between the gonads of males and females is detectable. In mice, and probably in other mammals, the c-Kit/ steel receptor–ligand system directs migration of germ cells from the yolk sac to the gonad and may also be responsible for their survival in their new milieu. Once development of the primordial gonadal and genital structures has occurred, the next steps diverge between the sexes. In the karyotypic male a primary “switch” that initiates specific testicular development is the Y chromosome-encoded transcrip-

CHAPTER 57 / SEX DETERMINATION AND DIFFERENTIATION

531

Figure 57-4 Control and timing of gonadal and sexual differentiation. (A) Hypothetical scheme for genetic control of sex determination. The genetic determinants of gonadal and genital differentiation are shown in approximate sequence. c-Kit receptor and steel factor are involved in germ cell migration, one of the earliest events in gonad formation. SF-1 and WT-1 appear to be required for normal formation of the primordial bipotential gonad. SRY directs differentiation of the bipotential gonad as a testis, probably with interaction of other genes such as SOX9. It is presumed that DAX-1 (or a nearby gene) is repressed in male development, since duplication of the locus suppresses testis development. After definitive testis formation has occurred, the testicular hormones AMH and testosterone direct the processes of Müllerian regression and masculinization of internal and external genitalia, each acting via a specific receptor (AMH receptor and androgen receptor). (B) Timetable. In both male and female development the first event of sex differentiation is the migration of germ cells, at around 5–6 weeks’ gestation. This is followed by testicular development in the male. After onset of AMH and testosterone secretion at approximately 7–8 weeks’ gestation, the Müllerian ducts regress and Wolffian ducts differentiate. The pattern of development is parallel, but essentially opposite, and somewhat delayed for female fetuses. In particular, there is a marked lag between testicular and ovarian development.

532

SECTION VII / ENDOCRINOLOGY

Table 57-1 Factors Involved in Sex Determination and Differentiation Gene name

Locus

Protein name

Protein type

SRY

Yp11.3

SRY

SOX9

17q24–25

SOX9

WT1

11p13

WT1

DAX-1

Xp21

DAX-1

HMG box-type ligand-independent transcription factor HMG box-type ligand-independent transcription factor Ligand-independent transcription factor Tumor repressor Orphan nuclear receptor

SF-1/Ftz-F1

9q33

SF1

Orphan nuclear receptor

c-Kit

4q12

c-Kit

Transmembrane tyrosine kinase receptor

slf

12q22

Steel factor

LH/CGR

2p21

LH/CG receptor

StAR

8p11.2

StAR

AMH

19p13.2–13.3

AMH (MIS)

AMHR

12q13

AMH Receptor

SRD51A AR

5p15 Xq11–12

5α-Reductase 2 AR

Membrane-bound and soluble peptides Ligand for c-Kit receptor Seven transmembranespanning G protein-coupled peptide hormone receptor Mitochondrial protein transport factor Glycoprotein homodimer, member of TGF-β family Transmembrane serine– threonine kinase receptor of TGF-β receptor type Microsomal enzyme Ligand-dependent nuclear transcription factor

tion factor, SRY. SRY appears to direct the process of seminiferous tubule organization by Sertoli cells and to regulate transcription of the Sertoli cell glycoprotein AMH; AMH in turn may play a role in directing undifferentiated interstitial cells to develop as Leydig cells. Once testicular differentiation is established, other Y-encoded genes are required to maintain spermatogenesis. A number of other genes have been identified as having involvement in testis development; however, their exact positions in the path of testis development, their functions, and the factors that they regulate or by which they are regulated remain to be elucidated. These include the SRY-related gene SOX9, the Wilms tumor suppressor gene WT1, and the gene encoding the orphan nuclear receptor or transcription factor DAX-1. X-chromosomal sequences are also likely to affect testis development, since the presence of one or more additional X chromosomes (in Klinefelter syndrome and its variants) causes reduced testicular size. At present, the factors directing ovarian development are poorly characterized; however, there is presumably an equally complex network of controls guiding this process. One such gene likely resides on the X chromosome and is probably a dosage-sensitive locus, as evidenced by girls and women with Turner syndrome, whose ovaries regress in the absence of two functional copies of the

Action Bends DNA

Possible targets

Sertoli cell development Transcription regulation

AMH SOX9 P450arom DAX-1 Unknown

Transcription regulation

Unknown

Inhibits SOX9 Regulation of gonadotropes Development of adrenals Development of hypothalamic nucleus Development of gonads/adrenals Regulation of steroidogenesis Migration/proliferation of stem cells ? Suppresses apoptosis As for c-Kit

SOX9 P450scc LHβ DAX-1 AMH — c-Kit receptor

Transduces LH signal Activates cAMP production

—

Mediates cholesterol transport to inner mitochondrial membrane Müllerian duct regression ? Regulates testicular descent Müllerian duct regression ? ? Regulates testicular descent

Unknown

Converts testosterone → DHT Transcription regulation

AMHR — — AMHR ?P450arom

X chromosome. A candidate for this role is DAX-1. Duplication of the region of the X chromosome containing this gene results in development of streak ovaries in 46,XY individuals. In addition, the murine homolog of DAX-1 is expressed in developing ovary at a time consistent with a regulatory role. Another ovarian determinant is likely autosomal, evidenced by the characterization of familial ovarian dysgenesis. The current hypothesis to explain the control of ovarian development is that it occurs when there is active function of as yet unidentified genes that repress testis-determining genes. SEX DIFFERENTIATION Differentiation of the internal and external genitalia along male or female lines depends on the presence or absence of functional testicular tissue. Once function of the testis is established, at around 8 weeks of human gestation, development of the male sexual phenotype is under control of fetal hormones and the receptors that mediate their action. These include fetal luteinizing hormone (LH), the luteinizing hormone-chorionic gonadotropin (LH/CG) receptor, at least five steroidogenic enzymes involved in testosterone biosynthesis, steroid 5α-reductase 2 required for conversion of testosterone to dihydrotestosterone, the androgen receptor, AMH, and the AMH receptor. Development of Wolffian structures requires testosterone, which is produced by testicular Leydig cells after activation of

CHAPTER 57 / SEX DETERMINATION AND DIFFERENTIATION

their cell surface receptor, the LH/CG receptor, a member of the seven transmembrane domain G-protein coupled class of receptors. During the first trimester of gestation the ligand for this receptor is probably CG, produced in large quantity by the placenta. Subsequently, endogenous fetal LH is the primary ligand. The lack of dependence on fetal LH during genital morphogenesis is evidenced by the normal penile development of male infants with hypopituitarism; however, the subnormal penile size of these infants highlights the role of LH and testosterone action in the penile growth that occurs in the third trimester. Action of CG or LH at the Leydig cell LH/CG receptor stimulates testicular steroidogenesis. Sertoli cells may also help regulate Leydig cell secretion in a paracrine fashion. The biochemistry and molecular biology of the steroidogenic enzymes is described in further detail below. The outcome of this multistep process is a high local, and to a lesser extent systemic, concentration of testosterone. Testosterone action is mediated by the androgen receptor (AR), a nuclear receptor-transcription factor found in high concentration in the tissues of the Wolffian ducts and external genitalia. Activation of AR by androgen binding results in interaction of the receptor–ligand complex as a dimer with the promoter regions of target genes, to regulate their transcription. The exact targets of the AR in genital development remain to be determined; however, the gene for the AMH receptor may be one of them. In external genital tissues testosterone (T) is converted by steroid 5α-reductase 2 to dihydrotestosterone, which has greater affinity for the AR. The same molecular events occur after interaction of either T or DHT with AR. In the absence of high androgen concentration, as in the female or androgen-deficient fetus, there is insufficient activated AR to induce transcription of target genes required for stabilization and development of the Wolffian system and for masculinization of the external genitalia. The female fetus bears the same AR as the male; thus, it is primarily the available concentration of androgens that is the major determinant of morphogenesis. AR function is described in further detail in Chapter 60. Müllerian duct regression is induced by the Sertoli cell glycoprotein hormone, AMH, a member of the transforming growth factor β (TGF-β) class. The stimuli for AMH secretion by Sertoli cells are largely unknown, although there is evidence for regulation by both SRY and SF-1. The actions of AMH require a specific receptor, the AMH type II receptor (AMHR), a member of the serine–threonine kinase group of transmembrane proteins. Action of AMH at its receptor, which localizes in the mesenchyme surrounding the Müllerian ducts, results in involution of Müllerian structures by apoptosis. Expression of the AMH receptor may be regulated in part by AR. Analysis of gene expression in wild-type and transgenic mice has led to the formation of a number of hypotheses to address the complex mechanisms of sex determination. Although these hypotheses are continually changing as more is learned about the timing and patterns of gene expression, and as new genes are discovered, the following represents a working hypothesis: SF-1 and WT1 are required to induce development of the primordial, bipotential gonad in both sexes; once this rudimentary structure is formed, specific sex-determining genes come into play; SOX9 is required for Sertoli cell development and testis differentiation. However, in the female, SOX9 is suppressed by DSS/DAX-1. In the male, SRY represses DSS (which is in single copy in the normal male), allowing expression of SOX9 and subsequent testis differentiation (Fig. 57-4).

533

Estrogens and their receptor do not appear to play a significant role in mammalian sex determination and differentiation in either sex. This is evidenced by the fact that during normal fetal development males and females are exposed to equally high levels of maternal estrogens. Similarly, there is no phenotype associated with estrogen deficiency or inability to respond to estrogen in utero: transgenic mice with targeted disruption of the estrogen receptor, both males and females, are phenotypically normal at birth. Their fertility is, however, reduced.

DISORDERS OF SEX DETERMINATION AND DIFFERENTIATION GENERAL BACKGROUND AND MOLECULAR PATHOPHYSIOLOGY Abnormalities of sex determination and differentiation comprise two major clinical groups: disorders of gonadal development, with secondary effects on genital development, and defects of genital morphogenesis in the presence of normal gonads. The terminology is sometimes confusing: sex reversal refers to the condition in which the individual’s genetic sex opposes the gonadal (and therefore generally the phenotypic) sex. These are individuals with 46,XY karyotype who have no testes (in their place usually are streak gonads) and have a female (or ambiguous) phenotype, and those with 46,XX karyotype who have testes and male (or ambiguous) phenotype. Pseudohermaphroditism (either male or female) refers to conditions in which karyotype and gonad are congruous; however, there is a discrepancy between the gonadal and the phenotypic sex: individuals with 46,XY karyotype and testes whose phenotype is female or ambiguous and those with 46,XX karyotype and ovaries whose phenotype is male or masculinized. Infants of either karyotypic sex with these latter disorders have phallic development that ranges from a completely formed penis to a diminutive clitoris; the labioscrotal region may be fully fused and rugose, or bifid and smooth; the internal structures may be mainly Müllerian (in the absence of AMH action), mainly Wolffian (in the presence of local androgen action), or a combination of the two. The genital morphology essentially reflects fetal production of, and response to, androgens and AMH during the critical period of gestation. In contrast to disorders of testicular development and function, disturbance of ovarian development or function does not adversely affect fetal development of the normal female internal or external genital phenotype. However, exposure to supranormal androgen levels during gestation can induce masculinization of a female fetus, but has no detrimental effect on male fetal development. The disorders of sex determination and differentiation are summarized in Table 57-2. Approach to diagnosis and management is discussed at the conclusion of the chapter. Because of the multiplicity of enzymes, hormones, receptors, and transcription factors involved, there are numerous opportunities for the usually well-coordinated processes of sex determination and differentiation to go awry. Hormones have no intrinsic action and must act via specific receptors. Thus hormone deficiencies are expressed by lack of function of the corresponding receptor. In general, steroid hormone deficiencies, resulting from defects of genes encoding biosynthetic enzymes, are manifest only in the presence of two defective gene alleles and are inherited as autosomal recessive traits. Compound heterozygosity occurs when the two alleles encoding an autosomal recessive trait each carry a different mutation. The functional outcome depends on the severity of the mutations independently, as well as on their interaction.

534

SECTION VII / ENDOCRINOLOGY

Table 57-2 Disorders of Sex Determination and Differentiation Sex determination disorders 46,XY sex reversal (46,XY karyotype with ovaries or streak gonads) SRY deletion/mutation SOX9 mutation DSS duplication WT1 deletion/mutation 46,XX sex reversal (46,XX karyotype with testes) XXY+—SRY translocation XXY–—probable mutation of downstream regulator of testis development True hermaphroditism (various karyotypes with ovotestes) Chromosomal anomaliesa Sex differentiation disorders 46,XY pseudohermaphroditism (46,XY karyotype with testes and female or ambiguous internal or external genitalia) Impaired testosterone production LH/CG receptor defect (Leydig cell hypoplasia) Defects of testosterone biosynthesis Defect of steroidogenic acute regulatory protein (StAR) 3β-Hydroxysteroid dehydrogenase deficiency 17α-Hydroxylase deficiency 17,20-Lyase deficiency 17β-Hydroxysteroid dehydrogenase deficiency Impaired androgen response 5α-Reductase deficiency Androgen insensitivity syndromes Impaired anti-Müllerian hormone production or action 46,XX pseudohermaphroditism (46,XX karyotype with ovaries and male or ambiguous internal or external genitalia) Fetal androgen excess Congenital adrenal hyperplasia 21-hydroxylase deficiency 11-β hydroxylase deficiency 3-β hydroxysteroid dehydrogenase deficiency Aromatase deficiency Maternal androgen excessa Drugs Adrenal/ovarian tumors aNot

discussed in this chapter.

Receptors that mediate peptide hormone action, such as membrane-associated receptors like the LH/CG receptor, act as transducers of the hormone signal, setting off an intracellular signal cascade. Mutations of transducer-type receptors cause either an increase or a decrease in the activity of the system, depending on whether the mutation is an activating or an inhibitory one: this is illustrated by the contrasting effects of mutations in the LH/CG receptor, inhibitory mutations being associated with the syndrome of Leydig cell hypoplasia (see below), whereas activating mutations are associated with the syndrome of familial male precocious puberty (see Chapter 59). Defects of these receptors are expressed only in the presence of two defective gene alleles. Nuclear transcription factor receptors, such as those for steroid hormones, interact directly with target genes to modify their transcription, usually in a ligand-dependent fashion. Absence of ligand or defective ligand binding translates to absent or reduced activity of the ligand-dependent transcription factors; complete absence of a transcription factor (such as occurs when the encoding gene is deleted) not surprisingly results in lack of transcription of target genes. A second class of transcription factors appears to be ligand independent (e.g., SRY, WT1). Mutations that alter protein–DNA interactions between transcription factors and their target genes may result in decreased, increased, or sometimes promiscuous

transcriptional activation, the latter being caused by loss of DNA-binding specificity. In the case of a heterozygous mutation, the abnormal protein produced from the mutant gene may interfere with the action of the normal protein produced from the wild-type allele of the gene, a process referred to as the “dominant-negative” effect. In this situation, the mutant protein impairs the action of the normal one, either by blocking access of the normal factor to its target DNA, by forming inactive dimers that are unable to bind the target DNA sequence, or by sequestering other critical transcription factors because of disturbed protein–protein interactions. Defects in genes encoding intracellular receptors and transcription factors are usually considered dominant conditions, as generally only one mutant allele is required for detrimental effect. An exception is the AR, defects of which are inherited in an X-linked recessive fashion. DISORDERS OF SEX DETERMINATION: SEX REVERSAL AND TRUE HERMAPHRODITISM This section describes disorders of genes known or thought to be involved in development of the gonads. Syndromes associated with sex chromosome aneuploidy are detailed in Chapter 58. Two major classes of disorders will be discussed—those involving genes located on the sex chromosomes, and those involving autosomal genes. The general clinical features of sex reversal and true hermaphroditism are described

535

CHAPTER 57 / SEX DETERMINATION AND DIFFERENTIATION

Table 57-3 Disorders of Sex Determination Disorder

Frequency

XX male

1:20,000

True hermaphrodite

1:20,000

XY gonadal dysgenesis Partial Complete (Swyer syndrome)

1:100,000

Phenotype Male Ambiguous (20%) Ambiguous

Ambiguous Female

initially and are summarized in Table 57-3. Clinical features unique to specific gene defects are described in the relevant subsections. General Clinical Features of Disorders of Sex Determination The archetypal and most common form of sex reversal is 46,XX maleness. Approximately 1:20,000 men has a 46,XX karyotype. The phenotype of most affected individuals with 46,XX maleness is similar to that seen in Klinefelter syndrome (47,XXY): structurally normal testes are present (sometimes cryptorchid) and the internal and external genitalia are male. Testicular size and histology are normal in infancy. Pubertal virilization occurs to a greater or lesser degree, but like those with Klinefelter syndrome, affected postpubertal individuals have small testes and are azoospermic, presumably because of the detrimental effect of the extra X chromosome. Histologically the testes have atrophic, hyalinized seminiferous tubules and Leydig cell hyperplasia. Up to 20% of 46,XX individuals with testes have subnormal masculinization, manifest by cryptorchidism, hypospadias, or frank genital ambiguity. 46,XX males are taller than average for females, but shorter than 46,XY males, presumably because of absence of stature-determining genes located on Yq. True hermaphroditism defines the condition in which there is coexistence of ovarian and testicular tissue in the same individual, usually in the same gonad (ovotestis), less often in opposite gonads. Approximately 70% of affected individuals have a 46,XX karyotype and some cases may represent a variant form of 46,XX maleness. Without thorough histologic evaluation of the gonads it may be impossible to distinguish undervirilized 46,XX males from 46,XX true hermaphrodites. In addition, true hermaphroditism and complete sex reversal can coexist within the same family. Approximately 20% of individuals with true hermaphroditism have chromosomal mosaicism for 46,XX/46,XY or less commonly 46,XX/47XXY; a small number of cases have a pure 46,XY karyotype and a few with 47,XXY karyotype have been reported. Notably, the testicular portions of the gonads are dysgenetic, with interstitial fibrosis and rare or absent spermatogonia, although the ovarian portions are histologically normal. Malignant degeneration of the gonad is reported in approximately 5% of cases. Most affected individuals with true hermaphroditism have ambiguous genitalia; however, the phenotypic spectrum is broad: for example, one case of a fully masculinized boy with bilaterally descended ovotestes has been reported. A hallmark, although not universal feature is asymmetric genital development—Müllerian structures

Gonad type

Karyotype

SRY

Testis

46,XX

SRY+ SRY–

Ovotestis

46,XX (70%) 46,XX/46,XY (20%) Other (46,XY rare) 46,XY

SRY– SRY+ (few)

Testis (dysgenetic) Streak

(90%)

SRY normal (90%) SRY mutation SRY deletion

(which may include a full or hemiuterus) on one side, and Wolffian structures on the other; a gonad-containing hemiscrotum on one side (more often the right) with a flat, empty labium majorum on the other. In general the pattern of development reflects the predominant functional nature of the gonad on the ipsilateral side. In association with an ovotestis, the internal genitalia may show elements of both Müllerian and Wolffian origin on the same side. The pattern of pubertal development reflects the function of the gonads, and fertility is not uncommon. The disorder that represents the converse of 46,XX maleness is complete gonadal dysgenesis in a 46,XY female (Swyer syndrome; XY gonadal dysgenesis); however, it is much less common than the former condition, occurring in only 1:100,000 females. Affected individuals have streak gonads with female internal and external genitalia, although significant phenotypic variation is found within affected families. There is a high incidence of gonadal neoplasia (gonadoblastomas and germinomas) in the streak gonads. Because of deficient estrogen production, breast development is poor and these individuals often present with delayed puberty or primary amenorrhea; gonadotropin concentrations are in the castrate range. Pubic hair is usually present. Probably because of the presence of Y chromosomal staturedetermining genes, affected women are of normal to tall stature compared with 46,XX females. They have no physical stigmata of Turner syndrome. However, individuals with deletion of the short arm of the Y chromosome (including the SRY locus), in addition to the general features described above, have certain features of Turner syndrome, such as lymphedema. The disturbances of testicular development (complete or partial gonadal dysgenesis) are associated with genital development that essentially reflects the functional state of the gonads at the critical period of sex differentiation. In contrast to the numerous disorders of testicular development, clearly defined disorders of ovarian development (other than Turner Syndrome) are rare. One example of abnormal ovarian development likely resulting from a single gene defect is 46,XX gonadal dysgenesis. This disorder does not cause sex reversal, internal and external genitalia being those of a normal female. Affected young women present with failure of normal pubertal development, without phenotypic features of Turner syndrome; they are of normal height for females, and shorter than those with 46,XY gonadal dysgenesis, highlighting the role of Y chromosomal genes in stature determination. Their gonads are streaks;

536

SECTION VII / ENDOCRINOLOGY

however, in contrast to those with 46,XY gonadal dysgenesis, no increase in incidence of gonadal neoplasia has been found. Gonadotropins, particularly follicle-stimulating hormone (FSH), are markedly elevated. A familial form of 46,XX gonadal dysgenesis, in which affected individuals also have sensorineural deafness, is known as Perrault syndrome. In a large Finnish study of familial, apparently autosomal recessive, cases, the disorder is associated with mutations of the gene encoding the FSH receptor. DISORDERS OF SEX DETERMINATION INVOLVING LOCI ON THE SEX CHROMOSOMES Sex-Determining Region of the Y Chromosome—SRY Gene The existence of a Y chromosomal “maleness-determining” gene was postulated in the 1930s, and in the 1960s was designated the “testis-determining factor” (TDF). Many candidates were proposed and rejected over the years. In 1990 the existence of such a gene was confirmed, with the discovery of the SRY gene (Sry in the mouse). Genetic and Molecular Pathophysiology The study of individuals with the syndromes of 46,XY gonadal dysgenesis and 46,XX maleness was the catalyst for the eventual localization and characterization of the SRY gene. The human SRY gene is a 1.0-kb intronless gene located just centromeric to the pseudoautosomal region of Yp (Yp11.3). The encoded protein acts as the dominant inducer of testis development in mammals. However, despite its preeminent role in testis determination, other Y-encoded sequences appear to regulate SRY expression. Expression of Sry occurs in the gonadal tissue of fetal mice in the earliest period of specific testis formation, being first detected at embryonic day 10.5 in pre–Sertoli cells in the developing gonadal ridge. This finding suggests that these cells are integral to the process of testis development; indeed, one of the primary events in testis development appears to be induction of Sertoli cell differentiation. Sry expression peaks at day 11.5 and declines once testicular development is established, at day 12.5. Notably Sry expression is quite restricted to the urogenital ridge, with no evidence for expression in any other tissue at this critical time. Sry expression and Sertoli cell differentiation are followed by testis differentiation, most notably, Sertoli cell-regulated organization of seminiferous tubules and Leydig cell differentiation. The human SRY gene comprises a single coding exon that predicts a 203-amino acid protein. Approximately the middle onethird of the protein represents the HMG (high-mobility group) domain, which endows SRY with its sequence-specific binding to a target nucleotide sequence—5'-AACAAAG-3'. The presence of the HMG box places SRY within a family of DNA-binding, transcription-regulating proteins, the HMG box proteins. The target genes of SRY are largely unknown, but may include those encoding AMH and the steroidogenic enzyme CYP11. Nuclear magnetic resonance spectroscopy analysis of the interaction between the SRY protein and the AMH gene promoter reveals that SRY binds in a sequence-specific manner in the minor groove of the DNA helix, intercalating the isoleucine residue at position 68 between the DNA strands, causing some unwinding of the strands and 70–80° bending of the DNA. In addition, SRY displays sequence-independent binding at four-way DNA junctions, which are intrinsically bent. Despite the absence of an activation domain, SRY–DNA interaction mediates transcriptional regulation, perhaps by bringing together distant DNA sequences or by facilitating access of other important transcription-regulating proteins to regulator sequences within specific testis-inducing target genes. It has been suggested that the spatial arrangement of the nucleopro-

tein complex organized by SRY is critical for the expression of these genes. There are two main subtypes of 46,XX maleness and 46,XX true hermaphroditism—XXY+ and XXY–. The majority of patients with complete 46,XX maleness are XXY+, resulting from translocation of all or part of the distal end of the Y chromosome, containing SRY, to the short arm of one X chromosome during paternal gamete meiosis (Fig. 57-6). In contrast, the majority of patients with 46,XX true hermaphroditism are XXY–. The molecular-phenotypic inference from these findings is that the presence of Y chromosome material results in a greater degree of masculinization. Two specific hot spots for Yp-Xp recombination within areas of high X-Y sequence homology have been reported to account for more than 50% of such recombination events. The presence of SRY in the genome of a 46,XX male can be confirmed by Southern blot or polymerase chain reaction (PCR) analysis. 46,XX males with unambiguous masculinization, more than 90% are SRY positive, whereas the converse holds for 46,XX males with genital ambiguity, approximately 5–10% of whom are positive for SRY. A small number of 46,XX true hermaphrodites have been found to be SRY positive. Nonrandom inactivation of the SRY-bearing X chromosome is the postulated explanation for the presence of ovarian and testicular tissue in the same gonad in individuals with SRY-positive true hermaphroditism. The coexistence of 46,XX complete maleness and 46,XX true hermaphroditism in a number of families is also postulated to reflect variations in the pattern of inactivation of the SRY-bearing X chromosome between affected individuals, affecting SRY expression. 46,XX maleness and true hermaphroditism are clearly genetically heterogeneous conditions. The finding that approximately 5–10% of 46,XX males with testes are completely negative for all Y-encoded sequences (XXY–) indicates that non-Y sequences must be responsible for testis determination in these cases, as discussed below. The absence of spermatogenesis in 46,XX males has been attributed to two factors, the presence of the extra X chromosome, and the absence of Y chromosomal genes involved in spermatogenesis. In parallel to the finding of SRY sequences in the majority of 46,XX males, it would be predicted that most sex-reversed 46,XY females would have defects of the SRY gene. However, in general this has not proven to be the case; 85–90% of 46,XY females do not have demonstrable defects within the coding region of SRY. Suggested explanations for this finding include the theoretical presence of mutations located outside the HMG box-encoding region of the gene, such as in sequences important for regulation of SRY expression, of inactivating mutations in (as yet uncharacterized) upstream regulators of SRY, or of mutations that induce constitutive activity in genes usually negatively regulated by SRY. SRY mutations are more common in 46,XY sex-reversed females with complete rather than partial gonadal dysgenesis. In the minority of individuals in whom an SRY defect can be detected, a variety of cytogenetic and molecular abnormalities have been reported. Phenotypic variation between individuals harboring the same mutation, both within and between kindreds, has been described. Variable penetrance is suggested as the explanation for the finding of the same SRY gene mutation in both affected and unaffected 46,XY individuals in certain families. Germline mosaicism for the SRY mutation may be associated with fertility in the fathers of affected individuals. Reported defects include deletion of Yp, isolated deletion of SRY, and at least 20 single-base mutations within the HMG box-encoding region of the SRY gene

CHAPTER 57 / SEX DETERMINATION AND DIFFERENTIATION

537

Figure 57-5 Clinical examples of male pseudohermaphroditism. These infants all have 46,XY karyotype and show varying degrees of subnormal masculinization, more severe moving from left to right. (A) This baby has a mild androgen receptor defect. The defect of androgen action is manifest by failure of fusion of the urethra, resulting in penile hypospadias. There is also cryptorchidism with subnormal scrotal development. (B) This infant has a small penis with perineoscrotal hypospadias and a completely bifid scrotum. (C) This child has marked genital ambiguity: the phallic structure is intermediate between a clitoris and a penis, and the labioscrotal folds are smooth and unfused. Nevertheless, the testes are present in the labioscrotal folds. (D) The only evidence of androgen action in this 46,XY child is the mildly enlarged clitoris. (E) Complete lack of androgen action in a 46,XY individual (complete androgen-insensitivity syndrome; a patient with Leydig cell LH receptor defect would have a similar appearance).

Figure 57-6 The SRY gene. (Upper) Mechanism of translocation of the SRY gene onto the X chromosome. During paternal gamete meiosis there is pairing of the X and Y chromosomes in their pseudoautosomal regions (shaded) located on their short arms, with obligatory crossing over of homologous sequences. Genes in this region escape X inactivation. The SRY gene is located just proximal to the pseudoautosomal boundary; therefore, an unequal crossover between the Y and X chromosomes could result in translocation of Y chromosomal material (shown here as the whole of the pseudoautosomal region), including the SRY gene, to the distal X chromosome. These recombination events occur with increased frequency at two hot spots that contain areas of high X-Y sequence homology. (Lower) Mutations in the human SRY gene. There are many missense mutations in the SRY gene, all occurring within the region encoding the conserved HMG box. These are shown below the gene, using the single letter amino acid code designation (see Fig. 8B). Nonsense mutations introducing premature termination codons are boxed; frameshift mutations are underlined. Only one mutation has been reported outside the HMG box-encoding region. This was a nonsense mutation predicting truncation of the SRY protein by 42 amino acids.

538

SECTION VII / ENDOCRINOLOGY

(Fig. 57-6). Severe SRY gene mutations resulting in failed protein expression include a frameshift because of a 4-bp deletion, and a nonsense mutation introducing a premature termination codon into the message. Only one mutation has been reported outside the SRY HMG box-encoding region, this being a nonsense mutation in the 3' region of the gene resulting in loss of 42 amino acids from the carboxyl-terminus of the protein. The majority of SRY mutations reported to date produce nonconservative amino acid substitutions at highly conserved sites within the HMG domain. In vitro analysis of mutant SRY proteins containing amino acid substitutions reveals abnormal DNA binding and bending: some mutant SRY proteins bind DNA less avidly but bend DNA normally, whereas others bind with near-normal affinity but bend the DNA to a different angle. The most severe defects appear to arise de novo, whereas the milder ones are compatible with fertility and thereby transmission to offspring. This is further evidenced by the fact that less dramatic missense mutations cause partial, rather than complete, gonadal dysgenesis, indicating that the mutant SRY protein likely retains some transcription-activating function. SRY gene mutations have also been reported in a few cases of 46,XY true hermaphroditism. One example was a postzygotic somatic mutation, because the affected individual had only the wild-type SRY sequence in leukocyte DNA but had both wild-type and mutant alleles in DNA derived from the gonads. Although SRY is probably the key positive regulator of testis determination, there are clearly other upstream and downstream factors that must be activated or repressed to allow testis development to occur. This is implied by the following: (1) the majority of 46,XY females with gonadal dysgenesis have an intact SRY gene; and (2) some 46,XX males with testes are completely Y negative, indicating that non-Y sequences must be responsible for testis determination in these cases. Furthermore, the tight ontogeny of Sry expression during embryogenesis in the mouse suggests the presence of a “controller,” that is, at least one upstream regulator of expression of the critical switch gene. In this regard, it is notable that a potential binding site for the early growth response-Wilms tumor 1 family of transcription factors has been reported upstream of the SRY open reading frame. Dosage-Sensitive Sex Reversal–Adrenal Hypoplasia Congenita Critical Region on the X Chromosome, Gene 1—DAX-1 The DAX-1 gene is implicated in development of the reproductive tract by its involvement in two unusual X-linked syndromes: dosage-sensitive sex reversal (DSS) and adrenal hypoplasia congenita-hypogonadotropic hypogonadism (AHC/HHG). Clinical Features Karyotypic males with a duplication of the region of the X chromosome containing the DAX-1 gene have typical features of 46,XY sex reversal, with streak gonads and a female phenotype. In contrast, individuals with deletion of this region have normal testes and normal male genitalia, although cryptorchidism has occasionally been reported. These males have congenital hypoplasia of the adrenal glands and many die unexpectedly in infancy or early childhood of undiagnosed adrenal failure, generally ascribed to “Addison disease.” Those who, with replacement therapy, survive childhood also commonly have hypogonadotropic hypogonadism resulting in failure of pubertal virilization. This compound syndrome is known as adrenal hypoplasia congenita–hypogonadotropic hypogonadism (AHC/HHG) (see Chapter 59). One report also describes late development of high-frequency hearing loss in affected individuals. Female carriers

of AHC/HHG have no clear abnormality of either adrenal gland development or gonadotropin secretion; however, the author has received some reports of menstrual irregularity and early menopause. Genetic and Molecular Pathophysiology The occurrence of sex reversal in the presence of two copies of sequences in the Xp21 region led to the designation of this locus as the dosagesensitive sex reversal region (DSS). Subsequently it was determined that this same region was deleted in individuals with adrenal hypoplasia congenita (AHC). The DAX-1 gene (DSS adrenal hypoplasia congenita locus on the X chromosome, gene 1) was cloned by analysis of this region in such patients. A well-described cytogenetic microdeletion syndrome encompassing the Xp21 region includes deletions of DAX-1, the dystrophin (DYS) gene (deletions of which cause Duchenne muscular dystrophy), and the gene encoding glycerol kinase (GK). The 5.0-kb DAX-1 gene located at Xp21.3–21.2 encodes a new, although probably rather primitive, orphan member of the nuclear hormone receptor superfamily of transcription factors. The gene demonstrates evolutionary conservation, homologous sequences being detected in an X-linked pattern of distribution in many species. It comprises two exons and one intron that encode a 470-amino acid protein. The amino-terminal region of the protein, encoded by the first exon of the gene, contains a novel DNA-binding domain that has little homology with that of other members of the family. However, the ligand-binding domain, encoded in part by the first and in part by the second exon of the gene, has strong similarities with those of the retinoic acid and retinoid X receptors. Expression of the homologous gene in the mouse occurs in the first stages of gonadal and adrenal differentiation and in the developing hypothalamus. In human studies, DAX-1 is expressed in adult testis and adrenal cortex, in the hypothalamus and pituitary gland, and at low level in adult ovary and liver. The expression of DAX-1 at all levels of the hypothalamic–pituitary–adrenal and gonadal axes supports its role in the coordinated development of the adrenal and reproductive systems and suggests that the DAX-1 protein may be directly or indirectly involved in gonadal regulation of hypothalamic–pituitary function. The promoter region of the DAX-1 gene contains a binding site for SF-1, linking the two transcription factors in the control of development of steroidogenic tissues and suggesting that SF-1 may directly regulate DAX-1. Furthermore, the fact that disruption of SF-1 inhibits development of gonads and adrenals (see below), whereas deletion of DAX-1 disturbs only adrenal development, suggests that DAX-1 is downstream of SF-1 in this cascade. Duplication of the DAX-1-containing region of the X chromosome (DSS) results in classic sex reversal, with ovarian development in 46,XY individuals. Overexpression of this region impairs testis formation despite the presence of a functional SRY gene; suggesting a “toxic” effect on testicular development. In contrast, carrier females with duplication of this region may be fertile, indicating that double dosage of this locus does not impair normal ovarian differentiation. Whether the sex-reversing gene is DAX-1 itself awaits analysis in transgenic animals. Recently, a cluster of potential candidates for the role of sex-reversal gene has been isolated from a region within 50 kb of DAX1, including MAGE-Xp or MAGEL1 (melanoma associated antigen gene), DAM6, and DAM10 (DSS/AHC critical interval genes belonging to the MAGE superfamily). These genes are expressed in adult testis and the latter two are expressed in lung tumors. Their developmental roles remain to be characterized.

CHAPTER 57 / SEX DETERMINATION AND DIFFERENTIATION

Although adrenal development does not proceed beyond the fetal stage, disruption of DAX-1 does not prevent testicular development in males, indicating that this gene does not appear to play a role in normal testis differentiation. However, potential interactions between DAX-1 and the primary testicular switch, SRY in testis formation remain to be examined. Deletion of DAX-1 has been reported in a number of patients with isolated AHC, whereas frameshift, premature termination, codon deletion, and missense mutations have been found in DAX-1 genes of individuals with combined AHC/HHG, suggesting a possible dominant-negative effect of a retained but structurally abnormal protein on hypothalamic–pituitary function. If DAX-1 is indeed the sex-reversal gene, it may represent a link between normal ovarian and testicular differentiation pathways. Different genes must be activated or repressed to allow differentiation of ovary rather than testis or vice versa. It is conceivable that DAX-1 (or a related locus) may act as a repressor of a gene or genes involved in testis determination; thus, absence of the gene product is inconsequential for testis development, but excess is detrimental. Loss of one allele likely has minimal effect on ovarian function because there is an active locus on the other X chromosome. It has been suggested that this locus may be a remnant of an ancestral sex-determining system that operated by dosage before the process of X-chromosome inactivation evolved. SEX DETERMINATION DISORDERS INVOLVING LOCI ON THE AUTOSOMES Analysis of a number of informative families with 46,XXY– maleness, true hermaphroditism, or both reveals transmission patterns consistent with an autosomal dominant trait. One postulated mechanism to explain sex reversal in such cases is the presence of an upregulatory mutation in an as yet uncharacterized downstream regulator of testis determination. SRY Homeobox-like Gene 9—SOX9 After the discovery of the SRY gene, many related genes were soon identified. Those that encode proteins with more than 60% similarity to the SRY HMG box region have been termed SOX genes. One such gene, SOX9, is significantly involved in the process of testis formation. Clinical Features The syndrome of campomelic dysplasia features a distinctive form of skeletal malformation that is present in 75% of 46,XY individuals affected by sex reversal (ovaries or streak gonads, and female internal and external genitalia). Affected 46,XX individuals have normal ovarian and external genital development. In most cases death occurs in infancy or early childhood because of respiratory compromise. The syndrome is uncommon, estimates of its frequency ranging from 1/5,000–1/ 200,000 live births. Genetic and Molecular Pathophysiology SOX9 was identified by cloning of a chromosomal translocation breakpoint from a sex-reversed patient with camptomelic dysplasia. The gene is located at 17q24–25, in a region termed sex-reversal autosomal 1 (SRA1); its cDNA is approximately 3.9 kb in size. The SOX9 gene displays strong sequence conservation throughout mammalian evolution. Furthermore, the sequence similarity between SOX9 and SRY suggests a possible evolutionary relationship between the two that may represent evolution from a dosage-dependent sexdetermination system to a dominant system. Notably, dosage sensitivity is a feature of many regulatory genes. In the human fetus, SOX9 mRNA is detectable by Northern analysis in brain, liver, and kidney; by in situ hybridization, SOX9 expression is detectable in the testis at week 18 of gestation in the area of the rete testis and seminiferous tubules. In the adult, SOX9

539

message is expressed most strongly in testis. It is also highly expressed in pancreas, prostate, kidney, brain, and the skeleton and at low level in most other adult tissues. In the mouse, the homologous Sox9 gene is highly expressed in fetal skeletal tissues, as well as neural, cardiac, and other tissues. Its expression can be detected in mRNA isolated from whole embryos at day 8.5. Consistent with a role in testis determination, Sox9 expression parallels Sertoli cell differentiation. SOX9 encodes a 509-amino acid protein with features of a transcription regulator: it contains a putative DNA-binding HMG box domain and a proline- and glutamine-rich domain in its carboxyterminal third, similar to activation domains of other transcription factors. In vitro deletion of this latter region destroys the transactivating function of the protein. SOX9 activates transcription via the nucleotide motif recognized by other HMG domain transcription factors, 5'-AACAAAG-3', and may represent another switchlike mechanism, because it appears to act in a dominant fashion, similar to SRY. SOX9 may be acting just downstream of SRY and may represent a target for SRY; however, this remains speculative, because the appropriate HMG box binding motif has not been identified within the SOX9 gene promoter. One postulated relationship between the two transcription factors in testis determination is that SOX9 expression in another cell type is required to complement SRY expression by Sertoli cells. Analysis of this hypothesis will require information regarding the comparative ontogeny of SOX9 and SRY expression. Inactivating mutations in one SOX9 allele have been identified in a number of 46,XY individuals with camptomelic dysplasia and sex reversal. No patient has been reported with mutations in both SOX9 alleles. In general SOX9 mutations appear to occur de novo, providing further evidence of the autosomal dominant nature of the disorder. Mutations reported to date likely destroy protein function; these include frameshifts and nonsense mutations that lead to premature chain termination and loss of large portions of the protein. Thus the phenotype of individuals with these mutations appears to be caused by loss of function of the encoded transcription factor (haploinsufficiency), rather than resulting from a dominant-negative or gain-of-function effect. Missense mutations have not yet been reported. Affected sibling pairs with normal parents have been reported, likely as a result of gonadal mosaicism for the mutant gene. Recently, a number of patients with autosomal sex reversal and campomelic dysplasia have been found to have chromosome 17 breakpoints at least 50–130 kb from the SOX9 locus, suggesting the presence of other genes responsible for the same phenotype, or perhaps a disturbance of SOX9 expression because of mutation of an upstream regulator. Positional cloning of the chromosome 17q breakpoint in one patient with sex reversal in the absence of campomelia (so-called acampomelic camptomelic dysplasia) identified a 3.5-kb cDNA that is expressed in testis but appears not to be translated. It was suggested by the authors that the mRNA itself has a functional role in sex determination. Wilms Tumor 1—WT1 The Wilms tumor 1 (WT1) gene product, a member of the early growth response (EGR) family of transcription factors, is implicated in gonadal and genital development by analysis of human mutations and transgenic mice with a deletion of the gene. Clinical Features Two clinical syndromes are associated with defects of the Wilms tumor 1 gene (WT1). The Denys-Drash syndrome comprises a triad of gonadal dysgenesis, congenital

540

SECTION VII / ENDOCRINOLOGY

nephropathy, and subsequent development of Wilms tumor. Patients with Denys-Drash syndrome have heterogeneous disorders of gonadal and genital development. The gonads range from streak gonads or ovotestes in either 46,XX or 46,XY individuals, to rudimentary or dysgenetic testes in 46,XY patients, or normal ovaries (in a 46,XX patient). Perhaps because of an ascertainment bias, the majority of patients (about 90%) have 46,XY karyotype. Their external genitalia are usually ambiguous, sometimes female, occasionally male. The internal genitalia are often incongruous with respect to the external genital phenotype: Wolffian structures may be present in phenotypic females, Müllerian structures in phenotypic males. Even in 46,XX females, there may be anomalous internal genital development, and one 46,XY patient with dysgenetic testes had neither Müllerian nor Wolffian structures. Gonadoblastomas and granulosa cell tumors have been reported in some cases. The congenital nephropathy (caused by diffuse or focal mesangial sclerosis) generally presents with proteinuria and hypertension in the first year of life, with progression to renal insufficiency and death by 2 years of age. Wilms tumor is reported in more than 50% of cases, most presenting with an abdominal mass in the second year of life. The incidence would probably be higher if the patients survived longer, and in fact most are found to have persistent intralobar renal blastema, which may be the precursor of Wilms tumor. Of interest, the incidence of hypospadias and cryptorchidism is 10-fold higher in patients with bilateral Wilms tumor (distinct from the Denys-Drash syndrome) than in the general population. In the WAGR syndrome, affected individuals, in addition to Wilms tumor and genitourinary abnormalities or gonadoblastoma, have aniridia and mental retardation. This disorder appears to be a contiguous gene defect, caused by heterozygous deletion of the 11p13 region that contains the WT1 gene. Thus, other genes located in this region may also be involved in pathogenesis. Genetic and Molecular Pathophysiology The WT1 gene located in chromosomal region 11p13 is approximately 50 kb in size, contains 10 exons, encoding four distinct mRNA species (of approximately 3 kb), generated by the presence or absence of two alternative splice sites. The fact that all transcripts are expressed at a similar level suggests that each encoded protein makes a significant contribution to normal function. Interactions between the four proteins, each of which may have distinct targets and functions, may be important in the control of cellular proliferation and differentiation exerted by WT1. WT1 expression occurs at a similar time to that of SRY, WT1 mRNA being detectable in pronephric and mesonephric tissues on embryonic day 10.5 in the mouse, at which time SRY expression is detectable in pre–Sertoli cells. By embryonic day 11.5 the nephrogenic cord, condensing metanephric tissue, and urogenital ridge display high levels of WT1 message. In the developing gonad WT1 expression is localized to the stromal cell components, and in mature gonads it is confined to the Sertoli cells of the testis and granulosa and epithelial cells of the ovary. Expression also has been detected in abdominal and lung mesothelium. The WT1 protein is a putative 45- to 49-kDa tumor suppressor with four contiguous Cys2-His2 zinc finger domains (encoded by exons 7–10) and an amino-terminus rich in proline and glutamine, consistent with a role as a transcription factor. WT1 has been shown to have separate domains that subserve transcriptional repression and transcription activation: residues 85–124 encompass the repressor domain, and 181–250 the activator domain. These regions are distinct from the DNA-binding domain and

their activities are probably mediated by protein–protein interaction. The WT1 protein shows homology to the early growth response-1 (EGR-1) gene product and has similar binding sites (5'-CGCCCCCGC-3'). These proteins are expressed early in the cell cycle, at G0 to G1 transition. The target genes for WT1 are at present unknown; however, it is suggested that the normal role of WT1 during embryologic development is to initiate a transcriptional program that controls differentiation of glomerular epithelial cells and gonadal primordium. The absence of both sets of internal genital duct anlagen in one patient would be compatible with a primary role of WT1 in early development of the bipotential internal duct systems. In renal tissue it acts as a tumor suppressor. Examination of transgenic mice homozygous for a knockout mutation of WT1 established a crucial role for WT1 in early urogenital development, with mutant embryos displaying failure of renal and gonadal development. Specifically, at day 11 of gestation, the cells of the metanephric blastema underwent apoptosis, the ureteric bud failed to grow out from the Wolffian duct, and the formation of the metanephric kidney did not occur. The mice were nonviable, probably because of abnormal development of the mesothelium, heart, and lungs. The Denys-Drash syndrome in humans is associated with heterozygous germline mutations in WT1 in approximately 95% of cases examined to date. The mutations cluster within or near the zinc finger (ZF) coding region (exons 7–10, particularly exon 9), most producing amino acid substitutions in ZF2 and ZF3. One mutation, encoding an arginine to tryptophan substitution at position 394 in the third zinc finger, has occurred recurrently, being present in approximately half of the cases reported to date. Other reported mutations include a variety of nonconservative missense, nonsense, frameshift, and splice-junction mutations, most of which appear de novo, since parental WT1 genes are normal. In one case, however, the normal father of an affected child was found to be heterozygous for the mutant allele, a finding that suggested either mosaicism or reduced penetrance of the mutant gene. The disorder is genetically dominant, because no patients have been described with mutations in both alleles of the gene. The WAGR syndrome likely results from heterozygous loss of contiguous genes within the 11p13 region including the WT1 gene, accounting for the aniridia and mental retardation that accompany the Wilms tumor and genitourinary abnormalities in WAGR cases. In patients in whom it has been examined, the mutant allele appears to have arisen in the paternally derived chromosome 11. The dominant phenotype associated with the null mutation indicates dosage sensitivity at this locus—that is, two functional copies of the gene are required to prevent the abnormalities associated with the syndrome. The molecular pathophysiology of WT1 mutations appears to be lack of normal DNA binding by the WT1 protein, because mutant WT1 proteins containing nonconservative amino acid substitutions within the zinc finger region do not bind to the EGR-1 consensus-binding sequence bound by wild-type WT1. Mutations that affect zinc-coordinating cysteine or histidine residues likely prohibit DNA binding by disrupting proper spatial organization of the zinc finger. WT1 appears to play a dominant-positive role in renal and gonadal development. Formation of inactive mutant– wild-type dimers could theoretically produce a dominant-negative effect, presumably via heterodimerization with wild-type WT1 protein. Interestingly, complete deletion of the WT1 gene produces milder genital abnormalities (cryptorchidism and/or hypospadias), than does a mutation that encodes expression of an

CHAPTER 57 / SEX DETERMINATION AND DIFFERENTIATION

abnormal WT1 protein, again suggesting a dominant-negative mechanism of action of mutant WT1. Although WT1 mutations are heterozygous at the germ cell level, tissue from the Wilms tumors of these patients generally demonstrates loss of heterozygosity for WT1, indicating that two mutant WT1 alleles may be required for loss of tumor suppressor function of WT1 (“two-hit” hypothesis). A role for interaction between WT1 and the ubiquitous tumor suppressor, p53, is suggested by an interaction between the two factors in transfected cells. The two may interact by direct protein– protein interaction or via an intermediate, and there is evidence that the tumor suppressor function of WT1 may be dependent on such an interaction, rather than be intrinsic to the protein itself. Steroidogenic Factor 1—SF-1 Defects of steroidogenic factor 1 (SF-1) have not yet been reported in humans; however, the phenotype of transgenic mice with targeted disruption of the homologous gene (fushi tarazu factor-1 [Ftz-F1]) is well-characterized, and provides important insights into aspects of mammalian gonadal development. SF-1 (also referred to as adrenal 4-binding protein [Ad4BP]), appears to hold a strategic position in inducing the development of the reproductive tract. Not only does SF-1 regulate gonadal and adrenal development, it also appears to control development of an important hypothalamic nucleus, and of pituitary gonadotropinsecreting cells. Furthermore, SF-1 regulates expression of the genes encoding three key steroidogenic enzymes: cholesterol sidechain cleavage enzyme, steroid 21-hydroxylase, and the aldosterone synthase isozyme of steroid 11β-hydroxylase. The human gene encoding SF-1 is located at 9q33 and its murine homologue is on mouse chromosome 2. Two distinct proteins— SF-1 and ELP (embryonal long terminal repeat)—are produced by alternative promoter usage and splicing; however, the critical protein with respect to reproductive tract development is SF-1. SF-1 is an orphan nuclear receptor or transcription factor that contains the typical two central zinc fingers and a carboxy-terminal ligandbinding domain. SF-1 is expressed in the urogenital ridge of male and female mice at embryonic day 9–9.5, the earliest stage of organogenesis of the gonads (still indifferent, and before SRY expression); it is subsequently expressed in fetal Sertoli cells and in the primordial cells of the adrenal glands. Consistent with its role in regulation of steroidogenesis, it is expressed before expression of the first enzyme of steroidogenesis, P450scc. Expression in fetal ovaries is quite low, and disappears between embryonic days 13.5 and 16.5, in keeping with the lack of ovarian steroidogenic activity at that time. SF-1 mRNA is detectable in the developing mouse pituitary at embryonic day 13.5–14.5 and can be specifically detected in gonadotrophs. In adult mice SF-1 is expressed in all primary steroidogenic tissues, including all zones of the adrenal cortex, testicular Leydig and Sertoli cells, ovarian theca and granulosa cells, and corpus luteum. There is sexually dimorphic expression of SF-1 postnatally. In male rats testicular SF-1 mRNA levels decline markedly after the first week of life, whereas ovarian SF-1 expression increases in females. The expression of SF-1 appears to be regulated by another factor, probably a helix-loop-helix protein similar to c-Myc, that binds to a region in the promoter of the rat Ftz-F1 gene (homologous to SF-1), designated the E box. Expression of the binding factor precedes that of SF-1. This fact, and its tissue-specific pattern of expression, is consistent with a role as a regulator of SF-1 expression. This factor may be the earliest member of the gonadal development cascade as yet identified.

541

SF-1 functions differently from most other members of the nuclear receptor family, because it appears to bind as a monomer, rather than as the more usual dimer, to a nonpalindromic DNA steroid response element half-site, 5'-AGGTCA-3'. Potential targets for SF-1 include the genes encoding P450scc, DAX-1, LHβ, AMH, and oxytocin, because SF-1 binding sites have been located within the promoter regions of these genes. The functional importance of SF-1 and Ftz-F1 is addressed by studies in transgenic animals. Both male and female mice lacking a functional Ftz-F1 gene had phenotypically female internal and external genitalia and died in the neonatal period of adrenal insufficiency. They failed to develop steroidogenic tissues, having neither gonads nor adrenal glands (the steroidogenic components of adrenal glands and gonads likely have a common embryologic origin in the urogenital ridge). Interestingly, these mice did display mesenchymal thickening in the gonadal ridge area at the earliest stages of gonadal development (embryonic day 10.5), but thereafter the cells of this region underwent apoptosis (programmed cell death), suggesting that the role of SF-1 and Ftz-F1 may be in maintaining rather than initiating gonadal development. These mice also had abnormal development of the ventromedial hypothalamic nucleus, a region important in control of pituitary gonadotropin secretion. Expression of proteins specific to gonadotroph cells, LHβ, FSHβ (follicle-stimulating hormone), and gonadotropin-releasing hormone (GnRH) receptor was absent, implicating SF-1 in gonadotrope development and LH and FSH expression. The presence of female internal genitalia in the null animals of both sexes indicates that the Müllerian ducts did not regress as expected in the males, reflecting lack of AMH action, because of absence of Sertoli cells. In females, expression of SF-1 during the critical period of genital development would be detrimental to normal sex differentiation. SF-1 induction of AMH expression would likely induce Müllerian regression and failure of oviduct and uterine development. The fact that SF-1 expression in developing ovaries is low supports this hypothesis. Interestingly, although SF-1 binds to the promoter of the AMH gene, it is unable to regulate AMH gene expression in cotransfection assays using a heterologous cell line, unless its ligand-binding domain is deleted. This finding suggests a role for a possible ligand for SF-1 in its regulation of AMH expression in vivo, perhaps a Sertoli cell-specific factor. Such a factor could prove to be critical in the sex determination cascade. In evaluating the role of SF-1 in reproductive tract development, a key question arises regarding the possible interactions between SRY and SF-1. The fact that the Ftz-F1 null mice have no gonadal tissue at all, whereas individuals with SRY mutations develop gonads that more or less resemble ovaries suggests that normal SF-1 or Ftz-F1 expression must occur before SRY can be effective in inducing testis determination. Furthermore, expression of the two genes is temporally dissociated, SF-1 being expressed in mouse testis at embryonic day 9–9.5, before transient gonadal SRY expression at day 10.5. However, in the older embryo expression of SRY correlates with sexually dimorphic expression of SF-1. In males SRY promotes upregulation of SF1; absence of SRY in females results in reduced SF-1 expression. In view of the discordant timing of expression of these proteins, this is likely an indirect effect. In this context it has been suggested that SRY could act by silencing a repressor of SF-1, thus facilitating SF-1 expression.

542

SECTION VII / ENDOCRINOLOGY

In summary, SF-1 has multiple key roles in reproductive tract development, controlling development of the gonads and adrenal glands themselves, a hypothalamic center that regulates gonadal function, gonadotropin expression, and production of the two hormones essential for male sex differentiation—testosterone (by regulation of key steroidogenic enzymes) and AMH. c-kit Receptor and Steel Factor In the mouse, the steel factor (Slf) and its receptor, c-kit, represent a growth factor–receptor system that is involved in development of three major cell lineages: hemopoietic cells, melanocytes, and germ cells. Mice with mutations at either the Slf or c-kit locus have white coat color, sterility, and anemia, attributable to failure of stem-cell populations (melanoblasts, primordial germ cells, and pluripotent hemopoietic stem cells) to migrate and/or proliferate effectively during development. The mouse c-kit gene is a proto-oncogene whose product is a 145- to 160-kDa transmembrane tyrosine kinase receptor of the same family as a number of growth factor receptors (such as insulin-like growth factor I and platelet-derived growth factor). The human c-kit gene is located at 4q12, spans more than 70 kb of DNA, and includes 21 exons, which are alternatively spliced. The longest transcript is approximately 5.2 kb. In mouse, Kit protein is expressed in primordial germ cells, spermatogonia, Leydig cells, and growing oocytes. Expression can be detected in primordial germ cells as early as mouse embryonic day 7.5, ceasing as germ cell proliferation is completed, after arrival at the gonadal ridge on day 13.5–15.5, and returning during the postnatal maturation of oocytes and spermatogonia. The Kit receptor appears to be active in a homodimer form. Thus, mutations of a single allele at the c-kit locus may cause defective receptor activity by combination of mutant receptor with wild-type, inhibiting the action of the normal receptor (dominantnegative effect). The ligand for the c-kit receptor, known as steel factor (Slf), stem cell factor (SCF), or Kit ligand (KL), is encoded by the Steel (Slf) locus. The gene maps to human chromosome 12q22. By alternative splicing the Slf gene codes for two different peptides, a membrane-bound form and a soluble form. In the testis, steel factor is produced primarily by the Sertoli cell. Deletion of the entire Slf gene is lethal in mice, whereas a mutation that allows production of only the soluble form of steel factor, while resulting in abnormal germ cell migration and spermatogenesis, is nevertheless compatible with life. These facts are interpreted as indicating that the membrane-bound form of Slf is crucial for differentiation of tissues. Slf and c-kit may in fact have their primary role as survival factors, because they appear to suppress apoptosis. Although c-kit and Slf are crucial for the development and migration of primordial germ cells in rodents, understanding of their role in human testicular development is still evolving. In normal human testicular development Kit protein is detectable up to 15 weeks of gestation. However, in fetal testes of individuals with intersex conditions Kit was detectable at a later stage, suggesting disturbance of the timing of germ cell development. Notably, individuals with intersex conditions are at increased risk for development of germ cell tumors, and Kit is expressed at high level by seminomas, one type of germ cell tumor. In adult testicular tissue Kit protein is detectable in undifferentiated spermatogonia, whereas Slf/SCF is found on the membrane of the seminiferous tubules and on the surface of Sertoli cells, suggesting that the c-kit and Slf system may act in a paracrine fashion. SCF is also expressed

in human ovary and cultured granulosa/luteal cells, in which it is downregulated by gonadotropins. Interestingly, both proteins are also expressed in the human central nervous system, particularly in the cerebellum, in addition to a wide variety of other fully differentiated tissues, suggesting that this system may have roles that extend well beyond fetal life. Despite the known role of c-kit in migration and proliferation of germ cells (in addition to hemopoietic stem cells and melanoblasts) in the mouse, human c-kit mutations have not been associated with reduced fertility in humans, but have been found only in association with piebaldism and in a human mast cell leukemia cell line. In summary, in the mouse the c-kit–steel receptor–ligand system plays a crucial role in migration, differentiation, and survival of primordial cells destined to become germ cells, and there is evidence for a role for these factors in human testicular development and carcinogenesis. It has been hypothesized that c-kit and SCF may have a regulatory function in normal testicular tissue by providing environmental factors necessary for spermatogenesis. The relevance of this system to human fertility is yet to be addressed. DISORDERS OF SEX DIFFERENTIATION: 46,XY AND 46,XX PSEUDOHERMAPHRODITISM These disorders represent abnormalities of genital morphogenesis, in the presence of gonads that are appropriate for karyotype. 46,XY Pseudohermaphroditism Male (46,XY) pseudohermaphroditism describes the condition in which an individual with male karyotype and normally formed testes has abnormal masculinization of the internal and/or external genitalia. The causes are numerous, but in essence there are two major classes—defects of production or response to testosterone, and defects of production or response to AMH. The causes of male pseudohermaphroditism are summarized in Table 57-2. Clinical examples are provided in Fig. 57-5. Disorders of Testosterone Production or Action: General Clinical Features The genital phenotype in these conditions generally reflects the severity of the defect at the functional level. 46,XY individuals with profound deficiency of androgen production or action have an entirely female phenotype. Those in whom some degree of androgen production and action is retained have a wide spectrum of genital phenotypes, including minor degrees of posterior labial fusion or clitoral enlargement; ambiguous genitalia characterized by incomplete labioscrotal fusion and a clitorislike phallus; and more significant masculinization with reasonable penile size accompanied by urethral hypospadias, penile chordee, and cryptorchidism. The testes may be located anywhere from the abdomen to the inguinal canals or the labioscrotal region. Internal genital structures range from fully feminized, separate vaginal and urethral structures in cases in which there is minimal androgen action, to more masculinized structures, such as a urogenital sinus with ectopic vaginal orifice, in milder cases in which a moderate degree of androgen production and action has occurred. The development of Wolffian duct structures (epididymis and vas deferens) is also variable, depending on the degree of local testosterone secretion and action in early gestation. The vaginal pouch is blind and Müllerian duct structures are absent or diminutive because AMH is secreted normally by the Sertoli cells. Disorders of Testosterone Production or Action: Specific Disorders—Leydig Cell Hypoplasia Clinical Features and Diagnosis. Leydig cell hypoplasia (LCH) is a rare cause of male pseudohermaphroditism, in which Leydig cell differentiation and

CHAPTER 57 / SEX DETERMINATION AND DIFFERENTIATION

testosterone production are impaired because of absence or defective function of LH/CG receptors on Leydig cell progenitors. This condition is generally inherited in a male-limited, autosomal recessive pattern. Inability to respond to CG or LH because of absence of the appropriate receptor results in inadequate testosterone production in utero and thereafter. The genital phenotype is generally female; however posterior labial fusion, a small vagina, or a urogenital sinus may be present, and occasional patients with small penis are reported, suggesting significant T production in the first trimester. Minimal development of either male or female secondary sexual characteristics occurs at puberty, a feature that distinguishes this condition from the androgen insensitivity syndromes, in which high estrogen production at puberty results in excellent breast development in most cases. In postpubertal individuals the testes are slightly small. Although Sertoli cells are normal there is hyalinization of the seminiferous tubules and interrupted spermatogenesis and absent or reduced Leydig cells. The latter finding indicates that LH is required for development and survival of Leydig cells. Testicular membrane preparations do not bind labeled hCG, a fact that, until recently, has served as the primary evidence of a defect of the LH/ CG receptor. The diagnosis of Leydig cell hypoplasia would be suggested by low testosterone levels, despite high serum LH, in a 46,XY infant with female or ambiguous external genitalia; this hormonal profile should allow diagnostic differentiation from the androgen insensitivity syndromes, because in these latter conditions, at least theoretically, testosterone concentrations are high. Unlike infants with a testosterone biosynthetic defect, who may also have low serum testosterone and high LH, there is no elevation of testosterone precursors. In addition, there is no increase in testosterone or its precursors in response to hCG administration. Failure of pubertal development in a phenotypic female with high serum LH and no measurable sex steroids would also suggest this diagnosis. Genetic and Molecular Pathophysiology. LCH is an apparently rare autosomal recessive disorder with an estimated prevalence of approximately 1:1,000,000, although this low prevalence may in part reflect lack of ascertainment. The 60-kb gene encoding the LH/CG receptor (CH/CG-R) is located at chromosomal position 2p21 and comprises 11 exons and 10 introns. The encoded receptor is a member of the seven transmembrane domain class of Gprotein-coupled receptors in which ligand binding induces increased intracellular production of cyclic AMP, the principal mediator of hormone action in this system. Exons 1–10 encode the major portion of the amino-terminal extracellular domain, containing a leucine-rich repeat, whereas exon 11 codes for a small part of the extracellular domain, as well as all 7 transmembrane loops and the carboxy-terminal intracellular region. The LH/CG receptor has significant homology with the FSH and thyroid-stimulating hormone (TSH) receptors. Testicular Leydig cells constitutively express LH/CG receptor in the absence of hormone. Interestingly, the receptor appears to be regulated by its ligand (LH or CG) via two processes: uncoupling (a fairly immediate process that results in reduced cAMP production in response to ligand, without reducing receptor number) and downregulation (a slower, biphasic process resulting in reduced receptor number because of internalization and degradation of already formed receptors, followed by reduced receptor mRNA transcription, via a cAMP-mediated process). These regulatory events are believed to contribute to the phenomenon of

543

desensitization that occurs in the presence of continuous hormonal stimulation. In contrast to the constitutive receptor expression seen in Leydig cells, ovarian cells require hormonal stimulation for receptor expression: in the preovulatory follicle estrogen and FSH synergystically induce receptor mRNA expression in rat granulosa cells, also via cAMP-mediated events. Subsequently, the high levels of LH that occur at ovulation downregulate receptor expression. There appears to be some constitutive expression of LH/CG receptor in theca cells, in addition to hormonal regulation similar to, though less tight than, that seen in granulosa cells. To date, reports of inactivating mutations in the LH/CG-R gene are scarce. In a pair of phenotypic female 46,XY siblings with Leydig cell hypoplasia a nonsense mutation was found in exon 11 of one allele of the LH/CG-R gene, introducing a premature termination codon that predicts truncation of the protein within the 5th transmembrane domain. Interestingly, the mutation was also present heterozygously in the phenotypically normal father and was absent in the mother. The affected siblings were therefore presumed to be compound heterozygotes, harboring a second undetected mutation on the other allele of the gene. In vitro studies of this mutant receptor revealed reduced cell surface expression of mutant receptor protein compared with the normal protein. In addition, hCG stimulation of cAMP production was impaired. In another family, three phenotypically female 46,XY siblings who had lack of breast development and primary amenorrhea, high serum LH, and absent Leydig cells on testicular histology, were found to be homozygous for a nonsense mutation that predicts protein termination in the third cytosolic loop of the receptor. Interestingly, their 46,XX sister, who had undergone menarche at 20 years of age, thereafter developing secondary amenorrhea, and who also had increased serum LH concentration, was homozygous for the same mutation. Another pair of affected siblings, born to consanguineous parents, were found to have a homozygous missense mutation resulting in an alanine to proline change at amino acid 593 in the 6th transmembrane domain of the receptor. In vitro studies revealed that although the receptor had normal affinity of binding for CG, the ligand-bound state did not induce cAMP production. The impaired cAMP production by Leydig cells harboring this mutant receptor explains the lack of hormonal action, since transduction of the hormonal signal requires cAMP-mediated events. The final patient whose mutation has been identified to date is a phenotypically male child with subnormal penile size, whose testes were normally descended. His serum LH was at the upper limit of normal for age during childhood. Analysis of genomic DNA revealed a missense mutation that converted the native serine to tyrosine at position 616 in the seventh transmembrane domain of the receptor. Surprisingly, in view of the normal masculinization of the affected child, in vitro studies of the mutant receptor revealed absent LH binding and cAMP production. This apparent paradox suggest either that testosterone secretion during the first trimester of gestation, when external genital morphogenesis occurs, is not LH or CG dependent, or that the in vitro finding of a nonfunctional LH/CG receptor does not reflect its activity in vivo. In contrast, penile growth during the third trimester is clearly dependent on function of the LH/CG receptor. Defects of Testosterone Biosynthesis A defect in any of the enzymatic steps of testosterone biosynthesis can cause ambiguous genitalia because of insufficient testosterone production during male sexual differentiation. Depending on the location of the enzymatic defect in the steroidogenic pathway, these disor-

544

SECTION VII / ENDOCRINOLOGY

ders may also affect glucocorticoid and mineralocorticoid biosynthesis by the adrenal gland (Fig. 57-7). Five defects of testosterone biosynthesis have been described; however, they are relatively uncommon causes of male pseudohermaphroditism. Because all of the genes encoding these enzymes are located on autosomes, these disorders are inherited as autosomal recessive traits. Table 57-4 summarizes molecular information regarding enzymes of steroidogenesis. The molecular and biochemical heterogeneity within each defect and the consequent variability of the testosterone biosynthetic defect makes it impossible to distinguish these enzyme disorders on clinical grounds. Internal and external genital development is variable, covering the full gamut of features described above. Wolffian derivatives also may be normal or hypoplastic, depending on the severity of the testosterone deficit. Congenital Lipoid Adrenal Hyperplasia: Steroidogenic Acute Regulatory Protein—StAR; P450 Side-Chain Cleavage Enzyme—P450scc The first and rate-limiting step in gonadal and adrenal steroidogenesis (Fig. 57-7) is the transfer of cholesterol from the outer mitochondrial membrane to the inner membrane, mediated by a recently discovered protein, steroidogenic acute regulatory protein (StAR). Thereafter, conversion of cholesterol to pregnenolone involves three distinct biochemical reactions: 20α-hydroxylation, 22-hydroxylation, and side-chain cleavage (the C20—C22 bond). These reactions are mediated by a mitochondrial mixed function oxidase, cytochrome P450scc (previously known as 20,22-desmolase), located on the inner mitochondrial membrane. A defect in the conversion of cholesterol to pregnenolone causes congenital lipoid adrenal hyperplasia (CLAH), the rarest and most severe form of congenital adrenal hyperplasia. In this condition, neither the gonads nor the adrenals have steroidogenic capability. Clinical Features and Diagnosis. Infants with this disorder usually present in the first few weeks of life with hyponatremia, hyperkalemia, and metabolic acidosis reflecting severe salt-wasting adrenal insufficiency, often resulting in neonatal death. Affected 46,XY infants usually have completely female external genitalia although minimal masculinization has been reported. Other features are typical for a defect of androgen production, as described above. 46,XX infants have normal female internal and external genital development. Computed tomography or magnetic resonance imaging scan of the abdomen reveals very large, lipid-laden adrenal glands as a result of the accumulation of cholesterol and cholesterol esters. Interestingly, the testes contain normal Leydig cells without lipid accumulation. Serum concentrations of adrenocorticotropic hormone (ACTH) are markedly increased (causing increased skin pigmentation in some cases), as is plasma renin activity, whereas serum and urinary concentrations of all adrenal and gonadal steroids and their metabolites, even after ACTH or hCG stimulation, are profoundly low. Adrenal and gonadal tissues from these patients are unable to convert cholesterol to pregnenolone in vitro. The disease has been said to be extremely rare in European populations; however, this may result in part from failure of ascertainment and early neonatal death precluding the diagnosis in many cases.

In contrast, it is thought to be the second most common form of adrenal hyperplasia in Japan and more than 50% of cases reported to date are of Japanese or Korean heritage. Unlike other forms of congenital adrenal hyperplasia (CAH), heterozyotes display normal steroid responses to ACTH stimulation, since the defect is caused by absence or dysfunction of the StAR protein and not to a steroidogenic enzyme deficiency. Genetic and Molecular Pathophysiology. The cholesterol transport protein StAR is encoded by an 8-kb gene located at 8p11.2 that comprises seven exons and six introns and contains an 855-bp open reading frame. A related pseudogene maps to chromosome 13. StAR mRNA expression is stimulated by trophic hormones, such as LH and ACTH, via cyclic AMP-mediated events. In addition, the promoter of the StAR gene contains binding sites for SF-1 and the estrogen receptor, implicating these important transcription factors in regulation of StAR expression. StAR mRNA expression in the human is limited to the ovary, testis, adrenal cortex, and kidney. Interestingly, although the placenta is an abundantly steroidogenic tissue, StAR is not expressed there, nor in brain, reflecting the fact that these tissues do not exhibit acute regulation of steroidogenesis. StAR is not expressed in any nonsteroidogenic tissues. The StAR protein comprises 285 amino acids and contains a hydrophobic amino-terminal region typical of a mitochondrial targeting sequence. It is produced as a 37-kDa cytosolic precursor that is imported to the mitochondria, where it is processed to four mature 30-kDa forms. The acute response of steroidogenic cells to stimulation is the transport of cholesterol to the inner mitochondrial membrane and the cholesterol side-chain cleavage enzyme. The mechanism of action of StAR in this process is thought to involve formation of contact sites between the inner and outer mitochondrial membranes. In three apparently unrelated Asian patients, homozygous mutations have been detected in the gene encoding StAR: perhaps because of a founder effect, two individuals (one Japanese, one Korean) had the same mutation converting the codon for glutamine at position 258 to a termination codon. The third subject had a similar nonsense mutation, converting arginine 193 to a termination codon. One further patient with CLAH was found to have a mutation that disturbed mRNA splicing, resulting in deletion of the 185 bp of sequence corresponding to exon 5 of the gene. The deleterious effect of these mutations is presumably caused by absence of a functional protein and subsequent cellular damage because of the accumulation of cholesterol esters. P450 Side-Chain Cleavage Enzyme—P450scc The first enzyme of steroidogenesis, P450scc, is encoded by the 20-kb CYP11A gene localized to the q23–q24 region of chromosome 15. The gene is expressed in steroidogenic tissues, such as adrenal cortex, ovarian granulosa cells, testicular Leydig cells, and placenta. Expression of the P450scc gene is enhanced by ACTH, gonadotropins, cAMP, and SF-1, which may be acting via other factors. Expression patterns are tissue-specific and involve the use of alternate promoter sequences. The P450scc protein complex is located on the inner mitochondrial membrane. Electrons are transferred from NADPH to adrenodoxin reductase, a membrane-bound flavoprotein, then to adrenodoxin, a soluble iron-sulfur protein,

Figure 57-7 (continued on next page) Pathway of steroid biosynthesis and defects in steroidogenesis. The first step in steroidogenesis is the transport of cholesterol across the mitochondrial membrane, mediated by the transport protein StAR. The remaining steps in adrenal and gonadal steroidogenesis are accomplished by a number of enzymes. Most are of the P450 cytochrome family. However, 3β-hydroxysteroid dehydrogenase and 17β-hydroxysteroid dehydrogenase are of the short-chain alcohol dehydrogenase family.

CHAPTER 57 / SEX DETERMINATION AND DIFFERENTIATION

545

Fig. 57-7 (continued). Molecular defects that cause male pseudohermaphroditism are designated by triangles. Defects that cause female pseudohermaphroditism are designated with stars. 3β-Hydroxysteroid dehydrogenase deficiency can cause either male or female pseudohermaphroditism.

546

Table 57-4 Enzymes Involved in Steroidogenesis, Their Encoding Genes, and Chromosomal Loci Gene name

Locus

Enzyme name

Enzyme type/location

Tissue

Action

15q23–24

P450scc

Cytochrome P450 mixed function oxidase Mitochondria

Adrenals, gonads

20α-Hydroxylation 22α-Hydroxylation 20–22 Side-chain cleavage

HSD3B2

1p13–11

3βHSD

Adrenals, gonads

CYP21

6p21.3

P450C21

Conversion of 3β-hydroxy-∆5-steroids to 3-keto-∆4-steroids 21-Hydroxylation

CYP11B1 CYP11B2

8q22 8q22

P450C11 P450cmo

Dehydrogenase Membrane bound Ctyochrome P450 Mitochondria Cytochrome P450 Ctyochrome P450

CYP17

10q24.3

P45017α

HSD17 (EDH17B1) HSD17B2 (EDH17B2) HSD17B3 (EDH17B2) CYP19

17q12–21 16q24.1–24.2 9q22 15q21.1

17βHSD 17βHSD2 17βHSD3 P450arom

SRD5A1 SDR5A2

5p15 2p23

5α-reductase 1 5α-reductase 2

Adrenals, gonads Fascic, glomerulosa Glomerulosa only

NADPH

NADPH Reductase NADPH

Adrenals, gonads

Cytochrome P450 Endoplasmic reticulum

Gonads, placenta

Conversion C19–C18 steroids

Liver, skin Liver, external genitalia

Reduction C19 and C21 steroids

Microsomes

NADPH

11-Hydroxylation 11 Hydroxylation, 18-Hydroxylation 18-Oxidation 17α-Hydroxylation 17-20 Side-chain cleavage (lyase) Interconversion androstenedione/ testosterone; esterone/estradiol

Cytochrome P450 Endoplasmic reticulum Dehydrogenase isozoymes Membrane bound

Gonads, placenta, other tissues

NADPH Adrenodoxin Adrenodoxin Reductase NAD+

NADPH NADPH

SECTION VII / ENDOCRINOLOGY

CYP11A

Cofactors

CHAPTER 57 / SEX DETERMINATION AND DIFFERENTIATION

and finally to the enzyme P450scc itself. No molecular defects in CYP11A have been found to date, despite extensive molecular studies; limited studies of adrenodoxin and adrenodoxin reductase mRNAs have also been normal in individuals with CLAH. 3 β -Hydroxysteroid Dehydrogenase Deficiency The 3β-hydroxysteroid dehydrogenase/∆5–∆4 isomerases (3β-HSD1 and 3β-HSD2) are two highly homologous noncytochrome, membrane-bound short chain alcohol dehydrogenases that require NAD+ as a cofactor. Both have two separate enzymatic activities: dehydrogenase activity and isomerase activity (conversion of ∆5-steroids to ∆4-steroids), the net result of which is the conversion of 3β-hydroxy-∆5-steroids (pregnenolone, 17-hydroxypregnenolone, and dehydroepiandrosterone) to the 3-keto-∆4-steroids (progesterone, 17-hydroxyprogesterone, and ∆4-androstenedione, respectively). This is an essential step in the biosynthesis of all classes of steroid hormones; deficiency of 3β-HSD activity in steroidogenic tissues impairs both adrenal and gonadal (testicular and ovarian) steroidogenesis and is reported to be the second most common cause of congenital adrenal hyperplasia. Clinical Features and Diagnosis. Affected 46,XY individuals have variable impairment of internal and external genital masculinization resulting from inadequate fetal testosterone biosynthesis because of defective function of the testicular isozyme, 3β-HSD2. In some cases there is also salt wasting adrenal insufficiency as a result of reduced synthesis of aldosterone and cortisol. Less severe “late-onset” forms of 3β-HSD present with more complete masculinization and without salt-wasting, reflecting genetic heterogeneity within this disorder. Because of lack of estrogen production, there is no breast enlargement at puberty, either in 46,XX or 46,XY individuals. The classic hormonal profile is a markedly increased concentration of the ∆5-steroids (pregnenolone, 17-hydroxypregnenolone, and dehydroepiadrosterone) and their metabolites in the serum and urine. However, somewhat confusingly, ∆4-steroids may also be increased, because of peripheral conversion of ∆5-steroids to ∆4-steroids resulting from 3β-HSD1 activity in peripheral tissues. Nevertheless, despite peripheral conversion, the ratio of ∆5- to ∆4-steroids and their metabolites is generally elevated. The peripheral ∆5- to ∆4- steroid conversion explains the paradoxical finding of ambiguous genitalia in conjunction with apparently normal ∆4-steroid levels in some of these patients. Although ∆4- steroid concentrations may be normal peripherally, it is probable that they were inadequate at the tissue level in the developing genitalia, and that specific activity of the testicular isoform, 3β-HSD2, is required to generate adequately high androgen levels during embryogenesis. Genetic and Molecular Pathophysiology. There are two highly homologous human genes encoding isoenzymes responsible for 3β-HSD activity—HSD3B1 and HSD3B2, encoding 3β-HSD1 and 3β-HSD2, respectively. The genes are approximately 7–8 kb in size and contain 4 exons and 3 introns. There is evidence for tight linkage between these genes, which are both located within band p13.1 on chromosome 1. Three pseudogenes containing stop codons or deletions have also been identified. The HSD3B1 gene is the predominant form expressed in nonsteroidogenic tissues, such as liver and kidney, and encodes a 371-amino acid (42-kDa) protein. No mutations have yet been detected in the type I gene in individuals with 3β-HSD deficiency. The type II isoform of 3βHSD has the same enzymatic activities but is expressed almost exclusively in the steroidogenic cells of the adrenals and gonads. In bovine adrenocortical cells the enzyme colocalizes with P450scc

547

at the inner mitochondrial membrane, suggesting a possible functional association between the two enzymes in regulation of steroidogenesis. Studies in cultured human adrenal cells reveal regulation of 3β-HSD mRNA and protein levels by ACTH and angiotensin II. In Leydig cells there is some evidence from rodent studies that 3β-HSD activity is regulated by LH/CG. More than 20 different mutations have been reported in the HSD3B2 genes of individuals with 3β-HSD deficiency, including 14 different single-base mutations in one study. Other mutations include a single-base insertion resulting in a frameshift, an intronic mutation causing aberrant splicing, and single-base mutations causing either premature termination (nonsense mutations) or amino acid substitutions (missense mutations). Gene conversion events (transfer of deleterious mutations from associated pseudogenes) have been suggested to underlie some of the mutations in HSD3B2. In vitro studies using site-directed mutagenesis to recreate a number of mutant 3β-HSDs, reveal significantly reduced enzyme affinity and specific activity for the substrates pregnenolone and dehydroepiandrosterone, and in some cases for the enzyme cofactor NAD+. The finding of a specific abnormality of NAD+ binding associated with substitution of aspartic acid for glycine at position 15 raises the speculation that this region may contribute to the NAD-binding domain of the enzyme. Although absence of salt losing in some patients with severe undermasculinization suggests that impairment of enzyme activity may differ between gonads and adrenals, molecular data indicate that the enzyme is similarly dysfunctional for either of its two main substrates, pregnenolone (adrenal) and dehydroepiandrosterone (gonad). These data suggest that the absence of salt wasting in some individuals results from weak residual enzyme activity sufficient to prevent salt loss, but insufficient to produce the testosterone levels required for male sex differentiation. The clinical heterogeneity of affected individuals (i.e., salt losing vs non–salt losing, varying degrees of masculinization) is at least in part explained by the genetic heterogeneity. Not surprisingly, mutations that completely destroy enzyme activity are associated with a more severe syndrome than are those that allow retention of some enzyme function. In general patients with non–salt wasting disease have up to 10% retained enzyme activity; however, there is no clear correlation between the molecular defect, enzyme activity, and clinical phenotype. The gene is located on an autosome and affected individuals are generally homozygous for the mutant allele; consanguinity has been reported in a number of pedigrees. Compound heterozygosity for two different mutations has been reported in a number of classically affected patients, and minor manifestations of simple heterozygosity have been described, including mild overproduction of ∆5-steroids and mild ovarian dysfunction. 17 α -Hydroxylase Deficiency and 17,20-Lyase Deficiency A single microsomal enzyme, cytochrome P450c17, catalyzes two consecutive oxidation reactions, the 17α-hydroxylase and 17,20-lyase (17,20-desmolase) reactions of adrenal and gonadal steroidogenesis. The 17α-hydroxylase reaction converts pregnenolone to 17-hydroxypregnenolone and progesterone to 17-hydroxyprogesterone, respectively (Fig. 57-7). This reaction is the rate-limiting step in androgen biosynthesis. The 17,20-lyase reaction cleaves the C17,20 bond to convert the C21 steroid 17-hydroxypregnenolone to the C19 steroid dehydroepiandrosterone (DHEA) and likely also catalyzes the equivalent conversion of 17OH-progesterone to ∆4-androstenedione, although this

548

SECTION VII / ENDOCRINOLOGY

latter is somewhat controversial, with discrepant findings in various studies. DHEA and androstenedione are the major precursors of testosterone and estradiol. Clinical Features and Diagnosis. 17α-Hydroxylase is required for cortisol, androgen, and estrogen synthesis and the enzyme is expressed in adrenal cortex and gonads. 17α-Hydroxylase deficiency may be characterized by deficiency of either or both of this enzyme’s activities; however, the molecular basis for this differential expression of the defect is at present unclear. Because of fetal androgen deficiency, 46,XY individuals with 17α-hydroxylase deficiency in the adrenals and gonads have female external genitalia, although a few individuals with limited masculinization have been described. In 46,XX individuals the external genitalia are normal; however, there is failure of development of secondary sexual characteristics at puberty (breast development, pubic and axillary hair) because of deficiency of both adrenal androgens and ovarian estrogen (see Chapter 37). Deficiency of 17α-hydroxylase activity also impairs adrenal cortisol production, and the resultant compensatory ACTH hypersecretion stimulates synthesis of progesterone, deoxycorticosterone (DOC), corticosterone, and 18-hydroxycorticosterone. The excess of the latter three compounds causes salt and water retention, hypertension, and hypokalemia because of their mineralocorticoid activity. Renin activity is therefore suppressed and aldosterone production reduced. Despite impaired cortisol production, evidence of glucocorticoid deficiency is unusual, perhaps because corticosterone has weak glucocorticoid activity. In addition to catalyzing the 17α-hydroxylation reaction in the adrenals and gonads, P45017α also catalyzes the 17,20-lyase reaction resulting in conversion of 17-hydroxypregnenolone to DHEA, and probably 17-hydroxyprogesterone to androstenedione (Fig. 57-7). Thus, deficiency of the lyase activity in the gonads prohibits synthesis of DHEA, ∆4-androstenedione, testosterone, and estrogen. Glucocorticoid and mineralocorticoid production are retained because the enzymatic defect is distal to these pathways. hCG stimulation produces an increased ratio of 17-hydroxy-C21-steroids (17-hydroxyprogesterone and 17-hydroxypregnenolone) to C19-steroids (DHEA and ∆4-androstenedione). Genetic and Molecular Pathophysiology. CYP17 is a single 6.6-kb, 8 exon, 7 intron gene located on chromosome 10 at band q24.3. The CYP17 gene is very similar to CYP21 (encoding the 21-hydroxylase enzyme), and the two may have originated from a common ancestral gene. P45017α is expressed in human adrenal, testis, and ovarian theca cells, but not in ovarian granulosa cells or placenta. P45017α mRNA is upregulated by ACTH via cAMP, and cAMP regulatory regions have been identified in the 5'-flanking region of the bovine P45017α gene; expression is also regulated by the inhibin-activin system in the ovary. Gene transcription is inhibited by activators of protein kinase C. The CYP17 gene encodes cytochrome P45017α a 508-amino acid, 57-kDa protein that is part of the P45017α complex anchored to the smooth endoplasmic reticulum of steroidogenic cells. The complex includes a 78-kDa flavoprotein reductase termed P450 reductase, containing binding domains for nicotinamide adenine dinucleotide phosphate (NADPH), flavin mononucleotide (FMN), and flavin adenine dinucleotide (FAD); and the cytochrome P45017α enzyme itself, which contains heme-binding and steroid-binding regions, accepts electrons from NADPH via the P450 reductase, and catalyzes the overall oxidation reaction. The ratio of 17α-hydroxylase to 17,20-lyase activity differs between adrenal and testis and is

developmentally regulated at adrenarche, an increase in the ratio of lyase to hydroxylase activity occurring at this time. Site-directed mutagenesis has elucidated specific amino acid residues critical for either the 17α-hydroxylase or the 17,20-lyase activity; however, the basis for the time- and tissue-specific differential regulation of the enzyme’s two main activities is not yet fully understood. There is evidence that serine phosphorylation of the enzyme increases lyase activity, whereas dephosphorylation eliminates this activity. Other regulatory features may include variation in the ratio of the NADPH P450 reductase to the P45017α enzyme itself, resulting in alterations of electron transfer to the enzyme. The presence of specific endoplasmic reticulum membrane lipids in certain cell types has also been suggested as a factor involved in the differential activity of this enzyme’s functions. A variety of molecular defects has been described in the P45017α genes of at least 15 individuals with 17α-hydroxylase deficiency. These include large deletions or insertions, small deletions and duplications, and single-base mutations causing frameshifts, premature termination, or amino acid substitutions. In vitro studies of a recreated mutant P45017α, in which histidine at position 373 was replaced by leucine, revealed that the mutant enzyme failed to bind the heme moiety critical for catalytic activity. Other mutations, however, are located distant from the critical heme-binding region or the active site, suggesting that the enzyme is sensitive to structural change. Interestingly, one specific mutation, a 4-base duplication that alters the reading frame of the CYP17 gene, was found in six Dutch families and two families of Canadian Mennonites, suggesting that a founder mutation had presumably occurred in the Dutch antecedents of the Mennonites. Compound heterozygosity has been reported in this form of congenital adrenal hyperplasia, as well as in the more common forms. Individuals heterozygous for a single 17α-hydroxylase mutation are clinically normal; however, ACTH stimulation produces increased serum concentrations of DOC, corticosterone, and 18-hydroxycorticosterone. A number of cases of 17,20-lyase deficiency have been reported in which 17α-hydroxylase activity was said to have been normal. Theoretically, different mutations in the P45017α gene could result in either 17α-hydroxylase deficiency, 17,20 lyase deficiency, or both. However, in vitro protein expression of mutant genes of individuals with the clinical diagnosis of isolated 17α-hydroxylase deficiency has revealed the presence of combined enzyme deficiencies in all cases studied thus far and there is no clear molecular explanation for the apparently retained 17α-hydroxylase activity in vivo. In vitro analysis of mutant P45017α enzymes predicted from the mutations detected in one compound heterozygote 46,XY patient with ambiguous genitalia reveals that retention of 20% of normal 17,20-lyase activity is adequate for some degree of masculinization. 17β-Hydroxysteroid Dehydrogenase 3 Deficiency The 17β-hydroxysteroid dehydrogenases (17β-HSDs, also known as 17β-ketosteroid reductases and as estradiol 17β-dehydrogenases) are NAD+/NADPH-dependent, membrane-bound enzymes not of the cytochrome P450 type, which catalyze the only reversible steps in the steroid biosynthetic pathway: interconversion of ∆4-androstenedione testosterone, DHEA ∆5-androstenediol and estrone estradiol) (Fig. 57-7) by oxidation or reduction of C-18 and C-19 steroids. There are at least three or four tissuespecific isozymes with differing specificity for either the oxidative or the reductive reaction. The isozymes are encoded by at least three distinct genes. 17β-HSD3 is the testicular isozyme, respon-

CHAPTER 57 / SEX DETERMINATION AND DIFFERENTIATION

sible for the final step in testosterone synthesis, reduction of androstenedione to testosterone. Deficiency of this enzyme is the most common defect of androgen production. Clinical Features and Diagnosis. Interestingly, although 46,XY individuals with deficiency of testicular 17β-HSD3 activity have female or mildly masculinized external genitalia, their internal genitalia are well-masculinized, with normally formed epididymis, vas deferens, and seminal vesicles. Reduced activity of testicular 17β-HSD3 is inferred by the finding of elevated ratios of androstenedione to testosterone and estrone to estradiol in spermatic venous samples. In addition, testicular tissues of affected individuals demonstrate impaired conversion of labeled ∆4-androstenedione to testosterone in vitro. This defect is isolated to the gonads—the site of expression of 17β-HSD3. Other tissues have normal 17β-HSD activity because of normal function of the type 1 and 2 isozymes. The relatively normal masculinization of internal structures may result from activity of 17β-HSD isozymes other than 17β-HSD3 in the proximity of the Wolffian ducts producing adequate local testosterone by conversion of testis-derived androstendione. In contrast there is insufficient testosterone at the level of the external genital primordia to act as substrate for 5α-reductase conversion to dihydrotestosterone. A striking feature of this disorder is the marked virilization that occurs at puberty. Affected 46,XY individuals develop a male body habitus with abundant body and facial hair, enlargement of the penis and testes to normal adult size, and pigmentation and rugation of the labioscrotal folds. Gynecomastia is variably present. Many affected individuals spontaneously adopt the male gender role, with apparently adequate sexual function, but with infertility. These clinical observations correlate with normalization of peripheral and spermatic vein testosterone levels. Part of this effect appears likely to be related to increased LH secretion and Leydig cell hyperplasia, because ∆4-androstenedione levels remain markedly elevated, indicating persistent enzyme dysfunction. Peripheral testosterone production as a result of retained extragonadal 17β-HSD activity may also contribute. The phenotype and clinical course of individuals with 17β-HSD deficiency is similar to that of individuals with 5α-reductase deficiency (described below). The diagnosis of 17β-HSD3 deficiency is suggested by a 10- to 15-fold elevation of the ratios of ∆4-androstenedione to testosterone and of estrone to estradiol in the perinatal period, after hCG stimulation in childhood, or after puberty. Because of a secondary increase in 3β-HSD activity, DHEA may be normal or only mildly elevated. Glucocorticoid and mineralocorticoid synthesis remain normal because the enzymatic defect is distal to these pathways. Genetic and Molecular Pathophysiology. There are at least four HSD17B enzymes encoded by separate genes and expressed in a tissue-specific fashion, some being expressed predominantly in estrogenic or androgenic tissues and others more widely. The nomenclature of this group of genes and enzymes is quite daunting: the type 1 enzyme, 17β-HSD I, is also known as estradiol-17β dehydrogenase II. This isozyme is encoded by a gene located at 17q11–12, referred to as HSD17B1 or EDH17B2. This locus in fact contains two genes in tandem: h17b-HSDI and II. The active gene is h17b-HSDII; h17b-HSDI is a pseudogene, also referred to as EDH17BP1, which is 89% homologous to the active gene. 17b-HSD I appears to be the isozyme responsible for ovarian 17β-HSD activity. The type 2 isoform, 17β-HSDII, is encoded by HSD17B2, located at 16q24. This isozyme, which is a 387-amino

549

acid, approximately 43-kDa progestin-regulated protein, is responsible for 17β-HSD activity in placenta and endometrium. Located predominantly in the endometrium, placenta, liver, and small intestine, this isozyme has a preference for NAD+ as its cofactor in oxidation of estradiol and testosterone to estrone and androstenedione, respectively. The testicular isoform, 17β-HSD3 (III), which is the isozyme implicated in male pseudohermaphroditism as a result of 17β-HSD deficiency, is encoded by HSD17B3 (EDH17B3), which shares 23% sequence homology with the HSD17B1 and 2 genes. The gene spans 60 kb, contains 11 exons, and is located at 9q22. This isozyme preferentially uses NADPH as its cofactor in the reduction of androstendione to testosterone. In addition to being found in ovary, testis, and placenta, significant levels of various 17β-HSD mRNAs are found in peripheral sites, such as uterus, breast, prostate, and adipose tissue. In these peripheral locations 17β-HSD may play a major role in regulating the levels of active androgens and estrogens, by using either their oxidative or reductive functions. The 17β-HSDs are not expressed in the adrenals. Mutations have been identified in the HSD17B3 genes of approximately 20 individuals with 17β-HSD deficiency. These include single-base mutations causing amino acid substitutions, frameshift mutations, and splice junction mutations causing abnormal mRNA splicing and thereby disturbed protein structure and enzyme function. One 46,XY phenotypic female was found to have compound heterozygosity for two distinct amino acid substitutions and another had compound heterozygosity for a splice acceptor mutation and a missense mutation. When expressed in cultured mammalian cells in vitro the mutant enzymes displayed impaired enzyme activity. The unusual finding of normal Wolffian structures in these 46,XY phenotypic females is potentially explained by activity of 17β-HSD isozymes other than 17β-HSD3 converting testis-derived androstenedione to testosterone in the local area of the internal genital structures. Mutations have not been reported in the HSD17B1 or HSD17B2 genes to date. 17β-HSD deficiency is inherited in an autosomal recessive fashion and a number of extensive inbred Arab kindreds with numerous affected individuals have been reported. Disorders of Androgen Action 5α-REDUCTASE DEFICIENCY The 5α-reductases are microsomal enzymes that catalyze the 5α-reduction of many C19 and C21 steroids, using NADPH as a cofactor. In the context of disorders of sexual differentiation, the most critical of these reactions is the conversion of testosterone to DHT, mediated by the isozyme 5α-reductase 2 (II). Deficiency of this enzyme results in a form of male pseudohermaphroditism previously referred to as pseudovaginal perineoscrotal hypospadias. This disorder is described in greater detail in Chapter 60. Clinical Features and Diagnosis. Deficiency of 5α-reductase 2 in the tissues of the bipotential fetal external genitalia and urogenital sinus results in inadequate local DHT concentrations. As in other forms of defective androgen production or action, there is subnormal masculinization of these structures, generally presenting as ambiguous external genitalia. The genital phenotype varies widely between and even within affected kindreds, the most consistent findings being underdevelopment of the penis and prostate. It is on the basis of these clinical findings that the requirement for DHT in external genital masculinization has been inferred. Because differentiation of the gonads and Wolffian ducts is not dependent on DHT, the testes, epididymides, vasa deferentia, and seminal vesicles develop normally.

550

SECTION VII / ENDOCRINOLOGY

One of the most intriguing and well-documented features of this disorder is the striking virilization, including increased muscularity and deepening of the voice, that occurs at puberty in many affected individuals. In addition, the testes often enlarge and descend; however, spermatogenesis is absent or severely impaired. Whether this virilization is mediated by improved DHT levels or by supranormal T levels remains unclear. Acne, facial hair, temporal hair recession, and prostatic enlargement do not develop, presumably because these events require higher concentrations of DHT. In contrast to patients with androgen insensitivity syndromes, described below, gynecomastia does not develop because there is no increase in testicular estrogen production. Many affected individuals initially raised as females change their gender role to male around the time of puberty, for reasons that have not been fully elucidated, but likely include cultural as well as biologic factors. Reduced conversion of T to DHT in the target tissues results in the marked increase in the ratio of T to DHT diagnostic of 5α-reductase 2 deficiency. In normal infants (2 weeks to 6 months of age), the hCG-stimulated T:DHT ratio is approximately 5 ± 2 (mean ±SD). Affected infants have markedly increased T:DHT ratios, in the 20–60 range or higher. Serum concentrations of LH may be mildly increased. 5α-Reductase 2 deficiency can be differentiated from 17β-HSD deficiency by the characteristic increase of serum androstenedione in the latter condition. A more direct, but more invasive diagnostic method is the in vitro demonstration of decreased 5α-reductase activity in genital skin fibroblasts; however, because post-hCG serum T:DHT concentrations are diagnostic, this procedure is not required to verify the diagnosis. Genetic and Molecular Pathophysiology. There are two isozymes of 5α-reductase encoded by separate genes. 5α-Reductase 1 is encoded by the SRD5A1 gene located at 5p15, and 5α-reductase 2, the enzyme defective in the clinical syndrome of 5α-reductase deficiency, is encoded by the SRD5A2 gene located at 2p23. The genes are structurally similar, with five coding exons and four introns each. The two enzymes share approximately 50% amino acid identity. Androgen binding is mediated by the ends of the enzyme molecule and binding of NADPH, required as the cofactor for the reduction reaction, by the carboxy-terminal half. The 5α-reductase isozymes have differential expression patterns. The type 1 isozyme is not detectable in the fetus, but is transiently expressed in newborn skin and scalp, and permanently expressed in the liver, and in skin after puberty. The type 2 isozyme is expressed in the liver and in androgen target tissues, including the external genitalia, accessory sex organs, and prostate. The ontogeny of 5α-reductase 2 expression is incompletely characterized; however, the enzyme is not expressed in the Wolffian ducts at the time of their differentiation, a finding that supports the contention that T rather than DHT is the critical androgen involved in this process. Its expression is upregulated by androgens, as demonstrated by the marked increase in 5α-reductase 2 mRNA level in the prostate of castrate animals after testosterone administration. Expression appears to be regulated in the opposite fashion in the liver. Interestingly, 5α-reductase activity has also been detected in neurons. 5α-Reductase 2 deficiency is inherited in an autosomal recessive fashion, and there is a high frequency of consanguinity within affected kindreds. A wide variety of mutations has been identified in affected families from more than 20 different ethnic groups. Mutations range from complete deletion of the SRD5A2 gene in a Papua New Guinea kindred, to single-base mutations distributed

throughout the SRD5A2 gene that result in gene-splicing defects, amino acid substitutions, or premature termination. A number of affected kindreds demonstrate compound heterozygosity for different mutations. In vitro expression of mutant proteins reveals a variety of causes of deficient enzyme activity: deletions, premature termination codons, and splice junction defects prevent expression of a functional enzyme; whereas amino acid substitutions impair binding of testosterone or produce an unstable enzyme. Alterations in both androgen and NADPH binding may be produced by such mutations. Differences in enzyme stability and affinity for testosterone and NADPH between kindreds reflect the genetic heterogeneity of the enzyme defect. There is no abnormality of the isozyme 5α-reductase 1 in individuals with the clinical syndrome of 5α-reductase deficiency, a fact that may in part explain the marked pubertal virilization of affected individuals. Androgen Insensitivity Syndromes (Androgen Receptor Disorders) The androgen receptor (AR) is a nuclear transcription factor that is activated by binding of androgen to a conformation capable of binding to elements in the promoter regions of androgen regulated genes to regulate their transcription. Defective function of this transcription factor results in resistance to the effects of androgens—the well-characterized androgen insensitivity (or resistance) syndromes. These conditions are described in greater detail in Chapter 60. Clinical Features and Diagnosis. The syndromes of androgen insensitivity (AIS), referred to by some authors as testicular feminization, are believed to comprise the most common definable cause of male pseudohermaphroditism. Affected individuals display variable degrees of defective masculinization. In the complete form of AIS not only is the genital phenotype unequivocally female, the labia minora and majora and clitoris in fact may be underdeveloped. Individuals with partial forms of AIS, who have some retained AR activity, have external genital phenotypes ranging from female with pubic hair at puberty, to almost completely masculinized. The complete form of AIS has a prevalence of approximately 1:20,000 male births. The prevalence of the partial forms of AIS is unknown. Diagnosis of AIS in infancy currently relies on clinical features, because there have been no definitive hormonal studies in infants. Preliminary data in a few infants with complete AIS (CAIS) suggest absence of the typical postnatal LH and T surge seen at around 6–12 weeks of age in normal male infants. Diagnosis of CAIS is generally based on the classic clinical findings of 46,XY karyotype, testes, and absent uterus in an otherwise normal phenotypic female. The only condition with which complete AIS could be confused in infancy is Leydig cell hypoplasia (described above). However, the brisk T response to hCG in patients with CAIS should clearly differentiate these two conditions. Diagnosis of partial forms of AIS in infancy is notoriously difficult, because there is no clear hormonal profile; however, a few case reports have noted increased concentrations of LH and T. Lack of suppression of sex hormone-binding globulin after androgen administration may aid in the diagnosis of these conditions; however, such studies have been undertaken in only a limited number of cases to date. The diagnosis of AIS should be suspected in a 46,XY infant with female or ambiguous genitalia if either the baseline or the hCG-stimulated concentrations of T and DHT are normal or exaggerated without disproportionate excess of T precursors. Demonstration of abnormal androgen binding in cultured genital skin fibroblasts or identification of a

CHAPTER 57 / SEX DETERMINATION AND DIFFERENTIATION

mutation in the AR gene of an affected individual confirms the diagnosis, but is impractical outside of research institutions. Furthermore, measured receptor levels and androgen binding affinity often correlate poorly with the degree of masculinization. Depending on the severity of the androgen resistance, varying degrees of virilization and/or feminization occur at puberty. At and beyond puberty, LH, T, and DHT increase to supranormal levels and serum estrogen concentrations are also enhanced, because of testicular estrogen secretion (driven by LH) and peripheral aromatization of T. Genetic and Molecular Pathophysiology. The approximately 90-kb gene encoding the AR contains eight coding exons and seven introns and is localized to the q11–12 region of the human X chromosome. AIS is thus inherited in an X-linked fashion. The AR is expressed in a wide array of genital and nongenital tissues, reflecting its role as a fairly ubiquitous transcription factor. The AR gene encodes a 110-kDa protein that is a member of the steroid receptor superfamily, comprising of three major functional domains: amino-terminal (transcription-regulating), central DNA-binding (zinc finger), and carboxy-terminal steroid-binding. After binding androgen, the AR, as a homodimer, interacts with androgen response element (ARE) DNA sequences in the promoter regions of target genes to regulate their transcription. Molecular analysis of the AR genes of more than 100 hundred unrelated individuals with various forms of AIS has revealed a highly heterogeneous genetic basis for the spectrum of clinical disorders. Defects include rare complete and partial gene deletions, insertions and deletions of a few basepairs, and single-base mutations that disrupt splice junctions, introduce premature termination codons, or cause amino acid substitutions. Complete and subtotal deletions of the AR gene, as well as frameshift and nonsense mutations, result in absence of a functional AR protein. In contrast, missense mutations resulting in amino acid substitutions, of which at least 150 have now been described, allow production of a structurally normal protein. The majority of missense mutations (approximately 80%) are located in the five exons encoding the steroid-binding domain, with almost all of the remainder occurring in the two exons encoding the DNA-binding domain. The function of many mutant receptors has been analyzed in vitro: when the mutation affects the steroid-binding domain androgen binding may be absent, reduced (affinity or capacity or both), or qualitatively abnormal (thermolability of binding, increased ligand dissociation, or altered binding specificity). Mutations located within the zinc finger DNA-binding domain are generally associated with normal (occasionally increased) androgen binding; however, there is altered affinity, capacity, or specificity of receptor binding to appropriate DNA sequences. In vitro studies indicate that the underlying abnormality of defective ARs, whether the mutation affects DNA or androgen binding, is loss of transcriptional competence of the mutant receptor—reduced ability to regulate transcription of androgen-dependent target genes. Disorders of Anti-Müllerian Hormone Production or Action—The Persistent Müllerian Duct Syndrome—Types I and II DISORDERS OF ANTI-MÜLLERIAN HORMONE PRODUCTION Anti-Müllerian hormone ([AMH] or Müllerian inhibiting substance [MIS]) is a glycoprotein product of testicular Sertoli cells that mediates regression of the Müllerian duct structures during normal male embryogenesis. Deficiency of this hormone or abnormality of its receptor results in persistence of the Müllerian ducts, resulting in the mildest form of male pseudohermaphroditism.

551

Clinical Features and Diagnosis. Karyotypic males with deficiency of AMH have normal male external genitalia, normal testicular histology, and normal Wolffian duct differentiation. The two specific defects found in these individuals are persistence of the Müllerian ducts, and abnormalities of testicular descent (the persistent Müllerian duct syndrome—PMDS). The typical case is that of a phenotypically normal male infant with bilateral cryptorchidism (normal testicular histology) and inguinal hernias, who is found, at the time of hernia repair, to have a uterus, cervix, and fallopian tubes in the inguinal canal. Serum concentrations of AMH, which is readily measurable in the serum of normal males until puberty, are reduced in most affected individuals with PMDS, referred to as type I. In some, AMH concentrations are normal, suggesting either bioinactivity of the hormone (not yet reported), or an AMH receptor defect (discussed below), referred to as PMDS type II. A recent study determined that mutations of AMH and AMHR each account for approximately 50% of cases of PMDS. Genetic and Molecular Pathophysiology. The human gene encoding AMH contains five exons and is located on chromosome 19, at position p13.3–13.2. Persistent Müllerian duct syndrome is therefore inherited in an autosomal recessive, male-limited, pattern. After SRY, AMH is the first molecular marker specific for testis differentiation. Its expression in developing mouse testis begins significantly later than that of SRY, at about day 13.5. In fact, SRY may regulate AMH production in males, because the AMH gene contains a regulatory element within its promoter that likely binds SRY. Notably however, the 48-h temporal dissociation between the expression of these two proteins suggests involvement of one or more intermediate factors. AMH expression is also regulated by SF-1, which binds to the AMH gene promoter, likely using a cofactor for optimal gene regulation. After puberty, AMH production is downregulated by androgens (a feature absent in individuals with AIS). AMH is not expressed prenatally in the ovary. However low amounts of AMH are released into the follicular fluid by mature granulosa cells in normal postpubertal females, who do not express SRY. AMH, a member of the TGF-β family of growth factors, is a 560-amino acid, approximately 70-kDa glycoprotein that forms a 140-kDa homodimer. AMH is produced by Sertoli cells not only during the critical period, when it induces regression of the Müllerian ducts (weeks 9–11 of human gestation), but also in late gestation, after birth, and even in adulthood (at a reduced level), suggesting that AMH may have physiologic roles other than its control of Müllerian duct regression. In its role as the inducer of Müllerian duct regression in males during embryogenesis, AMH acts in a paracrine fashion, mediating only the regression of the ipsilateral Müllerian duct. Given the fact that the testes are invariably undescended in PMDS, one postulated role for AMH in the latter part of gestation is in the regulation of testicular descent. However, this speculation is tempered by the finding of variable AMH levels in boys with simple cryptorchidism. A postulated role of AMH in postpubertal females is to inhibit oocyte meiosis, allowing for enhanced oocyte maturation before selection for ovulation. AMH acts via the type II AMH receptor at target sites. Mutations of the AMH gene have been identified in approximately 20 patients with the PMDS type I phenotype who have low or undetectable serum AMH. These include premature termination codons, frameshift mutations, and missense mutations present in the homozygous or compound heterozygous state. Surprisingly, the first exon of the AMH gene appears particularly mutation-

552

SECTION VII / ENDOCRINOLOGY

prone although it codes for an amino-terminal part of the AMH protein, which is not essential to bioactivity. The deleterious effect of AMH gene mutations reflects loss of function of the encoded peptide, presumably because of the markedly abnormal structure created by protein truncation, the predominant form of mutation reported to date. DISORDERS OF ANTI-MÜLLERIAN HORMONE ACTION The clinical features of this disorder are identical to those of the originally described PMDS. Since the availability of bioassay and radioimmunoassay and techniques for measurement of AMH, this group of patients has been identified by the finding of normal or increased serum concentrations of bioactive AMH (PMDS type II). Genetic and Molecular Pathophysiology. The gene encoding human AMH receptor contains 11 exons and is located at 12q13. Surprisingly, the expression of AMHR is not detected until mouse embryonic day 15, substantially later than that of its ligand AMH (day 13.5). If this is a true biologic phenomenon, rather than a methodologic artifact, its basis remains to be explained. It is similarly enigmatic that expression of AMHR is regulated by androgens; however, androgens are not produced until after the onset of AMH secretion, in the hiatus between AMH and AMHR expression, suggesting that initial regulation of AMHR is not by androgens or AR. In mouse AMHR is expressed in mesenchymal cells adjacent to the Müllerian ducts, suggesting that AMH may alter some aspect of the mesenchyme surrounding the Müllerian ducts, which in turn induces duct regression. The receptor protein is a membrane-bound serine–threonine kinase similar to those for TGF-β and activin, containing a single transmembrane domain. Although it binds ligand directly, it requires the presence of a type I receptor for signal transduction. Since its recent cloning, mutations in the human AMHR gene have been elucidated. The first reported case had a homozygous mutation at the splice donor site of intron 2, resulting in a downstream amino acid substitution and insertion of four new residues. In a study of 16 AMHR mutations, 10 patients had a 27-bp deletion in exon 10. Four patients were homozygous for this mutation, which predicts the loss of nine amino acids from the protein, and 6 patients were compound heterozygotes, having missense mutations on their other allele. 46,XX Pseudohermaphroditism Essentially, disorders of female sex differentiation (female pseudohermaphroditism) represent the converse of male pseudohermaphroditism, in general resulting from excessive androgen exposure. Although intuitively likely, exaggerated response to normal androgen levels has not yet been described as a cause of female pseudohermaphroditism. There are far fewer causes of virilization of a female fetus than there are of undermasculinization of a male fetus. Abnormalities of Müllerian development, such as the Mayer-Rokitansky-KüsterHauser syndrome, represent the mildest form of female pseudohermaphroditism. Lack of fetal estrogen effect does not affect gonadal or genital development, clearly evidenced by the fact that transgenic mice deleted for the estrogen receptor are phenotypically normal. Disorders Producing Excessive Fetal Androgenization The most common cause of excess fetal androgen production is congenital adrenal hyperplasia (of the 21-hydroxylase, 11β-hydroxylase, or 3β-hydroxysteroid dehydrogenase deficiency types). Molecular defects in these disorders are summarized in Figs. 57-8 and 57-9. A much rarer cause of female pseudohermaphroditism is aromatase deficiency in which fetal virilization results from high levels of

testosterone and androstenedione because of the inability to convert these precursors to estrogens (Fig. 57-10). Female pseudohermaphroditism as a result of maternal androgen excess will not be discussed here. Disorders Producing Excessive Fetal Androgenization: General Clinical Features Virilization of a karyotypic female can produce variable genital phenotypes, depending primarily on the timing, duration, and degree of androgen excess. The phenotype ranges from limited clitoromegaly and posterior labial fusion in the milder cases, to intermediate degrees of virilization with labioscrotal fusion, a urogenital sinus, and significant clitoral enlargement, to severely virilized genitalia that are indistinguishable from those of a cryptorchid male infant. The gonads are normal ovaries and Müllerian duct derivatives develop normally, reflecting absence of AMH. Even in the most virilized infants, Wolffian development does not occur, perhaps because of differences in timing of adrenal versus gonadal androgen production or the presence of less significantly increased androgen concentrations in the vicinity of the Wolffian ducts. Disorders Producing Excessive Fetal Androgenization: Specific Disorders—Congenital Adrenal Hyperplasia 21-HYDROXYLASE DEFICIENCY Steroid 21-hydroxylase is a microsomal cytochrome P450 enzyme (P450c21), located on the smooth endoplasmic reticulum, that catalyzes the hydroxylation of progesterone to deoxycorticosterone and 17-hydroxyprogesterone to 11-deoxycortisol, using an NADPH-dependent cytochrome reductase as a cofactor. Deficiency of this enzyme is the most common cause of congenital adrenal hyperplasia in both sexes (see Chapter 52). Clinical Features and Diagnosis. Hydroxylation is impaired in the zona fasciculata of the adrenal glands of those with 21-hydroxylase deficiency, so that 17-hydroxyprogesterone (17-OHP) is not converted to 11-deoxycortisol, the immediate precursor of cortisol. Cortisol deficiency leads to a compensatory increase in ACTH secretion that in turn drives the adrenal gland, resulting in overproduction of cortisol precursors, particularly 17-OHP. These steroids are shunted down the remaining functional steroidogenic pathway, producing a surfeit of androgens (mainly testosterone) that results in virilization of 46,XX fetuses but has no untoward effect on the male fetus. In more than half of cases, 21-hydroxylase is also deficient in the zona glomerulosa and there is failure of conversion of progesterone to 11-deoxycorticosterone, resulting in deficiency of aldosterone. Shock or death may result from severe salt wasting and resultant hypovolemia. There are four major clinical forms of 21-hydroxylase deficiency: salt wasting (which accounts for approximately 75% of cases), simple virilizing, nonclassic “late onset” (also called attenuated or acquired), and cryptic (asymptomatic). Only the two most severe forms, which produce genital ambiguity, will be discussed. Female infants with the classic virilizing form of 21-hydroxylase deficiency have variable masculinization of the genitalia as described above. The virilizing form of 21-hydroxylase deficiency is the most likely diagnosis in an infant with genital ambiguity in the presence of 46,XX karyotype. This should be suspected in any partially virilized infant in whom gonads can not be located, either by palpation or by ultrasonography. The presence of a uterus enhances the likelihood of this diagnosis. Increased serum concentrations of 17-hydroxyprogesterone, often as high as 40,000–60,000 ng/dL, are invariably present in infants with enzyme deficiency severe enough to cause virilization; this is accompanied by hyponatre-

CHAPTER 57 / SEX DETERMINATION AND DIFFERENTIATION

553

Figure 57-8 Genomic organization and molecular defects of CYP21B (encoding steroid 21-hydroxylase). (A) Genomic organization. CYP21B, the gene that encodes 21-hydroxylase, is located in complex a region of chromosome 6 that has undergone a duplication event. There are thus a number of pairs of tandemly arranged genes in this region. On the upper DNA strand, reading from the 5' direction, the arrangement is as follows: C4A (pseudogene for the 4th component of complement), CYP21A (the so-called pseudogene of CYP21), C4B (the active gene encoding the 4th component of complement), CYP21B (the gene encoding P450c21). On the lower strand the arrangement is even more complex: two genes (XB and XA) that overlap the 3' ends of CYP21B and CYP21A, respectively, are transcribed from the 5' to 3' direction. Two contiguous genes (YB and YA), which likely also arose by gene duplication, are transcribed in the opposite direction, using promoter sequences in the 5' region of CYP21A. The orientation of transcription is indicated by the arrows. (B) Molecular defects of CYP21B. Missense mutations, resulting in amino acid substitutions, are shown above the gene structure. The standard single letter amino acid code has been used. Other deleterious mutations, including nonsense mutations, splicing defects, and nucleotide deletions causing frameshifts, are shown below the gene. (Single letter amino acid code: A, alanine; C, cysteine; D, aspartic acid; E, glutamic acid; F, phenylalanine; G, glycine; H, histidine; I, isoleucine; K, lysine; L, leucine; M, methionine; N, asparagine; P, proline; Q, glutamine; R, arginine; S, serine; T, threonine; V, valine; W, tryptophan; Y, tyrosine. X denotes a nonsense mutation introducing a termination codon.)

mia, hyperkalemia, and hyperreninemia in those with the saltwasting form. An ACTH stimulation test is not necessary to make the diagnosis of CAH in a virilized female infant. Genetic and Molecular Pathophysiology. P450c21 is encoded by the CYP21 gene (CYP21B) located on chromosome 6 at position p21.3, within about 2 centiMorgans of the HLA complex, with which it has tight linkage. CYP21 and CYP17 (encoding P45017α) are thought to have arisen from a common ancestor gene. CYP21 contains 10 exons and its 2.0-kb cDNA encodes a 55-kDa protein. There is a homologous gene, generally considered a pseudogene, designated CYP21P (or CYP21A), located 30-kb upstream. This gene contains three deleterious mutations: an 8-bp deletion in exon 3, a T insertion in exon 7, and a stop codon in exon 8, such that it does not encode P450c21. However, the finding that sequences within this gene are transcribed has led to speculation that it may not in fact be a pseudogene. The gene encoding the fourth component of complement (C4B), is located centromeric to CYP21 and it too has a homologous pseudogene (C4A) near the CYP21 pseudogene. These two gene-pseudogene pairs are believed to have arisen through a duplication event. The genetic arrangement is further complicated by the presence of two additional genes, termed XB, and YB, that overlap CYP21 on the complementary DNA strand. XB is transcribed in the opposite direction to produce an extracellular matrix protein termed tenascin-X; the product of

YB is unclear. A homologous pair of genes, XA and YA, overlap CYP21A on the complementary DNA strand. CYP21 expression is regulated by ACTH via cAMP, using a specific cAMP-responsive sequence in the 5'-flanking region of the gene that binds a putative adrenal-specific nuclear protein (ASP). In addition, SF-1 activates the P450c21 promoter, as does another homologous orphan nuclear receptor nerve growth factor-induced gene-B (NGFI-B). Expression of this latter transcription factor increases dramatically in the adrenal cortex in response to stress. Interestingly, expression of P450c21 has recently been demonstrated in skin. Inheritance of 21-hydroxylase deficiency is autosomal recessive and the incidence of the disorder ranges from 1:15,000 (some white populations) to 1:700 (Yupik Eskimo tribe), the general incidence being reported as approximately 1:5,000. Molecular defects in CYP21 have been elucidated in approximately 100 patients with 21-hydroxylase deficiency. These almost all appear to have arisen as a result of recombination events between CYP21 and its homologous pseudogene, CYP21P, and result in either deletion of CYP21, or transfer of mutations from the pseudogene to the functional gene (a process termed “gene conversion”). This phenomenon has been suggested to account for the predominance of 21-hydroxylase deficiency over other forms of CAH. Complete deletion of CYP21 occurs in about 20% of cases of the classic saltwasting form. In a few cases the deletion has included the neigh-

554

SECTION VII / ENDOCRINOLOGY

Figure 57-9 Molecular defects of CYP11B1. Mutations in CYP11B1 causing 11β-hydroxylase deficiency and female pseudohermaphroditism have been reported in a number of individuals. Missense mutations are shown above the gene; frameshift (FS) and nonsense mutations are shown below the gene. Mutation R448H is found commonly in Moroccan Jews. This residue is probably required for heme binding. The majority of mutations have been found in exons 6–8.

Figure 57-10 Molecular defects of CYP19 gene. The CYP19 gene (encoding the aromatase enzyme, P450arom) has a complex structure in its 5' region: there are six untranslated exons (shaded boxes, locations not clearly known) and six alternate promoters used in different tissues (designated “P”); translation begins within exon 2. Rare mutations causing female pseudohermaphroditism have been reported. The two missense mutations shown above the gene denoted with asterisks were present in the same patient, who was therefore a compound heterozygote. Because of its role in heme binding, cysteine 437 is highly conserved. A splice junction mutation resulting in an 87-bp insertion is depicted below the gene. The patient was homozygous for this mutation.

boring C4 gene. However, as in most other genetic disorders, single-base mutations are common, and cause truncation, aberrant splicing, or amino acid substitutions. In a comprehensive study of 88 families, Speiser et al. determined that the most common mutation was an A-G change in the second intron affecting mRNA splicing (26%); large deletions had occurred in about 21% of affected individuals; substitution of isoleucine 172 by asparagine was found in 16%; and replacement of valine 281 by leucine in 11%. Homozygosity for severe mutations is present in about 50% of those with classic salt-wasting 21-hydroxylase deficiency. Correlations between enzyme activity (phenotype) and mutation (genotype) were noted, although some patients’ enzymatic phenotypes were not as predicted from their genotype. Compound heterozygosity for different CYP21 mutations is common. The clinical and enzymatic findings of such patients results from the combined effect of different mutations in each allele. In vitro analysis of mutant 21-hydroxylase enzymes has been performed in a number of cases. A mutant containing threonine at position 428 in place of cysteine has complete loss of enzymatic activity and heme binding. Of note, cysteine 428 is invariant among all P450 enzymes and is believed to be the heme-binding site. Other amino acid substitutions depress enzyme activity to varying degrees. 11β-HYDROXYLASE DEFICIENCY 11β-Hydroxylase is a cytochrome P450 enzyme that catalyzes the final step in cortisol synthesis, 11β-hydroxylation of 11-deoxycortisol to cortisol in the zona fasciculata of the adrenal gland. A related isozyme in the zona glomerulosa, aldosterone synthase, catalyzes the final three

steps in aldosterone synthesis, hydroxylation and dehydrogenation (oxidation) of 11-deoxycorticosterone to aldosterone (these latter activities are also referred to as corticosterone methyl oxidase types I and II [CMO I; CMO II]). Deficiency of 11β-hydroxylase results in a relatively common form of congenital adrenal hyperplasia, accounting for approximately 5% of cases, depending on ethnic background. This enzyme is not involved in gonadal steroidogenesis. Clinical Features and Diagnosis. There is a relatively high frequency of 11β-hydroxylase deficiency in Saudi Arabia and in Moroccan and Iranian Jews because consanguinity in these families is common. Affected female (46,XX) infants present with variable genital virilization, similar to those with 21-hydroxylase deficiency. In a number of kindreds, affected females have been so severely virilized that they have been reared as males, the diagnosis delayed until puberty, when breast development and menses occurred. 11Deoxycortisol is massively elevated and serum concentrations of 11-deoxycorticosterone, adrenal androgens, testosterone (by conversion from androstenedione), and ACTH are also increased. The distinguishing clinical feature of this condition is the presence of hypertension, induced by increased secretion of 11-deoxycorticosterone, the mineralocorticoid activity of which causes sodium retention and suppression of plasma renin activity. Hypokalemia resulting from the mineralocorticoid excess is variably present. There is minimal correlation between the severity of the virilization and the hypertension. Precocious pseudopuberty and advanced skeletal maturation occurs in untreated cases of both sexes.

CHAPTER 57 / SEX DETERMINATION AND DIFFERENTIATION

Genetic and Molecular Pathophysiology. There are two genes encoding 11β-hydroxylase isozymes. The 6.5-kb, 9-exon CYP11B1 gene encoding P450C11(P450XIB1) is localized to 8q21. A contiguous gene, CYP11B2, located approximately 40 kb away within the same chromosomal locus, encodes a second 11β-hydroxylase isozyme (P450cmo) that displays 93% amino acid homology with the first. CYP11B1 and CYP11B2 are structurally homologous to CYP11A, the gene that encodes P450scc, and the trio represent a subfamily within the P450 superfamily. CYP11B is expressed in the zona fasciculata of the adrenal cortex, under regulation by ACTH via cAMP. Like genes encoding other steroidogenic enzymes, the CYP11B promoter also contains a binding site for SF-1. P450C11 is a mitochondrial protein predicted to contain 479–503 amino acids, including an amino-terminal mitochondrial signal sequence. P450C11 catalyzes 11β-hydroxylation of 11-dexoxycortisol to cortisol in the zona fasciculata of the adrenal gland. It also has limited 18-hydroxylase activity. CYP11B2, the gene expressed in the zona glomerulosa, encodes the isozyme P450cmo (P450aldo; P450XIB2), which catalyzes the three-step conversion of 11-deoxycorticosterone to aldosterone (11-hydroxylation, 18-hydroxylation, and 18-oxidation). Its expression is regulated by angiotensin II. Mutations in CYP11B1 have been reported in at least 12 individuals with 11β-hydroxylase deficiency. Reported mutations include a 2-bp insertion and a 1-bp deletion (both of which cause a frameshift, deleting the enzyme’s heme-binding domain); singlebase mutations introducing termination codons (nonsense mutations), and a number of missense mutations causing amino acid substitutions. Substitution of arginine by histidine at position 448 is the predominant defect in P450C11 in Moroccan Jews, suggesting a founder effect. This amino acid is highly conserved and believed to be required for heme binding. Transient transfection studies reveal that this mutation abolishes enzymatic activity. Despite this, individuals heterozygous for this mutation have no demonstrable hormonal abnormalities, indicating that approximately 50% of enzyme activity is adequate for normal steroidogenesis. Hydroxylase activity is also abolished in other mutant enzymes containing different amino acid substitutions. The majority of mutations cluster in exons 6–8 of the CYP11B1 gene, suggesting that this region encodes residues critical for enzymatic activity. Notably, clinical variation has been observed between individuals with the same mutation. The mutations reported to date are de novo point mutations. The relatively high frequency of mutations in this gene is suggested to result from its high frequency of CpG dinucleotides, a well-recognized mutational hot spot in the human genome, rather than from recombination between CYP11B1 and CYP11B2 (as occurs between CYP21 and its adjacent pseudogene CYP21A). Mutations in CYP11B2 cause aldosterone synthase deficiency, which can present in infancy with hyponatremic, hypovolemic shock, but does not affect sexual differentiation in either sex. 3Β-HYDROXYSTEROID DEHYDROGENASE DEFICIENCY 3β-Hydroxysteroid dehydrogenase (3β-HSD) is a noncytochrome, membranebound enzyme that requires NAD+ as a cofactor to catalyze the conversion of 3β-hydroxy-∆ 5 -steroids (pregnenolone, 17hydroxypregnenolone, and dehydroepiandrosterone) to the 3-keto∆ 4-steroids (progesterone, 17-hydroxyprogesterone, and ∆4-androstenedione; Fig. 57-7). Deficiency of 3β-HSD, which is less common than 21-hydroxylase deficiency, impairs steroidogenesis in the adrenals and gonads, resulting in reduced synthesis of cortisol, aldosterone, and gonadal steroids.

555

Clinical Features and Diagnosis. Deficiency of 3β-HSD produces significant genital ambiguity in males, as described in detail above. However, in karyotypic females 3β-HSD deficiency is associated with only mild genital virilization (mild clitoromegaly or posterior labial fusion) not of sufficient severity to suggest genital ambiguity. Further clitoral enlargement and premature development of pubic and axillary hair may occur during childhood in untreated patients. The androgenization in females results from exaggerated levels of the relatively weak androgen dehydroepiandrosterone. Some cases have associated salt wasting because of reduced synthesis of aldosterone and cortisol. Genetic and Molecular Pathophysiology. Details of the genetics and molecular biology of this gene are provided in the preceding section on 3β-HSD deficiency as a cause of male pseudohermaphroditism and will not be reviewed here. One affected female without genital ambiguity was found to be homozygous for a nonsense mutation that produced a truncated protein of 169 amino acids, compared with the usual 371 amino acids. Compound heterozygosity for a missense mutation and an intronic mutation likely causing a splicing defect was found in another female with normal genitalia, whose affected brother had ambiguous genitalia. Other mutations in affected females will likely be similar to those found in affected males and their effect on genital development will probably reflect the nature and severity of the mutation. Aromatase Deficiency The conversion of androgens to estrogens is controlled by aromatase (P450arom), a cytochrome P450 enzyme located in the endoplasmic reticulum of estrogenproducing cells. Using NADPH-cytochrome P450 reductase as a cofactor, this microsomal enzyme catalyzes conversion of C19 steroids (androstenedione and testosterone) to C18 estrogens (estradiol, estrone, estriol), by modification of the steroid A ring to a phenolic ring. Clinical Features and Diagnosis. Androgens produced by the fetal adrenal gland and then desulfated and aromatized by the placenta are the major source of circulating estrogens during pregnancy. An apparently rare cause of female pseudohermaphroditism is the inability to convert fetal androgens to estrogens because of lack of placental aromatase activity. Three cases of aromatase deficiency have been described. In two cases, the mothers developed progressive virilization in the latter part of pregnancy, which resolved after delivery; serum androgens were high and estrogens low. Despite this, growth and development of the fetuses and placentas throughout gestation were normal. In the first case in vitro assays of the placenta after delivery revealed negligible aromatase activity. Absence of maternal virilization in the third case suggested that some aromatase activity was retained. The 46,XX affected infants in each case had male-appearing or ambiguous external genitalia with marked clitoral enlargement, rugation and fusion of labioscrotal folds, and a single meatus at the base of the phallic structure. The virilization results from conversion of dehydroepriandrosterone sulfate (DHEAS) to androstenedione and testosterone by the placenta; its inability to aromatize these steroids to estrogens results in massive elevations of these compounds. Most recently, a pair of affected 46,XX and 46,XY siblings have been reported. The female had ambiguous genitalia, similar to those described above; the male was normally masculinized. The two affected 46,XX subjects who are now postpubertal exhibited features of androgen excess from puberty onwards. Clitoral enlargement and development of facial acne were noted, in

556

SECTION VII / ENDOCRINOLOGY

addition to absence of breast development, attributed to deficiency of ovarian aromatase. Gonadotropins were modestly elevated, accompanied by high adrenal androgen concentrations, high plasma testosterone, low estradiol concentrations, and ovarian cysts on pelvic ultrasound. Because estrogens are the hormones primarily responsible for skeletal maturation, this was delayed, resulting in tall stature. There was also significant osteoporosis in the affected adult male. These cases, as well as the single reported case of an estrogen receptor (ER) mutation and the evidence from ER-deleted transgenic mice, indicate that contrary to long-standing beliefs, estrogens are not required for fetal survival. Genetic and Molecular Pathophysiology. Aromatase (P450arom) is a cytochrome P450 enzyme encoded by a 9-exon, 70-kb gene designated CYP19, whose chromosomal locus is 15q21.1. P450arom is expressed in a wide variety of human tissues, including the granulosa and luteal cells of the ovary, testicular Sertoli and Leydig cells, the placenta, adipose tissue, brain, muscle, and liver and in the preimplantation blastocyst. Expression levels and transcript sizes are quite different in these various tissues and expression is regulated in part by the use of tissue-specific alternative promoters, in ovary, placenta, brain, and adipose tissue. However, because the translational start site is conserved in the various mRNA species, the same protein is expressed in all tissues. Like that of other steroidogenic enzymes, the aromatase gene promoter contains a binding site for SF-1, and it appears that this site is involved in the cAMP-mediated regulation of gene expression. The expressed protein is similar to other P450 enzymes, containing a carboxyterminal heme-binding region encoded by the 9th exon. The defects in the above cases were inherited in an autosomal recessive pattern and there was known parental consanguinity in one case. In the first case described, the affected girl was homozygous for a splice junction point mutation that resulted in translation of an abnormal peptide containing an extra 29 amino acids. In vitro analysis revealed that the mutant enzyme retained only a minimal level of activity. In the second case, the affected individual was found be a compound heterozygote for two single-base mutations that introduced two separate amino acid substitutions into the enzyme: arginine to cysteine at position 435 and cysteine to tyrosine two residues downstream, at amino acid 437. In vitro analyses of the mutant enzymes revealed extremely low activity in the presence of the R435C substitution, and complete absence of activity with the C437Y defect. Cysteine 437 is very highly conserved, and is apparently involved in heme binding, hence the destructive effect of the mutation on enzyme function. In the final case, the mutation replaces a highly conserved arginine residue with cysteine at position 375 of the peptide. The mutant protein expressed in vitro had only 0.2% of the activity of wild-type P450arom. The region of the protein in which arginine 375 is located is postulated to be involved in anchoring the enzyme to the cell membrane. Molecular defects in CYP19 are summarized in Fig. 57-10. Disorders of Müllerian Development MAYER-ROKITANSKYKÜSTER-HAUSER (MRKH) SYNDROME Although individuals with this disorder clearly are phenotypic females, the condition technically represents the mildest form of female pseudohermaphroditism, in parallel with PMDS in the male. This developmental abnormality appears to result from defective Müllerian duct fusion in early gestation; however, the exact nature of the defect remains unclear. Affected girls usually present in their teens with primary amenorrhea in the presence of otherwise normal secondary sexual

development. Examination reveals absence or severe hypoplasia of the vagina; uterine agenesis is usual; however, some uterine development (uni- or bicornuate uterus), or occasionally a normal uterus, may be present. The ovaries are normally developed. There appear to be two subtypes of the disorder—the typical (isolated) and atypical forms, frequency of each being approximately equal. The typical form is characterized by the laparoscopy or laparotomy findings of symmetric muscular buds (the Müllerian remnants) and normal Fallopian tubes. The atypical form has asymmetric hypoplasia of one or both buds, with or without dysplasia of the Fallopian tubes. The atypical form has associated anomalies including renal defects (agenesis or ectopia in about 30–50% of patients) and skeletal abnormalities (vertebral malformations; Klippel-Feil anomaly) of varying severity. The MURCS association comprises Müllerian duct aplasia, renal agenesis or ectopia, and cervical somite dysplasia (Klippel-Feil). This represents the most severe form of the disorder and may represent a mesodermal malformation spectrum. Laparoscopy is required to distinguish the typical from the atypical form of the MRKH syndrome. Because of these associated anomalies, all patients with vaginal atresia should have skeletal radiographs and renal and pelvic ultrasound performed. The MRKH syndrome occurs in 1 in 4000–5000 females. Most cases are sporadic; however, a genetic defect likely underlies familial cases (approximately 5%), which have been reported in patterns consistent with sex-limited autosomal dominant or autosomal recessive inheritance. Galactose-1-phosphate uridyl transferase activity is reduced in some patients and may represent a candidate gene for MRKH syndrome.

DIAGNOSIS AND MANAGEMENT OF DISORDERS OF SEX DETERMINATION AND SEX DIFFERENTIATION DIAGNOSIS Abnormalities of sexual development require evaluation by an experienced team, including a pediatric endocrinologist, urologist, and geneticist. In the newborn period, rapid but careful diagnosis and early, appropriate sex assignment are essential to optimize parental adjustment to the child’s apparently incongruous genital appearance and to minimize subsequent psychosocial problems for the child. The principal aims of the diagnostic evaluation are to determine (1) presence of potentially life-threatening adrenal steroid deficiencies and electrolyte derangements; (2) chromosomal sex; (3) type and functional status of the gonads; (4) internal genital anatomy; and (5) most appropriate sex-of-rearing. Consanguinity between parents is suggestive of autosomal recessive conditions, such as CAH and 5α-reductase deficiency. Likewise, the presence of similarly affected siblings or a family history of ambiguous genitalia, infertility, amenorrhea, lack of pubertal development, or sudden death in infancy is suggestive of a genetically determined defect. A history of genital abnormalities, severe gynecomastia, or infertility in maternal relatives suggests an X-linked condition, such as androgen insensitivity. A history of maternal virilization during pregnancy could suggest aromatase deficiency. Clinical, ultrasound, and radiographic examination should define the following features: 1. External genital anatomy. Specifically, the size of the phallic structure (stretched penile length, excluding foreskin,

557

CHAPTER 57 / SEX DETERMINATION AND DIFFERENTIATION

Table 57-5 Steroid Profiles in Disorders of Sex Differentiation Disorder LCH CLAH (StAR) 3β-HSD 17α-OH 17β-OH 5α-Red AIS 21-OH 11β−OH Aromatase deficiency

Prog

DOC

Aldo

17OHPreg

17OHP

11-Deoxycortisol

Cortisol

DHEA

∆4A

Testo

DHT

N

N

N

N

N

N

N

N

N

↓↓

↓↓

↓ N or ↑ ↑ N N N ↑ N N

↓ ↓ ↑ N N N ↓ ↑ N

↓ ↓ ↓ N N N ↓ ↓ N

↓ ↑↑ ↓ N N N ↑ N or ↑ N

↓ N or ↑ ↓ N N N ↑↑↑ N or ↑ N

↓ ↓ ↓ N N N ↓ ↑↑↑ N

↓ ↓ ↓ N N N ↓ ↓ N

↓ ↑↑ ↓ N or ↑ N N ↑ ↑ ↑↑↑

↓ N or ↑ ↓ ↑ N N ↑ ↑ ↑↑↑

↓ N or ↑ ↓ ↓ ↑ N or ↑ ↑↑ ↑ ↑↑↑

↓ N or ↓ ↓ ↓ ↓ N or↑ NA NA ↑↑↑

LCH, leydig cell hypoplasia; CLAH, congenital lipoid adrenal hyperplasia; StAR, steroidogenic acute regulator protein; 3β-HSD, 3βhydroxysteroid dehydrogenase deficiency; 17α-OH, 17α-hydroxylase deficiency; 17β-OH, 17β-hydroxylase deficiency; 5α-red, 5α-reductase deficiency; AIS, androgen insensitivity syndrome; 21-OH, 21-hydroxylase deficiency; 11β-OH, 11β-hydroxylase deficiency. Prog, progesterone; DOC, deoxycorticosterone; Aldo, aldosterone; 17OHPreg, 17α-hydroxypregneneolone; 17OHP, 17α-hydroxyprogesterone; DHEA, dehydroepiandrosterone; 䉭4A, androstenedione; Testo, testosterone; DHT, dihydrotestosterone.

2.

3.

4.

5.

and diameter at midshaft), location of the urethral meatus, presence of a separate vaginal orifice, size, and fusion and rugation of labioscrotal folds provide information regarding the degree of fetal androgenization. Hyperpigmentation of the genitalia represents evidence of ACTH excess as a result of some form of CAH. Location of gonads. Gonads in the inguinal region are highly likely to be testes (ovarian herniation occurs exceptionally rarely), suggesting that the infant is probably a genotypic male (or much less likely, another karyotype with presence of SRY). If no gonads are palpable in a partially masculinized infant, ultrasound examination is required to locate the gonads (which may be anywhere from the inguinal ring to the abdomen) and to examine for associated developmental abnormalities of the renal tract. Presence or absence of a uterus. Because of exposure to maternal estrogens, the uterus is enlarged and readily detectable by rectal examination or ultrasonography in newborn females. Presence of a uterus reflects absence of AMH effect during gestation, generally indicating that the gonads are not testes (or, if testes are present, that AMH action was absent during gestation). Internal genital anatomy. Cystoscopy or voiding cystourethrogram is required to define the anatomy of the lower urogenital structures, including type of urethra (long, maletype vs short, female-type), presence of a vagina or more rudimentary structure such as a prostatic utricle, and presence and location of the vaginal orifice (perineal vs urethral). The longer the urethra, smaller the vaginal structure, and higher its entry to the urethra, the greater the degree of androgenization that occurred in utero. Presence of other dysmorphic features or intrauterine growth retardation. These findings may suggest that the genital defect is part of a more generalized disturbance of morphogenesis.

Despite careful examination, it is almost impossible to differentiate between many of the disorders of sex determination and

differentiation on clinical grounds alone. Detailed laboratory evaluation is urgent if dealing with an infant with ambiguous genitalia in whom the sex of rearing is yet to be determined. Hormone secretion in the immediate postnatal period is dynamic, with hormone concentrations changing rapidly over the first few days of life. Therefore, optimally, blood should be drawn within the first 36 h of life for hormone analysis. Primary care institutions should be instructed that for any infant with a concern regarding genital development, 10 mL of blood should be drawn immediately and the serum frozen for later use. The following studies should be undertaken urgently on fresh blood: 1. Karyotype. In all cases a formal karyotype is required. A buccal smear is inadequate because of frequent false-negative results. 2. Electrolytes and adrenal steroids for evidence of congenital adrenal hyperplasia. 17-Hydroxyprogesterone (for 21-hydroxylase deficiency); 11-deoxycortisol (for 11-hydroxylase deficiency); 17-hydroxypregnenolone and dehydroepiandrosterone (for 3β-hydroxysteroid dehydrogenase deficiency); pregnenolone; progesterone; 17-hydroxypregnenolone; and 17-hydroxyprogesterone (for 17-hydroxylase deficiency); and in most cases, plasma renin activity should be measured on the initial blood sample. Other more discriminating steroids can be measured on the stored, frozen serum once the karyotype and initial critical values are obtained. Normal electrolytes in the immediate postnatal period do not exclude congenital adrenal hyperplasia, and until definitive results are obtained, these should be monitored for development of hyponatremia and hyperkalemia. 3. Testosterone (T), dihydrotestosterone (DHT), and T precursors to determine the presence of a defect of T biosynthesis or 5α-reductase deficiency in karyotypic males (Table 57-5). Serum T is high (>200 ng/dL) in cord blood, but plummets to become almost undetectable at the end of the first week of life, rising from about 2–3 weeks, to peak at >200 ng/dL at around 8–12 weeks of age. Serum T falls

558

SECTION VII / ENDOCRINOLOGY

again from 12 to 16 weeks, finally becoming essentially unmeasurable by 6 months of age. Steroid analysis should therefore be performed during periods of maximal testicular activity, or, if this is not possible, after testicular stimulation with hCG (gonadal steroids and DHT are measured before, and 24 h after, administration of 1500 IU hCG every other day for three doses). Normal infants respond with T values >200 ng/dL, often much higher. An increased ratio of T to DHT (>20:1) during periods of active testicular steroidogenesis or after hCG is found in 5α-reductase deficiency. Hormonal profiles found in defects of T biosynthesis are summarized in Table 57-5. T and DHT concentrations are probably within normal limits in AIS; however, increased concentrations have been reported in a few infants with partial AIS. Minimal data are available for infants with complete AIS. 4. LH and FSH. LH concentration should be markedly increased in infants with Leydig cell hypoplasia and may also be increased in infants with partial AIS. FSH is increased in those with gonadal dysgenesis, but may not be elevated in early infancy. 5. ACTH stimulation test. This is required to characterize steroidogenic defects that affect the adrenals as well as the gonads (CLAH, 3β-hydroxysteroid dehydrogenase deficiency, and 17-hydroxylase deficiency). In virilized karyotypic females, an ACTH stimulation test is not required for diagnosis of 21-hydroxylase deficiency, but may aid diagnosis of other forms of CAH by exaggerating precursor steroid levels. 6. Gonadal biopsy. This may be required in cases of suspected gonadal dysgenesis. If concentrations of testosterone, DHT, and steroid precursors are normal in a karyotypic male with abnormal masculinization, the differential diagnosis then likely rests between a partial form of AIS and the perplexing “black box” of undiagnosable forms of male pseudohermaphroditism. MANAGEMENT Sex of Rearing Although an accurate diagnosis is important, sex assignment is based primarily on gonadal and genital morphology rather than diagnosis. It has been usual practice to assign female sex-of-rearing to any infant with ovaries, no matter how virilized, since she is potentially fertile. The converse principle has not generally been applied for infants with testes, since it has been argued that successful functional outcome after reconstruction of severely undermasculinized genitalia is unlikely. This is clearly a vital area of management, and one that is currently undergoing re-examination. Unfortunately, a detailed discussion of the principles, practice, and outcome of sex assignment of infants with disorders of sex determination and differentiation is beyond the scope of this chapter. Gonadectomy Gonadectomy is required for all individuals reared as females who have testes or gonadal dysgenesis and a Y-bearing cell line. However, opinion regarding the optimal timing of this procedure varies. For individuals with 46,XY or mosaic forms of gonadal dysgenesis, gonadectomy is advisable at the time of diagnosis, because of the high risk of malignancy. In those reared as females who have retained production of and responsiveness to testosterone, gonadectomy should be undertaken at the time of initial genital reconstructive surgery to prevent potential virilization at puberty. In those with complete AIS (therefore no

response to T), the timing of gonadectomy is the subject of debate. Some physicians recommend early gonadectomy for psychological reasons; others prefer to defer this until after puberty to take advantage of the excellent spontaneous feminization that results from endogenous testicular estrogen secretion. In following the latter course it is important that the young woman be made aware in early adolescence that the gonads will need to be removed at the completion of puberty because of the risk of malignancy. If inguinal hernia repair is required in early childhood, simultaneous gonadectomy is considered advisable, to obviate the need for a second surgical procedure. Hormone Replacement Therapy Gonadectomized individuals and females with CLAH or aromatase deficiency require estrogen supplementation at the appropriate time for puberty and thereafter, to induce and maintain feminization. The addition of cyclic progesterone is generally recommended once feminization is established, and is mandatory for those with a uterus. For partially virilized infants reared as male, testosterone therapy is usually required in infancy to enhance penile size before surgery, and may be beneficial at puberty to optimize virilization. Certain individuals with partial forms of AIS may require and respond to highdose testosterone therapy. Despite the enzymatic defect in conversion of testosterone to DHT, this therapy may also normalize serum DHT concentrations and improve virilization in some men with 5α-reductase deficiency. Affected individuals of either sex with any form of congenital adrenal hyperplasia require standard replacement therapy with glucocorticoids and mineralocorticoids. Treatment with physiologic doses of glucocorticoids suppresses ACTH production, thus reducing levels of precursor steroids that otherwise produce detrimental effects, such as virilization, advanced skeletal maturation, and, in the 17-hydroxylase and 11-hydroxylase deficiency forms of CAH, hypertension. The addition of antiandrogens to the standard therapy for CAH is currently under investigation. Prenatal Diagnosis and Treatment of CAH Prenatal diagnosis and treatment of female fetuses affected by virilizing 21hydroxylase deficiency has been undertaken successfully in a number of cases. This procedure requires that the molecular defect has been characterized in a prior affected sibling. Dexamethasone treatment is initiated as soon as pregnancy is confirmed. At 10 weeks’ gestation, a chorionic villus biopsy is performed and karyotype determined. If the karyotype is 46,XY, dexamethasone is suspended because masculinization is normal in affected males. If the karyotype is 46,XX, dexamethasone is continued until the status of the fetus is determined by molecular analysis of CYP21. If the 46,XX fetus is determined to be unaffected, dexamethasone treatment is suspended; if affected, the treatment is continued until term. Although not yet reported, this procedure would also be applicable in other forms of CAH, provided the molecular defect has been determined in an older sibling. Maternal complications of dexamethasone treatment are not insignificant (often quite severe Cushingoid changes), and such management should be undertaken only under the guidance of a team experienced with this therapy.

SELECTED REFERENCES Andersson S, Geissler WM, Wu L, et al. Molecular genetics and pathophysiology of 17 beta-hydroxysteroid dehydrogenase 3 deficiency. J Clin Endocrinol Metab 1996;81:130–136. Andersson S, Bishop RW, Russell DW. Expression cloning and regulation of steroid 5alpha-reductase, an enzyme essential for male sex differentiation. J Biol Chem 1989;264:16,249–16,255.

CHAPTER 57 / SEX DETERMINATION AND DIFFERENTIATION

Bose HS, Pescovits OH, Miller WL. Spontaneous feminization in a 46, XX female patient with congenital lipoid adrenal hyperplasia due to a homozygous frameshift mutation in the steroidogenic acute regulatory protein. J Clin Endocrinol Metab 1997;82:1511–1515. Cramer DW, Goldstein DP, Fraer C, Reichardt JK. Vaginal agenesis (Mayer-Rokitansky-Kuster-Hauser syndrome) associated with the N314D mutation of galactose-1-phosphate uridyl transferase. Mol Hum Reprod 1996;2:145–148. Curnow KM, Slutsker L, Vitek J, et al. Mutations in the CYP11B1 gene causing congenital adrenal hyperplasia and hypertension cluster in exons 6, 7 and 8. Proc Natl Acad Sci USA 1993;90:4552–4556. Foster JW, Dominguez-Steglich MA, Guioli S, et al. Campomelic dysplasia and autosomal sex reversal caused by mutations in an SRY-related gene. Nature 1994;372:525–530. Goodfellow PN, Lovell-Badge R. SRY and sex determination in mammals. Annu Rev Genet 1993;27:71–92. Ingraham HA, Lala DS, Ikeda Y, et al. The nuclear receptor steroidogenic factor 1 acts at multiple levels of the reproductive axis. Genes Dev 1994;8:2302–2312. Josso N. AntiMullerian hormone: new perspectives for a sexist molecule. Endocr Rev 1986;7:421–433. Labrie F, Luu-The V, Lin SX, et al. The key role of 17 beta-hydroxysteroid dehydrogenases in sex steroid biology. Steroids 1997;62:148–158. Lin D, Sugawara T, Strauss JF III, et al. Role of steroidogenic acute regulatory protein in adrenal and gonadal steroidogenesis. Science 1995;267:1828–1831. Luo X, Ikeda Y, Parker KL. A cell-specific nuclear receptor is essential for adrenal and gonadal development and sexual differentiation. Cell 1994;77:481–490. Luu-The V, Labrie C, Simard J, et al. Structure of two in tandem human 17beta-hydroxysteroid dehydrogenase gene. Mol Endocrinol 1990;4: 268–275. MacLean HE, Warne GL, Zajac JD. Intersex disorders: shedding light on male sexual differentiation beyond SRY. Clin Endocrinol 1997;46: 101–108. McPhaul MJ, Marcelli M, Zoppi S, Wilson CM, Griffin JE, Wilson JD. Mutations in the ligand-binding domain of the androgen receptor gene cluster in two regions of the gene. J Clin Invest 90:2097–2101. Morishima A, Grumbach MM, Simpson ER, Fisher C, Qin K. Aromatase deficiency in male and female siblings caused by a novel mutation and the physiological role of estrogens. J Clin Endocrinol Metab 1995;80: 3689–3698. Morrison-Graham K, Takahashi Y. Steel factor and c-Kit receptor: from mutants to a growth factor system. Bioessays 1993;15:77–83.

559

Muscatelli F, Strom TM, Walker AP, et al. Mutations in the DAX-1 gene give rise to both X-linked adrenal hypoplasia congenita and hypogonadotropic gonadism. Nature 1994;372:672–676. Newfield RS, New MI. 21-Hydroxylase deficiency. Ann N Y Acad Sci 1997;17:219–229. Pelletier J, Bruening W, Kashtan CE, et al. Germline mutations in the Wilms tumor suppressor gene are associated with abnormal urogenital development in Denys-Drash syndrome. Cell 1991;67:437–447. Penning TM. Molecular endocrinology of hydroxysteroid dehydrogenases. Endocr Rev 1997;18:281–305. Quigley CA, De Bellis A, Marschke KB, El-Awady MK, Wilson EM, French FS. Androgen receptor defects: Historical, clinical and molecular perspectives. Endcocr Rev 1995;16:271–321. Ramkissoon T, Goodfellow P. Early steps in mammalian sex determination. Curr Opin Genet Dev 1996;6:316–321. Segaloff DL, Ascoli M. The lutropin/choriogonadotropin receptor…4 years later. Endocr Rev 1993;14:324–347. Shen W-H, Moore CCD, Ikeda Y, Parker KL, Ingraham HA. Nuclear receptor steroidogenic factor 1 regulates the Mullerian inhibiting substance gene: a link to the sex determination cascade. Cell 1994;77:651–661. Shozu M, Akasofu K, Harada T, Kubota Y. A new cause of female pseudohermaphroditism: placental aromatase deficiency. J Clin Endocrinol Metab 1991;72:560–566. Simard J, Rheaume E, Sanchez R, et al. Molecular basis of congenital adrenal hyperplasia due to 3-beta hydroxysteroid dehydrogenase deficiency. Mol Endocrinol 1993;7:716–728. Simpson ER, Mahendroo MS, Means GD, et al. Aromatase cytochrome P450, the enzyme responsible for estrogen biosynthesis. Endocr Rev 1994;15:342–355. Simpson ER, Zhao Y, Agarwal VR, et al. Aromatase expression in health and disease. Recent Prog Horm Res 1997;52:185–213. Speiser PW, Dupont J, Zhu D, et al. Disease expression and molecular genotype in congenital adrenal hyperplasia due to 21-hydroxylase deficiency. J Clin Invest 1992;90:584–595. Wagner T, Wirth J, Meyer J, et al. Autosomal sex reversal and campomelic dysplasia are caused by mutations in and around the SRY-related gene SOX9. Cell 1994;79:1111–1120. Waterman MR. A rising StAR: an essential role in cholesterol transport. Science 1995;267:1780,1781. Wilson JD, Griffin JE, Russell DW. Steroid 5alpha-reductase 2 deficiency. Endocr Rev 1993;14:577–593. Yanase T, Simpson ER, Waterman MR. 17alpha-hydroxylase/17,20-lyase deficiency: from clinical investigation to molecular definition. Endocr Rev 1991;12:91–108.

CHAPTER 58 / SEX CHROMOSOME DISORDERS

58

561

Sex Chromosome Disorders ANDREW R. ZINN

INTRODUCTION Because sex chromosome disorders frequently result in abnormalities of sexual differentiation, the two subjects are usually discussed together in medical texts. The biology of human sexual differentiation and the pathophysiology of associated defects were discussed in the previous chapter. This chapter examines sex chromosome disorders from a genetic perspective, emphasizing unique features, such as X inactivation and dosage compensation, X-Y recombination, and male-specific functions of the Y chromosome. Table 58-1 summarizes karyotype, genotype, and phenotype data for selected disorders included in this chapter. The reader should consult an endocrinology textbook for more information about the clinical features of the various disorders.

STRUCTURE AND FUNCTION OF HUMAN SEX CHROMOSOMES The structures of the human X and Y chromosomes and the approximate positions of some genes of interest are depicted in Fig. 58-1. The X chromosome is about 160 Mb in length. It contains probably thousands of genes, encoding a variety of enzymes and structural and regulatory proteins. A distinguishing feature of most of these genes is that males have only one copy, accounting for X-linked recessive inheritance of many genetic diseases (see Transmission of Human Genetic Diseases, Chapter 4). By contrast, the much smaller Y chromosome is believed to contain a paucity of genes. The Y chromosome is divided into two parts. The euchromatic portion is about 30 Mb in length and contains all known Y-linked genes. The heterochromatic portion on the long arm can vary in length among normal men, averaging approximately 20 Mb. This region is composed of simple repetitive sequences and probably does not contain any genes. Because females do not require any Y-specific gene products, the Y chromosome has generally been regarded as functioning only in sexually dimorphic processes, such as testis formation and male gametogenesis. This notion is incorrect. As the discussion of Turner’s syndrome will show, the Y chromosome also contains genes involved in general viability, growth, and morphogenesis. Normally there is no recombination between the X and Y chromosomes except in the most distal regions, where the chromosomes pair and recombine during male meiosis. This pairing may be important for proper chromosome segregation. It is important From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

to note that the X and Y chromosomes bear identical copies of these recombining regions, and therefore genetic markers in these “pseudoautosomal” regions do not show strictly sex-linked inheritance. The short arm (p) pseudoautosomal region spans about 2.6 Mb and appears to be gene-rich. The region is known to contain genes for two cytokine receptor α-subunits (CSF2RA, IL3RA), an adenine nucleotide translocase (ANT3), two lymphocyte cell surface antigens (CD39, MIC2), the enzyme acetylserotonin methyltransferase (ASMT), a nuclear protein of unknown function (XE7), and the 5’ part of the gene for a red blood cell antigen (Xg). Thus far, the only gene assigned to the 0.4-Mb long arm (q) pseudoautosomal region is IL9R, which encodes another cytokine receptor. Only a handful of genes have been discovered in the strictly sex-linked portion of the Y chromosome. Known Y-specific genes encode the testis-determining factor (SRY), a zinc-finger protein of unknown function (ZFY), a ribosomal protein (RPS4Y), a homolog of the tooth-bud protein amelogenin (AMELY), the male transplantation antigen H-Y (SMCY), and several testis-specific proteins that may function in spermatogenesis (TSPY, RBM, and DAZ). Interestingly, the ZFY, RPS4Y, AMELY, and SMCY genes all have closely related X homologs, perhaps reflecting divergence of the sex chromosomes from an ancestral autosome pair. Alternatively, these X-Y homologous genes may be the result of subsequent transposition events. Data on the origin of X-Y genes from evolutionary comparisons of other species are equivocal.

X INACTIVATION AND DOSAGE COMPENSATION Unlike the genes just mentioned, the vast majority of X-linked genes do not have Y homologs. Thus males have only one copy of most X-linked genes, whereas females have two. This gender difference in the dosage of X-linked genes is balanced at the level of expression by the inactivation of one X chromosome in females during early embryogenesis, first postulated in the early 1960s by Mary Lyon. The choice of which X chromosome undergoes inactivation in each embryonic cell is normally random, but the pattern of inactivation propagates to daughter cells. Thus normal women are mosaic with regard to which X chromosome is active. The inactive X chromosome becomes hypermethylated and late-replicating, and in most tissues it condenses into the Barr body or sex chromatin. This process is irreversible except in oocytes, in which the inactive X is reactivated just before meiosis. Cytogenetic studies of X deletions and translocations indicate that a cis-acting region of the proximal long arm must be present for a chromosome to undergo X inactivation. The process appar-

561

562

Table 58-1 Features of Selected Sex Chromosome Disorders Disorder XY gonadal dysgenesis (Swyer syndrome, XY female, MIM 306100) XYp– Turner female XX male (MIM 278850) gene

Karyotype

Molecular cytogenetics Sexual phenotype

Gonads

Molecular defect

46,XY

Usually normal

Female

Streak gonads; risk of gonadoblastoma

Tall stature

SRY mutation in some cases

46,XY; 46,XYp–

Small Yp deletion, often with Xp;Yp translocation Y;X or Y;autosome

Female

Streak gonads; risk of gonadoblastoma

Turner-like, especially lymphedema

Deletion of SRY and nearby gene(s)

Male

Testes with azoospermia

Klinefelter-like

Presence of SRY

46,XX

translocation in most cases Xp21.2 disomy

in most cases Female

Streak gonads

No specific features

Overexpression of DAX1 gene?

Female

Streak gonads

Haploinsufficiency of SOX9 gene

Male

Hyalinized testes; azoospermia

Severe skeletal abnormalities, extraskeletal defects Poor to normal virilization; gynecomastia; long legs Short stature, webbed neck, aortic coarctation, cubitus valgus, others Turner-like (variable)

Turner’s syndrome

45,X and variants

Female

Streak gonads

45,X/46,XY mosaicism (including mixed gonadal dysgenesis) Gonadoblastoma (MIM 424500) Syndrome associated with small ring X chromosomes 46,XYXq syndrome

45,X/46,XY mosaic

Male, female, or ambiguous genitalia Female

Streak gonads or dysgenetic testis with risk of gonadoblastoma; normal testes in some cases Dysgenetic gonad

Female

46,XYq–

Ring X lacking X inactivation center in Xq13.2 Yq;Xq translocation

46,XY or 46,XYq-

Yq11.23 deletion

46,XY, 46,XYp–, 45,X/46,X,+mar 45,X/46,X,r(X) mosaic

Presence of Y material

Not known Not known

Variable

Not known

Streak ovaries

Mental retardation, multiple congenital anomalies

Male

Not reported

Male

Azoospermia; variable histology

Mental retardation, microcephaly, other anomalies Sometimes short stature

Functional disomy of unknown X-linked genes Functional disomy of unknown X-linked genes Deletion of DAZ gene?

MIM numbers from Online Mendelian Inheritance in Man, OMIM™, January 1996.

SECTION VII / ENDOCRINOLOGY

Dosage-sensitive sex 46,XY; 46,XY,dup(X); reversal (MIM 600191) unbalanced X translocation Camptomelic dysplasia 46,XY, sometimes with sex reversal abnormality involving (MIM 211970) 17q24.3-q25.1 Klinefelter’s syndrome 47,XXY and variants

Azoospermia (MIM 415000)

Extragonadal features

CHAPTER 58 / SEX CHROMOSOME DISORDERS

563

Figure 58-1 Structure and function of the human sex chromosomes. Cloned genes are italicized. Dashed lines indicate functional X-Y homologs. Vertical lines indicate loci not yet defined molecularly. pter, qter = telomores.

ently initiates within this region, termed the X inactivation center (XIC), and then spreads to the rest of the chromosome. A gene discovered fortuitously in the XIC region termed XIST (X-inactive specific transcript) has been implicated in the initiation of X inactivation. The XIST expression pattern is unique: the gene is transcribed only from the inactive X chromosome. Expression of XIST precedes chromatin condensation and transcriptional inactivation of other X-linked genes. The gene is transcribed as a >15-kb polyadenylated, alternatively spliced RNA that does not appear to encode any protein. The RNA is localized within the nucleus to the Barr body, where it may serve a structural function. Once XIST is expressed and X inactivation has occurred, the inactive state of the chromosome is maintained by DNA methylation and chromatin condensation. Although X inactivation is a chromosome-wide phenomenon, some specific genes escape X inactivation, i.e., they are transcribed from both the active and inactive X chromosomes. These genes are not dosage compensated. In some instances, Y-linked copies may serve instead to equalize the levels of gene products in males and females. For example, it appears that all short arm pseudoautosomal genes escape X inactivation. The X-linked homologs of the Y-linked genes ZFY, RPS4Y, and SMCY also escape X inactivation. However, some genes that escape X inactivation have no functional Y-linked homologs. Why such X-specific genes should escape inactivation is unclear (indeed, in most cases mouse orthologs undergo X inactivation). The signals that determine whether a gene is subject to X inactivation are unknown. Some genes that escape X inactivation are clustered, and local chromatin structure is likely to be important.

DISORDERS OF SEX DETERMINATION Classic studies indicated that phenotypic sex in the fruit fly Drosophila melanogaster is determined by the numerical ratio of X chromosomes to autosomes; the Drosophila Y chromosome is genetically inert with regard to sex determination. It was once thought that the same mechanism of sex determination would be true for mammals. However, karyotyping studies in the late 1950s showed that XO and XX humans are female, whereas XY and XXY individuals are male. These data established the Y chromosome’s primacy in human testis determination. Further cytogenetic and molecular genetic studies of rare “sex reversed” XX males and XY females culminated with the report in 1990 by Peter Goodfellow and colleagues of the single Y-linked gene, SRY (sex-determining region Y), that directs the differentiation of the bipotential gonad into testes. The human SRY gene encodes a protein of 204 amino acids that belongs to the high-mobility group (HMG) family of DNA-binding proteins. The gene is expressed in adult testis (where its function is unknown) and transiently in the gonadal ridge during embryogenesis. In all likelihood SRY acts by regulating transcription of other genes. The SRY protein binds to the minor groove of the double helix at specific sequences, and its binding induces a bend in the target DNA. It is not yet known which genes are regulated by SRY, nor is it certain whether SRY binding activates or inhibits transcription. One gene that might be activated by SRY binding is MIS, which encodes the Müllerian inhibiting substance that causes regression of the presumptive female reproductive tract. The discovery of SRY has clarified the etiology of most cases of XX sex reversal. The SRY gene and adjacent Y chromosome

564

SECTION VII / ENDOCRINOLOGY

Figure 58-3 Sexual phenotype of XY individuals with various Xp duplications. Figure 58-2

Aberrant Xp-Yp interchange.

sequences can be detected in most XX males by DNA hybridization or polymerase chain reaction (PCR) assays. The most frequent location of SRY sequences in these males is the short arm of the X chromosome. SRY’s propensity to be translocated from Yp to Xp can be explained by the gene’s proximity to the pseudoautosomal region. The mechanism of transposition involves aberrant X-Y interchange during male meiosis (Fig. 58-2). The translocated SRY gene is probably subject to X inactivation. Some cases of incomplete sex reversal, e.g., XX true hermaphroditism (presence of both testicular and ovarian tissue), may be due to mosaic SRY expression as a result of stochastic X inactivation. Mutations that eliminate SRY or affect the region of the protein involved in DNA binding cause XY sex reversal. However, only about 10% of XY females have detectable SRY mutations. The remaining 90% remain unexplained and are presumably caused by environmental factors or mutations in other genes required for testis determination. One such gene has already been identified because of its similarity to SRY. This gene, denoted SOX9 (SRY box), was cloned as a member of the SRY subfamily of HMG proteins. Lesions in one copy of the SOX9 gene, which maps to chromosome 17, have been discovered in several patients with camptomelic dysplasia, a rare skeletal disorder associated with XY sex reversal. The other copy of the gene in these patients is apparently normal. The mechanism by which half-normal dosage, or haploinsufficiency, of SOX9 causes both skeletal malformations and sex reversal is not yet known. Expression studies of the mouse homolog are consistent with the direct involvement of SOX9 in both abnormalities. Still another gene has been implicated in sex determination by studies of sex chromosome disorders. A number of XY females have been reported with partial X chromosome duplications that include band Xp21.1 of the short arm (Fig. 58-3). Because, in the absence of a second X chromosome, the duplicated X is not inactivated, sex reversal is thought to be caused by twofold increase in expression of a gene or genes within the duplication. Using molecular markers to study these subjects and 27 other unexplained 46,XY females, Bardoni et al. identified a submicroscopic Xp duplication in one subject and delimited the critical region of Xp21.1 associated with sex reversal to just 160 kb. They

designated the Xp locus causing sex reversal DSS (dosage-sensitive sex reversal). An attractive candidate gene for DSS known to lie within the critical region is DAX1, a member of the nuclear hormone receptor gene superfamily. Loss of function mutations in DAX1 cause congenital adrenal hypoplasia and hypogonadotropic hypogonadism but do not prevent testis determination. An interesting notion is that DSS may represent an ovary determining gene. Whether the DAX1 gene is indeed DSS should become clear from experiments using transgenic mice.

SEX CHROMOSOME ANEUPLOIDIES XY sex reversal associated with SOX9 haploinsufficiency or DSS duplication illustrates the critical importance of proper gene dosage. Abnormal gene dosage also plays a role in the phenotypes of numerical sex chromosome disorders, or aneuploidies. The most common sex chromosome aneuploidy is Klinefelter’s syndrome (47,XXY; 46,XY/47,XXY; and variants), with an incidence of about 1 in 500 male births. The extra X chromosome is usually the result of meiotic nondisjunction during paternal or maternal gametogenesis. The phenotype is variable but is characterized by small, firm testes, azoospermia, tall stature, eunuchoid habitus, gynecomastia, and elevated gonadotropin levels (Fig. 58-4). Azoospermia is caused by loss of spermatogenic cells. Animal studies suggest that this loss is somehow caused by the presence of an unpaired X chromosome at the pachytene stage of meiosis. Hyalinization of seminiferous tubules and impaired Leydig cell testosterone production follows loss of germ cells. Exogenous androgen is the treatment of choice for testosterone deficiency, whereas treatment of severe gynecomastia is surgical. The phenotype of Klinefelter’s syndrome is mild compared to disorders involving an extra autosome, for example trisomy 21 (Down’s syndrome). The reason for the milder phenotype in XXY men is that the additional X chromosome undergoes inactivation, whereas extra autosomes in other aneuploidy syndromes remain active. For this same reason, XXX females are usually normal, although they may be subfertile. Individuals with greater degrees of X aneuploidy (e.g., 48,XXXY; 48,XXXX; 49,XXXXY; or 49,XXXXX karyotypes) often show more severe abnormalities, even though all the supernumerary chromosomes undergo X inactivation. These abnormalities, which include mental retardation and skeletal malformations, are presumably caused by increased dosage of genes that escape X inactivation. Interestingly, some

CHAPTER 58 / SEX CHROMOSOME DISORDERS

565

Figure 58-4 Klinefelter’s syndrome. (Used with permission from Blackwell Science.)

Figure 58-5 Turner’s syndrome. (Used with permission from Blackwell Science.)

similar features have been noted in men with severe Y aneuploidy, such as 48,XYYY or 49,XYYYY karyotypes, suggesting that the culprit genes may be X-Y homologous or pseudoautosomal. By contrast to Klinefelter’s syndrome, the loss of one sex chromosome, or monosomy X, causes the more severe disorder known as Turner’s syndrome. Partial or complete monosomy X (45,X karyotype) is found in about 1 in every 3000 liveborn girls. The prenatal incidence is much greater: as many as 2% of all human conceptuses are estimated to be 45,X, but fewer than 1% survive to term. Paradoxically, liveborns with monosomy X have only modestly reduced life expectancy. An adult woman with Turner’s syndrome is shown in Fig. 58-5. As with Klinefelter’s syndrome, the phenotype is variable. Characteristic features include growth retardation, ovarian failure, and specific physical abnormalities,

such as webbed neck, aortic coarctation, increased carrying angle of the elbows (cubitus valgus), lymphedema, horseshoe kidney, and others. Metabolic and endocrine abnormalities are also frequent, including hypertension, glucose intolerance, and autoimmune thyroid disease. Girls with Turner’s syndrome are not typically mentally retarded, although as a group they show selective impairment of nonverbal cognitive skills such as visualspatial abilities. The pathophysiology of Turner’s syndrome is poorly understood. Growth retardation is not caused by growth hormone deficiency, although administration of pharmacologic doses during childhood accelerates growth and may increase final stature. Ovarian failure reflects rapid oocyte loss beginning around 6 months of gestation. Usually by early infancy there are few or no remain-

566

SECTION VII / ENDOCRINOLOGY

ing oocytes, and the ovaries become fibrous streaks. Infertility is the rule, with rare exceptions. Ovarian sex steroid production is usually deficient, and most girls with Turner’s syndrome require hormonal replacement therapy to induce puberty and to maintain cyclic menses. Some extragonadal manifestations, such as webbed neck, lymphedema of the extremities, and perhaps aortic coarctation, may be caused by delayed or defective lymphatic vessel development. The mechanism of chromosome loss in Turner’s syndrome is not known. Meiotic nondisjunction probably accounts for only a minority of cases. The single X chromosome is maternal in about three-fourths and paternal in about one-fourth of 45,X patients. There are no imprinting effects, i.e., the parental origin of the X does not appear to influence the phenotype. Many patients with Turner’s syndrome have karyotypic variants of monosomy X, most commonly mosaicism, the presence of two or more cell lines with different karyotypes (e.g., 45,X/46,XX). Mosaicism results from mitotic errors after conception. The second cell line often contains a structurally aberrant X chromosome. Mosaicism for a cell line with two or more copies of the X chromosome long arm enhances viability as compared with the nonmosaic 45,X constitution. Mosaicism for a Y-bearing cell line, most often 45,X/46,XY karyotype, deserves particular mention. The phenotype varies depending upon the proportion of XY cells and their tissue distribution. Bilateral testes may be present, or there may be unilateral testicular tissue with a contralateral streak gonad, a condition termed mixed gonadal dysgenesis (see Chapter 57). Regardless of the sexual phenotype, the presence of a Y chromosome confers a high risk of gonadoblastoma, a malignant tumor arising in the dysgenetic gonad. Prophylactic removal of streak gonads or histologically abnormal testes is recommended in any patient whose karyotype includes Y material. The Y-linked gene(s) predisposing to gonadoblastoma appears to be distinct from SRY, because XY females with SRY deletions or mutations are still at risk. Some studies using sensitive PCR assays have reported an alarmingly high prevalence of cryptic Y sequences in Turner’s syndrome patients. However, in the absence of testes or virilization, the clinical significance of low levels of Y material detected by PCR but not by karyotype is uncertain. Several genetic mechanisms underlie the phenotype of Turner’s syndrome. Oocyte loss may be caused in part by the effects of an unpaired X chromosome during meiosis, analogous to spermatogonia loss in Klinefelter’s syndrome. Haploinsufficiency, or halfnormal dosage of X-linked genes, may also contribute to ovarian failure. As previously noted, the inactive X chromosome is reactivated in oocytes; diploid expression of certain X-linked genes may be required for normal oocyte function. Consistent with this hypothesis, small deletions of Xq have been observed in some otherwise normal women with premature menopause. Anatomic abnormalities, such as lymphedema associated with Turner’s syndrome, suggest that diploid dosage of certain X-linked genes is important not only for ovarian function, but also for extragonadal development. Yet, after early embryogenesis, only one X chromosome is active in somatic tissues of normal women. This apparent contradiction can be explained if the genes responsible for extragonadal Turner features escape X inactivation. How then do we explain normal development in males, who have only one X chromosome? One possibility is male-specific upregulation of X-linked gene expression, the mechanism of dosage compensation in Drosophila. However, there is no evidence for such

upregulation in mammals. Another possibility is that in addition to its role in sex determination, the Y chromosome supplies a second copy of certain genes whose dosage is critical for normal somatic development. Present data, although not definitive, favor the latter hypothesis. Deletions of either the Xp or Yp pseudoautosomal region are associated with short stature, suggesting that at least part of the growth retardation in Turner’s syndrome is caused by haploinsufficiency of one or more pseudoautosomal genes. A strong candidate called SHOX (short stature homeobox-containing gene) was discovered by Rao and colleagues. Other evidence that X-Y homologous genes play a role in Turner’s syndrome comes from studies of Y deletions associated with sex reversal. Rare XY females with SRY deletions can arise by the same mechanism of aberrant Xp-Yp interchange that causes XX males (Fig. 58-2). In addition to streak gonads, these XYp– females almost invariably display extragonadal Turner’s syndrome features, most commonly lymphedema. By contrast, XY females with SRY point mutations or intragenic deletions have streak ovaries but no extragonadal manifestations of Turner’s syndrome. These data imply that there is at least one Y-linked Turner’s syndrome gene near SRY. David Page and colleagues have identified two candidate genes in this region, ZFY and RPS4Y. Both genes have X-linked homologs that escape X inactivation, as predicted for Turner’s syndrome genes. The ZFX and ZFY genes encode highly similar zinc finger proteins of unknown function that may be transcription factors. The RPS4X and RPS4Y genes encode isoforms of S4, a ubiquitous ribosomal small subunit protein. The S4X and S4Y proteins have been shown to function interchangeably in ribosomes, consistent with the hypothesis that RPS4Y serves to provide a second dose of S4 in males. However, the amount of S4Y in ribosomes is only about 10% that of S4X, and no biochemical consequences of S4 haploinsufficiency have been demonstrated. Thus the role of either RPS4X/RPS4Y or ZFX/ZFY in Turner’s syndrome is speculative. Interestingly, the mouse Zfx and Rps4 genes are both subject to X inactivation. Moreover, there do not appear to be functionally interchangeably mouse Y homologs. Even though the overall complement of X-linked genes in humans and mice is very similar, the XO mouse has good viability, is fertile, and is anatomically normal. Species differences in X inactivation and X-Y homology may explain the striking contrast in the phenotype of monosomy X in mice versus humans.

OTHER SEX CHROMOSOME DISORDERS One variant of Turner’s syndrome that has received much attention involves mosaicism for a ring X chromosome, or 45,X/ 46,X,r(X) karyotype. The associated phenotype varies dramatically. Some individuals have typical Turner’s syndrome, whereas others show more severe abnormalities such as mental retardation and multiple congenital anomalies. The severity of the phenotype correlates inversely with the size of the ring; larger rings tend to be associated with a milder phenotype. It is remarkable that a partial deletion, e.g., a ring, can be more deleterious (at least in liveborns) than the complete absence of one X chromosome. The explanation may lie with X inactivation and dosage compensation. Some small ring X chromosomes lack the XIC region; others retain XIC but fail to express the XIST gene (Fig. 58-6). The absence of XIST expression correlates with the severity of the phenotype. Barbara Migeon and coworkers have shown that

CHAPTER 58 / SEX CHROMOSOME DISORDERS

Figure 58-6

567

Two types of ring X chromosomes.

Figure 58-8 Deletion mapping a Yq azoospermia locus. (YRRM genes have been renamed RBM.) Figure 58-7

Aberrant Xq-Yq interchange.

rings that fail to express XIST express some genes that are normally X inactivated. Thus the severe phenotype associated with mosaicism for a small ring X chromosome is probably caused by overexpression of one or more X-linked genes, as yet unidentified. The importance of dosage compensation is also illustrated by another sex chromosome disorder. About one in every 1000 males lacks a portion of Yq, including the heterochromatic region. Men with the 46,XYq– karyotype are often normal, consistent with the absence of genes in this region. Deletions extending into the Yq euchromatic region are sometimes associated with short stature or infertility. In rare cases there are severe abnormalities, such as mental retardation, microcephaly, and other dysmorphic features. The severity of the phenotype does not generally correlate with the extent of the deletion. After the discovery of long arm pseudoautosomal region, David Page and colleagues speculated that aberrant Xq-Yq interchange might be one cause of Yq deletions, just as aberrant Xp-Yp recombination causes Yp deletions (Fig. 58-2). To test this hypothesis, they studied ten 46,XYq– males and found evidence of aberrant Xq-Yq recombination in three subjects. All three had severe phenotypes. The other seven subjects in whom there was no evidence of abnormal X-Y recombination showed only mild or moderate abnormalities. In the three severely affected subjects, portions of distal Xq as large as 10 Mb were translocated to the Y chromosome, resulting in partial X disomy (Fig. 58-7).

One X-linked gene within the translocated portion, G6PD (glucose-6-phosphate dehydrogenase), was shown to be expressed from both the normal X and the translocated portion, leading to a twofold increase in enzyme activity. Page et al. concluded that diploid dosage of one or more genes in distal Xq caused the severe phenotypes of the three severely affected subjects. They termed this disorder the “XYXq syndrome.” Molecular characterization of Yq deletions has also yielded insight into the function of the Y chromosome in male reproduction. As noted above, some men with cytogenetically visible Yq deletions have azoospermia (the absence of sperm in semen), suggesting that the euchromatic portion of Yq contains one or more genes important in spermatogenesis. This Yq locus is denoted AZF (azoospermia factor) (Fig. 58-8). Several groups have found using DNA assays that up to 10% of men with unexplained azoospermia bear de novo Yq microdeletions. Two candidates for AZF have been identified by positional cloning. The first is a family of Y-specific genes originally called YRRM (Y RNA recognition motifs), discovered by Ann Chandley, Howard Cooke, and colleagues. These genes, now designated RBM, are members of a larger family of RNA-binding proteins. There are at least three functional genes and numerous Ylinked pseudogenes. Most RBM sequences are clustered near the AZF region of distal Yq, but related sequences are also present near the centromere on both Yp and Yq. One of the functional genes is absent in most normal Japanese men and thus appears to

568

SECTION VII / ENDOCRINOLOGY

represent a normal polymorphism. The RBM genes are expressed specifically in germ cells, where they may play a role in RNA splicing or packaging during spermatogenesis. The presence of multiple RBM genes, at least some of which are polymorphic, makes it difficult to infer their importance from naturally occurring deletions. The second AZF candidate, termed DAZ (deleted in azoospermia), was discovered by Rene Reijo, David Page, and colleagues. Like the RBM genes, DAZ encodes an RNA recognition motif and shows testis-specific expression. Reijo et al. found that DAZ but not RBM was deleted in 12 of 89 azoospermic men with de novo Yq deletions. Interestingly, testicular biopsies from some of these men showed variable defects, even among adjacent seminiferous tubules in some instances. The defects ranged from the complete absence of germ cells (Sertoli-cell-only syndrome) to meiotic arrest with occasional mature condensed spermatids. DAZ is a strong candidate for AZF, the locus whose frequent deletion is associated with azoospermia, but this does not rule out a role for the RBM genes as well in spermatogenesis. Understanding the precise functions of these male-specific genes is the subject of active investigation.

SUMMARY The study of human sex chromosome disorders has been richly rewarding. Classic chromosome studies of Klinefelter’s syndrome, Turner’s syndrome, and mixed gonadal dysgenesis revealed the mechanism of mammalian testis determination. Genetic studies are now yielding molecular explanations for disorders of sex determination. Identification of genes responsible for other phenotypes associated with sex chromosome disorders, such as the ring X syndrome, the XYXq syndrome, and Turner’s syndrome, should be forthcoming. Molecular studies of these “experiments of nature” will also provide further insight into the unique biology of the mammalian sex chromosomes.

SELECTED REFERENCES Online Mendelian Inheritance in Man, OMIM™. 1996. Center for Medical Genetics, Johns Hopkins University, Baltimore, MD, and National Center for Biotechnology Information, National Library

of Medicine, Bethesda, MD. World Wide Web URL: http:// www3.ncbi.nlm.nih.gov/omim/ Affara NA, Lau YF, Briggs H, et al. Report and abstracts of the first international workshop on Y chromosome mapping 1994. Cytogenet Cell Genet 1994;67:359–402. Anonymous [editorial]. SOX9 and the switch hitting genes. Nat Genet 1995;9:1,2. Bardoni B, Zamaria E, Guioli S, et al. A dosage sensitive locus at chromosome Xp21 is involved in male to female sex reversal. Nat Genet 1994;7:497–501. Bogan JS, Page DC. Ovary? Testis?—a mammalian dilemma. Cell 1994;76:603–607. Cooke HJ, Elliott DJ. RNA-binding proteins and human male infertility. Trends Genet 1997;13:87–89. Foresta C, Ferlin A, Garolla A, Rossato M, Barbaux S, De Bortoli A. Ychromosome deletions in idiopathic severe testiculopathies. J Clin Endocrinol Metab 1997;82:1075–1080. Herzing LB, Romer JT, Horn JM, Ashworth A. Xist has properties of the X-chromosome inactivation centre. Nature 1997;386:272–275. Lahn BT, Ma N, Breg WR, Sratton R, Surti U, Page DC. Xq-Yq interchange resulting in supernormal X-linked gene expression in severely retarded males with 46,XYq– karyotype. Nat Genet 1994; 8:243–250. Migeon BR. X-chromosome inactivation: molecular mechanisms and genetic consequences. Trends Genet 1994;10:230–235. Pevny LH, Lovell-Badge R. Sox genes find their feet. Curr Opin Genet Dev 1997;7:338–344. Rao E, Weiss B, Fukami M, et al. Pseudoautosomal deletions encompassing a novel homeobox gene cause growth failure in idiopathic short stature and Turner syndrome. Nat Genet 1997;16:54–63. Reijo R, Lee TY, Salo P, et al. Diverse spermatogenic defects in humans caused by Y chromosome deletions encompassing a novel RNA-binding protein. Nat Genet 1995;10:383–393. Schwartz ID, Root AW. The Klinefelter syndrome of testicular dysgenesis. Endocrinol Metab Clin North Am 1991;20:153–163. Skuse DH, James RS, Bishop DV, et al. Evidence from Turner’s syndrome of an imprinted X-linked locus affecting cognitive function. Nature 1997;387:705–708. Willard HF, Cremers F, Mandel JL, Monaco AP, Nelson DL, Schlessinger D. Report and abstracts of the fifth international workshop on human X chromosome mapping 1994. Cytogenet Cell Genet 1994;67: 295–358. Zinn AR. Growing interest in Turner syndrome. Nat Genet 1997;16:3,4. Zinn AR, Page DC, Fisher EMC. Turner syndrome—the case of the missing sex chromosome. Trends Genet 1993;9:90–93.

CHAPTER 59 / DISORDERS OF PUBERTAL DEVELOPMENT

59

569

Disorders of Pubertal Development KAREN D. BRADSHAW AND CHARMIAN A. QUIGLEY

INTRODUCTION Puberty is a complex developmental process that culminates in sexual maturity. This transitional period begins in late childhood and is characterized by maturation of the hypothalamic–pituitary– gonadal axis, the appearance of secondary sexual characteristics, acceleration of growth, and ultimately the capacity for fertility. Significant endocrinologic changes accompany these developmental events. This chapter reviews the physiologic changes of normal puberty and examines the causes of precocious and delayed sexual maturation.

NORMAL PUBERTAL DEVELOPMENT The age at which the somatic changes of puberty begins is variable. In industrialized countries, pubertal changes usually begin between 8 and 13 years of age in girls, and between 9 and 14 years of age in boys. This variability in the time of onset is likely a reflection of the number of distinct influences that can affect the time at which it begins, both genetic and environmental. Approximately 5% of a given population will have the onset of puberty at an age outside of this range and will be considered to have either precocious or delayed puberty. In girls, the first somatic change that occurs is usually the beginning of breast development (thelarche), although in a minority of instances, pubic hair growth (pubarche) is the initial event. Thelarche and pubarche occur at mean ages of 10.9 and 11.2 years, respectively. Although the process of pubertal development is in fact a continuum, for descriptive purposes it is usually described in terms of a series of distinct stages, the five stages of breast and pubic hair development outlined by Marshall and Tanner being the most commonly employed scheme (Table 59-1). In parallel with the somatic changes of puberty, the volume of the ovaries and uterus enlarges, and the mucosa of the vagina thickens and becomes keratinized, evidenced clinically by lightening of the color of the mucosa from a deep red to a pale pink. After a variable period, averaging a little less than 2 years from the onset of breast development, most of the processes of pubertal maturation are complete, and menarche (the first menstrual period) occurs. Although most girls achieve menarche a little before their 13th birthday, the timing of this event is variable, and may occur as late as 141⁄2 years in normal girls.

From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

As in girls, puberty in boys is often described as a series of five distinct stages, based on testicular size, penile development, and distribution and character of pubic hair. In boys, the first physical evidence of puberty is an increase in testicular size. Measurement of the testes (length and width) or estimation of testicular volume using a Prader orchidiometer allows for early detection of pubertal onset. Testicular volume of 4.0 mL (or length of 2.5 cm), representing the onset of puberty, is noted at an average age of about 111⁄2 years. Testicular enlargement is followed by scrotal thinning, development of pubic hair, and penile enlargement. Adult testicular volumes and penile dimensions are generally achieved by about 16 years of age; however, there is quite marked individual variation, with young men completing their sexual maturation anywhere between the ages of 14 and 18 years. In addition to the appearance and development of secondary sexual characteristics in both boys and girls, puberty represents a period during which marked changes in body size and composition occur. Before the onset of puberty, the bodies of boys and girls are composed of similar proportions of adipose tissue and muscle. By the end of puberty, boys have a higher percentage of muscle and a lower percentage of fat relative to girls. It is during this period that the most rapid increases in bone mineralization occur.

PHYSIOLOGY OF PUBERTY HORMONES AND PUBERTY Puberty is a period during which many dramatic hormonal changes occur. Of these, it is clear that changes in the axes controlling the secretion of growth hormone and gonadal steroids play central roles in this process. Growth hormone (GH) is produced by the somatotrophs of the anterior pituitary gland. Although its control is complex, its synthesis and release is under the principal control of growth hormone releasing hormone (GHRH), which is released by the nerve endings of the hypothalamic GHRH neurons into the hypophyseal portal circulation. In response to pulses of GHRH, the pituitary releases pulses of GH into the systemic circulation. GH exerts its effects by binding to high-affinity receptors on the surfaces of responsive cells. Although GH certainly modulates some biologic processes directly, in many tissues the actions of GH are modulated indirectly through the action of growth factors that are produced in response to the action of GH, specifically insulin-like growth factor-1 (IGF-I, previously known as somatomedin C) and its complex series of binding proteins. Serum levels of IGF-I rise with age in concert with age-related increases in mean GH levels.

569

570

SECTION VII / ENDOCRINOLOGY

Table 59-1 Stages of Development of Secondary Sexual Characteristics Boys: genital (penis) development Stage 1 Prepubertal: testes, scrotum, and penis of about same size and proportion as in early childhood. Stage 2 Enlargement of scrotum and testes. Skin of scrotum reddens and changes in texture. Stage 3 Enlargement of penis, at first mainly in length. Further growth of testes and scrotum. Stage 4 Increased size of penis with growth in breadth and development of glans. Testes and scrotum larger; scrotal skin darkened. Stage 5 Genitalia adult in size and shape. Girls: breast development Stage 1 Prepubertal: elevation of papilla only. Stage 2 Breast bud stage: elevation of breast and papilla as small mound. Enlargement of areola diameter. Stage 3 Further enlargement and elevation of breast and areola, with no separation of their contours. Stage 4 Projection of areola and papilla to form a secondary mound above level of breast. Stage 5 Mature stage: projection of papilla only, related to recession of areola to general contour of breast. Both sexes: pubic hair Stage 1 Prepubertal: vellus over pubes is not further developed than over abdominal wall. Stage 2 Sparse growth of long, slightly pigmented, downy hair, straight or slightly curled, chiefly at base of penis or along labia. Stage 3 Considerably darker, coarser, and more curled hair. Hair spreads sparsely over junction of pubes. Stage 4 Hair now adult in type, but area covered is still considerably smaller than in adult. No spread to medial surface of thighs. Stage 5 Adult in quantity and type with distribution of horizontal (or classically “feminine”) pattern. Stage 6 Spread up linea alba (male-type pattern).

At the onset of puberty, increased activity of the gonadotropin releasing hormone (GnRH) pulse generator causes a progressive rise in mean concentrations of gonadotropins, resulting from an increase in the frequency and amplitude of GnRH pulses (see below). These increases are first detected as nocturnal gonadotropin pulses, but as puberty progresses gonadotropin pulses also increase during daytime, until adult mean gonadotropin levels are achieved. More recently, a role for leptin, the recently described hormone produced by adipose tissue, in the onset of puberty has been described. In leptin-deficient ob/ob mice, treatment with leptin restores puberty and fertility. Subsequent studies have demonstrated that treatment of normal prepubertal mice with leptin accelerates maturation of the reproductive tract and results in earlier reproduction. Similarly, in prepubertal boys, leptin levels increase several months before the onset of puberty, as determined by the initial rise in testosterone. Concentrations of a variety of other hormones also change with onset and progression of puberty. As a result of increased sex steroid concentrations, sex hormone-binding globulin is lower during puberty than in childhood. The gonadal peptide inhibin, structurally related to transforming growth factor-β (TGF-β), is regulated by and involved in regulation of follicle-stimulating hormone (FSH). This product of Sertoli and granulosa cells shows a progressive increase in mean concentration with advancing puberty in both sexes. Concentrations of the glycoprotein antiMüllerian hormone (AMH) show quite marked sexual dimorphism. AMH is produced in Sertoli cells and is relatively high in newborn boys, but undetectable in girls. In contrast, the hormone becomes very low in boys during puberty, at which time AMH concentrations increase in girls. The role of estrogens in the process of skeletal maturation in both boys and girls had been postulated for some time. As an example, effective suppression of the rapid skeletal maturation seen in boys with gonadotropin-independent forms of precocious puberty requires inhibition of aromatase activity to reduce serum estradiol concentrations, in addition to antiandrogens to interfere with androgen action. Furthermore, a female patient with a genetic

deficiency of aromatase activity had no pubertal growth spurt and exhibited delayed skeletal maturation, indicating that estrogens are required for these events. Inferences drawn from such cases regarding the importance of estrogen in promoting skeletal maturation in both sexes are reinforced by contrasting the patterns of pubertal development in syndromes of androgen and estrogen resistance. Studies of patients with complete testicular feminization (complete androgen insensitivity) have documented that a normal pubertal growth spurt is observed with the onset of pubertal gonadal function. Such findings suggest that the effects of gonadal steroids in male pubertal development are not mediated via the androgen receptor, but are instead exerted indirectly after the conversion of testosterone to estrogen. The study of a male patient with estrogen resistance, in whom estrogens are unable to exert their effects at the tissue level, confirms these deductions. In this patient with estrogen resistance, an inactivating mutation of the estrogen receptor was shown to have several discernible physiologic consequences. First, the epiphyseal growth plates of this man demonstrated no evidence of epiphyseal fusion. Consequently, at the time of diagnosis (age 28) the patient was still growing. In addition to these effects on bone maturation, skeletal mineralization was also abnormal. At age 28, the patient’s mineral bone density was markedly decreased, even when corrected for his retarded bone age of 15 years. These considerations emphasize the complex hormonal interactions that characterize the process of normal puberty. It is clear that the normal functioning of each of these components is necessary for normal pubertal development to occur. As described below, abnormalities at many levels can disrupt normal pubertal growth, development and maturation. MATURATION OF THE HYPOTHALAMIC–PITUITARY– GONADAL AXIS Maturation of the reproductive system occurs in a phasic manner in humans and in higher primates, and can be viewed as occurring in several distinct stages. The first stage, which begins during fetal life and lasts until late infancy, is characterized by development of the neuroendocrine systems responsible for regulation of the reproductive system. The area in the arcuate nucleus of the hypothalamus destined to

CHAPTER 59 / DISORDERS OF PUBERTAL DEVELOPMENT

become the GnRH pulse generator develops intrinsic and unregulated pulsatile activity by about 11 weeks of gestation. During this stage the reproductive system appears to be fully active, with gonadotropin and sex steroid hormone concentrations being measurable in fetal plasma. Concentrations of luteinizing hormone (LH) and FSH peak at about 4–5 months’ gestational age. Later in gestation, negative feedback from gonadal steroids begins to regulate the unrestrained pulsatile activity of the hypothalamic pulse generator and by term, gonadotropin levels (and by inference the activity of the GnRH pulse generator) are low. At the time of delivery the infant is separated from the dominant source of sex steroids, the placenta, and owing to the withdrawal of this negative feedback, the levels of gonadotropins rise. This increase is responsible for a transient secondary stimulation of the infant’s gonads occurring in the first months after birth. Although occurring in both boys and girls, this is observed most readily in female infants in whom there may be prolonged neonatal breast budding. By 6 months of postnatal age, gonadotropin and sex steroids concentrations in plasma have again declined to low levels and the third stage of maturation begins. This stage lasts throughout childhood, and is characterized by low plasma concentrations of LH, FSH, and sex steroids. From a physiologic perspective, the prepubertal stage of development presents apparent contradictions. Measurements of gonadotropins and sex steroids during fetal development suggest that the hypothalamic–gonadal axis has completely developed in utero and that it is regulated by steroid hormones during the latter stages of pregnancy. Despite this, during the prepubertal period, gonadotropins remain low even when sex steroid concentrations are extremely low, such as in patients with Turner’s syndrome or in castrate children. That serum gonadotropin concentrations remain low under such conditions suggests that additional inhibitory mechanisms in the central nervous system or hypothalamus have developed. Early studies to explain these different regulatory behaviors focused on examining the sensitivity of the hypothalamus and pituitary to feedback inhibition by gonadal steroids. Such investigations demonstrated that the levels of estrogen and androgen required to inhibit LH and FSH secretion in young prepubertal animals and in humans are consistently lower than those required to suppress gonadotropin levels to an equivalent extent in adult castrate animals. Such differences in the sensitivity of regulation of gonadotropin secretion in young prepubertal and adult animals have been described in a number of different species. Although such observations have substantial power in explaining the prepubertal quiescence of gonadotropin secretion, other observations suggest that additional mechanisms might also be operative. Although mean serum concentrations of gonadotropins are low during the prepubertal period, the reproductive system is not completely inhibited during this stage, because small spontaneous LH pulses occur at a low frequency in normal children. The fourth stage, puberty itself, occurs as the result of reactivation of the reproductive axis. Although a great deal of effort has been expended to identify the signals that control the onset of puberty, it appears that the mechanisms responsible for the initiation of pubertal events are extremely complex and likely involve the integration of numerous different signals, including attainment of a certain body mass or composition and perhaps neural signals derived from centers within the central nervous system that serve as a biologic clock.

571

The onset of puberty is heralded by striking increases in nocturnal LH secretion, manifest by an increase in amplitude and frequency of LH pulses. These increases of LH precede rises of sex steroid concentrations and the development of secondary sex characteristics. As pubertal maturation progresses, the amplitude and frequency of gonadotropin pulses also increase during the day, in a pattern similar to that seen at night, until the final stage of sexual maturation, adulthood, is reached. In this period, regular pulses of GnRH establish the mature pattern of gonadal steroid secretion. In females, this results in the regular cyclical variations of gonadotropins, estrogen, and progesterone characteristic of the menstrual cycle. In males, the same regular pulses of GnRH establish a pattern characterized by relatively constant levels of testosterone and gonadotropins, with minimal diurnal variation.

VARIATIONS OF NORMAL PUBERTY EARLY Fairly common forms of partial premature pubertal development are the isolated development of pubic hair (premature pubarche) and the isolated development of breasts (premature thelarche). Although these are benign conditions, such patients must be followed closely to monitor for progression to constitutional precocious puberty (CPP). Precocious pubarche is most often a benign condition secondary to early adrenarche. Balducci et al. studied 171 subjects with isolated precocious pubarche. Mild abnormalities of steroidogenesis (nonclassic forms of 21-hydroxylase or 3-β hydroxysteroid deficiency) were present in 12% of patients as diagnosed by adrenocorticotropic hormone (ACTH) stimulation tests. In the majority of those, basal 17α-hydroxyprogesterone (17α-OHP) levels were higher than in pubertal norms and were thought to be a good screening test to determine which patients should undergo ACTH stimulation testing. In some patients premature pubarche may predict the development of functional hyperandrogenism in the mid-teenage years associated with polycystic ovarian syndrome in girls. Premature thelarche is typically associated with a degree of FSH secretion, antral follicular development, and ovarian function that is greater than that measured in age-matched prepubertal control subjects. The prevalence of ovarian microcysts detected by ultrasonography is increased in these girls; however, plasma estradiol is commonly unmeasurable despite genitourinary cytology that shows evidence of estrogenization. Premature thelarche usually occurs in the first 2 years of life and regresses before puberty. Children who present with breast development later are more likely to have some degree of continued breast development, representing an early stage of precocious puberty. Once breast development is stimulated much beyond the breast-bud stage and reaches early adolescent proportions, breast contour generally does not regress. Occasionally (approximately 10–15%) girls go on to develop complete precocious puberty, but in the majority of patients the breast bud is a transient event that warrants only close follow-up for the appearance of other pubertal signs.

PRECOCIOUS PUBERTAL DEVELOPMENT The development of isosexual secondary sexual characteristics before the age of 8 years in girls and before the age of 9 years in boys is termed precocious puberty. Precocious puberty is characterized by early and progressive sexual development accompanied by advancement of skeletal maturation as measured by bone age. The rapid linear growth that characterizes precocious puberty is

572

SECTION VII / ENDOCRINOLOGY

Table 59-2 Causes of Gonadotropin-Dependent (Central) Precocious Pubertya Idiopathic precocious puberty CNS tumors Craniopharyngioma Hypothalamic hamartoma Optic glioma, astrocytoma, and others Other CNS disorders Static encephalopathy (secondary to infection, hypoxia, trauma, etc.) Low-dose cranial radiation Hydrocephalus Arachnoid cyst Septo-optic dysplasia Secondary central precocious puberty After late treatment of congenital adrenal hyperplasia Hypothyroidism aAn overview of the causes of gonadotropin-dependent precocious puberty.

associated with premature and rapid skeletal maturation and fusion of the epiphyses, in most cases resulting in short adult stature compared with genetic height potential. There are two major classes of precocious puberty: disorders that result from early reactivation of the hypothalamic–pituitary– gonadal axis (generally referred to as gonadotropin-dependent or central precocious puberty, Table 59-2) and those that do not (referred to as gonadotropin-independent precocious puberty, Table 59-3). Most girls and approximately half of boys who present with precocious puberty have central, gonadotropin-dependent precocious puberty which results from the secretion of GnRH from the hypothalamus. Although central precocious puberty is much more common among females than males, boys with this form of precocious puberty are more likely to be found to have an underlying central nervous system (CNS) abnormality. By contrast, the majority of girls have no discernible structural CNS lesion and are thus said to have an “idiopathic” form of the disorder. The overall incidence of gonadotropin-dependent precocious puberty has increased in both sexes because of the survival of children who have received CNS irradiation for brain tumors or leukemia. GENERAL EVALUATION Evaluation of patients with early pubertal development should begin with a detailed history focused on the occurrence of prior CNS trauma, radiation, or seizure activity, exposure to exogenous sex steroids in cosmetics or food, or a positive family history. In girls, physical examination should focus on determining whether the development reflects androgen action, estrogen action, or both. Girls with isolated androgenization most likely have premature adrenarche or congenital adrenal hyperplasia, whereas those with estrogenization or evidence of both androgen and estrogen action are more likely to have precocious puberty. In boys, the presence of isolated androgenization most likely represents premature adrenarche, whereas the finding of testicular volumes of more than 4.0 mL is diagnostic of precocious puberty. Diagnostic evaluation should begin with an X-ray to assess bone age as a marker for sex steroid hormone action. In general establishing that skeletal age is concordant with chronological age allows for continued observation. When secondary sexual characteristics are associated with an advanced bone age, measurements of estradiol, testosterone, and thyroid hormone should be obtained, and a GnRH stimulation test

Table 59-3 Gonadotropin-Independent Precocious Pubertya Boys Testicular disorders Familial male precocious puberty (testotoxicosis) McCune-Albright syndrome Leydig-cell adenomas Human chorionic gonadotropin-secreting tumors Androgen-secreting teratomas Girls Ovarian disorders McCune-Albright syndrome Granulosa or theca-cell tumors Simple follicular cyst Other estrogen-secreting tumors (teratomas, dysgerminomas) aAn overview of the causes of gonadotropin-independent precocious puberty. This term is used interchangeably with the term peripheral precocious puberty.

is indicated to differentiate between central (gonadotropin-dependent) and peripheral (gonadotropin-independent) causes of precocious puberty. In most cases, the diagnosis of gonadotropindependent precocious puberty, particularly in younger children, warrants obtaining a magnetic resonance imaging (MRI) or computerized tomography head scan.

GONADOTROPIN-DEPENDENT PRECOCIOUS PUBERTY IDIOPATHIC PRECOCIOUS PUBERTY In girls, most cases of precocious puberty are central (gonadotropin-dependent) in origin, and are believed to be caused by premature maturation of the hypothalamic–pituitary–ovarian axis. In perhaps two-thirds of cases, no recognizable cause of the disorder can be found. In these cases, the pattern of development and progression parallels that of normal pubertal development, although the onset is at an early age. CNS DISORDERS Lesions of the CNS are well-recognized as causing central precocious puberty. Common causes include static encephalopathy as a result of infection, hypoxia, trauma, or irradiation during infancy or early childhood. A less common, but nonetheless important, cause of central precocious puberty is a CNS tumor. These tumors can be viewed as causing precocious puberty by one of two distinct mechanisms. Hypothalamic hamartomas are benign tumors that have been shown to contain measurable GnRH. As such, they may be considered to be acting as ectopic GnRH pulse generators that have escaped from the normal inhibitory influences exerted in the prepubertal period on the centers that normally secrete GnRH. These small tumors are more frequently diagnosed in boys than in girls and are most easily visualized using MRI, as some may be only 2 or 3 mm in size. These tumors tend to grow slowly—if at all—and rarely cause neurologic symptoms. The chance of finding CNS pathology in either sex is inversely proportional to the age of the child, with the greatest yield in children younger than 4 years old. Kappy found that in girls whose pubertal development began after 6 years of age, any CNS pathology was already known or clinically evident, suggesting that routine MRI in these children will less likely have positive findings. In contrast, Pescovitz and coworkers reported that in a series of 4000 children referred to the National Institutes of Health, about one-third of the girls and more than 90% of the boys had an identifiable lesion of the

CHAPTER 59 / DISORDERS OF PUBERTAL DEVELOPMENT

CNS visible on computed tomography or MRI scans. This high prevalence of CNS lesions reflects the referral population. Cranial irradiation has dose-dependent effects on many hypothalamic–pituitary functions. Although doses of cranial irradiation exceeding 50 g to the hypothalamic–pituitary axis may render a child gonadotropin deficient, lesser doses of irradiation have been associated with early puberty in both sexes. Low-dose cranial irradiation (18–24 g) used in the CNS prophylactic treatment of acute lymphoblastic leukemia is associated with a downward shift in the distribution of ages at pubertal onset and menarche in girls. Central precocious puberty is rare in boys treated in this manner. The pathophysiology of central precocious puberty in nonhamartomatous lesions of the CNS is not yet established. It is possible that neural defects located near the hypothalamus cause precocious puberty by interfering with tonic central nervous system inhibition of the hypothalamic–pituitary–gonadal axis. It is also possible that focal derangements of the cellular environment in the vicinity of GnRH neurons may be causally related to the premature activation of GnRH secretion. Junier and Ojeda have speculated that neurotrophic or mitogenic activities produced locally in response to brain injury may be involved in the process. They suggest that the response of GnRH neurons to hypothalamic injury comprises three phases: an initial stage during which some of the neurons lose mature morphologic characteristics, reverting to a presumably more immature condition; an intermediate phase in which morphologic differentiation is reestablished and an increased synthesis of TGF-β by reactive astrocytes enhances GnRH release without affecting GnRH gene expression through a process that may involve glial release of prostaglandins; and a third phase in which the rise in sex steroids caused by the GnRHdependent increase in basal LH release activates GnRH neurons to secrete gonadotropins in an episodic nature. TREATMENT OF GONADOTROPIN DEPENDENT PRECOCIOUS PUBERTY Historically, medroxyprogesterone and cyproterone acetate were used to attempt to suppress activation of the hypothalamic–pituitary–gonadal axis. Neither of these agents was satisfactory, because they were not fully effective in inhibiting pubertal or skeletal maturation or in improving adult height. The recognition that the continuous, nonpulsatile presentation of GnRH to the pituitary gonadotrophs induced a state of secondary hypogonadism led to the development and use of potent GnRH agonists in the therapy of precocious puberty. These gonadotropin-releasing hormone analogs (GnRHa; see also Chapter 34) represent the first truly effective treatment for central precocious puberty. Significant reductions in basal and peak (GnRH-stimulated) serum FSH and LH concentrations occur within the first month of GnRHa therapy. In parallel with these changes, a reduction in plasma concentrations of estradiol (in girls) and testosterone (in boys) occurs after the first month and persists while the drug is administered. Importantly, the effects of these drugs are reversible. After withdrawal of GnRHa therapy, gonadotropin and gonadal steroid concentrations return to their pretreatment levels. Institution of GnRHa therapy results in a decrease in growth velocity, usually within a range that is appropriate for the child’s skeletal maturation. The slowing of linear growth is accompanied in most cases by slowing of skeletal maturation—one of the primary aims of such treatment. Preservation of, or an increase in, adult height can be achieved in some children with gonadotropindependent precocious puberty treated with GnRHa. Most investigators, using the tables of Bayley and Pinneau, reported increases

573

in the predicted final height of patients during the course of GnRHa therapy with a mean increase of 5 cm. Most studies suggest that the greatest improvement in predicted adult height is obtained in children whose bone ages are relatively young at the onset of treatment, indicating the need for early diagnosis and intervention. The length of time that such therapy is continued depends on bone age and estimates of final height in individual cases. Although concern has been raised regarding the possible effect of pubertal suppression on skeletal mineralization, an important feature of normal puberty, evidence available to date indicates no significant problem. Baens et al. evaluated the bone mineral density (BMD) in female patients treated with GnRHa for gonadotropin-dependent precocious puberty who had completed therapy and had subsequently attained a bone age of greater than 14 years. They found that BMDs were not different from those of a control population of girls with the same bone ages. The physical effects of pubertal suppression are not limited to the effects on bone development. The majority of girls experience no increase in breast development, and a third show regression to an earlier Tanner stage, correlating with a reduction in ovarian and uterine size. Some girls experience transient vaginal bleeding approximately 2–4 weeks after initiation of GnRHa therapy because of estrogen withdrawal. Effects on pubic hair are less predictable, although most children show either no progression or a minor degree of regression. Some children show an increase in pubic hair that correlates with normal adrenarche.

GONADOTROPIN-INDEPENDENT PRECOCIOUS PUBERTY Gonadotropin-independent forms of precocious puberty are about one-fifth as common as gonadotropin-dependent forms of precocious puberty. Gonadotropin-independent precocious puberty is characterized by increased production of gonadal steroids, causing the typical physical changes of puberty, in the absence of reactivation of the hypothalamic–pituitary axis (Fig. 59-1). This form of precocious puberty includes conditions that mimic the effect of pituitary gonadotropins on gonadal function, such as those in which there is secretion of gonadotropins from nonpituitary sources. Table 59-3 lists conditions associated with gonadotropinindependent precocious puberty. Molecular mechanisms underlying two forms of gonadotropin-independent precocious puberty, familial Leydig-cell hyperplasia and McCune-Albright syndrome, have recently been described. McCUNE-ALBRIGHT SYNDROME McCune-Albright syndrome (MAS) is characterized classically by the clinical triad of cutaneous hyperpigmentation, polyostotic fibrous dysplasia, and endocrine dysfunction. Although it is stated that at least two of the features of the triad must be present to make the diagnosis, this guideline should be interpreted with caution, since the features may develop over time. The condition occurs distinctly less frequently in boys than in girls, who comprise 90% of affected patients. In contrast to those with central precocious puberty, these girls not uncommonly present with vaginal bleeding as the first sign of their sexual development. A pattern of variable involvement of hormone-secreting cells occurs in subjects with MAS, and the endocrine abnormalities are characterized by excessive function of these cells. The most common endocrine manifestation of MAS is precocious puberty, often associated with ovarian cysts. A waxing and waning course of the precocious puberty is not uncommon. Depending on the specific cell types affected, a variety of

574

SECTION VII / ENDOCRINOLOGY

Figure 59-1 A schematic representation of the hypothalamic–pituitary–gonadal axis indicating the levels disturbed in selected forms of precocious puberty. Testitoxicosis (familial male precocious puberty, FMPP) is caused by amino acid substitution mutations in the membrane receptor for luteinizing hormone (LH) that result in its constitutive, ligand-independent activation. The gonadotropin-independent gonadal function that is observed in individuals with McCuneAlbright syndrome is caused by mutations in a component of the FSHregulated signaling cascade in granulosa cells of the ovary. In these individuals, replacement of one of two critical amino acid residues in a G protein subunit (Gsα) results in the constitutive activation of adenylate cyclase, mimicking the effect of FSH receptor activation in stimulating steroidogenesis in granulosa cells.

other endocrine disturbances may occur, including hyperthyroidism (caused by nodular or follicular thyroid hyperplasia) in 20–40% of patients, hypercortisolism (caused by nodular adrenal hyperplasia), and growth hormone- or prolactin-secreting pituitary adenomas. The peculiar bone lesion of MAS (polyostotic fibrous dysplasia) may occur in any bone, including long bones, the skull, and pelvis, and result in pathologic fracture or severe disfigurement. Malignant degeneration occurs in up to 4% of lesions. The usually large, pigmented cutaneous lesions (café-au-lait patches) have irregular, “serrated” outlines. Their distribution, like the variable involvement of endocrine tissues, reflects an underlying genetic mosaicism that results from a postzygotic somatic cell mutation occurring after a number of cell divisions have taken place. Common sites of occurrence of the café-au-lait patches include the forehead, the neck or upper back, the shoulder and upper arm, the lumbosacral region, and the buttocks. They often follow a distribution that is not dermatomal, but in fact relates to the lines of Blaschko, which are thought to represent the patterns of dorsal and ventral outgrowth of two different cell populations during early embryogenesis. The hyperpigmentation is present in the basal

epidermis; the number and size of melanocytes is normal, but the melanosomes are enlarged. The skeletal and cutaneous involvement in MAS is commonly asymmetrical and the skin lesions often stop abruptly at the midline. The diverse clinical abnormalities of MAS have in common the involvement of cells that respond to extracellular signals through activation of the hormone-sensitive adenylate cyclase system, the membrane-bound enzyme that catalyzes the formation of the intracellular second messenger cyclic adenosine monophosphate (cAMP). However, these endocrine disturbances are not accompanied by increased plasma concentrations of the relevant trophic or stimulatory hormones. Thus, girls with precocious puberty caused by MAS have ovarian enlargement and follicular hyperplasia but have low serum levels of LH and FSH and a prepubertal response of LH to administration of GnRH. The basis of the gonadotropin-independent nature of the precocious sexual development in MAS was elucidated by the identification of mutations that activate the intracellular second messager pathways by which the trophic hormones such as gonadotropins and thyrotropin signal. The activity of the hormone-sensitive adenylate cyclase system is primarily regulated by two guanine-nucleotide binding proteins (G proteins)—one inhibitory and the other stimulatory (see Chapter 46). Schwindinger et al. identified a mutation in the gene encoding the α-subunit of the stimulatory G protein that regulates adenylate cyclase activity. A heterozygous guanine to adenine transition was found in exon 8 of the gene encoding the α-subunit of Gs, predicting the replacement of arginine by histidine at position 201 of the mature protein. This amino acid is located within a critical region of the Gsα protein and the substitution causes a marked decrease in the intrinsic GTPase activity of Gsα, prolonging the survival of the active conformation of the enzyme and resulting in constitutive activation of adenylate cyclase activity. The consequent increased production of cAMP explains the increases in endocrine organ function typical of this disease, since this system mediates the stimulatory effects of many hormones (gonadotropins, thyrotropin, adrenocorticotropin, and GH-releasing hormone). In addition, cAMP modulates cellular proliferation in certain tissues, providing a basis for the hyperplasia seen in some tissues in MAS. Subsequent work by Weinstein and coworkers confirmed and extended these findings. Using PCR and allele-specific hybridization, these investigators identified two types of mutation at the same position implicated by Schwindinger (the codon encoding amino acid residue 201), resulting in substitution of either a histidine or a cysteine residue for the native arginine at this position. In the pathologic specimens from each of four patients with MAS studied by Weinstein, the mutant allele was found to be present in greatest abundance, relative to the normal allele, in the tissues histologically most affected, but in low levels in most of the tissues examined. The mutation has been found within the dysplastic bone lesions of affected patients, and may produce these lesions by inducing proliferation of mesenchymal progenitor cells. The mutation has also been detected in the hyperpigmented skin lesions, which commonly colocalize with the bone lesions, but its role in their pathogenesis remains unclear. Inactivating mutations of Gsα are associated with pseudohypoparathyroidism. The findings reported above are consistent with the hypothesis that a spontaneous mutation early in gestation produced an abnormal monoclonal cell population and that the variable clinical presentation of MAS is caused by the varying degrees of somatic

CHAPTER 59 / DISORDERS OF PUBERTAL DEVELOPMENT

mosaicism of this genetic alteration that would otherwise represent an autosomal dominant lethal defect. Because of this mosaicism, MAS is not inherited. FAMILIAL MALE PRECOCIOUS PUBERTY Although the endocrine disturbances are confined to the testes, a similar mechanism characterizes another GnRH-independent form of precocious puberty, familial male precocious puberty (FMPP, familial testotoxicosis). This disorder is inherited in an autosomal dominant, male-limited pattern and is characterized by the onset of puberty (testicular enlargement) before 4 years of age. Testosterone production and Leydig cell hyperplasia occur autonomously, in the context of low, prepubertal levels of LH. The clinical hallmark of this disorder is the relatively modest enlargement of the testes for the degree of virilization. Affected adult men are often short because of early epiphyseal fusion; however, they are otherwise healthy and have normal fertility. Although LH concentrations are low in childhood, they are normal in adulthood, indicating that normal maturation of the hypothalamic–pituitary–gonadal axis occurs. The disorder is not expressed in heterozygous females, since estrogen production by ovarian follicles requires coordinated stimulation by both LH and FSH; thus, activation of the LH pathway alone has no significant effect. The pathogenesis of this gonadotropin-independent form of precocious puberty has been traced to heterozygous mutations within specific segments of the gene encoding the receptor for luteinizing hormone and chorionic gonadotropin (the LH/CG receptor). The LH/CG receptor is a member of the seven transmembrane domain family of G-protein coupled receptors. Unlike the inactivating mutations of the LH receptor, which can occur in many parts of the molecule, most of the activating mutations of the LH receptor have resulted in amino acid substitutions within the fifth and sixth transmembrane helices and the cytoplasmic loop that connects them. Shenker et al. found a mutation that resulted in replacement of the native aspartate with glycine, at position 578, in the sixth transmembrane helix of the protein. In vitro studies of the mutant receptor containing the Asp 578 to Gly substitution demonstrated marked increase in cyclic AMP production in the absence of ligand. It is presumed that the structural changes induced in the LH receptor by these amino acid substitutions shift these portions of the receptor to a conformation that mimics the conformation of the normal LH receptor activated by LH. This conformation of the mutant LH receptor leads to the agonist-independent production of cAMP, and thereby to autonomous testosterone production by the Leydig cells of affected boys. The presence of the Asp 578 to Gly mutation in most affected kindreds studied to date is believed likely to reflect a common genetic ancestry in many cases. However the finding of this same mutation in a number of ethnically diverse individuals suggests the occurrence of fresh mutations in some cases. Other less common mutations include replacement of alanine 572 by valine, threonine 577 by isoleucine, aspartate 582 by glycine, and methionine 575 by isoleucine. This latter amino acid is located in the sixth transmembrane domain of the receptor, close to a region important for binding of G proteins. Recently, a mutation that converts methionine 398 to threonine has been reported. In contrast with the majority of mutations in this disorder, this amino acid is located in the second transmembrane domain of the receptor. Notably, there was phenotypic heterogeneity associated with this mutation, since one boy carrying the mutant allele had no evidence of precocious puberty.

575

ECTOPIC HORMONE PRODUCTION Human chorionic gonadotropin (hCG)-secreting germ cell tumors have been thought to cause precocious puberty exclusively in boys. Because of the extensive structural homology of the β-subunits of hCG and LH, the excess hCG acting through the LH receptor stimulates Leydig cell production of testosterone. Non-CNS tumors causing gonadotropin-independent precocious puberty in females have also been described. In such instances, the pubertal development is most often traced to the presence of a granulosa cell ovarian tumor or an adrenal tumor that produces excess estrogen. Although the cell types and hormones differ in the different tumors described above, they do share a common characteristic. That is, in each instance a hormone that is normally produced by specific cell types and regulated in a very precise fashion is expressed by the neoplastic cells in an inappropriate or unregulated fashion. Although not well-studied, it appears that the genetic events leading to the genesis of the neoplasms in some way mimic the signals controlling hormonogenesis. Although not strictly a cause of precocious puberty, congenital adrenal hyperplasia, most commonly because of deficiency of 21-hydroxylase, is another cause of early virilization. It is distinguished by absence of testicular enlargement in boys (hence the designation “pseudopuberty”), and virilization without feminization in girls. TREATMENT OF GONADOTROPIN-INDEPENDENT FORMS OF PRECOCIOUS PUBERTY Because patients with these forms of precocious puberty demonstrate pubertal development that is independent of pituitary gonadotropin secretion, GnRH agonist therapy is ineffective, at least as part of initial management. As such, even though the molecular lesions are different, treatments for FMPP and MAS are instead focused on directly inhibiting the synthesis or action of sex steroids. In boys with FMPP, this has been approached through administration of inhibitors of steroidogensis, such as the antifungal agent ketoconazole. This compound inhibits the enzyme that catalyzes the cleavage of the 20,22 bond of the cholesterol molecule, and thus inhibits the synthesis of both androgens and estrogens. An alternative approach has been to employ drugs that block the actions of androgen at the level of the androgen receptor itself. Although more potent drugs are now available, published reports have used spironolactone, a drug that has demonstrated modest activity as a competitive inhibitor of androgen receptor function. Results using this drug have been somewhat disappointing, but have been more encouraging when coupled with testolactone. This compound is an aromatase inhibitor that blocks the enzyme (P450 aromatase) that catalyzes aromatization of testosterone, the ratelimiting step in estrogen biosynthesis, thereby reducing the high estrogen concentrations responsible for the accelerated skeletal maturation typical of this condition. Of note, the early maturation of the hypothalamus induced by exposure to high levels of sex steroids may result in the activation of the endogenous GnRH pulse generator, leading to the development of central, gonadotropin-dependent precocious puberty, secondary to the preexisting gonadotropin-independent precocious puberty. In such instances, the addition of GnRHa therapy is required to prevent progression of pubertal development and skeletal maturation. Treatment of girls with precocious puberty because of MAS uses drugs that inhibit the synthesis of estrogen (aromatase inhibitors). The longest-term studies of this type have used testolactone.

576

SECTION VII / ENDOCRINOLOGY

Although demonstrating some efficacy, the benefits were minimal and were likely caused by imperfect patient compliance (because of side effects of the drug) and the relatively low potency of testolactone itself in inhibiting aromatase activity. The availability of more potent aromatase inhibitors (e.g., arimidex) may improve the effectiveness of such approaches in the future. The approach to inhibition of testosterone synthesis in boys with precocious puberty caused by MAS is similar to that employed in the treatment of patients with FMPP. Addition of a GnRH analog is required for those patients with MAS who develop secondary central precocious puberty.

CONTRASEXUAL DEVELOPMENT Contrasexual (or heterosexual) development refers to feminization in males or virilization in females, and can occur at any age, either prepubertally or postpubertally. These disorders are not forms of puberty. Evaluation consists of careful history and physical examination followed by measurement of gonadal and adrenal steroids, and in more subtle cases, ACTH stimulation testing to identify cryptic or atypical forms of congenital adrenal hyperplasia. Depending on the pattern of steroid hormone excess detected, imaging of the adrenals, gonads, or liver may be required. Gynecomastia is the most common manifestation of feminization and in pubertal boys is usually a benign self-limited condition; however, gynecomastia in a prepubertal boy is clearly pathologic. When associated with hypogonadism, Klinefelter’s syndrome should be considered and a karyotype obtained. In most cases, estrogen concentrations are not increased. However, in those patients in whom there is measurably high estrogen concentration, the source of excess estrogen may be a neoplasm (e.g., adrenal, hepatic, or testicular), an abnormality of steroid metabolism, or increased extraglandular conversion of C-19 steroids to C-18 estrogens, such as occurs in the presence of significant obesity (gynecomastia is more common in obese boys). Exposure to exogenous estrogens is another uncommon cause of feminization in males. Females with virilization may have adrenal or gonadal sources of androgen excess, causing development of sexual hair, clitoromegaly, acne, and hirsutism. The most common cause of virilization in prepubertal girls is congenital adrenal hyperplasia (CAH) because of deficiency of 21-hydroxylase activity (“late-onset” or “attenuated” CAH). Other virilizing forms of CAH include 3β-hydroxysteroid dehydrogenase and 11-hydroxylase deficiency. Severely advanced skeletal maturation is a common accompaniment of the virilization in affected girls, and these conditions can be further complicated by the development of secondary central precocious puberty. A variety of molecular defects underlie these enzyme deficiencies, as described in Chapter 57. Virilizing adrenal adenoma or ovarian neoplasms are associated with elevated levels of dehydroepiandrosterone sulfate, androstenedione, or testosterone. β-hCG and α-fetoprotein may also be increased and should be measured as markers for ovarian, testicular, and hepatic tumors in virilized girls and feminized boys.

DELAYED PUBERTAL DEVELOPMENT Delayed puberty is defined by lack of physical changes of sexual maturation at a chronological age that is two standard deviations above the mean age of onset of puberty. In the United States, the ages of 12 years in girls and 14 years in boys serve as practical guidelines to determine the need for evaluation. Diagnostic evaluation should differentiate between constitutional

Table 59-4 Causes of Delayed Pubertal Development Constitutional delay Hypogonadotropic hypogonadism Isolated gonadotropin deficiency Kallmann’s syndrome and variants Functional gonadotropin deficiency Chronic systemic disease and malnutrition Anorexia nervosa Exercise-induced amenorrhea CNS disorders Tumors Radiation therapy Anatomic defects (e.g., septo-optic dysplasia) Miscellaneous disorders CHARGE association Noonan’s syndrome Aarskog syndrome Prader-Willi syndrome Laurence-Moon-Biedl or Bardet-Biedl syndromes Hypergonadotropic hypogonadism Klinefelter’s syndrome Primary testicular failure Gonadal dysgenesis and variants Primary ovarian failure

delay, hypogonadotropic hypogonadism, and hypergonadotropic hypogonadism. Causes of delayed puberty are listed in Table 59-4. VARIATIONS OF NORMAL PUBERTY—DELAYED (CONSTITUTIONAL DELAY) Constitutional delay of puberty is a common, benign condition that represents a variant of normal puberty. The pattern of pubertal progress in affected children is quite normal; however, its time of onset is delayed with respect to the population as a whole. Boys are referred for evaluation of this condition significantly more often than are girls. In conjunction with their delayed pubertal development, these children generally have mildly short stature, approximately two to three standard deviations below the mean for age throughout childhood. Their short stature is accompanied by delayed skeletal maturation (2–4 years behind chronological age) and these children generally demonstrate sexual maturation that is more commensurate with their bone age than their chronological ages. Adult height and sexual maturation are achieved significantly later, many young men reporting continued linear growth in their late teens or early 20s. Final height is generally appropriate for the genetic background, but is commonly in the low normal range. The family history often reveals other affected individuals in the family, most often males.

HYPOGONADOTROPIC HYPOGONADISM Hypogonadotropic hypogonadism describes the deficiency of pulsatile release of gonadotropins, which may result from a variety of hypothalamic or pituitary disorders. In the presence of a hypothalamic defect, or absence of GnRH-secreting neurons, failure of GnRH secretion results in lack of stimulation of pituitary gonadotropin secretion. In contrast, pituitary disorders, such as tumors or hypophysitis, cause direct failure of pituitary gonadotropin secretion. During embryogenesis, the olfactory nerves and terminal nerve develop from the olfactory placode in the nose. The olfactory sensory neurons have short dendrites that terminate at the olfcatory

CHAPTER 59 / DISORDERS OF PUBERTAL DEVELOPMENT

epithelium in dendritic knobs with cilia that contain G-protein coupled, seven transmembrane domain, odorant receptors. The olfactory nerves associate with the terminal nerve and vomeronasal nerve to produce a bridge between the olfactory epithelium and the forebrain. The cells that will become GnRH-synthesizing neurons arise within the region of the olfactory placodes and migrate from the nasal epithelium, through the cribriform plate of the nose and then along the olfactory tract–forebrain axis to reach the preoptic and hypothalamic areas, where they differentiate to become the GnRH-secreting neurons. The exact cellular origin of the GnRH neurons has yet to be established; however, there is evidence in studies of embryonic chicks that they in fact derive from an embryonic precursor in the region that will become the nasal epithelium, rather than the olfactory placode itself. These GnRHsecreting neurons, which are dispersed through the medial basal hypothalamus rather than being located in a discrete nucleus, transduce neural signals into a hormonal signal—the periodic secretion of GnRH. Given the developmental connection between olfactory and GnRH neurons, it is of particular interest to note the relationship between olfactory acuity and reproduction in animals, evidenced by the importance of pheromones in sexual attraction. KALLMANN’S SYNDROME AND VARIANTS Isolated gonadotropin deficiency may occur sporadically or in a familial pattern. In contrast to patients with delayed puberty because of CNS tumors or constitutional delay, those with gonadotropin deficiency usually have appropriate or tall stature for their age. Untreated adults and individuals of pubertal age commonly have eunuchoid proportions. Kallmann’s syndrome was first described in 1856. It is uncommon, occurring with approximately one-tenth the frequency of Klinefelter’s syndrome (approximately 1:10,000 in males and 1:50,000 in females). As originally described, it comprised the association of hypogonadotropic hypogonadism with anosmia or hyposmia resulting from agenesis or hypoplasia of the olfactory bulbs and tracts. The disorder results from failure of fetal GnRH secretory neurons to migrate from the olfactory placode, where they arise, to the medial basal hypothalamus. Three modes of transmission have been described: X-linked, autosomal recessive, and autosomal dominant. Typical clinical features of Kallmann’s syndrome include delayed puberty, eunuchoid habitus, gynecomastia, and reduced sense of smell (often unknown to the patient). Cryptorchidism and a small penis are present in infancy in some affected individuals. Carrier females may display partial defects, such as reduced sense of smell, delayed menarche, or irregular menses, but are normally fertile. Other clinical features of Kallmann’s syndrome include unilateral renal agenesis in up to 40% of patients; midline facial anomalies, such as cleft lip, high-arched or cleft palate, or other forms of imperfect facial fusion; short metacarpals; talipes cavus; cerebellar ataxia; sensorineural deafness; epilepsy; and synkinesia (mirror movements of the hands). Variant forms of Kallmann’s syndrome manifest varying combinations of these features. MRI scans may demonstrate hypoplasia of the olfactory gyri and absent olfactory bulbs and tracts. When analyzed histologically, the olfactory epithelium of a patient with Kallmann’s syndrome was thinner than normal and contained fewer neurons. Those that were present lacked cilia, were immature, or showed signs of degeneration. The X-linked form of Kallmann’s syndrome has been most extensively studied. Positional cloning studies by Legouis et al. led to the isolation and characterization of the gene defective in

577

this form of Kallmann’s syndrome. The Kallmann gene (KAL1 or KALIG1), which is located at Xp22.3, and (at least partially) escapes X inactivation, encodes a 680-amino acid protein that is apparently extracellular in location. An inactive homologous pseudogene (KALP) is located on the Y chromosome at Yq11. The derived amino acid sequence of KAL1 predicts a protein that contains a leader peptide and no transmembrane or anchoring regions are present. The protein contains four fibronectin type III motifs (FNIII), suggesting a role in cell adhesion. Its general structure is similar to that of known neural-cell adhesion molecules (NCAMs) that mediate neurite outgrowth or axon–axon interactions. Such a function would be consistent with the well-characterized defect of embryonic neuronal migration in patients with Kallmann’s syndrome. In addition to these motifs, the protein sequence also contains a region (the WAPS domain, or whey acidic protein core domain) that has been associated with proteinase inhibitor activity in a number of proteins. This association of domains in the predicted amino acid sequence of the KAL1 gene suggests that this protein may also play a role in the tissue remodeling that accompanies this neuronal migration. Examination of the pattern of expression of the gene during embryogenesis demonstrates that it can be detected in various neuronal populations of the central nervous system, including cells of the olfactory bulbs. These findings have suggested that the KAL1 protein might be involved both in directing neuronal migration, as well as influencing the differentiation of specific cell types. KAL is conserved in other mammals, including monkey, cow, rabbit, and sheep, and in xenopus, zebra fish, and chicken, in which it is expressed in a wide range of tissues. Of note, it is not conserved in mice or rats, a fact that has hampered its characterization. Studies of patients with the X-linked form of Kallmann’s syndrome indicate heterogeneity at the genetic level, although all but one of the mutations reported to date predict the absence or marked truncation of the encoded protein. In a study of 21 male patients with the X-linked form of the disorder, Hardelin and coworkers detected a range of abnormalities within the KAL1 gene, including large deletions in two families, and nine different point mutations in other affected individuals, resulting in introduction of premature termination codons into the sequence. These findings support the involvement of the KAL1 gene in the pathogenesis of Kallmann’s syndrome. It is suggested that defects of KAL1 gene expression may underlie the varied clinical findings in Kallmann’s syndrome. The relationship of KAL1 to disorders such as epilepsy or synkinesia is suggested by knowledge of its role in neuronal migration. The role of KAL1 in events such as kidney formation and migration is more obscure; however, KAL is expressed in the Wolffian duct, which is involved in renal development by interaction with the metanephric mesenchyme. To date, the genetic causes of the other forms of Kallmann’s syndrome (i.e., non-X-linked) have not been defined. However, studies of individuals with Kallmann’s syndrome associated with cytogenetic abnormalities have suggested possible loci for genes responsible for the autosomal dominant form of Kallmann’s syndrome (KAL2) at 1q, 7q, and 12q. It is likely that the genetic defects will be traced to other genes that participate in the pathways affecting migration of the GnRH-secreting neurons. Treatment of individuals with Kallmann’s syndrome with supplemental testosterone will induce virilization, whereas intranasal GnRH or intramuscular chorionic gonadotropin can stimulate endogenous sex steroid production and even reproductive capability.

578

SECTION VII / ENDOCRINOLOGY

ISOLATED GONADOTROPIN DEFICIENCY A number of families have been reported in which affected individuals have isolated gonadotropin deficiency without the features of Kallmann’s syndrome. The inheritance pattern is autosomal recessive. Although not yet reported in humans, it is likely that this defect results from GnRH deficiency as a result of mutations in the gene encoding GnRH, as has been observed in the hypogonadal (Hpg) mouse. ISOLATED FSH DEFICIENCY A small number of patients have been reported to have isolated deficiency of FSH. A woman with isolated deficiency of FSH who presented with primary amenorrhea and infertility was found to have a mutation in the gene encoding the β-subunit of FSH. The 2-bp sdeletion caused a frameshift in the coding sequence, predicting the formation of a truncated protein that lacks regions required for association with the α-subunit of FSH and for binding to the FSH receptor. FUNCTIONAL GONADOTROPIN DEFICIENCIES Severe systemic and chronic disorders and malnutrition are associated with delayed puberty or failure to progress through the stages of puberty. In general, when body weight is less than 80% of ideal weight for height a functional gonadotropin deficiency may occur. Examples include the functional gonadotropin deficiencies associated with anorexia nervosa and exercise-induced amenorrhea. In such instances, normal hypothalamic–pituitary–gonadal function accompanies restoration of normal body mass.

CNS DISORDERS Mass lesions, such as sellar or suprasellar tumors (e.g., craniopharyngioma), commonly disturb the processes of pubertal development, causing either precocious puberty (as described above) or pubertal delay. Such tumors cause pubertal delay by impairing hypothalamic or pituitary function. In addition to abnormalities of pubertal development, growth failure, polydipsia, polyuria, and visual disturbances may also be part of the presentation of children with craniopharyngioma.

HYPERGONADOTROPIC HYPOGONADISM Primary gonadal failure is associated with elevated gonadotropins because of the absence of negative feedback effects of gonadal sex steroids. The most common causes of hypergonadotropic hypogonadism are associated with karyotypic and somatic abnormalities, but isolated gonadal failure can also present with delayed puberty without other abnormal physical findings. MUTATIONS OF THE LH AND FSH RECEPTORS The receptors for the gonadotropins LH/CG and FSH are both members of the seven transmembrane domain family of G-protein coupled receptors. Selective defects in the gonadal response to gonadotropins have been traced in several pedigrees to mutations of the genes encoding the FSH and LH/CG receptors. In males, rather than being associated with delayed puberty, mutations in the gene encoding the LH/CG receptor are classically associated with a form of male pseudohermaphroditism termed Leydig-cell hypoplasia (LCH). This disorder is characterized by female phenotype in the presence of a 46,XY karyotype, low serum testosterone, increased serum LH , and lack of testosterone secretion in response to hCG administration. Mutations in the LH/CG gene, particularly in the transmembrane regions of the receptor protein, causing this phenotype include those that result in premature termination or amino acid replacement.

The discovery of inactivating mutations in the LH receptor in patients with LCH and defective fetal masculinization is to be expected. Several interesting aspects of LH receptor mutations deserve comment. First is that a 46,XX sibling of two patients with LCH was a phenotypically normal adult female with amenorrhea and cystic ovaries, suggesting that such a defect was consistent with normal pubertal development, but impairment in the normal ovarian cycle. Second, the identification of a missense mutation of the LH receptor in a phenotypic male infant evaluated for a small but normally formed penis suggests that the range of altered phenotypes associated with abnormalities of the LH receptor may be broader than those initially identified. An uncommon form of hypergonadotropic hypogonadism is 46,XX gonadal dysgenesis. Affected girls, who typically present with pubertal failure, are of normal stature and have no phenotypic features of Turner’s syndrome. The condition appears to be genetically heterogeneous and both sporadic and familial cases have been reported. One familial variant—Perrault syndrome—has sensorineural hearing loss associated with the hypogonadism. A candidate recessive gene on chromosome 2, the ODG1 gene (ovarian dysgenesis 1) was identified by analysis of a large cohort of affected Finnish patients. The disorder is common in this population (1: 8300 females) and is inherited in an autosomal recessive manner. Aittomaki et al. have recently discovered a missense mutation in the gene encoding the FSH receptor in affected families. The FSH receptor gene is located at 2p21, coinciding with the ODG1 locus. A limited number of families with such defects have been described to date. The mutation in the FSHR gene in this group is an arginine to valine substitution at amino acid 189, located in the extracellular ligand-binding domain of the receptor. The mutation segregated with the affected phenotype and had a dramatic effect on the binding of ligand and the stimulation of cAMP production. Of interest, the affected males in the pedigrees were phenotypically normal and half were fertile, suggesting that variable defects in spermatogenesis might be the only discernible effect in males with the disorder. KLINEFELTER SYNDROME The most common form of primary testicular failure is Klinefelter’s syndrome (47,XXY karyotype), which occurs with an incidence of 1:1000 males. Male sexual differentiation is normal; testicular function remains relatively normal until about the age of puberty, declining thereafter. Patients with Klinefelter’s syndrome do not usually present with delayed puberty, although affected patients may demonstrate a slowing or arrest of pubertal development as testicular function declines. Supplemental testosterone is then required. Klinefelter’s syndrome and its variants are described in Chapters 57 and 58. TURNER’S SYNDROME Turner’s syndrome, 45,X gonadal dysgenesis, is associated with short stature, female phenotype, and delayed or absent pubertal development. Patients have “streak” gonads consisting of fibrous tissue without germ cells (germ cells may be present in infancy). Other classic but variable phenotypic features include ptosis, low-set ears, micrognathia, a short “webbed” neck, a broad shieldlike chest, hypoplastic areolae, short fourth or fifth metacarpals, cubitus valgus, structural anomalies of the kidney, extensive pigmented nevi, hypolastic, hyperconvex fingernails and toenails, and left-sided cardiovascular anomalies, coarctation of the aorta being the most common. Disordered growth is a major clinical feature of this syndrome, beginning in utero and worsening progressively from early childhood

CHAPTER 59 / DISORDERS OF PUBERTAL DEVELOPMENT

onward. Patients have no pubertal growth spurt and reach a mean final height approximately 20 cm shorter than that of the reference population. Although short stature is a classic feature of Turner’s syndrome, it is not a feature of other forms of hypergonadotropic hypogonadism that occur without karyotypic abnormalities. Because growth hormone secretion is usually normal in Turner’s syndrome, it is likely that the short stature is related to a subtle form of skeletal dysplasia, reflecting imbalances of gene expression caused by the absent X chromosome segments. Growth hormone treatment improves growth rates and the adult height of many affected patients. Pubic hair may appear late and is usually sparse in distribution. Serum gonadotropin concentrations in Turner syndrome are high between birth and 4 years of age. They decrease toward the normal range in prepubertal patients and then rise to castrate levels after 9 or 10 years of age. Variant forms of gonadal dysgenesis are associated with a variety of mosaic phenotypes, usually including a 45,X line, in addition to more complex X chromosomal rearrangements. Girls and women with these karyotypes may have phenotypic features typical of those with the classic syndrome of 45,X gonadal dysgenesis, or may have fewer manifestations. Patients with Swyer syndrome (46,XY complete gonadal dysgenesis) have streak gonads and pubertal delay similar to those with Turner’s syndrome. They do not, however, have the short stature typical of Turner’s syndrome and, in most cases, do not have the other phenotypic features. Gonadal dysgenesis and other sex chromosome defects are outlined in Chapters 57 and 58. Other causes of primary ovarian failure include radiation therapy, chemotherapy, and premature menopause. Patients with Addison’s disease may have autoimmune oophoritis and other features of autoimmune disease, in addition to their adrenal failure. Primary amenorrhea is a common presenting feature of women with androgen-insensitivity syndrome (phenotypic females with 46, XY genotype and androgen resistance). A sex steroid biosynthetic defect as a result of 17α-hydroxylase deficiency will be manifested by delayed puberty and primary amenorrhea in a phenotypic female (regardless of genotype) with hypokalemia and hypertension. Breast development in female patients with pubertal delay may be induced by estrogen replacement therapy using either a conjugated equine estrogen (such as Premarin) or synthetic estrogen (such as ethinyl estradiol) at slowly increasing doses until feminization is achieved. Thereafter, a progestin is required to induce uterine endometrial cycling, and for convenience, an appropriate relatively low-dose combined oral contraceptive pill may be used. Estrogen replacement therapy in Turner’s syndrome is more complex, requiring coordination of timing with respect to growth hormone therapy to optimize adult height achievement.

SELECTED REFERENCES Ahima RS, Dushay J, Flier SN, Prabakaran D, Flier JS. Leptin accelerates the onset of puberty in normal female mice. J Clin Invest 1997;99:391–395. Aittomäki K, Lucena JLD, Pakarinen P, et al. Mutation in the folliclestimulating hormone receptor gene causes hereditary hypergonadotropic ovarian failure. Cell 1995;82:959–968. Burstein S. Editorial: growth disorders after cranial radiation in childhood. J Clin Endocrinol Metab 1994;78:1280,1281. Balducci R, Boscherini B, Mangiantini A, Morellini, Toscano V. Isolated precocious pubarche: an approach. J Clin Endocrinol Metab 1994;79: 582–589. Chehab FF, Mounzih K, Lu R, Lim ME. Early onset of reproductive function in normal female mice treated with leptin. Science 1997;275:88–90.

579

Clayton RN. Molecular genetics, hypogonadism and luteinizing hormone. Clin Endocrinol 1992;34:201,202. Duke VM, Winyard PJ, Thorogood P, Soothill P, Bouloux PM, Woolf AS. KAL, a gene mutated in Kallmann’s syndrome, is expressed in the first trimester of human development. Mol Cell Endocrinol 1995;110:73–79. Furui K, Suganuma N, Tsukahara S-I, et al. Identification of two point mutations in the gene coding luteinizing hormone (LH) β-subunit, associated with immunologically anomalous LH variants. J Clin Endocrinol Metab 1994;78:107–113. Feuillan PP, Foster CM, Pescovitz OH, et al. Treatment of precocious puberty in the McCune-Albright syndrome with the aromatase inhibitor testolactone. N Engl J Med 1986;315:1115–1119. Georgopoulos NA, Pralong FP, Seidman CE, Seidmen JG, Crowley WF Jr, Vallejo M. Genetic heterogeneity evidenced by low incidence of KAL-1 gene mutations in sporadic cases of gonadotropin-releasing hormone deficiency. J Clin Endocrinol Metab 1997;82:213–217. Hardelin JP, Levilliers J, Blanchard S, et al. Heterogeneity in the mutations responsible for X chromosome linked Kallman syndrome. Hum Mol Genet 1993;2:373–377. Hardelin JP, Petit C. A molecular approach to the pathophysiology of the X-linked Kallman’s syndrome. Baillieres Clin Endocrinol Metab 1995;9:489–507. Holland FJ, Fishman L, Bailey JD, Fazekas AT. Ketoconazole in the management of precocious puberty not responsive to LHRH analogue therapy. N Engl J Med 1986;312:1023–1027. Jameson JL. Inherited disorders of the gonadotropin hormones. Mol Cell Endocrinol 1996;125:143–149. Junier MP, Hill DF, Costa ME, Felder S, Ojeda SR. Hypothalamic lesions that induce female precocious puberty activate glial expression of the epidermal growth factor receptor gene: differential regulation of alternatively spliced transcripts. J Neurosci 1993;13:703–713. Kappy MS, Ganong CS. Advances in the treatment of precocious puberty. Advances in Pediat 1994;41:223–261. Kawate N, Kletter GB, Wilson BE, Netzloff ML, Menon KMJ. Identification of constitutively activating mutation of the luteinizing hormone receptor in a family with male limited gonadotrophin independent precocious puberty (testotoxicosis). J Med Genet 1995;32:553,554. Kosugi S, Van Dop C, Geffner ME, et al. Characterization of heterogeneous mutations causing constitutive activation of the luteinizing hormone receptor in familial male precocious puberty. Hum Mol Genet 1995;4:183–188. Kraaij R, Post M, Kremer H, et al. A missense mutation in the second transmembrane segment of the luteinizing hormone receptor causes familial male-limited precocious puberty. J Clin Endocrinol Metab 1995;80:3168–3172. Latronico AC, Anasti J, Arnhold IJ, et al. A novel mutation of the luteinizing hormone receptor gene causing male gonadotropin-independent precocious puberty. J Clin Endocrinol Metab 1995;80:2490–2494. Latronico AC, Anasti J, Arnhold IJP, et al. Testicular and ovarian resistance to luteinizing hormone caused by inactivating mutations of the luteinizing hormone. N Engl J Med 1996;334:507–512. Laue L, Chan WY, Hsueh AJ, et al. Genetic heterogeneity of constitutively activating mutations of the human luteinizing hormone receptor in familial male-limited precocious puberty. Proc Natl Acad Sci USA 1995;92:1906–1910. Laue L, Jones J, Barnes KM, et al. Treatment of familial male precocious puberty with spironolactone, testolactone, and deslorelin. J Clin Endocrinol Metab 1993;76:151–155. Laue L, Wu SM, Kudo M, et al. A nonsense mutation of the human luteinizing hormone receptor gene in Leydig cell hypoplasia. Hum Mol Genet 1995;4:1429–1433. Lee PA. Advances in the management of precocious puberty. Clin Pediatr 1994;33: 54–61. Legouis R, Cohen-Salmon M, Del Castillo I, Petit C. Isolation and characterization of the gene responsible for the X chromosome-linked Kallmann syndrome. Biomed Pharmacother 1994;48:241–246. Matthews CH, Borgato S, Beck-Peccoz P, et al. Primary amenorrhoea and infertility due to a mutation in the β-subunit of follicle-stimulating hormone. Nat Genet 1993;5:83–86.

580

SECTION VII / ENDOCRINOLOGY

Mantzoros CS, Flier JS, Rogol AD. A longitudinal assessment of hormonal and physical alterations during normal puberty in boys. V. Rising leptin levels may signal the onset of puberty. J Clin Endocrinol Metab 1997;82:1066–1070. Nachtigall LB, Boepple PA, Pralong FP, Crowley WF Jr. Adult-onset idiopathic hypogonadotropic hypogonadism—a treatable form of male infertility. N Engl J Med 1997;336:410–415. Neely EK, Bachrach LK, Hintz RL, et al. Bone mineral density during treatment of central precocious puberty. J Pediatr 1995;127:819–822. Ogilvy-Stuart AL, Clayton PE, Shalet SM. Cranial irradiation and early puberty. J Clin Endocrinol Metab 1994;78:1282–1286. Paul D, Conte FA, Grumbach MM, Kaplan SL. Long term effect of gonadotropin-releasing hormone agonists therapy on final and near-final height in 26 children with true precocious puberty treated at a median age of less than 5 years. J Clin Endocrinol Metab 1995;80:546–551. Pescovitz OH, Comite F, Hench K, et al. The NIH experience with precocious puberty: diagnostic subgroups and response to short-term luteinizing hormone releasing hormone analogue therapy. J Pediatr 1986;108:47–54. Plant TM. Puberty in primates. In: Knobil E, Neill JD, eds. The Physiology of Reproduction, 2nd ed. New York: Raven, 1994; pp. 453–485. Rao E, Weiss B, Fukami M, et al. Pseudoautosomal deletions encompassing a novel homebox gene cause growth failure in idiopathic short stature and Turner syndrome. Nat Genet 1997;16:54–63. Schwindinger WF, Francomano CA, Levine MA. Identification of a mutation in gene encoding the subunit of the stimulatory G protein of adenylyl cyclase in McCune-Albright syndrome. Proc Natl Acad Sci USA 1992;89:5152–5156. Shankar RR, Pescovitz OH. Precocious puberty. Adv Endocrinol Metab 1995;6:55–89. Shenker A. G protein-coupled receptor structure and function: the impact of disease-causing mutations. Baillieres Clin Endocrinol Metab 1995;9:427–451. Shenker A, Laue L, Kosui S, Merendino JJ Jr, Minegishi T, Cutler GB Jr. A constitutively activating mutation of the luteinizing hormone receptor in familial male precocious puberty. Nature 1993;365:652–654. Smith EP, Boyd J, Frank GR, et al. Estrogen resistance caused by a mutation in the estrogen-receptor gene in a man. N Engl J Med 1994;331:1056–1061.

Styne DM. New aspects in the diagnosis and treatment of pubertal disorders. Pediatr Clin North Am 1997;44:505–529. Tapanainen JS, Aittomaki K, Min J, Vaskivuo T, Huhtaniemi IT. Men homozygous for an inactivating mutation of the follicle-stimulating hormone (FSH) receptor gene present variable suppression of spermatogenesis and fertility. Nat Genet 1997;15:205,206. Weinstein LS, Shenker A, Gejm an PV, Merino MJ, Friedman E, Spiegel AM. Activating mutations of the stimulatory G protein in the McCuneAlbright syndrome. N Engl J Med 1991;325:1688–1695. Weiss J, Axelrod L, Whitcomb RW, Harris PE, Crowley WF, Jameson JL. Hypogonadism caused by a single amino acid substitution in the β subunit of luteinizing hormone. N Engl J Med 1992;326:179–183. Whitney EA, Layman LC, Chan PJ, Lee A, Peak DB, McDonough PG. The follicle-stimulating hormone receptor gene is polymorphic in premature ovarian failure and normal controls. Fertil Steril 1995;64: 518–524. Winter JSD, Hughes IA, Reyes FI, Faiman C. Pituitary-gonadal relations in infancy: Patterns of serum gonadal concentrations in man from birth to two years of age. J Clin Endocrinol Metab 1976;42:679–686. Winters SJ. Expanding the differential diagnosis of male hypogonadism. N Engl J Med 1992;326:193–195. Yano K, Hidaka A, Saji M, et al. A sporadic case of male-limited precocious puberty has the same constitutively activating point mutation in luteinizing hormone/choriogonadotropin receptor gene as familial cases. J Clin Endocrinol Metab 1994;79:1818–1823. Yano K, Kohn LD, Saji M, Kataoka N, Okuno A, Cutler GB Jr. A case of male-limited precocious puberty caused by a point mutation in the second transmembrane domain of the luteinizing hormone choriogonadotropin receptor gene. Biochem Biophys Res Commun 1996; 220:1036–1042. Yano K, Saji M, Hidaka A, et al. A new constitutively activating point mutation in the luteinizing hormone/choriogonadotropin receptor gene in cases of male-limited precocious puberty. J Clin Endocrinol Metab 1995;80:1162–1168. Zachman M, Prader A, Sobel EH, et al. Pubertal growth in patients with complete androgen insensitivity: indirect evidence for the importance of estrogen in girls. J Pediatr 1986;108:694–697.

CHAPTER 60 / DEFECTS OF ANDROGEN ACTION

60

581

Defects of Androgen Action MICHAEL J. MCPHAUL

INTRODUCTION Since 1935 it has been recognized that the principal androgen secreted by the testes is testosterone. The seminal observations of Bruchovsky and Wilson and of Anderson and Liao, however, led to the recognition that although testosterone is the most abundant circulating androgen, 5α-dihydrotestosterone is the predominant hormone found complexed to the androgen receptor (AR) in the nuclei of target cells, such as the prostate. This finding opened a new perspective on androgen physiology and focused attention on the enzyme that mediated this conversion, 5α-reductase, as a potential modulator of androgen action in selected tissues. As described below, this inference has been confirmed by the later recognition that in selected patients abnormalities in male development are caused by defects in the conversion of testosterone to 5α-dihydrotestosterone.

residues, and one of repeated glycine residues). Polymorphisms of these regions appear to have little effect on AR function in normal individuals, but expansions of the glutamine repeat have been implicated in the pathogenesis of spinal and bulbar muscular atrophy (Kennedy’s syndrome, see below). The non–ligand-bound AR is believed to exist in the cell in association with a number of ancillary proteins, particularly members of the heat shock protein family. After the binding of hormone, the receptor dissociates from these proteins and binds to specific DNA sequences within or adjacent to androgen-responsive genes. This ligand-activated receptor is believed to interact with components of the transcriptional apparatus to stimulate formation of active transcription complexes. In some models, the AR appears to modulate genes in a negative fashion or to alter mRNA stability, but these phenomena have been less-well-characterized.

ANDROGEN RECEPTOR DEFECTS

STRUCTURE AND MECHANISM OF ACTION OF THE ANDROGEN RECEPTOR Abundant biochemical and genetic data demonstrated that the actions of androgen were mediated by a single receptor that was encoded on the X-chromosome, and that defects in this receptor protein, the AR, could present as a range of abnormalities of male phenotypic sexual development. The isolation of cDNAs encoding the AR revealed it to be a member of a large group of related transcription factors, the steroid-thyroid hormone-retinoid family of nuclear receptors. This family includes members that are ligand responsive and others that are believed to be constitutively active or modulated by other influences, such as phosphorylation. All exhibit a modular structure (displayed for the AR in Fig. 60-1) comprising a highly conserved DNA-binding domain, a less highly conserved carboxy-terminal hormone-binding domain, and an amino-terminal segment that is poorly conserved between individual family members, both in terms of primary amino acid sequence and length. The DNA-binding domain is composed of two elements (termed “zinc fingers”) that mediate the sequencespecific DNA binding of the AR. This segment is the most highly conserved region between members of this gene family. The carboxy-terminus is approximately 250 amino acids long and encodes the portion of the protein that binds androgens with high affinity. Of note, the amino-terminus of the AR is somewhat atypical in that it contains three segments composed of repeated amino acids (one of repeated glutamine residues, one of repeated proline From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

CLINICAL FEATURES A spectrum of phenotypes can result from defects of AR function (Table 60-1). Patients completely unresponsive to the actions of androgen (referred to as complete testicular feminization or complete androgen insensitivity) have a 46,XY karyotype, but an external phenotype that is completely female in appearance, despite normal or elevated levels of circulating androgens. Owing to the secretion and action of Müllerianinhibiting substance (MIS) by the functional testes present in these patients, the uterus and fallopian tubes are absent. Such individuals are usually raised as females and may first seek attention for evaluation of primary amenorrhea. Gonadectomy is often performed, because the intra-abdominal testes show an increased rate of malignant tumor development. With estrogen replacement, these individuals often lead completely normal lives as women, although they are infertile. The term incomplete testicular feminization has been applied to individuals with nearly complete forms of androgen resistance who demonstrate only slight evidence of virilization (such as clitoromegaly). These patients are usually managed in a fashion similar to that of patients with complete testicular feminization. Partial androgen resistance or Reifenstein’s syndrome is a constellation of features that includes severe defects of male urogenital development (perineal or penoscrotal hypospadias) and gynecomastia. Reifenstein’s syndrome represents a far more difficult clinical problem. As noted, the developmental abnormalities are substantial and the surgical correction of these defects often requires multiple separate surgical sprocedures. Such efforts are hampered by the small size of the genitalia of affected children. After

581

582

SECTION VII / ENDOCRINOLOGY

Figure 60-1 Mutations of the human androgen receptor that cause abnormalities of androgen receptor function. A schematic representation of the human androgen receptor (AR) is shown. The approximate boundaries of the DNA- and hormone-binding domains are indicated, as are the locations of the repeated stretches of glutamine, proline, and glycine residues within the amino terminus. (Below) The mutations causing androgen resistance are grouped according to the type of genetic lesion (amino acid substitution or premature termination codon, left margin) or the effect that mutation has on receptor function (right margin). Also included in this figure are three mutations reported in the literature detected in prostate cancer specimens that appear to alter the binding characteristics and activities of the mutant ARs. Not depicted in this summary are deletions and insertions of AR gene segments that account for 5–10% of AR mutations. Expansions of the glutamine repeat that causes spinal and bulbar muscular atrophy are also not represented. (Adapted from McPhaul et al, J Clin Endocrinol Metab 1993;76:17–23.)

surgical correction, most individuals with this phenotype are raised as males, although many exhibit difficulty with gender identity. At the mildly affected end of the spectrum, a small number of individuals have been described in which normal development of the external genitalia has occurred, but in which some degree of undervirilization is clinically evident. In some, this phenotype has been associated with azoospermia and infertility without undervirilization, whereas in others normal sperm density and fertility occur in undervirilized men. Patients with these phenotypes have been identified as having AR defects on the basis of a family history and abnormalities of in vitro ligand-binding assays of the androgen receptor. Although specific AR defects have now been reported in association with these syndromes, it is not clear how frequently such phenotypes are caused by AR defects in the general population.

An unusual variation of the undervirilized phenotype is that presented by patients with spinal and bulbar muscular atrophy (SBMA, Kennedy’s syndrome) (see Chapter 100). Patients with this syndrome have normal male development and normal male secondary sexual characteristics throughout much of their lives but develop signs of clinical androgen resistance beginning in middle age. Although these signs of androgen resistance that they display are unmistakable (usually gynecomastia), the difficulties presented by the symptoms of anterior motor neuron degeneration pose a far more serious threat to the health of such patients and represent the usual cause of death in affected individuals. ANDROGEN RECEPTOR MUTATIONS AND DEFECTIVE AR FUNCTION Molecular defects of the AR that cause the syndromes of androgen resistance have been studied by a number

583

CHAPTER 60 / DEFECTS OF ANDROGEN ACTION

Table 60-1 Phenotypes Associated with Defects in the Genes Encoding the Androgen Receptor and 5α-Reductase II Gene AR

Syndrome Complete testicular feminization Complete androgen insensitivity Incomplete testicular feminization Reifenstein syndrome

5α-Reductase II

Under-virilized, fertile male Infertile male 5α-Reductase deficiency

External genitalia

Wolffian duct derivatives

Müllerian duct derivatives

Completely female

Not virilized

Absent

Predominantly female; some clitoromegaly, closure of the posterior fourchette Substantial phallic development, but with severe forms of hypospadias Male phenotype with small phallus

Usually partially virilized

Absent

Variable defects

Absent

Male

Absent

Male Normal male

Absent Absent

Male phenotype Can be variable: ranging from female external genitalia at birth to a predominantly male phenotype with hypospadias

of groups. The causative mutations have been identified in more than 100 pedigrees and now include all of the major clinical syndromes. A database listing the mutations causing androgen resistance is accessible via the Internet. Deletions or insertions of the AR gene have been found to occur with a frequency of approximately 5–10% of patients with androgen resistance and range in size from single or multiple nucleotides to deletions of the entire gene. Because such mutations disturb the open reading frame of the AR, patients with this type of mutation do not express an intact receptor protein and, with few exceptions, lack detectable androgen binding in their cells and tissues. In addition, because the intact AR is not present in the tissues of such patients, AR function is absent and affected individuals always demonstrate the phenotype of complete testicular femininization. In contrast to the frequency of deletions and insertions in the AR gene, single nucleotide substitutions are much more frequent and account for the bulk of patients with androgen resistance as a result of AR mutations. In some cases, such substitutions result in large changes of receptor structure, as when they result in alterations of AR mRNA splicing or introduction of premature termination codons within the AR open reading frame. These instances, as with gene deletion or insertion mutations that interrupt the primary sequence of intact AR protein, are usually associated with a lack of detectable androgen binding in tissues and a phenotype of complete testicular feminization. Single nucleotide substitutions that result in single amino acid changes within the AR protein are the most frequent type of mutation in the AR. To this point, these defects have been found to fall into two large categories: those within the DNA-binding domain and those that have been localized to the hormone-binding domain of the receptor. Unlike the other mutation categories described above (deletions, insertions, premature termination), the effect of these substitutions on AR function can be quite variable, and thus the entire spectrum of androgen-resistant phenotypes has been traced to single amino acid substitutions within the AR protein. In most instances, the principal effect of the amino acid substitution is on AR function, whereas major effects on the level of AR abundance, as measured in patient fibroblasts, are uncommon.

Amino acid substitutions in the DNA-binding domain have little effect on the binding of ligand by the mutant receptor. Despite the capacity of these mutant ARs to bind ligand normally, receptor function is reduced after ligand stimulation, compared with the activity of the normal AR. This reduced function can be shown to be caused by a decreased capacity of the mutant AR to bind to target sequences within responsive genes. In this category of mutation, the degree of impairment of receptor binding to DNA appears to have a direct relationship to the degree that receptor function is reduced. Similar conclusions hold true for larger alterations of receptor structure occurring within the DNA-binding domain that maintains the reading frame of the receptor (e.g., deletions of single amino acid residues within the DNA-binding domain or the single reported instance caused by in-frame deletion of exon 3). Amino acid substitutions in the hormone-binding domain (HBD) can result in a variety of changes in the capacity of the AR to bind its ligand. In a surprisingly small proportion of cases, the amino acid replacement renders the receptor completely unable to bind ligand. Under these circumstances, it is presumed that the alteration of the HBD structure is distorted. This conformational change blocks the capacity of the AR to bind its ligand and the mutant receptor is thus “locked” into an inactive conformation. This type of mutant AR is functionally equivalent to mutations that result in the synthesis of truncated forms of the AR, because even pharmacologic doses of potent synthetic androgens are unable to restore receptor function in vivo or in functional assays performed in vitro. More frequently, however, amino acid substitutions in the HBD lead to the synthesis of mutant ARs that exhibit ligand-binding properties that are abnormal compared with the binding properties of the normal AR (e.g., bind ligand with a reduced affinity or stability). Although the exact type of ligand-binding abnormality differs depending on the nature and location of the amino acid substitution within the hormone-binding domain, it appears that the formation and stability of the hormone–receptor complex is the final determinant of the degree of impairment of mutant AR function. In vitro studies of such mutant receptors demonstrate that the use of multiple high doses of physiologic ligands

584

SECTION VII / ENDOCRINOLOGY

(testosterone or dihydrotestosterone) or the use of potent nonmetabolizable androgen agonists can overcome the defective function of many mutant ARs of this type. In addition to its mechanistic implications, this finding may have substantial clinical implications as well, since it suggests that the pharmacologic manipulation of many mutant ARs may be possible. Studies have now been performed to study the levels of expression and the function of normal and mutant ARs. It seems clear that the major effect of most mutations in the AR is not at the level of AR abundance, but at the level of AR function. This is not true in all cases, however, with the most obvious exceptions being mutations that result in truncation of the receptor protein (e.g., termination codons or alterations of mRNA splicing). There also appears to be a relationship between the level of mutant AR function and the phenotype observed in the patient: mutant ARs devoid of function being associated with complete testicular feminization, those with partial activity associated with incomplete forms of androgen resistance. This relationship is most clearly seen using assays that measure the level of AR function in fibroblasts established from the patient under study (e.g., delivery of an androgen-responsive reporter gene directly into the cultured fibroblasts). ANDROGEN RECEPTOR MUTATIONS—SPECIAL CASES The androgen-resistance syndrome described above is caused by mutations that impair receptor function to various degrees. Two additional types of AR mutation have been identified that result in an alteration of AR responsiveness or an apparent “gain” of function. The first example of a “gain of function” mutation in the androgen receptor is the genetic defect causing Kennedy’s syndrome. The pathogenesis of this extraordinary disease was traced to the expansion of a triplet nucleotide repeat (CAG) encoding a repeated sequence of glutamine residues within the amino-terminus of the receptor (see Fig. 60-1). The expansion of this glutamine repeat, the first of an increasing number of similar diseases that have also traced to the expansion of trinucleotide repeats (see Chapter 100), is believed to have two different effects on AR function. First, it is clear that the expansion of this segment of the AR diminishes the capacity of the AR to activate responsive genes after androgen stimulation. This diminished AR function can be demonstrated using a variety of functional assays and is presumably the cause of the subtle signs of androgen resistance observed in affected individuals. It is clear from the study of rare patients with complete deletions of the AR that the SBMA phenotype (progressive death of motor neurons in bulbar nuclei and in the spinal cord) is not caused by a simple lack of functional AR. Instead, this disease appears to be caused by some type of toxic “gain of function” that is caused by the expression of androgen receptors containing the expanded glutamine repeats in specific cell types. Analogy to results obtained from the study of the Huntington disease gene (also caused by a glutamine repeat expansion) would suggest that the glutamine expansion may permit interaction of the mutant AR with intracellular targets in spinal motor neurons that somehow mediate the observed cell-type-specific toxicity. Studies have also raised the possibility that a lower number of CAG repeats may correlate with the age of onset of prostate cancer. The second type of AR mutation that causes a gain or alteration of receptor function are the mutations that have been identified in human prostate cancer specimens. First identified in a prostate tumor cell line, LNCaP, amino acid substitutions have been identified in a number of clinical prostate cancer specimens. Although a number of these mutations have been identified, fewer have been

completely studied as to the effect that they have on the ligand responsiveness of the mutated AR. In those instances in which detailed studies have been performed, tests of receptor function using androgen-responsive reporter genes demonstrate that the mutant ARs can be stimulated by ligands that cannot ordinarily activate the normal unmutated AR. Such ligands include adrenal androgens and even compounds (such as hydroxyflutamide) that normally act to antagonize the function of the unmutated AR. It is not yet possible to conclude how important these mutant receptors are in the progression of prostate cancers toward the androgenindependent phenotype.

5α-REDUCTASE DEFICIENCY PHYSIOLOGIC AND CLINICAL STUDIES The observation that 5α-dihydrotestosterone was the principal hormone bound to the AR in the nuclei of target cells suggested the potential importance of 5α-reductase in androgen physiology. The subsequent identification of rare patients with specific defects of virilization who formed reduced quantities of 5α-reduced androgen metabolites served to emphasize the importance of this metabolic step. The normal virilization of other tissues, such as the epididymis in patients with clinical 5α-reductase deficiency, led to the concept that the action of testosterone was sufficient to effect the actions of androgen in selected tissues, but that the formation of 5α-dihydrotestosterone was crucial in others, such as the prostate, a dichotomy that as yet is still unexplained mechanistically. CLINICAL PHENOTYPE The clinical features of infants with a deficiency of 5α-reductase are consistent with a marked defect of androgen action (Table 60-1). The external genitalia are characterized by a microphallus, severe hypospadias, and a bifid scrotum. A blind vaginal pouch is present and opens either directly onto the perineum or onto a urogenital sinus. Owing to the production of MIS by the functional testes, no Müllerian structures are present. It is of interest that at puberty, several changes take place in the phenotype of affected individuals. The phallus enlarges and some male secondary sexual characteristics appear, such as changes in voice and muscle mass. Notably, this increase in testosterone has not been reported to result in acne, prostate growth, or male pattern baldness. It has also been reported that in such individuals, raised initially as females, gender identity may change after the pubertal rise in androgen levels. 5α-REDUCTASE STRUCTURE AND MECHANISM OF ACTION Attempts to purify 5α-reductase using classic techniques of protein purification failed and its structure remained elusive until Anderson and colleagues employed expression cloning in Xenopus oocytes to isolate a cDNA encoding a steroid 5α-reductase from rat prostate in 1992. This advance permitted a number of studies by this same group and resulted in the isolation of cDNAs encoding two related isozymes (Fig. 60-2) encoded by two distinct, related genes from humans and from all vertebrate species examined to date. Inspection of the structures of the cDNAs indicate substantial sequence divergence between the two isozymes, both within and between different species (Fig. 60-2). Both enzymes are extremely hydrophobic and are believed to be imbedded in the nuclear membrane of cells, accounting for the failure of extensive efforts to purify the enzyme in an active form. Structural differences between the two isozymes confer substantial differences in their physical properties that have been exploited to develop compounds that inhibit one or the other isozyme preferentially.

585

CHAPTER 60 / DEFECTS OF ANDROGEN ACTION

Figure 60-2 Comparison of the structures of human steroid 5α-reductase I and II and mutations that alter the binding of testosterone and NADPH. (Above) Alignment of the predicted protein sequences for steroid 5α-reductase I (259 amino acids) and steroid 5α-reductase II (254 amino acids). The degree of sequence identity is shown for each of the five coding exons (boxed percentages). (Below) Mutations identified within the coding sequence of steroid 5α-reductase deficiency. Although mutations have been detected in each coding exon that result in 5α-reductase II deficiency, only mutations that cause alterations in the binding of testosterone or NADPH are indicated. (Drawn after Russell et al, Annu Rev Biochem 1994;63:25–61.)

Studies of the distribution of the two isozymes (Table 60-2) have demonstrated that the patterns of expression of these two proteins differ, both in terms of tissue, cell type, and stage of development. These differences suggest substantially different physiologic roles for the two enzymes. GENETIC DEFECTS OF STEROID 5α-REDUCTASE II The genetic lesions causing clinical 5α-reductase deficiency have now been identified in a number of families from around the world. In all instances, when a genetic defect has been detected, it has been traced to a mutation within the 5α-reductase II gene. Unlike the AR gene, the gene encoding 5α-reductase II is autosomal and each individual possesses two copies of the 5α-reductase II gene. For this reason, mutations causing clinical 5α-reductase deficiency are recessive and the inheritance of two defective 5α-reductase II gene alleles is required for the defects to be manifest. 5α-Reductase II deficiency is infrequently caused by deletions or insertions in the gene. More often, defects in the gene are caused by mutations that result in premature termination of the protein or single amino acid substitutions within the open reading frame. Identification of the locations of these substitutions and determination of the physical properties of the mutant enzymes have permitted the identification of amino acid residues important for binding of steroid substrates and for the binding of nicotinamide adenine dinucleotide phosphate (NADPH), a cofactor required for the reduction reaction. Of interest, these studies have also identified sites that have been mutated repeated in apparently unrelated pedigrees. These sites are apparently more susceptible to mutation. As a consequence of the recessive nature of clinical 5α-reductase II deficiency, it would be expected that the genetic basis of this rare trait would be traced most frequently to homozygosity for a single defective allele (e.g., consanguinity). Surprisingly, a high proportion of patients (40%) have been found to be compound heterozygotes, suggesting either an unexpectedly high mutation rate of the 5α-reductase II gene or a high frequency of individuals in the general population that carry single defective alleles. GENETIC DEFECTS OF STEROID 5α-REDUCTASE I An abnormal phenotype caused by a deficiency of 5α-reductase I has not yet been described in humans. The differential expression of steroid 5α-reductase I in selected tissues (see Table 60-2) suggests that specific phenotypes in humans that result from lesions in this gene may well be identified in the future.

Table 60-2 Distribution of 5α-Reductase I and II Expression in Rat and Human Tissues Rat

Skin Testes Epididymis Vas deferens Seminal vesicle Ventral prostate Prostate Ovary Adrenal Brain Colon Heart Intestine Kidney Liver Lung Muscle Spleen Stomach Pons Cerebral Hypothalamus

Human

Type I

Type II

Type I

Type II

— D* D*e D* D D d*

— 1+* 5+*e 1+* 1+* D d*

+* ND† ND† — ND†

+a† +b†,* — +†

ND†,*

+c†,*

2+* 3+* 3+* 3+* ND* 3+* 3+* 4+* 2+* ND* D* D* — — —

ND* D* ND* D* ND* D* ND* ND* ND* ND* ND* ND* — — —

+†,*

+†,*

ND ND ND

+ + +

Summary of studies measuring 5α-reductase I and II expression using measurements of the corresponding RNA (*) or protein (†) are derived from published studies of Russell and coworkers. The rating scales (1+ least, 5+ highest) used in this summary are designed to convey a sense of the relative abundance of the 5α-reductase isozymes in the two species— rat and human—that have been studied most carefully. Because the rat and human measurements have been conducted separately, they can only be compared with one another in a qualitative fashion. D is detectable (i.e., at the limits of detection). ND, is not detected; —, values not reported. In addition to the results tabulated here, the studies of Thigpen et al. clearly indicate that in the human substantial changes in abundance can be demonstrated in the expression of the type I and type II isoenzymes in a tissue at different times in development. aHistochemical studies indicate that the expression of 5α-reductase II is localized to the cells of the dermal papilla. bStaining localized to epithelial cells in histochemical studies. cStaining localized to basal epithelial and stroma cells. dType I is detected in basal epithelial cells and type II in stroma cells in in situ hybridization studies of the regenerating rat central prostate. eType I and II are localized to the epithelial cells using in situ hybridization.

586

SECTION VII / ENDOCRINOLOGY

SUMMARY The development of the male phenotype is a complicated process that involves the active participation of genes involved at many levels, from gonadal differentiation to the androgen receptor itself. Defective virilization can be caused at defects anywhere along the pathway displayed schematically in Fig. 60-3. Taking into account the clinical syndromes caused by known defects of genes such as steroid 5α-reductase II and the androgen receptor, the pathogenesis of a large proportion of defects in virilization remains unexplained. It is conceivable that some may represent defects in genes required for normal AR function or at steps beyond the site of action of the AR itself (e.g., coactivators, or defects in genes activated by the AR).

ACKNOWLEDGMENTS This work was supported by NIH grants DK03892 and DK52678.

SELECTED REFERENCES Figure 60-3 Schematic the pathways controlling the development of the normal male phenotype. Male sexual development is a complex cascade of events that can be affected by lesions in a number of genes. Some, such as lesions in the AR or SRY, can cause disturbances of virilization of all androgen-responsive tissues. The effects of defects in other genes, however (such as mutations of 5α-reductase II), are manifest only in selected tissues. It is likely that defects in other genes can contribute to defects in virilization, as a large proportion of the subjects with abnormalities of male phenotypic development cannot be accounted for by defects in the genes most carefully studied to date, such as the AR and steroid 5α-reductase II.

Owing to the lack of mutations in the human population, the first insights into the nature of defects that may be expected in humans have come from experiments in mice in which steroid 5α-reductase I gene has been disrupted by homologous recombination. Male mice homozygous for this targeted null allele develop normally, both prenatally and postnatally. Unexpectedly, although female mice homozygous for this same null allele develop normally into adulthood, when pregnant, such mice fail to initiate parturition normally. Experiments in which different steroids were administered to these steroid 5α-reductase I-deficient mice suggested that the synthesis and actions of 5α-reduced androgens were important elements of the normal parturition process in mice. Although the mechanisms and generality of these findings remain to be determined, such results suggest that 5α-reductase I may play important—and even unexpected—roles in human physiology as well.

Giovannucci E, Stampfer MJ, Krithivas K, et al. The CAG repeat within the androgen receptor gene and its relationship to prostate cancer. Proc Natl Acad Sci USA 1997;94:3320–3323. Gottlieb B. Androgen Receptor Gene Mutation Data Base [email protected] Hardy DO, Scher HI, Bogenreider T, et al. Androgen receptor CAG repeat lengths in prostate cancer: correlation with age of onset. J Clin Endocrinol Metab 1996;81:4400–4405. Hiort O, Sinnecker GH, Holterhus PM, Nitsche EM, Kruse K. The clinical and molecular spectrum of androgen insensitivity syndromes. Am J Med Genet 1996;63:218–222. Imperato-McGinley J. 5 Alpha-reductase-2 deficiency. Curr Ther Endocrinol Metab 1997;6:384–387. Lumbroso S, Lobaccaro JM, Vial C, et al. Molecular analysis of the androgen receptor gene in Kennedy’s disease. Report of two families and review of the literature. Horm Res 1997;7:23–29. Marcelli M, Zoppi S, Wilson CM, Griffin JE, McPhaul MJ. Amino acid substitutions in the hormone-binding domain of the human androgen receptor alter the stability of the hormone receptor complex. J Clin Invest 1994;94:1642–1650. McPhaul MJ, Marcelli M, Zoppi S, Griffin JE, Wilson JD. Genetic basis of endocrine disease 4. The spectrum of mutations in the androgen receptor gene that causes androgen resistance. J Clin Endocrinol Metab 1993;76:17–23. Quigley CA, DeBellis A, Marschke KB, El-awady MK, Wilson EM, French FS. Androgen receptor defects: historical, clinical, and molecular perspectives. Endocr Rev 1995;16:271–321. Mahendroo MS, Cala KM, Russell DW. 5α-Reduced androgens are required for parturition in mice. Mol Endocrinol 1996;10:380–392. Russell DW, Wilson JD. Steroid 5α-reductase: two genes/two enzymes. Annu Rev Biochem 1994;63:25–61. Thigpen AE, Silver RI, Guileyardo JM, Casey ML, McConnell JD, Russell DW. Tissue distribution and ontogeny of steroid 5α-reductase isozyme expression. J Clin Invest 1993;92:903–910. Wilson JD, Griffin JE, Russell DW. Steroid 5α-reductase 2 deficiency. Endocr Rev 1993;14:77–93.

CHAPTER 61 / TESTICULAR DISEASES

61

587

Testicular Diseases MARCO MARCELLI

TESTICULAR EMBRYOGENESIS Between 1947 and 1952 Alfred Jost formulated the concept that sexual differentiation is an ordered process in which chromosomal sex determines gonadal sex, and gonadal sex in turn directs the development of phenotypic sex. Since 1959 it has been known that in humans and other mammals the Y chromosome determines maleness. These insights led to the postulate of a gene or genes on the Y chromosome that would act as a “master switch” to set in motion the cascade of events that leads to the development of the male gonads. The discovery and the roles played by the genes involved in testicular embryogenesis and their involvement in human diseases and experimental models characterized by abnormal testicular development are discussed in Chapter 57.

BASIC ENDOCRINOLOGY OF THE TESTIS MICROSCOPIC ORGANIZATION From a functional point of view, the microscopic architecture of the testis consists of two regions, the interstitial and tubular compartments, which are separated by the basal lamina. The interstitial compartment contains loose connective tissue with interspersed epithelial cells, the Leydig cells, which are specialized for the synthesis and secretion of testosterone. The tubular compartment consists of cylindrical structures, the seminiferous tubules, which are lined with Sertoli and germinal cells. Sertoli cells create support for the germ cells by resting on the basal membrane of the tubule and extending apically, and are believed to produce several growth factors affecting the process of testicular development and sperm maturation. In addition, they offer a critical contribution to the process of male sexual differentiation through the production of anti-Müllerian hormone (AMH), which causes regression of the Müllerian ducts. The growth and maturation of the germinal epithelium into mature spermatozoa is a highly synchronized process. According to the degree of maturation, germ cells can broadly be classified as spermatogonia, primary spermatocytes, secondary spermatocytes, spermatids, and spermatozoa. However, in post pubertal individuals more than a dozen different stages of germinal cell maturation have been described on the basis of subtle histologic differences. Cytoplasmic projections originating from the Sertoli cells between spermatogonia and primary spermatocytes are connected by tight junctions, and divide the tubule into the basal compartment (containing the Leydig cells, the basal lamina, the myoid cells, and the spermatogonia) and the adluminal compartment From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

(containing primary and secondary spermatocytes, spermatids, and mature spermatozoa). Diffusion of protein, ions, or steroid hormones between the adluminal and basal compartments is impossible, since, similar to the blood–brain barrier, the tight junctions between these two compartments of the testis create an environment that is impermeable to macromolecules. PHASES OF SPERMATOGENESIS The germinal epithelium undergoes continuous replication to regenerate multipotent stem cells, as well as to produce cells that can proceed through subsequent maturative changes that result in the formation of mature spermatozoa. The mechanisms regulating sperm maturation are largely unknown; however, it is well-established that spermatogenesis occurs only in the presence of androgens and a functional androgen receptor, and that Sertoli cells have a dual role of providing physical support for the maturing germinal cells and of producing, under the stimulation of follicle-stimulating hormone (FSH) and testosterone, factors important for the stimulation of germ-cell maturation. Two human conditions, congenital hypogonadotropic hypogonadism (HH) and men after hypophysectomy, have been very instructive in understanding the role played by the two gonadotropins in inducing sperm maturation. In congenital HH, the administration of human chorionic gonadotropin (hCG), a luteinizing hormone (LH)-like molecule that induces high intratesticular level of testosterone, is followed by the induction of germ-cell maturation from spermatogonia to spermatids. To obtain the full maturation of spermatids to spermatozoa, however, the simultaneous administration of FSH is required. Similarly, replacement with both gonadotropins is necessary to restore spermatogenesis after hypophysectomy, although once spermatogenesis is restored, it can be maintained by injecting hCG alone. Hence, the actions of both gonadotropins are necessary for normal spermatogenesis to occur. Three main events are necessary for the completion of spermatogenesis: (1) stem-cell renewal by mitosis, (2) reduction of chromosomal number by meiosis, and (3) transformation of conventional cells into specialized elements that can reach and fertilize the egg. Spermatogonial Stem-Cell Renewal The most primordial cells of the germinal lineage are called spermatogonia, of which three types are usually recognizable morphologically. The type A dark spermatogonium, which is thought to be the most primordial, is characterized by a densely stained chromatin containing a centrally located pale area, known as the nuclear vacuole. The type A pale spermatogonium, has a palely staining chromatin and one or two nucleoli. The type B spermatogonium is smaller, has coarse granules of heavily stained chromatin, and is believed to represent a more advanced level of maturation when compared with the type A cells.

587

588

SECTION VII / ENDOCRINOLOGY

Meiotic Maturation A minority of type B spermatogonia commit themselves to further differentiation, and, by responding to unknown stimuli, divide and detach from the basement membrane of the tubule to form the primary spermatocytes. These cells undergo a complex maturation process involving two meiotic divisions: the first from primary spermatocytes into secondary spermatocytes and the second from secondary spermatocytes to spermatids. When primary spermatocytes reach the time of the first meiotic division, they have already engaged in DNA synthesis, and each of their chromosome pair consists of four chromatids. Hence, when the first meiotic division occurs, the resulting secondary spermatocytes contain a haploid number of chromosomes, but each is composed of a pair of chromatids (haploid number of chromosomes and diploid DNA content). Only after the second meiotic division do the developing germ cells (at this level of maturation termed spermatids) have both haploid chromosomal number and haploid DNA content. Spermiogenesis In the final phase of germ-cell maturation spermatids undergo a number of morphologic changes, and are transformed into specialized cells that can reach and fertilize the ovum. These can be conveniently followed in five phases, consisting of (1) nuclear changes, (2) acrosome formation, (3) flagellum development, (4) redistribution of cytoplasm, and (5) spermiation. During the final phases of sperm maturation the nuclear chromatin condenses, the nucleus becomes progressively smaller and more peripheral, and is positioned in close contact with the cell membrane (nuclear changes), from which it is separated by the acrosomal cap. The latter is formed from large vacuoles produced by the Golgi complex, and contains enzymatic activities important for the penetration of the ovum during fertilization (acrosomal formation). The next most important change observed during this phase consists in the development of the sperm tail (flagellar development), the structure capable of generating the motion of mature spermatozoa. Its architecture is highly conserved among species, and consists of nine peripheral doublet microtubules (known as subfibers A and B), surrounding an additional central pair of microtubules, and is arranged in pattern termed the “9 + 9 + 2” structure. Connections between each peripheral doublet are provided by dynein arms, which project from subfibers A toward subfibers B, and by nexin links. Dynein arms are made of dynein, a protein with ATPase activity that is vital to the generation of flagellar motility. Other connections between each peripheral doublet and the two central microtubules are provided by radial spikes. The intimate connections between the central and peripheral tubular structures of the sperm tail are critical for the achievement of the typical wavelike motion necessary for the propulsion of the mature spermatozoa. Diseases associated with absence of the dynein arms, and with sperm immotility, have been described, and are discussed in Table 61-1. The eccentric migration of the nucleus toward the acrosomal cap occurs simultaneously with the caudal migration of the cytoplasm toward the developing tail (redistribution of the cytoplasm). The latter is facilitated by a system of cytoplasmic filaments, which extend from the nuclear membrane close to the acrosome to the caudal end of the developing sperm. The seminiferous tubules empty into a network of ducts called the rete testis. By the time the spermatids reach the rete testis, they have been shed of almost the entire cytoplasm, which at this stage is termed the residual body, and is phagocytosed and degraded by the surrounding Sertoli cells (spermiation).

FINAL PHASES OF SPERM MATURATION The interval from the beginning of spermatogenesis to the delivery of sperm into the rete testis is approximately 74 days. During the next 12 days, spermatozoa transit through the epididymis and vas deferens, undergo a number of final morphologic and functional changes, and acquire motility and full capacity to fertilize an ovum. PHYSIOLOGY OF TESTICULAR FUNCTION A discussion of the hormonal regulation of testicular function, of the synthesis and mechanism of action of testicular steroids, and of the feedback mechanisms regulating the hypothalamic–pituitary–testicular axis is presented in Chapter 56.

MALE HYPOGONADISM Disorders of testicular function result from abnormalities involving the endocrine (Leydig cells) or reproductive (germ-cell maturation) compartments of the testis. If low testosterone production is present, it is usually accompanied by infertility, because normal spermatogenesis requires normal production of testicular androgens (hypogonadism with undervirilization and infertility). However, there are situations in which hypogonadism is restricted to the germinal compartment (hypogonadism with infertility and normal virilization). The serum levels of gonadotropins indicate at what site within the hypothalamic–pituitary–gonadal (HPG) axis the defect is localized. In those instances in which gonadal function is abnormal (hypergonadotropic or primary hypogonadism), gonadotropin levels are elevated. In those instances in which the hypothalamopituitary structures are defective (hypogonadotropic or secondary hypogonadism), gonadotropin levels are inappropriately normal or low. Depending on the serum level of gonadotropins, hypogonadism can be classified as hypergonadotropic or hypogonadotropic (if caused by testicular or hypothalamic-pituitary disorders, respectively) and eugonadotropic (if the abnormality lies only in the germinal cells) (Tables 61-1 through 61-3). Both congenital and acquired forms of Leydig cell dysfunction have been described, and the clinical picture that results when testosterone synthesis is impaired is different depending on whether androgen deficiency developed prenatally, before puberty, or after puberty (Table 61-4). If the developing fetus was not exposed to adequate level of androgen as a result of testicular failure during fetal development, the infant will be affected with pseudohermaphroditism, characterized by a large spectrum of potential abnormalities, including ambiguous genitalia, micropenis, and rudimentary testes (Table 61-4). If reduced testosterone production developed before puberty, but testicular function was normal during embryogenesis, androgen-related somatic changes that are observed at puberty are absent or incomplete and the patient develops an eunuchoid habitus because of failure of closure of the epiphyses and poor development of the skeletal muscles and body hair (Table 61-4). If reduced testosterone production developed after puberty, the first symptom of the patient is impotence, whereas loss of secondary sexual characteristics may take 5–10 years to become complete (Table 61-4). Hypogonadism manifested by abnormal sperm production (Table 61-1 and 61-3) can be associated with normal virilization, since many infertile men produce normal or only minimally abnormal amounts of testosterone. However, in other cases infertility is associated with a reduced production of testosterone and clinical signs of androgen deficiency (infertility from hypergonadotropic or hypogonadotropic disorders [Table 61-2]). In such cases, infertility is a direct consequence of the reduced testosterone production.

589

CHAPTER 61 / TESTICULAR DISEASES

Table 61-1 Eugonadotropic Germinal Cell Failure Clinical features and causes Ductal obstruction Varicocele

Treatment

See text. Most frequently recognized cause of infertility; however, infertility is not always associated with its presence. It is still debated why varicocele causes infertility. CAH (congenital adzenal Suppression of LH and FSH by supraphysiologic hyperplasia) production of androgens. Infections The microorganisms most frequently detected are Ureaplasma urealyticum and Chlamydia trachomatis. They are thought to induce infertility by impairment of germ-cell maturation, inflammatory occlusion of the ejaculatory ducts, induction of autoimmune phenomenon, and dysfunction of ejaculated sperm. Immotile cilia syndrome Sperm of normal number and morphology, but with no motility. Associated with bronchiectasis and sinopulmonary infections because of immotility of respiratory cilia. Kartagener’s syndrome is the association of immotile cilia syndrome and situs inversus. Immotile cilia syndrome is caused by deletion or shortening of dynein arms and absence of central microtubules. Autoimmunity Autoantibodies can be detected against the basement membrane of the seminiferous tubule or against the sperm in 5–10% of men evaluated for infertility. Environmental factors Lead, mercury, cadmium, dibromochlorpropane (DBCP), 1,1,12-trichloro-2,2-bis(p-chlorophenylethane) (DTT), p,p'-DDE Drugs and radiation Chemotherapeutic agents, anabolic steroids, cocaine, marijuana, ethanol, radiation >1.0 gy, 131I treatment >100 mCi, high scrotal temperature. Retrograde ejaculation Associated with diabetes, transurethral prostatectomy. Typically with an ejaculate of ANDROSTENEDIONE ↓ 17β-Hydroxysteroid dehydrogenase 17β-Hydroxysteroid dehydrogenase ↓ ANDROSTENEDIOL > TESTOSTERONE 3β-Hydroxysteroid dehydrogenase (3β-HSD)

lopian tubes, hypoplastic Wolffian derivatives, and cryptorchid testes. Heterozygous parents of affected siblings develop normally. Their condition is not unmasked by provocative testing. Diagnosis This disease should be suspected at birth in patients with a salt-losing crisis and the phenotype discussed above. Further aid to a correct diagnosis is obtained by showing the typical enlargement of the adrenals and testes, which is visible using magnetic resonance imaging (MRI) or computed tomographic scanning. Typically the serum levels of cortisol, aldosterone, and testosterone are undetectable adrenocorticotropic hormone (ACTH), FSH, and LH are elevated, and no response is evident after stimulation with ACTH or β-hCG. At puberty a minimal degree of virilization can occur in 46,XY individuals. Genetic Basis of Disease For a long time this disease has been ascribed to defects of the P450scc system, which consists of the cholesterol side-chain cleavage enzyme and two electron transfer proteins, termed adrenodoxin reductase (AR) and adrenodoxin (AD). This conclusion was based on hormonal studies of affected patients who were unable to produce adrenal and gonadal steroids, and on the inability of mitochondria from affected tissues to convert cholesterol to pregnenolone. However, a role for P450scc has been ruled out in molecular studies of affected individuals, in which abnormalities in the coding sequence of this gene have never been detected. In addition, P450scc, AR, and AD are normally transcribed and translated in the steroidogenic tissue of affected individuals, suggesting that the lesion responsible for lipoid CAH is located elsewhere. When the role of the StAR protein became clear, this factor appeared immediately to be the candidate gene responsible for lipoid CAH. Molecular analysis of the StAR protein, now performed in at least 22 families, has confirmed that this factor has a central role in the etiology of lipoid CAH. Mutations affecting both alleles of the StAR gene have been detected in 95% of the patients. Most families have homozygous mutations, and a minority appear to be compound heterozygotes. The importance of these mutations in the etiology of lipoid CAH has also been demonstrated by showing that StAR cDNAs carrying these mutations lack the ability to promote pregnenolone synthesis in transfected cells. Therapy Glucocorticoid and mineralocorticoid replacement is the first therapeutic step for these patients. Given the sexual phenotype, most patients are raised as females. Prophylactic orchidectomy followed by estrogen replacement should be performed to prevent the neoplastic degeneration of the retained gonads and to avoid any possible virilization at the time of puberty.

3β-HYDROXYSTEROID DEHYDROGENASE (3β-HSD) ∆ 5 ,∆ 4 -ISOMERASE DEFICIENCY Background 3β-HSD regulates the conversion of ∆ 5 -3β-hydroxysteroids (pregnenolone, 17-OH pregnenolone, dehydroepiandrosterone, and androstenediol) into the corresponding ∆4-3-ketosteroids (progesterone, 17-OH-progesterone, androstenedione, and testosterone, respectively). Absent or reduced activity of this enzyme is potentially associated with impaired formation of all steroid hormone classes, including progesterone, mineralocorticoids, glucocorticoids, androgens, and estrogens. Clinical Features There is a remarkable heterogeneity in the clinical presentation of 46,XY individuals affected by this disease. The two major clinical features of these patients are incomplete virilization and salt-wasting. The degree of salt-wasting is heterogeneous, and ranges from life threatening to clinically inapparent syndromes. In a similar fashion, the defects of virilization can range from the development of various genital ambiguities (including hypospadias, micropenis, and presence of a blind-ending vagina) to, in rare cases, the development of normal external genitalia. In most instances, the Wolffian duct derivatives develop normally, suggesting that although the prenatal testes were capable of some androgen production, these quantities were insufficient to permit normal development of the sexual phenotype. The Müllerian duct derivatives involute normally indicating normal production of AMH by the Sertoli cells. At puberty these patients have typical hypergonadotropic hypogonadism with high gonadotropin and low testosterone. These endocrine abnormalities are associated with gynecomastia, spermatogenic arrest, and phenotypic features of undervirilization. Interestingly, no correlation has been made between the degree of salt-wasting and of impairment of male sexual differentiation, suggesting that 3β-HSD mutations can affect the different steroidogenic pathways differentially. Diagnosis The diagnosis of this disorder should be suspected in patients presenting with adrenal and gonadal insufficiency. An elevated ratio of ∆5- to ∆4-steroids is characteristic of this disorder; however, the interpretation of the endocrine data can be complicated by the increased level of some ∆4-steroids, including serum 17-OH-progesterone and ∆ 4 -androstenedione, and urinary pregnanetriol. This is caused by the presence of two 3β-HSD enzymes, of which only one is mutated (see Genetic Basis of Disease). Genetic Basis of Disease This disease is transmitted as an autosomal recessive character, and the gene is located in the region

596

SECTION VII / ENDOCRINOLOGY

p11–p13 of chromosome 1. Two isoenzymes with 3β-HSD activity have been cloned, and have been designated type I and type II 3β-HSD. Their structural organization consists of four exons and three introns, which are included in a genomic DNA fragment of about 7.8 kb. The type I 3β-HSD mRNA is transcribed in the breast, placenta, and skin, and to a much lower degree in the gonads. The type II isoenzyme is transcribed exclusively in the adrenals, testis, and ovary. The identification of two isoenzymes with the same activity and with a different pattern of expression has the potential of explaining the presence of normal to elevated serum levels of some ∆4-3-ketosteroids in the patient population with 3β-HSD deficiency. Since the disease causes salt-wasting and ambiguous genitalia, it would be expected that the type II isoenzyme (which is exclusively transcribed in the adrenals and gonads) is the mutated gene and the type I isoenzyme (which is transcribed in the periphery) would maintain its activity and would be responsible for the residual conversion of ∆5-3β-hydroxysteroids to ∆4-3-ketosteroids. In agreement with this expectation, the molecular abnormalities causing clinical 3β-HSD deficiency have so far been localized exclusively to the type II isoenzyme. Numerous mutations in the coding region of the 3β-HSD gene have been identified by sequence analysis of SSCP (single-strand conformational polymorphism) variants detected in genomic DNA obtained from these patients. They include missense, nonsense, and frameshift mutations, and have been detected in homozygous or compound heterozygous patients. Their functional characterization has been very useful in defining the functionally important regions of the 3β-HSD gene. Molecular Pathophysiology There is a reproducible correlation between the impairment of the enzymatic activity caused by each mutation and the clinical abnormalities of the patients. This is particularly true with regard to the presence of salt-wasting forms of the disease in which no residual enzymatic activity is detected. Molecular analysis of the 3β-HSD gene of these patients has shown the presence of nonsense mutations (causing the synthesis of prematurely truncated proteins) and missense mutations (causing the replacement of critical amino acids) (Table 61-7). The only identified patient of this group with minimal residual enzymatic activity (0.1–0.3%) is a compound heterozygote carrying two missense mutations (Leu108Trp and Pro186Leu). The presence of such low level of activity is most likely inadequate to prevent the development of salt-wasting. In patients with non–salt-wasting forms of the disease, there is residual 3β-HSD activity that ranges from 1 to 10%, a level that appears to permit sufficient mineralocorticoid synthesis to prevent salt-wasting. Interestingly, in patients with this milder form of the disease premature stop codons have not been described, which completely abolish 3β-HSD activity, rather only missense or splicing mutations have been found (Table 61-8). Of note is the observation that the amino acid residues that are mutated are conserved in all members of the vertebrate 3β-HSD isoenzymes described to date. It is not yet clear why the clinical abnormalities of genital development and salt-wasting do not correlate. A potential explanation for this phenomenon is that different mutations affect C21 and C19 steroidogenesis in different ways. Future molecular investigations of more patients with 3β-HSD deficiency should help to clarify this point. Therapy Therapy involves replacement with glucocorticoids and, if required, mineralocorticoids. Sufficient androgens should be given early in life to correct the microphallus and allow surgical

Table 61-7 Mutations of the Type II 3β-HSD Gene Detected in 46,XY Patients with the Salt-Wasting Form of 3β-HSD Deficiency Origin

Karyotype

Mutation

Spain/Portugal

46,XY

United States

46,XY

Holland

46,XY

United States

46,XY

Afghanistan/Pakistana Algeria

46,XY 46,XY

Leu108Trp Pro186Leu Glu124Lys Trp171Stop 1-bp insertion at 186 Tyr253Asn Trp171Stop 1-bp insertion at 186 2-bp deletion at 273 Gly15Asp

aThis mutation was detected in three families originating from the same region in northwest Pakistan. Modified from Simard J, et al. J Steroid Biochem Mol Biol 1995; 53:127–138.

Table 61-8 Mutations Detected in the Type II 3β-HSD Gene in 46XY Patients with the Non–Salt-Wasting Losing Form of 3β-HSD Deficiency Origin

Karyotype

Mutation

Statesa

46,XY

Algeria United Statesa

46,XY 46,XY

Scotlandb Brazil

46,XY 46,XY

Gly129Arg G-A 6651 (new splice site) Asn100Ser Tyr254Asp No mutations in the second allele Leu173Arg Ala82Thr

United

aThis bThis

mutation was detected in two siblings of the same family. mutation was detected in two members of the same family. Modified from Simard J, et al. J Steroid Biochem Mol Biol 1995; 53:127–138.

treatment of hypospadias. Normal replacement doses should be given at the time of puberty. 17α-HYDROXYLASE/17,20-LYASE DEFICIENCY Background Two important steps of the steroidogenic pathway are regulated by the enzyme 17α-hydroxylase/17,20-lyase (P450c17). The first reaction consists of the 17α-hydroxylation of pregnenolone and progesterone into 17-hydroxypregnenolone and 17hydroxyprogesterone. These C21 steroids undergo cleavage of the C-17,20 carbon bond via the 17,20-lyase reaction to yield the C19 steroids dehydroepiandrosterone and androstenedione. To date, more than 120 cases with P450c17 deficiency have been described in the worlds literature. The large majority of these patients are affected by complete 17α-hydroxylase deficiency; however, about 20 cases with a partial form of this disease have been described. There is clear evidence that P450c17 catalyzes both the 17α-hydroxylase and 17,20-lyase reactions, but the description of 14 patients with only 17,20-lyase deficiency indicates that a complete understanding of this disorder has not yet been achieved. Clinical Features Of the enzymatic activities catalyzed by P450c17, 17α-hydroxylase is critical for the production of both glucocorticoid and sex steroids, whereas 17,20-lyase, which is located one step downstream in the steroidogenic pathway, is critical for the

CHAPTER 61 / TESTICULAR DISEASES

production of sex steroids only. In theory, defects of P450c17 activity could occur as isolated deficiencies of 17α-hydroxylase or 17,20-lyase or as combined 17α-hydroxylase/17,20-lyase deficiency. In practice it is almost impossible to differentiate between combined 17α-hydroxylase/17,20-lyase and isolated 17α-hydroxylase deficiencies on clinical and biochemical grounds. However, isolated 17,20-lyase deficiency has been described as an independent entity in 14 patients, and two types of enzymatic defects have been proposed for this syndrome. In the first one there is an incomplete defect of the ∆4 and ∆5 pathways, whereas in the second type, associated with normal serum levels of DHEA, there is a complete defect only in the ∆4 pathway. The deficient production of cortisol observed in these patients is associated with a compensatory overproduction of ACTH, which stimulates the production of large amounts of corticosterone, usually found at concentrations 50- to 100-fold higher than normal. Although corticosterone has a lower affinity for the glucocorticoid receptor (GR) than does cortisol, its activity is sufficient to prevent the development of overt adrenal insufficiency. As a result of the increased production of deoxycorticosterone (DOC), corticosterone, and 18-hydroxycorticosterone, affected patients develop retention of water and salt, hypokalemic alkalosis, and hypertension. Suppression of renin and aldosterone production is a feature of most patients; however, a few individuals with normal or elevated aldosterone have been reported in the Japanese literature, and this finding has still not been adequately explained. Deficient production of sex steroids is associated in these patients with hypogonadism and a compensatory overproduction of gonadotropins. Since normal levels of testosterone synthesis are necessary during fetal life for the development of the normal male sexual phenotype, affected patients manifest different abnormalities of the external genitalia, ranging from an apparent female phenotype with a blind-ending vagina to hypospadias and maleappearing genitalia with micropenis. The Wolffian derivatives can be hypoplastic or normal, the Müllerian derivatives are normally involuted, and the hypoplastic testes may be located intra-abdominally, in the inguinal canal, or in the scrotum. There is lack of correlation between the severity of hypertension and hypokalemia and that of hypogonadism, again suggesting that the various mutations of P450c17 can affect the C21 and C19 steroidogenic pathways in different ways. Diagnosis 17α-Hydroxylase/17,20-lyase deficiency should be considered in every 46,XY patient with a family history of pseudohermaphroditism that suggests an autosomal recessive inheritance. It is usually recognized in young adults with hypertension, hypokalemia, and ambiguous genitalia. The diagnosis of 17α-hydroxylase deficiency can be confirmed by demonstrating increased serum concentrations of pregnenolone, progesterone, corticosterone, 18-hydroxy-DOC, and DOC, associated with increased urinary excretion of their respective glucoronidate metabolites. Plasma renin activity, serum cortisol, and aldosterone are usually suppressed and ACTH is elevated. Affected males have low testosterone and elevated gonadotropins. The response to hGC and ACTH stimulation tests are impaired. This steroid pattern applies to patients with combined 17α-hydroxylase/17,20-lyase deficiency. In patients with 17,20-lyase deficiency the clinical symptoms are usually limited to the sexual phenotype. An important feature of this latter group of patients is the presence of normal serum level of cortisol and DOC and of their urinary metabolites. However, because of the presence of 17,20-lyase deficiency, these

597

patients have been found to have increased serum level of progesterone, 17OH-progesterone, and OH-pregnenolone, increased urinary level of pregnanetriolone, and decreased serum level of testosterone and its precursors. Genetic Basis of Disease The human P450c17 gene (CYP17) is a single copy gene, located on chromosome 10q24–25. It consists of eight exons and seven introns spanning 6569 bases. Fulllength cDNA clones have been isolated from adrenal and testicular sources and have the same nucleotide sequence. When the cDNA clones encoding this enzyme are expressed in cultured cells, the expressed protein displays both 17α-hydroxylase and 17,20-lyase activities. The molecular and functional analyses of 20 mutations of the CYP17 gene have been published (Table 61-9). They show that the genetic lesions of 17α-hydroxylase deficiency are random events, and that there are no structural features of the gene predisposing to specific abnormalities. The functional analyses of these mutations have shown that they impair both 17α-hydroxylase and 17,20-lyase activities, giving further demonstration that this gene encodes a protein possessing both enzymatic activities. Only a limited number of mutations have been reported more than once and were detected in unrelated individuals from Guam and of German-Dutch descent, indicating that this disorder is genetically heterogeneous, unless associated with a specific ethnic group. Molecular Pathophysiology Introduction The description of the molecular defects and functional abnormalities affecting patients with the complete and incomplete forms of 17α-hydroxylase deficiency and with isolated 17,20-lyase deficiency have elucidated several new aspects of this disease. The biochemical events associated with 17α-hydroxylase activity consist of a chain of events involving proper anchorage of the enzyme into the microsomal membrane, heme binding, substrate binding, transfer of electrons from NADPH to cytochrome P450 reductase, and O2 binding. Mutations specifically affecting heme and substrate binding have been identified. Targeted site-directed mutations of the rat CYP17 have been very informative in dissecting the amino acids of this protein that contribute to the 17α-hydroxylase and 17,20-lyase activities and showed the importance of residues Arg346 and Arg363 in inducing 17,20-lyase and 17α-hydroxylase activities, respectively. Complete Combined 17α-Hydroxylase and 17,20-Lyase Deficiency Fifteen mutations resulting in loss of 17α-hydroxylase and 17,20-lyase activity have been identified in the CYP17 gene. They include insertion of premature termination codons, amino acid replacements, or small deletions. Premature termination codons resulting from nonsense mutations or from other abnormalities of the open reading frame (including deletions or insertions of DNA) lead to truncated forms of the P450c17 protein, which, by being deprived of the C-terminal heme-binding domain, become functionally inactive. Complete loss of enzymatic function in three patients with missense mutations has identified residues Ser106, His373, and Arg440 as critical for both 17α-hydroxylase and 17,20-lyase activities. Of note is the speculation that the mutation occurring in amino acid residue 106 may affect substrate binding, whereas the two other mutations (His373 and Arg440) affect heme binding. Partial Combined 17 α -Hydroxylase and 17,20-Lyase Deficiency Partial deficiency of 17α-hydroxylase and 17,20-lyase has been studied in two patients. One of these two patients was a boy with ambiguous genitalia, carrying two different CYP17 mutant alleles: one resulting in an Arg239Stop nonsense mutation

598

SECTION VII / ENDOCRINOLOGY

Table 61-9 Mutations of the CYP17 Gene in 46,XY Patients with 17α-Hydroxylase/17,20-Lyase Deficiency Origin

Karyotype

Sex of Rearing

Mutation

Switzerland

46,XY 46,XY

F F

Italy Guamb Caucasian

46,XY 46,XY 46,XY

F F F

Canada

46,XY

M

4 bp duplication Gln461→Stop Arg496→Cys Deletion and insertion in exons 2 and 3 Ser106→Pro Tyr64→Ser Duplication of Ile112 Arg239→Stop Pro342→Thr

Canadaa

Netherlandsa

aThis mutation was described in two unrelated Canadians of German origin and in 6 unrelated families residing in the Friesland region of the Netherlands. bThis mutation was described in two unrelated individuals from Guam. Modified from Yanase T. J Steroid Biochem Mol Biol. 1995;53:153–157.

and the other causing a Pro342Thr missense mutation. The residual activity of these two alleles was respectively 0 and 20% of normal. These data, together with the functional analysis of a 46,XY female with homozygous premature termination codons (with residual functional activity of 0%) and of her normally virilized heterozygous father (thought to have a residual functional activity of 50%), have been very useful in understanding the amount of 17α-hydroxylase activity that is necessary for the development of the male sexual phenotype. Because the boy with ambiguous genitalia had a residual activity of 20% and the normally virilized individual described above was thought to have a residual activity of 50%, it would appear that the threshold activity necessary to change the sexual phenotype from female to ambiguous male is between 0 and 20%, and to change from ambiguous to normal is between 20 and 50%. Isolated 17,20-Lyase Deficiency Only one patient with this form of the disorder has been analyzed at the molecular level. This patient was found to be a compound heterozygote, with one allele containing an Arg496Cys missense mutation and the other allele containing a Glu461Stop nonsense mutation. In contrast to the clinical study, both alleles of this patient showed a dramatically reduced activity of both 17α-hydroxylase and 17,20-lyase. This finding prompted a clinical reevaluation of the patient, who was found to have deficient activity of both 17α-hydroxylase and 17,20-lyase when reexamined at age 25 years. These results led to speculation that the levels of P450c17 activity expressed in vivo were affected by unknown age-dependent factors or that a modifying effect had been exerted by the estrogen replacement that had been initiated in the interim. Treatment Treatment consists of replacement with cortisol, which accomplishes suppression of ACTH followed by the normalization of DOC, corticosterone, and 18-hydroxycorticosterone, and by the normalization of renin and aldosterone levels. Adult 46,XY individuals reared as females require estrogen replacement and removal of the abdominal gonads to prevent their malignant degeneration. Adult 46,XY individuals raised as males require androgen replacement, surgical correction of their external genitalia, and, if necessary, orchiopexy. 17β-HYDROXYSTEROID DEHYDROGENASE DEFICIENCY (17β-HSD DEFICIENCY) Background The enzyme 17β-hydroxysteroid dehydrogenase (17β-HSD) catalyzes the conversion of androstenedione and estrone into testosterone and estradiol, respectively. As such, it converts two weak precursors,

androstenedione and estrone, into two potent hormones, testosterone and estradiol, that, on interacting with high affinity with their receptors, induce their biologic effects. The reaction catalyzed by 17β-HSD is reversible. The reduction of the 17-keto group is believed to occur mainly in the testis and ovary, where it is required for the synthesis of testosterone and estradiol. The oxidation reaction, which is believed to occur in several peripheral tissues, is important for the inactivation of these two hormones. Five different isoenzymes with 17β-HSD activity have been isolated and designated 1 through 5 according to the chronological order of their identification. These isoenzymes have different biochemical features, and are expressed in different tissues (Table 61-10). Interestingly, the isoenzymes 1 and 3 favor the reduction reaction, whereas the type 2 and 4 isoenzymes favor the oxidation reaction. The isoenzymes 2 and 4 have ubiquitous tissue distribution, whereas the isoenzymes 1 and 3 are expressed in a more selective way. The mRNA of the type 1 isoenzyme is present in the ovary and placenta, whereas the type 3 mRNA has been detected almost exclusively in the testis. Information concerning the type 5 isoenzyme is at present less detailed. Mutations of the type 3 isoenzyme have recently been associated with 17β-HSD deficiency. Clinical Features Approximately 42 patients from 23 families with classic 17β-HSD deficiency have been reported. This number probably does not reflect the high incidence of this syndrome in the Gaza strip, where it is estimated that 1 in 100–150 individuals is affected by this disorder. The pattern of inheritance is autosomal recessive. At birth these patients present with female external genitalia, absent Müllerian and normal Wolffian duct derivatives, and undescended testes. Based on this phenotype, gender assignment is female in almost every case. At puberty, both virilization (developing a deep voice, large phallus, and various degree of body and facial hair) and feminization (developing variable degree of gynecomastia) occur. Some of these individuals have undergone a change in gender role behavior in parallel with the virilization occurring at the time of puberty. Impaired 17β-HSD activity explains the abnormal virilization of the external genitalia at birth, the elevated serum level of androstenedione and decreased testosterone, and the enhancement of pituitary gonadotropin production. A milder form of late-onset 17β-HSD deficiency has recently been identified in 3 adult patients with gynecomastia and hypogonadism in a study of 48 subjects with idiopathic pubertal gynecomastia. The frequency of this syndrome is unknown. It is

599

CHAPTER 61 / TESTICULAR DISEASES

Table 61-10 Features of Five Isoenzymes with 17β-HSD Activity 17β-HSD deficiency

Amino acid size

Exons

Normal Normal

327 387

6 5

17q21 16q24

Mutated

310 736 323

11

9q22

9

Chromosome localization

Tissue localization

Cellular localization

Ovary, placenta Endometrium, placenta, liver Testis Ubiquitous

Cytosol Microsomes Microsomes Peroxisomes

10p14–15

Preferred substrate Estrogens Androgens, estrogens, progestogens Androgens, estrogens Estrogens

Preferred cofactor

Catalytic preference

NADPH NAD+

Reduction Oxidation

NADPH NAD+

Reduction Oxidation Reduction

Modified from Andersson S, et al. Trends Endocrinol Metab 1996;7:121–126.

equally unclear whether these patients are heterozygous for an inactivating mutation of type 3 17β-HSD or whether they are homozygous for a mutation of type 3 17β-HSD causing a milder enzymatic defect. Endocrine Features and Diagnosis The diagnosis should be considered in 46,XY patients presenting with pseudohermaphroditism inherited with an autosomal recessive pattern. Plasma androstenedione, estrone, and gonadotropin levels are elevated, whereas testosterone and dihydrotestosterone are decreased or in the low normal range. Steroid hormone metabolism has been studied in the testes of several affected subjects. In these tissues it was found that the conversion of androstenedione to testosterone was consistently abnormal. However, other authors have found that the oxidative reaction catalyzed by 17β-HSD was normal in cultured fibroblasts obtained from one patient. Most of these patients are diagnosed at puberty, when they are referred for failure to menstruate, or because they develop a mixed pattern of virilization and feminization. The diagnosis of 17β-HSD deficiency is more challenging in the newborn. However, the correct interpretation of the endocrine tests should permit the diagnosis of this disorder in any age group. Genetic Basis of Disease Because of its biochemical characteristics and tissue distribution, type 3 17β-HSD was considered to be a good candidate to cause the clinical abnormalities of 17β-HSD deficiency. Subsequent analysis has identified 14 mutations of this gene in 17 families with the typical phenotypic, endocrine, and genetic features of the syndrome. The mutations detected include 10 missense mutations, 3 splice junction abnormalities and 1 small deletion resulting in a frameshift. Patients with the syndrome were most frequently homozygous (12), but heterozygotes (1) and compound heterozygotes (4) have been described as well (Table 61-11). The mutation Arg80Gln has been detected in three unrelated families, including a member of the large kindred from the Gaza strip and two Brazilian families. Considering the high incidence of 17β-HSD deficiency in the Gaza strip, screening programs for this mutation may be worthwhile in that part of the world to detect heterozygosity or for prenatal diagnosis. The real frequency of this mutation in the entire Arab population of the Gaza strip could be very high, if one considers that homozygous females are asymptomatic, have normal internal and external genitalia, undergo normal sexual development, and have unimpaired fertility. The frequency of compound heterozygosity underscores the possible high frequency of heterozygous carriers, a potentially important fact, considering the recent identification of individuals with partial deficiency of testicular 17β-HSD activity that was reported in a group of patients with gynecomastia.

Molecular Pathophysiology of Disease Nine of the ten substitution mutations described in Table 61-11 have been recreated in vitro, and the activity of the mutated enzyme was studied in transfected cells. Eight of the nine mutations are completely inactivating. However, in agreement with biochemical studies performed in cultured genital skin fibroblasts, the ninth mutation (Arg80Gln) retains a small amount of activity. Detailed analysis of the enzymatic activity of this mutant has revealed a 100-fold decreased affinity for the cofactor NADPH, localizing at least a portion of the NADPH-binding domain to the region surrounding residue 80. These recent developments in the molecular biology of 17βHSD deficiency have permitted understanding of why these patients undergo virilization at the time of puberty. The consensus is that androstenedione, which is produced by the testes in supraphysiologic concentrations at the time of puberty, provides the substrate for the extraglandular production of testosterone by one of the four 17β-HSD isoenzymes that are not impaired in this disorder. However, two clinical features of 17β-HSD deficiency await adequate explanation. It is not clear why inadequate virilization is more complete during embryonic development, when the external genitalia do not virilize at all, than at the time of puberty, when a substantial masculinization of the phenotype occurs. In addition, it is not clear why the embryonic structures of the developing male genitalia respond to the hormonal abnormalities created by 17β-HSD deficiency in a different way. In this disorder the Wolffian derivatives undergo almost normal virilization during embryogenesis, whereas the external genitalia do not virilize at all. Further studies of subjects affected by 17β-HSD deficiency should help to answer these important biologic questions. Treatment Patients are usually raised as females and treatment consists of gonadectomy followed by estrogen replacement at the time of expected puberty. When the diagnosis is correctly identified in early infancy and gender reassignment is possible, the patients should receive testosterone treatment at pediatric doses and genitoplasty early in infancy and be replaced with adult doses of testosterone at the time of expected puberty. The basic enzymatic defect persists, and impaired spermatogenesis is present during adulthood. TYPE 2 5α-REDUCTASE DEFICIENCY The enzyme 5αreductase is involved in the conversion of testosterone into the powerful 5α-reduced metabolite dihydrotestosterone (DHT). Two isoenzymes have been isolated that share this enzymatic activity, and have been designated types 1 and 2 5α-reductase. The clinical syndrome is associated with mutations of the type 2 isoenzyme.

600

SECTION VII / ENDOCRINOLOGY

Table 61-11 Summary of the Mutations Detected in Type 3 17β-HSD Ethnic background

Family history

Homozygous patients Lebanese Brazilian Syrian Greek American American German German Iranian Polish Brazilian Brazilian Compound heterozygous patients American

Comment

Yes Yes No Yes No No No No No No No Yes

Arg80Gln Arg80Gln 655-1, G-A 655-1, G-A 325+4, A-T 325+4, A-T 325+4, A-T Phe208Ile Ser65Leu ∆ 777–783 Ala203Val Glu215Asp

Residual enzymatic activity Residual enzymatic activity Disrupt splice acceptor Disrupt splice acceptor Disrupt splice donor Disrupt splice donor Disrupt splice donor Inactivates enzyme Inactivates enzyme Frame shift truncates protein Inactivates enzyme Inactivates enzyme

No

Ser232Leu Met235Val 325+4, A-T Pro282Leu Arg80Gln 326-1, G-T 325+4, A-T Gln176Pro

Both inactivate enzyme

American

Yes

Brazilian

Yes

Italian German-Irish Heterozygous patients American

Mutation

No

No

Val205Glu

Disruption of splice donor; inactivates enzyme Residual enzyme activity; disrupts splice acceptor Disrupt splice donor

Inactivates enzyme

Modified from Andersson S, et al. J Clin Endocrinol Metab 1996;81:130–136.

ACQUIRED CASES OF HYPERGONADOTROPIC HYPOGONADISM None of these conditions of primary hypogonadism has a genetic basis. They are listed according to their etiology in Table 61-12.

HYPOGONADOTROPIC DISORDERS Several congenital and acquired disorders (listed in Table 61-2) account for HH. Their common denominator consists of the impaired production of LHRH or of gonadotropins by the hypothalamus or pituitary gland. This in turn results in failure to stimulate the correct production of gonadal steroids and in a spectrum of clinical manifestations that vary depending on whether the abnormality developed before birth, before puberty, or after puberty (Table 61-4).

HYPOGONADOTROPIC DISORDERS ASSOCIATED WITH ISOLATED GONADOTROPIN DEFICIENCY KALLMANN’S SYNDROME Background The first description of Kallmann’s syndrome (KS) dates back to 1856 by Maestre de San Juan, but it was not until the original description by Kallmann in 1944 that the disease was recognized as an inherited entity. It is found in 1 of every 10,000 newborn males, and in 1 of every 50,000 newborn females, and occurs in both sporadic (in up to 50% of the cases) and familial forms. Because of its predominance in the male sex, it was initially thought that KS is an X-linked disorder. However, there is now convincing evidence that additional forms exist that demonstrate autosomal recessive and autosomal dominant forms of transmission. In agreement with this heterogeneous inheritance, chromosomal abnormalities have been

detected both in the X chromosome and, in patients with the autosomally inherited form, in autosomes. Within the X chromosome, the KS gene has been located in the Xp22.3 region. At least five different syndromes are caused by abnormalities involving this region of the X chromosome, including short stature, chondrodysplasia punctata, mental retardation, ichthyosis, and KS. The location of the gene(s) responsible for the autosomally transmitted KS is not yet known. Embryology of the LH-RH Neurons The molecular events associated with the migration of the LHRH-secreting neurons during embryogenesis are intimately connected with the pathogenesis of KS. These neurons originate in the epithelium of the olfactory placode, from which they migrate into the brain by the 12th week of gestation along the pathway of the developing olfactory tract, represented by the olfactory, terminalis, and vomeronasal nerves. The LHRH neurons enter the brain through the nervus terminalis and vomeronasal, and thence into the septal-preoptic area and the hypothalamus. The normal migration of the LHRH neurons depends on genes located in Xp22.3, the same area of the X chromosome that is deleted in patients with KS. This was demonstrated by Schwanzel-Fukuda in a fetus with KS and a deletion of Xp22.3, in whom LHRH-expressing neurons were not found in the hypothalamus, but were instead identified beneath the forebrain, within the dural layers of the meninges and on the cribriform plate. When the migration is completed, there are fewer LHRH neurons in the brain (approximately 1500), than there were in the nose at the beginning of this process, probably because during the migration many cells die or differentiate into other cell types. Only after reaching their final destination in the brain do the LHRHcontaining neurons acquire the capacity to secrete LHRH.

CHAPTER 61 / TESTICULAR DISEASES

601

Table 61-12 Acquired Causes of Hypergonadotropic Hypogonadism Viral orchitis Sixty percent of postpubertal individuals with mumps orchitis exhibit damage to the germinal epithelium or both the germinal and interstitial epithelia Drugs causing hypergonadotropic hypogonadism grouped according to their mechanism of action Competition with the androgen receptor Spironolactone Cimetidine Flutamide Nilutamide Bicalutamide Cyproterone Direct toxic effect on Leydig cells Antineoplastic drugs Cyclophosphamide MOPP combination chemotherapy (mechlorethamine, vincristine, procarbazine, and prednisone) ABVD combination chemotherapy (doxorubicin, bleomycin, vinblastin, and dacarbazine) Ethanol Inhibitors of testosterone synthesis Ketoconazole Spironolactone Cyproterone Tetracycline Ethanol Stimulators of estradiol synthesis or activity Estrogenic compounds Digoxin Spironolactone Environmental toxins Lead DDT (1,11,1-trichloro-2,2-bis(p-chlorophenyl)ethane) p,p'DDE (1,1-dichloro-2,2-bis(p-chlorophenyl)ethylene) Radiation 15 rads: transient germinal epithelium damage causing ↑ FSH >600 rads: permanent germinal damage 2000–3000 rads: permanent germinal and interstitial damage Autoimmunity As part of the autoimmune polyglandular syndrome type II Trauma Frequent acquired cause of hypogonadotropic hypergonadism because of the relatively unprotected intrascrotal position of the testes

Clinical Features Individuals affected by KS can sometimes be recognized before puberty because of the presence of microphallus or cryptorchidism. Most patients, however, are recognized at puberty when they present with delayed appearance of the secondary sex characteristics or for the development of eunuchoid features. Abnormalities observed in KS other than those related to hypogonadism include hyposmia, anosmia, cleft lip, or cleft palate. Other symptoms occasionally described with the disease are sensory neural deafness, ocular motor abnormalities, abnormal spatial visual attention, cerebellar dysfunctions, talipes cavus deformity, and unilateral renal aplasia. It should be kept in mind that anosmia, which is considered one of the cardinal symptoms of KS, is not present in every patient, and that in some large series this feature is absent in up to 50–60% of the cases. In addition, some KS patients also manifest other diseases, like chondrodysplasia punctata, mental retardation, ichthyosis, short stature, and ocular albinism, which are associated with the deletion of the distal short arm of the X chromosome.

The hypogonadism of KS is caused by reduced secretion of LHRH by the hypothalamus, as shown by the observation that in almost every KS individual there is adequate response of LH and FSH after sufficient priming of the pituitary with exogenous LHRH. The decreased secretion of LHRH results in an abnormal secretory pattern of LH pulses, ranging from decreased or absent amplitude to diminished frequency. In the complete form of LHRH deficiency, LH and FSH deficiency is severe, and no evidence of sexual maturation is evident. In the incomplete form (known as the fertile eunuch syndrome), there is either normal FSH and low LH, or partial defects in both, with some degree of germ cell maturation. Although it is widely believed that this syndrome represents a continuum with the complete form of KS, the genetic studies to establish whether the fertile eunuch syndrome is a variable manifestation of KS or a separate genetic entity have not been performed. Microphallus and olfactory disturbances appear to be less common in this variant. Another variant form of isolated gonadotropin deficiency has been described in men with isolated FSH deficiency.

602

SECTION VII / ENDOCRINOLOGY

Laboratory tests reveal low serum FSH and LH levels and low testosterone. However, a large degree of heterogeneity in the serum levels of FSH and LH can be observed. Diagnosis The diagnosis is suggested by the typical clinical picture and laboratory tests that demonstrate hypogonadotropic hypogonadism. The most challenging differential diagnosis is with delayed puberty. Important criteria are the ability to show the presence of anosmia or hyposmia, any other associated congenital defect or positive family history for KS. However, the separation of these two entities may require a prolonged period of observation. Genetic Basis of Disease Using positional cloning techniques, two independent groups have isolated a candidate gene for KS (KAL) that is located at the Xp22.3 interval. Evidence that this gene is responsible for KS comes from the observation that in some affected patients it contains extensive deletions, or point mutations. The presence of a highly conserved KAL homolog in the Y chromosome has somehow delayed the development of mutation scanning strategies; however, point mutations within the 14 exons of the gene have been detected in at least 14 patients. Molecular Pathophysiology Sequence analysis of the KAL cDNA has provided interesting insights into the pathogenesis of Kallmann’s syndrome. The predicted protein is 680 amino acids long. Owing to the lack of a transmembrane domain or a phosphatidyl inositol anchorage site, it is believed to be an extracellular matrix molecule. Search for homologies has shown the presence of two important motifs. The first is a 4-disulfide core domain, found in a number of proteins with protease inhibitory activity, that is located in the N-terminal part of the molecule. The second is a region of similarity with the fibronectin type III repeat, found in numerous adhesion molecules involved in axon to axon interaction or in neuronal migration. Based on these sequence homologies, the KAL protein may represent an extracellular matrix factor that possesses both antiprotease and adhesion (or antiadhesion) functions. This protein may play a critical role in the migration process involving the olfactory and LHRH neurons during embryogenesis. Its absence or abnormal function in Kallmann’s syndrome may help to explain the double clinical defect (hypogonadism, anosmia) observed in many of these patients. In addition, abnormalities of the KAL protein could also explain some of the other, less frequent, manifestations of the disease. RNA in situ hybridization studies have demonstrated KAL mRNA expression in the Purkinje cells of the cerebellum, the nucleus of the oculomotor nerve, the mesonephros, and the facial mesenchyme, which correlates with the presence in some KS patients of cerebellar dysfunctions, eye movement defects, unilateral renal aplasia, and cleft palate. Despite these important new developments, there are still a number of unanswered questions. The KAL gene does not explain the autosomal types of KS, which is probably caused by abnormalities of autosomal genes that influence the migration of the LHRH neurons to the hypothalamus. In addition, it is not clear what the defect is in patients with hypogonadotropic hypogonadism without anosmia or in patients affected by the sporadic form of KS. If the sense of smell of patients without anosmia is normally developed, one can hypothesize that their olfactory bulb has developed normally and that their KAL gene may be normal, or that they represent a group with variable penetrance of the LHRH migratory defect, with mutations that do not completely impair function of the KAL gene product. In patients affected by the sporadic form of KS there is no supportive family history of an X-linked trait. Initial reports in abstract form suggest that the KAL

gene is mutated only in a minority of these patients. Therefore the sporadic form of KS seems to be a heterogeneous disorder, arising in part from de novo mutations of the KAL gene, and in part from mutations of an unidentified autosome.

HYPOGONADOTROPIC HYPOGONADISM FROM BIOLOGICALLY INACTIVE LH HYPOGONADISM CAUSED BY MUTATIONS IN THE GENES OF THE HPG AXIS Abnormalities occurring in the sequences of the genes encoding LHRH, the LHRH receptor, LH, or FSH have the potential of explaining some cases of idiopathic hypogonadotropic hypogonadism (HH) or infertility in males. However only a few such patients have been identified. It is likely that heterozygous patients with these mutations are subfertile; therefore, one can speculate that patients carrying homozygous mutations in the genes of the HPG axis are very rare because they would be the product of two subfertile individuals. Evidence supporting the possibility that mutations of the LHRH gene are associated with hypogonadotropic hypogonadism is illustrated by the hpg mouse, an animal model with hypogonadism and infertility as a result of a deletion of the LHRH gene. Two studies have investigated the sequence of gonadotropin-releasing hormone (GnRH) in patients with idiopathic HH. Weiss et al. studied the coding region, the promoter, and the 3' untranslated tract of the GnRH gene in three male patients. No mutations were detected in this study. In another investigation reported in abstract form, Layman and collaborators identified a homozygous A-to-G transition 85 bp downstream from the splice donor site of intron 2 in 1 of 117 patients with HH. These authors speculated that this mutation could create a potential alternate donor splice site or affect transcription of the GnRH gene; however, no data have been presented in this regard. Mutations of the FSH gene have not been reported in men, but are known to account for hereditary hypergonadotropic ovarian failure in women. Individuals of male sex with mutations of the FSH receptor have been identified in these pedigrees. These individuals are predicted to have a spectrum of abnormalities ranging from azoospermia to normospermia; however, the characteristics of their sexual phenotype have not yet been described in detail. The FSH receptor can also undergo activating mutations, as recently described by Gromoll and collaborators in a man who was able to sustain spermatogenesis despite being hypophysectomized. HYPOGONADOTROPIC HYPOGONADISM CAUSED BY MUTATIONS OF THE LH GENE In males, only one case of hypogonadism with infertility has been associated with mutations of the LH gene. The propositus of this study was a 17-year-old patient with a hormonal profile showing low testosterone, high immunoreactive LH, and normal FSH. A testicular biopsy specimen revealed arrest of spermatogenesis and the absence of Leydig cells. When this patient was treated with exogenous hCG, a normal increase in testosterone secretion occurred, which was in contrast with the initial diagnosis of primary hypogonadism. Subsequently, the LH of this patient was found to be devoid of biologic activity using a dispersed Leydig cell bioassay, suggesting the production of an immunologically active but biologically inactive LH molecule. Sequence analysis of the coding sequence of the LH β gene revealed the patient to be homozygous for a missense mutation, Gln to Arg at amino acid residue 54. Functional experiments have shown that this mutation impairs the ability of the resulting LH to bind and activate its receptor. This study was extended to other

CHAPTER 61 / TESTICULAR DISEASES

603

Table 61-13 Syndromes Associated With Central Nervous System Disorders Clinical and genetic features Prader-Willi syndrome Möbious syndrome Laurence-Moon-Biedl syndrome Börjeson-Forssman-Lehman syndrome LEOPARD syndrome Rud’s syndrome Lowe’s syndrome Carpenter’s syndrome

See text See text See text X-linked mental retardation (XLMR) syndrome, mapped to Xq26–27. Associated with supraorbital ridges, ptosis, hypotonia, mental retardation, hypogonadism Lentigenes, ECG conduction defects, ocular hypertelorism, pulmonic stenosis, retarded growth, deafness, hypogonadism Mental retardation, epilepsy, hypogonadism, ichthyosis XLMR syndrome mapped to Xq26. Associated with cataracts, renal tubular acidosis, hypogonadism, hypotonia, mental retardation Obesity, acrocephaly, craniosynostosis, agenesis of the hands and feet

members of the family, including three maternal uncles and the mother of the propositus, and has elucidated how heterozygous carriers of both sexes manifest clinically. The mother of the patient, an obligate heterozygote, underwent normal puberty and was fertile. The three heterozygous maternal uncles underwent normal puberty, reported normal libido and sexual performance, and had a normal physical examination. However, their serum testosterone concentration was low in several measurements, and at the time of the study they were childless. Therefore, given the limited nature of the pedigree, it would appear that heterozygous mutations of the LH gene do not manifest with important clinical abnormalities in the female sex, but may be associated with infertility in males. Although mutations of the LH gene causing hypogonadism and infertility have only been detected in one male patient, the description of this case report has permitted a better understanding of the physiologic role played by LH during the embryologic development of the male sexual phenotype. The proband of this study was born with normal masculinization of his genitalia and with descended testes. Thus one can argue that LH is not critical for normal male sexual development, and that during embryogenesis adequate amounts of androgens from the adrenals or the testes are produced, probably under the stimulation of placental hCG.

HYPOGONADOTROPIC HYPOGONADISM ASSOCIATED WITH CENTRAL NERVOUS SYSTEM DISORDERS Several congenital syndromes have been described in which central nervous system (CNS) disorders are associated with hypogonadotropic hypogonadism (Table 61-13). One hypothesis is that these disorders may be the consequence of a spectrum of congenital abnormalities disrupting at the same time the neuronal pathways leading the LHRH neurons into the hypothalamus and the normal development of the CNS. The identities and function of the gene(s) responsible for this group of syndromes are unknown. PRADER-WILLI SYNDROME (PWS) Introduction PWS is an unusual condition occurring in 1 of every 25,000 newborns that is associated in about 50% of the cases with an interstitial deletion of chromosome 15, involving 15q11–q13. It is generally sporadic, although 10 families with more than one case have been described. The origin of the deleted chromosome 15 is paternal in almost all the individuals investigated by cytogenetic and DNA studies, making PWS one of the best examples for parental imprinting in humans.

Clinical Features The PWS phenotype is characterized by childhood obesity, carbohydrate intolerance, short stature, mental deficiency, infantile hypotonia, small hands and feet, a characteristic face, and hypogonadism. The latter is present in 95% of the cases, is mostly caused by GnRH deficiency, and is associated with cryptorchidism, micropenis, and scrotal hypoplasia. The response of serum gonadotropin concentrations to a single GnRH bolus is usually attenuated; however, long-term treatment with clomiphene or GnRH may stimulate gonadotropin secretion. Genetic Basis of Disease The large size of the 15q11–q13 deletion has hampered the identification of candidate PWS genes in this region; however, the identification of two atypical PWS deletions, which greatly reduces the common region of deletion, has helped in focusing the quest for candidate genes to a smaller portion of 15q11–q13. As a result, three genes displaying paternal allele-specific expression have been recently identified in the area commonly deleted in PWS. One of these genes is the small nuclear ribonucleoprotein-associated polypeptide N gene (SNRPN), a molecule involved in mRNA splicing, that is transcribed only from the paternal allele. The two other candidate genes are PAR-5 and PAR-1, which were isolated from a YAC contig of the region deleted in PWS, and cluster together with SNRPN to a roughly 250-kb subregion of 15q11–q13. Molecular Pathophysiology of Disease Whether any of these three genes is directly involved in the phenotype of PWS is currently under investigation. Considering the role played by SNRPN in mRNA splicing and the complex phenotype of PWS, one is tempted to speculate that a malfunctioning SNRPN contributes to PWS by altering the expression of several different and unrelated genes. This theory awaits confirmation in the years to come. LAURENCE-MOON-BIEDL SYNDROME This entity has traditionally been considered as the association of retinitis pigmentosa, obesity, mental retardation, polydactyly, spastic paraparesis and hypogonadism. Currently two syndromes with slightly different manifestations are recognized: the Laurence-Moon syndrome, in which spastic paraparesis dominates and polydactyly is very rare; and the Bardet-Biedl syndrome, with very rare neurologic complications and frequent occurrence of dystrophic extremities and renal disease. Hypogonadism is present in both entities, and it manifests with microphallus, hypospadias, and undescended testes in prepubertal boys. There is disagreement on the origin of testicular failure as cases of both primary and secondary hypogonadism have been reported in the literature. The dis-

604

SECTION VII / ENDOCRINOLOGY

Table 61-14 Functional Causes and Systemic Diseases Associated with Hypogonadotropic Hypogonadism Functional causes Hyperprolactinemia: Patients affected by hyperprolactinemia can develop HH because of either the direct effect of hyperprolactinemia, which impairs the hypothalamic release of LHRH, or the mass effect of the adenoma. Cushing’s syndrome: Increased cortisol impairs normal production of LHRH. Exercise: In men strenuous exercise can associate with decreased testosterone levels, without any change in LH and Testesterone Binding Globulin (TeBG) production. Usually without clinical signs. Melatonin: One case of HH with increased melatonin production has been described, in which spontaneous normalization of melatonin hypersecretion was associated with reversal of HH. Systemic diseases AIDS: Frequent association between AIDS and HH (observed in 30% of HIV-positive patients in one study). In addition, preliminary data suggest that hypogonadism is probably associated with the wasting syndrome frequently observed in these patients. Obesity: associated with ↓ TeBG, ↓ total testosterone, normal free testosterone, and ↑ estradiol (the latter because of increased peripheral conversion of aromatizable compounds). Liver diseases: Multifactorial hypogonadism, with primary and secondary components, characterized by ↓ testosterone, ↑ estradiol, and gonadotropin levels from normal to moderately elevated. Renal failure: Multifactorial hypogonadism with primary and secondary components. Hyperprolactinemia, a cause of HH, is present in 25% of men with chronic renal insufficiency undergoing dialysis. Hemochromatosis: Multifactorial hypogonadism, with primary and secondary components. The secondary form of hypogonadism is predominant and it is caused by accumulation of iron in the pituitary gland. It is associated with diabetes, liver cirrhosis, and congestive heart failure.

ease is transmitted in an autosomal recessive manner, as indicated by a consanguinity rate of 48% among parents of affected patients in one series, and by a male to female ratio of 47:41 by combining two large studies. MÖBIOUS SYNDROME This entity is also known as congenital oculofacial paralysis and is associated with multiple cranial nerve paralyses involving the 3rd, 4th, 5th, 9th, 10th, and 12th nerves, mental retardation, gait disturbances, peripheral neuropathy, and hypogonadism. Gonadotropin deficiency has been documented in several patients; however, considering the normal gonadotropin response to GnRH, GnRH deficiency is likely to constitute the primary defect. It is transmitted as an autosomal dominant trait, although sporadic cases have also been described. The clinical characteristics of the remaining syndromes belonging to this group of diseases are summarized in Table 61-13.

HYPOGONADOTROPIC HYPOGONADISM ASSOCIATED WITH ADRENAL INSUFFICIENCY ADRENAL HYPOPLASIA CONGENITA Background Adrenal hypoplasia congenita (AHC) is a rare congenital disorder in which hypogonadotropic hypogonadism is associated with adrenal insufficiency. Two forms of this disorder have been described, the miniature adult form, secondary to ACTH deficiency transmitted as an autosomal recessive character, and the X-linked primary, or cytomegalic form. Clinical Features ACH usually manifests with symptoms of adrenal insufficiency early in infancy. Gonadotropin deficiency is commonly associated with the X-linked primary form of AHC, and is noted at the expected time of pubertal maturation. The nature of the hypogonadotropic hypogonadism reported in these patients may be caused by pituitary gland dysfunction, based on lack of gonadotropin responses to prolonged pulsatile stimulation with synthetic GnRH. Genetic Basis of Disease The gene for AHC has been mapped to chromosome Xp21 based on the deletion of this region in patients with a contiguous gene syndrome that includes AHC, glycerol kinase deficiency, and Duchenne muscular dystrophy.

The gene responsible for this syndrome, designated DAX-1, has recently been isolated. It is a new member of the nuclear hormone receptor superfamily, and has been found to be deleted in some AHC patients or to have point mutations in patients with AHC, but no evidence of a deletion. The exact role played by this gene in the pathogenesis of the hypogonadotropic hypogonadism observed in AHC is currently under investigation.

ACQUIRED FORMS OF GONADOTROPIN DEFICIENCY FROM FUNCTIONAL CAUSES OR SYSTEMIC DISEASES A brief discussion of these diseases, none of which has a genetic etiology, is presented in Table 61-14.

ANATOMIC DISORDERS ASSOCIATED WITH HYPOGONADOTROPIC HYPOGONADISM A large number of anatomic disorders are associated with HH (Table 61-2). Infiltrations of the pituitary by inflammatory lesions, as in the case of lymphocytic hypophysitis, or by granulomatous diseases, such as in sarcoidosis, tuberculosis, and histiocytosis X, are well-described causes of gonadotropin deficiency. Other disorders consist of tumors developing in the hypothalamic–pituitary region. They act by the local compression and destruction of the gonadotroph. They include nonsecretory or secretory pituitary adenomas, craniopharyngiomas, dysgerminomas, metastases, meningiomas, hamartomas, or hypothalamic tumors. Panhypopituitarism can also be the consequence of pituitary apoplexy. This condition has been described in patients affected by pituitary tumors or diabetes mellitus, in patients harboring lesions producing increased intracranial pressure, and in the setting of radiotherapy and anticoagulation. Residual LH and FSH deficiency after pituitary apoplexy has been described in 76 and 58% of the patients, respectively, and diminished testosterone secretion in 85% of the subjects investigated in a large study. The mechanism is believed to reflect the direct destruction of the adenohypophysis by the apoplexy.

605

CHAPTER 61 / TESTICULAR DISEASES

Table 61-15 Systemic Disorders Associated With Infertility Diseases impairing fertility Infections Gonorrhea Syphilis Tuberculosis Infectious parotitis Chlamydial epididymitis Brucellosis Filariasis Typhoid Endocrine Disorders (other than hypogonadism) Thyrotoxicosis Diabetes Systemic illnesses Renal failure Hepatic failure Sickle cell anemia Neurologic diseases Paraplegia Chronic respiratory tract diseases Bronchiectasis Chronic sinusitis Chronic bronchitis

Mechanism Obstruction of the seminal pathway Orchitis Obstruction of the seminal pathway Orchitis Obstruction of the seminal pathway Orchitis Obstruction of the seminal pathway Orchitis Increased estrogen level Impotence and ejaculatory disturbances Hypogonadism Hypogonadism Impaired sexual maturation Impotence, ejaculatory dysfunctions, decreased testosterone level Associated with immotile cilia syndrome or with Young’s syndrome

GERMINAL CELL FAILURE INTRODUCTION Infertility is usually defined as lack of conception after 1 year of adequate and unprotected intercourse. In many Western countries, approximately 15% of couples are unable to conceive within 12 months. Female and male factor disorders are equally distributed, and each are present in about 50% of the cases. Significant abnormalities are found in the male alone in one-third of the cases, while in an additional 20% of cases abnormalities are found in both members of the couple. Therefore, given the high incidence of concomitant disease in both partners, basic evaluation is warranted in both the man and the woman. In the male partner it is useful to proceed in a stepwise way, starting with (1) a careful medical history, which is then followed by (2) a physical exam, (3) a semen analysis, and, if useful, (4) a testicular biopsy. By questioning the patient, one should always keep in mind that a group of treatable causes of infertility are associated with decreased testosterone production as a result of a hypothalamic or pituitary cause. To ascertain this the patient should be asked whether he underwent normal puberty, and whether he is affected by symptoms of impotence or decreased libido. If this is the case, an appropriate work-up, including measurement of LH, FSH, testosterone, and prolactin and imaging of the pituitary, is warranted. In addition to hypogonadotropic disorders, numerous systemic disorders should be considered a possible secondary cause of male infertility (Table 61-15). Another possible cause of infertility is caused by a direct anatomic damage to the testis or seminal pathways. Consequently it is important to investigate whether the patient has a history of testicular trauma or torsion or of genitourinary or testicular infections, and whether he underwent genitourinary surgery. Finally, it is important to ascertain whether the patient has been taking medications or illicit drugs or has been exposed to environmental factors known to affect normal gonadal function (Table 61-1).

It is useful to classify disorders of the germinal compartment of the testes according to the serum FSH level. Based on this, male infertility can be of the hypogonadotropic, hypergonadotropic, or eugonadotropic type. Hypogonadotropic disorders fall in the group of diseases discussed in Table 61-2. Patients with a hypogonadotropic disorder are affected by the simultaneous failure of the interstitial and germinal cells and belong to a group of infertile patients for which successful treatments are available. Hypergonadotropic and eugonadotropic disorders represent categories of infertility that are characterized by increased or normal levels, respectively, of FSH (Table 61-3). In such states, the interstitial compartment is usually not affected by the disease, and therefore infertility will be the only symptom. Although a large number of conditions have been associated with male infertility, specific etiologies have been defined for only a small percentage of the conditions recognized. In many series, 60–80% of the infertile population consists of individuals with varicocele or idiopathic disease (Table 61-16). Since the exact mechanism causing infertility in men with varicocele is not understood, one can conclude that the etiology of infertility has not been defined in as many as 80% of patients. It is reasonable to assume that a better understanding of the basic biology of male infertility will permit a more successful treatment of these patients in the future.

EUGONADOTROPIC GERMINAL CELL DYSFUNCTION OBSTRUCTION OF THE SEMINAL PATHWAY Introduction Acquired or congenital obstruction of the epididymis, vas deferens, or ejaculatory duct leads to azoospermia in up to 6% of infertile men. Up to 40% of men with azoospermia are eventually found to have an obstruction at some level of the seminal pathways, which is typically associated with normal testicular development, normal FSH, and normal histology of the testicular biopsy specimen. A classification based on the findings at explor-

606

SECTION VII / ENDOCRINOLOGY

Table 61-16 Relative Frequency of Diseases Associated with Male Infertility Idiopathic Varicocele Cryptorchidism Epididymal or vas deferens obstruction Klinefelter’s syndrome Mumps orchitis Hypogonadotropic hypogonadism Nonmotile sperm Irradiation/chemotherapy Coital disorders Other

38.9% 40.3% 6.4% 4.1% 1.9% 1.6% 0.6% 0.6% 0.5% 0.5% 4.5%

Modified from Baker HGW, et al. Relative incidence of etiological disorders in male infertility. In: Santen RJ, Swerdloff RS, eds. Male Reproductive Dysfunction: Diagnosis and Management of Hypogonadism, Infertility and Impotence. New York: Marcel Dekker, 1986; pp. 341–372.

atory scrototomy in 365 patients with azoospermia and normal serum FSH levels is shown in Table 61-17. Of these entities, only those with a genetic etiology will be discussed in detail. The others are grouped in Table 61-18. CONGENITAL BILATERAL ABSENCE OF THE VAS DEFERENS (CBAVD) Background The most frequent genetic abnormality causing ductal obstruction is known as congenital bilateral absence of the vas deferens, or CBAVD, an autosomal recessive condition associated with azoospermia, which accounts for at least 1–2% of infertility cases in men. A similar abnormality, known as congenital unilateral absence of the vas deferens (CUAVD) has also been described, and it is believed to be an incomplete form of CBAVD. Recently, both these entities have been shown to be a genital form of cystic fibrosis (CF). Clinical Features The typical patients with CBAVD presents with infertility as a result of absence of the vas deferens, and with a low-volume (>Na+

K+ = Rb+ = NH4 = Cs+>>Na+ Yes

Yes High υM K+ >> Na+ (20:1) Yes

Yes

and ROMK3); these spliced isoforms are both functional as channels and alternatively expressed along the mammalian nephron. The finding that the ROMK gene contains introns within the coding region is an unusual feature for a mammalian K+ channel, suggesting that these splicing events may serve some important role in altering channel function or regulation (or both). ROMK1– ROMK3 isoforms are produced by alternative splicing at the 5' end of the transcripts, which alters both the length and amino acid sequence of the encoded channel at initial segment of the NH2terminus. The role of the NH2-terminal variations on the ROMK isoforms in altering channel function and regulation by phosphorylation-dephosphorylation processes is currently under active investigation. As shown in Fig. 67-5, ROMK isoforms are differentially expressed along the nephron, with the overall distribution of ROMK being consistent with this IRK, representing the low-conductance secretary KATP channel observed in TAL and principal cells. ROMK2, which encodes the channel with the shortest NH2-terminus, is the most widely distributed of the ROMK transcripts. In contrast, the ROMK1 transcript is only expressed in collecting ducts (CCD, OMCD, and initial IMCD), whereas ROMK3 is expressed only in the earlier nephron segments (MTAL, CTAL, and DCT). Thus, all of these nephron segments, except the OMCD, apparently express at least two different ROMK isoforms: ROMK2 and one of the isoforms (ROMK1 or ROMK3) encoding a channel protein with a longer NH2-terminus. Do ROMK isoforms associate to form heteromultimeric complexes in some of these nephron segments? Would such heteromultimeric channels exhibit distinctive functional and regulatory properties? While these important questions are being actively pursued in several laboratories, other inward-rectifying K+ channels have recently been shown to heteromultimeric complexes: the G-proteingated cardiac atrial K+ channel formed from two IRK subunits and the pancreatic β-cell KATP formed from an IRK and the sulfonylurea receptor, which is a member of the P-glycoprotein family. The latter is quite interesting, since ROMK lacks significant sensitivity to glyburide, while the native low-conductance KATP in TAL or principal cells is inhibited by high micromolar concentrations of this potent sulfonylurea (Table 67-1). Could ROMK associate with a p-glycoprotein which would, like the β-cell KATP, impart sulfonylurea sensitivity? Mutations in ROMK1 cause the antenatal variant of Bartter syndrome.

FUNCTION OF ROMK CHANNELS AND CORRELATION WITH STRUCTURE ROMK K+ channels expressed in Xenopus laevis oocytes have been shown to exhibit many of the ion permeation and regulatory properties of the native lowconductance KATP channels observed on patch clamping of apical membranes of thick ascending limb and principal cells. These ROMK channels have the following basic properties: a unitary conductance of 30–45 pS in symmetrical 140–150 mM KCl; high K+:Na+ selectivity (K+ > Rb+ > NH4+> Na+, Li+); weak inward rectification resulting from block by internal Mg2+ and/or a polyamine such as spermine or spermidine; marked sensitivity to “cytosolic” side reductions in pH; modulation by arachidonic acid; run-down or loss of channel activity in excised patches in the absence of MgATP that involves dephosphorylation by a protein phosphatase; reactivation of channels after run-down by re-exposure to MgATP and the catalytic subunit of protein kinase A; and sensitivity (inhibition) by higher concentrations of MgATP. Importantly, ROMK channels exhibit no reduction in channel activity with addition of millimolar concentrations of TEA+ to extracellular media. This lack of sensitivity to TEA+ is in K+ secretion in kidney since both K+ secretion by the CCD and the low-conductance KATP channel found in principal cells are characteristically unaffected by TEA+. The aromatic amino acid, phenylalanine, at position 148 within the H5 region on ROMK1, could impart significant sensitivity to external TEA+ in ROMK channels, since site-directed mutagenesis studies in Shaker channels have shown that aromatic amino acids at a similar position confers a high sensitivity to external TEA+. It is plausible that the positively charged arginine, at position 147 in ROMK1, might interfere with TEA+ binding, and explain the lack of high TEA+ sensitivity in the ROMK1 channel. In this regard, recall that the intermediate conductance KATP channel found in rat TAL is sensitive to block by external TEA+. This pharmacological difference, together with the differences in single channel conductance and sensitivities to glyburide, provide strong evidence that ROMK does not, by itself, encode the –70 pS KATP channel. However, the similarity in many of the properties of the low- and intermediate-conductance channels suggest that these IRKs may be structurally similar. The possibility remains, however, that the intermediate conductance channel could be a heteromultimer of ROMK and some other IRK.

664

SECTION VIII / NEPHROLOGY

As discussed earlier in this chapter, inward rectification is also a characteristic of the apical secretary KATP channels in kidney as well as other members of the IRK family of channels. Recent studies have begun to identify important amino acid residues or segments critical for both types of rectification. The COOH-, but not NH2–, terminus appears to be important in determining inward rectification. Moreover, a single negatively charged amino acid residue (aspartic acid) in the second membrane-spanning helix of the strong inward rectifiers like IRK1 appears to be important in imparting this type of rectification. Importantly, ROMK lacks this charged amino acid residue, consistent with its weak type of inward rectification. As discussed above, the secretory K+ channels in distal nephron segments are ATP-sensitive or KATP channels. A segment following the second membrane-spanning region contains a Walker A site motif (GXG[H in ROMK]XXGK) within a so-called P04-binding or P loop that has been linked to ATP binding in many other proteins. In fact, a 27-amino acid region containing the Walker type A site exhibits significant similarity to the catalytic subdomain I of ERBB3, a member of the epidermal growth factor (EGF) receptor tyrosine kinase family. Based on this potential ATP-binding segment and the presence of several potential protein kinase A and protein kinase C phosphorylation sites in the COOH-terminus of ROMK channel proteins, we have suggested that this region forms a regulatory domain. Recent studies are consistent with the role of this COOH-terminal domain in regulation by nucleotides and protein serine-threonine kinases. In summary, molecular cloning has begun to identify important ion channels in the mammalian kidney. One of these, ROMK, appears to form the ion permeation pore of the ATP-sensitive inwardly rectifying K+ channel that mediates potassium secretion into tubular urine of TAL and CCD. Future studies will clearly provide new information on the function and regulation of this channel and provide new insights into renal potassium handling.

ACKNOWLEDGMENT This work was supported in part by a grant from the National Institutes of Health (DK37605) to SCH.

SELECTED REFERENCES Ashcroft FM. Adenosine 5'-triphosphate-sensitive potassium channels. Annu Rev Neurosci 1988;11:97–118. Ashcroft SJH, Ashcroft FM. Properties and functions of ATP-sensitive K-channels. Cell Signalling 1990;2:197–214. Boim MA, Ho K, Shuck ME, et al. The ROMK inwardly rectifying ATPsensitive K+ channel. II. Cloning and intrarenal distribution of alternatively spliced forms. Am J Physiol (Renal Fluid Electrolyte Physiol) 1995;268:F1132–F1140. Gebremedhin D, Kaldunski M, Jacobs ER, Harder DR, Roman RJ. Coexistence of two types of Ca(2+)-activated K+ channels in rat renal arterioles. Am J Physiol 1996;270:F69–F81. Hebert SC. An ATP-regulated, inwardly rectifying potassium channel from rat kidney (ROMK). Kidney Int 1995;48:1010–1016.

Hebert SC, Andreoli TE. Control of NaCl transport in the thick ascending limb. Am J Physiol (Renal Fluid Electroltye Physiol) 1984;247– 36:F745–F756. Hille B. Ionic Channels of Excitable Membranes, 2nd ed. Sunderland, MA: Sinauer, 1992. Ho K, Nichols CG, Lederer WJ, et al. Cloning and expression of an inwardly rectifying ATP-regulated potassium channel. Nature 1993;362:31–38. Inagaki N, Gonoi T, Clemant JP, et al. Reconstitution of IKATP: An inward rectifier subunit plus the sulfonylurea receptor. Nature 1995; 270:1166–1170. International Collaborative Study Group for Bartter-Like Syndromes. Mutations in the gene encoding the inwardly-rectifying renal potassium channel, ROMK, cause the antenatal variant of Bartter syndrome: evidence for genetic heterogeneity. Hum Mol Genet 1997;6:17–26. Kawahara K, Anzai N. Potassium transport and potassium channels in the kidney tubules. Jpn J Physiol 1997;47:1–10. Ketchum KA, Joiner WJ, Sellers AJ, Kaczmarek LK, Goldstein SAN. A new family of outwardly rectifying potassium channel proteins with two pore domains in tandem. Nature 1995;376:690–695. Krapivinsky G, Gordon EA, Wickman K, Vellmirovic B, Krapivinsky L, Clapham DE. The G-protein-gated atrial K + channel I KAch is a heteromultimer of two inwardly rectifying K+-channel protein. Nature 1995;374:135–141. Kubokawa M, Wang W, McNicholas CM, Giebisch G. Role of Ca2+/ CaMK II in Ca2+-induced K+ channel inhibition in rat CCD principal cell. Am J Physiol (Renal Fluid Electrolyte Physiol) 1995;268: F211–F219. Lee WS, Hebert SC. The ROMK inwardly rectifying ATP-sensitive K+ channel. I. Expression in rat distal nephron segments. Am J Physiol (Renal Fluid Electrolyte Physiol) 1995;268:F1124–F1131. McNicholas CM, Yang Y, Giebisch G, Hebert SC. Molecular site for nucleotide binding on an ATP-sensitive renal K+ channel (ROMK2). Am J Physiol 1996;271:F275–F285. Misler S, Giebisch G. ATP-sensitive potassium channels in physiology, pathphysiology, and pharmacology. Current Opinion in Nephrology and Hypertension 1992;1:21–33. Rossier BC, Canessa CM, Schild L, Horisberger JD. Epithelial sodium channels. Current Opinion in Nephrology and Hypertension 1994; 3:487–496. Simon DB, Karet FE, Rodriquez-Soriano J, et al. Genetic heterogeneity of Bartter’s syndrome revealed by mutations in the K + channel, ROMK. Nat Genet 1996;14:152–156. Suprenant A, North RA. P2X receptors bring new structure to ligand-gated channels. TINS 1995;18:224–229. Wang W. Two types of K+ channel in TAL of rat kidney. Am J Physiol (Renal Fluid Electrolyte Physiol) 1994;267–36:F599–F605. Wang W, Cassola A, Giebisch G. Arachidonic acid inhibits the secretory K+ channel of cortical collecting duct of rat kidney. Am J Physiol (Renal Fluid Electrolyte Physiol) 1992;262–31:F554–F559. Wang W, Cassola A, Giebisch G. Involvement of actin cytoskeleton in modulation of apical K channel activity in rat collecting duct. Am J Physiol (Renal Fluid Electrolyte Physiol) 1994;267–36:F592–F598. Wang W, Giebisch G. Dual effect of adenosine triphosphate on the apical small conductance K+ channel of the rat cortical collecting duct. J Gen Physiol 1991;98:35–61. Wang W, Hebert SC, Giebisch G. Renal K+ channels: structure and function. Annu Rev Physiol 1997;59:413–436. Xu ZC, Yang Y, Hebert SC. Phosphorylation of the ATP-sensitive, inwardly rectifying K+ channel, ROMK, by cyclic AMP-dependent protein kinase. J Biol Chem 1996;271:9313–9319.

CHAPTER 68 / ALPORT SYNDROME

68

665

Alport Syndrome KARL TRYGGVASON AND PIRKKO HEIKKILÄ

BACKGROUND Alport syndome, also termed hereditary nephritis, was initially described in 1927 by AC Alport as an inherited kidney disease characterized by hematuria and sensorineural deafness. Later, ocular lesions were also associated with the syndrome and, with the introduction of the electron microscope, irregularities and disruptions in the glomerular basement membrane (GBM) were shown to be typical for this disorder as well. The disease is progressive, usually leading to renal failure during adolescence or before middle age. Alport syndrome is primarily inherited as an X-chromosomelinked dominant trait with an estimated gene frequency of 1:5000, but autosomal forms also exist. The defective gene in X-linked Alport syndrome was located in 1988 and 1989 to the long arm of the X chromosome. In 1990, a gene encoding a novel basement membrane (type IV) collagen α5 chain was discovered and localized to the Alport gene region on chromosome X, and this was soon followed by identification of mutations in this gene in Alport patients. More recently, mutations have also been reported in the genes for the α3 and α4 type IV collagen chains in the rarer autosomal forms of Alport syndrome. Presently, well over 100 different mutations are now identified in type IV collagen genes in Alport patients. These mutations can be considered responsible for abnormalities in the structural framework of the GBM, resulting in kidney manifestations. In this chapter we review clinical features of the disease, the molecular properties of the GBM and type IV collagen, which are the prime targets of the disease, as well as recent advances in the molecular genetics and therapy.

CLINICAL FEATURES AND DIAGNOSIS Characteristically, Alport patients have recurrent microscopic or gross hematuria in childhood, earlier in males than in females. It usually leads to end-stage renal disease in affected males and in rare cases also in females. The hearing loss is sensorineural and primarily affects high tones. Electron microscopy usually reveals thinning and thickening of the GBM with longitudinal splits into thin layers with a basket-weave pattern (Fig. 68-1). These changes are most evident in male patients, except in boys of very young age. Lenticonus, a peculiar change of lens shape, is also frequently observed in Alport patients. Recurrent corneal erosion has also been reported. The disease is inherited and family history is present From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

in 85% of cases. The remaining 15% may represent new mutations. Recent work has revealed a large proportion of patients with renal failure and mutations in type IV collagen without hearing loss or eye lesions. Thus, the classical clinical definition for the Alport syndrome, i.e., hereditary nephritis with hearing loss, does not apply to all patients with type IV collagen defects. Clinicians should, therefore, consider Alport syndrome as a possible diagnosis for patients with hematuria and/or renal failure, even though they lack symptoms such as hearing loss, ocular lesions or family history.

STRUCTURE AND MOLECULAR BIOLOGY OF THE GLOMERULAR FILTRATION BARRIER FILTRATION BARRIER The glomerulus is a specialized vascular organ that functions as the filtration unit of plasma. Glomerular filtration is distinguished from transcapillary exchange in other organs by two characteristics: First, the glomerulus almost completely excludes plasma proteins of the size of albumin (mol wt approximately 70,000, radius 3.6 nm) and larger from the filtrate. Second, the glomerulus exhibits an extraordinary high permeability to water and small solutes. The filtration effect across the glomerulus depends on the size of the molecules; radius, which is called a size-dependent permeability barrier in the glomerulus. The filtration decreases with increasing effective molecular radius. In addition to the size, the glomerulus discriminates between molecules according to their charge, allowing greater penetration of neutral and cationic molecules than of anionic molecules of the same size. The actual filtration barrier consists of the fenestrated capillary endothelium, the GBM of 300–350 nm in thickness, and the epithelial podocytes that are separated by the slit and connected by a thin diaphragm. Similar to basement membranes elsewhere in the body, the GBM consists of type IV collagen, laminin, proteoglycan, and nidogen/entactin. Type IV collagen forms the structural framework of the GBM together with laminin, which also plays a role in cell adhesion. Nidogen interconnects the type IV collagen and laminin networks, and proteoglycans are believed to serve as an anionic filtration barrier. TYPE IV COLLAGEN Type IV collagen is a basement membrane specific protein that belongs to the large family of collagens that are the main extracellular structural proteins in all multicellular organisms. Collagen molecule is composed of three α chains, characterized by the repeating Gly-X-Y sequence, where X and Y can be any amino acid. The glycine residue, as every third amino acid enables the assembly of three chains into a triple-helical struc-

665

666

SECTION VIII / NEPHROLOGY

Table 68-1 Diseases with Linkage to Type IV Collagen Chains Disease Genetic X-linked Alport syndrome X-linked Alport syndrome associated with diffuse esophageal lyomyomatosis Autosomal Alport syndrome Acquired Goodpasture syndrome

Figure 68-1 Electron micrograph of the glomerular filtration barrier in an 8-year-old Alport syndrome patient. The figure displays typical findings for the disease with irregular thinning and thickening, as well as lamellation of the GBM. C, capillary lumen; En, endothelial cell; GBM, glomerular basement membrane; Ep, epithelial cell; U, urinary space. (Courtesy of Dr. Helena Autio-Harmainen, University Hospital of Oulu, Finland.)

Type IV collagen α chain involved α5 α5 + α6 α3 and/or α4 α3

α chains are known. Of those, the α1 and α2 chains are ubiquitous and can be found in all basement membranes. In contrast, the α3, α4, α5, and α6 chains are minor component chains with a more restricted tissue distribution and thus probably specialized functions. For example, molecules containing α3, α4, and α5 chains are particularly abundant in the GBM. Basement membranes containing these chains are believed to be stronger as a result of the presence of a high content of cysteine residues and, thus, disulfide crosslinks. The strength of the structural network is considered to be particularly important in basement membranes, such as those of renal glomeruli and the lens capsule, which do not have supportive collagen fibers. Molecules containing the α3–α6 chains have been shown to be involved in the pathogenesis of diseases such as Goodpasture syndrome, Alport syndrome, and/or diffuse esophageal leiomyomatosis (Table 68-1). Isoform switching (α4, α5) of type IV collagen is developmentally arrested in Alport syndrome, leading to increased susceptibility of renal basement membranes to endoproteolysis. TYPE IV COLLAGEN GENES The mammalian type IV collagen genes are located pairwise in an unique head-to-head fashion on three different chromosomes: chromosomes 13, 2, and X (Fig. 68-3). This implies that the six genes have evolved through duplication and inversion of an ancestral gene. Subsequently, the duplicated genes have undergone two further rounds of duplication, resulting in the three head-to-head located gene pairs on different chromosomes. The type IV collagen genes are large, over 100 kb, and complex, containing 46–52 exons. The genes are termed COL4A1, COL4A2, COL4A3, COL4A4, COL4A5, and COL4A6, respectively, for the type IV collagen α1, α2, α3, α4, α5, and α6 chains.

MOLECULAR PATHOLOGY OF THE DISEASE Figure 68-2 Schematic illustration of the structure and extracellular assembly of type IV collagen. (A) Each a chain has a 400-nm-long collagenous domain characterized by a frequently interrupted G-X-Yrepeat sequence. The interruptions, which are not shown in the figure give flexibility to the molecule and the type IV collagen network. (B) Three α chains form a triple-helical molecule. (C) Individual molecules assemble by end-to-end associations into a network-like structure into which other basement membrane proteins are bound in a largely unknown fashion. In reality, the type IV collagen network structure also contains more complex laterally aligned molecules.

ture. All collagens consist of three α chains that form the triplehelical molecule inside the cell. Once secreted from the cells, the triple-helical molecules assemble by end-to-end aggregation and lateral alignment into a tightly crosslinked network structure (Fig. 68-2). Presently, six genetically distinct type IV collagen

Mutations have been described in the COL4A5 gene in more than 100 cases of X-linked Alport syndrome. Furthermore, mutations have been identified in the COL4A3 and COL4A4 genes in autosomal recessive forms of the disease. Almost all of the mutations identified to date differ among familial groups. This, together with the complexity of the collagen genes, makes DNA-based diagnosis of Alport syndrome particularly difficult. About 15% of the mutations are large gene rearrangements, such as deletions, insertions, inversions, or duplications. The rest are small mutations, mainly large proportion of single base changes, in addition to small deletions, insertions, or duplications. The mutations can result in a complete absence of the protein (α chain) in question, a truncated protein, or a malfunctional protein. A loss of the carboxyterminal noncollagenous domain would prohibit the formation of heterotrimers. Several mutations in Alport syndrome involve replacement of a glycine residue in the collagenous domain

CHAPTER 68 / ALPORT SYNDROME

667

Figure 68-3 Evolution of type IV collagen genes. Analysis of the gene and protein sequences indicate that an ancestor type IV collagen α chain gene was first duplicated and inversed, followed by divergence of the genes to α1- and α2-like genes. This gene pair was later duplicated to two other chromosomes so that the α3(IV) and α4(IV) genes were formed prior to those of α5(IV) and α6(IV).

by another amino acid. Glycines at every third position are required for the formation of collagen triple helices, since this amino acid is the only amino acid small enough to fit into the interior of the tightly wound triple helix. Glycine mutations are also frequently the cause of other collagen disorders such as osteogenesis imperfecta. Based on the knowledge obtained from studies on osteogenesis imperfecta, it has been suggested that the abnormally folded triple helices are very susceptible to degradative enzymes, leading to the partial or total absence of all the α3, α4, and α5 chains. As a result, the structural framework of the GBM, which requires type IV molecules containing these polypeptide chains, becomes structurally weak and disrupted. The use of monoclonal antibodies against the α3, α4, or α5 chains in immunofluorescense microscopy of kidney biopsies may become a method to confirm the diagnosis of Alport syndrome in some cases, as an absence of these antigens can sometimes be seen in the skin basement membrane. A mouse model for the autosomal form of Alport syndrome has been produced by a collagen COL4A3 knockout. The mice develop progressive glomerulonephritis. An interesting fact is that a clear correlation cannot be found between the nature of a mutation and the respective phenotype. A small single amino acid substitution can cause as severe symptoms as a large deletion of almost the entire collagen gene.

THERAPY GENERAL Although the onset of hematuria occurs during early childhood, the disease usually progresses slowly. A large number of male patients enter terminal renal failure during adolescence (juvenile form), but the onset also occurs beyond 25 years of age (adult type). Usually affected males and homozygous

Figure 68-4 Expression of β-galactosidase in porcine kidneys following in vivo perfusion with an adenovirus containing the β-galactosidase reporter gene under the cytomegalovirus promoter. Intense expression (dark color) can be observed in almost all glomeruli, whereas little, if any, is observed in the tubular cells.

females develop end-stage renal disease before the fifth decade of life, while heterozygous females rarely develop renal failure. There is no satisfactory and curative conservative treatment available. Patients developing end-stage renal disease are treated by hemodialysis and also by kidney transplantation whenever possible. About 5% of transplanted patients develop anti-GBM nephritis and reject the allografted kidneys.

668

SECTION VIII / NEPHROLOGY

PROSPECTS FOR GENE THERAPY As a result of this advances in molecular genetics and biology research, gene therapy may be developing into a real possibility for the treatment of hereditary diseases in the future. Although gene therapy has not yet come of age as a real therapeutic alternative, extensive research efforts are being made in that direction. Hereditary kidney diseases such as Alport syndrome, primarily affecting the renal glomeruli, could potentially be treated by somatic gene therapy, which involves the introduction of normal cDNAs or complete genes for the α3, α4, or α5 chains into glomerular cells. Recent work has already shown it to be possible to transfer foreign genes into over 85% of the glomeruli in vivo (Fig. 68-4). Thus, the stage has been set for actual gene therapy experiments in animal models of Alport syndrome. There are, however, many obstacles to overcome before we can expect the disease to be cured in humans by gene therapy. But if gene therapy is going to become a viable alternative for the treatment of genetic diseases in general, it is likely to become an option for Alport syndrome.

SELECTED REFERENCES Adler SG, Cohen AH, and Glassock RJ. Secondary glomerular diseases. In: Brenner BM, ed. The Kidney, 5th ed. Philadelphia, PA: W.B. Saunders, 1996; pp. 1555–1558. Alport AC. Hereditary familial congenital haemorrhagic nephritis. Br Med J 1927;1:504–506. Atkin CL, Gregory MC, Border WA. Alport syndrome. In: Schrier WW, Gottschalk, CW, eds. Diseases of Kidney. Boston: Little, Brown, 1988; pp. 617–641. Atkin CL, Hasstedt SJ, Menlove L, et al. Mapping of Alport syndrome to the long arm of the X-chromosome. Am J Hum Genet 1988;42: 249–255. Barker DF, Hostikka SL, Zhou J, et al. Identification of mutations in the COL4A5 collagen gene in Alport syndrome. Science 1990;248: 1224–1227. Brunner H, Schroder C, van Bennekom C. Localization of the gene for X-linked Alport’s syndrome. Kidney Int 1988;34:507–510. Cosgrove D, Meehan DT, Grunkemeyer JA, et al. Collagen COL4A3 knockout: a mouse model for autosoma Alport syndrome. Genes Dev 1996;10:2981–2992. Ding J, Stitzel J, Berry P, Hawkins E, Kashtan CE. Autosomal recessive Alport syndrome: mutation in the COL4A3 gene in a woman with Alport syndrome and posttransplant antiglomerular basement membrane nephritis. J Am Soc Nephrol 1995;5:1714–1717. Flinter FA, Abbs S, Bobrow M. Localization of the gene for classic Alport syndrome. Genomics 1989;4:335–338.

Heikkilä P, Parpala T, Lukkarinen O, Weber M, Tryggvason K. Adenovirus-mediated gene transfer into kidney glomeruli using an ex vivo and in vivo kidney perfusion system—First steps towards gene therapy of Alport syndrome. Gene Therapy 1996;3:21–27. Heiskari N, Zhang X, Zhou J, et al. Identification of 17 mutations in ten exons in the COL4A5 collagen gene, but no mutations found in four exons in COL4A6: a study of 250 patients with hematuria and suspected of having Alport syndrome. J Am Soc Nephrol 1996;7:702–709. Hostikka SL, Eddy RL, Byers MG, et al. Identification of a distinct type of IV collagen α chain with restricted kidney distribution and assignment of its gene to the locus of X chromosome-linked Alport syndrome. Proc Natl Acad Sci USA 1990;87:1606–1610. Hudson BG, Reeders ST, and Tryggvason K. Type IV collagen: structure, gene organization, and role in human diseases: molecular basis of Goodpasture and Alport syndromes and diffuse leiomyomatosis. Minireview J Biol Chem 1993;268:26,033–26,036. Kalluri R, Shield CF, Todd P, Hudson BG, Neilson EG. Isoform switching of type IV collagen is developmentally arrested in X-linked Alport syndrome leading to increased susceptibility of renal basement membranes to endoproteolysis. J Clin Invest 1997;99:2470–2478. Knebelmann B, Breillat C, Forestier L, et al. Spectrum of mutations in the COL4A5 collagen gene in X-linked Alport syndrome. Am J Hum Genet 1996;59:1221–1232. Lemmink HH, Mochizuki T, van den Heuvel LPWJ, et al. Mutations in the type IV α3 (COL4A3) gene in autosomal recessive Alport syndrome. Hum Molec Genet 1994;3:1269–1273. Lemmick HH, Schroder CH, Monnens LA, Smeets HJ. The clinical spectrum of type IV collagen mutations. Hum Mutat 1997;9:477–499. Miner JH, Sanes JR. Molecular and functional defects in kidneys of mice lacking collagen alpha 3(IV): implications for Alport syndrome. J Cell Biol 1996;135:1403–1413. Mochizuki T, Lemmink HH, Mariyama M, et al. Identification of mutations in the α3(IV) α4(IIV) collagen genes in autosomal recessive Alport syndrome. Nature Genet 1994;8:77–82. Renieri A, Bruttini M, Galli L, et al. X-linked Alport syndrome: an SSCPbased mutation surve over all 51 exons of the COL4A5 gene. Am J Hum Genet 1996;58:1192–1204. Rhys C, Snyers B, Pirson Y. Recurrent corneal erosion associated with Alport’s syndrome. Rapid Communication. Kidney Int 1997;52:208–211. Saito A, Yamazaki H, Nakagawa Y, Arakawa M. Molecular genetics of renal diseases. Intern Med 1997;36:81–86. Tryggvason K. Mutations in type IV collagen genes and Alport phenotypes. In: Tryggvason K, ed. Molecular Pathology and Genetics of Alport Syndrome, vol 117. Basel: Karger, 1996; pp. 154–171. Tryggvason K, Zhou J, Hostikka SL, Shows T. Molecular genetics of Alport syndrome. Kidney Int 1993;43:38–44. Yurchenco PD. Assembly of basement membranes. In: Fleischmajer R, Olsen BR, Kühn, K, eds. Structure, Molecular Biology and Pathology of Collagen, vol. 580. Ann New York Acad Sci, 1990; pp. 195–213.

CHAPTER 69 / NEPHROGENIC DIABETES INSIPIDUS

69

669

Nephrogenic Diabetes Insipidus DENNIS BROWN AND DENNIS A. AUSIELLO

INTRODUCTION The neurohypophyseal antidiuretic hormone, arginine vasopressin (AVP), stimulates urinary concentration in mammals by increasing the water permeability of renal collecting ducts and by stimulating NaCl reabsorption by thick ascending limbs of Henle into the medullary interstitium. In the collecting duct, AVP binds to specific V2 receptors (V2R) on the basolateral plasma membrane of principal cells and stimulates adenylyl cyclase, which increases cytosolic cAMP levels. The increase in cAMP activates protein kinase A (PKA) and protein phosphorylation ensues. While the nature and role of the PKA substrates remains generally obscure, one phosphorylated protein is the vasopressin-sensitive water channel, aquaporin 2 (AQP2), which relocates from a pool of intracellular vesicles to the apical plasma membrane of principal cells upon vasopressin stimulation. This exocytotic insertion of water channels, illustrated in Fig. 69-1, greatly increases the water permeability of the principal cell apical membrane. Because the basolateral membrane of principal cells has a constitutively high water permeability because of the presence of two other water channels, AQP3 and AQP4, this vasopressin-induced apical exocytotic process is the rate-limiting step for increasing collecting duct epithelial cell water permeability. The urine in the tubule lumen equilibrates osmotically with the hypertonic interstitium by the bulk flow of water across the collecting duct principal cells, and the urine is concentrated. Defective urinary concentration has long been recognized in man and in animal models such as the Brattleboro DI rat and the nephrogenic diabetes insipidus (NDI) mouse. The molecular basis for many of these related disorders has been examined and elucidated, thanks to the cloning and sequencing of two of the key proteins involved, the V2R and the vasopressin-sensitive collecting duct water channel, AQP2. This brief review will discuss cell and molecular biological evidence showing that mutations resulting in defective targeting and/or function of the V2R and the AQP2 water channel can both lead to distinct forms of NDI. Type I congenital nephrogenic DI (CNDI), the more frequent X-linked form, is caused by mutations in the vasopressin receptor, whereas type II CNDI is an autosomal recessive disease resulting from mutations in the AQP2 water channel.

From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

CONGENITAL NDI—LOCATION OF THE VASOPRESSIN V2 RECEPTOR (V2R) GENE CNDI was first described over 50 years ago, and genetic linkage studies in several families established that it is an X-linked trait. In CNDI, arginine vasopressin production is normal, but the target cells in the kidney do not respond to the presence of circulating hormone. Thus, the transepithelial water permeability of the collecting duct remains in its low basal state, and a large volume of hypotonic urine is produced. Clinically, this condition is recognizable soon after birth and if not corrected can result in severe dehydration, hypernatremia, and damage to the central nervous system. Linkage studies showed that the gene responsible is located in the subtelomeric region of the long arm of the X chromosome in region 28 (Xq28), where the V2R gene is also located.

STRUCTURE OF THE V2R The V2R is a 371-amino acid protein that is a member of the family of seven membrane-spanning domain receptors that couple to heterotrimeric G proteins. Other members of this large family include rhodopsin and the β2-adrenergic receptor. Homologs of the V2R have been cloned from human, pig, and rat, and the receptor sequences are more than 90% identical. The membrane topology of the receptor, as well as several functionally important features, are illustrated in Fig.69- 2. These include: (1) an extracellular N-terminus with a consensus site for N-linked glycosylation; (2) a cytoplasmic carboxy-terminus and large intracellular loop that contain several sites for serine and threonine kinase phosphorylation, and which probably play a role in receptor internalization and sequestration; (3) conserved sites for fatty acylation, which may serve as an additional membrane anchor in the C-terminal tail; and (4) two highly conserved cysteine residues in the second and third extracellular loops, which may form a disulfide bridge that is important for correct folding of the molecule and stabilization of the ligand-binding site.

CNDI AND MUTATIONS IN THE V2R A series of reports have now identified mutations in the coding sequence of the V2R gene that can account for the production of a nonfunctional receptor by target cells in the kidney. A vast array of missense, nonsense, and frameshift mutations in various domains of the receptor coding sequence have been reviewed in detail previously, and some are illustrated in Fig. 69-3. Some result in the appearance of premature stop codons, which produce an effective null mutation by causing early truncation of the growing

669

670

SECTION VIII / NEPHROLOGY

Figure 69-1 Proposed pathway of AQP2 recycling in collecting duct principal cells. AQP2 is located on intracellular vesicles that move to and fuse with the apical plasma membrane following vasopressin stimulation and a subsequent rise in intracellular cAMP levels. The water channel is then internalized via clathrincoated pits and eventually recycles back to the apical membrane in a complex trafficking pathway that has not yet been fully elucidated. In the same cell type, two other aquaporins, AQP3 and AQP4 (also referred to as the mercurial insensitive water channel [MIWC]), are located on the basolateral plasma membrane. The membrane localization of these two proteins is not regulated by vasopressin. They are constitutively present in this membrane domain, and they are assumed to be responsible for its permanently high water permeability. Figure 69-3 Reported mutations in the V2R, showing the diversity of affected sites and types of mutations. (A) Missense mutations. (B) Nonsense and frameshift mutations. (Taken from Holtzman et al.)

Figure 69-2 Membrane topology of the vasopressin receptor (V2R) molecule. The 371-amino acid protein has seven membranespanning domains, an extracellular N-terminus, and a cytoplasmic C-terminus. Several features of the molecule are illustrated, with some key residues and potential phosphorylation sites shown in black. One mutation (R113W), close to the disulfide bridge, is shown. In work from our laboratory, a FLAG epitope added to the N-terminus allows efficient immunolocalization studies to be performed and was used to demonstrate retention of the R113W mutation in the RER.

polypeptide chain. Others are single-point mutations that cause a single amino acid replacement at a functionally critical site in the receptor molecule. One such example is an arginine (R) 113 to tryptophan (W) mutation caused by a C -> T base transition. This point mutation at the end of the first extracellular loop is one amino acid toward the carboxyl-terminus from a critical cysteine residue and may interfere with the formation of an important disulfide bond within the vasopressin receptor. Interestingly, an analysis of other G-protein coupled receptors has shown that tryptophan is never found at position 113 in any functional receptor, indicating that its presence is incompatible with normal receptor function. Mutations in the V2R sequence that still allow the production of a full-length or near full-length protein could result in CNDI by interfering with different aspects of the receptor-ligand signal transduction cascade. For example: (1) the receptor could be expressed normally at the plasma membrane, but fail to bind vasopressin; (2) the receptor could be expressed normally at the membrane, bind its ligand normally, but fail to couple to its stimulatory GTP-binding protein, so that adenylyl cyclase is not activated;(3) the mutated receptor may be incorrectly folded and might be retained for degradation in the rough endoplasmic reticulum (RER), and may never reach the cell surface; (4) changes in the ability of the receptor to be phosphorylated may affect several aspects of function, including trafficking and desensitization. So far, functional analyses of the potential cell biological mecha-

CHAPTER 69 / NEPHROGENIC DIABETES INSIPIDUS

671

nisms that result in CNDI have lagged behind the rapid rate at which various V2R mutants have been identified at the molecular level.

THE R133W MUTATION MODIFIES INTRACELLULAR TRAFFICKING OF THE V2R The experiment in nature provided by the R113W V2R mutation has provided a model for the study of structural elements within receptors that permit proper folding and/or movement through the intracellular biosynthetic pathway. It has been repeatedly shown that misfolded proteins are triaged by a rigorous quality-control mechanism within the RER, and in most cases abnormally folded or assembled proteins are targeted for intracellular degradation without leaving the RER. When expressed at normal levels in vivo, the R113W V2R probably never reaches the plasma membrane in detectable amounts. Experimental overexpression of the R113W V2R in transfected LLC-PK1 cells appears to “saturate” the quality-control process, allowing at least some mutated V2R to escape degradation and reach the cell surface. Once at the cell surface, the R113W V2R can bind vasopressin, albeit at lower affinity than the wild type receptor, and can stimulate intracellular cAMP accumulation. However, even in transfected cells, the bulk of the mutated receptor is located within the RER by immunocytochemistry (Ausiello et al., unpublished results). An R337V2R mutation provides another example of receptor dysfunction because of arrested maturation and degradation before expression on the cell surface. Thus, this particular mutation disables the V2R and the urinary concentrating mechanism, not by preventing receptor-ligand binding (although there is a shift in binding affinity), but by modifying the intracellular trafficking of the abnormal protein, such that it never reaches the plasma membrane and never comes into contact with its stimulatory ligand. A similar situation occurs in cystic fibrosis. The most common mutation in the CFTR gene product is the so-called ∆F508 mutation, again a single-point mutation. This mutation also leads to the trapping of a misfolded CFTR protein in the RER and prevents its delivery to the cell surface where it normally functions as a PKA-regulated chloride channel. Under experimental conditions that allow delivery of the ∆F508 CFTR to the cell surface, it functions normally and is regulated normally by PKA. Functional rescue of mutant V2 receptors has been demonstrated in transfected COS-7 cells by coexpression of the normal receptor sequences. Thus, it may be possible to devise therapeutic strategies that result in the correct delivery of misfolded, but otherwise functionally active, proteins to their correct location at the cell surface in vivo.

AUTOSOMAL RECESSIVE NDI—MUTATIONS IN AQUAPORIN 2 Some cases of NDI are not of the typical X-linked type, but are autosomal-recessive in nature. Alterations in the vasopressin receptor have not been found in these individuals, but instead, the defect lies within the vasopressin-regulated water channel, aquaporin 2 (AQP2). Based on two decades of studies, it was postulated that vasopressin exerts its antidiuretic effect by causing the exocytotic insertion of water channels into the apical plasma membrane of collecting duct principal cells (Fig. 69-1). The resulting increase in membrane water permeability allows osmotic equilibration to occur between the luminal fluid (urine) and the hypertonic medullary interstitium of the inner medulla.

Figure 69-4 Membrane topology of the aquaporin 2 (AQP2) water channel. This 271-amino acid protein spans the lipid bilayer six times. Both N- and C-termini are in the cytoplasm. There is a consensus site for N-glycosylation in the second extracellular loop, and phosphorylation sites for PKC (Pc) and PKA (Pa) are located in the C-terminal tail. Two point mutations (R187C and S216P) that were found in NDI patients are indicated with arrows. (Taken from Deen et al.)

AQP2 was cloned and sequenced based on its homology to AQP1, and a series of studies showed that AQP2 is the vasopressin-sensitive water channel in collecting duct principal cells. In collecting duct principal cells, the membrane localization of AQP2 is tightly regulated by the antidiuretic hormone, vasopressin. In Brattleboro homozygous rats, which lack vasopressin and which have hypothalamic diabetes insipidus, AQP2 is located primarily on intracellular vesicles but is delivered to the apical plasma membrane by exocytosis following vasopressin treatment in vivo. A similar translocation was seen in isolated perfused tubules from normal rat. More recently, translocation of AQP2 from intracellular vesicles to the plasma membrane has been demonstrated in transfected renal epithelial cells in culture. This latter result shows that the AQP2 protein contains sorting information that allows it to be incorporated into a regulated pathway of exocytosis. Together, these data directly support the shuttle hypothesis of vasopressin action, which invokes a cycle of exo- and endocytosis of water channels to explain the stimulatory effect of vasopressin on collecting duct water permeability. In contrast, the proximal tubule and thin limb water channel, AQP1, is delivered to the plasma membrane in a constitutive, nonregulated pathway that results in a permanently high membrane water permeability in these renal epithelial cells. When expressed in epithelial cells in culture, AQP1 is also delivered to the cell surface without the need for hormonal stimulation. Thus, AQP1 and AQP2 contain distinct targeting motifs that direct them to different and physiologically important intracellular transport pathways. These distinct pathways can be maintained in vitro in transfected epithelial cells, providing powerful new model systems for the investigation of aquaporin cell biology. Important data on the role of AQP2 in non–X-linked NDI was obtained by examining the AQP2 molecule in human disease. A few examples of AQP2 mutations have now been described (Fig. 69-4), but in most cases the cause of the AQP2 inactivity is not clear. In the first patient studied, two distinct point mutations were found that resulted in a substitution of Cys for Arg187, and Pro

672

SECTION VIII / NEPHROLOGY

for Ser216. The R187C mutation occurs in a region of the third extracellular loop that is strongly conserved in members of the aquaporin family. The S216P mutation is located in the last transmembrane domain of AQP2. When mRNAs coding for these mutated proteins (which had the predicted size of 29 kDa) were expressed in Xenopus oocytes, no increase in membrane osmotic water permeability was detected above that seen in water-injected controls. However, whether the mutated proteins actually reached the cell surface of oocytes is unknown. Mutations in the AQP2 gene that result in production of a fulllength protein could have at least two consequences that would result in a loss of principal cell vasopressin-sensitivity: (1) the production of an AQP2 channel that is still vasopressin-sensitive, and is inserted into the plasma membrane after vasopressin action, but which has lost the capacity to function as a water channel; (2) the production of a water channel that is still functional, but that is no longer targeted to the plasma membrane after vasopressin action. Other mutations could, of course, be null mutations that result in an absent or severely truncated protein. In support of the second possibility, oocyte expression studies have shown that the missense mutations R187C and S216P are impaired in their delivery to the cell surface. The trapped form of these AQP2 mutants is a 32-kDa high mannose form, indicating that the protein is blocked in the RER, in much the same way as the ∆F508 CFTR mutation is trapped inside the RER. However, it could not be determined whether these mutants were also functionally inactive water channels. That the water permeability of AQP2 can be modified by mutagenesis was shown by recent data indicating that some point mutations, including mutations in asparagine residue 123, significantly affect channel water permeability. The plasma membrane expression of this mutation was only slightly decreased from that shown by the wild type protein in oocytes.

ACQUIRED NDI Several pathological conditions or drugs result in the syndrome of acquired NDI. Among these are lithium-induced NDI and hypokalemia (which occurs in 40% of all patients receiving lithium, polyuria in up to 35%, and NDI in 12–20%). Many studies have shown that a major cellular effect of lithium is an impairment of cAMP production that would, in principal cells, normally occur after vasopressin stimulation. While the basis for this inhibition is likely to be complex, acute lithium loading in renal epithelial cells directly inhibits adenylyl cyclase activation by competing for activation of the stimulatory GTP-binding protein subunit, Gs. The effect appears to be at the magnesium-sensitive rate limiting step in the generation of dissociated and GTP-activated Gαs-subunit, which normally activates adenylyl cyclase. How might this modify collecting duct water permeability? Acutely, the effect would be to inhibit vasopressin-induced insertion of the AQP2 water channel into the plasma membrane of principal cells, which is a cAMP-dependent process. However, lithium also has chronic effects, and it has recently been shown that the cellular content of AQP2 protein is decreased in lithiumtreated rats. This effect probably results from decreased transcription of the AQP2 gene, which is known to have a cAMP-responsive element in its 5' flanking region. Other studies have shown that dehydration of normal rats, or AVP treatment of Brattleboro rats, increases cellular AQP2 levels, implying that increased cellular cAMP “switches on” AQP2 production in principal cells. In lithium intoxication, the cellular cAMP levels would be chroni-

cally reduced, thus lowering AQP2 production at the transcriptional level. Hypokalemia also results in NDI, and experimental hypokalemia in rats also reduces the cellular content of AQP2. Again, the precise cause is unknown, but hypokalemia causes cellular resistance to AVP, as well as directly stimulating thirst. At the cell biological level, it is known that lowering cellular potassium inhibits clathrin-mediated endocytosis and thus may interfere with the exo- and endocytosis of key proteins involved in the antidiuretic response. In particular, AQP2 is probably internalized and recycled by a process involving clathrin-coated pits, and most receptors, including the vasopressin receptor, is also dependent on clathrin-mediated endocytosis as part of its recycling pathway. Thus, interference with the correct recycling of either or both of these proteins could result in an NDI-like syndrome. Studies on tissues from potassium-depleted animals have shown a significant reduction in AVP-induced cAMP generation, which would, as in the case of lithium toxicity, result in lower transcription on the AQP2 gene and lower AQP2 levels in principal cells.

SUMMARY With the cloning and sequencing of key proteins that are involved in the renal medullary concentrating mechanism, cell biological explanations for nephrogenic diabetes insipidus are now emerging. In congenital X-linked NDI, many mutations in the vasopressin receptor result in varying degrees of receptor inactivation, either by altering the function of the receptor itself or by preventing its proper targeting and trafficking to the plasma membrane of principal cells in the collecting duct. The rarer cases of non–X-linked NDI can, in some cases, be attributed to defects in the AQP2 water channel. Again, it is likely that functional as well as trafficking mutations will explain the NDI phenotype in cases where vasopressin-receptor activation and cAMP generation are normal. Finally, two examples of acquired NDI were presented, and potential causes of the defect based on the cellular actions of lithium and hypokalemia on AQP2 production and trafficking have been proposed.

SELECTED REFERENCES Ausiello DA, Holtzman EJ, Gronich GH, Ercolani L. Cell signalling. In: Seldin DW, Giebisch G, eds. The Kidney: Physiology and Pathophysiology. New York: Raven, 1992; pp. 645–692. Bai L, Fushimi K, Sasaki S, Marumo FG. Structure of aquaporin-2 vasopressin water channel. J Biol Chem 1996;271:5171–5176. Bichet DG. X-linked nephrogenic diabetes insipidus mutations in North America and the Hopewell hypothesis. J Clin Invest 1992;92: 1262–1268. Bichet DG, Cameron S, Davison AM. Nephrogenic diabetes insipidus. In: Graunfeld JP, Kerr D, Ritz E, eds. Oxford Textbook of Clinical Nephrology. New York: Oxford Medical, 1992; pp. 789–800. Birnbaumer M, Seibold A, Gilbert S. Molecular cloning of the receptor for human antidiuretic hormone. Nature 1992;357:333–335. Bonifacio JS, Suzuki CK, Klausner RD. A peptide sequence confers retention and rapid degradation in the endoplasmic reticulum. Science 1990;247:79–82. Brown D, Orci L. Vasopressin stimulates the formation of coated pits in rat kidney collecting ducts. Nature 1983;302:253–255. Brown D, Shields GI, Valtin H, Morris JF, Orci L. Lack of intramembranous particle clusters in collecting ducts of mice with nephrogenic diabetes insipidus. Am J Physiol 1985;249:F582–F589. Brown D, Stow JL. Protein trafficking and polarity in kidney epithelium: from cell biology to physiology. Physiol Rev 1996;76:245–297. Brown D, Weyer P, Orci L. Vasopressin stimulates endocytosis in kidney collecting duct epithelial cells. Eur J Cell Biol 1988;46:336–340.

CHAPTER 69 / NEPHROGENIC DIABETES INSIPIDUS

Cheng S, Gregory DR, Marshall J, et al. Defective intracellular transport and processing of CFTR is the molecular basis of most cystic fibrosis. Cell 1990;63:827–834. Cheong HI, Park HW, Ha IS, et al. Six novel mutations in the vasopressin V2 receptor gene causing nephrogenic diabetes insipidus. Nephron 1997;75:431–437. Deen PM, Verdijk MA, Knoers NV, et al. Requirement of human renal water channel aquaporin-2 for vasopressin-dependent concentration of urine. Science 1994;264:92–95. Deen PMT, Croes H, van Aubel RAMH, Ginsel LA, van Os CH. Water channels encoded by mutant aquaporin-2 genes in nephrogenic diabetes insipidus are impaired in their cellular routing. J Clin Invest 1995;95:2291–2296. Dohlman HG, Thorner J, Caron MG, Lefkowitz RJ. Model systems for the study of seven-transmembrane-segment receptors. Annu Rev Biochem 1991;60:653–688. Ecelbarger CA, Terris J, Frindt G, et al. Aquaporin-3 water channel localization and regulation in rat kidney. Am J Physiol 1995;269:F663–F672. Echevarria M, Windhager EE, Tate SS, Frindt G. Cloning and expression of AQP3, a water channel from the medullary collecting duct of rat kidney. Proc Natl Acad Sci USA 1994;91:10,997–11,001. Forssmann H. On hereditary diabetes insipidus. Acat Med Scand 1994;121 (Suppl. 159):9–46. Frigeri A, Gropper MA, Kawashima M, Brown D, Verkman AS. Localization of MIWC and GLIP water channel homologs in neuromuscular, epithelial and glandular tissues. J Cell Sci 1995;108:2993–3002. Fushimi K, Uchida S, Hara Y, Hirata Y, Marumo F, Sasaki S. Cloning and expression of apical membrane water channel of rat kidney collecting tubule. Nature 1993;361:549–552. Hayashi M, Sasaki S, Tsuganezawa H, et al. Expression and distribution of aquaporin of collecting duct are regulated by vasopressin V2 receptor in rat kidney. J Clin Invest 1994;94:1778–1783. Hochberg Z, Van Lieburg A, Even L, et al. Autosomal recessive nephrogenic diabetes insipidus caused by an aquaporin-2 mutation. J Clin Endocrinol Metab 1997;82:686–689. Holtzman EJ, Ausiello DA. A molecular defect in the vasopressin V2receptor gene causing nephrogenic diabetes insipidus. N Engl J Med 1993;328:1534–1537. Holtzman EJ, Kolakowski LF, Ausiello DA. The molecular biology of congenital nephrogenic diabetes insipidus. In: Schlondorff D, Bonventre JV, eds. Molecular Nephrology. New York: Marcel Dekker, 1995; pp. 887–910. Hozawa S, Holtzman EJ, Ausiello DA. cAMP motifs regulatingtranscription of the aquaporin 2 gene. Am J Physiol 1996;270:C1695–C1702. Ishibashi K, Sasaki S, Fushimi K, et al. Molecular cloning and expression of a member of the aquaporin family with permeability to glycerol and urea in addition to water expressed at the basolateral membrane of kidney collecting duct cells. Proc Natl Acad Sci USA 1994;91:6269–6273. Katsura T, Verbavatz JM, Farinas J, et al. Constitutive and regulated membrane expression of aquaporin-CHIP (AQP1) and aquaporin 2 water channels in stably transfected LLC-PK1 epithelial cells. Proc Natl Acad Sci USA 1995;92:7212–7216. Lolait SJ, O’Carroll AM, Konig M, Morel A, Brownstein MJ. Cloning and characterization of a vasopressin V2 receptor and possible link to nephrogenic diabetes insipidus. Nature 1992;357:336–339. Marples D, Christensen S, Christensen EI, Ottosen PD, Nielsen S. Lithium-induced downregulation of aquaporin-2 water channel expression in rat kidney medulla. J Clin Invest 1995;95:1838–1845. Marples D, Frokiaer J, Dorup J, Knepper MA, Nielsen S. Hypokalemiainduced downregulation of aquaporin-2 water channel expression in rat kidney medulla and cortex. J Clin Invest 1996;97:1960–1968. Marples D, Knepper MA, Christensen EI, Nielsen S. Redistribution of aquaporin-2 water channels induced by vasopressin in rat kidney inner medullary collecting duct. Am J Physiol 1995;269:C655–C664.

673

Mulders SM, Knoers NV, Van Lieburg AF, et al. New mutations in the AQP2 gene in nephrogenic diabetes insipidus resulting in functional but misrouted water channels. J Am Soc Nephrol 1997;8:242–248. Nielsen S, Chou CL, Marples D, Christensen EI, Kishore BK, Knepper MA. Vasopressin increases water permeability of kidney collecting duct by inducing translocation of aquaporin-CD water channels to plasma membrane. Proc Soc Natl Acad Sci USA 1995;92:1013–1017. Nielsen S, DiGiovanni SR, Christensen EI, Knepper MA, Harris HW. Cellular and subcellular immunolocalization of vasopressin-regulated water channel in rat kidney. Proc Natl Acad Sci USA 1993;90: 11,663–11,667. Nielsen S, Smith BL, Christensen EI, Knepper MA, Agre P. CHIP28 water channels are localized in constitutively water-permeable segments of the nephron. J Cell Biol 1993;120:371–383. Pang Y, Metzenberg A, Das S, Jing B, Gitschier J. Mutations in the V2 vasopressin receptor gene are associated with X-linked diabetes insipidus. Nat Genet 1992;2:103–106. Robinson MG, Kaplan SA. Inheritance of vasopressin-resistant nephrogenic diabetes insipidus. Am J Dis Child 1960;99:164–174. Rosenthal W, Seibold A, Antaramian A, et al. Molecular identification of the gene responsible for congenital nephrogenic diabetes insipidus. Nature 1992;359:233–235. Sabolic I, Katsura T, Verbavatz JM, Brown D. The AQP2 water channel: effect of vasopressin treatment, microtubule disruption, and distribution in neonatal rats. J Memb Biol 1995;143:165–175. Sabolic I, Valenti G, Verbavatz J, et al. Localization of the CHIP28 water channel in rat kidney. Am J Physiol 1992;263:C1225–C1233. Sadeghi HM, Innamorati G, Birnbaumer M. An X-linked NDI mutation reveals a requirement for cell surface V2R expression. Mol Endocrinol 1997;11:706–713. Schoneberg T, Yun J, Wenkert D, Wess J. Functional rescue of mutant V2 vasopression receptors causing nephrogenic diabetes insipidus by a co-expressed receptor polypeptide. EMBO J 1996;15:1283–1291. Skorecki KL, Brown D, Ercolani L, Ausiello DA. Molecular mechanisms of vasopressin action in the kidney. In: Windhanger EE, ed. Handbook of Physiology: Section 8, Renal Physiology. New York: Oxford University Press, 1992; pp. 1185–1218. Strange K, Willingham MC, Handler JC, Harris HW Jr. Apical membrane endocytosis via coated pits is stimulated by removal of antidiuretic hormone from isolated, perfused rabbit cortical collecting tubule. J Membr Biol 1988;103:17–28. Terris J, Ecelbarger CA, Marples D, Knepper MA, Nielsen S. Distribution of aquaporin-4 water channel expression within rat kidney. Am J Physiol 1995;269:F775–F785. Valtin H. The discovery of the Brattleboro rat, recommended nomenclature, and the question of proper controls. Ann NY Acad Sci 1982; 394:1–9. van den Ouweland AMW, Dressen JCFM, Verdijk M, et al. Mutations in the vasopressin type 2 receptor gene (AVPR2) associated with nephrogenic diabetes insipidus. Nat Genet 1992;2:99–102. van Lieburg AF, Verdijk MA, Knoers VV, et al. Patients with autosomal nephrogenic diabetes insipidus homozygous for mutations in the aquaporin 2 water-channel gene. Am J Hum Genet 1994;55:648–652. Wade JB, Stetson DL, Lewis SA. ADH action: evidence for a membrane shuttle mechanism. Ann NY Acad Sci 1981;372:106–117. Wenkert D, Schoneberg T, Merendino JJ Jr, et al. Functional characterization of five V2 vasopressin receptor gene mutations. Mol Cell Endocrinol 1996;124:43–50. Williams RH, Henry C. Nephrogenic diabetes insipidus transmitted by females and appearing during infancy in males. Ann Int Med 1947;27:84–95. Yamamoto T, Sasaki S, Fushimi K, et al. Localization and expression of a collecting duct water channel, aquaporin, in hydrated and dehydrated rats. Exp Nephrol 1995;3:193–201.

CHAPTER 70 / POLYCYSTIC KIDNEY DISEASE

70

675

Polycystic Kidney Disease GREGORY G. GERMINO AND LUIZ F. ONUCHIC

BACKGROUND Renal cysts are common clinical findings, often incidentally discovered in the course of evaluating other problems. They may be either acquired or seen in association with a number of inherited and congenital disorders. The most common disorder, autosomal dominant polycystic kidney disease (ADPKD), is an important cause of end-stage renal disease (ESRD), accounting for 5–8% of the entire ESRD population. At the time of Dalgaard’s sentinel description of polycystic kidney disease (PKD), death from renal failure was a frequent outcome for affected individuals. While transplantation and dialysis have improved the prognosis for those suffering from PKD, no therapies have yet been discovered that prevent or inhibit cystogenesis. Hindering their development has been an incomplete knowledge of the pathogenesis of this process. Although most types of renal cysts share some common features (cell proliferation, basement membrane abnormalities, fluid secretion), the broad range of disorders associated with renal cysts suggests that defects in multiple, possibly intersecting, pathways are responsible for cyst formation and expansion. Recent molecular genetic studies have yielded important breakthroughs that are likely to lead to the unraveling of this mystery. In this chapter, we will review these data and how they have improved our understanding of renal cystic disease.

CLINICAL FEATURES The hallmark of this group of disorders is the replacement of normal renal parenchyma by cysts in a process that frequently results in renal impairment. The cysts vary greatly in number and size, ranging from microscopic lesions of several millimeters in diameter to very large, macroscopic structures >10 cm in width. In general, larger cysts are seen in ADPKD, tuberous sclerosis (TS), and von Hippel Lindau disease (VHL), whereas smaller cysts are more often associated with autosomal recessive polycystic kidney disease (ARPKD) and medullary cystic disease/nephronophthisis (MCDNP). The cysts increase in both number and volume in most of the disorders as the individual ages. Table 70-1 summarizes the features associated with each disorder. The nephron segment from which cysts are derived varies among the disorders and can, on occasion, be used to help establish a diagnosis. In ARPKD, the principal renal histopathologic abnormality is cystic expansion of all generations of the collecting ducts. From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

In the neonatal form of the disease, up to 90% of the collecting ducts are involved with cysts found in both the cortex and medulla. The kidneys tend to be massively but symmetrically enlarged and are usually palpable on exam. Ultrasonography typically reveals increased echogenicity with loss of corticomedullary differentiation. The cysts often are not detectable using radiologic imaging techniques. These severely affected children typically die within the first year of life. Children affected with a less severe form have a 10–60% involvement of their collecting ducts. Their kidneys are usually less enlarged and may develop macroscropic cysts. Hypertension, progressive renal insufficiency, and liver disease (portal tract fibrosis) are the primary manifestations of the disease in this older group. The cysts in MCDNP arise from the distal convoluted tubules and are found in the medulla and the cortico-medullary junction. Tubulo-interstitial inflammation and fibrosis are universal features and may be present even in the absence of renal cysts. The kidneys typically are small and hyperechoic on ultrasound. The cysts, which are small and heterogeneous in size, may be difficult to visualize without special methods that maximize resolution of the abdominal ultrasound or CT scan. Affected individuals often present with polyuria, polydypsia, and hyposthenuria, and develop progressive renal impairment. Autosomal recessive forms of the disease present in childhood and are important causes of ESRD in this age group. Rare families have been reported with a later onset of disease and an autosomal dominant pattern of inheritance. Renal cysts develop from all nephron segments in ADPKD. The cysts begin to form in utero and slowly increase in size and number during the lifetime of the individual. In early stages of the disorder, the kidney may be either normal in size or slightly enlarged. Progressive cystic disease typically distorts the structure of the kidney and results in chronic renal insufficiency in approximately 50% of affected individuals. Hypertension is a common problem and is associated with a greater likelihood of developing ESRD. There is considerable intra- and interfamilial variation in the severity of disease, with a number of families displaying progressive severity in successive generations. ADPKD occasionally presents with severe manifestations in childhood. These cases tend to have inherited the disease from an affected mother and often cluster within families. It is important to note that most of the diseases discussed above have associated extrarenal manifestations that may be major causes of morbidity or premature death (Table 70-1). Congenital hepatic fibrosis is an important cause of morbidity in children with ARPKD

675

676

Table 70-1 Genetic Renal Cystic Disorders—Clinical Features Disease

Inheritance

Kidney pathology

Extrarenal manifestations Liver cysts (40%), occasional cysts in other organs (pancreas, spleen), intracranial aneurysms, aortic aneurysms, cardiac valvular abnormalities, (mitral valve) prolapse) colonic diverticiculi Hepatic fibrosis, portal hypertension, hypertension, pulmonary hypoplasia

ADPKD

AD

All nephron segments cyst size variable, often >1 cm

Recurrent infections, hematuria, nephrolithiasis, renal failure

ARPKD

AR

Abdominal mass, hematuria, concentrating defect renal failure

Bardet-Biedl Syndrome

AR

MCDNP

AR

Senior-Loken

AR AD AD

All portions of the collecting duct 90% involvement, infantile form 10–60% involvement, later forms cysts usually 4 cm) cutaneous squamous-cell carcinomas were common among patients in their late 20s and early 30s and less than 10% survived beyond 30 years. Lastly, genetic counseling should be made available to all families. Prenatal diagnosis of type IA OCA can be performed as outlined above. Because there is currently no treatment that will reverse the major forms of this genetic disorder, both patients and their parents can receive valuable information and support by joining NOAH, the National Organization for Albinism and Hypopigmentation (1530 Locust Street #29, Philadelphia, PA

CHAPTER 79 / OCULOCUTANEOUS ALBINISM

19102-4415). Allogeneic bone marrow transplantation has been used successfully in patients with CHS and 1-desamino-8D-arginine (DDAVP) has been reported as a potential treatment for the bleeding diathesis associated with HPS.

FUTURE DIRECTIONS Just as genetic studies in albino mice and pink-eyed dilution mice shed a great deal of light on the molecular pathophysiology of type IA OCA and type II OCA, respectively, so will the studies of brown and beige mice provide insights into types IV and VIB OCA. Given the dysfunction of several organelles including melanosomes, platelet-storage granules, and lysosomes in patients with HPS and CHS, it is anticipated that the protein products of the genes responsible for these disorders will play an important role in the formation and/or function of intracytoplasmic organelles. The genetic mutation in the silver mouse leads to an abnormal melanosomal matrix protein as well as altered pigmentation, and it is possible that a similar situation will exist in humans. Lastly, the slaty mouse has been shown to have a mutation in the TRP-2 gene, whose protein product is dopachrome tautomerase (Table 79-2), an enzyme involved in the melanin biosynthetic pathway that converts dopachrome to DHICA (Figs. 79-1 and 79-8). The human TRP-2 gene has been cloned and is located on chromosome 13. Although the albino phenotype has been rescued in transgenic mice (see above), gene therapy for humans with type IA OCA represents futuristic thinking, especially since the majority of pigment production in the eye occurs prior to birth.

SELECTED REFERENCES Bailin T, Oh J, Feng GH, Fukai K, Spritz RA. Organization and nucleotide sequence of the human Hermansky-Pudlak syndrome (HPS) gene. J Invest Dermatol 1997;108:923–927. Barbosa MDFS, Nguyen QA, Tchernev VT, et al. Identification of the homologous beige and Chediak-Higashi syndrome genes. Nature 1996;382:262–265. Barrat FJ, Auloge L, Pastural E, et al. Genetic and physical mapping of the Chekiak-Higashi syndrome on chromosome 1q42-43. Am J Hum Genet 1996;59:625–632. Bassi MT, Schiaffino MV, Renieri A, et al. Cloning of the gene for ocular albinism type I from the distal short arm of the X chromosome. Nature Genet 1995;10:13–19. Boissy RE, Zhao H, Oetting WS, et al. Mutation in and lack of expression of tyrosinase-related protein-1 (TRP-1) in melanocytes from an individual with brown oculocutaneous albinism: a new subtype of albinism classified as “OCA3”. Am J Hum Genet 1996;58:1145–1156. Bolognia JL, Shapiro PE. Albinism and other disorders of hypopigmentation. In: Arndt KA, LeBoit PE, Robinson JK, Wintrobe BU, eds. Cutaneous Medicine and Surgery: An Integrated Program in Dermatology. Philadelphia: Saunders, 1995; pp. 1219–1232. del Marmol V, Beermann F. Tyrosinase and related proteins in mammalian pigmentation. FEBS 1996;381:165–168. Durham-Pierre D, Gardner JM, Nakatsu Y, et al. African origin of an intragenic deletion of the human P gene in tyrosinase positive oculocutaneous albinism. Nature Genet 1994;7:176–179. Falik-Borenstein TC, Holmes SA, Borochowitz Z, Levin A, Rosenmann A, Spritz RA. DNA-based carrier detection and prenatal diagnosis of tyrosinase-negative oculocutaneous albinism (OCA1A). Prenatal Diagnosis 1995;15:345–349. Feng GH, Bailin T, Oh J, Spritz RA. Mouse pale ear (ep) is homologous to human Hermansky-Pudlak syndrome and contains a rare ‘AT-AC’ intron. Hum Mol Genet 1997;6:793–797. Fukai K, Holmes SA, Lucchese NJ, et al. Autosomal recessive ocular albinism associated with a functionally significant tyrosinase gene polymorphism. Nature Genet 1995;9:92–95.

743

Fukai K, Oh J, Karim MA, et al. Homozygosity mapping of the gene for Cheiak-Higashi syndrome to chromosome 1q42-q44 in a segment of conserved synteny that includes the mouse beige locus (bg). Am J Hum Genet 1996;59:620–624. Gahl WA, Pottere B, Durham-Pierre, Brilliant MH, Hearing VJ. Melanosomal tyrosine transport in normal and pink-eyed dilution murine melanocytes. Pigment Cell Res 1995;8:229–233. Giebel LB, Spritz RA. The molecular basis of type I (tyrosinase-deficient) human oculocutaneous albinism. Pig Cell Res 1992;2(Suppl): 101–106. Giebel LB, Tripathi RK, King RA, Spritz RA. A tyrosinase gene missense mutation in temperature-sensitive type I oculocutaneous albinism. J Clin Invest 1991;87:1119–1121. King RA, Townsend D, Oetting W, et al. Temperature-sensitive tyrosinase associated with peripheral pigmentation in oculocutaneous albinism. J Clin Invest 1991;87:1046–1053. Kobayashi T, Urabe K, Winder A, et al. DHICA oxidase activity of TRP1 and interactions with other melanogenic enzymes. Pigment Cell Res 1994;7:227–234. Körner A, Pawelek J. Mammalian tyrosinase catalyzes three reactions in the biosynthesis of melanin. Science 1982;217:1163–1165. Kwon BS. Pigmentation genes: the tyrosinase gene family and pmel 17 gene family. J Invest Dermatol 1993;100:134S–140S. Lee S-T, RD Nicholls, S. Bundey, Laxova R, Musarella M, Spritz RA. Mutations of the P gene in oculocutaneous albinism, ocular albinism, and Prader-Willi syndrome plus albinism. N Engl J Med 1994; 330:529–534. Lee S-T, Nicholls RD, Jong MTC, Fukai K, Spritz RA. Organization and sequence of the human P gene and identification of a new family of transport proteins. Genomics 1995;26:354–363. Luande J, Henschke CI, Mohammed N. The Tanzanian human albino skin. Natural history. Cancer 1985;55:1823–1828. Nagle DL, Karim MA, Woolf EA, et al. Identification and mutation analysis of the complete gene for Chediak-Higashi syndrome. Nature Genet 1996;14:307–311. Oetting WS, King RA. Molecular basis of type I (tyrosinase-related) oculocutaneous albinism: mutations and polymorphisms of the human tyrosinase gene. Hum Mutation 1993;2:1–6. Oetting WS, King RA. Analysis of tyrosinase mutations associated with tyrosinase-related oculocutaneous albinism (OCA1). Pig Cell Res 1994;7:285–290. Oh J, Bailin T, Fukai K, et al. Positional cloning of a gene for HermanskyPudlak syndrome, a disorder of cytoplasmic organelles. Nat Genet 1996;14:300–306. Rinchik EM, Bultman SJ, Horsthemke B, et al. A gene for the mouse pinkeyed dilution locus and for human type II oculocutaneous albinism. Nature 1993;361:72–76. Rosemblatt S, Durham-Pierre D, Gardner JM, Nakatsu Y, Brilliant MH, Orlow SJ. Identification of a melanosomal membrane protein encoded by the pink-eyed dilution (type II oculocutaneous albinism) gene. Proc Natl Acad Sci USA 1994;91:12,071–20,075. Shimizu H, Niizeki H, Suzumori K, et al. Prenatal diagnosis of oculocutaneous albinism by analysis of the fetal tyrosinase gene. J Invest Dermatol 1994;103:104–106. Spritz RA. Molecular genetics of oculocutaneous albinism. Hum Mol Genet 1994;3:1469–1475. Spritz RA, Bailin T, Nicholls RD, et al. Hypopigmentation in the PraderWilli syndrome correlates with P gene deletion but not with haplotype of the hemizygous P allele. Am J Med Genet 1997;71:57–62. Spritz RA, Fukai K, Holmes SA, Luande J. Frequent intragenic deletion of the P gene in Tanzanian patients with type II oculocutaneous albinism (OCA2). Am J Hum Genet 1995;56:1320–1323. Stevens G, Ramsay JM, Jenkins T. Oculocutaneous albinism (OCA2) in sub-Saharan Africa: distribution of the common 2.7-kb P gene deletion mutation. Hum Genet 1997;99:523–527. Wagstaff J, Hemann M. A familial “balanced” 3;9 translocation with cryptic 8q insertion leading to deletion and duplication of 9p23 loci in siblings. Am J Hum Genet 1995;56:302–309.

CHAPTER 80 / BASAL-CELL NEVUS SYNDROME

80

745

Basal Cell Nevus Syndrome ERVIN EPSTEIN, JR.

BACKGROUND The basal cell nevus syndrome (BCNS) (nevoid basal-cell carcinoma syndrome, Gorlin syndrome, McKusick MIM 109400) is a rare, multisystem, heritable disorder. Evidence exists for its presence in Egyptian mummies, and cases were described initially a century ago. The association of multiple basal-cell carcinomas with jaw cysts and skeletal abnormalities was well-described by midcentury, and subsequent descriptions have produced an extensive list of phenotypic abnormalities. Case reports by Howell and Caro in 1959 and by Gorlin and Goltz in 1960 focused much more widespread attention on the condition, and more recently the interest of several groups in molecular analysis of the syndrome has stimulated the collection of larger groups of patients and consequent better clinical characterization. Two recent estimates of disease frequency are 1:56,000 and 1:164,000 or greater.

CLINICAL FEATURES Phenotypic abnormalities affect many organ systems. Individual patients have varying combinations of abnormalities, and intrafamilial variation is prominent. A listing of some of the abnormalities is given in Table 80-1. SKIN The most notable skin abnormalities are basal-cell carcinomas (BCCs). These are the commonest human tumor—perhaps a half million are treated each year in the United States, and in sunny areas their incidence plus that of cutaneous squamouscell carcinomas may exceed the combined incidence of all other cancers. Although the dose–reponse relationship is not straightforward, it is clear that skin damage by sunlight predisposes to BCCs. Thus, they occur frequently on sun-exposed areas and much less frequently on covered sites; they occur more commonly in sunnier areas of the world; they occur commonly in individuals with white skin and seldom in patients with darker skin (although albinos of African or Asian descent are susceptible to their development); and they occur in high numbers in patients who have defective repair of ultraviolet radiation-induced DNA damage (i.e., patients with xeroderma pigmentosum). Typically, BCCs occur in individuals of middle-to-later ages and are one to several in number. Some individual BCCs in patients with BCNS may have the pearly, translucent, telangectactic, papular appearance of BCCs in sporadic cases but others may appear as tiny tag-like papules that may be so widespread that they are dismissed unless the examinaFrom: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

tion is quite careful. Histologically they are indistinguishable from BCCs in nonfamilial cases. The role of sunlight in producing BCCs in BCNS patients, like that in sporadic cases, is not straightforward. An active role of sunlight is suggested by the reported earlier age of appearance of the first BCCs in BCNS patients in Australia than in the North West of England (47 vs 4% by age 20), the small numbers of BCCs in black-skinned BCNS patients, and the predominant localization of BCCs on sun-exposed skin sites of patients with BCNS. However, the latter localization is less strict than in patients without BCNS, for the percentage of BCCs occurring in normally covered areas (i.e., not the face and arms) is higher in BCNS patients than in sporadic patients—42 vs 13%. BCCs in BCNS patients have metastasized to the lungs and heart, as in sporadic cases, such occurrences are quite rare. Uncontrollable local invasion is less rare but still very uncommon—4 of 70 patients in one series suffered facial mutilation such as loss of an eye or an ear. The other very common skin abnormality in BCNS patients is the presence of pits of the palms and soles (Fig. 80-1). These typically appear during the second decade, are 1–2 mm in diameter, and may be made more prominent by soaking the skin in tapwater for 15 min, thus causing swelling of the surrounding normal stratum corneum. The pits are formed by local absence of the stratum corneum. Generally, the epidermis underlying the pits is unremarkable, but BCCs have arisen from them. BCNS patients also have a higher than normal incidence of epidermal inclusion cysts—perhaps 50%. Usually these are histologically unremarkable. JAW The most diagnostic extracutaneous abnormality is the presence of one or more cysts in the jaws. These cysts are lined by a stratified-squamous epithelium that produces parakeratotic keratinous debris and may have satellite epithelial cysts and/or islands resembling BCCs. They may be clinically silent and detected only by X-ray examination or may cause localized swelling or pain of the jaw. Their histology closely overlaps that of the odontogenic keratocysts of sporadic cases. Recurrences postoperatively are common—10/16 in one series—and loss or displacement of teeth because of local invasion of bone can occur. Malignant changes (development of squamous-cell carcinoma) in the lining are very rare. These keratocysts can develop at any age but they appear most commonly in the later part of the first decade through the second and third decade, and they are more common in the mandible than the maxilla. SKELETAL Calcification of the falx is present commonly in patients—in one report in 100% of patients 20 years of age and older as compared to a percentage that increases with age in nor-

745

746

SECTION IX / DERMATOLOGY

Table 80-1 Diagnostic Findings in Adults with BCNS 50% or greater frequency Enlarged occipitofrontal circumference (macrocephaly, frontoparietal bossing, in index cases only?) Multiple basal-cell carcinomas Odontogenic keratocysts of jaws Epidermal cysts of skin High-arched palate Palmar or plantar pits Rib anomalies (e.g., splayed, fused, partially missing, bifid) Spina bifida occulta of cervical or thoracic vertebrae Calcified falx cerebri Calcified diaphragma sellae (bridged sella, fused clinoids) Hyperpneumatization of paranasal sinuses 49 to 15% frequency Calcification of tentorium cerebelli and petroclinoid ligament Calcified ovarian fibromas Short fourth metacarpals Kyphoscoliosis or other vertebral anomalies Lumbarization of sacrum Narrow sloping shoulders Prognathism Pectus excavatum or carinatum Pseudocystic lytic lesion of bones (hamartomas) Strabismus (exotropia) 14% or less, but not random Medulloblastoma Inguinal hernia (?) True ocular hypertelorism Meningioma Lymphomesenteric cyst Cardiac fibroma Fetal rhabdomyoma Ovarian fibrosarcoma Marfanoid build Anosmia Agenesis of corpus callosum Cyst of septum pellucidum Cleft lip and/or palate Low-pitched female voice Polydactyly, postaxial in hands or feet Sprengel deformity of scapula Syndactyly Congenital cataract, glaucoma, coloboma of iris, retina, optic nerve, medullated retinal nerve fibers Subcutaneous calcifications of skin (possibly underestimated frequency) Minor kidney malformations Hypogonadism in male subjects Mental Retardation Reprinted with permission from Gorlin, R.J. Nevoid basal cell carcinoma syndrome. Dermatologic Clinics 1995;13:113–125.

mal individuals from 10 to 50% between ages 20 and 70. Calcification is more dense, even exuberant, in BCNS patients and is described as lamellar. Other intracranial membranes also may be calcified. Dysmorphic changes also are reported frequently. One of the most common is an increased head size, but one recent study found increased head size only in index cases and not in their affected relatives. Large supraorbital ridges and wide-set eyes also are reported frequently as part of a characteristic facies. Ribs commonly are abnormal—bifid, flared, fused, hypoplastic, or miss-

Figure 80-1 Right palm of patient with BCNS. The typical, distinctive, 2-mm in diameter pits in the stratum corneum are visible as darker spots.

ing—and rib abnormalities and/or spina bifida occulta of the cervical or thoracic spine occur in approximately 50% of patients. The sternum may be concave or convex, and numerous other malformations have been cataloged. Some patients have a marfanoid habitus, and many adult patients are considerably bulkier than their parents. Phalangeal lucencies were common in one series in which hand X-rays were studied, and sclerotic bone changes resembling those of metastases have been reported. VISCERAL TUMORS Medulloblastomas are the most serious extracutaneous tumors present in clearly higher frequency in BCNS patients. Like BCCs, they tend to arise at a younger age in BCNS patients than in sporadic cases (most before age 5 and half before age 2, at which age sporadic medulloblastomas are unusual). In larger series of BCNS patients, medulloblastomas have been found in 1–5% of patients, and the incidence of BCNS in patients with medulloblastomas is also 1–5%. BCNS patients with meningiomas have been reported but less commonly than those with medulloblastomas. By contrast, ovarian fibromas are common, and these tumors frequently are calcified. They usually are asymptomatic and do not interfere with ovarian function. Cysts of the mesentery may be present, sometimes with calcification leading to notable radiological abnormalities. Several BCNS patients have had cardiac fibromas, some with onset in infancy, and in one instance a fibrous histiocytoma became so large as to necessitate cardiac transplantation.

DIAGNOSIS Diagnosis of the classical case is straightforward, but the phenotypic variability of affected patients and the fact that all of the abnormalities may be seen, albeit less frequently, in otherwise normal individuals may make it difficult to be certain of the diagnosis, even in relatives of patients whose own diagnosis is indubitable. Both of two recent larger series proposed four major diagnostic criteria— BCCs, odontogenic keratocysts, palmoplantar pits, and lamellar calcification of the falx—and minor criteria including rib anomalies, medulloblastomas, cardiac or ovarian fibromas, and several of the other abnormalities listed in Table 80-1. For diagnosis, two major or one major plus two minor criteria are required. Without an inexpensive and sensitive screening for the underlying molecule defect, phenotype-based diagnosis is likely to continue to be difficult for many patients with less than classical findings.

CHAPTER 80 / BASAL-CELL NEVUS SYNDROME

GENETIC BASIS OF DISEASE BCNS is inherited as an autosomal-dominant trait. New mutations are common (14–81% in one large series, this wide range reflecting the difficulty of being certain that unexamined parents lack all phenotypic abnormalities of the syndrome). The lack of obvious differences in phenotype according to whether disease has been inherited from the mother or the father suggests that imprinting does not underlie the phenotypic variability. The autosomal-dominant inheritance and the cardinal manifestation of the development of BCCs in higher number and at an earlier age than in sporadic cases suggested to several observers more than a decade ago the possibility that the gene whose mutations underlie the BCNS might be a tumor-suppressor gene and that, like retinoblastomas, BCCs might require two “hits” for their development. Since one frequent mechanism of tumor-suppressor gene inactivation is allelic deletion, A. Bale and colleagues assessed BCCs arising in sporadic cases for such deletions. Indeed, they found a high incidence of loss of heterozygosity at chromosome 9q and then found significant linkage of the inheritance of BCNS to the inheritance of 9q22.3–9q31. This was confirmed rapidly by other groups. Further studies of loss of heterozygosity have confirmed an approximately 50% incidence of loss of this region with relatively little loss at other loci except at chromosome 1q. In patients with BCNS, the copy of chromosome 9q22.3–9q31 lost in BCCs is that predicted by linkage to contain the normal allele. Thus, these tumors are left with a single copy of this region, and that copy is predicted to harbor a nonfunctional allele. No BCNS kindred has been reported in which the mutant gene fails to map to the same site on chromosome 9q, suggesting that there is a single locus whose mutations underlie all BCNS patients. Subsequent genetic mapping narrowed the region to an estimated 1.8–3.6 cM; YAC-BAC-cosmid contigs spanning the genetically delimited region were assembled; and several groups embarked on a search for isolating genes in this region. This search was capped in 1996 by the simultaneous reports by two groups of mapping of the human homolog of the Drosophila patched gene to the genetically defined BCNS region and of mutations of this gene in DNA both from leukocytes of BCNS patients and also from sporadic basal-cell carcinomas. Thus, it appears that this gene functions, at least in the skin, as a tumor-suppressor gene, that patients with the BCNS inherit one defective copy, and that somatic mutations that inactivate in keratinocytes the one functional allele in BCNS patients or both normal alleles in sporadic cases are a crucial step underlying the development of basal-cell carcinomas that arise sporadically. The nature of the mutations described in the skin tumors is consistent with some of them having been caused by ultraviolet radiation, as expected. The fly patched gene has been studied extensively since its identification in 1989. It is a membrane-bound member of a signaling pathway that is critical in development. It acts to suppress expression of several genes, including cubitus interruptus, a member of the GLI gene family (so named because of the identification of GLI as an oncogene in human glioblastomas); wingless, a member of the WNT oncogene family; decapentaplegic, a member of the TGF-β gene family; as well as patched itself. It is opposed in these functions by the secreted product of the hedgehog gene. This pathway helps control segmentation and wing development in flies and appears to play analogous roles in vertebrates. Mutations also have been described in BCCs in the genes encoding sonic hedgehog and smoothened, the other two members

747

of this pathway that interact with the patched protein at the cell surface. The mutant proteins appear to act dominantly as the products of activated oncogenes, and this identification provides evidence that it is activation of the signaling pathway that determines the abnormal behavior of keratinocytes. Thus it seems reasonable now to consider the phenotypic abnormalities in BCNS patients in two categories—developmental (e.g., rib and spine abnormalities and generalized overgrowth) and tumors (e.g., BCCs, medulloblastomas, fibromas, cysts) and to consider that the PATCHED gene functions in humans embryologically to direct development and postnatally to control proliferation and/ or differentiation, i.e., as a tumor suppressor. Many questions are raised by this gene identification, and the role of the hedgehog/patched signaling pathway in humans is likely to be a subject of intense investigation during the next several years.

TREATMENT Individual BCCs in BCNS patients can be treated by the surgical approaches used for BCCs in sporadic cases. The often high number of tumors may make their excision impractical, and, e.g., cryotherapy with liquid nitrogen may be helpful for the many small lesions, especially those that are well-removed from the eye, ear, and other sites where BCCs may be particularly destructive. X-irradiation should be used with extreme caution because BCNS patients clearly are susceptible to X-ray carcinogenesis. Thus, patients successfully irradiated for medulloblastomas typically develop large numbers of BCCs in the overlying skin within a half decade. Also, instances of squamous-cell carcinomas and of fibrosarcomas with metastases have occurred soon after therapeutic irradiation of BCCs. Several reports have documented significant reduction in the appearance of new BCCs during systemic treatment with 13-cisretinoic acid but typically the side effects at the doses required to inhibit BCC growth eventually become intolerable. Topical retinoic acid and topical 5-fluorouracil may slow the growth of tumors.

SELECTED REFERENCES Bale SJ, Amos CI, Parry DM, Bale AE. Relationship between head circumference and height in normal adults and in the nevoid basal cell carcinoma syndrome and neurofibromatosis type I. Am J Med Gen 1991;40:206–210. Binkley GW, Johnson HHJ. Epithelioma adenoides cysticum: basal cell nevi, agenesis of the corpus callosum and dental cysts. Arch Dermat Syph 1951;63:73–84. Bonifas JM, Bare JW, Kerschmann RL, Master SP, Epstein EH Jr. Parental origin of chromosome 9q22.3-q31 lost in basal cell carcinomas from basal cell nevus syndrome patients. Hum Mol Genet 1994;3: 447,448. Chenevix-Trench G, Wicking C, Berkman J, et al. Further localization of the gene for nevoid basal cell carcinoma syndrome (NBCCS) in 15 Australasian families: linkage and loss of heterozygosity. Am J Hum Genet 1993;53:760–767. Compton JG, Goldstein AM, Turner M, et al. Fine mapping of the locus for nevoid basal cell carcinoma syndrome on chromosome 9q. J Invest Dermatol 1994;103:178–181. Dunnick NR, Head GL, Peck GL, Yoder FW. Nevoid basal cell carcinoma syndrome: radiographic manifestations including cystlike lesions of the phalanges. Radiology 1978;127:331–334. Evans DGR, Farndon PA, Burnell LD, Gattamaneni HR, Birch JM. The incidence of Gorlin syndrome in 173 consecutive cases of medulloblastoma. Br J Cancer 1991;64:959–961. Evans DGR, Ladusans EJ, Rimmer S, Burnell LD, Thakker N, Farndon PA. Complications of the naevoid basal cell carcinoma

748

SECTION IX / DERMATOLOGY

syndrome: results of a population based study. J Med Genet 1993;30: 460–464. Farndon PA, Morris DJ, Hardy C, et al. Analysis of 133 meioses places the genes for nevoid basal cell carcinoma (Gorlin) syndrome and Fanconi anemia group C in a 2.6-cM interval and contributes to the fine map of 9q22.3. Genomics 1994;23:486–489. Forssell K, Forssell H, Kahnberg K-E. Recurrence of keratocysts: a longterm follow-up study. Int J Oral Maxillofacial Surg 1988;17:25–28. Gailani MR, Bale SJ, Leffell DJ, et al. Developmental defects in Gorlin syndrome related to a putative tumor suppressor gene on chromosome 9. Cell 1992;69:111–117. Goldstein AM, Bale SJ, Peck GL, DiGiovanna JJ. Sun exposure and basal cell carcinomas in the nevoid basal cell carcinoma syndrome. J Am Acad Dermat 1993;29:34–41. Goldstein AM, Pastakia B, DiGiovanna JJ, et al. Clinical findings in two African-American families with the nevoid basal cell carcinoma syndrome (NBCC). Am J Med Genet 1994;50:272–281. Goldstein AM, Stewart C, Bale AE, Bale SJ, Dean M. Localization of the gene for the nevoid basal cell carcinoma syndrome. Am J Hum Genet 1994;54:765–773. Golitz LE, Norris DA, Luekens CA Jr, Charles DM. Nevoid basal cell carcinoma syndrome: multiple basal cell carcinomas of the palms after radiation therapy. Arch Dermat 1980;116:1159–1163. Goodrich LV, Milenkovic L, Higgins KM, Scott MP. Altered neural cell fates and medulloblastoma in mouse patched mutants. Science 1997;277:1109–1113. Gorlin RJ. Nevoid basal cell carcinoma syndrome. Dermat Clinics 1995;13:113–125. Gorlin R.J. Nevoid Basal-Cell Carcinoma syndrome. Medicine 1987; 66:98–113. Gorlin RJ, Goltz RW. Multiple nevoid basal-cell epithelioma, jaw cysts and bifid ribs: a syndrome. N Engl J Med 1960;262:908–912. Hahn H, Wicking C, Zaphiropoulous PG, et al. Mutations of the human homologue of Drosophila patched in the nevoid basal cell carcinoma syndrome. Cell 1996;85:841–851. Herman TE, Siegel MJ, McAlister WH. Cardiac tumor in Gorlin syndrome. Pediatric Radiol 1991;21:234,235. Howell JB. Nevoid basal cell carcinoma syndrome: profile of genetic and environmental factors in oncogenesis. J Am Acad Dermat 1984;11:98–104. Howell JB, Caro MR. Basal cell nevus: its relationship to multiple cutaneous cancers and associated anomalies of development. Arch of Dermat 1959;79:67–80. Johnson AD, Hebert AA, Esterly NB. Nevoid basal cell carcinoma syndrome: bilateral ovarian fibromas in a 3 1/2-year-old girl. J Am Acad Dermat 1986;14:371–374. Johnson RL, Rothman AL, Xie J, et al. Human homolog of patched, a candidate gene for the basal cell nevus syndrome. Science 1996;272:1668–1671. Leppard BJ. Skin cysts in the basal cell naevus syndrome. Clin Exp Dermat 1983;8:603–612.

Morris DJ, Reis A. A YAC contig spanning the nevoid basal cell carcinoma syndrome, Fanconi anaemia group C, and xeroderma pigmentosum group A loci on chromosome 9q. Genomics 1994;23:23–29. Oro AE, Higgins KM, Hu Z, Bonifas JM, Epstein EH Jr, Scott MP. Basal cell carcinomas in mice overexpressing sonic hedgehog. Science 1997;276:817–821. Peck GL, DiGiovanna JJ, Sarnoff DS, et al. Treatment and prevention of basal cell carcinoma with oral isotretinoin. J Am Acad Dermat 1988;19:176–185. Ratcliffe JF, Shanley S, Chenevix-Trench G. The prevalence of cervical and thoracic congenital skeletal abnormalities in basal cell naevus syndrome; a review of cervical and chest radiographs in 80 patients with BCNS. Br J Radiol 1995;68:596–599. Ratcliffe JF, Shanley S, Ferguson J, Chenevix-Trench G. The diagnostic implication of falcine calcification on plain skull radiographs of patients with basal cell naevus syndrome and the incidence of falcine calcification in their relatives and two control groups. Br J Radiol 1995;68:361–368. Shanley S, Ratcliffe J, Hockey A. Nevoid basal cell carcinoma syndrome: review of 118 affected individuals. Amer J Med Genet 1994;50: 282–290. Southwick, GJ, Schwartz RA. The basal cell nevus syndrome: disasters occurring among a series of 36 patients. Cancer 1979;44:2294–2305. Strange PR, Lang PG. Long-term management of basal cell nevus syndrome with topical tretinoin and 5-fluorouracil. J Am Acad Dermat 1992;27:842–845. Vorechovsky I, Tingby O, Hartman M, et al. Somatic mutations in the human homologue of Drosophila patched in primitive neuroectodermal tumours. Oncogene 1997;15:361–366. Wicking C, Berkman J, Wainwright B, Chenevix-Trench G. Fine genetic mapping of the gene for nevoid basal cell carcinoma syndrome. Genomics 1994;22:505–511. Wicking C, Shanley S, Smyth I, et al. Most germ-line mutations in the nevoid basal cell carcinoma syndrome lead to a premature termination of the PATCHED protein, and no genotype-phenotype correlations are evident. Am J Hum Genet 1997;60:21–26. Wolter M, Reifenberger J, Sommer C, Ruzicka T, Reifenberger G. Mutations in the human homologue of the Drosophila segment polarity gene patched (PTCH) in sporadic basal cell carcinomas of the skin and primitive neuroectodermal tumors of the central nervous system. Cancer Res 1997;57:2581–2585. Xie J, Johnson RL, Zhang X, et al. Mutations of the PATCHED gene in several types of sporadic extracutaneous tumors. Cancer Res 1997; 57:2369–2372. Xie J, Quinn A, Zhang X, et al. Physical mapping of the 5 Mb D9S196D9S180 interval harboring the basal cell nevus syndrome gene and localization of six genes in this region. Genes Chromosomes Cancer 1997;18:305–309. Xie J. Nature 1998, in press.

CHAPTER 81 / XERODERMA PIGMENTOSUM

81

749

Xeroderma Pigmentosum and Related Disorders W. CLARK LAMBERT, HON-REEN KUO, AND MURIEL W. LAMBERT

INTRODUCTION Inherited diseases with cutaneous manifestations currently believed to be associated with defective DNA repair and/or chromosomal instability, of which xeroderma pigmentosum is the prototype, are all very rare. The more important of these, discussed here, are listed in Table 81-1. Diseases above the open space are all (mainly) autosomal recessive disorders; those below it, autosomal dominant. Of the recessive diseases, there are four major entities: (1) xeroderma pigmentosum, which may occur together with a distinctive autosomal recessive disorder, Cockayne disease, and the cells of which may show some very similar changes, with failure of complementation, with those of cells from some cases with yet a third disorder, trichothiodystrophy; (2) Fanconi anemia, an important variant of which, lacking only the congenital anomalies characteristic of Fanconi anemia, is the DiamondBlackfan syndrome (also known as the Estren-Dameshek variant of Fanconi anemia); (3) Ataxia-Telangiectasia, which shares distinctive cell culture but not clinical characteristics with two other disorders, the Nijmegen breakage syndrome and the Berlin breakage syndrome, which complement each other in complementation assays but are otherwise identical; and (4) Bloom syndrome. There is also less-well-documented evidence for defective DNA repair and/or chromosomal instability in a number of other disorders, such as progeria of various subtypes and dyskeratosis congenita. It is likely that the genetics of some of these rare recessive diseases is much more complex than the simple phrase “autosomal recessive” would imply. A number of cell lines from different ones of these disorders test positive in at least one assay, in which chromosome abnormalities are scored following irradiation in the G2 phase of the cell cycle. This implies that these diseases are indeed related to each other. There is convincing evidence for both intragenic and intergenic genetic heterogeneity within several of these entities. Thus, different involved loci or combinations of involved loci may produce different genetic subtypes of the same disease, or even subtypes of different diseases occurring together. It is possible that the extremely complex nature of chromatin structure, with different proteins often playing more than one role, interacting with each other and possibly partially compensating for any loss or deficiency of one another, may result in specific From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

diseases being traced only to very severe—or combined— deficiencies in these proteins. Many of the nuclear proteins identified to date as defective in one or more of these diseases are involved not only in DNA repair and maintenance of chromatin stability but also in such other cellular processes as initiation of transcription, DNA replication, and cell-cycle checkpoint controls. This may provide an explanation for some of the pleiotropic features of these disorders. A number of much more common diseases may also be related etiopathologically to less extreme defects in chromatin proteins and nuclear enzymes, but this is yet to be conclusively proved.

XERODERMA PIGMENTOSUM BACKGROUND Xeroderma pigmentosum is a rare, [usually] autosomal recessively transmitted disease characterized by variable but usually extreme sun sensitivity associated with a marked tendency to develop precancerous lesions and, subsequently, skin cancers at a [usually] very early age in sun-exposed areas of the skin, lips, conjunctivae, and, sometimes, anterior tongue. These changes are absolutely dependent on sun exposure; sun-protected areas do not develop these changes. A subset of patients develop a progressive neurodegenerative disease, not related to sun exposure. The disease is genetically heterogeneous, with eight complementation groups known, labeled A, B, C, D, E, F, G, and V (see below). Individuals in groups A, B, and G are the most sun-sensitive, whereas those in groups A and D are most prone to develop neurological disease. The first description of the disease, under the name “xeroderma” was in 1870 by Moritz Kaposi in a textbook edited by his father-in-law, Ferdinand von Hebra, professor of dermatology at the University of Vienna. Twelve years later he described additional patients and added the term “pigmentosum” to the name of the disease. The first case with neurological signs was described by Dr. Albert Neisser, better known for his work on the Neisseria species of bacteria named for him. In 1932, De Sanctis and Cacchione described a severely neurodegenerate case with microcephaly and mental retardation. The term “De-Sanctis-Cacchione syndrome” has subsequently been used to designate cases of xeroderma pigmentosum with a neurologic component. However, we believe that this term should only be applied in cases of severe neurologic deficiency, such as described by De Sanctis and Cacchione, with the remainder simply known as xeroderma pigmentosum with a neurological defect or component.

749

750

SECTION IX / DERMATOLOGY

Table 81-1 Skin Disorders Associated with Defective DNA Repair and/or Chromosomal Instability Xeroderma pigmentosum Cockayne syndrome Some trichothiodystrophies Fanconi anemia Diamond-Blackfan syndrome (Estren-Dameshek variant of Fanconi anemia) Ataxia-telangiectasia (Louis-Barr syndrome) Nijmegen breakage syndrome Berlin breakage syndrome Ataxia-telangiectasia Fresno Bloom syndrome Gardner syndrome (variant of Familial adenomatous polyposis [FAP] syndrome) Turcot syndrome (variant of familial adenomatous polyposis [PAP] syndrome) Muir-Torre syndrome (variant of the Human nonpolyposis colon carcinoma [HNPCC] syndrome) Other diseases are also known with less well characterized DNA repair defects and/a chromosomal instabilities

Figure 81-1 Face of a 22-month-old white male with severely sun-sensitive xeroderma pigmentosum. Note early changes, as well as a neoplasm (keratoacomthoma) on the cheek and a dysplastic melanocytic lesion on his nose. (Reproduced with permission from Lambert et al., 1998.)

The study of xeroderma pigmentosum at the cellular/biochemical level has generated many justified “firsts” in the study of human genetic diseases. In 1968, James Cleaver reported that fibroblasts derived from a patient with xeroderma pigmentosum were unable to perform the nucleotide excision type of DNA repair following exposure to short wavelength ultraviolet C light emitted by a germicidal lamp. This finding was confirmed shortly afterwards by Richard Setlow and his coworkers in cultured fibroblasts, and by John Epstein and his collaborators in an in vivo assay carried out in the skin of a xeroderma pigmentosum patient. This was the first indication of a DNA repair defect associated with a human disease. In 1972, Dirk Bootsma and his laboratory modified the autoradiography assay, which had been used to detect defective DNA repair in fibroblasts, to develop a complementation assay, which demonstrated genetic heterogeneity within xeroderma pigmentosum. This was the first such cellular/biochemical assay developed for examination of genetic heterogeneity in a human genetic disease. Rufus Day and his coworkers modified a host-cell reactivation viral assay developed in bacteria and Michael Sideman and Kenneth Kraemer and their colleagues used a further modification of this assay, the shuttle vector assay, to examine DNA repair and mutagenesis in normal and xeroderma pigmentosum cells in culture. In 1982 our laboratory, and in 1988 Richard Wood and his colleagues developed cell-free assay systems for nucleotide excision repair and used them to demonstrate defective DNA repair in xeroderma pigmentosum cell extracts in cell-free systems. The former assay has been used to show defective interaction of xeroderma pigmentosum chromatin protein complexes involved in DNA repair with nucleosomal DNA (i.e., addition of histones to damaged DNA, so as to reconstitute the DNA into nucleosomal DNA, creates a substrate on which normal chromatin DNA repair protein complexes show enhanced activity, whereas the corresponding xeroderma pigmentosum protein complexes show decreased activity on this nucleosomal DNA). The latter system has been used extensively to examine the roles of different proteins and protein complexes involved in several cellular processes, particularly initiation of transcription, and DNA

repair, and to determine which of them are defective in different complementation groups of xeroderma pigmentosum. In 1992, Tanaka and his colleagues cloned a gene defective in patients in group A of xeroderma pigmentosum; five other xeroderma pigmentosum “correcting” genes have subsequently been cloned. Many other “firsts” related to cellular/biochemical studies of xeroderma pigmentosum are beyond the scope of this review. CLINICAL FINDINGS Xeroderma pigmentosum has a worldwide distribution and has been reported in all major racial groups, including Blacks. In the United States and Europe, and probably most parts of the world, the incidence is approximately 2–4 per 106 live births, but it is significantly more common in certain areas, especially Japan, North Africa, and the Middle East. It has also been reported to be especially common among the Bantus of Africa. In Japan, the incidence has been estimated to be about one in 40,000 live births. Xeroderma pigmentosum typically presents in infancy or early childhood with a sunburn following sun exposure. Less commonly, the child may cry when exposed to sunlight to the extent that this is called to the attention of a physician. Sometimes the initial sunburn may be so severe that the parents are suspected of some sort of child abuse. Depending on the sensitivity of the patient to sunlight and the extent of prior cumulative sun exposure, physical examination shows variable changes consisting of dryness (xerosis), induration, and dyspigmentation consisting of irregularly mottled freckling. The changes are absolutely limited to sun-exposed areas, and a sharp line of demarcation between involved and uninvolved skin is consistently present. As this process progresses, the freckles become larger, very irregular in shape, and some of them become darker. (Fig. 81-1). Areas of hypopigmentation and/or atrophy or scarring as well as telangiectasias also appear. These changes resemble those seen in the photoaging observed in older persons with light-complexioned skin who have undergone chronic sun exposure. They are often accompanied by eye lesions, such as extropion and entropion. A conspicuously absent feature, seen in these people but not in patients with xeroderma pigmentosum, however, is elastosis of dermal collagen.

CHAPTER 81 / XERODERMA PIGMENTOSUM

751

Figure 81-3 Face of a 47-year-old woman with xeroderma pigmentosum. Hundreds of neoplasms have been removed, as have both eyes.

Figure 81-2 Face of a 29-year-old man with xeroderma pigmentation. Note progression of changes (compared to Fig. 81-1).

As these changes continue to progress, premalignant and, subsequently, malignant skin lesions begin to appear (Fig. 81-1). Solar keratoses and keratoacanthomas usually appear first, followed by basal and squamous cell carcinomas and, sooner or later, melanomas (Fig. 81-2). The melanomas may arise in progressively larger and more irregularly shaped freckles; since there may be hundreds to thousands of these freckles, early detection of melanomas is often difficult. Long survival of severely sun-sensitive cases is uncommon; patients who reach middle age often show numerous neoplasms as well as extensive scarring and loss of tissue due to numerous surgical procedures carried out over the years on limited areas of skin (Fig. 81-3). A recent review of data obtained by the Xeroderma Pigmentosum Registry has provided the first reliable information regarding the distribution of melanoma and nonmelanoma skin cancers in patients with xeroderma pigmentosum. In both normal subjects and in patients with xeroderma pigmentosum, nonmelanoma skin cancers appear over areas of greatest cumulative sun exposure (Fig. 81-4). In contrast, melanomas appear at sites of intermittent sun exposure, closely correlating to areas exposed to the sun during sunbathing (upper body in men and upper body [sparing the breasts] and legs in women) (Fig. 81-5). These data, which for Xeroderma Pigmentosum Registry patients contrast sharply with those obtained from perusal of multiple case reports of xeroderma pigmentosum in the literature, are consistent with the hypothesis that xeroderma pigmentosum is a valid model for skin cancer in the general population. There are, however, some interesting differ-

ences. Some patients with xeroderma pigmentosum develop multiple carcinomas and only a few melanomas, but others develop numerous melanomas and few or no carcinomas. At least two patients have developed numerous melanomas, which, unaccountably, have failed to metastasize, and another has shown a dramatic response of melanomas to injection with interferon-α. Computerized morphometry has found that the nuclei of melanomas arising in xeroderma pigmentosum patients are smaller and less pleomorphic than those arising in otherwise normal subjects, and in many cases these melanomas arise from solar lentigos, whereas this is rarely the case in melanomas that arise in the general population. Despite these discrepancies, however, these data indicate at least one major difference in the pathogenesis of nonmelanoma vs melanoma skin cancers in both normal subjects and in patients with xeroderma pigmentosum, even though they appear to share the same etiology, ultraviolet light damage due to sunlight. Skin cancers of all types are about 1000 or more times more frequent in patients with xeroderma pigmentosum and arise about 40 years earlier than in the general population. If patients survive to adulthood they may have had hundreds to thousands removed, with extensive morbidity due to both the lesions, themselves, and the surgery necessary to remove them. Often, enucleation of the eyes is necessary (Fig. 81-3). In addition to skin cancers, premalignant and malignant conjunctival lesions often develop, as well as premalignant and malignant lesions of the lips and anterior tongue (Fig. 81-6). Sooner or later, patients with xeroderma pigmentosum usually die from complications due to these cancers, either from metastatic disease in developed countries or, in more primitive countries where it is impossible to treat so many tumors, from sepsis due to the tumors. Xeroderma pigmentosum is an extremely heterogeneous disease, both clinically and genetically (see below). Despite the fact that the typical patient shows extensive changes early in life, there are others with much milder disease that develops much more slowly. These patients often present much later in life, and are more difficult to diagnose, than their more typical counterparts. In sharp contrast to the above skin, ocular, and oral stigmata, which are absolutely related to cumulative exposure to ultraviolet

752

SECTION IX / DERMATOLOGY

Figure 81-4 Location of basal cell and squamous cell cancers in normal subjects and patients with xeroderma pigmentosum. Data from 401 cancers in xeroderma pigmentosum patients and 26, 817 cancers in the US white population. (Reproduced with permission from Kraemer et al. Arch Dermatol 1994.)

Figure 81-5 Location of melanomas in normal subjects and in patients with xeroderma pigmentosum. Data from 58 cutaneous melanomas in xeroderma pigmentosum patients and 5844 cutaneous melanomas in the general US population. (Reproduced with permission from Kraemer et al. Arch Dermatol 1994.)

radiation, neurologic abnormalities in patients with xeroderma pigmentosum occur independent of any known environmental influence. These changes affect only a minority of patients; most show either no abnormalities or subclinical abnormalities detectable only using special methods. When they do occur, these changes are progressive. The most common abnormality is loss of high-frequency hearing. Also, an abnormality in REM sleep has been reported. In most cases, however, these changes, although progressive, are mild, and the worst that most patients can expect is a need to use a hearing aid. A very small subset of patients with xeroderma pigmentosum shows severe neurologic abnormalities, as well as microcephaly, present at birth. As noted above, we believe that the term “De Sanctis-Cacchione syndrome” should be reserved for these unfortunate patients. Severe mental retardation is usually present. A few patients with xeroderma pigmentosum also have evidence of a second inherited disease with neurologic signs, Cockayne syndrome, as discussed below. The most important diagnostic assay, for xeroderma pigmentosum, the unscheduled DNA synthesis (UDS) test, is often not done, simply because it is expensive and not really necessary for the diagnosis in many cases. Even when the test is done, further tests, using fused cells, are needed to establish the complementation group;

and since these tests are also expensive, they are often not performed. Sometimes an assignment of “complementation group” is made based on clinical characteristics and the degree of depression of UDS following ultraviolet light exposure, rather than on tests using fused cells, so that the complementation group assignment is not as definitive as it should be. Nevertheless, despite these limitations, it is possible to state some significant general clinical facts about xeroderma pigmentosum complementation groups, as follows: Complementation Group A Patients in this group are severely sun-sensitive, and some but not all of them develop, prior to age 21 years, neurodegenerative disease, which is often severe. Group A is a relatively common complementation group of xeroderma pigmentosum, and is the most common group in Japan. Complementation Group B Only three patients in this group, one American and two in the same family in Europe, have been reported. The American patient was severely sun-sensitive, the European patients much less so. All three patients also had Cockayne syndrome. Complementation Group C Patients in this group are moderately to severely sun-sensitive but do not usually have neurological disease, although special testing reveals a subclinical hearing loss of high frequency sounds. Group C is probably the

CHAPTER 81 / XERODERMA PIGMENTOSUM

Figure 81-6 Lip and anterior tongue lesions in late xeroderma pigmentosum (same patient as in Fig. 81-3). (Reproduced with permission from Lambert et al., 1998.)

most common group worldwide, and is the most common group found in the United States, Europe, and the middle East. Two patients in this group have shown typical skin changes of xeroderma pigmentosum but have developed only melanomas that have not metastasized, and no other skin cancers. Complementation Group D Patients in this complementation group have mild-to-moderate sun-sensitivity and typically develop neurodegeneration that is mild and of late onset. It consists mainly of deafness. It is moderately common, comprising about 20% of cases of xeroderma pigmentosum worldwide. A few patients in group D have also had unequivocal stigmata of Cockayne disease. Some patients with an entirely different disease, trichothiodystrophy, have also tested positive for the XP-D trait in the complementation assay (see below). Complementation Group E Patients in this complementation group have only mild sun sensitivity and are not usually neurologically abnormal or are minimally so. Only a few cases have been reported. Complementation Group F Most group F patients have been moderately sun-sensitive, with development of only a few skin tumors and no neurologic abnormalities. However, one patient has been markedly sun-sensitive with severe neurological disease. Only a handful of patients have been reported, most of them Japanese. Complementation Group G Group G patients are severely sun-sensitive. Only five patients have been reported, two of which also had unequivocal stigmata of Cockayne disease. The remaining three did not show clinically evident neurologic defects. Complementation Group V (Variant group) Group V patients’ cells test normal in the UDS assay following exposure to ultraviolet light but can be induced to show low UDS by treatment with caffeine. These patients are moderately sun-sensitive and usually are not neurologically defective. This is a relatively common group, comprising about one-third of xeroderma pigmentosum patients worldwide. It is especially common in Japan. Dominantly Inherited Cases Two documented families— one Scottish, the other Australian—have shown an autosomal dominant pedigree for a mild type of xeroderma pigmentosum. Complementation studies have not been carried out on these patients’ cells.

753

DIAGNOSIS In children in whom skin tumors have begun to develop, the diagnosis of xeroderma piqmentosum can usually be made with confidence purely on clinical grounds (dry, oddly pigmented, often slightly scaly skin sharply limited to sun-exposed areas, along with the tumors). In younger children, in whom tumors have not yet developed, or in older patients with milder forms of the disease, however, diagnosis may be more difficult. Children with xeroderma pigmentosum can be differentiated from those with other light-sensitive disorders by the usual tests for those disorders. For example, porphyrias can be eliminated by testing for porphyrins in the blood, urine, and feces. Although Gorlin (nevoid basal-cell carcinoma) syndrome may show multiple basalcell cancers at an early age, the other stigmata of Gorlin syndrome (palmer pits, jaw cysts seen on X-ray) are usually easily recognizable. The dysplastic nevus syndrome may be associated with development of melanoma at a relatively young age, but the numerous dysplastic nevi usually make diagnosis relatively straightforward. Although, ideally, one would like to perform the UDS assay (see below) as well as complementation group analysis on all patients with xeroderma pigmentosum, these tests are expensive, and probably need only be done when the diagnosis is in doubt. A positive UDS test is a highly specific, although not absolutely specific, indication that the patient has xeroderma pigmentosum. The UDS test is currently performed at the following laboratories: c/o Dr. David Bush Department of Environmental and Toxologic Pathology Armed Forces Institute of Pathology 14th Street N.W. and Alaska Avenue Washington, DC 20306-6000 Tel: 202-576-0222 Fax: 202-576-2164 c/o Dr. James E. Cleaver (by special arrangement only) Laboratory of Radiobiology University of California at San Francisco Room LR 102 3rd and Parmassus Avenue San Francisco, CA 94143-0750 Tel: 415-476-4563 Fax: 415-476-0721 XP patients should contact, for information and support group activity: Xerderma Pigmentation Society, Inc. c/o Ms. Caren Mahar PO Box 4759 Poughkeepsie, NY 12601 Tel/Fax: 914-473-9735 Physicians with patients with Xeroderma patients should contact: Xeroderma Pigmentosum Registry c/o W. Clark lambert, MD Rm C-520 Medical Science Building UMDNJ-New Jersey Medical School 185 South Orange Avenue Newark, NJ 07103-2714 Fax: 973-972-7293 GENETIC BASIS OF DISEASE With the exception of a few families—one Scottish and one Australian family on which published data are available, a third family in North Carolina, and a possible fourth family in Ohio—in which xeroderma pigmentosum has been transmitted as an autosomal dominant trait, and one pedi-

754

SECTION IX / DERMATOLOGY

Figure 81-7 Formation of cyclobutane pyrimidine dimers and (6-4) photoproducts in cellular DNA by ultraviolet light. (From Lambert WC, Andrews AD, German J, et al. Xeroderma pigmentosum. In Thiers BH, Dobson RL, eds. Pathogenesis of Skin Disease. New York: Churchill Livingstone, 1986; pp. 576–599; with permission.)

gree in Italy in which the disease may have been sex-linked, xeroderma pigmentosum has been found to arise in pedigrees consistent with an autosomal recessive condition. The autosomaldominant cases have all been of mild degree; none have been confirmed by UDS studies or analyzed for complementation group to date. The sex-linked cases were found to complement in xeroderma pigmentosum group A. It has been known for many years that “classical” cases of xeroderma pigmentosum, those in complementation groups A–F, but not V, are defective in the initial, damage recognition and incision step in the nucleotide incision repair (NER) pathway of adducts introduced into cellular DNA by ultraviolet light. These adducts consist mainly of cyclobutane pyrimidine dimers and

6-4 pyrimidine-pyrimidone linkages, both of which occur between successive pyrimidines on the same strand of the duplex DNA molecule (Fig. 81-7). The dimers have outnumbered the 6-4 photoproducts by a ratio of at least two to one in most studies, but the 6-4 photoproduct causes a much greater degree of distortion in cellular DNA. Cells derived from patients with xeroderma pigmentosum have been found to be defective in repair of both adducts, but not in repair of other adducts due to reactive oxygen species generated by ultraviolet light, such as thymidine glycols, repaired by the base excision repair (BER) pathway. Figure 81-8 outlines the base and nucleotide excision repair pathways in both mammalian cells and in cells of many other species, including yeast and Escherichia coli. On the left a mecha-

CHAPTER 81 / XERODERMA PIGMENTOSUM

755

Figure 81-8 Alternate excision repair pathways proposed for both prokaryotes and eukaryotes. (From Lambert WC, Andrews AD, German J, et al. Xeroderma pigmentosum. In: Thiers BH, Dobson RL, eds. Pathogenesis of Skin Disease. New York: Churchill Livingstone, 1986; pp. 576–599; with permission.)

nism is depicted in which a nick is created by an endonuclease on one side of the adduct, followed by digestion of the part of the same strand containing the adduct by an exonuclease, creating a single strand gap in the DNA. This was once thought to be the way in which mammalian NER proceeded, but is now known to be incorrect. The correct mechanism for NER in both mammalian cells and bacteria is shown on the right of the figure. A nick is made endonucleolytically on both sides of the adduct and the portion of the strand containing the adduct is removed, not necessarily by an exodonuclease. In bacteria this is mediated by the products of the UvrABC and D genes (called the UvrABC system) and the term “exinnuclease” has been used to describe the resulting combinations of enzyme activities. In mammalian cells and in yeast, these steps are mediated by protein complexes, more about which we shall explore below. The end result is again a single-stranded gap in the double-stranded DNA molecule (Fig. 81-8). In the center of Fig. 81-8 is depicted the base excision repair (BER) pathway in which an adduct, usually a simpler single-base alteration, is removed first by the action of a glycosylase, which removes the adduct so as to create an apurinic/apyrimidinic (AP) site, followed by the action of an AP endonuclease, which cuts the phosphodiester bond at the AP site. The two activities are often present in the same enzyme molecule. This eventuates, again, in the creation of a single-stranded gap in the double-stranded DNA molecule, although it is believed to be a shorter gap than that created by the NER pathway. Whatever the mechanism, the single-stranded gap that results is then filled in by a DNA polymerase and then sealed by a DNA ligase. This affords a way to assay the process in intact cells in culture, since a small amount of DNA synthesis must occur. Replicative DNA synthesis occurs in a discrete phase of the cell cycle known as S phase, with phases both before and after it, called respectively G1 and G2 phases, in which no DNA synthesis occurs. However, when NER or BER are active, a small amount of DNA

Figure 81-9 Autoradiograms of normal (A) and xeroderma pigmentosum, complementation group A (B) fibroblasts following irradiation with UV-C (254 nm) light and treatment with 3H-thymidine. S-phase cells show large numbers of grains (A, lower left, and B, lower right, one cell in each), whereas cells undergoing unscheduled DNA synthesis show moderate numbers of grains (seven cells in A). Xeroderma pigmentosum, complementation group A cells show few or no grains (four cells in B). (From Lambert WC, Lambert MW. Diseases associated with DNA and chromosomal instability. In: Alper JC, ed. Genetic Disease of the Skin. St. Louis, Mosby-Year Book, 1991; pp. 320–358; with permission.)

synthesis, called unscheduled DNA synthesis (UDS) occurs in G1 and G2 as well. This can be measured in UV-irradiated cells using a 3H thymidine pulse followed by autoradiography, thus providing a measure of repair related DNA synthesis. The UDS assay has become the gold standard for establishing a diagnosis of xeroderma pigmentosum in cases where the clinical diagnosis is uncertain. In this assay, a patient’s cells are irradiated in culture with ultraviolet (UVB or C) light, pulsed with tritiated thymidine (an exclusive DNA precursor), and subjected to autoradiography. High-grain-count cells are discounted (as S phase cells) and low-grain-count nuclei scored as cells undergoing a limited amount of non–S-phase unscheduled DNA synthesis (UDS). Figure 81-9 shows an example of a normal and abnormal UDS test in fibroblasts; and Fig. 81-10 shows this test in lymphblastoid cells. Recently our laboratory has applied a new computerized image analysis system to these autoradiograms, which has allowed a vast improvement in the results obtained. Figure 81-11 shows a

756

SECTION IX / DERMATOLOGY

Figure 81-10 (A and B) Autoradiogram of human lymphoblastoid cells in culture following 1-h exposure to 1 mM methyl methanesulfonate, a DNA-damaging agent, and labeling with 3H-thymidine. Cells show both scheduled (heavily labeled) and unscheduled (lightly labeled) DNA synthesis. This test, with ultraviolet radiation (UVR) or UVR-mimetic chemicals, is frequently employed in lymphoblasts or fibroblasts to confirm a clinical suspicion of xeroderma pigmentosum. This autoradiogram was examined by direct (transmission) microscopy in which the silver grains appear dark (A) and also by incident polarized light fluorescence microscopy in which the sliver grains appear bright (B), and the two images integrated by electronic image analysis (see Figs. 81-11 and 81-12). This eliminates most nonspecific background.

computer-generated image of a UDS assay; Fig. 81-12 shows the results generated from the assay. This system greatly facilitates and improves the accuracy of these assays. The UDS assay is, however, not completely specific; at least one normal person and one individual with hydroa vacciniforme have been found to have very low UDS levels, and, as noted above, a number of patients with trichothiodystrophy have also had low UDS results. Thus the UDS assay must be evaluated in the proper clinical context to be meaningful. The UDS assay has also been used as the basis for the complementation assay in xeroderma pigmentosum, the first such assay to be developed in human genetics. In this assay, cells from different patients are fused and the UDS assay performed. In cells from different complementation groups, both nuclei show normal or near-normal UDS, which is not seen in cells from either patient fused with themselves (Fig. 81-13). This is interpreted as evidence

for genetic heterogeneity. Nuclei of fused cells from patients in the same complementation group fail to show this phenomenon: UDS remains low. MOLECULAR PATHOPHYSIOLOGY OF DISEASE The correlation between decreased nucleotide excision repair capability and development of skin lesions, especially skin cancer, following ultraviolet light exposure has been demonstrated in several ways. Sun protection leads to avoidance of these cutaneous lesions, which also occur in sun-exposed areas (see above). Within the cutaneous lesions, mutation analysis shows that critical cancerrelated genes, such as p53, are mutated with a markedly increased incidence in xeroderma pigmentosum patients. Moreover, these mutations can be traced to the types of adducts produced (and not repaired properly in xeroderma pigmentosum patients) by ultraviolet light. The roles of various products of xeroderma pigmentosum genes in the nucleotide excision repair pathway have recently become elucidated. This has been made possible by several recent developments. The first development is the successful cloning of several xeroderma pigmentosum “correcting” genes. Successful cloning of the XPA gene by Tanaka et al. has been alluded to above. Some other human genes have been identified based on their ability to restore mutant, ultraviolet-sensitive rodent cells to normal following their introduction by artificial means. These genes were first termed ERCC (Excision Repair Cross-Complementing), followed by the number of the rodent cell complementation group (the rodent groups are labeled 1–11) that the human gene can complement. A number of these genes have now been cloned and found to also function to correct ultraviolet light response of cells in a xeroderma pigmentosum complementation group, at which point, by convention, they are renamed XP (complementation group) C genes, as XPAC, XPBC, and so on. Recently the final “C” has been deleted, so that XPAC is now XPA, and so on (Table 81-2). The second development has been the improvement of in vitro cell-free systems for examination of nucleotide excision repair. The first of these systems, developed in our laboratory in 1983, has been most useful in elucidating the role of chromatin proteins in enhancing the repair process. The second major system, developed by Richard Wood and Tomas Lindahl in 1988, has been used to examine the role of individual proteins in nucleotide excision repair (NER). Using this system, it has been found that NER in bacteria requires 6 proteins, in mammalian cells at least 30 proteins. The third major development has been the realization that a major protein complex serving as a regulator of initiation of human RNA polymerase II, called TFIIH, is required for NER, and that NER progresses much more rapidly in the transcribed strand of actively transcribed genes than in other DNA. TFIIH contains the products of the xeroderma pigmentosum-correcting genes XPB, XPD, and XPG. This fact may help to explain why some or all patients in these complementation groups have also had Cockayne syndrome, and why those Cockayne syndrome patients have been of short stature. Furthermore, those patients with trichothiodystrophy who have tested positive in the UDS assay have fallen into complementation group D (most patients) or B of xeroderma pigmentosum. A scheme for the current concept of how these genes interact in NER has been proposed by Alan Lehmann, as follows (Table 81-2): First, the XPA gene product recognizes and binds to the damaged site in DNA via a zinc-finger domain on the XPA molecule. A

CHAPTER 81 / XERODERMA PIGMENTOSUM

757

Figure 81-11 Computer-generated image generated from analysis of Figs. 81-9 and 81-10. Virtually all background has been eliminated and accurate counts of 0 to several hundred grains per nucleus may be obtained.

different domain on this same XPA protein molecule also binds to the single-stranded DNA binding protein, RPA (also known as SSB). Second, the XPA and RPA proteins act together to recruit the large TFIIH protein complex, which contains both the XPB and the XPD proteins. These latter proteins are helicases with opposite polarity, which unwind the double-helial DNA molecule, respectively, 5' and 3' to the adduct, creating a loop or “bubble” in the DNA molecule. Third, two different DNA endonucleases, one consisting of a complex of two proteins, XPF and ERCC 1, and the other consisting of the XPG protein, cut the strand of the DNA molecule containing the adduct on its 5' and 3' sides, respectively. Fourth, DNA polymerase epsilon and accessory proteins, replication factor C (RPC) and proliferating cell nuclear antigen (PNCA), fill in the resulting gap in the damaged DNA molecule. This is the step that produces the UDS phenomenon. Finally, the newly synthesized DNA is joined at its 3' end to the DNA beyond the repair gap by a DNA ligase. This pattern of activities for the xeroderma pigmentosum gene products is, however, almost certainly quite simplistic, and should be regarded as only a first approximation of the events that occur in NER. Our laboratory, for example, has produced evidence that complementation group A cells are lacking a different product, which functions not in adduct recognition but rather in interaction of the repair proteins with chromatin. Indeed, it appears likely that, for xeroderma pigmentosum and/or some of these other disorders of DNA and chromatin instability (see below), more than one biochemical defect may be necessary to produce the clinical disease. We have proposed a model, corecessive inheritance, which accounts for many of the discrepancies found in these diseases, such as, for example, the finding of a single genetic defect, in the ATM gene, in several different complementation groups of ataxiatelangiectasia (see below). This model proposes an overlap in the function of some of these DNA repair/surveillance genes, so that

the product of one can take effect where another may fail. Thus, to clinically have the disease, one must be homozygous for defective alleles (or hemizygous, should a sex-linked gene be involved) at two or more loci simultaneously. This model, should it or something like it prove correct, has profound implications for human biology. For example, it predicts extremely high carrier frequencies for these defective alleles in the general population (Fig. 81-14), so that the genes responsible for these very rare diseases may also play important roles in the etiopathogenesis of common disorders, such as cancer, ageing, and even age-related neurodegenerations. This model is by no means accepted as proven or even probable, however. It may, for example, be argued that the existence of XP “knockout” (KO) mice mitigates against it. On the other hand, KO mice are generated in inbred strains that have lost nearly half the genetic material present in the wild type, and the phenotypes of these KO mice are incomplete. MANAGEMENT/TREATMENT Avoidance of sun exposure and vigilant surgical removal of tumors in xeroderma patients are the mainstays of treatment. It is important that surgery be performed with as little loss or destruction of normal tissue as possible, noting a high likelihood that surgery may have to be performed in the same site again and again, as malignancies continue to arise. Particularly in severely affected small children, there is usually an interval in which skin damage due to sun exposure is evident in xeroderma pigmentosum patients, yet skin cancers have either not yet appeared or have just begun to appear. Parents at this time are hopeful that this situation may persist, but it rarely does. At this time they are loath to believe the dire prognosis their physicians present to them, and are extremely vulnerable to proponents of “alternate medicine,” further compounding their problems and often draining their pocketbooks. Moreover, even if sufficient sun protection is afforded to prevent acute discomfort, this may be

758

SECTION IX / DERMATOLOGY

Figure 81-12 Unscheduled and scheduled (S phase) DNA synthesis in a lymphoblastoid cell line, derived from a normal individual, treated with progressively increasing doses of the DNA-damaging agent and mutagen, methyl methansulfonate (MMS). This plot was computergenerated based on grain counts and cell morphology analysis of autoradiograms of cells incubated for two hours in 3H-thymidine following a 1-h exposure to MMS, performed in our laboratory using data generated by the Leitz Quantimet 500 Plus Image Analysis System (Leica, Cambridge, UK). The back plane corresponds to untreated cells; the front plane to cells treated with 3 mM MMS. On the far left, the proportion of cells with 0–25 grains, indicating little or no DNA synthesis, is seen. On the right, a ridge, indicative of the population of cells in S phase, each showing multiple grains, is seen. Between them is a trough in which there is overlap between low-labeled S-phase cells (mostly cells entering and leaving S phase during the radiolabeled thymidine pulse) and cells undergoing UDS. (For this reason, in this experiment, UDS was measured as loss of cells from the unlabeled population on the far left.) In untreated cells, approximately 35% of the cells showed little or no DNA synthesis (a). With progressively higher doses of MMS, this proportion fell to near zero as non–S-phase, intephase cells underwent UDS, making a modest amount of new DNA (b). At still higher doses of MMS, cells lost viability and were unable to synthesize DNA by either mechanism (c, d, and e). At the same time, cells in S phase, making large amounts of DNA progressively diminished their rate of replicative DNA synthesis (f in untreated cells to g in cells treated with 3 mM MMS). Determinations were made at 0.0, 0.1, 0.5, 1.0, 1.5, 2.0, 2.5, and 3.0 mM MMS with over 2500 cells counted. (Reprinted with permission from Lambert et al., 1995.)

insufficient to prevent the appearance of tumors later. Conversely, parents with limited resources and/or other healthy children may elect to limit the special care that they afford the child with xeroderma pigmentosum. These are difficult choices that only the family can make, and there is usually no correct or incorrect solution. Several other modes of treatment are emerging, but they remain experimental at present. High-dose retinoids have been shown to retard rates of tumor growth. They also produce a “phasing” effect, allowing multiple tumors to be removed at one time. These high doses, however, are not without side effects, especially bone abnormalities. Application of bacterial DNA repair enzymes via topically applied microcoeles, which allow penetration of the superficial cornified layer of the skin, are also a promising

Figure 81-13 Complementation assay (UDS, fibroblasts) showing a heterokaryon (fusion product of one cell from each patient) with normal UDS and a homokaryon (fusion product of two cells from one patient) with depressed UDS. (Courtesy Dr. H. Takebe.)

approach. Dermabrasion of the skin has also been recently applied, with some success. Finally, surgical replacement of severely damaged skin by undamaged skin has been done in selected cases with some success (Fig. 81-15). Even though the replaced skin is still sun-sensitive, the accumulated mutagenic damage has not occurred, and new tumor growth is very much diminished. In families with an affected child, prenatal diagnosis should be done whenever possible, and acceptable to the parents. FUTURE DEVELOPMENTS Improved prenatal diagnosis of xeroderma pigmentosum using cloned genes or even in vitro selection of nonaffected embryos may become possible in the near future. Effective gene therapy, possibly using cutaneous application of corrective genes in vehicles such as microcoeles, may make it possible for patients with this disease to live nearly normal lives. If these genes play important roles in the etiopathogenesis of more common diseases, screening of the normal population for these defective alleles may allow earlier diagnosis of some of these diseases to be made. In any case, further studies on xeroderma pigmentosum are certain to provide additional information regarding human DNA repair processes and chromatin biology.

ATAXIA-TELANGIECTASIA (LOUIS-BARR SYNDROME) BACKAROUND Ataxia-telangiectasia is an uncommon disorder characterized by development of large telangiectatic lesions on the conjunctivae and face, progressive cerebellar ataxia and other neurological deficiencies, immunological defects, and marked hypersensitivity to ionizing radiation. Although the first report of this disease is attributed to Madame Louise Barr in 1931, it may have been described earlier. Peripheral blood lymphocytes of these patients show somatic translocations, primarily involving chromosomes 7 and 14. Cultivated cells derived from these individuals show a tendency for chromosomes to fuse at their telomeres and defective cell cycle checkpoint controls, especially following exposure to ionizing radiation.

759

CHAPTER 81 / XERODERMA PIGMENTOSUM

Table 81-2 Human DNA Repair Genes and Their Nonhuman Homologs Homologous nonhuman genes Disease complementation group/gene XP-A XP-A XP-B/CSa ATPase XP-C XP-D XP-E XP-F XP-G XP-V CS-A CS-B Bloom syndrome FA (FA-A-D) – HHR6A HHR6B

Human chromosome

Rodent

Yeast S. cerevisiae

S. pomp

RAD14

Putative function

9Q34 8 2q21

ERCC3

(ERCC3sc)

(ERCC3sp)

5 19q13.2

ERCC2

RAD14 RAD3

rad14

DNA endonuclease; DNA-dependent ATPase DNA helicase; transcription factor Damage-specific DNA-binding protein; photolyase

RAD2

rad13

DNA helicase; transcription factor

RAD10 RAD1 RAD6 RAD6

swil0 rad16

15 13q32-38

b

10q11 15 9 19q13

ERCC6

ERCC5

ERCC1

Damage-specific DNA-binding protein DNA helicase; DNA-dependent

Damage-specific DNA-binding protein (FAOA) RAD1-RAD10 complex: double-stranded DNA endonuclease Ubiquitin-conjugating enzyme

aXP-B/CS is also homologous to the haywire gene in drosophila. bERCC1 and ERCC11 are candidates based on recent studies in cell-free

systems. From Lambert WC, Lambert MW. DNA repair deficiency and skin cancer in xeroderma pigmentosum. In: Mukhtar H, ed. Skin Cancer, Mechanisms and Human Releance. Boca Raton: CRC, 1994, with permission.

The disease is transmitted as an autosomal recessive disorder with four complementation groups, A, C, D, and E, currently recognized, although a number of others have also been proposed. In 1995, an international consortium identified a single gene, known as the ATM (Ataxia-Telangiectasia Mutated) gene, which, surprisingly, is mutated in patients in all of these complementation groups as well as some other patients with ataxia-telangiectasia not assigned to a specific group. Heterozygotes for the ATM gene are believed to be at increased (two - to threefold) risk to develop certain types of cancer, particularly breast cancer in women. The breast cancer risk in these women has been proposed to be four to five times that of women in the general population, and it has been estimated that the ATM gene is responsible for approximately 8% of human breast cancer. However, some studies have shown lower associations between ATM gene carrier state and breast cancer, and the subject is controversial. Three related diseases—The Nijmegen breakage syndrome, the Berlin breakage syndrome, and ataxia-telangiectasia Fresno— have clinical or cellular features closely related to those of ataxiatelangiectasia. CLINICAL FEATURES Ataxia-telangiectasia is by far the most common entity among those discussed in this chapter, affecting approximately one individual per 30,000 live births. The pedigrees of these cases have been consistent with an autosomal recessive mode of inheritance. The Nijmegen breakage syndrome and the Berlin breakage syndrome are both characterized by microcephaly and mental retardation, but by none of the other clinical features of ataxia-telangiectasia. Ataxia-telangiectasia Fresno has clinical features of all of these diseases. If microcephaly, mental retardation, or both are present, this essentially rules

out a diagnosis of ataxia-telangiectasia. However, a patient with ataxia-telangiectasia may appear to be retarded when in reality he or she is only socially repressed. The first sign of the disease is invariably ataxia, which typically begins at about the time the child learns to walk and is progressive, so that by the age of 10 years the patient usually requires a wheelchair. The ataxia is truncal and of cerebellar type. It initially affects gait but later is of intention type as well. A second, also very characteristic clinical finding in ataxia telangiectasia is ocular apraxia. When a patient is asked to follow an object moved across the visual fields with his or her eyes, the head is turned to follow the object, the eyes following afterward. The characteristic telangiectatic oculocutaneous stigmata of ataxia-telangiectasia often appear years later than the neurologic signs, but by the age of 10 years almost all patients have markedly dilated, thin-walled blood vessels visible over the bulbar conjunctivae. These large vessels contrast sharply with the finely telangiectatic rash seen over the face of patients with Bloom syndrome, and do not show blistering, ulceration, or any association with sunlight. These dilated vessels are often also present on the eyelids, the bridge of the nose, on the pinnae of the external ears, and, less commonly, over the antecubital and popliteal fossae. In addition to these stigmata, very young patients may also, occasionally, show a few gray hairs and hypertrichosis, the latter especially over the forearms. Ataxia-telangiectasia patients are markedly prone to develop malignancies involving the lymphoid system. About 75% of these are lymphomas, most commonly B-cell lymphomas. These often occur in young patients; of the approximately one-fourth of these malignancies that arise later, most are leukemias, usually of the T-cell chronic lymphatic type. Often the malignant cells show

760

SECTION IX / DERMATOLOGY

Figure 81-15 Total Facial skin replacement by surgical graft in a 7-year-old girl with severe xeroderma pigmentosa.

Figure 81-14 Carrier frequencies in the general population for defective alleles for DNA repair and chromosomal stability genes associated with xeroderma pigmentosum and related disorders predicted by the classical autosomal recessive (AR) and corecessive inheritance (C-RI) models. n, number of genes. (For C-RI: number of equally prevalent autosomal recessive genes which must be homozygously defective for the disease to be observed.) Z, proportion of live births in which the disease is present; F, frequency (1.00 is 100% for the population) who are carriers; FAR, for AR; FOAR, for oligogenic AR (i.e., AR with complementation groups); FC-RI, for C-RI. (Reproduced with permission from Lambert and Lambert, 1992.)

inversions or translocations involving chromosome 14. Nonlymphoid malignancies, when they occur, almost always are found in older patients with ataxia-telangiectasia. These patients are extremely radiosensitive, and may even develop an erythematous rash or other symptoms after diagnostic X-ray examination. Occasionally a patient develops a leukemia or lymphoma prior to being diagnosed with ataxia-telangiectasia. This may lead to administration of a dose of X-irradiation or of one or more chemotherapeutic agents that are contraindicated in ataxiatelangiectasia. The results are usually dramatic and tragic. It is imperative that the diagnosis of ataxia-telangiectasia be considered, and appropriate diagnostic steps taken, if necessary, prior to treating a very young patient for cancer. Although some type of immunodeficiency is present in most patients with ataxia-telangiectasia, there is great heterogeneity in both type and degree of this between patients and often even

between siblings. Four-fifths of patients have decreased serum levels of IgG2, IgE, or IgA. Although blood T-cell counts are often normal, T-cell responsiveness to mitogens or allogenic cells is often low. At autopsy, the thymus is typically rudimentary. About 95% of patients with ataxia-telangiectasia have elevated levels of α-fetoprotein in their sera. The 5% of patients who do not have this abnormality do not appear to differ from those that do in any other respect. In 5–10% of peripheral blood lymphocytes of most patients with ataxia-telangiectasia, there is a translocation involving chromosomes 7 and 14. Skin fibroblasts do not show these characteristic translocations. DIAGNOSIS If early onset cerebellar ataxia, ocular apraxia and the characteristic telangiectasia of the bulbar conjunctive are all recognized, the diagnosis of ataxia-telangiectasia may be considered to be virtually established. In these cases laboratory tests are simply confirmatory. In cases in which not all of these signs are present, however, the laboratory plays a more important role. This is rather frequently the case in young children, since the telangiectatic eruption often arises years later than the ataxia and ocular apraxia. The indicated tests are serum α-fetoprotein and immunoglobulin determinations, and karyotyping and radiosensitivity assays on mitogen-stimulated peripheral blood lymphocytes and/ or cultured cells. Tests for radiation-resistant DNA synthesis may also be done, but are usually a research procedure. The recent cloning of the ATM gene (see below) may soon provide molecular assay for this disease, but this is not yet available. α-Fetoprotein levels are elevated in about 95% and serum IgE, IgG2, or IgA levels are depressed in about 80% of patients with ataxia-telangiectasia. As for most clinical laboratory tests, these studies should be done at least twice. Cytogenetics assays of mitogen-stimulated peripheral blood lymphocytes may be complicated by the fact that ataxia-telangiectasia lymphocytes do not show an optimal mitogenic response to phytohemagglutinin. Thus a cytogenetics laboratory may return a report of “no metaphases observed.” To avoid this, the laboratory should be advised to use at least one additional (higher) dose of phytohemagglutinin in separate cultures and also to harvest some of the lymphocyte cultures at 72 h rather than 48 h. This will allow metaphases to be visualized in most cases of ataxia-telangiectasia.

CHAPTER 81 / XERODERMA PIGMENTOSUM

Figure 81-16 Chromosome preparation of a metaphase cell from a patient with ataxia-telangiectasia. Note fusion of telomeres (arrows).

Karyotypes of ataxia-telangiectasia cells show abnormalities, especially telomere–telomere linkages (Fig. 81-16). Most ataxiatelangiectasia patients’ peripheral blood lymphocyte karyotypes show translocations between chromosomes 7 and 14 in 5–10% of metaphases examined. Survival studies of cells derived from patients suspected of having ataxia-telangiectasia following radiation or other treatments are not routine procedures in most centers, but can provide valuable information when available. Colony-forming ability (CFA), colony survival assay (CSA), and cloning efficiency (CE) assays of peripheral blood lymphocytes irradiated with 1 Gy of X-irradiation typically show CFA values in ataxia-telangiectasia cells that are less than 20% of control levels and also less than 20% of levels of ataxia-telangiectasia heterozygotes. In some assays, heterozygote cells may show levels of activity intermediate between those of normal and those of homozygote cells. Prenatal diagnosis is an extremely important aspect of diagnosis in a family in which a child with ataxia-telangiectasia has been identified. Until very recently the standard technique has been based on haplotype analysis of chromosome band 1 lq.22.3 in both parents and the affected child, to determine which is associated with the defective ataxia-telangiectasia gene, and then testing the DNA of the fetus, obtained either from amniocytes or from chorionic villus cells, to identify which of the parents’ haplotypes have been inherited. This is then compared to the haplotypes of the affected sibling. An estimate of the likelihood of the fetus being affected is then computed. This procedure has provided extremely accurate prenatal diagnosis, due to the fact that the gene has been very well localized. Now that the ATM gene has been identified and cloned (see below), prenatal diagnosis should become simpler and even more accurate. GENETIC BASIS OF DISEASE Pedigrees of individual families of patients with ataxia-telangiectasia have consistently shown an autosomal recessive pattern of inheritance. However, an exten-

761

sive epidemiologic study of incidence of ataxia-telangiectasia in a defined region of central-north England produced data that could not be well-fitted by any inheritance model. Subsequent examination of those reported data by two of us (WCL and MWL) indicate that they may be fitted by the corecessive inheritance model (see above). However, there may be other explanations for these findings. Cells in culture derived from patients with ataxia-telangiectasia—as well as those from patients with the Nijmegen breakage syndrome, Berlin breakage syndrome, and ataxia-telangiectasia Fresno—show a number of distinctive features: Peripheral blood lymphocytes, but not skin fibroblasts, show inversions and recombinations in chromosomes 7 and 14 in 5–10% of metaphases. All cell types tend to show chromosome breaks and a tendency of chromosomes to fuse at their telomeres (Fig. 81-16). All cells are radiosensitive and, in some but not all assays, cells derived from ataxia-telangiectasia heterozygotes show an intermediate level of radiosensitivity between that of normal cells and ataxia-telangiectasia cells. The increase in stability of p53 seen in normal cells after irradiation is significantly delayed in ataxia-telangiectasia cells and fails to reach normal levels. Ataxia-telangiectasia cells show additional breaks and other abnormalities in their metaphase chromosomes following irradiation, in excess of those seen in normal cells. Unlike normal cells, ataxia-telangiectasia cells fail to slow their rate of DNA synthesis following X-irradiation, a distinctive phenomenon termed by Painter and his colleagues “radiation-resistant DNA synthesis” (RRDR); several cell-cycle checkpoint controls, including G1 and G2 checkpoint controls, are defective. This phenomenon (RRDS) has been used, in fused cells, to identify complementation groups in ataxia-telangiectasia. Nuclei of cells fused from different individuals with ataxia-telangiectasia, which separately show RRDR, show normal responses to X-irradiation when the patients are in different complementation groups, but show RRDS when they are in the same group. On this basis, four complementation groups of ataxia-telangiectasia have been proposed; A, C, D, and E. What was formerly group B has been incorporated into group A. Not all of the cellular studies have produced unequivocal results, however, and additional groups have been proposed. In addition to these groups, cells from patients with the Nijmegen breakage syndrome and the Berlin breakage syndrome have been found to show similar changes in culture to those of ataxia-telangiectasia cells, but complement both each other and all ataxia-telangiectasia cell lines tested in the RRDS complementation assay. Therefore cells derived from patients with the Nijmegen breakage syndrome and the Berlin breakage syndrome are said to be in ataxia telangiectasia complementation groups V1 and V2, respectively. The complementation group analysis is essentially the only criterion that separates these latter two disorders, which are otherwise very similar in both clinical and cell culture characteristics, although different in the former from ataxia-telangiectasia. In 1995, a large, international consortium of coinvestigators isolated and cloned a gene, which they termed the ATM (AtaxiaTelangiectasia Mutated) gene, which was found to be defective in each of a large group of ataxia-telangiectasia cell lines examined. Surprisingly, the ATM gene was defective in cells derived from patients in each of the four well-characterized complementation groups. They attributed this result to intragenic complementation, but the issue has not been resolved. They did not, in fact, explore this issue in any sort of detail, and consequently did not consider that defective alleles at one or more other genetic loci, in

762

SECTION IX / DERMATOLOGY

Figure 81-17 Sequence homology of the COOH-terminal region of the human ATM gene with other members of the PI3K family. At the right the % identity (I) and similarity (S) of both the kinase domains and the adjacent 103 amino acids is given. (Reproduced with permission from Zakian VA. Cell 1995;82:685–687.)

addition to the ATM gene, might be necessary to produce clinical ataxia-telangiectasia. Based on linkage studies performed to date, the ATM gene is unlikely to be associated with either the Nijmegen breakage syndrome on the Berlin breakage syndrome. The ATM gene is on chromosome band 1 lq22-23 and has a transcript of 12 kb. Following discovery, it was quickly found to be homologous to a number of other known eukaryotic and prokaryotic genes (see below) and was immediately recognized as having a gene product similar to phosphatidylinositol-3-kinase. MOLECULAR PATHOPHYSIOLOGY OF DISEASE The ATM gene product, predicted from its cDNA, was found to have, in the approximately 400 amino acids composing its carboxyl (-COOH) end (believed to have seronine/threonine protein kinase activity), strong homology to a group of proteins in humans and other species known as phosphatidylinistol-3-kinases (PI3Ks). PI3Ks phosphorylate both lipids and—in proteins, serines, and threonines—play a central role in a number of signal transduction pathways, including those that regulate certain cell-cycle checkpoints and such other cellular processes as transport of certain substances across membranes. Figure 81-17 shows the degree of sequence homology between the predicted ATM gene product and a number of products of other genes in the PI3K gene superfamily in various species. As is immediately apparent, the protein products of certain genes, particularly the Drosophila melanogaster (fruit fly) gene MEI41, the Saccromyces cerevisiae (yeast) genes TEl 1 and Esr 1/Mec 1, the S. pombe (yeast) gene Rad 3, and the human gene ATM share significantly more sequence homology with each other than they do with the other gene products listed. There is also close homology with the products of the S. cerevisiae genes Tor 1 and Tor 2 and with those of their mammalian homologs, FRAP(mTor) and rRAFT 1. Slightly less homologous to all of these is the catalytic subunit (cs) of human DNA-protein kinase (DNA-PK), an important DNA repair enzyme (see below). Least homologous to the

other gene products within this group are those of the most homologous PI3Ks known, p110 (bovine) and Vps 34 (S. cerevisiae). Table 81-3 lists some of the properties of ataxia-telangiectasia cells along with those of cells mutant in the most homologous of these genes, Tel 1, Esr 1/Mec 1, Rad 3 and D. melanogaster MEI41. Phenotypes of S. pombe Rad 3- and D. melanogaster MEI41mutant cells clearly resemble those of ataxia-telangiectasia cells, as do those of S. cerevisiae cells mutant in both Tel 1 and Esr 1/Mec 1. All of them are hypersensitive to several DNA damaging agents, including X-irradiation, and all show defective cell-cycle checkpoint controls, particularly following X-irradiation. Interestingly, yeast cells with a truncated mutant form of Esr 1/Mec 1 but normal Tel 1 show partial resistance to the lethal effects of X-irradiation when they have one or two extra normal copies of Tel 1. Cell lines for which this information is available also show chromosomal abnormalities, particularly defective telomere stability, and these defects are exacerbated by X-irradiation. Human ataxia-telangiectasia cells are hypersensitive to treatment with bleomycin or streptonigrin, hypersensitivities to which are present in S. cerevisiae cells only if mutant in both Tel 1 and Esr 1/Mec 1, further reinforcing the idea that both genes together function rather like the ATM gene in man. The degree of overlap in these phenotypic characteristics is not perfect, however. For example, all of these nonhuman mutated cell lines are hypersensitive to ultraviolet radiation, whereas most ataxia-telangiectasia cell lines show normal sensitivity to this agent. Esr 1/Mec 1 apparently has some function in yeast which it lacks in humans, since Esr 1/Mec 1 mutants are lethal. The yeast TOR 1 and TOR 2 gene products, working together, and their mammalian homologs, FRAP and rRAFT, working together, are required for the G1 and S phase cell cycle progression in their respective species. DNA-PKcs, when bound to the remaining subunits of DNA-PK, kDa 70 and kDa 80, which together comprise the Ku nuclear antigen important in certain autoimmune diseases, forms an active DNA-protein kinase (DNA-PK). However this binding is in turn dependent on recognition and binding of the Ku subunits to DNA double strand (ds) breaks. Such DNA ds breaks are an important consequence of X-irradiation. DNA-PK is a serine/threonine protein kinase which activates a number of different proteins, including those that modulate certain cell-cycle checkpoints, so as to arrest the cell cycle, and the nuclear protein, p 43, which leads to stabilization of p 53, a tumor suppressor gene product the stabilization of which, following X-irradiation, is delayed in ataxia-telangiectasia cells. Activity of DNA-PK is also important in B and T lymphocytes undergoing V(D)J gene recombination. Thus a defective ATM gene, related to the DNA-PK gene, may be responsible for the immunodeficiency seen in ataxiatelangiectasia. Human and mouse SCID (Severe Combined ImmunoDeficiency) mutant cells with deficient DNA-PK activity are hypersensitive to X-irradiation, show DNA repair defects, and are unable to support V(D)J recombination. Interestingly, S. cerevisiae telomeres also interact with DNA double strand breaks, resulting in cell cycle arrest at the G2 to M checkpoint. Notably, DNA-PK does not phosphorylate lipids (such as phosphatidylinositol). It is likely that many or all of the above gene products, except the PI3Ks (p 110 and VPS-34), exert part of their effects by phosphorylation of proteins but not lipids. This further distinguishes them from PI3Ks. Thus the ATM gene is a reasonable candidate to play a role in the pathophysiology of numerous cellular and clinical features of ataxia-telangiectasia. However, much needs to be done to delin-

763

CHAPTER 81 / XERODERMA PIGMENTOSUM

Table 81-3 Properties of Ataxia-Telangiectasis Cells and of Cells with Mutant Genes with Product Homology to the ATM Gene

Cell type Human Ataxia-telangiectasia S. cerevisiae Mecl+ Tell+ Mecl– Tell+ Mecl– Tell+ S. pombe Rad3– D. melanogaster Mei-41–

Lethal

Cell-cycle checkpoint defects

No

Yes

No Yes Yes

Chromosomal abnormalities

Hypersensitivities UVa

X-Raysb

Chem.c

B and Sd

No

Yes

No

Yes

No Yes Yes

Yes, with impaired telomere function Yes, with short telomeres — —

No Yes Yes

No Yes Yes

No Yes Yes

No No Yes

No

Yes

—

Yes

Yes

Yes

—

No

Yes

Yes

Yes

Yes

Yes

—

aultraviolet C irradiation (254 nm). bX-irridation. cChemical alkylating agents (e.g., methylmethansulfonate). dBleomyein and Streptonigrin.

eate this. It is notable that the amino (-NH3) end of all of the gene products discussed above are very different, often with little or no homology present. The presence of the same mutated gene in at least four different complementation groups of ataxia-telangiectasia is also yet to be explained. TREATMENT/MANAGEMENT Because the ataxia and ocular apraxia usually occur in the first few years of life, whereas the characteristic oculocutaneous telangiectasias often appear much later, at about 6 or 8 years of age, the diagnosis of ataxia-telangiectasia may be suspected as one of several possible causes of the neurologic signs, but not confirmed as the cause, for an interval of as long as several years during childhood. This may complicate management as well as genetic counseling to the family. As soon as possible appropriate cytogenetic studies should be carried out. Once the diagnosis has been made, the patient should be carefully followed during childhood for evidence of development of a lymphoma, especially a B-cell lymphoma in early childhood and a T-cell leukemia later. Older patients should be screened for development of other types of cancer as well. Infections should be treated promptly and aggressively because of the immune dysfunction often present in these patients. X-ray exposure must be carefully limited, especially therapeutic X-irradiation, as well as chemotherapeutic drugs used to treat lymphomas, which these patients are prone to develop. These modalities are often valuable and useful in ataxia-telangiectasia patients, but at carefully controlled, lower doses. An underlying ataxia-telangiectasia should be considered in all young patients with lymphoma or leukemia prior to administration of therapeutic X-irradiation or chemotherapy. Obligate heterozygotes and other family members should be advised that the risk of developing cancer in an ataxia-telangiectasia heterozygote may be two to three times that of age-matched normal persons. This risk in female heterozygotes may be as high as five times normal for breast cancer. However, how large this increased risk is, and even whether it is present, are subjects of controversy in the current literature, and patients should be advised of the uncertainty due to this as well. Since the prime mode of early detection of mammary cancer at present is the mammogram, which uses X-rays (which are also carcinogenic), and since ataxia-telangiectasia heterozygote cells

are hypersensitive to X-rays in some (although not all ) assays, this may pose a dilemma for female ataxia-telangiectasia heterozygotes. However, since the energy level of the X-rays used in mammograms is extremely low, it would seem prudent for these women to obtain regular mammograms, as well as to practice frequent periodic self-examination of their breasts. FUTURE DIRECTIONS Many more studies of the ATM gene need to be, and soon will be, done. Why is it defective in all complementation groups of ataxia-telangiectasia? Are other genes involved in the etiopathogenesis of this disease as well? What is the phenotype of ATM knockout mice? The answers to these and other questions should be forthcoming. Perhaps even more important than its use in elucidating the pathophysiology of ataxia-telangiectasia is the role the cloned ATM gene is expected to play in delineating the significance of the carrier state for this gene. What proportion of patients who are carriers develop cancer, and what kinds of cancer do they develop? Do the cancers that arise in these patients have a different course, prognosis, or biologic behavior? Is the carrier state associated with an increased sensitivity to X-irradiation or chemotherapy, and, if so, can the effectiveness of radiotherapy and chemotherapy in all patients be improved by culling these patients out and designing different dose regimens for them, with other patients able to receive higher (and thus more effective) doses?

NIJMEGEN BREAKAGE SYNDROME/ BERLIN BREAKAGE SYNDROME/ ATAXIA-TELANGIECTASIA FRESNO These three disorders all show the same cellular abnormalities seen in ataxia-telangiectasia. However, the first two (the breakage syndromes) have different clinical manifestations (microcephaly and mental retardation, without the neurologic changes or the telangiectatic eye and skin changes seen in ataxia-telangiectasia), whereas ataxia-telangiectasia Fresno shows the clinical manifestations of all three diseases. Presumably all of these entities are associated with a predisposition to develop lymphomas and cancer, but the case numbers reported have been so small that the data base is inadequate. The term “breakage” in the first two disorders refers to chromosome breakage in their karotypes. They are clinically very similar,

764

SECTION IX / DERMATOLOGY

and are separated on the basis of ability of their cells to complement each other in the complementation assay discussed above for ataxia-telangiectasia. Since cells from both breakage syndromes can complement those from all complementation groups of ataxiatelangiectasia they have been assigned their own complementation groups V1, for the Nijmegen breakage syndrome, and V2, for the Berlin breakage syndrome, respectively. Linkage studies indicate that neither breakage syndrome is defective at the ATM locus; this should (and presumably will) become elucidated by molecular studies using the cloned ATM gene.

BLOOM SYNDROME BACKGROUND Bloom syndrome is a rare, autosomal recessive disorder characterized by small body size, one or both of two distinctive types of cutaneous lesions, characteristic cytogenetic abnormalities in cultured cells, and a marked predisposition to develop cancer at a young age. The cutaneous eruptions consist of a telangiectatic rash over sun exposed areas appearing in early childhood and more variable hyper- and hypopigmented areas appearing primarily on the trunk. The cancers appear in multiple organs and tissues. Bloom syndrome was first described by Dr. David Bloom, a New York City dermatologist, in 1954. Of the multiple abnormalities present in cultured cells derived from Bloom syndrome patients (discussed below), the most characteristic are a high rate (about 10 or more times normal) of sister chromatid exchanges and multiple quadriradial figures representing abnormal linkages between two homologous chromosomes seen in cultured cells treated with bromo-deoxy-uridine. In 1995, Dr. James German and his colleagues identified and cloned the gene responsible for these and other anomalies in Bloom syndrome, which they named the BLM gene. This was achieved using a new cloning method, somatic crossover point mapping (SCPM) based on the BLM gene itself. SCPM should allow extremely rapid and precise localization of a recessive gene responsible for virtually any phenotype identifiable in cultured cells. It thus should be applicable to the study of many other recessively inherited diseases. Therefore, in addition to being of interest in the etiopathogenesis of Bloom syndrome, the BLM gene is likely to become an important tool in the study of inherited diseases generally, and to provide a rapid means for gene localization and cloning. CLINICAL FEATURES Bloom syndrome is extremely rare. As of August 1993, fewer than 170 patients had been entered into the Bloom Syndrome Registry maintained by Dr. James German at the New York Blood Center in New York City. Small, but proportionate, body size is the most characteristic and consistent sign of Bloom syndrome. Only four of 82 males and three of 53 females in the Bloom Syndrome Registry for whom this information was available as of 1995 had birth weights within two standard deviations of the normal mean. This small size persists throughout life. Almost as characteristic as small body size is the marked tendency of Bloom syndrome patients to develop cancer at a moderately early age. These cancers are of virtually all types, with 41 carcinomas, 21 leukemias, and 4 sarcomas reported in the 165 individuals in the Bloom Syndrome Registry as of August 1993. A telangiectatic cutaneous eruption, varying from severe to mild to undetectable, occurs over sun exposed areas, especially of the face where it tends to involve the cheeks and nose in a strikingly “butterfly” distribution (Figs. 81-18 and 81-19). Less commonly, a similar, but milder, eruption is present on the dorsa of the hands and forearms. When the facial rash is severe, it may also

Figure 81-18 Young mother with Bloom’s syndrome, demonstrating the typical telangiectatic eruption in a butterfly distribution over the malar eminences and nasal bridge. (From Friedberg EC. DNA Repair. New York: WH Freeman, 1985, with permission; and courtesy of Dr. James German.)

involve the ears, neck, and suprasternal area. In these cases, the shoulders and chest may show a mild telangiectatic eruption as well. This telangiectatic eruption does not involve other areas, however, and a telangiectatic rash on the legs, upper arms, trunk or buttocks essentially rules out Bloom syndrome as a diagnosis. This eruption is not congenital, as once thought, but rather usually appears in the first two years of life. When it is severe, loss of inferior eyelids and blistering of the inferior lip are common. Younger affected siblings often show less severe skin changes, due to being better protected by the parents. The eruption tends to be more severe in boys, possibly accounting for the fact that slightly more boys than girls have been reported to the Bloom Syndrome Registry (94 vs 71 cases). In marked contrast to the sun-damaged skin seen in xeroderma pigmentosum and to the other eruption characteristic of Bloom syndrome (see below), there is no dyspigmentation associated with this eruption. Regardless of whether the eruption is present, Bloom syndrome patients tend to be hypersensitive to sunlight, in areas exposed to sun in infancy, throughout their lives (Fig. 81-20). In other areas, however, they are not particularly sun-sensitive. The second characteristic cutaneous stigma seen in many cases of Bloom syndrome consists of well-circumscribed areas of (usually) coffee-colored, cafe au fait type hyperpigmentation, with or without areas of hypopigmentation, located mainly over the trunk but which can occur anywhere. These areas vary widely in size and shape. The hypopigmented areas are most easily visualized using a Wood’s lamp. These changes are more easily detected in dark complexioned and Black patients, and are most often noted in early childhood, rather than in infancy.

CHAPTER 81 / XERODERMA PIGMENTOSUM

Figure 81-20 German.)

Figure 81-19 Young man with Bloom’s syndrome. With much less severe “butterfly rash” but present some related lip lesion. (Courtesy of Dr. James German.)

Immunodeficiency, which is generalized and very variable, is characteristic of Bloom syndrome. Most patients experience at least one severe bacterial pulmonary infection during infancy or childhood, which, if inadequately treated, may lead to severe chronic lung disease. After cancer, this has been the most common cause of death in Bloom syndrome patients. A variable degree of vomiting and diarrhea also occurs during infancy; whether it is related to an immunodeficiency is unknown. This can rapidly produce life-threatening dehydration. Other, less constant characteristics of Bloom syndrome include a high-pitched voice, a tendency to develop diabetes mellitus, a tendency to have minor anatomic anomalies, mildly restricted intellectual ability, and, in men, small testes and a total failure of spermatogenesis; in women, reduced fertility and a short menstrual life. DIAGNOSIS The diagnosis of Bloom syndrome usually depends on recognition of the small but proportionate stature in combination with the characteristic telangiectatic facial rash. It is unusual for the diagnosis to be made if these two features are not recognized. The small stature can be detected as in utero growth deficiency. Conversely, if the birth weight is within the normal range and the postnatal height is in the third percentile or greater, it is very unlikely that the patient has Bloom syndrome. The diagnosis of Bloom syndrome can be confirmed cytogenetically using the Sister Chromatid Exchange (SCE) assay (see below). Because the characteristic telangiectatic cutaneous eruption is not always recognizable, Dr. James German has recommended that cells obtained from the following groups of patients with abnormally small, but proportionate, stature also be examined with this assay: 1. Those with excessive numbers (i.e., more than five) cafe au lait macules, especially if hypopigmented macules are also present.

765

Lip of patient in Fig. 81-19. (Courtesy of Dr. James

2. Those with unexplained immunodeficiency. 3. Those with an unexplained restriction in intelligence. 4. Those in whom diabetes mellitus develops later than the usual age of onset for Type I and earlier than this age for Type II diabetes mellitus. 5. Infertile men with abnormally small testes for which no explanation can be found. 6. Women with early onset of menopause. 7. Those patients who develop any type of cancer, especially if this occurs at a relatively young age for the type of cancer in question. Not surprisingly, Bloom syndrome is often misdiagnosed as another type of dwarfism. Of these, it is perhaps most commonly confused with Russel-Silver dwarfism. GENETIC BASIS OF DISEASE Bloom syndrome is transmitted by autosomal recessive inheritance with no evidence of genetic heterogeneity. This distinguishes it from most of the other major entities (xeroderma pigmentosum, ataxia telangiectasia, and Fanconi Anemia) discussed in this chapter, all of which are genetically heterogeneous. Cultured cells derived from patients with Bloom syndrome show a number of distinctive features. Some of these are also detected in circulating lymphocytes in Bloom syndrome patients. These abnormalities include chromosome breaks, gaps, translocations, and an increased frequency of intra- and inter-chromosomal strand exchanges. The latter are detected in the Sister Chromatid Exchange assay. The results of an SCE assay on cells derived from a patient with Bloom syndrome are shown in Fig. 81-21. Note the presence of numerous sister chromatic exchanges (SCEs) as well as a balanced (i.e., symmetrical) quadriradial figure, indicative of a recombination event between two homologous chromosomes. These recombinations between homologous chromosomes are responsible for “crossovers” which occur in Bloom syndrome cells, similar to the crossovers between portions of homologous chromosomes that otherwise occur mainly in meiosis. These crossovers are exploited in a novel gene localization technique, Somatic Crossover Point Mapping (SCPM), which has recently allowed identification and cloning of the gene responsible for Bloom syndrome, the BLM gene (see below).

766

SECTION IX / DERMATOLOGY

Figure 81-21 Chromosome preparation of a metaphase cell from a patient with Bloom syndrome, processed through two cell cycles so as to reveal sister chromatic exchanges. Note the presence of numerous exchanges per chromosome (identified as exchanges between dark and light strands) as well as a homologous chromosome exchange (quadriradial figure) between the long arms of chromosome 1. (Reproduced with permission from Cell cover image, V.83, #4, Nov. 17, 1995.)

Bloom syndrome cells in culture have accumulated markedly elevated numbers of mutations, compared to normal cells, in all loci in which this has been examined. These loci have included both coding sequences for actively transcribed genes and noncoding repetitive DNA. Bloom syndrome cells in culture show a prolonged generation time and in particular a long DNA synthetic (S) phase. The rate of nascent DNA chain elongation is retarded and the distribution of DNA replication intermediates is abnormal. Bloom syndrome cells also show increased sensitivity to DNA damaging agents, including ultraviolet light, mitomycin C, ethyl methanesulfonate and N-nitroso-N-ethylurea. However, not all Bloom syndrome cells show these hypersensitivities and they do not show up in some standard assays. Moreover, no defect in a DNA repair pathway has been found in Bloom syndrome cells. A number of enzymes associated with DNA replication and/or repair have been reported to have abnormal activities in Bloom syndrome cells. These include DNA ligase I, topoisomerase II, thymidylate synthetase, uracil-DNA glycosylase, N-methylpurine DNA glycosylase, 0 6-methylguanine methyltransferase, and superoxide dismutase. Interestingly, some but not all Bloom syndrome cell lines have shown these defects, and the nature of abnormalities observed has been very variable. For example, both increased and decreased activities of DNA ligase I have been reported in different cell lines by different authors, and studies on a purified uracil glycosylase from Bloom syndrome cells showed that a number of properties of the enzyme, itself, are markedly abnormal. In 1994 James German and his colleagues noted that, in Bloom syndrome patients who are the product of nonconsanguineous matings, a proportion of peripheral blood lymphocytes are phe-

notypically normal. Based on this observation they developed a new technique, which they termed “Somatic Crossover Point Mapping” (SCPM). In this system, peripheral blood lymphocytes of Bloom syndrome patients are analyzed very much like progeny meiotic cells, using crossovers to obtain extremely precise mapping information. It allowed rapid localization and then cloning of the BLM gene, which is mutated in Bloom syndrome cells. The normal phenotype is produced by a crossover event within this gene between the parental mutations. Previously, conventional methods had already localized the BLM gene to chromosome band 15q26.1. The BLM gene encodes a 4437 bp cDNA encoding a 1417 amino acid peptide with homology to the RecQ helicases of E. coli. The SCM technique is extremely interesting because it offers a way to exploit the BLM gene to study other inherited diseases in which the cells show associated phenotypic abnormalities. MOLECULAR PATHOPHYSIOLOGY OF DISEASE Whether the defective BLM gene can account for all of the clinical, cellular, and biochemical abnormalities present in Bloom syndrome is at present unknown; however it would appear to be an excellent candidate for many of them. The BLM gene is homologous to the RecQ helicases in E. coli; these helicases are a subfamily of DExH box-containing DNA and RNA helicases. Although no universal function for the RecQ helicases has been postulated to date, a number of genes in the DExH family have been implicated in DNA repair processes. The BLM gene product may also have homologous functions to those of the yeast SGS 1 gene. SGS 1 mutants show slow growth, poor sporulation, chromosomal nondisjunction in mitosis, missegregation in meiosis, and elevated recombination frequencies. Also, an interaction between the BLM protein and topoisomerase II has been suggested. MANAGEMENT/TREATMENT Upon establishing a diagnosis of Bloom syndrome, the physician should immediately notify the Bloom Syndrome Registry (even though the Registry was officially closed several years ago). The Registry (c/o Dr. James German, The New York Blood Center, 213 East 31st St., New York, NY 10023) is a very valuable source of information and support. Patients with Bloom syndrome should avoid sunlight, although the importance of this is not nearly as great as in xeroderma pigmentosum. The clinician should be constantly on the alert for development of cancer in these patients, paying special attention to such signs as hoarseness or throat pain, mild dysphagia, hematochezia or melena, a breast lump, a positive Papanicolaou smear, abdominal discomfort or pain, intussusception, or convulsions, which might otherwise be given less notice, especially in a young patient. Unless or until the prognosis of leukemia is shown to be better if the disease is diagnosed and treated before symptoms develop, however, periodic hematologic surveillance of children with Bloom syndrome is unnecessary and may be inappropriate. Since children with Bloom syndrome are of small stature, and also tend to be poor eaters; they should be encouraged but never coerced to eat well, and a vitamin supplement provided each day. Gastric intubation is currently being tried in a few patients, but no long-term benefit has yet been demonstrated. Growth hormone levels in Bloom syndrome patients have been found to be normal to low normal. Exogenous administration of growth hormone has resulted, in a few cases, in development of cancer, both in Bloom syndrome patients and in other patients. This has tempered enthusiasm for administration of growth hormones to these patients.

767

CHAPTER 81 / XERODERMA PIGMENTOSUM

Table 81-4 Clinical and Molecular Characteristics of Fanconi Anemia Complementation group

Relative frequency

Chromosome location of gene

A

66.0

16q24.3

B C

4.3 12.7

Unknown 9q22.3

D E

4.3 12.7

3q22-26 Unknown

Characteristics of gene product 163 h Dalton protein Primary nuclear localization Unknown 63-kDa protein Primary cytoplasmic localization Unknown Unknown

Several additional coplementation groups likely.

Counseling of patients and their families, carefully advising them of the nature of the disease, is probably the most important service the physician can provide. It is important that parents of a Bloom syndrome child be advised of the risk of having another affected child should they conceive again. FUTURE DIRECTIONS More precise methods for diagnosing Bloom syndrome using the BLM gene in molecular assays will now be developed. Also the precise etiopathogenesis of this disease can now be examined in greater detail in molecular studies using this gene. By inactivating the BLM gene in other types of mutant cells in culture, perhaps by using incisional mutagenesis or antisense DNA or RNA, the SCPM method should be applicable to rapid precise mapping of many other genes unrelated to the BLM gene.

FANCONI ANEMIA BACKGROUND Fanconi anemia is a rare, autosomal recessive disorder characterized by diverse congenital anomalies, a predisposition to develop aplastic anemia, and an increased risk to develop cancer, particularly acute myelogenous leukemia. The clinical features of the disease are quite variable, making a diagnosis on basis of clinical findings quite difficult. This clinical heterogeneity is both interfamilial and intrafamilial. Cases in which no congenital anomaly is recognized, but who nonetheless have cell culture abnormalities and develop bone marrow changes characteristic of Fanconi anemia, are said to have the DiamondBlackfan syndrome, also known as the Estren-Dameshek variant of Fanconi anemia. Cells in culture derived from patients with Fanconi anemia show several distinctive characteristics, among which are an increased number, compared to cells derived from normal individuals or first-degree relatives of Fanconi anemia patients, of chromosome breaks in untreated cultures, as well as hypersensitivity to both the cytotoxic and clastogenic effects of DNA crosslinking agents such as mitomycin C, psoralen plus UVA light, and diepoxybutane. The latter property has formed the basis for development of a diagnostic test for the disease in cultured cells, known as the DEB, or diepoxybutane, test. Because the clinical manifestations of the disease are so variable, the definitive diagnosis of Fanconi anemia is currently made based on this test. Five genetic complementation groups, A, B, C, D, and E, have been defined (Table 81-4). It is very likely that several others will be found. Fanconi anemia was first described in three brothers with congenital anomalies, anemia, and a fatty aplastic bone marrow by Guido Fanconi in 1927. The diepoxybutane test for diagnosis of

Fanconi anemia was developed by Arleen Auerbach and her colleagues in 1977. Androgen therapy, one of the clinical mainstays for treatment of hypoplastic anemia short of bone marrow transplantation or gene therapy, was introduced by Nasrollah Shahidi, previously a junior associate of Fanconi. In 1992 one of the genes responsible for Fanconi anemia, the Fanconi anemia [group] C gene, was cloned by Manuel Buchwald and his associates. In 1996 two groups simultaneously reported cloning of the FAA gene. Quite recently, Lin and Walsh and their associates at the National Institutes of Health (US) Clinical Center in Bethesda, MD, performed the first successful gene therapy for Fanconi anemia using the FAC gene. It has been proposed that Fanconi anemia heterozygotes comprise at least 0.5% of the general population. It has also been proposed that many cases of aplastic anemia without Fanconi anemia may be etiopathogenically related to this carrier state. CLINICAL FEATURES The incidence of Fanconi anemia is thought to be similar to that of xeroderma pigmentosum, approximately 2–4 per million live births. The true incidence may be significantly higher, however, with many cases not being diagnosed. The disease appears to occur worldwide in all ethnic groups. It is much more common among Ashkenazi Jews. A recent study showed that over 1% of normal Jewish individuals of Ashkenazi ancestry are carriers of a gene for Fanconi anemia, complementation group C. This would indicate (Hardy-Weinberg law) that matings within this group would produce affected individuals in over 1/40,000 births. Clinical features seen in some but not all patients with Fanconi anemia include the following: (1) short stature, with over 50% of patients below the fifth percentile in height, sometimes associated with low birth weight or “failure to thrive”; (2) congenital anomalies of the thumb and radius (Fig. 81-22); (3) other congenital musculoskeletal anomalies, including hip, spinal and rib malformations, and scoliosis; (4) congenital structural renal malformations, including agenesis of a kidney, fused kidneys, and misshapen kidneys; (5) hyperpigmented areas on the skin, sometimes resembling coffee colored cafe au fait macules. These are often large and irregular in shape (Fig. 81-23); (6) microcephaly; (7) microphthalmia; (8) congenital gastrointestinal malformations, with many patients with no observable internal defects also experiencing digestive problems; (9) hypogonadism; (10) congenital cardiac defects, especially involving the atrial or ventricular septum; and (11) mental retardation, sometimes manifested only as a learning disability without other evidence of mental deficiency. Expression of these features is extremely variable, even between affected siblings; some patients who develop aplastic anemia or acute

768

SECTION IX / DERMATOLOGY

Figure 81-22 Hands and arms of patient with Fanconi anemia showing radial deformities. Such deformities are heterogeneous and not present in every patient (courtesy of Dr. N. Shahidi, University of Wisconsin, Madison, WI).

myelogenous leukemia and who have a positive DEB test show none of these features. Viruses, including those commonly encountered in childhood, particularly the herpes varicella/zoster virus, may elicit abnormal responses in patients with Fanconi anemia, particularly thrombocytopenia and hypoplastic anemia. These viruses include, besides herpes varicella/zoster, parvovirus B 19 (the cause of fifth disease), hepatitis viruses, the Epstein-Barr virus, and cytomegalovirus. Not uncommonly the disease is first suspected and diagnosed based on a child’s response to infection with one of these, particularly the former. Since most children are immunized against a number of other viruses, including rubella (German measles), rubeola (measles), mumps, and influenza A, it is unknown whether these viruses might also induce thrombocytopenia or hypoplastic anemia in Fanconi anemia patients. Bacterial infections may also induce thrombocytopenia or aplastic anemia in patients with Fanconi anemia. In some cases, the diagnosis of Fanconi anemia is only suspected after development of aplastic anemia or, less commonly, myelodysplastic syndrome or even acute myelogenous leukemia. Anemia in Fanconi anemia patients is characteristically macrocytic, often with elevated fetal hemoglobin levels. Patients with Fanconi anemia should be expected, sometime during childhood or early adulthood, to develop bone marrow failure and/or leukemia. Moreover, especially in patients who survive childhood, there is an increased risk of development of cancer not involving the bone marrow. These, as far as is known, may occur anywhere and may be of any type; especially common are squamous-cell carcinomas of the oral cavity, esophagus, vagina, and cervix uteri. A more controversial issue is whether patients with Fanconi anemia have an increased susceptibility to oxygen free radicals in their cells and tissues. Hypersensitivity to agents which produce oxygen free radicals has been demonstrated convincingly in Fanconi anemia patients’ cells in tissue culture (discussed below), but this hypersensitivity is lost in some of these cell lines after undergoing transformation (immortalization) in tissue culture. These transformed cells continue to test positive in other assays for Fanconi anemia. It has been proposed that patients with Fanconi anemia also show clinical evidence of this hypersensitivity, such as defective responses to bacterial infections, which induce oxygen free radical formation by neutrophils and other inflammatory

Figure 81-23 Large pigmented macule in chest of boy with Fanconi anemia (courtesy of Dr. N Shahidi).

cells, but this evidence can also be explained by other defects in Fanconi anemia cells, as noted above. Patients with Fanconi anemia eventually require therapy for their disease, and thus develop side effects of this therapy. Most patients receive blood transfusions, leading to a risk of iron overload. Those who receive androgen therapy develop the usual side effects of this therapy, including virilization in females and a risk of hepatotoxicity in both sexes. Increased susceptibility to infection and graft vs host disease are important potential complications of bone marrow transplantation. DIAGNOSIS Every author of a report of a large series of cases of Fanconi anemia has emphasized the enormous clinical heterogeneity of the disease, even among siblings and, in one report, even between monozygotic twins. It is thus very difficult to make specific recommendations as to when a diepoxybutane (DEP) test should be obtained. Certainly a patient with congenital anomalies affecting the thumb or radius should be suspected of having the disease, as should a small patient with thrombocytopenia or hypoplastic, macrocytic anemia not due to folate or vitamin B 12 deficiency. Beyond this it is difficult to make a specific recommendation. The DEB test is the gold standard for the diagnosis of Fanconi anemia, and should be carried out whenever the diagnosis is seriously entertained. This test is performed by the following laboratory:

CHAPTER 81 / XERODERMA PIGMENTOSUM

Figure 81-24 Chromosome preparation of a metaphase cell from a patient with Fanconi anemia. Note acentric (a), dicentric (d), and quadriradial (q) chromosomes (courtesy of Dr. N. Shahidi, University of Wisconsin).

c/o Dr. Arleen Auerbach The Rockefeller University 1230 York Avenue New York, NY 10021 212-327-7533 It is strongly recommended that the physician telephone in advance regarding optimal ways to submit samples, which must be live (never fixed) cells. A more challenging problem is what to do with patients who test negative in the DEB assay but who have other manifestations of Fanconi anemia. Despite the great helpfulness of this test, we do not know for certain that it is 100% sensitive in detecting cases of Fanconi anemia, or 100% specific in excluding patients with other disorders. The DEB test has been used very successfully to identify affected fetuses in prenatal diagnostic examinations of amniocytes or chorionic villi cells, and should be carried out in all cases where the parents are known heterozygotes. A more specific test, based on a known mutation in the Fanconi anemia C gene, has been used successfully in Ashkenazi Jews to identify heterozygotes (see below). GENETIC BASIS OF DISEASE Pedigrees of families of patients with Fanconi anemia have been consistent with autosomal recessive inheritance. We are unaware of any reported case in this is untrue or in doubt. Cultured cells derived from patients with Fanconi anemia are both hypersensitive to killing by agents that produce interstrand crosslinks in cellular DNA and show markedly increased numbers of breaks and other abnormalities in metaphase chromosomes following exposure to such agents. This latter characteristic has formed the basis for the most definitive test for this disease yet available, the diepoxybutane (DEB) assay, in which cells derived from a patient suspected of having Fanconi anemia are exposed to the DNA crosslinking agent, diepoxybutane, and subsequent metaphases arrested and analyzed cytogenetically. Fanconi anemia cells show increased numbers of acentric and dicentric, as well as quadriradial, chromosomes, which result from chromosome breaks, compared to normal cells, both with and without drug treatment, although the increased frequency of abnormalities is much easier to detect in treated cultures (Fig. 81-24). In up to

769

30% of cases, cells derived from circulating lymphocytes fail to show these abnormalities, which are then detected in cells derived from skin fibroblasts. This is thought to be due to recombination events in the precoursers of the former cell type. If the patient is a compound heterozygote, such recomination could produce phenotypically normal cells doubly mutated or one allele and wild type on the others. With a selective growth advantage, these cells would then become predominant among circulating lymphocytes. Interestingly, the Fanconi anemia cells show a wide range in their numbers of chromosomal abnormalities, although these are consistently outside of the range of numbers seen in normal cells. Despite this variability, the diepoxybutane (DEB) test remains the gold standard on which the diagnosis of Fanconi anemia is made, due to the very variable clinical presentation of the disease, which virtually precludes use of clinical criteria to make the diagnosis in most cases. In addition to diepoxybutane, Fanconi anemia cells are hypersensitive to other DNA crosslinking agents, including mitomycin C, nitrogen mustard (mechlorethamine), cyclophosphamide, cisplatin, and psoralen plus ultraviolet A (long wavelength) radiation. Elegant cell cycle studies have shown that this hypersensitivity is associated with prolongation of, or arrest in, the G2 phase of the cell cycle. Not only the first G2 phase but also subsequent G2 phases are prolonged after exposure to these agents. In addition to its use in establishing the diagnosis of Fanconi anemia, the diepoxybutane test is also the basis for a complementation assay in patients known to have Fanconi anemia. Fused cells (heterokaryons) from patients in different complementation groups show numbers of metaphase chromosome abnormalities, following treatment with diepoxybutane or other DNA crosslinking agents, similar to those of similarly treated normal cells, whereas fused cells from patients in the same complementation group continue to show increased levels of chromosome abnormalities following this treatment. In this way five complementation groups, A, B, C, D, and E, of Fanconi anemia patients have now been identified. Initially, Fanconi anemia cell lines were assigned to complementation groups strictly on the basis of whether or not they complemented, in this assay, within group A. Unfortunately, all other cell lines were initially classified as “group B,” regardless of whether or not they complemented each other in this assay. In consequence, some of these cell lines, once classified as in “complementation group B,” are now in different complementation groups. One example is cell line HSC 62 (GM 13023), once classified as in group B, now in group D. This has led to some confusion in the literature. In reviewing the research literature on Fanconi anemia cell lines, it is important to determine precisely which cell line was examined in each study, and to corelate this information with the current complementation group classification of each cell line. Such a classification is to be found in Liu, et al., Table 1. Since only a rather small number of cell lines have been analyzed by complementation group analysis to date, it is likely that additional ones exist, beyond the five already identified, in Fanconi anemia. A gene that corrects hypersensitivity to crosslinking agents in Fanconi anemia group C, and therefore named FACC now FAC, was isolated and portions of it cloned by Manuel Buchwald and his colleagues in 1992, who used this property as a selection method, which hastened the progress of their work. Unfortunately, this has not proved a successful strategy, to date, for isolation and cloning of other Fanconi anemia genes. The FAC gene does not comple-

770

SECTION IX / DERMATOLOGY

ment Fanconi anemia cells in other complementation groups. The cDNA codes for a protein of 63 kDa, with alternative 5' and 3' untranslated sequences. Examination of available data banks has failed to identify any different human or nonhuman protein with significant amino acid sequence homology. The gene itself appears to be larger than 100 kb and has 14 exons. By in situ hybridization it has been mapped to chromosome band 9q22.3. The gene product localizes primarily in the cytoplasm. A number of specific mutations have now been identified in Fanconi anemia, complementation group C, patients. These include a splice mutation in intron 4 (IVS4+4A to T), which is the predominant mutation in Ashkenazi Jews, and mutations 322delG, Q13X, R185X, and D19SV. Among the Ashkenazi, the IVS4 mutation is associated with much more severe disease than is the 322delG mutation. Alterations introduced into the gene product by site-directed mutagenesis, along with the locations of the Fanconi anemia mutations discovered to date, indicate that the carboxy (-COOH)-terminal portion of the protein is essential for activity. There is no known homolog to the human FAC gene product. The mouse FAC gene, found using probes based on the human sequence, is 68% homologous to it. The mouse FAC protein is identical at 65% and similar at 78% of its amino acid residues to the human FAC protein; where examined, it has a similar number of exons. The FAA gene cDNA codes for a protein of 163 kDa with two overlapping bipartite nuclear localization signals and a partial lencine zipper consensus region. It localizes primarily in the nucleus. Again, examination of available data banks has failed to identify any different human or nonhuman protein with significant amino acid homology. The gene localizes to chromosome 16q24.3. MOLECULAR PATHOPHYSIOLOGY OF DISEASE The molecular basis for the increased numbers of chromosome breaks and for the hypersensitivity of Fanconi anemia cells in culture to DNA crosslinking agents remains incompletely understood. It has been suggested that both defects may be due to inappropriate activation of the gene rearrangement mechanism, which normally is functional only in T and B cells undergoing final differentiation. Our laboratory has identified two endonuclease complexes in the chromatin of normal cells, pI 4.6 and pI 7.6, which cleave psoralen plus ultraviolet A light induced monoadducts as well as interstrand crosslinks. The complex, pI 7.6, acts primarily on the former type of adduct; the complex, pI 4.6, acts primarily on the latter. Using a molecularly defined substrate DNA in which the adducts were precisely localized, the exact sites of cleavage of these adducted DNAs were determined. The endonuclease complex pI 4.6, from Fanconi anemia group A cells was deficient in ability to incise DNA containing intrastrand crosslinks activity, whereas the complex, pI 7.6, was able to incise DNA containing the monoadducts. This correlates with a higher degree of sensitivity of these cells to DNA interstrand crosslinking agents. Moreover, when normal endonuclease complexes were introduced into these cells by electroporation, repair function, as measured by a computerized UDS assay based on autoradiography, was restored to normal. Fanconi anemia group A cells are also deficient in a DNA binding protein; how this relates to the endonuclease defect is under investigation. Other laboratories have also identified DNA repair deficiencies in Fanconi anemia cell lines, but results have been inconsistent, perhaps because well-defined systems were not used. At present, the function of the FAC gene product remains unknown.

While it would seem likely that it, too, is active in a DNA damage recognition and/or repair process, this has been thrown into question by the discovery that the FAC protein is mainly localized to the cytoplasm. Even if the FAC protein is primarily cytoplasmic, however, it could still function as a DNA repair enzyme. A number of other proteins that act on DNA, such as some hormone receptors and some enzymes active on DNA, are known to be localized mainly in the cytoplasm, entering the nucleus only when specifically activated to do so. Alternatively, the FAC protein has been reported to bind to the cyclin-dependent kinase, cdc2, suggesting a possible role of this protein in cell-cycle regulation. The function of the FAA gene product, which localizes to the nucleus, is similarly unknown. In addition to increased sensitivity to both the cytotoxic and clastogenic effects of DNA crosslinking agents, Fanconi anemia cells in culture show at least three additional general features: (1) prolongation, compared to normal cells, of the G2 phase of the cell cycle (this is in addition to the additional prolongation of G2 following treatment with DNA crosslinking agents; (2) hypersensitivity to oxygen; and (3) overproduction of tumor necrosis factor (TNF-α). In addition to these features, X-irradiation of Fanconi anemia cells in G2 also produces increased numbers of chromatic aberrations seen in the following metaphase, a feature Fanconi anemia cells share with a number of other genetic diseases reviewed in this chapter, as discussed above. Fanconi anemia cells in culture grow very poorly at ambient oxygen levels (atmospheric 20%) but well at reduced oxygen tensions (atmospheric 5%). Cell-cycle/flow cytometric studies indicate that this oxygen sensitivity is associated with the elongation of the G2 phase also seen in these cells; the G2 phase becomes shortened at lower oxygen tensions. Even in normal cells, treatment with oxygen or reactive oxygen radicals produces elongation of the G2 phase. It has been suggested that this effect may be due to interference of oxygen with the function of one or more DNA topoisomerases. The hypersensitivity of Fanconi anemia cells to oxygen and to reactive oxygen species might be due to one or more of three mechanisms: (1) overproduction of reactive oxygen species in Fanconi anemia cells due to malfunction of the very complex cellular system that controls this rate of production; (2) deficient antioxidant defense mechanisms; or (3) an intrinsic inability of Fanconi anemia cells to tolerate oxygen-induced damage. Data have been reported (from a number of different laboratories) that are consistent with all three mechanisms at work in Fanconi anemia cells. However, not all cell types in Fanconi anemia patients show these deficiencies, or show them to the same extent or in the same way. In erythrocytes and leukocytes of these patients, activities of superoxide dismutase (SOD) are decreased; but in their fibroblasts, activity of at least one SOD, magnesium-SOD, is paradoxically increased. Also, transformation of Fanconi anemia fibroblasts in culture with the SV40 large T-antigenic protein has been reported to cause them to lose their hypersensitivity to oxygen. For these reasons, Liu et al. have argued that oxygen hypersensitivity is a secondary, rather than a primary, effect of mutations in Fanconi anemia genes. Whether primary or secondary, the hypersensitivity of Fanconi anemia cells in culture to oxygen, and the associated abnormalities in these cells in antioxidant defense mechanisms, has prompted clinical use of antioxidants in Fanconi anemia patients. Following a report that treatment of Fanconi anemia fibroblasts so as to

CHAPTER 81 / XERODERMA PIGMENTOSUM

increase their activity of copper-zinc superoxide dismutase (CuZn-SOD)—an important antioxidant that reduces the oxygen species, superoxide anion (O2)—decreased the cytotoxic effect of mitomycin C, pilot clinical trials using CuZn-SOD were initiated. These produced results suggesting that, indeed, chromosome breakage, at least, is reduced in Fanconi anemia patients receiving this treatment. These findings have also prompted Shahidi to recommend high daily doses of dietary antioxidants for Fanconi anemia patients. It appears likely that the prolongation of the G2 phase of the cell cycle, increased numbers of G2 irradiation induced chromatic anomalies, and even TNF-α overproduction by Fanconi anemia cells in culture are also secondary events. The G2-to-M cell-cycle transition is genetically controlled to proceed only if DNA damage has been repaired. Thus excess DNA damage, or inadequate DNA repair, might be responsible for G2 prolongation, rather than a primary cell-cycle disturbance. Also, TNF-α acts preferentially during G2 and is known to be able to prolong it. Thus TNF hypersecretion by these cells may be due to an autocrine stimulatory loop. We thus agree with Liu et al. that “A defective response (recognition or processing) to DNA lesions is likely to be a key feature of Fanconi anemia.” TREATMENT/MANAGEMENT Once a diagnosis of Fanconi anemia has been established by the diepoxybutane (DEB) test, the patient should be entered into the International Fanconi Anemia Registry (IFAR). If the DEB test is done at the Rockefeller University, this is virtually automatic, since the IFAR is in the same location. An excellent manual for patients and physicians (Frohnmayer and Frohumayer: Fanconi Anemia. A Handbook for Families and their Physicians) is published by the Fanconi Anemia Research Fund, Inc., 1902 Jefferson Street, Eugene, OR 97405 (tel: 503-687-4658; fax: 503-687-0548). This is also an excellent support group with which families of patients should interact. Management of patients with Fanconi anemia has, until very recently, consisted largely of monitoring the blood and bone marrow for appearance of a premalignant or malignant clone or for development of aplastic anemia, treatment with androgens where appropriate to delay the latter, and bone marrow transplantation when one or the other of the above events seems to have progressed to the point to warrant this high-risk procedure. This entails some challenging decisions. For example, bone marrow transplants are less likely to succeed after long androgen therapy; however, androgen therapy also becomes less effective after long-term application. When, then, does one decide to discontinue androgens and opt for bone marrow transplantation? The introduction of gene therapy, for patients in complementation group C and in the future for other Fanconi anemia patients, promises to brighten this picture. Using an adenovirus vector, Liu, Walsh, and their associates have recently successfully repopulated group C patients’ bone marrows with their own bone marrow cells with a corrected FACC gene introduced via an adenovirus vector into CD34+ bone marrow stem cells. At present, bone marrows of Fanconi anemia patients are examined yearly, unless a clone has been identified, in which case they are examined every 3 months. At each examination, chromosome analysis is carried out. The diepoxybutane test, already done to establish the diagnosis, should be extended to determine the complementation group. This is especially important now that it appears that it may soon become possible to offer group C patients the possibility of gene therapy. HLA typing, in anticipation of finding a bone marrow donor, is, of course, necessary.

771

In addition to these considerations, it is important to examine the status of other organ systems that may be affected by Fanconi anemia or its treatment. Ultrasound examination of the kidneys and urinary tract should be done. Periodic developmental assessments of children should be carried out, as well as formal hearing testing, and in all patients, blood chemistry studies, that include those for liver and kidney functions as well as iron status. Patients should probably avoid noxious fumes, such as in automobile exhaust, gasoline, formaldehyde, herbicides, pesticides, organic solvents, and tobacco smoke. Supplementation of the diet with antioxidants, such as vitamins C and E, is also probably a wise step, even though it is uncertain whether oxygen hypersensitivity is really significant in this disease (see above). Even if it is only a secondary problem, it may be well worth treating. In addition to treating patients with Fanconi anemia, it is important that patients at risk be screened as well. The DEB test should be done on parents and siblings of patients. Also, the cloning of the FAC gene and identification of specific mutations among Ashkenazi Jews have made it possible to do molecular screening of large numbers of individuals to determine carrier status. This is now being done. This has proved to be more difficult in the FAA gene because of great heterogeneity in the mutations found. FUTURE DIRECTIONS Since the most important problems that arise in Fanconi anemia patients are in their bone marrow cells, and since adenovirus vectored gene therapy of hematogenous stem cells has already been shown to be effective in some Fanconi anemia group C patients, it is possible that identification and cloning of other Fanconi anemia genes will lead to effective gene therapy for patients in other Fanconi anemia complementation groups. However, in some patients with Fanconi anemia, intragenic recombination, or some other event, has produced phenotypically normal clones of cells in their blood, yet they still have the disease. This indicates that gene therapy for this disease may be less effective than currently hoped. Alternatively, it is now possible to imagine that this disease may be “cured” within a decade. The term “cured” is in quotes because these patients will still have other problems, including increased tendency to develop carcinomas and whatever consequences they may have from their congenital structural abnormalities. As they come to live much longer, other problems, presently unanticipated, may also arise.

SELECTED REFERENCES Aboussekhra A, Biggerstaff M, Shiuji MKK, et al. Mammalian DNA nucleotide excision repair reconstituted in purified protein components. Cell 1995;80:859–868. Alter BP. Fanconi’s anemia and its variability. Br J Haematol 1993;85:9–19. Andrews AD, Halosy CL, Poh-Fitzpatrick MB. Abnormally low UVinduced unscheduled DNA synthesis in cells from a patient with hydroa vacciniforme. Photodermatology 1985;2:315–318. Auerbach AD. Fanconi’s anemia. Dermatol Clin 1985;13:41–49. Auerbach AD, Rogatko A, Schroeder-Kurth TM. International Fanconi Anemia registry: relation of clinical symptoms to diepoxybutane sensitivity. Blood 1989;73:391–397. Barlow C, Hirotsune JS, Paylor R, et al. ATM-deficient mice: a paradigm of ataxia telangiectasia. Cell 1996;86:159–171. Baskaran R, Wood LD, Whitaker LL, et al. Ataxia telangiectasia mutant protein activates c-Abl tyrosine kinase in response to ionizing radiation. Nature 1997;387:516–519. Bootsma D. The genetic defect in DNA repair deficiency syndromes. Eur J Cancer 1993;29A:1482–1488. Brookman KW, Lamerdin JE, Thelen MP, et al. ERCC4 (XPF) encodes a human nucleotide excision repair protein with eukaryotic recombination homologs. Mol Cell Biol 1996;16:6553–6562.

772

SECTION IX / DERMATOLOGY

Broughton BC, Thompson AF, Harcourt SA, et al. Molecular and cellular analysis of the DNA repair defect in a patient in xeroderma pigmentosum complementation group D who has the clinical features of xeroderma pigmentosum and Cockayne syndrome. Am J Hum Genet 1995;56:167–174. Brzoska PM, Chen H, Zhu Y, et al. The product of the ataxia-telangiectasia group D complementing gene, ATDC, interacts with a protein kinase C substrate and inhibitor. Proc Natl Acad Sci USA 1995; 92:7824–7828. Butturini A, Gale RP, Verlander PC, Adler-Brecher B, Gillio AP, Auerbach AD. Hematologic abnormalities in Fanconi Anemia: an International Fanconi Anemia Registry study. Blood 1994;84: 1650–1660. Cleaver JE, Kraemer KH. Xeroderma pigmentosum. In: McKusick VE, ed. Mendelian inheritance in man, 11th ed. Baltimore: Johns Hopkins University Press, 1994;1560:2275–2281. Cleaver JE, Volpe JP, Charles WC, Thomas GH. Prenatal diagnosis of xeroderma pigmentosum and Cockayne syndrome. Prenatal Diagn 1994;14:921–928. Cooper PK, Nouspikel T, Clarkson SG, Leadon SA. Defective transcription-coupled repair of oxidative base damage in Cockayne syndrome patients from XP group G. Science 1997;275:990–993. D’Andrea AD. Fanconi anaemia forges a novel pathway. Nature Genet 1996;14:240–242. de Vries A, van Oostrom CT, Hofhuis FM, et al. Increased susceptibility to ultraviolet-B and carcinogens of mice lacking the excision DNA repair gene XPA. Nature 1995;337:169–173. Ellis NA, Groden J, Ye T-Z, et al. The Bloom’s syndrome gene product is homologous to RecQ helicases. Cell 1995;83:655–666. Ellis NA, Lennon DJ, Proytcheva M, Alhadeff B, Henderson EE, German J. Somatic intragenic recombination within the mutated locus BLM can correct the high SCE phenotype of Bloom syndrome cells. Am J Hum Genet 1995;57:1019–1027. Elson A, Wang Y, Daugherty CJ. et al. Pleiotropic defects in ataxiatelangiectasia protein-deficient mice. Proc Natl Acad Sci USA 1996;93:13,084–13,089. Evens E, Bourre F, Qulliet X, et al. Different removal of ultraviolet photoproducts in genetically related xeroderma pigmentosum and trichodystrophy diseases. Cancer Res 1995;55:4325–4332. Fanconi anaemia/Breast cancer consortium (35 authors). Positional cloning of the Fanconi anaemia group A gene. Nature Genet 1996; 14:324–328. Fazaa B, Pierard-Franchimont C, Zghal M, Kamoun MR, Pierard GE. Nuclear morphometry in xeroderma pigmentosum-associated malignant melanomas. Am J Dermatopathol 1994;16:611–614. Fitzgerald MG, Bean JM, Hegde SR, et al. Heterozygous ATM mutations do not contribute to early onset of breast cancer. Nat Genet 1997; 15:307–310. Freidberg EC. DNA Repair, 2nd ed. New York, 1994; 317-365, 633–655. Fritz E, Elsea SH, Patel PI, Meyn MS. Overexpression of a truncated human topoisomerase III partially corrects multiple aspects of the ataxia-telangiectasia phenotype. Proc Natl Acad Sci USA 1997; 94:4538–4542. German J. Bloom’s syndrome. Dermatol Clin 1995;13:7–18. German J. Bloom’s syndrome. XX. The first 100 cancers. Cancer Genet Cytogenet 1997;93:100–106. Gillio AP, Verlander PC, Batish SC, Giampietro PF, Auerbach AD. Phenotypic consequences of mutations in the Fanconi anemia FAC gene: an International Fanconi Anemia Registry study. Blood 1997;90:105–110. Gotti RA. Ataxia-telangiectasia. Dermatol Clin 1995;13:1–6. Greenwell PW, Kronmal SL, Porter SE, Gassenhuber J, Obermaier B, Petes T. Tel 1: a gene involved in controlling telomere length in S. Cerevisiae, is homologous to the human ataxia-telangiectasia gene. Cell 1995;82:823–840. Hanawalt PC. Transcription-coupled repair and human disease. Science 1994;266:1957,1958. Hang B, Yeung AT, Lambert MW. A damage-recognition protein which binds to DNA containing interstrand cross-links is absent or defective in Fanconi Anemia complementation group A, cells. Nucleic Acids Res 1993;21:4187–4192.

Hari KL, Santerre A, Sekelsky JJ, McKim KS, Boyd JB, Hawley RS. The met-41 gene of D. melanogaster is a structural and functional homolog of the human ataxia-telangiectasia gene. Cell 1995;82:815–821. Hartley KO, Gell D, Smith GCM, et al. DNA-dependent protein kinase catalytic subunit: a relative of phosphatidylinositol 3-kinase and the ataxia-telangiectasia gene product. Cell 1995;82:849–856. Hoeijmakers JHJ. Nucleotide excision repair II: from yeast to mammals. Trends Genet 1993;9:211–217. Huo YK, Wang Z, Hong JH, et al. Radiosensitivity of ataxia-telangictasia, X-linked agammaglobulinemia and related syndromes using a modified colony survival assay. Cancer Res 1994;54:2544–2560. Jaspers NGJ, Bootsma D. Genetic heterogeneity in ataxia-telangiectasia: studies by cell fusion. Proc Natl Acad Sci USA 1982;79:2641–2646. Jaspers NGJ, Gatti RA, Boan C, et al. Genetic complementation analysis of ataxia-telangiectasia and Nijmegen breakage syndrome: a survey of 50 patients. Cytogenet Cell Genet 1988;49:259–269. Joenje H, Lo Ten Foe JR, Oostra AB, et al. Classification of Fanconi Anemia patients by complementation analysis: evidence for a fifth genetic subtype. Blood 1995;86:2156–2163. Jorgensen TJ, Russell PS, Mc Ray DA. Radioresistant DNA synthesis in SV-40-immortalized ataxia-telangiectasia fibroblasts. Radiat Res 1995;143:219–223. Kasten M. Ataxia-telangiectasia: broad implications for a rare disorder. N Engl J Med 1995;333:662,663. Kleijer WJ, van der Kraan M, Los FJ, Jaspers NG. Prenatal diagnosis of ataxia-telangiectasia and Nijmegen breakage syndrome by the assay of radioresistant DNA synthesis. Int J Radiat Biol 1994;66:S 167–S 174. Kohyama J, Shimohira M, Kondo S, et al. Motor disturbance during REM sleep in group A xeroderma pigmentosum. Acta Neurol Scand 1995;92:91–95. Kraemer KH, Lee MM, Andrews AD, Lambert WC. The role of sunlight and DNA repair in melanoma and nonmelanoma skin cancer. Arch Dermatol 1994;130:1018–1021. Kraemer KH, Lee MM, Scotto J. Xeroderma pigmentosum: cutaneous ocular and neurologic abnormalities in 830 published cases. Arch Dermatol 1987;123:241–250. Kraemer KH, Levy DD, Parris CN, et al. Xeroderma pigmentosum and related disorders: examining the linkage between defective DNA repair and cancer. Invest Dermatol 1994;103(5 Suppl):96S–101S. Kumaresan KR, Hang B, Lambert MW. Human endonucleolytic incision of DNA 3' and 5' to a site-directed psoralen monoadduct and interstrand cross-links. J Biol Chem 1995;270:30,709–30,716. Kupfer GM, Yamashita T, Naf D, Suliman A, Asano S, D’Andrea AD. The Fanconi anemia polypeptide, FAC, binds to the cyclin-dependent kinase, cdc2. Blood 1997;90:1047–1054. Lambert MW, Tsongalis GJ, Lambert WC, Hang B, Parrish DD. Defective DNA endonuclease activities in Fanconi’s Anemia cells, complementation Groups A and B [now reclassified as group D]. Mutat Res 1992;273:57–71. Lambert WC, Kuo H-R, Lambert MW. Xeroderma pigmentosum. Clinics Dermatol 1995;13:169–209. Lambert WC, Lambert MW. Co-recessive inheritance: a model for DNA repair and other surveillance genes in higher eukaryotes. Mutat Res 1992;273:179–192. Lambert WC, Lambert MW. DNA repair deficiency and skin cancer in xeroderma pigmentosum. In: Mukhtar H, ed. Skin cancer: Mechanisms and Human Relevance. Boca Raton: CRC, 1994; pp. 39–70. Lambert WC, Schwartz RA. The Muir-Torre syndrome. In: Demis DJ, ed. Clinical Dermatology. Philadelphia: JB Lippincott, in press. Lehmann AR. Nucleotide Excision repair and the link with transcription. Trends Biochem Science 1995;20:402–405. Lehman AR, Arlett CF, Broughton BC, et al. Trichothiodystrophy: a human DNA repair disorder with heterogeneity in the cellular response to ultraviolet light. Cancer Res 1988;48:6090–6096. Lehmann AR, Carr AM. The ataxia-telangiectasia gene: a link between checkpoint controls, neurodegeneration and cancer. Trends Genet 1995;11:375–377. Lehmann AR, Thompson AF, Harcourt SA, Stefanini M, Norris PG. Cockayne’s syndrome: correlation of clinical features with cellular

CHAPTER 81 / XERODERMA PIGMENTOSUM

sensitivity of RNA synthesis to ultraviolet radiation. J Med Genet 1993;30:679–682. Liu JM, Buchwald M, Walsh CE, Young NS. Fanconi anemia and novel strategies for therapy. Blood 1994;84:3995–4007. Lo Ten Foe JR, Kwee ML, Rooimans MA, et al. Somatic mosaicism in Fanconi anemia: Molecular bases and clinical significance. Eur J Hum Genet 1997;5:137–148. Lo Ten Foe JR, Rooimans MA, Bosnoyan-Collins L, et al. Expression cloning of a cDNA for the major Fanconi anaemia gene. FAA Nature Genet 1996;14:320–323. Matsumura Y, Sato M, Nishigori C, et al. High prevalence of mutations in the p53 gene in poorly differentiated squamous cell carcinomas in xeroderma pigmentosum. J Invest Dermatol 1995;105:399–401. Morrow DM, Tagle DA, Shilow Y, Collins FS, Heiter P. Tel 1, an S. cerevisiae homolog of the human gene, mutated in ataxia-telangiectasia, is functionally related to the yeast checkpoint gene MEC 1. Cell 1995;82:831–840. Nakane H, Takeuchi S, Yuba S, et al. High incidence of ultraviolet-B or chemical carcinogen induced skin tumors in mice lacking the xeroderma pigmentosum group A gene. Nature 1995;337:165–168. Nelson BR, Fader DJ, Gillard M, Baker SR, Johnson TM. The role of dermabrasion and chemical peels in the treatment of patients with xeroderma pigmentosum. J Amer Acad Dermatol 1995;32: 623–626. Nishigori C, Moriwaki S, Takebe H, Imamura S. Gene alterations and clinical characteristics of xeroderma pigmentosum group A patients in japan. Arch Dermatol 1995;130:191–197. Nouspikel T, Lalle P, Leadon SA, Cooper PK, Clarkson SG. A common mutational pattern in Cockayne syndrome patients from xeroderma pigmentosum group G: implications for a second XPG function. Proc Natl Acad Sci USA 1997;94:3116–3121. Pandita TK, Pathak S, Geord CR. Chromosome end associations, telomeres and telomerase activity in ataxia-telangiectasia cells. Cytogenet Cell Genet 1995;71:86–93. Parshad R, Tarone RE, Price EM, Sanford KK. Cytogenetic evidence for differences in DNA incision activity in xerodermal pigmentosum group A, C and D cells after X-irradiation during G2 phase. Mut Res 1993;294: 149–155. Paulovich AG, Hartwell LH. A checkpoint regulates the rate of progression through S phase in S. cerevisiae in response to DNA damage. Cell 1995;82:841–847. Randal J. ATM gene discovery may quiet cancer risk debate. J Natl Cancer Inst 1995;87:1350,1351. Reardon JT, Bessho T, Kung HC, Bolton PH, Sancar A. In vitro repair of oxidative DNA damage by human nucleotide excision repair system: Possible explanation for neurodegeneration in xeroderma pigmentosum patients. Proc Natl Acad Sci USA 1997;94:9463–9468. Robbins JH, Brumback RA, Moshell AN. Clinically asymptomatic xeroderma pigmentosum neurological disease in an adult: evidence for a neurodegeneration in later life caused by defective DNA repair. Eur Neurol 1993;33:188–190. Robert C, Sarasin A. Xeroderma pigmentosum: what may be expected from biological studies? Ann Dermatol Venereol 1995;121:434–439. Sancar A. Mechanisms of excision repair in mammalian cells. Science 1994;266:1954–1956. Sands AT, Abuin A, Sanchwz A, Conti CJ, Bradley A. High susceptibility to ultraviolet-induced carcinogenesis in mice lacking XPC. Nature 1995;337:162–165. Savitsky K, Bar-Shira A, Gilad S, et al. A single ataxia-telangiectasia gene with a product similar to PI-3 kinase. Science 1995;268: 1749–1753.

773

Seal G, Tallarida RJ, Sirover MA. Purification and properties of the uracil DNA glycosylase from Bloom’s syndrome. Biochem Biophys Acta 1991;1097:299–308. Stern JB, Peck GL, Haupt HM, Hollingsworth HC, Beckerman T. Malignant melanoma in xeroderma pigmentosum: search for a precursor lesion. J Am Acad Dermatol 1993;28:591–594. Strathdee CA, Duncan AMV, Buchwald M. Evidence for at least four Fanconi anemia genes Including FACC on chromosome 9. Nature Genet 1992;1:196–198. Strathdee CA, Garrish H, Shannon WR, Buchwald M. Cloning of cDNAs for Fanconi’s anemia by functional complementation. Nature 1992; 356:763–766. Stumm M, Gatti RA, Reis A, et al. The ataxia-telangiectasia variant genes 1 and 2 are distinct from the ataxia-telangiectasia gene on chromosome llq23.1. Am J Hum Genet 1995;57:960–962. Swift M. Ionizing radiation, breast cancer, and ataxia-telangiectasia. J Natl Cancer Inst 1994;86:1571,1572. Swift M, Chase CL, Morrell D. Cancer predisposition of ataxia-telangiectasia. N Engl J Med 1990;46:21–27. Swift M, Morrell D, Massey RB, Chase CL. Incidence of cancer in 161 families affected by ataxia-telangiectasia. N Engl J of Med 1991; 325:1833–1836. Swift M, Reitnauer PJ, Morrell D, et al. Breast and other cancers in families with ataxia-telangiectasia. N Engl J Med 1987;316:1289–1291. Tanaka K, Miura N, Satokata I, et al. Analysis of a human DNA excision repair gene involved in group A xeroderma pigmentosum and containing a zinc-finger domain. Nature 1990;348:73–76. Taylor AM, Boyd PJ, Mc Conville CM, Thacker S. Genetic and cellular features of ataxia-telangiectasia. Int J Rad Biol 1994;65:65–70. Taylor EM, Broughton BC, Botta E. et al. Xeroderma pigmentosum and trichothyodystrophy are associated with different mutations in the XPD (ERCC2) repair/transcription gene. Proc Natl Acad Sci USA 1997;94:8658–8663. Tu Y, Bates S, Pfeifer GP. Sequence-specific and domain-specific DNA repair in xeroderma pigmentosum and Cockayne syndrome cells. J Biol Chem 1997;272:20,747–20,755. Turner ML, Moshell AN, Corbett DW, et al. Clearing of melanoma in situ with intralesional interferon alfa in a patient with xeroderma pigmentosum. Arch Dermatol 1995;130:1491–1494. Walsh CE, Nienhuis AW, Samulski RJ, et al. Phenotypic correction of Fanconi anemia in human hematopoietic cells with a recombinant adeno-associated virus vector. J Clin Invest 1994;94:1440–1448. Weemaes CMR, Hustinx TWJ, Scheres JMIC, et al. A new chromosomal instability disorder: the Nijmegen breakage syndrome. Acta Paediatr Scand 1981;70:557. Wegner R-D, Metzger M, Hanefeld F, et al. A new chromosomal instability disorder confirmed by complementation studies. Clin Genet 1988;33:20–23. Weyl Ben Arush M, Rosenthal J, Dale J, et al. Ataxia-telangiectasia and lymphoma: an indication for individualized chemotherapy dosing: Report of treatment in a highly inbred Arab family. Pediatr Hematol Oncol 1995;12:163–169. Young BR, Painter RB. Radioresistant DNA synthesis and human genetic diseases. Hum Genet 1989;82:113–120. Youssoufian H. Localization of Fanconi anemia C protein to the cytoplasm of mammalian cells. Proc Natl Acad Sci USA 1994;91:7975–7979. Zakian VA. ATM-related genes: what do they tell us about functions of the human gene? Cell 1995;82:685–687. Zhang N, Chen P, Khanna KK, et al. Isolation of full-length ATM cDNA and correction of the ataxia-telangiectasia cellular phenotype. Proc Natl Acad Sci USA 1997;94:8021–8026.

CHAPTER 82 / GENE THERAPY

82

775

The Skin as a Vehicle for Gene Therapy SOOSAN GHAZIZADEH, TADEUSZ M. KOLODKA, AND LORNE B. TAICHMAN

INTRODUCTION: DNA AS A NEW FORM OF DRUG THERAPY Gene therapy involves the introduction and expression of new genetic material in a subset of somatic cells for therapeutic purposes and links the capacity for identifying and cloning new genes together with the ability to introduce and express these recombinant constructs in specific cell types. Gene therapy is a fast-moving, but young, enterprise. For the moment, much of the effort is concentrated on developing effective methods for gene transfer and gene expression. Clinical trials are underway covering a broad range of target cells and disorders, but in almost all instances the goal is to establish safety not efficacy. None of the current trials involves gene transfer to epidermal cells. The purpose of this chapter is to review the current state of knowledge for gene therapy, particularly as it applies to epidermal keratinocytes and dermal fibroblasts. Gene transfer to melanocytes and Langerhans cells is outside the scope of this review. Gene therapy has features that offer a new approach to the treatment of disease (Table 82-1). Unlike previous pharmacological agents, DNA has the potential to become a permanent part of the cell and the cell’s descendants by integrating into the chromosome and replicating along with the host chromosome. In this way, the newly introduced DNA is inherited by all descendant cells and therefore is a potential permanent therapy. If the gene or its expression is restricted to a specific subset of cells and if the gene product remains intracellular, unwanted toxic effects on nontargeted cells, commonly encountered with classical drug therapy, will not be seen. Gene therapy also enjoys another more subtle feature that simplifies development of new applications. For more classical drug-based therapies, each new drug has its own unique half-life, mode of action, distribution, and so on. In gene-based therapies, the same “drug” is used throughout, namely a polynucleotide. What differs is the sequence of nucleotides and the information encoded therein. This simplifies commercial manufacturing and implies that once the hurdles of safety and efficacy are overcome, new applications will be forthcoming at a very rapid rate.

FEATURES OF CUTANEOUS BIOLOGY RELEVANT TO GENE THERAPY RENEWAL In epidermis, loss of surface cells is balanced by mitotic replication of a progenitor population located primarily in From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

the basal compartment (Figure 82-1). This progenitor population consists of two types of replicating cells: a stem cell, the true progenitor; and amplifying cells, which undergo a limited number of rounds of replication followed by terminal differentiation and migration out of the basal compartment. As seen in Figure 82-1, the only cell that remains in the epidermis is the stem cell. Amplifying cells and terminally differentiated cells are eventually lost. If the goal of gene therapy is to provide a new gene product for prolonged periods, the target of gene insertion must be the stem cell. Putative stem cells have been localized both to the bulge region of the hair follicle by virtue of their prolonged cell cycle and to regions of interfollicular epidermis by virtue of intense staining with antibodies to β1 integrins. CULTURE MODELS Cutaneous biology is fortunate to have several culture models that permit in-depth analysis of gene transfer and gene expression under controlled conditions and allow for accurate preclinical testing prior to human clinical trials. These models range from the relatively simple submerged cultures in which keratinocyte growth is favored but differentiation is poor, to the more complex organotypic culture in which the in vivo architecture is approximated. In addition, these cultures have been used for autologous grafting in severe thermal injuries, suggesting that stem cells are present in culture. BARRIER A primary function of skin is to act as a barrier. Attempts, therefore, to introduce new genetic information into cells within the skin will encounter a permeability barrier and an interconnected, multilayered epidermis, making access to the progenitor population of keratinocytes problematic.

IN VIVO AND EX VIVO GENE THERAPY There are two basic approaches to gene therapy (Table 82-2). In vivo transfer involves insertion of the new gene directly into the tissue, whereas ex vivo transfer involves gene transfer to cells while in culture, followed by transplantation to the donor. Accessibility of epidermis makes the in vivo approach attractive, but the barrier property and the multilayered nature of the tissue prevent easy access to the basal or progenitor population. Ex vivo gene therapy is more complex, requiring additional cultivation and grafting steps. But the ex vivo approach does offer two advantages: first, gene transfer methods generally work more effectively in culture than in vivo; second, with ex vivo therapy, the performance of genetically altered keratinocytes can be assessed prior to transplantation. One drawback with ex vivo therapy is the requirement for grafting, especially for keratinocytes. Autologous grafting of cultured epithelial sheets has been established for burn

775

776

SECTION IX / DERMATOLOGY

Table 82-1 How Gene Therapy Differs from Classical Pharmacological Therapy DNA can become a permanent part of the cell and be inherited by its descendants. If gene transfer or gene expression is restricted to a particular cell type, unwanted side effects on other cell types are reduced or eliminated. In gene therapy, the “drug” is a polynucleotide, for all applications.

Figure 82-1 Keratinocyte renewal in a stratified squamous epithelium. The progenitor population of cells in epidermis consist of stem cells, the true progenitor cell, and amplifying cells that undergo a limited number of cycles of replication followed by terminal differentiation. Amplifying and terminal differentiating cells are eventually lost from the tissue by desquamation. Only the stem cell remains as a permanent resident of the epidermis. For long-term gene therapy, the stem cell must be the target of gene transfer.

therapy, but these techniques require full-thickness excision at the recipient graft site, a procedure that usually leads to scarring and contracture. Ex vivo gene therapy with fibroblasts may be more acceptable as autologous gene-altered fibroblasts can be implanted without having to excise normal tissue at the graft site.

GENE TRANSFER GENE TRANSFER BY TRANSFECTION Transfection involves gene transfer with purified DNA in a plasmid vector. There are several methods for transfection in vitro, each designed to facilitate DNA delivery to the interior of the target cell. For example, in calcium phosphate-mediated transfection, formation of a calcium phosphate precipitate entrains DNA and facilitates accessibility to the cell surface and endocytic uptake. In liposomeor lipid-mediated transfection, transport of the DNA through cell membranes is facilitated by membrane fusion. Electroporation works by creating temporary holes in the cell membrane through which DNA may enter. Transfer of DNA in vivo has been achieved by use of the “gene gun,” which accelerates microscopic gold

Table 82-2 Comparison of Ex Vivo and In Vivo Gene Therapy In vivo Direct transfer to tissue Preclinical testing not possible Gene transfer usually less efficient

Ex vivo Involves gene transfer to cells in culture followed by transplantation Allows preclinical testing of gene altered cells in advance of therapy Gene transfer techniques work optimally in culture

particles coated with DNA and forces the projectiles into the cell interior. Recently, intradermal injection of purified DNA into mouse or pig skin has been shown to result in expression of the encoded gene product in dermal fibroblasts as well as overlying epidermal keratinocytes. The mechanism of DNA uptake by any transfection method is unclear. Transfection methods are simple to use and can achieve transfer of relatively large molecules (>8 kbp). However, transfection in most cell types is inefficient, achieving gene transfer in a small percentage (usually 1–10%) of the target population. Transfected DNA does not ordinarily integrate into the chromosome of the target cell and, as a result, is diluted out by repeated cycles of cell replication. Expression, therefore, is transient, reaching peak levels in 2–3 days and abating after 5–10 days. However, expression can persist for much longer periods if the cell does not ordinarily cycle, such as mature muscle cells. GENE TRANSFER BY VIRAL TRANSDUCTION The use of viral vectors to transduce cells for gene therapy capitalizes on the capacity of viruses to insert and express their own genetic material in a cell with high efficiency. To harness the transducing powers of a viral vector, an infectious, but crippled virus is constructed in which viral genes required in trans for viral replication are replaced by a gene(s) of interest. The infectious property of the virus is maintained, but deletion of essential viral genes renders the vector replication incompetent, unable to complete its life cycle in the target cell. The end result is a system for efficiently introducing and expressing new genetic information in a cell without the risk of reinfection and viral spread. Assembly of viral vectors is carried out in packaging cell lines that are designed to constitutively express in trans the deleted viral genes. Three types of viruses have been approved for use in human gene therapy trials: adenovirus, retrovirus, and adeno-associated virus (see Table 82-3). Adenovirus vectors are noted for high titers and their ability to transduce nonreplicating cells, a feature that allows transduction of tissues with a low mitotic index. Adenoviral vectors have been used to transduce keratinocytes in vivo. As the vector does not have a mechanism for integration or self renewal, it does not persist and expression is lost within several days. Adenovirus vectors can be toxic at high multiplicities of infection and development of an immunological response limits duration of expression and prevents reinoculation. Attempts to reduce the immunological rejection by reducing the antigenicity of the vector or suppressing the immunoresponsiveness of the host have met with some success. Retrovirus vectors are currently the most widely used agents for gene transfer in clinical trials. These RNA vectors are derived from the Moloney murine leukemia virus and require cell replication for successful infection, as there is no mechanism to transport the viral replication complex into the nucleus. A reverse-tran-

777

CHAPTER 82 / GENE THERAPY

Table 82-3 Viral Vectors for Gene Therapy Virus structure Titer (CFU/mL) Insert size Target cell Integration In vivo transfer Safety concerns

Retrovirus

Adenovirus

AAV

RNA 106–107 7–8 kbp Replicating cells At random sites Not reported in skin Insertional mutagenesis

Linear, duplex DNA 1011–1012 ~10 kbp Replicating and nonreplicating cells No integration Very effective in variety of tissues Toxicity at high input

Linear, single-stranded DNA 103–106 5.2 kbp Replicating and nonreplicating cells Wild-type integrates at chromosome 19q13 Bronchial airway epithelial cells Without known pathology

scribed, double-stranded DNA copy of the viral genome integrates at high efficiency in the host-cell genome at random sites. Integration provides a mechanism for ensuring persistence of vector sequences in descendant cells. Retrovirus vectors are therefore quite suitable for transducing stem cells, if these cells can be induced to replicate. Recent advances in retrovirus vector design include: (1) high-titer stocks by encapsidating the viral genome in the capsid protein of vesicular stomatis virus, thereby enabling concentration to titers ≥109 colony forming units per mL (CFU/mL) and expanding its host range; (2) lentivirus vectors capable of transducing nondividing cells; (3) receptor-binding sequences allowing targeting to a specific cell type; and (4) vectors with inactivated 5' promoter sequences that may allow unfettered expression of an internal promoter. A concern with retrovectors is the potential for mutagenesis at the site of integration. However, the likelihood of disrupting a vital gene is remote and no untoward effects of retroviral transduction have been reported to date in human trials. Cultured keratinocytes are highly transducible with recombinant retroviruses. At titers above 5 × 106 CFU/mL, all clonogenic keratinocytes in the culture are transduced. In culture, putative stem cells are successfully transduced with no apparent loss of growth potential. No one has yet succeeded in using retrovirus vectors for direct gene transfer to epidermis or dermis, either for short-term or prolonged expression. Wild-type adenoassociated virus has several features that make it an attractive vector system including the lack of any known pathology associated with infection, its ability to infect nonreplicating cells, and its potential for integrating at a specific site in the chromosome. However, the size of the transgene it can transfer is limited and site-specific integration is likely to require the presence of the viral rep gene, further limiting the size of the transgene. Papillomaviruses induce benign warts and condyloma in stratified squamous epithelia. Papillomaviruses have not received much attention as a vector but two factors make these viruses particularly attractive for keratinocyte gene therapy. First, they have a natural tropism for keratinocytes of stratified squamous epithelium; second, their circular DNA genome replicates as a stable, multicopy episome. Three viral genes are required for episomal replication: an origin of replication, the E1 and E2 genes. The E5, E6, and E7 genes, which encode products with oncogenic potential, can be removed without loss of episomal replication. No packaging cell line for papillomavirus is available. However, infectious, but empty viral capsids have been produced by expressing viralcapsid genes at high levels. It may be possible to develop this system further to produce infectious, replication-incompetent papillomavirus.

Table 82-4 Potential Applications for Cutaneous Gene Transfer Treatment of a cutaneous disorder (inherited or acquired) Vaccination Systemic delivery of gene product Creation of an enzyme reservoir for processing of circulating metabolites

EXPRESSION OF THE THERAPEUTIC GENE For gene therapy to succeed, the therapeutic gene must not only be inserted in the target cell it must also be expressed at a level sufficient to achieve a therapeutic effect. Some gene products may need to be regulated to achieve a suitable response. Promoters from Moloney murine leukemia virus, SV40, and cytomegalovirus have been widely used and provide high level activity in a variety of cell types including keratinocytes. For reasons that remain unclear, activity from these viral promoters remains stable while the host cell is maintained in culture, but drops precipitously when the same cells are transplanted to an animal host. The mechanism behind this tissue-induced inactivation is unknown but may be related to methylation of cytosine residues in the viral promoter. Experiments with fibroblasts, myoblasts, and hepatocytes suggest that tissue-specific promoters normally active in the particular cell type are not subject to this inactivation. The use of keratinocyte-specific promoters to overcome this inactivation has not been reported.

GENE THERAPY FOR INHERITED AND ACQUIRED CUTANEOUS DISORDERS INHERITED DISORDERS An inherited cutaneous disorder in which the mutant gene has been identified and the normal allele cloned is a candidate for gene therapy (Table 82-4). However, there is a key difference between dominant and recessive disorders in terms of their suitability for gene therapy (Fig. 82-2). In a dominant disorder, the mutant phenotype is induced by a single copy of the mutant allele. The presence of the normal allele in the same cell has no sparing effect. Expressing an additional copy of the normal allele is therefore not likely to alter this situation. For example, epidermolysis bullosa simplex (see Chapter 73) and epidermolytic hyperkeratosis (see Chapter 74) are two dominantly inherited skin diseases in which mutation in keratin structural genes cause collapse of the intermediate filament network and cytolysis. Insertion of an additional normal allele is not likely to offset the disruptive effects of the mutant protein on the filament network. A more effective therapy is likely to require selective inactivation of the mutant gene by homologous recombination or by an antisense or ribozyme strategy.

778

SECTION IX / DERMATOLOGY

Table 82-5 Examples of Systemic Uptake of a Gene Product Released by Cells Located in Skin Cell Keratinocyte

Fibroblast

Figure 82-2 Gene therapy for dominant and recessive disorders. In a dominant disorder, a single mutation in one of the two alleles causes the mutant phenotype, whereas, in a recessive disorder, both alleles must be mutated for the mutant phenotype to be present. In carriers of a recessive disorder, the presence of one normal allele spares the cell from the mutant phenotype. Insertion and expression of a normal allele in a recessive disorder is therefore likely to result in reversion to the normal phenotype, whereas such a strategy is not likely to succeed for a dominant disorder.

Recently, a new methodology has been described that may allow targeted correction of a known dominant point mutation. A chimeric DNA/RNA oligonucleotide homologous to a specific host sequence in all but one nucleotide is introduced into a cell, and by a mechanism possibly involving DNA mismatch repair, the host sequence is altered to correspond with that of the oligonucleotide. In this way, the point mutation in the hemoglobin β allele was corrected. The use of chimeric DNA/RNA oligonucleotides has not been described in cutaneous cells. In a recessive disorder, the mutant phenotype appears when both alleles in a cell are mutant (Fig. 82-2). Inserting a normal allele into cells from such an individual is likely to recreate conditions resembling the carrier state in which there is a single copy of the mutant and the normal allele. Lamellar ichthyosis (LI) is a case in point. LI is a recessive disorder associated with mutations in keratinocyte transglutaminase I (TGaseI) and results in abnormal epidermal differentiation and defective barrier function. By inserting a copy of the normal TGaseI allele into LI-keratinocytes, a more normal phenotype is evident when these gene-altered cells are grafted to athymic mice. Among the changes seen in the genetically modified epithelium were reduced hyperkeratosis, a more normal distribution of filaggrin and a reduction in transepidermal water loss. Although gene therapy is not possible over the entire epidermis, it may be possible to treat selective areas of skin. If the gene-corrected cells possess a selective growth or attachment advantage, they may be capable of replacing the defective epithelium. ACQUIRED DISORDERS It is not clear at this time what acquired cutaneous disorders might be approached with gene therapy. Several factors are responsible for this lack of clarity. First, because of the untried nature of keratinocyte gene therapy, clinical trials are likely to be approved only for cutaneous disor-

Protein α1-antitrypsin apo E Factor IX Growth hormone IL-6 Growth hormone Erythropoietin Factor VIII Transferrin

ders in which all current therapies have proven to be unsatisfactory. Second, because of limitations on the size of the transferred gene, only a single gene can be effectively transferred. Therefore, the disease in question must be one that can be treated with a single gene product. Third, with current methods of in vivo gene transfer, expression is transient, thereby limiting applications to conditions such as wound healing.

CUTANEOUS GENE TRANSFER FOR SYSTEMIC DISORDERS Expressing new genetic information in skin may be used to induce effects that have systemic implications. Three types of systemic applications are described: VACCINATION Uptake and expression of naked plasmid DNA by epidermal and dermal cells can be followed by development of long-lasting humoral and cellular immune responses to the encoded protein, suggesting that intradermal gene vaccination may constitute an alternative to traditional immunization. For example, when mice receive an intradermal injection of a plasmid encoding the influenza nucleoprotein, antinucleoprotein antibodies and cytotoxic T lymphocytes developed and the mice gain resistance to challenge with a heterologous strain of influenza virus. The mechanism of immunization is not well-understood. Immunogenicity of plasmid DNA is not based on the level of protein expression, but rather the presence of short immunostimulatory DNA sequences in the plasmid backbone that elicits production of proinflammatory cytokines by keratinocytes and antigen-presenting cells. The induction of immune responses may also involve transfection of skin-derived dendritic cells that localize to the draining lymph node. SYSTEMIC DELIVERY Systemic delivery from epidermis was first demonstrated in the case of apolipoprotein E (apo E), a 34-kDa protein naturally secreted by keratinocytes. Insertion of a recombinant tagged apo E gene under the control of a viral promoter into keratinocytes with subsequent grafting to athymic mice resulted in the presence of both the endogenous and the tagged apo E in the animal’s bloodstream. Systemic delivery of apo E may be useful for the treatment of familial hypercholesterolemia III. Other gene products that have been induced in keratinocytes and delivered systemically include human growth hormone, factor IX, and α1 antitrypsin (Table 82-5). METABOLIC PROCESSING OF A CIRCULATING SUBSTRATE In inherited metabolic disorders with a nonfunctional enzyme and toxic accumulation of its substrate, one therapy has

CHAPTER 82 / GENE THERAPY

been to replace the missing enzyme by injection. Gene-transfer techniques may allow creation of a permanent intracellular enzyme reservoir to metabolize circulating substrate. A case in point is severe-combined immunodeficiency caused by mutation in adenosine deaminase (ADA). ADA normally deaminates deoxyadenosine (dAdo) to produce deoxyinosine. In ADA deficiency, dAdo accumulates in the bloodstream and is toxic to T- and B-cell development. Classical therapy for this disease has been to furnish ADA enzyme either by injection or by transfusion of red-blood cells. Recently, human gene therapy trials have successfully introduced the ADA gene into peripheral lymphocytes and hematopoietic stem cells. Dermal fibroblasts transduced with an ADA-retrovirus vector have been shown to express high levels of the enzyme and deaminate significant quantities of the dAdo substrate and the suggestion has been made to employ these cells to create a cutaneous enzyme reservoir.

CONCLUSION If gene therapy is an infant science, cutaneous gene therapy is at an earlier, embryonic stage. To date no clinical trials have been initiated or approved using skin as the target organ. The first clinical trials are likely to entail genetic vaccination as this application does not require sustained expression of the gene product and methods are in place for gene transfer and expression. For applications that require long-term expression of a gene product, two hurdles must be overcome: If ex vivo gene transfer is to be attempted, then methods for autologous transplantation of the genetically modified cells must be developed that are not destructive to tissues at the recipient site; if in vivo gene transfer is to be used, then vectors must be available that efficiently integrate into the host genome or undergo autonomous episomal replication. Cutaneous gene therapy will undoubtedly bring a new dimension to the treatment of cutaneous as well as systemic disorders.

779

SELECTED REFERENCES Blaese RM, Culver KW, Miller DA, Carter CS, Fleisher T, Clerici M. T lymphocyte-directed gene therapy for ADA-SCID: initial trial results after 4 years. Science 1995;270:475–480. Choate KA, Medalie DA, Morgan JR, Khavari PA. Corrective gene transfer in the human skin disorder lamellar ichthyosis. Nature 1996;2:1263–1267. Cole-Strauss A, Yoon K, Xiang Y, et al. Correction of the mutation responsible for sickle cell anemia by an RNA-DNA oligonucleotide. Science 1996;273:1386–1389. Fenjves ES. Approaches to gene transfer in keratinocytes. J Invest Dermatol 1994;103:70S–75S. Fenjves ES, Smith J, Zaradic S, Taichman LB. Systemic delivery of secreted protein by grafts of epidermal keratinocytes: prospects for keratinocyte gene therapy. Hum Gene Ther 1994;5:1241–1248. Freiberg RA, Choate KA, Deng H, Alperin ES, Shapiro LJ, Khavari PA. A model of corrective gene transfer in X-linked ichthyosis. Hum Mol Genet 1997;6:927–933. Garlick JA, Katz AB, Fenjves ES, Taichman LB. Retrovirus-mediated transduction of cultured epidermal keratinocytes. J Invest Dermatol 1991;97:824–829. Greenhalgh DA, Rothnagel JA, Roop DR. Epidermis: an attractive target tissue for gene therapy. J Invest Dermatol 1994;103:63S–69S. Khavari PA, Krueger GG. Cutaneous gene therapy. Dermatol Clin 1997;15:27–35. Krueger GG, Morgan JR, Jorgensen CM, et al. Genetically modified skin to treat disease: potential and limitations. J Invest Dermatol 1994;103: 76S–84S. Mathor MB, Ferrari G, Dellambra E, et al. Clonal analysis of stably transduced human epidermal stem cells in culture. Proc Natl Acad Sci USA 1996;93:10,371–10,376. Mulligan RC. The basic science of gene therapy. Science 1993;260:926–932. Muzyczka N. Use of adeno-associated virus as a general transduction vector for mammalian cells. Curr Topics Microbiol Immunol 1992;158:98–129. Raz E, Carson DA, Parker SE, et al. Intradermal gene immunization: the possible role of DNA uptake in the induction of cellular immunity to viruses. Proc Natl Acad Sci USA 1994;91:9519–9523. Vogel JC, Walker PS, Hengge UR. Gene therapy for skin diseases. Adv Dermatol 1996;11:383–398. Wilson JM. Adenoviruses as gene-delivery vehicles. N Engl J Med 1996;334:1185–1187.

CHAPTER 83 / ACQUIRED CUTANEOUS DISEASES

ACQUIRED DISEASES OF CUTANEOUS TISSUES

781

782

SECTION IX / DERMATOLOGY

CHAPTER 83 / ACQUIRED CUTANEOUS DISEASES

83

783

Acquired Diseases of Cutaneous Tissues Introduction THOMAS S. KUPPER

INTRODUCTION Over 2000 skin diseases have been described to date, and the majority of these are not genodermatoses, but rather are acquired over the course of the individual’s lifetime. Many of these are likely to have a genetic component that is presently undefined. In addition, the skin can be an important indicator of systemic disease, including metabolic, oncologic, and infectious disorders. A careful and encyclopedic description of all acquired skin diseases, and skin signs of systemic disease, is well beyond the scope of this section and is further precluded by considerations of space. The reader is referred to several excellent comprehensive textbooks of Dermatology (listed below). In compiling this section, we elected not to simply list a large number of skin diseases (for which the etiopathogenesis is often completely obscure); rather, we decided to focus on several selected important skin diseases for which the molecular genetic basis is currently being rigorously explored. Cutaneous oncology is a particularly important area in modern dermatology; much of what we know about the mechanism of carcinogenesis in animal models and humans stems from the study of murine models of cutaneous carcinogenesis. The molecular genetics of the most common skin cancers are discussed, including squamouscell carcinoma, basal-cell carcinoma, and melanoma. These cancers all share, at least to some extent, an etiopathogenesis based on ultraviolet B (UVB)-radiation-mediated DNA damage. Genodermatoses associated with excessive sensitivity to UVB irradiation with respect to cutaneous carcinogenesis are discussed in Chapter 81. Lesscommon skin cancers, including cutaneous T-cell lymphoma (CTCL), are not discussed in detail here. This interesting non-Hodgkins lymphoma, which, in its most common form (mycosis fungoides), represents a malignancy of mature (CD7–), memory (CD45RO), skin homing (CLA+), or helper (CD4+) T cells, often presents initially as an inflammatory skin disease. A comprehensive review of these lymphomas was recently published (see below). Merkel-cell carcinomas and fibrosarcomas are beyond the scope of this section. Psoriasis and atopic dermatitis (Chapters 86 and 87) are, collectively, the most common inflammatory skin diseases that appear also to have a genetic basis, and significant evidence has emerged implicating T cells and the immune system in their etiopathogenesis. Moreover, the type of T cell involved (CLA+, CD45RO, or CD3+ T cells) in the initiation of lesions of psoriasis and atopic dermatitis shares many cell-surface markers with the transformed From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

skin-homing T cell of CTCL (see above). Whereas T cells, antigen-presenting cells, and cytokines are important in the pathogenesis of psoriasis and atopic dermatitis, there is evidence that a more complex polygenic pattern of inheritance underlies these diseases. Although certain major histocompatibility complex (MHC) associations appear to be solid, other of these candidate genes are likely to not be directly related to the immune system. However, it is important to recall that immunosuppression mediated by drugs such as cyclosporin and FK506 can clinically suppress these diseases. Bullous diseases, discussed in Chapters 88 and 89, often involve an immune-mediated assault on structural elements of the skin. The relevant antigen in pemphigus vulgaris was recently found to be a cadherin-like adhesion molecule. The gene defect in a form of junctional epidermolysis bullosa involves the deficiency of a molecule first described in the context of the blistering disease bullous pemphigoid (the BPAG1) as an antigen. Finally, Chapters 90 and 91 describe two classes of disease long considered to have an autoimmune etiology: lupus erythematosis and scleroderma. It has been extremely difficult to dissect the pathogenesis of these complex diseases, and physicians have had to attempt to classify and treat these disorders in the absence of a clearly defined etiologic mechanism. These chapters review the current state of knowledge about these diseases, and suggest future directions for clinical investigation. It is important for internists and other physicians to appreciate that there are many important skin signs of systemic diseases, and it is unreasonable to expect nondermatologists to be conversant in these. In addition, as our pharmacologic armamentarium grows and new powerful systemic therapies emerge, the already large number of drug reactions involving skin will grow. Excellent texts exist that include photographs and descriptions of these disorders, and the reader more than casually interested is encouraged to seek these out. Unfortunately, the molecular bases for most of these skin manifestations of systemic disease or drug reactions remain completely unknown. As our understanding of the molecular basis of skin disease grows, it is anticipated that this state of affairs will change. Space precludes an exhaustive discussion of other acquired skin diseases in this section.

SELECTED REFERENCES Champion RH, Burton JL, Ebling FJG. Rook’s Textbook of Dermatology, 5th ed. London: Blackwell Scientific, 1992. Fitzpatrick TB, Eisen AE, Wolff K, Freedberg IM, Austen, KF. Dermatology in General Medicine, 4th ed. New York: McGraw Hill, 1993. Frank MM, Austen KF, Clamen HN, Unanue ER. Samter’s Immunologic Diseases, 5th ed. Boston: Little Brown, 1994. Koh H, Foss. Cutaneous T Cell Lymphoma. St. Louis: Mosby-Year Book, 1996.

783

CHAPTER 84 / BASAL- AND SQUAMOUS-CELL CARCINOMAS

84

785

Basal- and Squamous-Cell Carcinomas PAUL NGHIEM AND THOMAS S. KUPPER

EPIDEMIOLOGY In 1996, nearly one million skin cancers were diagnosed in the United States alone. This number approaches the sum of all other types of cancer combined. Malignant transformation can occur in each of the three major call types of the epidermis: the basal cells, the squamous cells, or the pigment-producing melanocytes. Cancer of the melanocytes, or malignant melanoma, is the most lethal and, fortunately, least common type of skin cancer. Of approximately 38,000 cases in 1996, there were approximately 7000 deaths. (Melanoma will be covered in Chapter 85.) The nonmelanoma skin cancers, basal-cell carcinoma (BCC) and squamouscell carcinoma (SCC), together comprise more than 95% of skin cancers, but cause few deaths as they rarely metastasize. Early diagnosis and excision is highly effective in treating all three forms of skin cancer, including melanoma. It is now well-established that ultraviolet (UV) radiation, especially in the UVB range from approximately 280–320 nanometers, is responsible for most skin cancers. However, the complex pathway from sun exposure encountered in youth to skin cancers developing years later is only beginning to be unraveled.

RISK FACTORS FOR NONMELANOMA SKIN CANCER ULTRAVIOLET LIGHT Australia has served as an inadvertent experiment on the effects of sunlight on minimally pigmented skin following the colonization of this sunny land with Caucasians from the British Isles. Whites in Australia have the highest rates of skin cancer of any people in the world, much higher than their British relatives who remained in the north or the Australian Aborigines who are protected by darker skin. Similar trends exist in the United States where southerners have higher rates of all three types of skin cancer than their less-exposed northern relatives. At the molecular level, sunlight has also left telltale signs of its association with skin cancer. One of several types of mutations that can be induced by UV radiation involves adjacent cytosines on the same strand of DNA and is essentially unique to this mutagen. Specifically, light induces the formation of aberrant chemical bonds between adjacent cytosines. These so-called cyclobutanepyrimidine dimers are then misread by DNA polymerase, causing them to be paired with adenine instead of guanine. When the next strand is copied, the adenines are paired with thymines instead of From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

the original cytosines resulting in a so-called C-to-T mutation in the original strand of DNA. C-to-T mutations occurring in adjacent pyrimidines are highly suggestive of UV radiation as no other mutagen routinely causes this change. A high percentage of DNA mutations detected in skin cancer involve such signature mutations. Interestingly, the vast majority of these mutations occur in the nontranscribed strand of DNA, as the transcriptional machinery has significant DNA repair capacity and typically corrects mutations in the transcribed strand prior to replication. INHERITED CANCER SYNDROMES Basal-cell nevus syndrome (BCNS) and xeroderma pigmentosum are examples of familial skin cancer syndromes. Both of these diseases have contributed to our understanding of the etiology of sporadic skin cancers and are covered in detail in separate chapters in this text. In basal-cell nevus syndrome, hundreds or thousands of basal-cell carcinomas (BCCs) develop in sun-exposed areas. In this autosomal dominant disorder, one copy of the human homolog of drosophila patched, a tumor-suppressor gene, is mutated, leading to multiple developmental abnormalities in addition to these tumors. The molecular pathway involving this tumor suppressor will be discussed below as it is relevant in sporadic BCCs as well. In xeroderma pigmentosum, defective DNA repair enzymes allow rapid accumulation of UV-induced mutations. Xeroderma pigmentosum patients suffer from profound sun damage after trivial UV-light exposures and develop all three types of skin cancer within the first decades of life. IMMUNE SUPPRESSION Through an important but poorly understood mechanism, the immune system plays a significant role in controlling skin cancers. This hypothesis is derived from the observation that organ transplant patients on prolonged immunosuppression with agents such as cyclosporine, prednisone, and azathioprine have markedly increased risks of developing basaland especially squamous-cell carcinomas. By 20 years after kidney transplantation and beginning immunosuppression, approximately 40% of patients have developed SCCs and/or BCCs as compared with 6% for the age-matched normal population. The rates are even higher following heart transplantation, as these patients require more intensive immunosuppression and undergo transplantation later in life when precancers are presumably more advanced. In addition to elevated rates of occurrence, SCCs especially tend to be highly aggressive when they develop in the setting of immune compromise, with a much higher rate of metastasis and lack of response to therapy, especially if immunosuppression is continued. Whereas these epidemiologic studies support the concept of tumor surveillance, we do not know the critical mecha-

785

786

SECTION IX / DERMATOLOGY

Figure 84-1 Proposed model of Sonic Hedgehog (Shh) signaling pathway in vertebrates, relevant in development and BCC. Shh is a diffusible signaling molecule that binds the 12 transmembrane-spanning receptor patched (Ptc), which is mutated in both BCNS and in many sporadic BCCs. The binding of Shh to wild-type Ptc causes disinhibition of the pathway, allowing the seven transmembrane-spanning receptor Smoothened (Smo) to transmit an intracellular signal affecting cellular differentiation. The critical nature of this pathway in vertebrate development is demonstrated by transgenic mice deficient in either Shh or Smo which have a phenotype of multiple skeletal and neurological defects and die before birth. Mutation or loss of Ptc (and its normal inhibitory action on Smo) causes constitutive signaling by this pathway, and leads to the development of BCCs. (Adapted from Stone DM, et al. Nature 1996;384:133.)

nisms by which the immune system recognizes most nascent skin cancers yet fails to detect others (the ones that progress). Are the products of mutant cancer-associated genes expressed as novel antigens? Does immunosuppression induced by UV light (which is potent enough to block immune-related diseases such as psoriasis for example) play a clinically relevant role in allowing tumor progression? Do tumors actively suppress the immune system by engaging the lymphocyte Fas apoptotic pathway or elaborating diffusible factors? Answers to these questions have biological and potential therapeutic implications. HUMAN PAPILLOMA VIRUS The E6 protein product of human papilloma virus (HPV) types 16 and 18 has been shown to induce ubiquitin-mediated degradation of the p53 tumor-suppressor protein. This mechanism is believed to be central in the strong association of these particular subtypes of HPV with SCC. Genital warts caused by HPV, the most common sexually transmitted disease, is also relevant in cervical carcinoma (HPV types 6, 11, 16, and 18). SCCs are also associated with some cases of immunocompromised patients with disseminated HPV lesions.

MOLECULAR ETIOLOGY OF NONMELANOMA SKIN CANCER BASAL-CELL CARCINOMA Two lines of investigation, genetic-disease linkage analysis and drosophila developmental biology, converged in 1996 to shed some light on the molecular basis of BCCs by identifying the affected gene in the BCNS. The developmental role of this gene product is suggested by the constellation of findings in this syndrome: jaw cysts, rib, vertebral, and shoulder abnormalities; characteristic pitting of the palms; and numerous BCCs. The affected gene, Patched, is 67% identical between human and fly at the nucleotide level and functions in the hedgehog (drosophila) or sonic hedgehog (vertebrate) signaling pathway. In vertebrates, sonic hedgehog (Shh) is involved in the development of motor, serotonergic, dopaminergic, and forebrain neurons, vertebrae, hindgut mesoderm, and distal limb structures. As depicted in Fig. 84-1, this pathway involves the diffusible signal protein, Shh, as well as Patched (Ptc) and Smoothened (Smo), both of which are transmembrane receptors. The normal function of Ptc appears to be the inhibition of intracellular signaling by Smo except in the presence of Shh. Loss of signaling in this pathway

through the targeted disruption of either the diffusible activator protein, Shh, or the intracellular signaling molecule, Smo, leads to numerous developmental defects and embryonic lethality in transgenic mice. Conversely, loss of Ptc is associated with constitutive signaling through this pathway and the development of BCCs. In BCNS, patients have lost a single copy of the Patched gene, which is encoded on chromosome 9. In a classic pattern analogous to that described for retinoblastoma and other tumor suppressor genes, this leads to a disease that is autosomal dominant at the level of the individual but autosomal recessive at the cellular level. The loss in all cells of the body of one functional copy of the tumor suppressor gene leaves only a single copy as a target of mutagens or possible errors in DNA replication. In a normal individual, a sporadic mutation of a given tumor suppressor in a somatic cell would have a minimal effect as the second copy would remain functional, but in an affected individual, a single sporadic mutation would leave the cell without function of that tumor suppressor, leading to cancer in that cell’s progeny. Although it appears that, in BCNS, loss of Patched function underlies development of BCCs, what is the mechanism in the much more common sporadic BCCs? Several early studies suggest that Patched is also mutated in a significant number of sporadic BCCs. In these cases, typically one copy is deleted by allelic loss of the 9q22 chromosomal region which encodes Patched, whereas the other is the target of an inactivating mutation in its sequence often as a result of a characteristic CC-to-TT mutation from UV light. Studies of Patched mutations in sporadic BCCs are difficult because of its large size (4330 bp) and the fact that virtually all tumors characterized thus far have distinct mutations. Recent data suggest that, although many mutations in sporadic BCCs occur in the Patched gene, mutation of other target proteins can also induce these tumors. Specifically, mutations have been found in the Smoothened gene in several sporadic BCCs as well. Indeed, when an activating mutant version of the Smoothened gene was overexpressed in the skin of transgenic mice, they developed BCC-like lesions. This mutation thus activated the Smoothened protein to generate its signal even in the absence of Sonic hedgehog. The nature of the signal from Smoothened has also recently been elucidated. A critical target of Smoothened signaling is a zinc finger transcription factor called Gli1. As one would predict for a

CHAPTER 84 / BASAL- AND SQUAMOUS-CELL CARCINOMAS

target of this pathway, overexpression of Shh and loss of Patched both cause an upregulation of Gli1. Moreover, Gli1 is upregulated in sporadic BCCs (46 of 47 cases) but not in SCCs (0 of 10 cases). Thus, this pathway can be disrupted at several sites, all of which appear to cause an upregulation of the Gli1 transcription factor, which mediates the effects of Sonic hedgehog. SQUAMOUS-CELL CARCINOMA The 1980s brought exciting progress in understanding the molecular basis of carcinomas arising in a classic model of multistage carcinogenesis in mouse skin. The experimental design involves shaving an area of mouse skin, applying a single dose of an “initiator” chemical such as DMBA (dimethylbenzanthracene), followed by twice-weekly applications of a “promoter” such as TPA (tetradecanoyl-phorbolacetate). This treatment leads to benign skin tumors called papillomas, a small percentage of which eventually become invasive carcinomas. Elegant experiments led to the discovery that the initiator causes mutations in the ras gene most often in the 12th, 13th, or 61st codons. When the normal cellular form of the Harvey ras (H-ras) gene is mutated in such a manner it loses its ability to act as a GTPase-activating protein (hence becoming unable to turn off G protein-mediated signaling) and is referred to as an activating H-ras mutant. Whereas activated H-ras is adequate to transform the immortalized mouse NIH-3T3 cell line, it is incapable of transforming primary cell lines. Presumably, H-ras is complementing pre-existing mutations in the immortalized 3T3-cell line not present in normal cells. For example, the combination of H-ras with an oncogenic version of fos is capable of transforming normal epidermal cells. In the mouse multistage paradigm, H-ras mutation by an initiator must be followed by weeks of promotion, involving (among other things) repeated activation of protein kinase C, known to alter the differentiation state of cells. Unfortunately, the critical targets of this second stage of mutagenesis, promotion, remain unknown in this model of chemically induced SCCs. A further limitation in this model system is that the vast majority of human SCCs are induced by UV light rather than chemicals. What is the relevance of ras mutations in typical human SCCs arising in sun-exposed areas? Ras mutations are present in a minority (13% in one study) of human SCCs, and evidence exists that this mutation favors an increased growth rate and diminished ability of keratinocytes to differentiate, both features that may be relevant at least to the early stages of tumorigenesis. On the other hand, a larger proportion (30%) of benign and self-regressing skin tumors called keratoacanthomas displayed ras mutations, arguing that activating ras mutations are neither capable of causing a malignant transformation nor adequate in maintaining growth of benign tumors. More recently, mutations have been discovered in the p53 gene which are likely to be relevant as they are present in a large fraction (between 60 and 90%) of human SCCs and bear the hallmark of UV-induced damage, the C-to T substitution. p53 is a tumor suppressor that functions in at least two important ways: (1) It halts the cell cycle following DNA damage allowing repair to occur prior to DNA synthesis; and (2) it induces programmed cell death in cells in which DNA damage is too extensive for reliable repair. The loss of p53 function would thus favor tumorigenesis as DNA mutations could accumulate in cells much more rapidly. Indeed, at least one copy of p53 may be lost at a surprisingly early stage in the UV induction of cancer as p53 mutations are detectable in more than 70% of “normal” skin samples from sun-exposed sites on patients who had developed skin cancers elsewhere. No p53 muta-

787

tions were detected in skin from sun-protected sites on control individuals without skin cancer. High rates of p53 mutations were also noted in the common premalignant lesions known as actinic keratoses. p53 mutant cells typically display nuclear accumulation of p53-immunoreactivity that represents nonfunctional p53 protein. Clones of cells displaying such immunoreactivity were present in sun-exposed skin at 10 times the rate of sun-protected skin. Interestingly, sun-exposed skin was five times more likely to have large clones than was sun-protected skin, arguing that sunlight was acting both to induce the mutations and to promote their subsequent growth. Whereas p53 mutations are likely to be relevant, there is ample evidence to suggest that they are not sufficient in keratinocyte carcinogenesis. In p53-knockout mice challenged with chemical carcinogens, both initiation and promotion steps were required, although carcinomas did develop more rapidly than in wild-type mice. Importantly, mutant ras, even in the setting of homozygous p53 loss, was not adequate to induce carcinoma in this model, which required promotion with TPA, an activator of protein kinase C. The relevant targets of TPA in this system remain unknown. Similar arguments about the insufficiency of p53 mutations in the induction of SCCs are derived from humans with Li-Fraumeni syndrome, in which one copy of p53 is lost in the germline. Patients with Li-Fraumeni syndrome suffer from extremely high rates of leukemia, breast carcinoma, and soft-tissue sarcoma, yet not of sun-induced skin cancers.

SUMMARY A series of molecular events is required to generate a true cutaneous carcinoma and, within the past few years, some of the critical steps have been identified. In the case of basal-cell carcinomas, the discovery of Patched is a development that will further our understanding of these most common cancers. Major questions include: Is Patched necessary for the development of BCCs or do other molecular pathways exist for generating these cancers? Is its homozygous loss in a cell sufficient for malignant transformation? What is the nature of the signal generated in the Sonic hedgehog pathway by Smoothened? Can it be pharmacologically inhibited in order to revert tumor cells to their normal phenotype? The puzzling observation that sunlight exposure in youth participates in carcinogenesis decades later may be partially explained by the following: (1) p53 mutations are detectable very early on in life following UV exposure and even when present on only one chromosome seem to lead to a modest growth advantage likely as a result of decreased apoptosis of these cells following UV. (2) Homozygous p53 mutations and, to a lesser degree, certain heterozygous mutations favor the accumulation of genetic mutations because of loss of function of this critical guardian-of-the-genome protein. (3) Estimates as to the number (n) of genetic mutations required for malignant transformation to carcinoma vary from 2 to 7, whereas the probability that a given cell has acquired mutations in all n genes needed to generate a cancer increases as the nth power of age. Taken together, these observations argue that youthful UV exposure may damage cells, allowing subsequent UV exposure to more effectively accumulate the mutations required to generate a cancer. For squamous-cell carcinoma, although p53 and activated H-ras appear relevant, ample evidence suggests that more steps are required for complete malignant transformation, as mutant H-ras, even in the setting of doubly p53-negative keratinocytes is inadequate for carcinogenesis. In squamous-cell carcinoma, a series of

788

SECTION IX / DERMATOLOGY

genetic hits, the accumulation of which is favored by the loss of p53 function, is likely to be the route to malignancy. It would appear that some of these critical hits by sunlight are directed against target proteins that have not yet been identified.

SELECTED REFERENCES Alcedo J, Ayzenzon M, Von Ohlen T, Noll M, Hooper JE. The Drosophila smoothened gene encodes a seven-pass membrane protein, a putative receptor for the hedgehog signal. Cell 1996;86:221–232. Chen Y, Struhl G. Dual roles for patched in sequestering and transducing Hedgehog. Cell 1996;87:553–563. Corominas M, Kamino H, Leon J, Pellicer A. Oncogene activation in human benign tumor of the skin (keratoacanthomas): is HRAS involved in differentiation as well as proliferation? Proc Natl Acad Sci USA 1990;86:6372–6376. Dahmane N, Lee J, Robins P, Heller P, Ruiz i Altaba A. Activation of the transcription factor Gli1 and the Sonic hedgehog signalling pathway in skin tumors. Nature 1997;389:876–881. Fan H, Oro AE, Scott MP, Khavari PA. Induction of basal cell carcinoma features in transgenic human skin expressing Sonic Hedgehog. Nat Med 1997;3:788–792.

Hahn H, Wicking C, Zaphiropolous PG, et al. Mutations of the human homolog of drosophila patched in the nevoid basal cell carcinoma syndrome. Cell 1996;85:841–851. Jonasen AS, Kunala S, Price GJ, et al. Frequent clones of p53-mutated keratinocytes in normal human skin. Proc Natl Acad Sci USA 1996;93:14,025–14,029. Leffell DJ, Brash DE. Sunlight and skin cancer. Sci Am 1996;275: 52–59. Stone DM, Hynes M, Armanini M, et al. The tumor-suppressor gene patched encodes a candidate receptor for Sonic hedgehog. Nature 1996;384:129–134. Unden AB, Zaphiropoulos PG, Bruce K, Toftgard R, Stahle-Backdahl M. Human patched (PTCH) mRNA is overexpressed consistently in tumor cells of both familial and sporadic basal cell carcinoma. Cancer Res 1997;57:2336–2340. van den Heuvel, Ingham PW. smoothened encodes a receptor-like serpentine protein required for hedgehog signalling. Nature 1996;382: 547–551. Xie J, Murone M, Luoh S, Ryan A, Gu Q, Zhang C, Bonifas J, Lam C, Hynes M, Goddard A, Rosenthal A, Epstein E, de Sauvage F. Activating Smoothened mutations in sporadic basal-cell carcinoma. Nature 1998;391:90–92.

CHAPTER 85 / MELANOMA GENETICS

85

789

Melanoma Genetics DANIEL B. DUBIN AND SAUMYEN SARKAR

INTRODUCTION Cancer is currently being viewed as the cellular accumulation of activated oncogenes and inactivated tumor suppressor genes (see Chapter 7). Detailed studies on the genetics of adenocarcinoma of the colon have provided the empirical framework for the stepwise genetic progression of benign proliferations to frank malignancy. Similar to adenocarcinoma of the colon, melanoma has a putative precursor lesion, the dysplastic nevus, and may arise sporadically or in clusters within predisposed families. Recently, germline mutations have been identified in several hereditary cancer syndromes, including Gardner’s syndrome (APC), familial breast and ovarian cancer (BRCA1/2), and basalcell nevus syndrome (BCNS) (PATCHED). Mutations in these cancer-predisposing genes are intriguing as they have been found in both sporadic and inherited malignancies. Identification of these genes has already resulted in the commercialization of a geneticscreening test (BRCA1/2) for familial breast and ovarian cancer. The discovery of veritable “melanoma-susceptibility genes” would facilitate the development of not only screening and diagnostic tests, but also gene-reconstitution treatment strategies. The increasing incidence of melanoma and the dismal prognosis for advanced-stage disease have spurred investigation for genetic alterations responsible for sporadic and inherited forms of melanoma. Linkage analysis in melanoma-prone kindreds and karyotyping of melanomas and their derived cell lines have implicated nonrandom abnormalities of chromosomes 1, 6, 7, 9, and 10 in the pathogenesis of melanoma. Several candidate melanomapredisposing genes and loci have been reported; however, controversy regarding the identity of veritable “melanoma-susceptibility genes” abounds.

CHROMOSOME 1 Investigators have identified alterations of chromosome 1 in primary melanomas and melanoma cell lines. The majority of these mutations have represented deletions or translocations involving 1p22–36. However, one group has reported t (1:19) translocations involving 1q in 3 cases of advanced stage melanoma. Another study identified 4/30 melanoma cell lines with transforming N-ras mutations that map to 1p22; however, given their relative infrequency, there is concern that such N-ras mutations may represent an epiphenomenon of transformation. The beta chain of NGF maps From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

to the 1p22 region, but no alterations in this gene have been found in melanoma. Using restriction-fragment-length polymorphism analysis, one group demonstrated loss of heterozygosity (LOH) for 1p in 15/35 melanomas and 11/21 melanoma cell lines, thus suggesting the residence of a melanoma suppressor gene in this region. Certain genetic studies have linked alterations in the 1p32-36 region, which contains the Rh gene, to melanoma and the dysplastic nevus syndrome in distinct families; however, this association has been questioned by other studies.

CHROMOSOME 6 Deletions in the 6q1 2–27 region are the most common chromosome 6 alterations that have been found in melanoma. Although the proto-oncogenes c-ros and c-myb have been mapped to the long arm of chromosome 6, mutations of these genes in melanoma are not common. Reported LOH of the 6q22–23 and 6q24–27 regions in melanoma implies that melanoma suppressor gene(s) may reside in these loci. Microcell melanoma cell hybrids into which a normal chromosome 6 has been introduced lose their capacity to form tumors in nude mice. The loss of the normal chromosome 6 from these melanoma-cell hybrids results in a restoration of their tumorigenicity in nude mice.

CHROMOSOME 7 Duplication of all or part of chromosome 7 has been reported in 36/58 advanced stage melanomas in a recent study. Similar duplications of chromosome 7 have been found in glioblastoma cell lines in which overexpression of c-erbB, a component of the epidermal growth factor receptor (EGF) occurs. The notion that overexpression of c-erbB could contribute to tumorigenesis in melanoma is intriguing but not proven.

CHROMOSOME 9 By genetic-linkage analysis, a locus responsible for melanoma predisposition has been mapped to a 2-cM interval in the chromosomal region 9p21. This locus could function to act as a somatically recessive tumor suppressor gene in the manner proposed by Knudson. Consistent with this notion, the 9p21 locus is the site of frequent somatic chromosomal deletions and/or translocations in melanoma tumor cells and cell lines. Linkage data and mutational analysis have suggested that the cyclin-dependent kinase inhibitors (CDKI), p16 and p15, both of which map to the 9p21–22 region, represent the mts-1 and mts-2 genes, respectively. In fact, p16 and/or p15 deletions and/or mutations

789

790

SECTION IX / DERMATOLOGY

have been detected in tumors such as pancreatic adenocarcinoma, esophageal adenocarcinoma, non–small-cell lung carcinoma, leukemia, lymphomas, osteosarcomas, bladder carcinomas, glioblastoma, and mesothelioma. Mechanistically, the CDKIs p15, p16, and possibly p18, the protein product generated from the alternative reading frames of the INK4a gene, represent attractive tumor-suppressor gene candidates (see Chapter 6). The CDKIs act to reinforce the START checkpoint that regulates G1 to S phase transition. By inhibiting cyclin-dependent kinases (CDKs), CDKIs prevent phosphorylation of Rb. Therefore, Rb continues to bind tightly and inactivate the transcription factor E2F, which is essential for driving the G1 to S phase transition. Both p15 and p16, when transfected into transformed mammalian-cell lines can diminish their proliferative capacity. Furthermore, p15 is upregulated by TGF-β, a cytokine which inhibits cellular proliferation in vivo. Therefore, it is an attractive hypothesis to consider that loss of p15 results in a loss of TGF-β–maintained restraint of proliferation. Since p15, p16 deletions/mutations are more common in either tumor-derived cell lines or tumor explants than in the primary tumors, there is concern that loss of p15/p16 function may represent an epiphenomenon of transformation. In fact, a recent report that cites in vitro upregulation of p16 as a key event in the terminal stage of growth arrest in senescence could explain why p16 but not p15 is commonly inactivated in established melanoma cell lines. However, the detection of p15/p16 mutations in some primary tumors has preserved hope that these genes do represent mts-1 and mts-2. Although p16 gene mutations in melanoma cell lines are well documented, linkage studies in kindreds with familial melanoma syndrome linked to 9p21–22 have been inconclusive. One study identified six relevant mutations of p16 in 9/18 kindreds. A second study found a single common p16 mutation in 13/15 families. Recently, germline mutations of p16, one of which has been independently tightly linked to two Australian melanoma pedigrees, have been reposted in 5/33 (18%) of patients with a family history of melanoma. On the other hand, in an analysis of 38 melanomaprone families, 13 of which segregate 9p linkage, only two relevant mutations of p16 were detected. Furthermore, p16/19 knockout mice demonstrated a greater than expected spontaneous development of lymphomas and sarcomas. The knockout data and the mixed-linkage studies raise doubt as to whether or not p16 represents the authentic MLM gene. Could p15 be MLM? Kamb et al. did not find relevant p15 (mts-2) mutations or deletions in their analysis of melanoma-cell lines and germline DNA derived from melanoma prone kindreds. However, we (SS) have found homozygous p15 deletions of exon 2 in 6/10 human melanoma cell lines; by contrast, no homozygous p16 deletions of exon 2 were detected in these same cell lines. Two of the cell lines in which p15 deletions were detected were primary tumor-cell cultures. In cell lines without homozygous deletion of p15, several mutations of both p15 and p16 were noted. The final word on p15’s role in melanoma is pending further study.

CHROMOSOME 10 Translocations and deletions of chromosome 10 have been described in a dysplastic nevus, early-stage melanoma, and latestage melanomas. No specific genes on chromosome 10 have been implicated. One group found both chromosome 10 deletion and chromosome 7 duplication in 10/58 of advanced-stage melano-

mas. This association may actually be greater since LOH of regions of chromosome 10 may occur without gross cytogenetic loss. Nowell has hypothesized that chromosome 10 may contain a transacting suppressor of c-erbB. In theory, the combination of loss of expression of the erbB suppressor and gain of an extra copy of c-erbB would provide a proliferative advantage.

OTHER RELEVANT GENES Two studies have found ras mutations in 5–25% of melanoma tumors. However, Gerhard et al. analyzed 58 patients with sporadic melanoma and concluded that the distribution of ras mutations did not differ from that in a normal population. Recently, a specific germline cyclin-dependent kinase 4 (CDK4) mutation that renders it resistant to CDKI inhibition has been detected in two melanoma pedigrees with an inheritance not linked to a locus on 9p21. This CDK4 mutation may subvert the delicate START checkpoint that is maintained in part by CDKIs such as p16 and p15 and, thus, confer a proliferative advantage to cells with such a mutation. Several lines of evidence have implicated p53 and Rb mutations, two guardians of the START checkpoint, as pathogenic in the development of melanoma. In one study, mutations of p53 were found in 85% of primary and metastatic melanoma. Additionally, patients with Li-Fraumeni syndrome (-/p53) as well as up to 7% of survivors of hereditary retinoblastoma (-/Rb) have been reported to develop melanoma. Further evidence that p53 and Rb alterations may be important in the pathogenesis of melanoma can be gleaned from a transgenic mouse line whose melanocytes express large T antigen, a putative inactivator of p53 and Rb function. These transgenic mice develop ocular and cutaneous melanomas.

CONCLUSIONS The literature is flush with circumstantial evidence implicating alterations of various genes and chromosomal regions in the pathogenesis of melanoma. Given the lessons of colon adenocarcinoma progression, manifestation of the melanoma phenotype likely requires the accumulation of several genetic hits, but not necessarily in any particular order. Furthermore, the array of genetic alterations necessary and sufficient for melanoma development in any one tumor may represent only a small subset of the genetic mutations that could potentially combine to cause melanoma. Cytogenetics and molecular analysis have defined several chromosomal regions in which putative melanoma suppressor genes may lie. However, the identification of true “melanoma susceptibility genes” will likely require further linkage analysis in melanomaprone kindreds, functional studies, and more detailed sequencing of candidate chromosomal regions.

SELECTED REFERENCES Cannon-Albright L, Goldgar D, Meyer L. Assignment of a locus for familial melanoma, MLM to chromosome 9p13-p22. Science 1992; 258:1148–1152. Cannon-Albright LA, Kamb A, Skolnick M. A review of inherited predisposition to melanoma. Semin Oncol 1996;23:667–672. Dracopoli N, Harnett S, Bale B, et al. Loss of alleles from the distal short arm of chromosome 1 occurs late in melanoma tumor progression. Proc Natl Acad Sci USA 1989;86:4614–4618. Fitzgerald M, Harkin D, Silva-Arrieta S. Prevalance of germ-line mutations in p16, p19 ARF and CDK4 in familial melanoma: analysis of a clinic base population. Proc Natl Acad Sci USA 1996;93:8541–8545. Greene MH. Genetics of cutaneous melanoma and nevi. Mayo Clin Proc 1997;72:467–474.

CHAPTER 85 / MELANOMA GENETICS

Hannon G, Beach D. p15 INK4B is a potential effector of TGF-β-induced cell cycle arrest. Nature 1994;371:257–261. Hunter T, Pines J. Cyclins and cancer II: cyclin D and CDK inhibitors come of age. Cell 1994;79:573–582. Hussussian C, Strvewing J, Goldstein A. Germline p16 mutations in familial melanoma. Nat Genet 1994;8:15–21. Kamb A, Liu Q, Hhapshaman K, et al. Rates of p16 (MTS1) mutations in primary tumors with 9p21 loss. Science 1994;265:416,417. Kamb A, Shattuck-Eidens D, Eeles R. Analysis of the p16 gene (CDKN2) as a candidate for the chromosome 9p melanoma susceptibility locus. Nat Genet 1994;8:22–26. Mintz B, Silvers W. Transgenic mouse model of malignant skin melanoma. Proc Natl Acad Sci USA 1993;90:8817–8821.

791

Parmiter A, Nowell P. Cytogenetics of melanocytic tumors. J Invest Dermatol 1993;100:2545–2585. Sarkar S, Kupper T. The putative 9p21 tumor suppressor gene for sporadic malignant melanoma: p15INK4b may be more important than p16INK4a (Abstract). J Invest Dermatol 1995;104:568. Serrano M, Hannon G, Beach D. A new regulatory motif in cell-cycle control causing specific inhibition of cyclin D/CDK4. Nature 1993;366:704–707. Wolfel T, Hauer M, Schneider J, et al. A p16INK4a_insensitive CDK4 mutant targeted by cytolytic T lymphocytes in a human melanoma. 1995;269:1281–1284. Zuo L, Weger J, Yang B, et al. Germline mutations in the p16INK4a binding domain of CDK4 in familial melanoma. Nat Genet 1996;12:97–99.

CHAPTER 86 / PSORIASIS

86

793

Psoriasis JAMES T. ELDER AND JOHN J. VOORHEES

BACKGROUND HISTORICAL PERSPECTIVES Psoriasis was first clearly described by Celsus (25 BC–45 AD), and conditions resembling psoriasis are present in the works of Hippocrates. In 1809 Robert Willan was the first to accurately describe the various forms of psoriasis, which were finally recognized as one entity and distinguished from leprosy by von Hebra in 1841. The association of psoriasis with a distinctive form of arthritis was first recognized by Alibert in 1818. Careful histopathologic studies of psoriasis were first carried out by Auspitz, Unna, and Munro in the late 1800s. DISEASE DEFINITION Psoriasis is a chronic, inflammatory, hyperproliferative disease of the skin, scalp, nails, and joints, affecting 1–2% of the United States population at an estimated annual cost of $1.6 billion in 1985. Psoriasis is found worldwide, although its frequency varies widely among different ethnic groups. It has a variety of clinical presentations, most of which eventuate into erythematous, scaly plaques with or without nail disease, and arthritis. Susceptibility to psoriasis is unmistakably heritable, but environmental factors, notably trauma, stress, and infections, are also important determinants of disease onset and severity. At the cellular level, psoriasis is characterized by: (1) markedly increased epidermal proliferation and incomplete differentiation; (2) elongation, dilatation, and “leakiness” of the superficial plexus of dermal capillaries; and (3) a mixed inflammatory and immune-cell infiltrate of the epidermis and papillary dermis. A multitude of plausible pathomechanisms can be envisaged for psoriasis. However, true molecular insight into the cause of psoriasis is lacking. As stated by Lomholt over 30 years ago:

Question upon question may be asked—the disease is capricious and refuses to part with its innermost secret. SCOPE In this chapter, we will review what has been accomplished to solve the long-standing riddle of psoriasis, particularly in the areas of immunology and genetics. In addition, we will attempt to summarize its protean clinical manifestations, and review its diagnosis, differential diagnosis, and management.

CLINICAL FEATURES SKIN Several types of psoriatic skin lesions are recognized, including chronic plaque, guttate, pustular, inverse, palmoplantar, and generalized (Fig. 86-1). Of these, chronic plaque lesions are From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

by far the most common, and over time, most other forms eventuate into chronic plaque disease. The typical chronic plaque is a well-demarcated, erythematous lesion with loose, silvery scales that can vary from a few millimeters to many centimeters in diameter. The scale can be lifted off the surface of lesions, revealing a shiny base and microscopic points of bleeding (Auspitz sign). Lesions are typically located over the extensor surfaces of the extremities, and often display striking symmetry. Other areas of predilection include the scalp, umbilicus, intergluteal cleft, and genitalia. Plaques may assume a variety of shapes, described as annular, circinate, gyrate, and serpiginous. Scalp lesions can be particularly thick and refractory to therapy, and often show a sharp cutoff at the hairline. With the exception of childhood cases, the face is usually spared. Eruptive, or guttate psoriasis (Latin gutta, “drop”) appears rapidly as tens to hundreds of 5- to 10-mm papules on the trunk and proximal extremities, usually in young patients following a streptococcal or other upper respiratory-tract infection. Generalized pustular psoriasis presents as waves of sterile pustules 2–3 mm in diameter that appear on erythematous skin, usually accompanied by fever. This is the form of psoriasis most likely to evolve into erythroderma, defined as total-body skin involvement. Inverse psoriasis presents as red plaques with moist scaling, typically located in the inframammary or inguinal folds, umbilicus, intergluteal cleft, axillae, or under the prepuce. Sebopsoriasis lesions feature a more yellowish, greasy scale in the scalp, facial, presternal, and upper back—a distribution reminiscent of seborrheic dermatitis. Localized pustular psoriasis of the palms and soles and geographic tounge are unusual forms, probably distinct from psoriasis. NAILS Psoriasis produces three distinctive nail changes, at least one of which can be found in approximately 50% of psoriasis patients (Fig. 86-2). These include pitting, onychodystrophy, (destruction of the nail plate), and “oil drop” spotting. Pits are typically well-demarcated, randomly distributed, 1- to 2-mm depressions in the surface of the nail plate, produced by defective terminal differentiation of keratinocytes in the nail matrix. Onychodystrophic changes arise from disease of the nail bed and range from separation of the nail plate (onycholysis) to thick, yellowish, crumbly nails, to complete loss of the nail, especially in pustular cases. Oil-drop spots are orangish-to-brown discolorations underneath the nail, resulting from the leakage of serum proteins through the nail-bed capillaries. JOINTS Essentially the only systemic manifestation of psoriasis, psoriatic arthritis typically presents between the ages of 35 and 45, usually but not always after onset of skin disease (Fig. 86-3).

793

794

SECTION IX / DERMATOLOGY

Figure 86-1 Morphology of psoriasis skin lesions. (A) Chronic plaque lesions of the knees. Note silvery scale that reveals a shiny base when removed. (B) Guttate and chronic plaque lesions of the lower back. (C) Inverse (intertriginous) psoriasis of the axilla. (D) Scalp psoriasis. Note thick, adherent scale and relatively sharp cutoff at the hairline. (Photos courtesy of Harrold Carter and James Rasmussen.)

The incidence of arthritis in psoriatics is unclear, but is at least three times that of the general population. Disease is oligoarticular and asymmetrical in over 80% of patients, and is most severe in the small joints of the hands and feet and the large joints of the legs. Involvement of the distal interphalangeal joints distinguishes psoriatic from rheumatoid arthritis, which tends to be more proximal and symmetrical. Fortunately, only approximately 5–8% of psoriatic arthritis patients develop highly severe, deforming joint disease (arthritis mutilans). Enthesopathy is the result of mononuclear cell accumulation at tendon or ligament insertions, and can progress to involve the entire digit (dactylitis, “sausage digit”). COURSE The natural history of psoriasis is highly unpredictable. Onset is usually in childhood to the early 20s, with a median age of 15–20 years and over 75% of cases presenting by the age of 40 (see below). Typically, lesions wax and wane, and may reappear in the same or in different locations. Approximately 25% of patients experience a complete remission at some point in their lives, which may last from 1 to 50 years. The Köebner phenomenon, also known as the isomorphic response, refers to the tendency of psoriasis to arise at sites of cutaneous injury. The phenomenon is reported by 25% of patients, requires epidermal and possibly high dermal injury, and is usually seen in the setting of flaring rather than stable or regressing disease.

DIAGNOSIS CLINICAL EVALUATION Based on the foregoing clinical features, the diagnosis of psoriasis is generally straightforward. The Auspitz sign is well-known but is neither sensitive nor specific for psoriasis. Fungal infection, seborrheic dermatitis, chronic

eczema, candidiasis, cutaneous T-cell lymphoma, drug eruptions, syphilis, pityriasis rosea, lichen planus, and lupus erythematosus can present an occasional diagnostic challenge. Skin biopsy can be very useful in these cases. HISTOPATHOLOGY The histopathologic features of psoriasis are variable yet distinctive (Fig. 86-4). These include: (1) epidermal edema and hyperplasia; (2) loss of the granular layer with retention of keratinocyte nuclei (parakeratosis) with or without collections of neutrophils in the stratum corneum; (3) dilatation and tortuosity of the superficial dermal capillaries; and (4) a moderate mononuclear cell infiltrate of the epidermis and dermis (see below). LABORATORY Several serologic, hematologic, and metabolic abnormalities may be present. However, they have no value as diagnostic tests because of lack of sensitivity and specificity. Whereas antistreptolysin O titers suggestive of recent infection are found in 30–40% of psoriasis patients, specific Streptococci can often be recovered from intertriginous sites, and Staphylococcus aureus colonization is increased in frequency, the diagnostic value of cultures and examination of skin scrapings is limited to ruling in other entities, rather than ruling out psoriasis. Mild anemia with folate and iron deficiency is common, and negative nitrogen balance is occasionally seen, because of loss of these nutrients in scale. Acute-phase reactants including C-reactive protein and α-2 macroglobulin correlate with elevated erythrocyte-sedimentation rate and disease severity. Serum uric acid levels are also increased in about one third of patients, although frank gout is rare. Psoriasis displays very strong HLA associations that correlate with age of onset and disease severity (see below); however, HLA typing is only a research tool at this time.

CHAPTER 86 / PSORIASIS

Figure 86-2 Psoriasis of the nails. (A) Fingernail involvement. Note presence of pits, onycholysis, and oil-drop spotting. (B,C) Toenail involvement, before (B) and after (C) 4 months of CsA therapy. (B) Note thickening and yellowish discoloration as well as periungual involvement. (C) Note that nail and skin involvement improve markedly after CsA treatment. (Photos courtesy of Harrold Carter and James Rasmussen.)

GENETIC BASIS OF DISEASE GENETIC EPIDEMIOLOGY Although the inheritance patterns found in certain psoriasis kindreds seem to suggest autosomal-dominant inheritance with reduced penetrance (Fig. 86-5), the weight of evidence from population-based studies indicates that susceptibility to psoriasis is determined by multiple genes as well as environmental stimuli (multifactorial inheritance). Two parameters are often used for estimating the genetic component of multifactorial diseases: heritability (h2), that portion of the variability in the manifestation of psoriasis ascribable to genetic factors; and risk ratio (λR), the risk of psoriasis in relatives of degree R relative to the general population. Psoriasis displays one of the highest h2 values known in the multifactorial diseases (0.8–0.9). Although identical-twin concordance rates were much lower in Australia than in Scandinavia, h2 (which depends on the ratio of concordance rates in identical vs fraternal twins) was very similar in both locations. Given the therapeutic effects of ultraviolet (UV) light (see below), it is possible that sun exposure may be an important environmental factor in psoriasis. At present, it is not clear how many genes are involved in psoriasis; however, the parameter λ R can be used to infer that the number must be greater than 1.

795

Figure 86-3 Psoriatic arthritis. (A) Note fusiform swelling of the middle digit (enthesopathy), accentuation of DIP involvement in association with severe nail disease. (B) Same patient as (A); note limited range of motion. (C) Arthritis mutilans. Note progressively more severe distal involvement. (Photos courtesy of Harrold Carter and James Rasmussen.)

λ R is also a useful measure of the practicability of identifying genes in multifactorial diseases. For specific genes to be identifiable by genetic-linkage methods, λR for siblings (termed λS) should exceed a value of 3. In psoriasis, estimates of λS vary from 4 to 10. JUVENILE-ONSET PSORIASIS Two types of psoriasis have been distinguished, largely on the basis of age of onset. The onset of psoriasis peaks between 15 and 25 years, with onset before the age of 40 in 75–90% of patients. A second, much smaller, peak is detectable at 65–70 years of age. Psoriatics who are under 40 years old at onset are much more likely to have affected first degree relatives (Fig. 86-6A), to express known HLA susceptibility alleles, and to experience severe and recurrent disease. As adult-onset psoriatics display a different HLA profile, this population may actually have a different disease, more closely related to the spondyloarthropathies. HLA ASSOCIATIONS AND THE EFFECT OF LINKAGE DISEQUILIBRIUM There can be no doubt that one or more genes in the HLA locus strongly influence psoriasis susceptibility.

796

SECTION IX / DERMATOLOGY

Figure 86-4 Histopathology of psoriasis. (A) Normal skin (scale bar = 60 µ). (B) Chronic plaque psoriasis (scale bar = 150 µ). Compared to normal skin, note striking regularity of epidermal downgrowth (acanthosis), loss of the granular layer, and retention of nuclei in the stratum corneum (parakeratosis). (C) Psoriasis lesion, high-power view (scale bar = 30 µ). Note prominent capillary dilatation (C), dermal dendritic cells, identifiable by their plump nuclei (D), mononuclear cell infiltrate in the dermo-epidermal junctional zone (M), widening of the extracellular spaces (E), and mitotic keratinocytes (arrowheads). (Photomicrographs courtesy of Harrold Carter, James Varani, and John Headington.)

HLA-B13, -B17, -B39, -B57, -Cw6, and -Cw7 (Class I), as well as HLA-DR4 and -DR7 (Class II), all show highly significant positive associations, especially in juvenile-onset disease (Fig. 86-6B). Whereas HLA-Cw6 has consistently produced the strongest associations with psoriasis worldwide, the frequencies of Cw6 and DR7 are approximately the same in psoriatic patients. The tendency of Cw6 and DR7 to be found together is not confined to psoriatics, but is also observed in the general population (allelic association). Population and family studies have revealed the basis of this allelic association: Cw6 and DR7 are linked together on the same chromosome much more often than would be predicted by chance in both affected and unaffected individuals (linkage disequilibrium). Linkage disequilibrium in HLA is not confined to Cw6 and DR7, but involves multiple alleles at HLA-A, -B, and -C in the Class I region, several loci in the Class III region, and HLA-DRA, -DRB, -DQA, and -DQB in the Class II region (Fig. 86-7). (Curiously, disequilibrium does not appear to extend to HLA-DP.)

Through the use of family studies, these allele combinations can be shown to be caused by cosegregation on the same haploid chromosome, and are therefore referred to as haplotypes. Certain haplotypes are particularly prevalent in one or more world populations, and appear to be of ancient origin. It is estimated that all HLA haplotypes found worldwide today represent one of approximately 50 of these so-called ancestral haplotypes or their recombinants. The ancestral HLA haplotype 57.1, which carries HLA-B57 in addition to HLA-Cw6 and -DR7, is 15 times more common in German psoriatics than in the German population. However, this haplotype is neither necessary nor sufficient for disease, as only about 35% of German psoriatics carry 57.1, and only 20–25% of the German 57.1+ population develops psoriasis. Studies of those rare individuals carrying only the Class I or Class II “ends” of this haplotype strongly suggest that the major susceptibility determinant resides on its Class I end, in the vicinity of HLA-C. However, several lines of evidence indicate the susceptibility determinant is not HLA-Cw6 itself, but rather an allele at one of at least 10 other genes in the centromeric HLA Class I region. At this writing, the HLA-linked susceptibility gene remains to be identified. NON-HLA GENES Several laboratories are conducting genome-wide searches for psoriasis susceptibility genes, using both recombination-based and affected sibling pair (ASP) methods of linkage analysis (see Chapter 4 for a discussion of genetic analysis methods). In insulin-dependent diabetes mellitus, at least six such non-HLA loci have been confirmed using these methods (see Chapter 57). Two large-scale genome-wide scans for psoriasis susceptibility loci have recently been published (Trembath et al., 1997; Nair et al., 1997). Both of these studies confirmed HLA involvement and identified a candidate region on the short arm of chromosome 20. One study (Nair et al.) confirmed an earlier report of a non-HLA locus on the distal long arm of chromosome 17 (Tomfohrde et al., 1994), and identified a region of linkage on the long arm of chromosome 16. Interestingly, this region coincides with a confirmed region of linkage to Crohn’s disease. Psoriasis is seven times more common in Crohn’s disease patients than in the general population. It is attractive to speculate that a proinflammatory genetic mutation or polymorphism common to both diseases may reside in this chromosomal region.

MOLECULAR PATHOPHYSIOLOGY OF DISEASE EVOLUTION OF LESIONS Early lesions of psoriasis display mast-cell degranulation, capillary dilatation, papillary dermal edema, and the appearance of a mononuclear infiltrate. However, the exact sequence of events in emerging lesions has been a subject of debate since the late 19th century. Most studies agree that as mononuclear cells begin to invade the epidermis, intercellular edema (spongiosis) of basal keratinocytes appears, followed by focal vacuolization, and patchy loss of granular layer keratinocytes and the appearance of parakeratotic cells with or without neutrophilic infiltration of the stratum corneum. During this transition, the time required for keratinocyte transit from the basal layer to the granular layer decreases from 2 weeks to 2 days, the keratinocyte mitotic rate increases approximately 10-fold, and a characteristic pattern of altered keratin synthesis ensues, with a transition in suprabasal keratins from keratins1 and 10 to keratins 6 and 16. More chronic lesions continue to be markedly hyperplastic, resulting in a regular pattern of epidermal thickening (acanthosis) with greatly elongated and edematous rete ridges, dilation, and tortuosity of superficial capillaries, and a complete loss of granular cells with diffuse parakeratosis.

CHAPTER 86 / PSORIASIS

797

Figure 86-5 Examples of multiplex psoriasis pedigrees. Note generation-to-generation and male-to-male transmission, suggestive of autosomal dominant inheritance with reduced penetrance. (Modified from Nair et al., Human Heredity 1995;45:219–230.) A

Figure 86-6 (A) Increased heritability of juvenile-onset (type I) psoriasis. (B) HLA associations in Type I vs Type II psoriasis. (From Elder et al., Arch Dermatol 1994;130:216–224.)

THE CASE FOR T-CELL INVOLVEMENT Given its known specificity for T cells, the marked efficacy of the immunosuppressant cyclosporine A (CsA) in psoriasis has provided important functional evidence of T-cell involvement. Besides cyclosporine, other T-cell–specific immunosuppressants including FK506, peptide T, anti-CD3 and -CD4 monoclonal antibodies (MAb), and diptheria toxin-coupled interleukin-2 (IL-2) have all proven efficacious in limited studies, albeit not without side effects. Additional evidence for T-cell involvement comes from reports of disease exacerbation after IL-2 and interferon-γ (IFN-γ) therapy, and after bone marrow transplantation from a psoriatic donor. Conversely, complete clearing of severe psoriasis has been reported after transplantation from a nonpsoriatic donor.

In recent years, it has become increasingly clear that psoriatic lesions contain substantial numbers of dendritic antigen-presenting cells (APCs) in addition to T cells. Most of the T cells present are not naive, but rather memory T cells capable of binding to skinspecific homing receptors on superficial dermal capillary endothelial cells. In addition, accessory factors required for efficient T-cell activation are present in the lesion, including costimulatory APC surface ligands such as B7-1 and B7-2, and a rich cytokine milleu produced by surrounding keratinocytes, mast cells, and mononuclear cells. Several lines of evidence argue for ongoing T-cell activation by antigen in the lesion itself. These include: (1) the presence of actively dividing memory T cells in contact with APCs in the dermis and epidermis of psoriatic lesions; (2)

798

SECTION IX / DERMATOLOGY

Figure 86-7 HLA region map, showing ancestral haplotypes. Psoriasis-associated ancestral haplotypes are boxed. Depiction of HLA loci are incomplete and not to scale. cen, centromere; tel, telomere; Mb, megabases; C, Caucasoid; M, Mongoloid; N, Negroid. “Hotspot” indicates approximate region of loss of linkage disequilibrium described in the text. (Data from Degli-Epsoti et al., Immunogenetics 1992;36:345–356, Hum Immunol 1992;34:242–252, and Wu et al., Hum Immunol 1992;33,89–97.)

oligoclonal usage of TCR Vβ chains in the CD8+ subset of lesional T cells; and (3) maintenance of the psoriatic phenotype after transplantation of lymph node-free, split-thickness skin to globally immunodeficient (SCID) mice. Thus, psoriatic lesions appear to feature all components of the classical “ternary complex” required for T-cell activation: MHC, antigen, and the responding T-cell receptor. It is attractive to speculate that preferential Vβ subset usage reflects an important role for superantigen as the initial stimulus for T-cell activation, followed by further activation by nominal antigen; however, this concept remains to be confirmed. Both helper (CD4+) and suppressor (CD8+) T cells are present in lesions, with most CD8+ cells being found in the epidermis and most CD4+ cells in the dermis. A case has been made for greater importance of CD8+ T cells, based on: (1) the strength of MHC class I associations in psoriasis; (2) the principle that Class I molecules usually present antigens to CD8+ T cells; (3) the evidence for oligoclonal expansion of CD8+ T-cells; and (4) the selective depletion of CD8+ by UVB light treatment. Whereas this is an attractive hypothesis, it does not explain why IL-2, IFN-γ, and IL-12 strongly predominate over IL-4, -5, and -10 in psoriatic lesions, fitting the so-called Th1 lymphokine profile usually considered characteristic of CD4+ helper T cells. However, it must be remembered that both CD4+ and CD8+ cells can contribute to the Th1 lymphokine profile. Adding to the complexity, we have already discussed the confounding effects of linkage disequilibrium on the interpretation of Class I vs Class II associations in psoriasis, and there is evidence that anti-CD4 MAbs can clear psoriasis. Therefore, until more is known, it seems prudent to acknowledge the potential roles of both CD4+ and CD8+ T cells in lesion development. Recently, T cells cloned from psoriatic lesions have been used to prepare lymphokine-rich conditioned media. Interestingly, these media stimulated the proliferation of short-term keratinocyte cultures from uninvolved skin of psoriasis patients, but not cells cultured from normal epidermis. Neutralization studies showed that IFN-γ was a major mitogen in the context of other as-yet-unidentified lymphokines in the conditioned medium, whereas it is growth inhibitory when used alone. In the long-term, however, keratinocyte proliferation in psoriasis appears to be maintained by autocrine stimulation of the EGF receptor via one of at least three EGF-like growth factors: TGF-α, amphiregulin, and HB-EGF.

Based on the foregoing observations, the following working model for maintenance of psoriatic lesions can be envisaged (Fig. 86-8): Memory T cells enter the skin, having been directed to the skin by upregulated adhesion molecules on activated endothelial cells. Once present in epidermis and dermis, these T cells may encounter bacterial and/or viral superantigens, leading to an expansion of a subset of cells with restricted utilization of TCR Vβ chains. These superantigen-activated cells will also encounter foreign or host-derived standard antigens on the surfaces of dendritic antigen-presenting cells, to which a limited number of the superantigen-activated T cells will respond by further expansion in the presence of appropriate costimulatory signals. T-cell activation causes release of IFN-γ, which may in turn stimulate keratinocyte proliferation, whereas IL-2 and other lymphokines may promote further activation of T cells, APCs, endothelial cells, and/or keratinocytes via a cytokine cascade. Alternatively, T cells may exert sublethal damage to keratinocytes, which respond by entering a wound-healing pathway of epidermal differentiation. Whether the psoriatic keratinocyte is genetically more susceptible to growth stimulation by lymphokines from these T cells requires further study.

MANAGEMENT/TREATMENT In view of its genetic basis, lifelong duration, and low mortality, avoidance of toxicity is a cornerstone of psoriasis therapy. Nevertheless, effective treatment of moderate to severe psoriasis can entail substantial toxicity and demands careful management. Because of incomplete terminal differentiation, the stratum corneum of the psoriatic plaque is fragile and presents a defective barrier to epidermal water loss. Recent research has shown that loss of epidermal barrier function is a trigger for keratinocyte production of various pro-inflammatory cytokines and adhesion molecules including IL-8, TNF-α, and ICAM-1. Emollients have long played an important role in psoriasis therapy, presumably by improving barrier function. Although the density of the bacterial flora in psoriasis lesions is double that of normal skin, antibiotic treatment is only occasionally helpful. CORTICOSTEROIDS/RETINOIDS/VITAMIN D Whereas the antipsoriatic efficacy of topical and systemic steroids has been known since the 1950s, their cellular target remains unclear. Irrespective of their site of action, it is now clear that steroids

CHAPTER 86 / PSORIASIS

799

Figure 86-8 Model for T-cell pathogenesis of psoriasis. APC, dendritic antigan presenting cell; MHC, HLA Class I or Class II molecules; SAg, superantigen; T, T cell; Tm, memory T cell; Tm*, superantigen-activated memory T cell; M, monocyte/macrophage; KC, keratinocyte; PMN, polymorphonuclear leukocyte; EC, endothelial cell; EC*, activated endothelial cell expressing E-selectin (receptor for CLA), ICAM-1 (receptor for LFA-1), and VCAM (receptor for VLA-4). (See text for details.)

exert their beneficial effects via binding to specific nuclear receptors, which serve as ligand-dependent transcription factors. These nuclear receptors are encoded by a superfamily of related genes that encompasses the receptors for not only corticosteroids, but also other effective antipsoriatic agents including vitamin D derivatives and retinoids. At present, the full spectrum of steroid-responsive genes is unknown; however, steroids are known to inhibit the transcription of major immunomodulatory genes such as IL-2 by interfering with the binding of other activating transcription factors such as AP-1. Systemic steroids were initially highly popular but are now little used because of their well-known complications as well as “rebound” flaring of disease upon their withdrawal. In their stead, many different topical corticosteroid preparations of varying potency and vehicle formulations have appeared. Systemic side effects and local effects such as atrophy and hypopigmentation are a concern for only the most potent of these compounds, generally fluorinated derivatives. Retinoids and vitamin D3 derivatives constitute the other two “legs” of the “therapeutic triad” of nuclear receptor ligands with antipsoriatic effects. The synthetic retinoid etretinate has been used for nearly 20 years, especially in pustular and guttate flares alone or in combination with ultraviolet light. Recently, a topical retinoid has been found to be effective; this compound may interfere with AP-1 activity in addition to activating retinoic acid receptors. Topical vitamin D3 derivatives have also emerged as useful compounds in the 1990s. At present, the major vitamin D3-related compound is calcipotriol, a short-lived structural analog of 1,25 dihydroxy vitamin D3, which lacks most of the hypercalcemic side effects of vitamin D3 itself. As with the corticosteroids, the cellular and molecular targets of retinoid and vitamin D3 action remain unclear. Both classes of compounds have profound effects upon keratinocyte differentiation, and must be used with care in psoriasis, as they produce their own characteristic patterns of skin irritation.

ULTRAVIOLET LIGHT Psoriasis therapy with combinations of high-energy ultraviolet B (UVB) light (290–320 nm) with crude coal tar was introduced in 1925, and the combination of psoralens with lower-energy UVA light (320–400 nm) followed in 1974. Initially, both regimens were thought to act directly on keratinocyte proliferation. However, it is now recognized that both therapies induce immunosuppression in addition to a complex pattern of keratinocyte effects, including immunosuppressive cytokine production and induction of apoptosis. CYCLOSPORINE A As discussed earlier, CsA exerts potent antipsoriatic effects, with nearly immediate clearing of psoriasis when used at transplantation dose of 14 mg/kg/d. The drug remains highly effective at less nephrotoxic doses as low as 3 mg/kg/d. Its mechanism of action is thought to involve binding to the protein cyclophilin, with subsequent inhibition of the calmodulin-dependent protein phosphatase regulatory subunit calcineurin B and subsequent inability to translocate a component of the transcription factor NF-AT from the cytoplasm to the nucleus of the T cell. METHOTREXATE As in the case of UV therapy, methotrexate was initially thought to exert its effects via inhibition of keratinocyte proliferation. However, it is effective under doses and schedules of administration that fail to inhibit keratinocyte proliferation in vitro. It now appears that this drug exerts profound antiinflammatory effects by promoting the extracellular accumulation of adenosine. OTHER AGENTS Anthralin is a strong reducing agent long used topically in psoriasis, either alone or in combination with UVB light. Although quite effective, its use is limited by skin staining and irritation. Its mechanism of action is likely to be multifactorial. Sulfasalazine is a systemic drug that is effective in a subset of psoriasis patients, with only limited toxicity and the advantage of low cost. Its mechanism of action is also unknown, but is thought to be anti-inflammatory and/or immunosuppressive.

800

SECTION IX / DERMATOLOGY

FUTURE DIRECTIONS As can be appreciated from the previous sections, it appears that the actions of nearly all antipsoriatic drugs can be understood in terms of immunosuppressive and/or antiinflammatory effects, with T-cell activation as a likely primary target via the ternary complex of APC, antigen, and T-cell receptor in the context of appropriate accessory signals. Thus it is not surprising that a variety of novel molecular approaches to blocking this activation event are currently under development. These include vaccination against specific peptides of the TCR Vβ families characteristic of oligoclonal and presumably pathogenic T cells; administration of soluble CTLA-4 to block the accessory signal; administration of IL-10 to shift the lymphokine phenotype away from the diseaseassociated Th1 and toward the countervailing Th2 state; and additional agents that can block T-cell signal transduction with fewer side effects than CsA and FK506. As a genodermatosis, the question remains whether psoriasis will prove to be amenable to gene therapy. The issue of direct correction of mutant gene function must await further identification of the genes involved. Assuming that defective gene function can be documented in bone-marrow-derived cells, it will not be unreasonable to expect to reintroduce the normal version of the gene via autologous transplantation of genetically engineered stem cells or long-lasting hematopoietic precursors of appropriate linage. Moreover, inasmuch as simple approaches to introduction of DNA into the skin by direct injection or even topical application are proving to be feasible, it is also reasonable to consider the introduction of various immunomodulators at the level of the gene, rather than the cognate protein. Under either scenario, the future prospects of gene therapy for psoriasis deserve serious consideration.

SELECTED REFERENCES Barker JN, Griffiths CE. Progress in psoriasis. Psoriasis: from gene to clinic. London, UK, 5-7 December 1996. Mol Med Today 1997;3:193,194. Bata-Csorgo Z, Hammerberg C, Voorhees JJ, Cooper KD. Kinetics and regulation of human keratinocyte stem cell growth in short-term primary ex vivo culture. Cooperative growth factors from psoriatic lesional T lymphocytes stimulate proliferation among psoriatic uninvolved, but not normal, stem keratinocytes. J Clin Invest 1995;95:317–327. Chang JC, Smith LR, Froning KJ, et al. CD8+ T cells in psoriatic lesions preferentially use T-cell receptor V beta 3 and/or V beta 13.1 genes. Proc Natl Acad Sci USA 1994;91:9282–9286. Christophers E, Sterry W. Psoriasis. In: Fitzpatrick TB, Eisen AZ, Wolff K, Freedberg IM, Austen KF, eds. Dermatology in General Medicine. New York: McGraw-Hill, 1993; pp. 489–514. Cronstein BN, Naime D, Ostad E. The antiinflammatory effects of methotrexate are mediated by adenosine. Adv Exp Med Biol 1994;370:411–416. Degli-Esposti MA, Leaver AL, Christiansen FT, Witt CS, Abraham LJ, Dawkins RL. Ancestral haplotypes: conserved population MHC haplotypes. Hum Immunol 1992;34:242–252. Elder J. Transforming growth factor-alpha and related growth factors. In: Luger T, Schwarz T, eds. Epidermal Growth Factors and Cytokines. New York: Marcel Dekker, 1994; pp. 205–240.

Elder J. Cytokine and genetic regulation of psoriasis. In: Dahl M, ed. Advances in Dermatology. St. Louis: Mosby-Year Book, 1995; pp. 99–134. Elder JT, Nair RP, Guo SW, Henseler T, Christophers E, Voorhees JJ. The genetics of psoriasis. Arch Dermatol 1994;130:216–224. Fry L. Psoriasis. Br J Dermatol 1988;119:445–461. Greaves MW, Weinstein GD. Treatment of psoriasis [see comments]. N Engl J Med 1995;332:581–588. Henseler T. The genetics of psoriasis. J Am Acad Dermatol 1997; 37:S1–S11. Henseler T, Christophers E. Psoriasis of early and late onset: characterization of two types of psoriasis vulgaris. J Am Acad Dermatol 1985;13:450–456. Krueger JG, Wolfe JT, Nabeya RT, et al. Successful ultraviolet B treatment of psoriasis is accompanied by a reversal of keratinocyte pathology and by selective depletion of intraepidermal T cells. J Exp Med 1995;182:2057–2068. Lander ES, Schork NJ. Genetic dissection of complex traits. Science 1994;265:2037–2048. Mangelsdorf DJ, Evans RM. The RXR heterodimers and orphan receptors. Cell 1995;83:841–850. Modlin RL. Th1-Th2 paradigm: insights from leprosy. J Invest Dermatol 1994;102:828–832. Morganroth GS, Chan LS, Weinstein GD, Voorhees JJ, Cooper KD. Proliferating cells in psoriatic dermis are comprised primarily of T cells, endothelial cells, and factor XIIIa+ perivascular dendritic cells. J Invest Dermatol 1991;96:333–340. Moss P, Charmley P, Mulvihill E, et al. The repertoire of T cell antigen receptor beta-chain variable regions associated with psoriasis vulgaris. J Invest Dermatol 1997;109:14–19. Nair RP, Henseler T, Jenisch S, et al. Evidence for two psoriasis susceptibility loci (HLA and 17q) and two novel candidate regions (16q and 20q) by genome-wide scan. Hum Mol Genet 1997;6:1349–1356. Nickoloff BJ, Kunkel SL, Burdick M, Strieter RM. Severe combined immunodeficiency mouse and human psoriatic skin chimeras. Validation of a new animal model. Am J Pathol 1995;146:580–588. Nickoloff BJ, Naidu Y. Perturbation of epidermal barrier function correlates with initiation of cytokine cascade in human skin. J Am Acad Dermatol 1994;30:535–546. Northrop JP, Crabtree GR, Mattila PS. Negative regulation of interleukin 2 transcription by the glucocorticoid receptor. J Exp Med 1992;175: 1235–1245. Stem RS. Psoriasis. Lancet 1997;350:349–353. Tomfohrde J, Silverman A, Barnes R, et al. Gene for familial psoriasis susceptibility mapped to the distal end of human chromosome 17q. Science 1994;264:1141–1145. Trembath RC, Clough RL, Rosbothem JL, et al. Identification of a major susceptibility locus on chromosome 6p and evidence for further disease loci revealed by a two stage genome-wide search in psoriasis. Hum Mol Genet 1997;6:813–820. Vallat VP, Gilleaudeau P, Battat L, et al. PUVA bath therapy strongly suppresses immunological and epidermal activation in psoriasis: a possible cellular basis for remittive therapy. J Exp Med 1994;180: 283–296. Winchester R. Psoriatic arthritis. In: Fitzpatrick TB, Eisen AZ, Wolff K, Freedberg IM, Austen KF, eds. Dermatology in General Medicine. New York: McGraw-Hill, 1993; pp. 515–527. Wong RL, Winslow CM, Cooper KD. The mechanisms of action of cyclosporin A in the treatment of psoriasis. Immunol Today 1993; 14:69–74.

CHAPTER 87 / ATOPIC DERMATITIS AND ATOPY

87

801

Atopic Dermatitis and Atopy DONALD Y. M. LEUNG AND LARRY BORISH

BACKGROUND Atopic dermatitis (AD) is a chronic inflammatory skin disease frequently seen in patients with a personal or family history of asthma and allergic rhinitis. The term atopic dermatitis was first introduced in 1933 by Hill and Sulzberger in recognition of this close association between AD and respiratory allergy. During the past 10 years, there have been extraordinary strides made in our understanding of the immunopathogenesis of allergic diseases. In particular, this constellation of inherited illnesses have now been demonstrated to be associated with activation of a specific group of cytokine genes encompassing interleukin (IL)-3, IL-4, IL-5, IL-13, and granulocyte-macrophage colony-stimulating factor (GM-CSF). The molecular basis for selective activation of this cytokine gene cluster and their immunologic consequences are now actively being pursued by many laboratories. However, it is clear that allergic diseases result from a polygenic inheritance pattern that not only involves cytokine-gene activation but other less-well-defined gene products as well. In addition, the clinical expression of allergic diseases is highly dependent on a complex interaction between the host and its environment, e.g., allergen exposure. This chapter will review some of these recent developments and their clinical implications.

DIAGNOSTIC AND CLINICAL FEATURES OF AD The diagnosis of AD is based on the constellation of clinical features in Table 87-1. Approximately 50% of patients develop AD by the first year of life, and an additional 30% between age 1 and 5 years. Nearly 80% of patients with AD eventually develop allergic rhinitis or asthma later in childhood. Many of these patients outgrow their skin disease as they are developing respiratory allergy. This observation is consistent with the concept that the clinical expression of allergic disease is determined in part from local tissue allergen sensitization and compartmentalization of the immune response in the skin vs the respiratory mucosa. SKIN REACTION PATTERNS IN AD Intense pruritus and cutaneous reactivity are the hallmarks of AD. Scratching may be intermittent throughout the day but is usually worse in the early evening and night. Patients with AD also have a reduced threshold for pruritus. Clinically, this is supported by the observation that allergens, reduced humidity, excessive sweating, and irritants— e.g., wool, acrylic, soaps, and detergents—can exacerbate pruritus and scratching. From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

Several types of inflammatory skin lesions are commonly seen in AD. Acute lesions are intensely pruritic, and characterized by erythematous papules over erythematous skin. These are associated with extensive excoriations, erosions, and serous exudate. Subacute dermatitis is characterized by erythematous, excoriated, scaling papules. Chronic dermatitis is characterized by thickened skin, accentuated skin markings (lichenification), and fibrotic papules. In chronic AD, all three stages of skin reactions usually coexist in the same individual. At all stages of AD, patients usually have dry skin. Skin distribution and reaction patterns vary according to the patient’s age and disease chronicity. During infancy, the AD is generally more acute and primarily involves the face, scalp, and the extensor surfaces of the extremities. In older patients who have had long-standing skin disease, the rash is characterized by the chronic form of dermatitis with lichenification and localization to the flexural areas of the extremities. ROLE OF ALLERGENS IN AD Serum IgE levels are elevated in 80–85% of patients with AD. Approximately 85% of patients have positive immediate skin tests or RAST for specific IgE to a variety of food and inhalant allergens. However, positive immediate skin tests to specific allergens do not always indicate clinical sensitivity and patients who outgrow AD frequently continue to have positive skin tests. Indeed, May first made a distinction between symptomatic and asymptomatic hypersensitivity based on the observation that AD patients with positive food skin tests did not always have positive challenges to the foods implicated by IgE responses. These clinical observations suggest that the relationship between IgE and clinical disease is not exclusively dependent on IgE-mediated mast-cell degranulation. Well-controlled, double-blind, placebo-controlled food challenge studies (DBPCFC) have, however, demonstrated that food allergens can exacerbate skin rashes in at least a subset of patients particular young children with AD. Importantly, elimination of putative food allergens in such patients results in significant improvement of their skin disease. As patients grow older, they outgrow their food allergy, but a majority of them become sensitized to inhalant allergens. Whereas sensitization to inhalant allergens in most cases reflects the development of coexisting respiratory allergy—e.g., asthma or allergic rhinitis—a number of clinical studies suggest that inhalation or contact with aeroallergen may play a role in the exacerbation of AD. CUTANEOUS INFECTIONS Patients with AD have an increased tendency to develop viral, fungal, and bacterial skin infections. These infections are generally localized to the skin. Deep-seated infections suggest the possibility of hyperimmunoglobulinemia E

801

802

SECTION IX / DERMATOLOGY

Table 87-1 Diagnostic Features of Atopic Dermatitisa

Table 87-2 Immunologic Features of Atopic Dermatitis

Major features Pruritus Typical appearance and distribution of skin lesions Facial and extensor involvement in infancy and early childhood Flexural involvement and lichenification by adolescence Chronic or frequently relapsing course (duration >6 weeks) Personal or family history of either AD or respiratory allergy Minor features Increased susceptibility to skin infections, particularly S. aureus Xerosis (dryness of the skin) Early age of onset Multiple positive immediate prick skin tests Elevated serum IgE levels Ichthyosis; keratosis pilaris; hyperlinearity of palms Nonspecific hand/foot dermatitis Scalp dermatitis, e.g., cradle cap

Increased IgE production Immediate skin-test reactivity to multiple allergens Increased basophil spontaneous histamine release Decreased CD8 suppressor/cytotoxic number and function Increased expression of CD23 on mononuclear cells Chronic macrophage activation with increased secretion of GM-CSF, PGE2, and IL-10 Expansion of IL-4–, IL-5–, and IL-13–secreting Th2-like cells Decreased numbers of IFN-γ–secreting Th1-like cells

aModified

from Hanifin and Rajka.

syndrome or another immunodeficiency syndrome. Viral infections include herpes simplex, vaccinia, warts, molluscum contagiosum, and papilloma virus. The most common viral infection is herpes simplex, which tends to spread locally or can become generalized. Superficial fungal infections also appear to occur more frequently in atopic individuals. The potential importance of these dermatophyte infections is suggested by the reduction of AD skin severity following treatment with antifungal agents such as ketoconazole. Considerable attention has focused on the contribution of Staphylococcus aureus colonization and infection to the severity of AD. S. aureus is found in over 90% of AD skin lesions. Honeycolored crusting, extensive serous weeping, folliculitis, and pyoderma indicate bacterial infection usually secondary to S. aureus in patients with AD. The importance of S. aureus in AD is supported by the observation that not only patients with impetiginized AD, but also AD patients without superinfection show clinical response to combined treatment with antistaphylococcal antibiotics and topical corticosteroids. Recent studies suggest that S. aureus exacerbates or maintains skin inflammation in AD by secreting a group of toxins known to act as superantigens that stimulate marked activation of T cells and macrophages. Since staphylococcal enterotoxins (SEs) are globular proteins of 24–30 kDa, the possibility that they could act as allergens has also been studied. In this regard, it has been found that nearly half of AD patients produce IgE directed to staphylococcal toxins, particularly SEA, SEB, and toxic shock syndrome toxin-1 (TSST-1). This is of interest because staphylococci isolated from AD skin lesions predominantly secrete one of these three exotoxins. These findings suggest the possibility that local production of SEs at the skin surface could cause IgE-mediated histamine release and thereby trigger the itch–scratch cycle that can exacerbate AD.

IMMUNOLOGIC FINDINGS IN ATOPIC DERMATITIS A number of observations suggest an underlying immunoregulatory abnormality in AD (Table 87-2). Studies of T-cell clones support the concept that activation of a subpopulation of helper cells leads to the release of cytokines important in the pathogenesis of AD. In mice, two types of CD4+ T-cell clones have been

described, based on their pattern of cytokine secretion. T-helper type 1 (Th1) cells secrete IL-2 and interferon (IFN)-γ, but not IL-4 or IL-5. In contrast, Th2 cells produce IL-4, IL-5, and IL-13, but not IFN-γ. IL-4 and IL-13 act as IgE isotype-specific switch factors, and induce the expression of VCAM-1, an adhesion molecule involved in the migration of mononuclear cells and eosinophils into sites of tissue inflammation. IL-5 promotes the differentiation, vascular endothelial adhesion, and survival of eosinophils and also enhances histamine release from basophils. In contrast, IFN-γ inhibits IgE synthesis as well as the proliferation of Th2 cells. The lack of IFN-γ production, as well as the concomitant activation of IL-4, IL-5, and IL-13 is thought to play a critical role in the pathogenesis of AD (Fig. 87-1). In support of the concept that this profile of cytokines arises from the selective activation of Th2, as compared to Th1, cells are a number of studies demonstrating increased frequency of allergen-specific T-cells producing increased IL-4 and IL-5, but little IFN-γ in the peripheral blood and skin lesions of patients with AD. IL-4 inhibits IFN-γ production and downregulates the differentiation of Th1 cells. In addition, atopic monocytes have been found to have elevated cAMP phosphodiesterase and secrete increased levels of IL-10 and prostaglandin E2. Both IL-10 and PGE2 inhibit IFN-γ production and may therefore contribute to the decreased IFN-γ production by AD PBMC.

IMMUNOHISTOLOGY OF ATOPIC DERMATITIS IMMUNOHISTOLOGIC FINDINGS The histologic features of AD depend on the acuity and therefore the duration of the skin lesion. Uninvolved or clinically normal skin of AD patients is histologically abnormal and demonstrates mild hyperkeratosis, epidermal hyperplasia, and a sparse dermal cellular infiltrate consisting primarily of T lymphocytes. Acute lesions are characterized by marked intercellular edema (spongiosis) of the epidermis, and intracellular edema noted as ballooning of the keratinocytes. A sparse epidermal infiltrate consisting primarily of T lymphocytes is frequently observed. In the dermis, there is a marked perivenular inflammatory-cell infiltrate consisting predominantly of lymphocytes, and occasional monocyte-macrophages. Eosinophils, basophils, and neutrophils are rarely present in the acute lesion. In chronic lichenified lesions, the epidermis is hyperplastic with elongation of the rete ridges, prominent hyperkeratosis, and minimal spongiosis. Increased numbers of Langerhans cells are present in the epidermis, and macrophages dominate the dermal mononuclear cell infiltrate. The number of mast cells are increased in number but are generally fully granulated. Endothelial cells of the superficial venular plexus and deep venules are hypertrophied with enlarged nuclei and prominent nucleoli.

CHAPTER 87 / ATOPIC DERMATITIS AND ATOPY

803

Figure 87-1 Cellular and cytokine interactions involved in the pathogenesis of atopic dermatitis. (Reproduced with permission from Leung et al. J Allergy Clin Immunol 1995;96:302–319.)

PATTERNS OF CYTOKINE EXPRESSION IN AD SKIN LESIONS The pattern of cytokines expressed locally play a critical role in modulating the nature of tissue inflammation. Therefore, an analysis of cytokine expression in AD is critically dependent on the acuity or duration of the skin lesion. The in situ hybridization (ISH) technique has been used to investigate the expression of IL-4, IL-5, and IFN-γ messenger RNA (mRNA) in skin biopsies from clinically normal (uninvolved), acute (erythematous AD lesions of 2 weeks’ duration) skin lesions of patients with AD. As compared with normal control skin, uninvolved skin of patients with AD had a significant increase in number of cells expressing IL-4, but not IL-5 or IFN-γ, mRNA. Acute and chronic skin lesions, when compared to normal skin or uninvolved skin of AD patients, had significantly greater numbers of cells that were positive for mRNA for IL-4 and IL-5. However, neither acute AD or uninvolved AD skin contained significant numbers of IFN-γ mRNA-expressing cells. As compared with acute AD, chronic AD skin lesions had significantly fewer IL-4 mRNA-expressing cells, but significantly greater IL-5 mRNA-expressing cells. T-cells constituted the majority of IL-5–expressing cells in acute and chronic AD lesions. Chronic lesions also expressed significantly greater numbers of

activated IL-5 mRNA-expressing eosinophils than acute lesions. These data indicate that although cells in acute and chronic AD lesions are associated with increased activation of IL-4 and IL-5 genes, acute skin inflammation in AD is associated with a predominance of IL-4 expression, whereas maintenance of chronic inflammation is predominantly associated with increased IL-5 expression and eosinophil infiltration. In addition, chronic AD lesions are associated with overexpression of GM-CSF and IL-10 expression.

MULTIFUNCTIONAL ROLE OF IgE IN AD SKIN INFLAMMATION Although the histologic features of AD closely resembles a type IV delayed-type hypersensitivity (DTH) reaction, several lines of evidence indicate that AD is not a conventional DTH reaction, and furthermore that IgE-mediated mechanisms play a role in the pathogenesis of AD. T cells that infiltrate into conventional DTH skin reactions (e.g., tuberculin skin reactions) secrete IFN-γ and therefore induce the expression of HLA-DR on skin keratinocytes. In contrast, keratinocytes in the AD skin lesion do not express HLA-DR. More direct evidence that the T cells in skin lesions of AD differ from the Th1 cells in conventional DTH has

804

SECTION IX / DERMATOLOGY

come from studies demonstrating increased IL-4 and IL-5, but not IFN-γ, mRNA expression in acute AD skin lesions. Furthermore, Th2 cells have been found to be just as effective in inducing skin inflammation as Th1 cells. However, the mediation of skin inflammation by Th2 cells is IL-4, but not IFN-γ, dependent. Taken together, these data suggest that two histologically indistinguishable cutaneous DTH reactions exist: The first is mediated by IFN-γ–secreting Th1 cells found in conventional DTH reactions. The second is mediated by IL-4– and IL-5–secreting Th2 cells and involves allergen-induced, cell-mediated reactions. There are several mechanisms by which IgE molecules can contribute to the induction of a mononuclear cell infiltrate. In this regard, clinically significant allergen-induced reactions are associated with an IgE-dependent biphasic response. In such reactions, following exposure to allergen, mast cells bearing IgE directed to the relevant allergen release various mediators and cytokines into local tissue within 15–60 min of allergen challenge. This immediate reaction likely contributes to the acute pruritus and erythema observed after exposure of AD patients to relevant food and inhalant allergens. Several hours after this immediate reaction begins to subside, there is onset of an IgE-dependent late-phase reaction (LPR) characterized initially by expression of leukocyte adhesion molecules on postcapillary venular endothelium, followed by the infiltration of eosinophils, neutrophils and mononuclear cells into the inflamed area. Granulocytes reach their maximum cell accumulation at 6–8 h, and by 24–48 h after onset of the reaction, the cellular infiltrate consists predominantly of mononuclear cells. Using ISH, Kay and coworkers have reported that the cellular infiltrate in allergeninduced late-phase skin reactions express increased mRNA for IL-3, IL-4, IL-5, and granulocyte-macrophage colony-stimulating factor but no mRNA for IFN-γ. These results suggest that the T cells infiltrating into the allergen-induced LPR, similar to allergen-specific T cells grown from AD skin lesions, are Th2-like cells. Langerhans cells and macrophages infiltrating into the AD skin lesion bear IgE antibody. Binding of IgE to Langerhans cells occurs via both high-affinity and low-affinity IgE (CD23) receptors. Macrophages can express CD23 in response to IL-4. Allergens have been demonstrated to activate IgE-bearing macrophages in an IgE-dependent manner with the formation of leukotrienes, PAF, IL-1, and TNF. The activation of IgE-bearing Langerhans cells and macrophages by allergens could thus contribute to the skin inflammation associated with AD. There is also mounting evidence that IgE-bearing Langerhans cells in AD skin play an important role in cutaneous allergen presentation to Th2 cells. Of note, IgE-bearing Langerhans cells from AD skin lesions, but not Langerhans cells that lack surface IgE, are capable of presenting house-dust-mite allergen to T cells. These results suggest that cell-bound IgE on Langerhans cells facilitate binding of allergens to Langerhans cells prior to their processing and antigen presentation. Although allergen challenges and experimental models suggest the participation of specific mechanisms of inflammation, it should be emphasized that an analysis of the AD skin lesion does not allow simple classification discretely into an IgE-mediated LPR or a T-cell–mediated immune reaction. Thus, in all likelihood, the mononuclear cell infiltrate in the AD skin lesion reflects a combination of both IgE-dependent mast cell/basophil degranulation and Th2-cell–mediated responses elicited during acute exposures to allergen and other antigens or superantigens.

Finally, although the release of a variety of mediators, e.g., histamine and proteases into the skin following challenge by allergens trigger acute pruritus in AD, clinical studies suggest that the actual development of eczematoid skin rashes is dependent on the skin trauma inflicted by scratching. Once the itch–scratch cycle is triggered, the mechanisms by which scratching promotes inflammation of the skin is suggested by a number of recent observations demonstrating the keratinocyte is an important epidermal source of cytokines including IL-1, IL-6, IL-8, GM-CSF, and TNF-α. In this regard, any injury including mechanical trauma to the keratinocyte will result in the release of cytokines that can induce inflammation through a number of actions. This includes the release of IL-1, TNF-α, and IL-4, which are critical cytokines in the induction of adhesion molecules, such as ELAM-1, ICAM-1, and VCAM-1 that attract lymphocytes, macrophages, and eosinophils into cutaneous sites of inflammation. At this stage, a wide variety of resident and infiltrating cells are then capable of secreting cytokines and mediators that sustain the inflammation. AD therefore results from a combination of specific and nonspecific cellular mechanisms that serve to trigger and maintain skin inflammation.

TARGETING OF Th2-LIKE CELL RESPONSE IN ATOPIC SKIN Taken together, AD and asthma are both associated with the local infiltration of Th2-like cells, allergen sensitization, the development of chronic local tissue inflammation and the presence of organ-specific (cutaneous vs bronchial) hyperreactivity that may be caused by underlying tissue inflammation. The potential mechanisms that determine tissue specificity of Th2 cell responses in different allergic diseases are therefore of interest. In this regard, studies in experimental animal models have demonstrated heterogeneity in the ability of memory T cells to migrate to mucosal vs nonmucosal tissues. This tissue-selective homing is regulated in large part at the level of T-lymphocyte recognition of vascular endothelial cells (EC) via the interaction of differentially expressed T-lymphocyte homing receptors (HR) and their EC ligands. In humans, lymphocyte/EC adhesion molecule pairs thought to participate in tissue-selective lymphocyte homing include the skin-selective HR, cutaneous lymphoid antigen (CLA), and peripheral lymph node HR, L-selectin. It has been found that T cells migrating into the skin are highly enriched for the CLA-expressing memory T-cell subset, whereas memory T cells isolated from the airways of asthmatics are predominantly CLA negative. Thus, the propensity of a given individual to develop AD as opposed to asthma may depend on differences in the skin- vs lung-seeking behavior of their memory/ effector T cells. Children with food-induced AD provide an opportunity for determining whether there is a relationship between the tissue specificity of a clinical reaction to an allergen and the expression of HR on T cells activated in vitro by the relevant allergen. In this regard, a recent study assessed the expression of CLA and L-selectin on peripheral blood T cells from patients with AD and milk-induced eczema, and compared their HR expression, following stimulation with casein to T cells collected from patients with allergic eosinophilic gastroenteritis, milk-induced enterocolitis, or nonatopic healthy controls. The casein-reactive T cells from patients with milk-induced eczema displayed significantly higher levels of CLA than Candida albicans-reactive T cells from the same patients, and either casein- or C. albicans-reactive T cells from nonatopic controls or noneczematous atopic patients.

CHAPTER 87 / ATOPIC DERMATITIS AND ATOPY

Further evidence for the relationship between CLA and cutaneous T-cell responses in AD has been provided by Babi et al. These investigators analyzed the CLA phenotype of circulating memory T cells in AD vs asthmatic patients who were sensitized with house dust mite. When peripheral blood CLA+CD3+CD45RO+ T cells were separated from CLA–CD3+CD45RO+ T cells, the mite-specific T-cell proliferation response in patients with AD sensitized to dust mite was localized to CLA+ T cells. In contrast, mitesensitive asthmatics had strong mite-dependent proliferation responses in their CLA– T-cell subset. The link between CLA expression and skin disease-associated T-cell effector function in AD was demonstrated by the observation that freshly isolated circulating CLA+ T cells in AD patients, but not normal controls, selectively demonstrated spontaneous production of IL-4, but not IFN-γ. These observations strongly support the concept that, in human allergic diseases, immunologic mechanisms exist to target memory Th2 cells to specific organs.

GENETICS OF ATOPIC DERMATITIS AND ATOPY When the classic studies of Cooke and Van der Veer in 1916 first introduced the concept of allergic sensitization, familial predisposition was considered an important component of these conditions. However, atopy and allergic disorders such as AD are complex genetic diseases that do not follow simple Mendelian inheritance patterns. For atopy and AD, the challenges involved in understanding their genetics include the difficulty of dissecting out the roles of many genes, each with variable degrees of involvement in any given individual. Genetic studies have been difficult because identification of atopic individuals is also confounded by the role of environmental influences. In this regard, atopic, genetically at-risk individuals do not necessarily develop an allergic disease nor are people with allergic disease always carriers of any given disease gene. Nevertheless, there has been considerable progress made in identifying potential genes involved in the inheritance of atopy. GENETIC STUDIES OF ATOPIC DERMATITIS A familial predisposition to AD as a manifestation of atopy is well-established. Among subjects with AD, 50–67% have a history of either one or both parents having an atopic disorder and the risk of a child developing AD is 25–30% with a history of a single parent with atopy and 50–75% with a dual history. In addition to an inherited predisposition to AD mediated through atopy, additional genetic factors may specifically predispose to the development of dermatitis. An extensive study of allergy in families in Switzerland confirmed that bronchial asthma, allergic rhinitis, and AD were genetically linked. However, the history of respiratory atopic diseases in relatives of probands with AD was significantly lower than that observed for probands with asthma or allergic rhinitis. Twin studies are a useful means to assess the presence of a genetic contribution in complex genetic diseases in which there are contributions from both genetic and environmental influences. Twins, whether monozygotic (MZ) or dizygotic (DZ), if raised in the same household, share much of the same environmental influences. However, whereas MZ twins have identical genomes, DZ twins on average have only half their chromosomes in common. Thus a higher concordance rate of a given condition in MZ twins provides evidence for the presence of genetic influences. These data provide the information to produce a heritability estimate, an estimate of the relative contributions to a given condition provided by genetic as opposed to environmental factors. The most exten-

805

sive investigation of twins for AD was the Swedish Twin Study, which identified a concordance rate of AD of 15% in MZ and 5% for DZ twin pairs. Studies of AD in Danish twins found much higher concordance rates for AD but also confirmed a higher rate in MZ (77%) than DZ twins (15%). These authors concluded that genetic factors play a decisive role in the development of AD. GENETICS OF IgE Genetic studies of AD have been problematic because of variable diagnostic criteria among different studies. The identification of IgE and establishment of its role as the basis for allergic sensitization has provided an improved basis for performing genetic studies of atopy. Both the ability to accurately quantify total IgE and to identify the presence of specific IgE toward a given allergen by either skin testing or serum immunoassays have become objective means for phenotyping atopic subjects. Twin studies of elevated serum IgE performed by Hopp and Townley in 61 monozygotic twin pairs and 46 dizygotic pairs revealed a heritability estimate of 0.61. Several additional studies have placed the heritability index of serum IgE in at 0.60–0.70. These twin studies unambiguously establish a genetic component to the inheritance of elevated serum IgE. Several additional studies have used segregational analysis to attempt to identify the mechanism for this inheritance. Cookson and Hopkin studied the familial occurrence of elevated IgE in 239 members of 43 families recruited by the presence of an asthmatic proband. Ninety percent of the atopic children had one or more atopic parents. These authors concluded that the high IgE phenotype appeared to be inherited as an autosomal-dominant trait with a variable clinical expression. In contrast, Meyers and Marsh studied the inheritance pattern of total serum IgE and concluded that there was evidence for a major IgEregulating gene that is inherited via an autosomal-recessive model, but confounded with other familial and polygenetic factors. Their studies demonstrated evidence of this major IgE-regulating gene effect only in some families. Parks et al. studied serum IgE levels from 42 families based on an asthmatic proband and determined that a polygenic heritability model best explained the data. These conflicting results are consistent with either an autosomal recessive gene acting in combination with a polygenic component or, alternatively, the presence of a single major gene that is only operative in some families or ethnic groups, but not in others. Thus, even though it provides a more objective basis for phenotyping subjects, the use of total IgE data in segregational analyses has failed to produce unambiguous conclusions regarding its mechanism of inheritance. This confusion results primarily from the impossibility of controlling for environmental factors that influence serum IgE levels. LESSONS FROM THE GENETICS OF ASTHMA AND BRONCHIAL HYPERREACTIVITY Similar to the results observed with AD and IgE, genetic studies of asthma have produced conflicting results. As with AD, this also reflects variable diagnostic criteria and the lack of a definitive biochemical or laboratory marker. Edfors-Lubs reviewed nearly 7000 twins in the Swedish Twin Study and reported a diagnosis of asthma in 3.8% of the study population. The concordance rate for asthma was 19% in monozygotic twins and only 4% in dizygotic twins. Most investigators have found higher concordance rates for asthma in twins. Falliers et al. reviewed the literature and reported 30–80% concordance rates for asthma in monozygotic twins and 4–45% in dizygotic twins. As with the IgE studies, these data support a hereditary component in the development of asthma.

806

SECTION IX / DERMATOLOGY

More importantly, several studies have now addressed the question of whether asthma is inherited in part separately from atopy. Sibbald recruited asthmatic individuals and investigated the presence of asthma and positive skin tests in the nuclear family. They reported that asthma was more frequent in parents (13%), offspring (9%), and siblings (8%) of the asthmatic probands and there were no significant differences whether or not the subjects were allergic (extrinsic) asthmatics or nonallergic (intrinsic) asthmatics. As would be expected, the prevalence of atopic disorders in general (asthma, allergic rhinitis, or atopic eczema) were higher in the relatives of extrinsic than intrinsic asthmatics. They concluded that asthma and atopy are inherited independently and that the presence of atopy enhances the likelihood of asthma being expressed. Other studies have investigated the genetics of bronchial hyperreactivity (BHR). Longo et al. performed familial analyses of BHR and documented increased occurrence of BHR in the parents of asthmatic probands. Konig and Godfrey studied exercise challenge and skin-test responses in the nuclear families of 12 children with asthma. A positive exercise challenge developed in 43% of family members and 45% of subjects with at least one positive skin test responded positively. Townley and coworkers found increased BHR in 47 nuclear families of patients with asthma when compared with 26 families with no asthma or family history. Subjects without a personal history of asthma, but from families with a history of asthma, showed bimodal distributions in their BHR, whereas control subjects from families without a history of asthma or allergy had a single uniform distribution. A bimodal distribution was also found in the nonasthmatic parents of asthmatic children. All of these observations supported an inherited component to BHR unrelated to atopy. The segregational analyses of Townley’s group supported a genetic component of BHR, but not by a single gene. The consensus of these investigations, however, suggests that in addition to a general susceptibility to the atopic diseases, there may also be a separate inherited tendency to develop asthma. Unfortunately, easily reproductive objective measures of cutaneous hyperreactivity do not exist, but given the parallels between asthma and AD the possibility that cutaneous hyperreactivity may be genetically transmitted in AD must be considered as well. POSITIONAL CLONING STUDIES OF ATOPIC DISORDERS With diseases such as the atopic disorders for which the responsible gene is unknown, the approach to identifying diseasecausing genes frequently utilizes positional cloning. This technique is based on the presence of highly polymorphic markers whose location on the human genome have been mapped. Markers located close to the disease-causing gene will link to the disease when multiple kindreds are analyzed. The frequency of crossover events permits estimation of the distance between the genome marker and the disease gene. Identification of a closely linked marker is followed by “chromosomal walking” until the mutant gene is identified. In the atopic disorders with their variable expression, polygenic inheritance, and the challenges of unambiguously phenotyping subjects, such an approach will be particularly challenging. In addition to all of these previously discussed difficulties that will confound attempts at positional-based cloning, a final challenge is derived from the high frequency of these disorders. This results in the incumbent danger that even though one parent may not be atopic or have an atopic disorder they could still be the carrier of the relevant disease gene.

Despite these difficulties, investigators have utilized positional cloning technology to define the genetic basis for atopy. Cookson and colleagues defined atopic individuals through the presence of either having allergen-specific IgE or a high total serum IgE. Their initial studies of 40 nuclear and 3 extended families was consistent with an autosomal dominant inheritance. Segregational analysis of 7 extended families showed a linkage to the chromosome 11q marker D11S97 and produced a maximal lod score of 5.58. Subsequently, additional studies of 64 families and sib-pair analysis on 743 subjects confirmed this linkage. Analysis of 11q demonstrated that this marker mapped close to the gene for the β chain of the high-affinity IgE receptor. These investigators were able to link atopy to 3 specific amino acid substitutions within the fourth transmembrane hydrophic domain of the β chain. The function of the β chain remains unclear insofar as the α and γ chains are sufficient for transducing activating signals and it is not obvious how the three conservative amino acid substitutions they describe may alter signal transduction. However, it is possible that hydrophobic interactions between transmembrane domains may be critical for proper assembly of the IgE receptor and that, although not necessary for triggering histamine release, the β chain may modulate different signaling pathways. Modulation of cytokine production by a base substitution would be consisting with the biological activity of an asthma gene. Unfortunately, several reports have been unable to confirm these data, although two other groups have confirmed a linkage to this locus The basis for these conflicting results is unclear, except to argue that the development of atopy is a complex genetic process with possibly many genes and pathways leading to the development of the clinical phenotype. CANDIDATE ATOPIC GENES Because of the difficulties inherent in positional-based cloning strategies, many investigators have opted for an approach based on performing linkage analysis with several specific candidate genes. The basis for this approach is to argue that the genes that cause the atopic disorders are, in fact, known. Specifically, certain proteins may be either abnormally regulated or otherwise function inappropriately to produce atopy and the allergic diseases. These candidate genes include cytokine genes such as IFN-γ and IL-12, cytokine-receptor genes, the ε heavy-chain gene, the low-affinity receptor for IgE (FceRII), CD40 and its ligand, and numerous other genes. Additional genes, such as those responsible for increased PGE2 production and cyclic AMP-phosphodiesterase activities are candidate genes more specific for AD. Similarly both the IL-10 gene and gene(s) controlling its expression may contribute to the diminished cellular immune responses characteristic of this disorder. However, this review will focus on studies of the major histocompatibility complex (MHC) and the cytokine gene complex on chromosome 5. Linkages have been proposed between MHC class I and class II antigens and both the presence of atopic disease and the specificity of allergen responses. MHC class II alleles are responsible for mediating the genetic component of immune responsiveness to specific antigenic epitopes. Whereas specific MHC alleles have been proposed to link to atopic phenotype, such a linkage has not been well-established. With what is known about the function of the MHC molecules, there is no obvious biochemical basis for the MHC to regulate development of either atopy or an atopic disease. Linkages to MHC are at best likely only to explain a heritable basis for the specific immune response to a given allergenic epitope. Linkages of immune responsiveness to Amb a V and the ryegrass

CHAPTER 87 / ATOPIC DERMATITIS AND ATOPY

antigens Lol p I and II have been linked to specific MHC loci. However, a clinically important role for MHC loci in regulating even specific immune responses to allergens is unlikely given the degeneracy of the binding of allergenic epitopes to class II antigens. Thus, both clinical observations and twin studies on the presence of IgE to specific allergens are more consistent with environmental influences (i.e., presence and concentration of antigen) than by genetic factors. Several other genes have proved to be better candidates for inducing atopy and allergic diseases. These include numerous genes that regulate IgE production, eosinophilia, and mast-cell proliferation. A surprising number of these genes are clustered on the long arm of chromosome 5, between 5q22 and 5q31. This cluster includes the genes for ILs-3, -4, -5, -9, -13, and GM-CSF. IL-4 and IL-13 are responsible for initiation of ε germline transcription, which in combination with signals generated by CD40 acting through the CD40 ligand leads to the IgE isotype switch. IL-5 is an eosinophilopoietin and is an activating factor for mature eosinophils. Additional factors contributing to eosinophil activation and survival are GM-CSF and IL-3. Proliferation and increased releasability of mast cells is a characteristic feature of atopic diseases and represents a T-cell–dependent process. Within this cluster, both IL-3 and IL-9 function as mast cell growth and differentiating factors, whereas IL-3 also enhances histamine release. Of these genes IL-4, IL-13, and IL-5, as well as IL-3 and GM-CSF are particularly tightly linked, separated by only approximately 300 kbp. To some extent, these genes are under shared regulatory control. Thus, GM-CSF, IL-3, IL-4, and IL-5 have all been identified at the sites of allergic reactions. The transcriptional dysregulation of these cytokines may be the basis for the atopic state. Thus, when T-lymphocyte clones are generated towards the dust mite allergen, the clones from atopic donors are characterized by the transcriptional activation of IL-4, whereas those obtained from nonatopic donors are characterized by IFN-γ and IL-2, but not IL4. That genes in close proximity may be coregulated has been observed for several other gene clusters such as has been described for the IL-1 gene cluster (IL-1a, IL-1b, and IL-1ra) and the MHC complex. Such coregulation does not preclude the possibility that under appropriate conditions one or more genes may be transcribed nonsynchronously. It has therefore been proposed that regulatory elements associated with these genes may be responsible for inducing the atopic state, that these elements may be linked to specific chromosomal markers, and, therefore, that the genetic predisposition to atopy can be identified by the presence of these markers. Two groups have now reported their results utilizing linkage analyses of polymorphisms present on chromosome 5q to asthma or elevated serum IgE. Marsh et al. analyzed 170 individuals from 11 extended Amish families by sib-pair analysis. Utilizing a series of markers within the region 5q31.1 these authors found evidence for genetic linkages to high IgE phenotype for five markers within a narrow region (1.4 cM) of 5q31.1. No linkage was found to specific IgE. Similar results were obtained by Bleeckers’ group in a study of 92 nuclear families from northern Holland. Via sib-pair analyses these investigators demonstrated highly significant linkages to both total IgE and to bronchial hyperreactivity with three markers within the chromosome 5q31.1 cytokine gene cluster. In summary, although much work remains to elucidate the genetics of atopic diseases such as AD and asthma, several potential genes have been identi-

807

fied and this is likely to be an exciting area of investigation in the future.

MANAGEMENT OF AD Current treatment of AD is directed at symptom relief and reduction of cutaneous inflammation and the reduction of exacerbating factors are critical for effective management. Factors that must be considered and eliminated include irritants, allergens, and emotional stresses. Systemic antimicrobial therapy, particular antistaphylococcal antibiotics, is often necessary because AD skin has enhanced binding properties for S. aureus and, because of the frequent scratching, becomes secondarily infected. Maintenance of daily skin care with hydration of the skin and appropriate use of topical steroids to reduce skin inflammation is critical. Therapy must be individualized and is dependent on whether the patient is experiencing an acute flare or dealing with the management of chronic AD. For patients who are poorly controlled on conventional therapy, alternative therapies should be considered. Ultraviolet light therapy may be a useful adjunctive therapeutic modality. Under professional supervision, UVB can be effective and has been found to have antiinflammatory effects in part because of its ability to inhibit lymphocyte trafficking and antigen-processing. Photochemotherapy with oral psoralen therapy followed by UVA (PUVA) may also be helpful in patients with severe disease. However, PUVA is reserved for patients with more recalcitrant disease because of the expense, and the potential increased risk of skin cancer. Since patients with AD manifest abnormalities in immune regulation, therapy directed toward correction of their immune dysfunction represents an alternative approach. In this regard, therapeutic trials using several experimental immunomodulators or immunosuppressive agents have been reported. IFN-γ, a cytokine that downregulates Th2-cell function, has been found in placebo-controlled trials to reduce clinical severity associated with AD and decrease total circulating eosinophil counts. Cyclosporine, a drug that downregulates cytokine production, has also been reported in double-blind, placebo-controlled trials to cause a significant improvement in AD. Cyclosporine therapy did, however, lead to mild renal and liver toxicity. Thus, the side effects associated with prolonged systemic cyclosporine therapy make it an unlikely candidate for chronic treatment of AD. Of note, FK506, an analog of cyclosporine, has recently been used topically in clinical trials and found to be efficacious with no significant side effects. In addition, the new high-potency phosphodiesterase inhibitors may be useful in targeting the increased PDE activity in atopic monocytes and have demonstrated promising preliminary clinical results.

FUTURE DIRECTIONS The current review has attempted to highlight several important advances in our understanding of the immunopathogenesis of AD. These include the observation that IgE has a multifunctional role in the pathogenesis of allergic inflammation. Aside from its involvement in IgE-mediated degranulation of mast cells/basophils, it is also involved in the activation of macrophage/monocytes and the stimulation of Th2 cells. Many of the critical cytokines involved in the pathogenesis of AD and the potentially important mechanisms by which inflammation occurs at the sites of allergic responses have now been identified. These important

808

SECTION IX / DERMATOLOGY

new advances will allow scientists in the future to target certain candidate genes to determine whether their abnormal regulation or altered function contributes to the molecular basis of atopy. Elucidation of the key genes involved in atopy is likely to provide new methods for classifying atopic diseases, particularly with respect to diagnosis, disease severity, and possibly response to therapy. Considering the complexity of multiple genes involved in atopy it is difficult to envision the use of gene therapy in the foreseeable future. Novel immunologic therapies for chronic allergic diseases are likely to include the modulation of Th2 cells and their cytokines targeting specific organs. Recent studies with experimental immunomodulatory and immunosuppressive agents are promising and serve as proof of the concept that therapies that downregulate T-cell function and cytokine secretion are effective in reducing the clinical severity of AD. As our understanding of the immunopathogenesis of allergic responses continues to grow, manipulation of the immune response in AD and other allergic diseases is likely to become an exciting new direction in the treatment of this fascinating group of illnesses.

ACKNOWLEDGMENTS The author would like to thank Maureen Sandoval for her assistance in the preparation of this manuscript. Supported in part by NIH grants AR 41256, RR00051, and HL36577.

SELECTED REFERENCES Abernathy-Carver KJ, Sampson HA, Picker LJ, Leung DYM. Milkinduced eczema is associated with the expansion of T cells expressing cutaneous lymphocyte antigen. J Clin Invest 1995;95:913–918. Babi LFS, Picker LJ, Soler MTP. Circulating allergen-reactive T cells from patients with atopic dermatitis and allergic contact dermatitis express the skin-selective homing receptor the cutaneous lymphocyte-associated antigen (CLA). J Exp Med 1995;181:1935–1940. Bieber T, de la Salle H, Wollenberg A, et al. Human epidermal Langerhans cells express the high affinity receptor for immunoglobulin E (FcεRI). J Exp Med 1992;175:1285–1290. Blumenthal M, Mendell N and Yunis E. Immunogenetics of atopic diseases. J Allergy Clin Immunol 1980;65:403–405. Bochner BS, Klunk DA, Sterbinsky SA, Coffman RL, Schleimer RP. IL-13 selectively induces vascular cell adhesion molecule-1 expression in human endothelial cells. J Immunol 1995;154:799–803. Chan SC, Hanifin JM. Immunopharmacologic aspects of atopic dermatitis. Clin Rev Allergy 1993;11:523–542. Chandrasekharappa SC, Rebelsky MS, Firak TA, LeBeau MM, Westbrook CA. A long-range restriction map of the interleukin-4 and interleukin5 linkage group on chromosome 5. Genomics 1990;6:94–99. Collee JM, de Vries HG, Gerritsen J. Allele sharing on chromosome 11q13 in sibs with asthma. Lancet 1993;ii:936. Cooke RA, van der Veer A. Human sensitization. J Immunol 1916;1:201–305. Cookson WOCM, Hopkin JM. Dominant inheritance of atopic immunoglobulin E responsiveness. Lancet 1988;1:86–88. Cookson W, Sharp PA, Faux JA, Hopkin JM. Linkage between immunoglobulin E responses underlying asthma and rhinitis and chromosome 11q. Lancet: 1989;1:1292–1295. Cooper KD. New therapeutic approaches in atopic dermatitis. Clin Rev Allergy 1993;11:543–560. Edfors-Lubs MI. Allergy in 7000 twin pairs. Acta Allergol (Kbh) 1971;26:207–219. Falliers CI, de A Cardoso RR, Bane HN, et al. Discordant allergic manifestations in monozygotic twins: genetic identify versus clinical, physiologic, and biochemical differences. Allergy 1971;47:207–219. Hamid Q, Boguniewicz M, Leung DYM. Differential in situ cytokine gene expression in acute vs. chronic atopic dermatitis. J Clin Invest 1994;94:870–876. Hanifin JM, Rajka G. Diagnostic features of atopic dermatitis. Acta Derm Venereol 1980;92:44–47.

Hill LW, Sulzberger MB. Yearbook of Dermatology and Syphilology. Chicago: Year Book Medical Publisher, 1933; pp. 1–70. Hopp RJ, Bewtra AK, Watt GD, Nair NM, Townley RG. Genetic analysis of allergic disease in twins. J Allergy Clin Immunol 1984;73:265–270. Jones SM, Sampson HA. The role of allergens in atopic dermatitis. Clin Rev Allergy 1993;11:471–490. Kay AM, Ying S, Varney V, et al. Messenger RNA expression of cytokine gene cluster, interleukin 3 (IL-3), IL-5, and granulocyte/macrophage colony-stimulating factor, in allergen-induced late-phase cutaneous reactions in atopic subjects. J Exp Med 1991;173:775–778. Konig P, Godfrey S. Prevalence of exercise-induced bronchial liability in families of children with asthma. Arch Dis Child 1973;48:513–518. Lacour M, Hauser C. The role of microorganisms in atopic dermatitis. Clin Rev Allergy 1993;11:491–522. Lee SK, Metrakos JD, Tanaka KR, Heiner DC. Genetic influence on serum IgE levels. Pediatr Res 1980;14:60–63. Leung DYM, Geha RS. Clinical and immunologic aspects of the hyper IgE syndrome. In: Hematology Oncology Clinics of North America. Philadelphia: Saunders, 1988; pp. 81–100. Leung DYM, Travers JB, Norris DA. Superantigens in skin disease. J Invest Dermatol 1995;105:37S–42S. Leung DYM, Harbeck R, Bina P, Hanifin JM, Reiser RF, Sampson HA. Presence of IgE antibodies to staphylococcal exotoxins on the skin of patients with atopic dermatitis: evidence for a new group of allergens. J Clin Invest 1993;92:1374–1380. Leung DYM. Atopic dermatitis: the skin as a window into the pathogenesis of chronic allergic diseases. J Allergy Clin Immunol 1995;96:302–319. Leung DYM, Rhodes AR, Geha RS, Schneider L, Ring J. Atopic dermatitis. In: Fitzpatrick TB, Eisen AZ, Wolff K, Freeberg IM, Austen KF, eds. Dermatology in General Medicine. New York: McGraw-Hill, 1993; pp. 1543–1564. Longo G, Strinati R, Puli F, Fumi F. Genetic factors in nonspecific bronchial hyperreactivity. Am J Dis Child 1987;141:331–334. Lympany P, Welsh K, MacCochrane G, Kemeny DM, Lee TH. Genetic analysis using DNA polymorphism of the linkage between chromosome 11q13 and atopy and bronchial hyperresponsiveness to methacholine. J Allergy Clin Immunol 1992;89:619–628. Marsh D, Hsu SH, Roebber M, et al. HLA-Dw2—a genetic marker for human immune response to short ragweed allergen RAS: I. Response resulting primarily from natural antigenic exposure. J Exp Med 1982;155:1439–1451. Marsh DG, Neely JD, Breazeale DR, et al. Linkage analysis of IL-4 and other chromosome 5q31.1 markers and total serum immunoglobulin E concentrations. Science 1994;264:1152–1156. May CE. Objective clinical laboratory studies of immediate hypersensitivity reactions to foods in asthmatic children. J Allergy Clin Immunol 1976;58:500–515. Meyers D, Marsh D. Report on a national institute of allergy and infectious diseases sponsored workshop on the genetics of total immunoglobulin E levels in humans. J Allergy Clin Immunol 1981;67: 167–170. Moffatt MF, Sharp PA, Faux JA, Young RP, Cookson WOCM, Hopkin JM. Factors confounding genetic linkage between atopy and chromosome 11q. Clin Exp Allergy 1992;22:1046–1051. Mosmann TR, Cherwinski H, Bond MW, Giedlin MH, Coffman R. Two types of murine helper T cell clones. I. Definition according to profiles of lymphokine activities and secretory proteins. J Immunol 1986;136: 2348–2357. Mudde GC, Van Reijsen FC, Boland GJ, DeGast GC, Bruijnzeel PLB, Bruijnzeel-Koomen CAFM. Allergen presentation by epidermal Langerhans cells from patients with atopic dermatitis is mediated by IgE. Immunology 1990;69:335–341. Parks T, Felix K, Rice T, Subbarao PV, Marimuthu KM, Rao DC. A genetic study of immunoglobulin E and atopic disease based on families ascertained through asthmatic children. Hum Hered 1990;40:69–76. Patore S, Fanales-Belasio E, Albanesi C, Chinni LM, Giannetti A, Girolomoni G. Granulocyte macrophage colony-stimulating factor is overproduced by keratinocytes in atopic dermatitis. Implications for sustained dendritic cell activation in the skin. J Clin Invest 1997;99: 3009–3017.

CHAPTER 87 / ATOPIC DERMATITIS AND ATOPY

Postma DS, Bleecker ER, Amelung PJ, et al. Genetic susceptibility to asthma- bronchial hyperresponsiveness coinherited with a major gene for atopy. N Engl J Med 1995;333:894–900. Rich SS, Roitman-Johnson B, Greenberg B, Roberts S, Blumenthal MN. Genetic analysis of atopy in three large kindreds; no evidence of linkage to D11S97. Clin Exp Allergy 1992;22:1070–1076. Schnyder UW. Neurodermititis—Asthma—Rhintis. Eine genetischallergologisch studie. Acta Genet Stat Med 1960;10(Suppl 18):1–106. Schultz Larsen F, Holm NV, Henningsen K. Atopic dermatitis. A geneticepidemiologic study in a population-based twin sample. J Am Acad Dermatol 1986;15:487–494. Shirakawa TS, Li A, Dubowitz M, et al. Association between atopy and variants of the β-subunit of the high affinity Immunoglobulin E receptor. Nat Genet 1994;7:124–129. Shirakawa TS, Morimoto K, Hashimoto T, Furuyama J. Linkage between severe atopy and chromosome 11q in Japanese families. Clin Genet 1994;46:228–232.

809

Sibbald B. Extrinsic and intrinsic asthma: influence of classification on family history of asthma and allergic disease. Clin Allergy 1980;10: 313–318. Sistonen P, Johnsson V, Koskenvuo M, Aho K. Serum IgE levels in twins. Hum Hered 1980;30:155–158. Townley RG, Bewtra AK, Nair NM, Brodkey FD, Watt GD, Burke KM. Methacholine inhalation challenge studies. J Allergy Clin Immunol 1979;64:569–574. Tsicopoulous A, Hamid Q, Varney V, et al. Preferential messenger RNA expression of Th1-type cells (IFN-γ+, IL-2+) in classical delayedtype (tuberculin) hypersensitivity reactions in human skin. J Immunol 1992;148:2058–2061. Vercelli D, Geha RS. Regulation of IgE synthesis: from membrane to the genes. Springer Semin Immunopathol 1993;15:5–16. Vercelli D, Jabara HH, Lee B, Woodland N, Geha RS, Leung DYM. Human recombinant interleukin-4 induces FcεR2/CD23 on normal human monocytes. J Exp Med 1988;167:1406–1416.

CHAPTER 88 / PEMPIGUS FOLIACEUS AND VULGARIS

88

811

Pemphigus Foliaceus and Pemphigus Vulgaris JANET A. FAIRLEY, XIANG DING, GEORGE J. GIUDICE, AND LUIS A. DIAZ

BACKGROUND Pemphigus is a group of autoimmune disorders characterized by spontaneous intraepidermal blisters and epidermal-specific autoantibodies. Blister formation occurs by the histologically distinctive loss of epidermal cell–cell adhesion, termed acantholysis. The two major forms of the disease are pemphigus vulgaris (PV) and pemphigus foliaceus (PF). In PV, the epidermal-cell separation occurs just above the basal layer of the epidermis, whereas, in PF, the cell separation occurs at the more superficial level of the granular-cell layer of the epidermis. Although PV and PF are diseases described all over the world, there is an endemic form of PF that has been observed in Brazil since the turn of the century and recently in Colombia. This endemic form of PF is also known as “fogo selvagem” (FS; wild fire). Prior to the advent of immunosuppressive therapy, the mortality caused by these disorders was extremely high. The histological hallmark of PV and PF, epidermal-cell detachment, has been known since the early 1940s. However it was not until the mid-1960s that a new insight into the pathogenesis of these disorders was gained with the discovery of PV and PF antiepidermal autoantibodies. The following years confirmed the diagnostic value of these autoantibodies and also brought convincing experimental data linking the epidermal-cell detachment seen in these patients with binding of antiepidermal autoantibodies to their target antigens. The true autoimmune nature of PV and PF was tested experimentally, addressing both the autoantibody side as well as the autoantigen side of the reaction. The autoantibody titers in patients were found to correlate well with disease activity and currently are used to follow the therapeutic response of these patients. It was also documented that neonates born to PV mothers presented a blistering disease probably caused by transplacental passage of maternal autoantibodies. These clinical observations suggested a pathogenic role for PV autoantibodies. Experimentally, it was found that the IgG fractions from PV and PF sera promoted cell detachment in primary epidermal-cell cultures and in organ cultures of human skin. In the early 1980s, other studies convincingly demonstrated that PV and PF autoantibodies were indeed pathogenic. In these experiments intraperitoneal injections of the IgG fraction of PV and PF sera were shown to induce From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

intraepidermal blisters in neonatal mice. The injected animals duplicated the key clinical and histological features of the human diseases. The search for epidermal antigens recognized by PV and PF autoantibodies lead to the characterization of those molecules as components of the desmosome. It had previously been established that desmosomes contain transmembrane glycoproteins and plaque proteins. The desmosomal glycoproteins were characterized as desmoglein 1 (Dsg1, 165 kDa), desmoglein 2 (identified in colonic epithelium), desmoglein 3 (Dsg3, 130 kDa mol wt), and the desmocollins. Furthermore, it is currently accepted that these glycoproteins are members of the cadherin family of cell adhesion molecules. PV and PF autoantibodies recognized calcium-dependent conformational epitopes on the extracellular domain of Dsg3 and Dsg1, respectively (Fig. 88-1). Finally, PF, and less commonly PV, may also be triggered by medications, most commonly penicillamine. However, captopril, penicillin, ampicillin, and rifampin have also been associated with pemphigus-like eruptions. A paraneoplastic form of pemphigus has been recently reported associated in a great number of cases with underlying lymphoproliferative disorders. The autoantibody response in these patients is complex and directed against desmosomal and hemidesmosomal antigens.

CLINICAL FEATURES PV is the most common form of pemphigus in the United States and is characterized by the spontaneous development of blisters arising on normal appearing skin (Fig. 88-2A). The lesions are typically flaccid bullae that rupture easily, producing denuded areas of skin and erosions that are slow to heal. The disease usually begins in the oral mucosa, where, because of the fragility of the blisters, the predominant lesions are erosions. Though the oral mucosa is the most common site of involvement, the disease may affect other squamous epithelium including the esophagus, conjunctiva, larynx, vagina, urethra, cervix, and rectum. The disease may remain localized to the mucosae, but generally progresses over a period of months to involve the glabrous skin. The onset of disease is most commonly between the fourth and sixth decade of life, however any age group may be affected. PF has a different clinical appearance than PV, exhibiting superficial vesicles that rupture easily, producing erosions and crusting (Fig. 88-3A). The skin lesions in PF appear most commonly on the head and neck regions. In chronic cases, these excori-

811

812

SECTION IX / DERMATOLOGY

Figure 88-1 Schematic of Dsg1 and Dsg3. Dsg1 and Dsg3 are both transmembrane proteins with an intracellular, extracellular domain, and a single transmembrane domain. The extracellular domains of both Dsg1 and Dsg3 contain four cadherins repeats (EC1-4). Within the cadherin repeats, the black bars represent calcium-binding sites, and the RAL is the putative cell–cell recognition site.

Figure 88-2 Clinical, histologic, and immunologic features of PV. PV patients typically have flaccid blisters and erosions that can become widespread (A). Histologically, biopsies of PV show suprabasilar acantholysis, with the retention of the basal cells on the dermis (B). Patients have both tissues-bound and circulating autoantibodies directed against the cell surface of the keratinocyte. (C) The results of indirect IF performed with PV serum using monkey esophagus as a substrate.

ated lesions may become more hyperkeratotic, verrucous, or vegetating plaques. In other cases, the disease may progress to an exfoliative erythroderm. In contrast to PV, the mucous membranes are spared. The clinical, immunological and histologic features of endemic PF are identical to sporadic nonendemic PF.

DIAGNOSIS The diagnosis of pemphigus is based on clinical, histologic, and immunologic criteria. Biopsies from the lesional skin of patients with PV reveals intraepidermal vesicles occurring just above the basal-cell layer. Basal keratinocytes remain attached to

CHAPTER 88 / PEMPIGUS FOLIACEUS AND VULGARIS

813

Figure 88-3 Clinical, histologic, and immunologic features of PF. This patient with endemic PF has widespread erosions and crusting on the head and neck (A). A skin biopsy specimen shows an intraepidermal vesicle occurring in the granular-cell layer (B). Indirect IF of this patient’s serum shows ICS staining, using human skin as a substrate (C).

the dermis, producing the histological “tombstone” sign (Fig. 88-2B). The dermis may show edema and inflammation with the presence of eosinophils in early lesions, and plasma cells in older lesions. The histology of PF also shows intraepidermal vesicles, but occurring high in the epidermis, within or just below the granular-cell layer (Fig. 88-3B). Older lesions of PF may also show hyperkeratosis and thickening of the epidermis. Immunofluorescence (IF) studies are extremely helpful in confirming the diagnosis of pemphigus. Direct IF, performed on perilesional biopsies shows typical deposition of IgG and/or C3 in the epidermal intercellular spaces (ICS) in 100% of patients with active disease. Indirect IF studies of patients’ sera demonstrate circulating autoantibodies that produce an ICS staining pattern on stratified squamous epithelial substrates. Indirect IF tests yield positive results in 80–90% of patients with active disease. Though occasionally PF autoantibodies may bind preferentially to the upper layers of the epidermis, and PV autoantibodies more suprabasilarly, in general the two disorders cannot be differentiated on the basis of indirect or direct IF (Figs. 88-2C and 88-3C). In the majority of patients, the serum autoantibody titers correlate well with extent and disease activity. Recent advances in the characterization of the antigens recognized by PV and PF autoantibodies have led to the application of

other techniques to precisely diagnose these disorders. Immunoblotting and immunoprecipitation have defined the PV antigen as Dsg3, a glycoprotein of 130 kDa and the PF antigen as a 160-kDa glycoprotein, Dsg1. Specialized laboratories can perform immunoblotting and immunoprecipitation (or ELISA) techniques to test patients’ sera. A positive result by immunoprecipitation is seen only in PF and PV sera if assayed with recombinant antigens (Dsg3 or Dsg1).

GENETIC BASIS OF DISEASE Like many other human autoimmune diseases, PV and PF have been associated with certain HLA specificities. Since there is a high frequency of PV cases in the Ashkenazi Jewish background population, the HLA associations have been examined in two groups: Jewish and non-Jewish PV patients. The Jewish group shares the (HLA-B38, SC21, DR4, DQw8) and (HLA-B35, SC31, DR4, Dqw8) extended haplotype, whereas the non-Jewish group share the (HLA Bw55, Drw6, SB45, Dqw5) extended haplotype. These susceptibility alleles are thought to be inherited as a dominant trait in the Jewish group of PV patients. HLA studies have demonstrated that FS is associated with a high frequency of HLA DR1 and/or HLA D4 genes. Moreover, FS patients appear to possess the DRB1*0102 gene as a susceptibility factor (relative risk

814

SECTION IX / DERMATOLOGY

[RR] = 7.3, p = 0.002) and DQB1*0201 as a gene that confers resistance to FS (RR = 0.04, p = 0.006). Recently it has been shown that these alleles are not present in Brazilian Indians affected by FS. Instead, 6 out of 10 Indians of the Xavante tribe with FS shared the HLA DRB1*0404 gene that was present in only 5 out of 74 of the Xavante controls, conferring a RR of 9.6 (p = 0.002). These findings have been confirmed recently in other Brazilian Indian tribes. It is interesting to note that the amino acid sequence of residues 67–74 of the third hypervariable region of the DRB1 gene, i.e., LLEQRRAA, is shared by DRB1*0102 (susceptibility gene of non-Indian FS patients) and DRB1*0404, DRB1*1402, and/or DRB1*1406 found with increased frequency in FS patients occurring in Brazilian Indians.

MOLECULAR PATHOPHYSIOLOGY OF DISEASE Following the experimental demonstration that PV and PF autoantibodies are pathogenic in the mouse model system (Figs. 88-4 and 88-5), efforts were concentrated in dissecting the molecular mechanisms leading to acantholysis. It has been wellestablished that PV and PF autoantibodies induce epidermal-cell detachment in the mouse model independent of complement activation, i.e., C5-deficient strains of mice and neonatal mice depleted of complement by cobra-venom treatment continue to develop extensive blistering disease when passively transfused with PF or PV IgG fractions. Furthermore, in a series of studies we have demonstrated that F(ab’)2 and Fab fragments from PF and PV IgG are also pathogenic, suggesting that these autoantibodies may trigger cell detachment by simple binding to their target epitopes on the ectodomain of Dsg1 and Dsg3, respectively. This binding may impair the adhesive function of these molecules. Other potential mechanisms involved in pemphigus acantholysis have been suggested. For example, it has been reported that activation of keratinocyte plasminogen activator (PA) by PV and PF autoantibodies may trigger epidermal-cell detachment. This hypothesis was reinforced by detecting elevated values of PA in epidermalcell cultures treated with PV or PF IgG. This elevation was abolished by previous treatment of the cultures with dexamethasone, which also abolishes the pathogenic effect of these antibodies. However, this finding could not be reproduced in vivo using the mouse model system. Recently other hypotheses have been put forward, i.e., that PV and PF autoantibodies may bind the keratinocyte surface domain of Dsg1 or Dsg3 and trigger intracellular responses leading to collapse of the cytoskeleton and impairment of cell adhesiveness. These theories need further experimental testing. Mapping the epitopes recognized by pathogenic PV and PF autoantibodies on Dsg3 and Dsg1 has been hampered by the conformational nature of these regions of the molecules. Furthermore, it appears that the calcium-binding domains of the extracellular regions of Dsg1 and Dsg3 also modulate the optimal recognition and binding of autoantibodies. However the availability of the cDNAs of both desmogleins have opened new approaches to further analyze these conformational epitopes. For example, in recent studies the cDNAs of Dsg1 and Dsg3 have been expressed in the baculovirus system as soluble glycoproteins that are highly immunoreactive with patients’ sera. Autoantibodies to Dsg3 have been purified by affinity chromatography from patients’ sera utilizing full-length recombinant Dsg3 and shown to be pathogenic in the mouse system. Moreover, preadsorption of patients’ sera with recombinant Dsg3 will abrogate the pathogenic effect in the passive-transfer experiments. These studies prove convincingly that

Figure 88-4 Passive-transfer mouse model of PV. Neonatal BALB/ c mice injected with purified IgG from PV patient’s sera reproduced the clinical (A), histologic (B), and immunologic features of the disease.

anti-Dsg3 autoantibodies in PV sera are indeed pathogenic and responsible for inducing the lesions in PV patients. Similarly, Dsg1 recombinant protein produced in the baculovirus system has been used to immunoadsorb PF autoantibodies. Like in the PV cases, it has been demonstrated that anti-Dsg1 autoantibodies are pathogenic and responsible for inducing the disease in patients. Both Dsg1 and Dsg3 share common features with the cadherin family of cell adhesion molecules. Dsg1 and Dsg3 are both transmembrane proteins that have an N-terminal extracellular domain and an intracellular C-terminal domain. The extracellular region contains a series of five repeat domains, termed EC1 to EC5 (Fig. 88-1). The first four of these domains are homologous cadherin repeats. A series of three pairs of calcium-binding sites are also located within these first four extracellular domains. These calcium-binding sites have been hypothesized to be crucial in cell– cell adhesion, and also appear to be important in autoantibody

CHAPTER 88 / PEMPIGUS FOLIACEUS AND VULGARIS

815

Figure 88-5 Passive-transfer mouse model of PF. Injection of purified IgG from the serum of a PF patient produces superficial erosions in neonatal BALB/c mice (A). A biopsy shows acantholysis in the granular-cell layer of the epidermis (B).

recognition. A three-amino acid sequence near the amino terminus (RAL) is the putative cell–cell recognition site. The region just outside the transmembrane domain (EC5) shows the greatest divergence between Dsg1 and Dsg3, as a portion of this region is deleted in Dsg1. The importance of the intracellular domain is its function in integrating the desmosomal proteins with the cytoskeleton. Within desmosomes, the 93-kDa desmosomal plaque protein, plakoglobin, is believed to be crucial in linking desmosomes with the cytoskeleton. Plakoglobin is part of both the PV and PF complex, and is believed to directly bind to both Dsg1 and Dsg3. The genes for both Dsg1 and Dsg3 are located on chromosome 18. In situ hybridization and immunofluorescence have been used to localize Dsg1 and Dsg3 in the epidermis. Dsg3 is expressed more strongly in the basal and lower spinous layers of the epidermis, whereas Dsg1 was expressed throughout the epidermis, but appeared to gradually increase during differentiation. This difference in the site of expression of Dsg1 and Dsg3 in the epidermis may help to explain the level at which blistering occurs in PF in comparison to PV.

MANAGEMENT/TREATMENT Despite the tremendous advances in understanding the molecular pathogenesis of PV and PF, the mainstay of treatment remains corticosteroids and immunosuppressive agents. Prior to immuno-

suppressive therapy, the 5-year mortality of PV approached 100%. The prognosis in PF was slightly better, with a 40–60% mortality rate. In recent years, despite the effective steroid and immunosuppressive therapy, the 5-year mortality remains between 5 and 15%. In mild cases, ultrapotent topical or intralesional steroids may be of value; however, the vast majority of cases will require systemic glucocorticoids. Prednisone is the most commonly used steroid, with an initial dose of approximately 1 mg/kg/d. If no response is seen, this may be increased to 1.5–2 mg/kg/d, not to exceed 100– 120 mg/d in our experience. A slow tapering is required to prevent relapse of the disease. Cytotoxic agents, including azathioprine, cyclophosphamide, and methotrexate may be used as the sole therapeutic agents in patients in whom steroids are contraindicated, but more commonly are used in conjunction with prednisone as steroidsparing agents. The dose of azathioprine and cyclophosphamide is 1–2 mg/kg/d, and 5–15 mg/wk for methotrexate. Plasmapheresis may be used as a adjunct therapy in patients with refractory disease to cause an acute drop in circulating antibody titers. However, a rebound rise in autoantibody titers will occur if the patient in not concurrently on adequate immunosuppressive therapy. Parenteral gold has been successfully used in the therapy of PV, but because of the incidence of side effects over long-term therapy (dermatitis, nephrotoxicity, blood dyscrasias), its use has been limited. Other drugs occasionally reported to be

816

SECTION IX / DERMATOLOGY

effective in pemphigus include cyclosporine, etretinate, and dapsone. Extracorporeal photopheresis has also been recently reported as an alternative therapy for pemphigus, although the results are too preliminary to currently recommend it as a treatment. Patients undergoing steroid and immunosuppressive therapy must be monitored with a variety of laboratory tests including CBC, liver enzymes, creatinine, urinalysis, and X-rays looking for toxicity, underlying infections, or evidence of osteoporosis. In addition, indirect IF studies of the patient’s serum every 6–8 weeks may provide valuable information in monitoring the effectiveness of the patient’s therapy.

FUTURE DIRECTIONS The past 10 years have shown remarkable progress in understanding the pathogenicity of the autoantibodies in PF and PV. Furthermore, the advances in desmosomal biology have furthered our understanding in how antidesmosomal autoantibodies may induce cell detachment in these patients. Furthermore, the availability of full-length recombinant protein for Dsg1 and Dsg3 will allow the dissection of epitopes recognized on these molecules not only by pathogenic autoantibodies by also by T cells from patients. It is expected that within the next few years the cellular mechanisms of autoantibody production in PV and PF will be disclosed. These relevant pathogenic epitopes on Dsg3 and Dsg1 should pave the way for more specific therapies, including induction of immunological tolerance to these molecules. Newer drugs may also be available to modulate the immunopathological events happening at the level of the target cells. Finally the endemic form of PF seen in Brazil might offer the unique opportunity of identifying the etiological agent of PF. This finding could have a major impact in understanding human autoimmunity.

ACKNOWLEDGMENTS This work was supported in part by grants R37 AR32081, R01 AR32599, R01 AR40410, and T32 AR07577 from the National Institutes of Health and by the Department of Veterans Affairs.

SELECTED REFERENCES Amagai M, Klaus-Kovtun V, Stanley JR. Autoantibodies against a novel epithelial cadherin in pemphigus vulgaris, a disease of cell adhesion. Cell 1991;67:869–877. Amagai M, Hashimoto T, Green KJ, Shimizu S, Nishikawa T. Antigenspecific immunoabsorption of pathogenic autoantibodies in pemphigus foliaceous. J Invest Dermatol 1995;104:895–901. Amagai M, Hashimoto T, Shimizu N, Nishikawa T. Absorption of pathogenic autoantibodies by the extracellular domain of pemphigus vulgaris antigen (Dsg3) produced by baculovirus. J Clin Invest 1994;94:59–67.

Ahmed AR, Yunis EY, Khatri K, et al. Major histocompatibility complex haplotypes studies in Ashkenazi Jewish patients with pemphigus vulgaris. Proc Natl Acad Sci USA 1990;87:7658–7662. Anhalt GJ, Labib R, Voorhees JJ, et al. Induction of pemphigus in neonatal mice by passive transfer of IgG from patients with the disease. N Engl J Med 1982;306:1189–1196. Anhalt GJ, Kim SC, Stanley JR, et al. Paraneoplastic pemphigus: an autoimmune mucocutaneous disease associated with neoplasia. N Engl J Med 1990;323:1729–1735. Beutner EH, Jordon RE. Demonstration of skin antibodies in the sera of pemphigus vulgaris patients by indirect immunofluorescent staining. Proc Soc Exp Bio Med 1964;117:505–510. Cerna M, Fernandez-Vina M, Friedman H, et al. Genetic markers for susceptibility to endemic Brazilian pemphigus foliaceus (fogo selvagem) in Xavante Indians. Tissue Antigens 1993;42:138–140. Diaz LA, Sampaio SAP, Rivitti EA, et al. Endemic pemphigus foliaceous (fogo selvagem): II. Current and historical epidemiological studies. J Invest Dermatol 1989;92:4–12. Fine J-D. Drug therapy: management of acquired bullous skin disease. N Engl J Med 1995;333:1475–1484. Koch PJ, Mahoney MG, Ishikawa H, et al. Targeted disruption of the pemphigus vulgaris antigen (desmoglein 3) gene in mice causes loss of keratinocyte cell adhesion with a phenotype similar to pemphigus vulgaris. J Cell Biol 1997;137:1091–1102. Korman NJ, Eyre RW, Klaus-Kovtun V, et al. Demonstration of an adhering-junction molecule (Plakoglobin) in the autoantigens of pemphigus foliaceus and pemphigus vulgaris. N Engl JMed 1989;321. 631–635. Lever WF. Pemphigus vulgaris. In: Lever WF, ed. Pemphigus and Pemphigoid. Springfield, IL: Charles C. Thomas, 1965; pp. 15–72. Lin MS, Mascaro JM Jr, Liu Z, Espana A, Diaz LA. The desmosome and hemidesmosome in cutaneous autoimmunity. Clin Exp Immunol 1997;107:9–15. Moraes JR, Moraes ME, Fernandez-Vina M, et al. HLA antigen and the risk for the development off pemphigus foliaceus (fogo selvagem) in endemic areas of Brazil. Immunogenetics 1991;33:388–391. Morioka S, Lazarus G, Jensen P. Involvement of urokinase-type plasminogen activator in acantholysis induced by pemphigus IgG. J Invest Dermatol 1987;89:474–477. Rock B, Labib RS, Diaz LA. Monovalent Fab’ immunoglobulin fragments from endemic pemphigus foliaceus autoantibodies are pathogenic to BALB/c mice by passive transfer. J Clin Invest 1990; 85:296–299. Roscoe JT, Diaz LA, Sampaio SAP, et al. Brazilian pemphigus foliaceus autoantibodies are pathogenic to BALB/c mice by passive transfer. J Invest Dermatol 1985;85:538–541. Takiechi M. Cadherins: a molecular family essential for selective cell-cell adhesion and animal morphogenesis. Trends Genet 1987; 3:213–217. Zhou S, Ferguson DJ, Allen J, Wojnarowska F. The location of binding sites of pemphigus vulgaris and pemphigus foliaceus autoantibodies: a post-embedding immunoelectron microscopic study. Br J Dermatol 1997;136:878–883.

CHAPTER 89 / PEMPHIGOID DISEASES

89

817

Bullous Pemphigoid, Cicatricial Pemphigoid, and Pemphigoid Gestationis GRANT J. ANHALT AND DIYA F. MUTASIM

BACKGROUND The term “pemphigoid” applies to several different clinical disorders that share the clinical appearance of blisters, the histologic finding of subepidermal blister, and circulating and skin-bound IgG antibodies against a component of the basement membrane zone (BMZ) (Table 89-1). Extensive skin involvement in the elderly population occasionally leads to significant morbidity. In addition, therapy with systemic corticosteroids and immunosuppressive agents may be associated with significant adverse effects.

CLINICAL FEATURES The exact incidence of pemphigoid is not known, but it is estimated to be twice that of pemphigus vulgaris (PV). Its incidence in Great Britain is estimated at 1 per 100,000 per year in 1985 and 7.63 per million in France in 1995. There are three closely related pemphigoid diseases, namely bullous pemphigoid (BP), cicatricial (mucosal) pemphigoid (CP), and pemphigoid (herpes) gestationis (HG). BULLOUS PEMPHIGOID This disease appears to be a phenomenon of the aging immune system, with peak incidence very late in life (Figs. 89-1 and 89-2). There is no known HLA Class II gene that confers susceptibility for the disease. However, in herpes gestationis, there appears to be an HLA Class II-linked susceptibility. Approximately 60–85% of HG patients are HLA-DR3 positive. More strikingly, approximately 45% of women affected with the disease are heterozygote HLA-DR3/DR4. The disease tends to be generalized with predilection for the lower abdomen, inner thighs, groin, and axillae. The characteristic blisters are pruritic and appear on either normal-appearing or erythematous, urticarial base. When blisters rupture, denuded areas may become crusted and occasionally invaded by bacterial pathogens. The course and prognosis of BP varies. Most treated patients go into a clinical remission within several months without need for further treatment. CICATRICIAL PEMPHIGOID This is a relatively rare condition characterized by chronic blistering and scarring of mucosal surfaces covered by stratified squamous epithelium. Involved From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

organs include the ocular conjunctiva, nasal pharynx, oral pharynx, larynx, upper esophagus, and anogenital mucosa. The marked tendency for scarring in this disorder leads to high morbidity secondary to loss of vision, esophageal strictures, and fibrotic scarring of other mucous membranes. PEMPHIGOID GESTATIONIS Also known as herpes gestationis (HG), this is a relatively rare disease seen during pregnancy. The disorder often begins in the second or third trimester of pregnancy. It tends to recur in subsequent pregnancies and may be elicited by challenge with oral contraceptives. Lesions tend to involve the abdomen and then spread peripherally. The clinical appearance of the lesions is similar to that of BP.

DIAGNOSIS HISTOPATHOLOGY Histopathologic examination of an early intact blister reveals a subepidermal blister with a mixed inflammatory infiltrate that is rich in eosinophils. The blister cavity contains serum, fibrin, and inflammatory cells. DIRECT IMMUNOFLUORESCENCE The ideal substrate for this test is a biopsy of normal-appearing skin adjacent to a lesion (Fig. 89-3). A properly performed test from such a specimen is positive in 100% of BP patients. There is continuous linear deposition of C3 and IgG along the epidermal BMZ. Direct IF in HG and CP is similar to that of BP. INDIRECT IMMUNOFLUORESCENCE The most commonly used substrate for this test is monkey esophagus. Other stratified squamous epithelia including normal human skin may also be used. Most patients with BP have circulating IgG autoantibodies against the BMZ. These antibodies bind to basement membrane molecules of stratified squamous epithelia. Indirect IF is usually negative in most patients with HG by the standard methods. The anti-BMZ antibodies can, however, be detected by complement-fixation amplification (HG factor assay) or Western blotting. Approximately 25% of patients with CP have circulating anti-BMZ antibodies as detected by standard indirect IF. IMMUNOELECTROMICROSCOPY (IEM) Direct IEM reveals the in vivo bound immune deposits (C3 and IgG) to be within the lamina lucida of the BMZ as well as basal-cell hemidesmosomes. Indirect IEM reveals that the circulating BP antibodies are directed against both the basal-cell hemidesmosomes as well as the lamina lucida. The findings in HG and CP are similar.

817

818

SECTION IX / DERMATOLOGY

Table 89-1 The Pemphigoid Disorders Disorder Distribution of lesions Tendency for scarring Subepithelial blister IgG and C3 at the BMZ IgA at the BMZ Serum anti-BMZ antibodies detected by: Indirect Immunofluorescence Complement fixation Western blotting or immunoprecipitation Target antigen in BMZ Location Molecule

Bullous pemphigoid

Cicatricial pemphigoid

Pemphigoid gestationis

Flexures, neck, abdomen – + + Rare

Oral, ocular, anogenital + + + Occasional

Abdomen, extremities – + + –

70% 80% 95%

25% ? 80%

20% 50% 80%

Hemidesmosome and lamina lucida BP 180, BP 230

Hemidesmosome, lamina lucida and lamina densa BP 180, BP 230, laminin 5, other

Hemidesmosome and lamina lucida BP 180, rare BP 230

BMZ, basement membrane zone; BP 180, bullous pemphigoid antigen, mol wt 180 kDa; BP 230, bullous pemphigoid antigen, mol wt 230 kDa.

Figure 89-2 This is a close-up of the lesions revealing urticarial plaques and bullae.

Figure 89-1 This is an elderly individual with generalized bullous pemphigoid.

PATHOPHYSIOLOGY THE BULLOUS PEMPHIGOID ANTIGENS It is established that there are two protein antigens recognized by autoantibodies in bullous pemphigoid. Recently it has also been established

by use of a passive-transfer animal model that that interaction of these autoantibodies with one of these proteins in vivo is directly responsible for blister formation. The first pemphigoid antigen to be cloned and sequenced has been called either the BPAg1 or the BP230 antigen, and has an estimated molecular weight of 230 kDa in sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE). The hemidesmosome is an organelle that mediates a stable interaction between the keratin intermediate filament proteins of the basilar pole of the keratinocyte and transmembrane adhesion molecules that anchor the basal cell to the underlying lamina densa of the basement membrane. The BPAg1 is the major protein of the cytoplasmic plaque of the hemidesmosome, and its presumed (but not proven) function is to mediate this interaction. It is notable that the BPAg1 has approximately 30% sequence homology with desmoplakin I, the major plaque protein of the desmosome—an organelle that serves a similar function in mediation of cell—cell adhesion. BPAg1 has been mapped to chromosome 6 (6p12-p11) in the human and to the proximal region of chromosome 1 in the mouse. The second pemphigoid antigen to be characterized is called the BPAg2 or the BP180 antigen, and has an estimated molecular

CHAPTER 89 / PEMPHIGOID DISEASES

Figure 89-3 Direct immunofluorescence of perilesional skin of patient with bullous pemphigoid. Note the continuous linear deposition of complement (C3) along the epidermal basement membrane.

weight of 180 kDa in SDS-PAGE. The BPAg2 has been mapped to chromosome 10 (10q24.3) in the human and to the distal end of chromosome 19 in the mouse. It is a unique molecule, with many important structural features. It is a transmembrane molecule with a type 2 orientation, the carboxy-terminus being in the extracellular space and the amino-terminus being intracellular. The extracellular domain has lengthy collagenous domains, characterized by the presence of Gly-X-Y repeats, and this finding has led to an additional proposed name for the molecule, collagen type XVII. This portion of the molecule is presumed to interact with adhesion molecules of the lower lamina lucida/lamina densa to anchor the basal cell in place. The hypothesis that this molecule is important in normal epithelial adhesion is supported by the recent observation of kindreds in which a mutation of the BP180 gene results in a nonlethal, but generalized form of congenital epidermolysis bullosa. Immediately adjacent to the transmembrane domain of the BPAg2 there is a noncollagenous domain that contains the immunodominant epitope of the protein. Giudice and Diaz identified a polypeptide of 14 amino acids from this epitope that is recognized by the majority of human sera in bullous pemphigoid and herpes gestationis. Subsequently, they identified the corresponding region of the murine BPAg1 and found no significant homology between the human and murine epitope. This was a critical observation that allowed these investigators to subsequently develop a passive-transfer model in the mouse to define the immunopathologic events that lead to blister formation in vivo. MECHANISMS OF BLISTER FORMATION For many years, it had been assumed that the sequence of events leading to blister formation in the human disease had been as follows: (1) Loss of tolerance to the BP antigens leads to autoantibody production and binding to the hemidesmosome. (2) Tissue-bound autoantibody activates the complement cascade. (3) Generation of chemotactic fragments of complement such as C3a and C5a recruit neutrophils and eosinophils to the site of bound autoantibody. (4) Release of proteolytic enzymes from the polymorphs degrades the BP antigens or other important adhesion molecules at the site. (5) Subepidermal blister formation ensues. This hypothesized sequence of events has now been confirmed and clearly defined in vivo by the use of an animal model. In this model, the 14-amino acid

819

polypeptide, encoding the immunodominant epitope of the murine BPAg2 was synthesized, and antibodies against it were raised in rabbits. These antibodies were then transfused into neonatal mice, and within 24 h, cutaneous inflammation and blister formation was apparent. The induced blisters showed subepidermal separation, deposition of rabbit IgG and murine C3 along the basement membrane zone, and infiltration of the base of the blister with polymorphonuclear leukocytes. Blister formation was not observed if the model was manipulated to inhibit complement activation by bound anti-BPAg2 antibody. This was demonstrated by transfusing the antibody into C5 deficient strains of mice, into Balb/c mice depleted of complement by pretreatment with cobra venom factor, or by the infusion of F(ab')2 fragments of the rabbit anti-BPAg2 antibodies into newborn mice. If the mice were depleted of circulating polymorphonuclear cells prior to antibody infusion, blister formation was also not observed. In related studies, the role of proteases released from eosinophils in the degradation of the BPAg2 was defined. Abundant message for a 92-kDa gelatinase was found in eosinophils at the base of blisters in humans with lesions of BP. In vitro studies confirmed that the eosinophil-derived gelatinase degraded the BPAg2 in the extracellular collagenous domains. In total, these studies provide compelling evidence for the proposed immunopathogenesis of lesions in BP. It is not clear what role, if any, antibodies specific for the BPAg1 play in the development of lesions. Passive transfer of human antibodies against the BPAg1 do not produce any inflammation in the skin of mice. Rabbits that are immunized to the antigen and have circulating anti-BPAg2 antibodies do not develop any cutaneous lesions. These animals do exhibit an unusual inflammatory reaction to ultraviolet radiation. In this circumstance, erythematous doses of ultraviolet light produce marked epidermal inflammation and keratinocyte necrosis. This suggests that antibodies against the BPAg1 may cause some secondary inflammatory events in vivo, but it appears that antibodies against the BPAg2 are really of primary pathogenic importance. It is not clear why patients should simultaneously make autoantibodies against two completely unrelated protein antigens, although a similar phenomenon is observed in autoimmune thyroiditis, where autoantibodies develop against discrete but unrelated antigens that are related to each other only by close physical proximity.

MANAGEMENT Pemphigoid is a disorder that results from an abnormal immune response and that has prominent inflammatory features. Therapy for pemphigoid should suppress inflammation and/or the immune response. If therapy fails, the elderly patient with extensive erosions may develop complications such as fluid loss, electrolyte imbalance, bacterial colonization with potential sepsis, scarring, and decubitus ulcers. Most patients with generalized BP require systemic therapy. The most commonly used systemic agents are the corticosteroids. Immunosuppressive drug therapy with chemotherapeutic agents should be considered for patients who require high maintenance doses of corticosteroids, patients who develop corticosteroid side effects, and patients whose disease does not respond completely to corticosteroid therapy. The most commonly used immunosuppressive agents are azathioprine, cyclophosphamide, methotrexate, and cyclosporine. Treatment of CP and HG follows the same principles as that of BP.

820

SECTION IX / DERMATOLOGY

FUTURE DIRECTIONS In the past decade, extensive information has been derived about the molecular structure of BP antigens. A recent animal model confirmed the pathogenic role of BP antibodies in the induction of skin lesions. Future research should be directed toward understanding the mechanisms for the immune response that results in the production of BP antibodies.

SELECTED REFERENCES Anhalt GJ, Bahn CF, Labib RS, et al. Pathogenic effects of bullous pemphigoid autoantibodies on rabbit corneal epithelium. J Clin Invest 1981;68:1097–1101. Anhalt GJ, Jampel HD, Patel HP, et al. Bullous pemphigoid autoantibodies are markers of corneal epithelial hemidesmosomes. Invest Ophthalmol Vis Sci 1987;28:903–907. Bedane C, McMillan JR, Balding SD, et al. Bullous pemphigoid and cicatricial pemphigoid autoantibodies react with ultrastructurally separable epitopes on the BP180 ectodomain: evidence that BP180 spans the lamina lucida. J Invest Dermatol 1997;108:901–907. Giudice GJ, Emery DJ, Diaz LA. Cloning and primary structural analysis of the bullous pemphigoid autoantigen BP 180. J Invest Dermatol 1992;99:243–250. Giudice GJ, Emery DJ, Zelickson BD, Anhalt GJ, Liu Z, Diaz LA. Bullous pemphigoid and herpes gestationis autoantibodies recognize a common non-collagenous site on the BP180 ectodomain. J Immunol 1993;151:5742–5750. Jonkman MF, deJong MC, Heeres K, et al. 180-kD bullous pemphigoid antigen (BP180) is deficient in generalized atrophic benign epidermolysis bullosa. J Clin Invest 1995;95:1345–1352.

Jordon RE, Kawana S, Fritz KA. Immunopathologic mechanisms in pemphigus and bullous pemphigoid. J Invest Dermatol 1985;85:72s–78s. Katz SI, Hertz KC, Yaoita H. Herpes gestationis: immunopathology and characterization of the HG factor. J Clin Invest 1976;57:1434–1441. Li KH, Sawamura D, Giudice GJ, et al. Genomic organization of collagenous domains and chromosomal assignment of human 180-kDa bullous pempigoid antigen-2, a novel collagen of stratified squamous epithelium. J Biol Chem 1991;266:24,064–20,469. Liu Z, Diaz LA, Troy JT, et al. A passive transfer model for the organspecific autoimmune disease, bullous pemphigoid, using antibodies generated against the hemidesmosomal antigen, BP180. J Clin Invest 1993;92:2480–2486. Liu Z, Giudice GJ, Swartz SJ, et al. The role of complement in experimental bullous pemphigoid. J Clin Invest 1995;95:1539–1544. Liu Z, Roopenian DC, Zhou X, et al. beta2-microglobulin-deficient mice are resistant to bullous pemphigoid. J Exp Med 1997;186:777–783. Mutasim DF, Diaz LA. The relevance of immunohistochemical techniques in the differentiation of subepidermal bullous diseases. Amer J Dermatopath 1991;13:77–83. Mutasim DF, Morrison LH, Takahashi Y, et al. Definition of bullous pemphigoid antibody binding to intracellular and extracellular antigen associated with hemidesmosomes. J Invest Dermatol 1989;92: 225–230. Stahle-Backdahl M, Inoue M, Giudice GJ, Parks WC. 92-kD gelatinase is produced by eosinophils at the site of blister formation in bullous pemphigoid and cleaves the extracellular domain of recombinant 180-kD bullous pemphigoid autoantigen. J Clin Invest 1994;93: 2022–2030. Stanley JR, Hawley-Nelson P, Yuspa SH, et al. Characterization of bullous pemphigoid antigen: a unique basement membrane protein of stratified squamous epithelia. Cell 1981;24:897–903.

CHAPTER 90 / CUTANEOUS LUPUS ERYTHEMATOSUS

90

821

Cutaneous Lupus Erythematosus RICHARD D. SONTHEIMER

BACKGROUND Lupus erythematosus (LE) is an extremely heterogenous autoimmune disease process that has the potential to involve every major organ system within the body. Involvement of internal organs is designated as systemic LE (SLE), whereas involvement of the skin is referred to as cutaneous LE. Some LE patients have only SLE and others only cutaneous LE, however, most will have features of both SLE and cutaneous LE. Polyclonal B-cell activation with resultant production of autoantibodies to a number of nuclear and cytoplasmic antigens is a hallmark of LE. Multiple genetic and environmental factors are thought to conspire to produce the mosaic of clinical and laboratory features that constitutes the LE disease process. The skin is a major target in LE—cutaneous disease is the second most common overall clinical disease manifestation, and skin disease is the second most frequent way that this autoimmune disorder first presents itself. Unfortunately, the molecular basis of most forms of skin involvement in LE is poorly understood owing in large part to the absence of reliable in vitro or in vivo models of this pattern of cutaneous inflammation. Thus, a discussion of this subject must be allowed some degree of freedom concerning the extrapolations that naturally flow from an analysis of a diverse set of isolated clinical, pathological, and immunopathological observations as has been reported for LE. This overview will focus especially on the immunological effector mechanisms that might be responsible for the characteristic pattern of cutaneous inflammation that is seen in this disorder. This chapter will review our current understanding of the etiology and pathogenesis of LE-specific skin disease (LESSD), the type of cutaneous involvement that is histopathologically unique to individuals having LE. Cutaneous vasculitis, a form of LE-nonspecific skin disease, can be seen in LE patients but also is seen in a number of other disease processes other than LE. Space does not allow discussion of the causal mechanisms of LE non specific skin disease. In its strictest sense, the term “cutaneous LE” might be used to refer to all forms of skin lesions that occur as the result of the underlying LE autoimmune process; however, in practice most use this term synonymously with LESSD.

CLINICAL FEATURES The classification system popularized by James N. Gilliam divided LESSD (cutaneous LE) into three subgroups—acute From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

cutaneous LE (ACLE), subacute cutaneous LE (SCLE), and chronic cutaneous LE (CCLE) (a comprehensive color atlas illustrating the many ways that LE can be expressed in the skin can be found in Sontheimer RD et al.). Similarities and differences exist in the clinical and laboratory features associated with these three forms of cutaneous LE (Table 90-1). LESSD lesions can exist as an isolated event (e.g., localized classical discoid LE [DLE], which is the most commonly encountered variety of CCLE) or can accompany underlying SLE disease activity (e.g., ACLE). However, LESSD lesions that occur in the context of SLE are indistinguishable clinically and pathologically from those that occur as an isolated event. Although it is possible that each of these three major forms of LESSD has a unique etiopathogenesis, it is much more likely that each represents a variation on a common underlying theme with host factors such as genetic differences and environmental stimuli such as exposure to ultraviolet (UV) light accounting for the differences in phenotypic expression. Several clinical observations concerning LESSD can serve as clues to the underlying pathogenesis of this pattern of cutaneous inflammation. When nonlesional skin of an LE patient is injured in any way, LESSD can be precipitated at the site of injury (i.e., the Köebner phenomenon). Autografting experiments have revealed LESSD to be recipient site dominant, i.e., grafts of nonlesional autologous skin transplanted into an area of LESSD inflammation assume the clinical and pathological features of the recipient site. In some patients, LESSD, especially SCLE, can be induced by exposure to certain drugs including hydrochlorothiazide, procainamide, D-penicillamine, sulfonylureas, oxyprenolol, griseofulvin, piroxicam, naproxen, PUVA, spironolactone, and diltiazem. Photosensitivity is a clinical hallmark of LE occurring in as many as 80–90% of certain cutaneous LE subsets such as SCLE. In addition to precipitating or exacerbating cutaneous LE activity, exposure to UV light is also capable of aggravating the underlying systemic autoimmune abnormalities. The abnormal cutaneous response to UV light in LE patients is delayed in time, usually taking days to weeks to become fully apparent. The degree of abnormal sensitivity to sunlight and artificial sources of UV light can wax and wane over an individual patient’s disease course. As previously mentioned, certain photosensitizing drugs such a hydrochlorothiazide and griseofulvin are capable of inducing some forms of LESSD such as SCLE. Some studies have indicated that circulating humoral factors may predispose to the development of cutaneous LE photosensitivity. Most LE patients respond normally to minimal erythema dose testing, however, some reports have suggested decreased minimal

821

822

SECTION IX / DERMATOLOGY

Table 90-1 Comparison of the Major Forms of LE-Specific Skin Diseasea Disease features

ACLE

SCLE

Classical DLE

0 0 + 0 + +++

0 0 ++ 0 ++ +++

+++ +++ +++ +++ +++ +

+++ + + ++ ++

+ + + +++ +++

+++ +++ +++ ++ ++

0 + +

+ ++ +

+++ +++ +++

++ ++ +++

++ + ++

+++ 0 +

+ ++ +++ +++ +++

+++ +++ + + ++

0 + 0 + +

Clinical features of skin lesions Induration Scarring Pigment changes Follicular plugging Hyperkeratosis Photosensitivity Distribution of lesions Face Scalp Ears Extensor arms, forearms V-area of neck Histopathology Thickened basement membrane Intensity of lichenoid infiltrate Periappendageal inflammation Lupus band Lesional Nonlesional Antinuclear antibodies Ro/SS-A antibodies By immunodiffusion By ELISA Antinative DNA antibodies Hypocomplementemia Risk for developing SLE

+++, strongly associated; ++, moderately associated; +, weakly associated; 0, negative, no association. a(Adapted with permission from Table 1.4 of Sontheimer RD, et al.)

erythema doses in some individuals. In the classical phototesting studies, the action spectrum of LESSD appeared to be restricted to the ultraviolet B (UVB) spectrum; however, more recent studies have suggested that ultraviolet A (UVA) radiation can contribute to the photosensitive states experienced by some subsets of LE, especially SCLE. Single- and repeated-dose exposure of nonlesional skin to UV light can result in the production of typical cutaneous LE lesions. These UV light-induced lesions have identical histopathological features to idiopathic LE skin lesions. The earliest histopathological change in a UV light-induced skin lesion is perivascular accumulation of mononuclear cells—these changes clearly predate the appearance of immunoglobulin and complement deposits at the dermal–epidermal junction. Some photochallenge studies have suggested that UVB exposure preferentially induces the epidermal changes of LESSD, whereas UVA predominantly produces the dermal changes.

DIAGNOSIS The diagnosis of LESSD is certain when the expected histopathological changes are seen in a skin lesion that is clinically consistent with a form of cutaneous LE. Dermatomyositis skin lesions can at times be difficult to distinguish from some forms of LESSD on routine histopathology alone. Whereas qualitative differences exist between ACLE, the pattern of pathological change in each of these forms of LESSD is that of a lichenoid tissue reaction. Experimental evidence suggests that other examples of the lichenoid tissue reaction such as lichen planus and graft-vs-host

skin disease are the result of T-lymphocyte–mediated autoimmune injury within the skin.

PATHOLOGY EPIDERMAL CHANGES Epidermal keratinocyte hyperproliferation in cutaneous LE lesions is associated with normal early differentiation and premature terminal differentiation. In addition, epidermal Langerhans cells are decreased in density and display a perturbed morphology. However, the primary focus of change within the epidermis is at the basal cell layer with hydropic or liquefactive degeneration of basal keratinocytes being the most prominent change. Activated lymphocytes and macrophages can be seen admixed with these degenerating basal keratinocytes. The intimate association of activated lymphocytes with the degenerating epidermal basal cells has led to the speculation that autoantigen-specific lymphocytes might mediate keratinocyte cytotoxicity in LESSD. The presence of elevated serum levels of soluble interleukin (IL)-2 receptors in certain cutaneous LE subsets also indirectly supports a role for activated T cells in the pathogenesis of this form of LESSD. Epidermal keratinocytes in LESSD lesions express HLA class II antigen, ICAM-1, and the B7-3 costimulatory ligand. In addition, preliminary in vitro studies have suggested the existence of phenotypic variation in tumor necrosis factor (TNF)-α–induced ICAM-1 expression by epidermal keratinocytes from cutaneous LE patients. DERMAL–EPIDERMAL JUNCTION CHANGES There is thickening of the epidermal basement-membrane zone in chronic

CHAPTER 90 / CUTANEOUS LUPUS ERYTHEMATOSUS

LESSD lesions such as DLE. This appears to result from basal lamina reduplication that occurs as a result of the hyperproliferation and premature death that occurs within the epidermal basal-cell compartment. DERMAL CHANGES Mononuclear-cell infiltration around blood vessels and dermal appendages is a constant feature of LESSD. CD4 T lymphocytes that express class II histocompatibility antigens comprise the majority of this infiltrate. These cells express other T-cell–activation markers such as the IL-2 receptor and thymidine incorporation to only variable degree. Monocytes and macrophages are seen to a lesser extent and B cells are quite rare. The infiltrating T cells are predominately of the helperinducer, memory phenotype (CD4+, CD45RA–, CD45RO+) and express CD28/CTLA-4. There do not appear to be significant qualitative differences in the inflammatory infiltrate seen in DLE and SCLE lesions. Some workers have reported that γ/δ T-cell receptor-positive T cells are virtually nonexistent in the epidermis and dermis of CCLE lesions, whereas others have suggested that T cells having a specific γ/δ receptor phenotype (V-γ2/V-δ2) are preferentially expanded within the epidermal and dermal T-cell infiltrates of CCLE lesions. The presence of γ/δ T cells would suggest the possibility of an antigen-specific response to stress proteins induced within the epidermis by environmental stimuli such as UV light. The dermal microvasculature is a major focus of cellular inflammation within the dermis. Several studies have suggested that HLA-DR antigens are expressed at lower levels on dermal microvascular endothelial cells than is normal in DLE and SLE skin lesions. Dermal mucin deposition can be a prominent feature of LESSD lesions. Recent studies have suggested the possibility that circulating factors in LE patients that parallel SLE disease activity can stimulate exaggerated production of glycosaminoglycans by dermal fibroblasts. SUBCUTANEOUS CHANGES Subcutaneous inflammation is relatively rare in LE; however, lupus panniculitis (profundus) can at times be the dominant cutaneous change. The histologic pattern is a lobular lymphocytic panniculitis. Perivascular infiltration with lymphocytes, plasma cells, and histiocytes in the deep dermis and subcutaneous fat is usually seen; this can extend to the point of lymphoid nodule formation. Vessel-wall thickening and invasion by mononuclear cells (“lymphocytic vasculitis”) can also be observed. There is a distinct absence of polymorphonuclear leukocytes. Hyaline fat necrosis can be seen as well as prominent fibrinoid degeneration of collagen. Mucinous degeneration and calcification in established lesions occurs over time.

IMMUNOPATHOLOGY Immunoglobulin (IgG, IgA, IgM) and complement (C3, C4, membrane-attack complex) components can be found in a bandlike array along the dermal–epidermal junction of both lesional and nonlesional skin of LE patients. Elution of such immunoreactants from the skin of SLE patients has suggested that they contain antinuclear and anti–basement membrane zone reactivity. Ultrastructural studies have localized these deposits to structures below the lamina densa that are associated with type VII collagen (autoantibodies to type VII collagen are produced by some patients with SLE who develop vesiculobullous skin lesions [“vesiculobullous SLE”]). These deposits have been shown to follow the appearance of the perivascular mononuclear-cell infiltrate in UV-induced LE skin lesions by 4–6 weeks and thus are

823

probably not causally involved as a primary factor in the pathogenesis of LESSD. Experimental studies in mice have shown that deposition of immunoglobulin at the dermal–epidermal junction can result from charge interaction alone when cationized antibody or immune complexes are present. A pattern of IgG deposition different from the LE band has been observed to occur in SCLE patients. This dust-like array of IgG deposition is seen most prominently over the epidermal basal cells and has been suggested to correlate with the presence of circulating Ro/SS-A autoantibodies. The ability to detect this subtle immunofluorescence finding appears to vary somewhat between laboratories, suggesting that technical issues might be important to its detection. In LE panniculitis/profundus, immunoglobulin and complement deposits are usually found in blood vessel walls of the deep dermis and subcutis by direct immunofluorescence.

ULTRASTRUCTURAL CHANGES Early electron microscopic studies identified paramyxoviruslike structures within lesions of DLE; however, similar tubuloreticular structures have also been found in other settings such as dermatomyositis and Sjögren’s syndrome. Considerable evidence now suggests that these “lupus inclusions” are not directly related to viral infection since they can be reproduced in cells in vitro as a result of excessive stimulation with interferon (IFN)-α or -β.

GENETIC BASIS OF DISEASE HLA associations have been reported for several forms of LESSD. SCLE has been associated most strongly with HLA-B8, DR3, DRW6, DRW52, and DQW1/DQW2. This the HLA haplotype that is even more strongly associated with Ro/SS-A autoantibody production, the serological marker for SCLE. Nucleotide sequence analysis of DQW1/DQW2 alleles in patients who produce Ro/SS-A antibody have demonstrated that 100% have a glutamine residue at position 34 of the outermost domain in the DQA1 chain and a leucine at position 26 of the DQB1 chain. Partial or complete genetic deficiency of C2 and C4 is also associated with SCLE. CCLE (DLE) has been a found to correlate with the presence of HLA-B7, B8, Cw7, DR2, DR3, DQW1, and DQA1*0102. Like SCLE, DLE has been associated with genetic deficiency of C2, C4, and C5. There is no specific autoantibody marker for CCLE. No specific HLA associations have been reported for ACLE; however, ACLE lesions usually occur in the context of SLE and SLE has been most closely associated with HLA-DR2 and DR3. Non-HLA genetic associations have also been reported for CCLE. DLE has been observed to be present at increased frequency in female carriers of X-linked chronic granulomatous disease, which is now known to result from genetic deficiency of reduced nicotinamide dinucleotide phosphatase (NADPH). In addition, IL-1 receptor antagonist and TNF-α gene polymorphisms have been associated with SLE; however, the role that this plays in the expression of cutaneous LE has not been fully determined. Preliminary evidence has argued against an association between SCLE and TNF-α polymorphism.

MOLECULAR PATHOPHYSIOLOGY OF DISEASE The foregoing clinical and laboratory observations suggest that LESSD is the result of a genetically determined autoimmune

824

SECTION IX / DERMATOLOGY

Figure 90-1 Modulation of Ro/SS-A autoantigen in epidermal keratinocytes following exposure to UVB irradiation. Lefeber, Norris, and coworkers first reported that UVB irradiation of human epidermal keratinocytes in culture resulted in the translocation of normally intracellular Ro/SS-A autoantigen to the surface of viable. Ro/SS-A antibody from the circulation would then be able to bind to these surface-expressed autoantigens, targeting the cells for injury through immune effector mechanisms such as antibody-dependent cell-mediated cytotoxicity. Golan et al. first presented data suggesting that greater degrees of Ro/SS-A autoantibody binding could be observed on the surface of cultured epidermal keratinocytes of LE patients following UVB compared to keratinocytes from normal individuals.

response to cutaneous autoantigens that is mediated by autoantibodies, immune complexes composed of autoantibodies and autoantigens, or autoantigen-specific T cells. The tempo and extent of this form of autoimmune inflammation can be impacted by various environmental stimuli foremost among which is exposure to UV light. The photoreceptor for cutaneous LE has not been identified; however, some have argued that it might be DNA. Earlier work attempting to explain the molecular basis of photosensitivity in LE patients suggested that the pathologic immune response might be directed against neoantigens generated in the skin by the action of UV light on DNA. A murine model of some of the histopathological and immunopathological changes associated with cutaneous LE was produced by exposing the skin of mice that had been immunized to thymidine dimers to UV light. In addition, several investigators have argued that LE patients have abnormal UV lightaltered DNA-repair capacity. However, the role played by UV-induced DNA injury and repair in the pathogenesis of human LESSD remains to be determined. Most attention over the past decade has been directed to the idea that some forms of Ro/SS-A autoantibody-associated, photosensitive cutaneous LE such as SCLE and neonatal LE might result from the interaction of Ro/SS-A antibody from the circulation with Ro/SS-A antigen that has been transported to the surface of epidermal keratinocytes as a result of exposure to UVB (Fig. 90-1). The following observations support this hypothesis. Ro/SS-A autoantibodies are extremely prevalent in SCLE and there is a strong correlation between Ro/SS-A antibody levels and skin disease activity in neonatal LE. The dust-like pattern of IgG deposi-

tion seen in human SCLE lesions has been experimentally reproduced in nude mice bearing human skin explants by the passive transfer of Ro/SS-A autoantibody derived from LE patient sera. UVB-induced Ro/SS-A antigen modulation in keratinocytes has now been confirmed by several groups, and it has been suggested that this phenomenon might be enhanced in keratinocytes from LE patients compared to those from normal individuals. It has not yet been determined whether differential UVB-induced expression of the molecular constituents of Ro/SS-A ribonucleoprotein particles (Ro60, Ro52, La/SS-B, calreticulin) might occur in the epidermal keratinocytes of LE patients compared to normal individuals. UVB-induced apoptosis might be responsible for the translocation of Ro/SS-A autoantigen to the cell surface. Autoantibody binding to the exposed antigens on the surface of keratinocytes could result in tissue injury through complementmediated lysis or antibody-dependent cell-mediated cytotoxicity. Infection by alphaviruses such as sindbis also appears to be capable of inducing cell-surface expression of Ro/SS-A antigens as a result of virus-induced apoptosis. It is also possible that UV light might cause the expression of stress-related autoantigens in keratinocytes of LE patients that could then become the target of a pathological immune response. Although quite attractive, the hypothesis that UVB-induced Ro/SS-A antigen modulation is a causative factor in photosensitive LESSD still remains unproven. Several arguments can be made against this hypothesis. Most Ro/SS-A antibody-positive Sjögren’s syndrome patients do not develop SCLE or deliver infants with neonatal LE skin lesions, even though they have equally high levels of Ro/SS-A autoantibody (most studies have

CHAPTER 90 / CUTANEOUS LUPUS ERYTHEMATOSUS

Figure 90-2

825

Possible mechanisms of T-cell interaction with epidermal keratinocytes of an LE patient following exposure to UVB irradiation.

Figure 90-3 Events occurring in epidermal keratinocytes following exposure to UVB that could be relevant to the pathogenesis of LE-specific skin disease.

not found distinctive differences in the Ro/SS-A polypeptide specificities of the autoantibodies found in SCLE and Sjögren’s syndrome). In addition, from the preliminary data that is available, there does not appear to be a positive correlation between Ro/SSA antibody levels and disease activity in adults with SCLE. Also only a very small percentage of newborns exposed to maternal Ro/SS-A autoantibody production in utero develop either neonatal LE or congenital heart block. To date, no in vitro or in vivo experimental system involving UVB-induced Ro/SS-A autoantigen modulation in epidermal keratinocytes has reproduced the pattern of histopathological change that is typical of LESSD. It is conceivable that Ro/SS-A antibodies in photosensitive cutaneous LE patients represent only a marker for Ro/SS-A antigen-specific T cells and that such autoreactive T cells may represent the primary immunological effector mechanism for photosensitive LESSD (Fig. 90-2). Increasing evidence suggests

that many of the autoantibodies seen in LE are antigen driven and preliminary studies have identified ribonucleoprotein autoantigenspecific T cells from LE patients, including 52-kDa Ro/SS-A-specific T cells. As previously mentioned, the pathology of LESSD shares many features with other T-cell–mediated skin disorders such as lichen planus and graft-vs-host skin disease. However, to date, Ro/SS-A antigen-specific T cells have not been identified in LESSD lesions. If such cells exist they might be expected to be increased in density in LESSD lesions compared to the peripheral blood. In addition to modulating intracellular ribonucleoprotein autoantigens, UV light also has a number of other proinflammatory effects on epidermal keratinocytes (Fig. 90-3). UV light can upregulate proinflammatory molecules such as IL-1α, TNF-α, GM-CSF, IL-6, IL-10, IL-12, MIP-1 and 2, prostaglandin E, proteases, oxygen free radicals, and histamine. It is conceivable that genetically determined enhanced expression or defective meta-

826

SECTION IX / DERMATOLOGY

bolic turnover of one or more of these products might occur in LE patients; however, there is currently little data that directly addresses this possibility. As previously mentioned, TNF-α polymorphism has been associated with SLE, and IL-1-receptor antagonist gene polymorphism has been implicated as a genetic factor in cutaneous LE. UVB irradiation of murine and human skin can result in the generation of immunological tolerance to antigens applied to the skin at the site of irradiation. Increasing evidence suggests that UVB-induction of IL-10 by epidermal keratinocytes or recruited dermal macrophages might play an important role in the generation of this state of active tolerance. Since T-suppressor cell function has often been observed to be defective in SLE patients and murine models of SLE, one could question whether UVB induction of normal amounts of IL-10 within the epidermis of an LE patient might fail to produce an adequate state of physiological tolerance to epidermal autoantigens that are modulated by UVB exposure or epidermal neoantigens that are generated by UVB. Alternatively, genetically determined aberrations in IL-10 expression following UVB might also result in inadequate generation of tolerance signals leading to autoimmune responses to cutaneous autoantigens. At the moment there is no direct data to either support either of these speculations.

MANAGEMENT/TREATMENT Effective treatment of LESSD usually requires systemic therapy with anti-inflammatory or immunosuppressive agents. Approximately 80% of patients respond to one or a combination of the aminoquinoline antimalarials such as hydroxychloroquine (Plaquenil), chloroquine (Aralen), or quinacrine. When antimalarial therapy is not successful in fully suppressing disease activity, alternative therapy with other anti-inflammatory drugs including dapsone, isotretinoin/etretinate, gold, clofazamine, and thalidomide can be beneficial. Long-term systemic corticosteroids and antimetabolite/cytotoxic immunosuppressives such as methotrexate, azathioprine, and cyclophosphamide are reserved for potentially disabling forms of cutaneous disease. Aminoquinoline antimalarials have been associated with a diverse set of actions on cell physiology and pathways of inflammation. Which of their many molecular effects is most important to their value in LESSD is not at all certain. One of the few common actions shared by the aminoquinoline antimalarials and the rather diverse group of other drugs that are of value in cutaneous LE (e.g., dapsone, retinoids, auranofin) is their ability to scavenge oxygen free radicals, perhaps yielding a clue to an important mechanisms of cellular inflammation in LESSD. Thalidomide is another drug that can have a profoundly beneficial effect on LESSD. Recent studies have determined that thalidomide powerfully inhibits the production of TNF-α, a UVB-inducible epidermal cytokine that has been implicated in the pathogenesis of cutaneous LE.

FUTURE DIRECTIONS The absence of a working in vivo or in vitro model of LESSD has greatly hindered our understanding of the molecular and genetic basis of this distinctive form of cutaneous inflammation. Whereas there are murine models of some of the clinical and immunological features of SLE, such animals, even when challenged with UV light do not reproducibly develop clinical skin lesions or develop the typical histopathological changes associated with LESSD. One exception is the MRL mouse. These mice, under some conditions, spontaneously develop patchy hair loss that is

associated with very mild pathological and immunopathological changes typical of LESSD. It would be quite interesting to determine whether MRL mice that are crossed with transgenic mice that overexpress UVB-inducible proinflammatory cytokines such as IL-1α, TNF-α, or MIP in epidermal basal cells might more reliably express an LESSD pathology and thus provide a better model for cutaneous LE inflammation. Much could also be done with the tools of modern molecular biology to determine whether variant expression of UVB-inducible proinflammatory cytokines might occur in the epidermal cells of LE patients compared to normal individuals. For example, RT-PCR and Northern blot analysis might be used to examine sequential suction blister-derived epidermal biopsies from UV light-challenged nonlesional skin of LE patients to determine whether the kinetics of UV light-induced cytokine production might be different in LE patients compared to normal individuals. Systematic studies to detect evidence of oligoclonal T-cellreceptor gene rearrangement are also needed to determine whether T-cell expansion of the type associated with antigen-driven T-cell stimulation might be present in LESSD lesions. Efforts to clone Ro/SS-A autoantigen-specific T cells from SCLE skin lesions could also prove rewarding.

SUMMARY AND CONCLUSIONS David Norris has proposed a four-step model for the pathogenesis of LESSD: (1) exposure to UV light induces the release of proinflammatory epidermal and dermal mediators such as IL-1 and TNF-α; (2) these mediators induce changes in epidermal and dermal cells including the induction of adhesion molecules and promotion of the translocation of normally intracellular autoantigen such as Ro/SS-A to the surface of epidermal cells; (3) autoantibody from the circulation binds to autoantigens such as Ro/SS-A that have been translocated to the surface of epidermal keratinocytes; and (4) keratinocyte cytotoxicity ensues as the result of lymphoid cells that have been recruited from the circulation, recognizing and responding to the Fc domains of autoantibody molecules bound to autoantigen expressed on the surface of keratinocytes (i.e., antibody-dependent cell-mediated cytotoxicity). Although this remains among the most attractive hypotheses for the explanation of Ro/SS-A antibody-associated forms of LESSD such as SCLE and neonatal LE, it does not address the pathogenesis of other forms of LESSD such as DLE that are not associated with Ro/SS-A antibody or other known autoantibody specificity. Better in vitro and in vivo experimental models will be required to more fully understand the molecular basis of all forms of LESSD.

ACKNOWLEDGMENTS This work was supported by NIH grant AR19101 and the resources of the University of Texas Southwestern Skin Disease Research Core Center (AR41940).

SELECTED REFERENCES Boumpas DT, Fessler BJ, Austin HA III, Balow JE, Klippel JH, Lockshin MD. Systemic lupus erythematosus: emerging concepts. Part 2: Dermatologic and joint disease, the antiphospholipid antibody syndrome, pregnancy and hormonal therapy, morbidity and mortality, and pathogenesis. Ann Inter Med 1995;123:42–53. Casciola-Rosen LA, Anhalt G, Rosen A. Autoantigens targeted in systemic lupus erythematosus are clustered in two populations of surface structures on apoptotic keratinocytes. J Exp Med 1994;179:1317–1330. Furukawa F. Animal models of cutaneous lupus erythematosus and lupus erythematosus photosensitivity. Lupus 1997: 6:193–202.

CHAPTER 90 / CUTANEOUS LUPUS ERYTHEMATOSUS

Golan TD, Elkon KB, Gharavi AE, Krueger JG. Enhanced membrane binding of autoantigens to cultured keratinocytes of systemic lupus erythematosus patients after ultraviolet B/ultraviolet A irradiation. J Clin Invest 1992;90:1067–1076. Kawashima T, Zappi EG, Lieu TS, Sontheimer RD. Impact of ultraviolet radiation on the cellular expression of Ro/SS-A-autoantigenic polypeptides. Dermatology 1994;189:6–10. Lee LA. Neonatal lupus erythematosus. J Invest Dermatol 1993;100:9S–13S.

827

Norris DA. Pathomechanisms of photosensitive lupus erythematosus. J Invest Dermatol 1993;100:58S–68S. Rosen A, Casciola-Rosen L, Ahearn J. Novel packages of viral and self-antigens are generated during apoptosis. J Exp Med 1995;181: 1557–1561. Sontheimer RD, Provost TT. Lupus erythematosus. In: Sontheimer RD, Provost TT, eds. Cutaneous Manifestations of Rheumatic Diseases, 1st ed. Baltimore: Williams & Wilkins, 1996; pp. 1–71.

CHAPTER 91 / SCLERODERMA AND MORPHEA

91

829

Scleroderma (Systemic Sclerosis) and Morphea EDWIN A. SMITH AND E. CARWILE LEROY

BACKGROUND The term “scleroderma,” referring to dermal fibrosis, has been applied to two distinct illnesses. One of these, systemic sclerosis, is a generalized vascular and fibrotic autoimmune disease affecting multiple organs, including the lungs, gastrointestinal tract, heart, and kidneys. In contrast, localized scleroderma is limited to the skin. Because there is little evidence that localized scleroderma can become systemic sclerosis and with the likelihood that these represent different etiologies and pathogeneses, systemic sclerosis and localized scleroderma will be treated separately.

SYSTEMIC SCLEROSIS (SSc) HISTORY Notwithstanding a suggestive case report of scleroderma by Curzio of Naples in 1753, the earliest definitive clinical and pathological reports of scleroderma date from the mid-19th century, the same era in which Raynaud described the phenomenon bearing his name (1862) and associated it with scleroderma. That cardiac fibrosis can be of clinical significance was documented by Weiss in 1943, and the association of renal failure and scleroderma was made by Moore and Sheehan in 1952. Recent emphasis has been on understanding the pathophysiology of the disease with a hope that knowledge of the immunologic features as well as the causes of both the fibrosis and the vascular disease will lead to prevention or better treatment. EPIDEMIOLOGY The incidence of systemic sclerosis is estimated at between 4 and 20 per million per year, a broad range because of the likelihood of unrecognized and misclassified cases. Estimates of the prevalence of disease meeting the ACR (American College of Rheumatology) preliminary criteria (see Table 91-1) range from 50,000 to 300,000 in the United States. It has been estimated, using more liberal criteria (Raynaud’s phenomenon, abnormal nailfold capillaries, and positive tests for antinuclear antibodies) to identify persons suffering from “scleroderma spectrum disorders,” that the population prevalence ranges from 67 to 265 per 100,000. Systemic sclerosis is unusual in children, but has its onset throughout adult years. There is a suggestion that African AmeriFrom: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

cans are at a moderately increased risk, but the illness is found in all racial groups and in all locales. Females are affected about three times more often than are males, and this difference is even more striking in young adult years. Associations have been made between the occurrence of systemic sclerosis and particular occupational exposures, including silica (gold and coal miners), vinyl chloride (industrial plastics workers), and organic solvents. Studies vary widely on the association of systemic sclerosis with HLA haplotypes, but several reports have found an association with HLA-DR5. Among selected scleroderma patients, the occurrence of certain autoantibody types has been associated with certain HLA class II alleles (see below). CLINICAL FEATURES Systemic sclerosis has been divided into two distinct types, differing in clinical course, types of organ involvement, and autoantibody profiles (Table 91-2). These two types, diffuse cutaneous systemic sclerosis (dcSSc) and limited cutaneous systemic sclerosis (lcSSc), differ in the extent of maximal skin involvement. In lcSSc, the skin thickening is limited to the face and areas distal to the elbows and knees, whereas, in dcSSc, the sclerosis is proximal as well as distal, and frequently involves the trunk. The clinical differences between limited and diffuse cutaneous systemic sclerosis are outlined in Table 91-3. LcSSc is the more insidious of the two forms of the disease with mild symptoms which may go unreported to a physician for years or decades. Raynaud’s phenomenon in the typical lcSSc patient has often been occurring since teenage or young adult years with symptoms of esophageal reflux or dysphagia occurring at a later date. It is often only with the occurrence of digital ulcers, telangiectases, or symptoms of pulmonary hypertension and right-sided heart failure that the diagnosis is made. Anticentromere antibodies are associated with lcSSc in Caucasian patients and their occurrence in a patient whose only symptom is Raynaud’s phenomenon suggests that this illness is likely to develop. However, these antibodies are found in only approximately 40% of patients with lcSSc and can occur in other diseases than SSc. Early dcSSc is very different from lcSSc, with a much more abrupt onset of symptoms. Although Raynaud’s phenomenon is also common in this form of the disease, it may follow other symptoms. The earliest skin manifestation is edema of the extremities followed over a period of months by progressive induration of the dermis. This induration usually proceeds from distal to proximal

829

830

SECTION IX / DERMATOLOGY

Table 91-2 Clinical Features of Systemic Sclerosis

Table 91-1 Preliminary Criteria for Classification of Systemic Sclerosis (Scleroderma) For the purposes of classifying patients in clinical trials, in population surveys, and for other studies, a person shall be said to have systemic sclerosis (scleroderma) if the one major or two or more minor criteria listed below are present. Localized forms of scleroderma, eosinophilic fascitis, and the various forms of pseudoscleroderma are excluded from these criteria. A. Major criterion 1. Proximal scleroderma: Symmetric thickening, tightening, and induration of the skin of the fingers and the skin proximal to the metacarpophalangeal or metatarsophalangeal joints. The changes may affect the entire extremity, face, neck, and trunk (thorax and abdomen). B. Minor criteria 1. Sclerodactyly: above-indicated skin changes limited to the fingers. 2. Digital pitting scars or loss of substance from the finger pad: depressed areas at tips of fingers or loss of digital pad tissue as a result of ischemia. 3. Bibasilar pulmonary fibrosis: bilateral reticular pattern or lineonodular densities most pronounced in basilar portions of the lungs on standard chest roentgenogram; may assume appearance of diffuse mottling of “honeycomb lung.” These changes should not be attributable to other primary lung disease.

Musculoskeletal Digital flexion contractures Tendon friction rubs Muscle weakness (when myositis present) Calcinosis Arthritis Gastrointestinal involvement Esophageal reflux Dysphagia (diminished peristalsis or stricture) Delayed gastric emptying Malabsorption Alternating diarrhea and constipation Pulmonary Interstitial fibrosis Pulmonary hypertension Cardiac Pericarditis Myocardial fibrosis Tachyarrhythmias Conduction blockade Heart failure Renal Hypertensive renal failure Neurologic Entrapment neuropathies (including trigeminal neuralgia)

Table 91-3 Comparison of Clinical Features of Limited and Diffuse Cutaneous Systemic Sclerosis Onset Initial symptom Skin involvement Lung involvement Heart involvement Renal involvement ANA profile

lcSSc

dcSSc

Insidious Raynaud Distal Pulmonary hypertension Secondary to pulmonary hypertension Very unusual Anticentromere

Abrupt Raynaud or swelling Distal and proximal Fibrosing alveolitis Myocardial fibrosis Hypertensive renal crisis Anti-Scl-70 (antitopoisomerase I) Anti RNA polymerase

on the extremities and often involves the trunk. These dcSSc patients are more likely than the lcSSc patients to develop the more severe manifestations of SSc, including pulmonary fibrosis, hypertensive renal crisis, and heart failure or arrhythmias. Antibodies to topoisomerase I (30–50% of dcSSc patients) or to RNA polymerases I, II, and III (10–40%) are associated with dcSSc. Vascular Manifestations The vascular lesion in systemic sclerosis, luminal narrowing of small arteries and arterioles as a result of intimal proliferation, is widespread and similar in all involved organs. Microvascular involvement is seen in the form of dermal telangiectases and loss of capillaries in all vascular beds. This capillary loss can be demonstrated by in vivo widefield nailfold microscopic examination of the capillaries just proximal to the cuticle. The characteristic changes of systemic sclerosis seen by this technique are capillary loop dilatation and areas of avascularity. Because these changes are not seen in primary Raynaud’s phenomenon, they are very useful from a prognostic point of view in evaluating persons who have recently developed symptoms. The vascular lesions are directly associated with sev-

eral distinct clinical problems, including Raynaud’s phenomenon, pulmonary hypertension, and renal insufficiency. Raynaud’s Phenomenon The nearly universal occurrence of Raynaud’s phenomenon (a reversible decrease in digital blood flow on exposure to cold with subsequent postischemic vasodilatation) underscores the vascular nature of systemic sclerosis. Although this is a common complaint among otherwise healthy persons (occurring in approximately 5% of the general population), it is also the most frequent initial symptom of systemic sclerosis. Patients complain of color changes (blanching or cyanosis), numbness, and/or pain in response to a cold stimulus and there is often a reactive hyperemia with rewarming. Raynaud’s phenomenon in systemic sclerosis results from the superimposition of physiologic cold-induced vasoconstriction on the already pathologically reduced luminal diameter. The ischemia resulting from the digital vascular lesions can result in irreversible changes, including tapering of the palmar fingertip pad, ulcerations of the fingertips and over the proximal interphalangeal joints, and infarction of digital tips.

CHAPTER 91 / SCLERODERMA AND MORPHEA

Renal Involvement The characteristic kidney involvement seen in systemic sclerosis is much more likely to occur in dcSSc than lcSSc, and usually occurs early in the illness when the skin is undergoing rapid fibrosis. A sudden onset of severe hypertension is followed by progressive azotemia. This “scleroderma renal crisis” results from secretion of large quantities of renin by the juxtaglomerular apparatus in response to the decreased blood flow caused by the arteriolar luminal narrowing. The resulting formation of angiotensin causes further vasoconstriction and reduction of renal blood flow, leading to severe hypertension and diminished glomerular filtration. Shearing of erythrocytes within the altered renal vasculature can result in a microangiopathic hemolytic anemia. Until the invention of angiotensin-converting enzyme inhibitors (see Management section, below), renal crisis led invariably to renal failure or death. Pulmonary Hypertension Pulmonary hypertension in systemic sclerosis can be the result of either of two pathologic processes. Fibrosing alveolitis (see below) resulting in restrictive pulmonary mechanics can result in chronic hypoxemia with secondary pulmonary hypertension. However, an increase in pulmonary vascular resistance can occur without fibrotic parenchymal lung disease as a result of the intimal lesion in the pulmonary arterioles. This isolated pulmonary hypertension is seen almost exclusively in lcSSc, where it often remains silent for decades after the occurrence of the initial manifestations of the disease. The earliest clinical manifestation is usually exertional dyspnea with overt right-sided heart failure developing later in the course. Telangiectasia The development of dilated cutaneous blood vessels in the skin is usually a later-developing manifestation of systemic sclerosis. These telangiectases are most common on the face, palms, fingers, and lips. Fibrotic Manifestations Dermal Fibrosis The early edematous phase of skin changes is followed by progressive thickening and tightening of the skin as a result of deposition of extracellular matrix (primarily collagen) in the dermis. This usually proceeds from distal to proximal locations on the extremities. The skin becomes indurated, difficult to pinch into a fold, and tight over bony prominences. The patient may complain of itching or burning sensations. Involvement of underlying structures such as tendons and muscles gives rise to flexion contractures of the fingers and a diminished maximal oral aperture. The maximal extent of skin fibrosis is usually reached in the first 2 or 3 years of disease, followed by progressive atrophy that is perceived as improvement by the patient and clinical examiners. However, this improvement generally does not result in diminished digital contractures. This spontaneous regression of dermal sclerosis has led to many therapies being heralded as beneficial to the course of SSc, only to be later discounted in more rigorous trials. Pulmonary Fibrosis The development of restrictive lung disease as a result of interstitial pulmonary fibrosis has become the most common factor accounting for increased mortality in systemic sclerosis. As stated above, it is much more common in patients with dcSSc than in limited disease. Symptoms (dyspnea on exertion, nonproductive cough) are often delayed until the disease is far advanced. Chest roentgenograms are insensitive to the detection of early lung disease, but show interstitial fibrosis when the disease has progressed. Less-advanced cases of restrictive lung disease may be established by pulmonary function testing in which decreases in vital capacity and gas diffusion are found. Computerized tomographic scanning of the pulmonary parenchyma by

831

high-resolution techniques can identify both early fibrosis and the presence of alveolitis which gives a “ground glass” appearance to the areas of inflammation. This alveolitis presumably represents the prefibrotic stage of the illness. Pulmonary fibrosis results in dyspnea, first on exertion and then at rest. The development of pulmonary hypertension followed by right heart failure frequently precedes death. Cardiac Involvement Several types of heart involvement occur in SSc. Symptomatic pericarditis is less common than is an echocardiographic finding of increased pericardial fluid. Reversible myocardial ischemia as a result of the vascular disease in small vessels results in contraction band necrosis, followed by patchy myocardial fibrosis. Clinical manifestations include varying degrees of heart block, arrhythmias, and congestive heart failure. Left ventricular pressure overload may result from the systemic hypertension associated with renal crisis, and right ventricular hypertrophy followed by failure often results from pulmonary hypertension. Digestive System Involvement Involvement of the gastrointestinal system in SSc follows only Raynaud’s phenomenon as a common clinical manifestation. Diminished smooth muscle motor activity of the distal two-thirds of the esophagus is the most frequent problem and results in dysphagia and esophageal reflux. With chronic reflux esophagitis, stricture can develop. Diminished peristalsis also occurs throughout the remainder of the GI tract. Delayed gastric emptying may worsen esophageal reflux. Small-bowel involvement resulting in malabsorption with weight loss is uncommon, but may be severe in the individual patient. More frequently, stasis results in bacterial overgrowth within the small intestine with deconjugation of bile salts and subsequent development of diarrhea. Colonic involvement results in stasis and constipation. Thinning of the colonic mucosa results in wide mouth diverticuli, which are usually asymptomatic. DIAGNOSIS The diagnosis of SSc rests on clinical grounds according to the preliminary classification criteria established by the American College of Rheumatology (ACR) (Table 91-1). Many persons have illnesses that do not fulfill these intentionally insensitive, but specific, criteria, yet clearly have either early or incomplete forms of SSc. GENETIC BASIS OF SSc As for all autoimmune diseases, the pathogenesis of SSc is felt to involve interacting genetic and environmental factors, although the precise nature of neither is completely understood. Whereas SSc is not inherited in the classical Mendelian sense, some reproducible serological manifestations are regulated by immune-response genes. DNA methods of typing HLA alleles have demonstrated close associations with disease-specific autoimmune manifestations. This link of the immunogenetic profile is much closer to the autoantibody profile than it is to the subset of SSc (lcSSc or dcSSc) or to any specific clinical or outcome parameter. The genetic hypothesis of SSc has been bolstered by rare reports of familial occurrence and by the fact that the autosomal-dominant animal model, the Tsk/+ mouse, expresses disease-specific autoantibodies (anti-topo I and antiRNAP I) in addition to dermal fibrosis. Whereas understanding of direct links between immune events and the vascular, cutaneous, or visceral fibrosis of SSc remains elusive, the presence of distinct autoimmune features in Ssc permits hypotheses similar to those for other autoimmune disorders. These hypotheses include: (1) an altered T-cell repertoire that fails to discriminate certain antigens as “self;” (2) one or more key

832

SECTION IX / DERMATOLOGY

immune events initiated by these self antigens and involving the trimolecular complex (MHC, TCR, antigen) that leads to the expansion and activation of self-reactive T cells; and (3) the elaboration of cytokines to provide help for B cells that express surface Ig molecules that recognize specific autoantigens such as centromere proteins, topoisomerase I, or RNA polymerase. In the specific case of SSc, these elaborated cytokines are also hypothesized to activate fibroblasts and cause endothelial cell damage and proliferation. When fully understood, many specifics of this general autoimmune paradigm should provide the basis for therapeutic immune intervention. Antigen-presenting cells employ HLA molecules to present MHC-restricted epitopes to the heteromultimeric T-cell receptor (TCR). The expression of both the HLA and TCR are under genetic control. TCR genes utilize combinatorial mechanisms (V-J-D) to express a remarkable diversity. Most epitopes are recognized by T cells containing α/β chains in their TCRs. To date, no clonal selection of α/β T cells has been observed in either the blood or target organs (gut or lung) of SSc patients. A less prevalent subset of T-cells expresses only γ/δ chains in their TCRs. These cells seem to favor intestinal subepithelial sites and can be clonally expanded by either antigens or superantigens, the latter being microbial products that activate and expand T cells by ligating TCRs outside the cleft in which they receive MHC-restricted antigens. Normally, individuals express 10–20 different types of delta chains on their γ/δ TCRs. In groups of Ssc patients, however, γ/δ TCRs from blood, lungs, or gut biopsies display five or fewer junctional region sequences, indicating oligoclonal expansion. These studies by White et al. show expanded Vd1 ⊕ γ/δ T cells and lead to the hypothesis that the gut serves as a reservoir of γ/δ T cells where they are exposed to external influences (e.g., toxic oil, conjugated tryptophan, trichloroethylene in drinking water, possibly viruses) in response to which they expand, perhaps migrate, and potentially respond to similar influences in inhaled air. The function of this oligoclonal expansion of γ/δ T cells, including cytokine secretion and B-cell help, is poorly understood in general, and particularly, in SSc. Because T lymphocytes respond to short peptides, precise epitope mapping of these Vd ⊕, γ/δ TCRs might be a fruitful future area of study to design peptides that engage this particular T cell but block its expansion. Both afferent (i.e., related to antigen processing or recognition) and effector (e.g., null complement haplotypes or differences in other immune-inflammatory mediators) immune-response genes have been shown to be associated with SSc to a statistically, but perhaps not clinically, significant degree. No conclusive link between class I HLA types and SSc has been demonstrated. Several class II associations have been made with different results, depending on ethnic origin of the population being studied. Positive associations with DR 3 and 5 (now 11) and negative associations with DR 2 (now 15), 7, and 9 have been made. Logistic-regression analysis has identified the primary class II HLA association to be with DQA1*0501, with the DR associations being explainable by linkage dysequilibrium. Ethnic differences between Europeans, Japanese, African Americans, and Native American Indians are substantial, but generally mirror autoimmune disease associations in those populations. When HLA molecules and autoantibody types are examined in groups of SSc patients regardless of clinical pattern (lcSSc or dcSS), impressive associations between serology and HLA alleles can be demonstrated. For example, in one study of 42 SSc patients

with anticentromere antibodies, 100% were either heterozygous or homozygous for Gly or Tyr (polar residues) rather than leucine (a nonpolar residue) at position 26 of the DQB1 molecule (a position in the peptide-binding groove of the HLA-DQ molecule). However, that 71% of the healthy population and 69% of SSc patients without anticentromere antibodies also have either one or two alleles containing such a polar residue at position 26 makes the likelihood of having anticentromere antibodies if one has such an allele very low. This indicates a major environmental influence. In the case of SSc patients expressing antitopoisomerase I autoantibodies, the polar residue in DQB1 conferring “susceptibility” is at position 30. These intriguing observations both indirectly strengthen the case that the autoimmune serology may be playing a pathogenetic role in the disease and raise the possibility of ultimately immunizing the highly susceptible individual to a blocking peptide. It should be noted that these susceptibility residues reside in the second hypervariable region of the DQB1 chain in contrast to the situation in rheumatoid arthritis or type I diabetes mellitus in which “shared epitopes” reside in the third hypervariable region of the DRB1, highlighting the lack of present understanding of functional differences between DQ and DR HLA molecules in antigen presentation. These class II HLA-autoantibody associations are strong, even if at present they are poorly understood. Class III HLA molecules, which influence effector functions such as complement activation and tumor necrosis factor (TNF)-α cytokine secretion, are encoded by genes located between those of classes I and II on chromosome 6. A null genotype for one of the isoforms of the fourth component of complement, C4A (tightly linked to the general autoimmune haplotype [B8, DR3]) is characterized by a large gene deletion in the class III region, and is associated with several autoimmune disorders, particularly lupus and SSc. In English patients with SSc, C4A null alleles are present in 50% of patients studied. The functional implications of this gene deletion are unknown but nonetheless intriguing. MOLECULAR PATHOPHYSIOLOGY OF DISEASE The pathophysiology of systemic sclerosis is incompletely understood, and the primary etiology(ies) remains completely obscure. Because evidence of the vascular lesion is seen so early in the course of the illness, theoretical constructs that attempt to explain all features of the illness as being secondary to the vascular lesion have been made. However, it is possible that both the vascular and fibrotic features of the illness result from some other cause, possibly immunologic or toxic. That the immunologic and microvascular lesions are already evident when the first clinical manifestations occur makes understanding the seminal event very difficult. The Vascular Lesion The marked abnormalities seen in the blood vessels prompted searches for the cause of the endothelialcell damage and proliferation. It is not clear whether alterations in platelet function (elevated plasma β-thromboglobulin levels and increased circulating platelet aggregates) are a cause or result of the vascular disease. Injury to or activation of the endothelium is supported by the elevated circulating levels of von Willebrand factor and endothelin (both synthesized and released by endothelial cells) and loss of endothelial reserve by diminished levels of angiotensin-converting enzyme. It has been proposed that the T lymphocytes found infiltrating the perivascular areas in early dermal lesions are the cause of the endothelial injury. Clearly, these interstitial lymphocytes leave the circulation by way of the endothelial cells. Increased expression of intercellular adhesion molecule 1 (ICAM-1) on endothelial cells of active dermal lesions

CHAPTER 91 / SCLERODERMA AND MORPHEA

Table 91-4 Antinuclear Antibodies in Systemic Sclerosis Antigen recognized Centromere Topoisomerase I (Scl-70) RNA polymerase I, II, III U3RNP (fibrillarin) Th U1RNP

Incidence (all SSc)

SSc Type (incidence)

10–20% 15–25% 10–25% 5–45% 5% 5–35%

lcSSc 40% dcSSc 25–50% dcSSc 20–45% dcSSc 5% lcSSc lcSSc

and the demonstration that infiltrating lymphocytes bear cell surface leukocyte adhesion factor 1 (LFA-1), the ligand for ICAM-1, indicate a mechanism for the presence of T cells in the perivascular infiltrates. The endothelial cells also express increased amounts of endothelial leukocyte adhesion molecule 1 (ELAM-1 or E-selectin). That T cells may be important in causing endothelial injury is supported by the finding of elevated levels of granzyme-1 and antibodies to granzyme-1 in sera of SSc patients. This enzyme is able to lyse endothelial cells possibly in association with perforin and via a programmed cell death (apoptosis) pathway. Immunologic Aspects Integral immunologic features of SSc include the presence of specific autoantibodies (see Table 91-4), infiltration of the skin by lymphocytes and macrophages, and the demonstration of abnormal amounts of certain cytokines in the blood, lungs, and dermis. Nests of lymphocytes (primarily T cells) and macrophages are found in the dermis, particularly in the deeper reticular area, where fibrosis is most intense. Subset analysis shows these infiltrating lymphocytes to be CD4+ cells bearing α/β T-cell receptors. The occurrence of antinuclear antibodies in the sera of patients with systemic sclerosis approaches 100%. All types of staining patterns may be seen, but nucleolar patterns are more specific. Associations have been made between the presence of antibodies to certain purified nuclear antigens and clinical manifestations of SSc. Antibodies to centromeres (kinetochore plates) occur in approximately 40% of patients with lcSSc, but in only 2–5% of patients with dcSSc. On the other hand antibodies to topoisomerase I (also known as Scl-70) are present much more frequently in dcSSc (30–40%). The presence of these latter antibodies is therefore also associated with rapid skin thickening, pulmonary interstitial fibrosis, and renal crisis. Antibodies to RNA polymerase types I, II, and III are also seen frequently in dcSSc, and account for the frequent occurrence of antinucleolar ANA staining. There are marked racial differences in the occurrence of the various autoantibodies in the clinical subsets of SSc, making the regulation of expression of these specific antibodies of interest. Linkage of certain autoantibody types with specific HLA loci has been demonstrated (see above). Inflammatory Aspects SSc, with only moderate elevations of circulating acute-phase reactants, is not an overwhelmingly inflammatory condition. The only area in which polymorphonuclear leukocytes have been demonstrated to be present in increased numbers is in pulmonary lesions. When alveolar cells are obtained for analysis by bronchoalveolar lavage, increased numbers of PMNs, eosinophils, and macrophages are seen. These macrophages are activated, as evidenced by increased production of fibronectin in cell culture. Analysis of the fluid obtained by

833

BAL has demonstrated increased levels of several cytokines. The PMN chemoattractant IL-8 is present in increased quantities in such fluid. Cytokines that increase fibroblast proliferation (platelet-derived growth factor, PDGF) and matrix synthesis (transforming growth factor type 1, TGF-1) are also present in increased concentrations. Systemic Sclerosis (SSc) and the Extracellular Matrix (ECM) Many abnormalities of the ECM have been documented in SSc, and it is likely that many more will be forthcoming. Recent awareness that cells derive informational signals directly from ECM via integrins as well as from cytokines and growth factors bound to ECM has both substantially increased the momentum to define ECM components and placed ECM modular motifs in a central position in the study of cell behavior in fibrosis (Fig. 91-1). As an example, skin fibroblasts behave quite differently with regard to collagen secretion and metalloproteinase production when cultured on three-dimensional collagen lattices in contrast to monolayer tissue-culture plastic. A detailed review of the components of the ECM is beyond the scope of this discussion. Suffice it to say that there are 19 collagens composed of 33 genes, 8 laminins, 5 thrombospondins, alternatively spliced fibronectins and cytotactins, and wide variations in the charge and size of a family of proteoglycans that have recently been completely renamed. That collagens are the most diverse of the modular protein families thus far characterized can be appreciated visually from Fig. 91-1. The common structural characteristic of collagens is the primary structure Gly-X-Y, which, by virtue of the absence of a side chain on glycine in peptide linkage, folds into a tight and strong α helix. When this helix is uninterrupted for 1000 or more residues, collagen fibrils with impressive tensile strength are formed, characteristic of skin, bone, tendon, and blood vessels. Collagens with interrupted helices and specialized globular modules may participate in other ways in ECM function. Collagens, fibronectins, and proteoglycans may interact with cells in adhesive ways, cytotactins and thrombospondins may function in antiadhesive ways. Cell–matrix interactions via a large number of integrin surface cell membrane receptors trigger signaling cascades and, via reorganization and clustering of intracellular cytoskeletal elements, can direct cell shape, movement, division, and secretion. The cell–matrix interactions of fibroblasts and endothelial cells are under intense study presently and understandings of direct relevance to SSc can be anticipated. The SSc lesional skin fibroblast has been known since 1972 to differ from normal fibroblasts, with overproduction of collagen being the first abnormality described. Whereas its in vitro proliferative response to PDGF is less than that of control fibroblasts, it proliferates in response to TGF-β1, whereas control cells do not, and it does not see bFGF as a mitogen, whereas control cells do. When control cells are exposed to TGF-β1, they express some of the phenotypic traits of the activated SSc fibroblast, including the heightened expression of the PDGF receptors α and β. In vivo healthy skin fibroblasts rarely express collagen α1(I) mRNA but, in SSc lesions, fibroblasts in the vicinity of mononuclear-cell perivascular inflammatory infiltrates express this message. Immunolocalization studies of this same microenvironment can identify TGF-β1 and PDGF AA. It is also the site of microvascular endothelial-cell abnormalities and presumably of perivascular fibrosis as well. Using a combination of in vitro and in vivo techniques, the matrix molecules shown to be overexpressed in the Ssc microlesion are collagens type I, III, V, VI, and VII,

834

SECTION IX / DERMATOLOGY

Figure 91-1 A graphic presentation of the diversity of collagens. (Reproduced with permission from Karger S, Basel AG. The collagen superfamily. Int Arch Allergy Immunol 1995;107:484–490.)

CHAPTER 91 / SCLERODERMA AND MORPHEA

Table 91-5 Cytokines Possibly Involved in the Pathogenesis of Systemic Sclerosis Monocyte/macrophage derived IL-1, IL-6, IL-8, TGF-β, PDGF, fibronectin Lymphocyte derived IL-2, Platelet derived TGF-β, PDGF Endothelial Endothelin Fibroblast Fibronectin, IL-6, VEGF (vascular endothelial growth factor)

fibronectin, and cytotactin (tenascin). Fibroblast exposure to TGF-β1 upregulates these same matrix molecules in a prolonged fashion. The data to-date suggest that the molecular site of upregulation of these matrix molecules is pretranslational in that there is an increased level of steady-state mRNA present. Evidence also exists that suggests both an increased rate of mRNA synthesis and a decreased rate of mRNA decay, in the case of the best-studied collagen, type I. Regarding mRNA synthesis, directed by 5' promoter elements, at least 4 cis-acting DNA response elements can be shown to bind at least three distinct protein-transcription factors, including a collagen-binding factor (CBF) binding a CCAAT motif, the ubiquitous Sp1 binding a series of G-C–rich boxes, and a c-Krox factor binding several G rich motifs. How these interactions combine to initiate transcription and which ones are critical to the fibrosis of Ssc remains to be elucidated. Nonetheless, the area of the transcriptional control of ECM synthesis, the interruption of the widespread and prolonged stimulation by TGF-β1, and ultimately the modulation of ECM production in SSc remains a potentially important area for therapeutic intervention. Cytokines Several cytokines have been implicated in the pathogenesis of systemic sclerosis (Table 91-5). Elevations of IL-2, IL-4, IL-6, and IL-8 have been demonstrated in the sera of patients. Both IL-2 and IL-2-receptor levels seem to mirror the activity of fibrosis that is occurring and therefore indicate involvement of the immune system. IL-2 has no direct effects on fibroblasts, but has been reported to increase TGF-β release by macrophages. Cytokines that directly affect fibroblast proliferation and matrix synthesis are involved in fibrotic lesion. Platelet-Derived Growth Factor (PDGF) PDGF, which exists as either homo- or heterodimers of A and B chains, is a potent mitogen for fibroblasts. Rich sources of this protein include platelet α granules and macrophages, both cell types known to be activated in SSc. PDGF α receptors, which recognize all PDGF isoforms, are increased by TGF-β in scleroderma fibroblasts but not in normal dermal fibroblasts. That PDGF AA, which signals only via α receptors, is found in scleroderma lesions makes it likely that this cytokine is important in fibroblast proliferation in systemic sclerosis. Both PDGF AA and BB are found in increased concentration in fluid removed from the lungs of systemic sclerosis patients by bronchoalveolar lavage. Transforming-Growth Factor β (TGF-β) TGF-β has quite contrasting effects on cells depending on cell type and on their state at the time of study. Interest in TGF-β in SSc stems from the fact that this is the most potent stimulator of fibroblast synthesis of extracellular matrix known. TGF-β is produced by megakaryo-

835

cytes, macrophages, and T cells, particularly TH-2 type. Whereas no increases in blood levels of TGF-β are seen in systemic sclerosis patients, other evidence implicates this cytokine in the areas of occurring lesions. In situ expression of TGF-β mRNA in dermal lesions has been demonstrated. As mentioned above, TGF-β causes dermal fibroblasts of scleroderma lesions to proliferate via increased expression of PDGF α receptors and response to PDGF A ligand. In the lung lesion, increases in TGF-β type 1, but not type 2, have been measured in bronchoalveolar lavage fluid and increases in TGF-β mRNA are seen in mononuclear cells obtained by this means. Interleukin-6 IL-6, a product of macrophages and fibroblasts, has been shown to be produced in vitro by scleroderma fibroblasts at levels up to 30 times that of normal fibroblasts. Although IL-6 is a stimulator of collagen synthesis only at very high concentrations, it inhibits fibroblast synthesis of tissue inhibitor of metalloproteases (TIMP). Produced locally, IL-6 may therefore promote fibrosis by inhibiting collagenase activity. Basic Fibroblast Growth Factor (bFGF) bFGF, a stimulator of fibroblast proliferation, has been found in dermal lesions. Fibroblasts themselves produce bFGF and it may thus be involved in an autocrine role in fibroblast proliferation in systemic sclerosis. Interferon-γ A product of T cells, interferon (IFN)-γ actively suppresses fibroblast collagen synthesis. Serum levels of IFN-γ have been found to be depressed in systemic sclerosis and in vitro stimulation of peripheral blood T cells fails to elicit the increase in interferon gamma production seen in normal T cells. Therefore, a defect in T cells of systemic sclerosis patients resulting in an inability to suppress fibrosis may exist. Interleukin-1 IL-1, primarily a product of macrophages, is able to stimulate fibroblast extracellular-matrix synthesis. Increased sensitivity of systemic sclerosis fibroblasts in culture to stimulatory effects of IL-1β has been demonstrated. There has been recent evidence that systemic sclerosis fibroblasts from clinically involved dermis are able to produce IL-1α and that this cytokine is responsible for increases in PDGF-A and IL-6 production by these cells. This intriguing possibility would link several of the abnormal cytokine productions seen in systemic sclerosis together into a pathogenetically meaningful cytokine “cascade.” Endothelin Endothelin is the potent smooth muscle contracting peptide derived from endothelial cells. Increased blood levels of endothelin have been described in both primary Raynaud phenomenon and in SSc, and some evidence indicates that levels are further increased with exposure to cold. Elevated levels are found in patients with pulmonary hypertension and also fibrotic lung involvement. There is evidence indicating that endothelin can increase fibroblast collagen-synthetic rates, a finding that links the vascular and fibrotic lesions seen in SSc. MANAGEMENT Because the etiology of SSc remains so obscure, treatment is directed to each of the clinical problems presented by the disease. It is hoped that with better understanding, treatment directed at the underlying disease processes can be developed. Skin Involvement Because spontaneous skin softening almost always occurs with time, uncontrolled trials are difficult to interpret, and enthusiasm over proposed therapies must be tempered. There are many reports of dramatic improvements, but these therapies have often failed when studied in a prospective and randomized fashion.

836

SECTION IX / DERMATOLOGY

D-Penicillamine is widely used to treat the dermal thickening, based on its inhibition of collagen crosslinking and facilitation of collagen degradation. Several uncontrolled studies have demonstrated improvement in the extent of skin thickening. A multicenter, randomized trial comparing high-dose and low-dose D-penicillamine is currently underway, and not until its results are known can one assuredly state that D-penicillamine is an effective therapy for SSc. If therapy with D-penicillamine is undertaken, the dose should be gradually increased to 1–1.5 g/d, a dose that must be maintained for 6 months to a year to see any clinically important improvements. Many patients experience side effects with gastrointestinal symptoms being the most common. Renal (hematuria, proteinuria) and hematological (leukopenia, thrombocytopenia) adverse reactions are of greater concern and close monitoring is required. Immunosuppression with glucocorticoids and/or cytotoxic chemotherapy is not effective in changing the skin disease. Chlorambucil and 5-fluorouracil has been shown to be ineffective. IFN-γ, which decreases fibroblast collagen synthesis, has been reported to decrease skin thickening in patients with SSc in uncontrolled trials, but has been associated with some exacerbation of vascular problems. Raynaud’s Phenomenon (RP) Modifications of both behavior and environment are necessary to avoid cool temperatures. β-blockers may worsen RP by allowing unopposed α-adrenergic stimulation. Behavioral therapy with biofeedback may reduce the frequency or severity of attacks, although the evidence suggests that this is more effective in primary than in secondary Raynaud’s phenomenon. Calcium-channel blockers are the most commonly used pharmacologic therapies, but there is no definitive proof of their efficacy in SSc. The sustained release form of nifedipine is prescribed most commonly. Calcium-channel blockade causes inhibition of gut smooth muscle contraction and may therefore increase esophageal reflux or delay peristalsis in the stomach or gut. Prazosin (an α-adrenergic blocker) can be used instead of, or in addition to, calcium-channel blockers. Platelet inhibitors (low-dose aspirin and dipyridamole) are used, although there is no definitive proof of their efficacy. Nitroglycerin paste can be applied at the bases of the most affected digits to promote arterial dilatation. Stellate ganglion blockade to bring about temporary sympathectomy may greatly alleviate pain and induce vasodilatation. Infection of a digital ulcer requires local cleansing, and application of an antibacterial ointment may be helpful. Occasionally infection is so rapidly progressive or advanced to the point of osteomyelitis that therapy with parenteral antibiotics is indicated. Surgical therapy has been used to treat severe RP. Whereas cervical sympathectomy may reduce attack frequency and severity for a short time, symptoms usually recur and it is, therefore, not recommended. Digital sympathectomy has been promoted as more selective and having longer-lasting effects than cervical sympathectomy, but this procedure is still investigational. Amputation of a phalanx or digit may quickly relieve pain, and is sometimes the most effective means of dealing with infection. Renal Involvement The development of hypertensive renal disease in a SSc patient is a medical emergency because irreversible loss of renal function may result. The discovery of angiotensin-converting enzyme (ACE) inhibitors has changed the natural history of SSc. Any elevation in blood pressure should be taken quite seriously, and when persons who have had low-to-

normal pressures develop elevations that are still within the normal range of < 140/90 mmHg, they should be evaluated for renal disease. During the early stages of treatment, captopril rather than other ACE inhibitors is recommended because its shorter half-life allows greater flexibility in dosing. For those who remain hypertensive, captopril should be maintained and other hypotensive agents added using whatever combination of drugs allows normal blood pressures to be obtained. Antihypertensive therapy should be continued, even if renal failure ensues, since there are reports of return of sustained renal function even after dialysis. Pulmonary Disease No controlled, prospective trial has shown any drug to alter the course of SSc interstitial lung disease. All patients should be vaccinated for Streptococcus pneumoniae and influenza. Patients with severe fibrosis may benefit symptomatically from oxygen. Because fibrotic interstitial lung disease is associated with alveolar inflammation, therapy to reduce this inflammation has been undertaken in uncontrolled trials. Daily oral cyclophosphamide (100 mg/d) significantly increased the forced vital capacity (FVC) of patients with alveolitis. This treatment must be undertaken advisedly, however, as no controlled trial has been undertaken. Pulmonary hypertension remains a difficult problem in the management of SSc, and is often involved in the demise of these patients. Pulmonary hypertension may occur either as a complication of interstitial fibrosis or as isolated pulmonary hypertension in the absence of interstitial lung disease. This latter form of severe pulmonary hypertension occurs almost exclusively in patients with limited cutaneous SSc. Echocardiography with Doppler allows estimation of peak systolic pulmonary artery pressure by a noninvasive means. Results of treatment studies of pulmonary hypertension complicating SSc have been quite variable. The long duration of pulmonary vascular disease before it becomes symptomatic leads to fixed vascular lesions that may not be amenable to vasodilation. Calcium-channel blockers may lower pulmonary artery resistance, but their effect on survival is unknown. Cardiac Disease Pericarditis usually responds to nonsteroidal antiinflammatory drugs or to low-dose corticosteroids. Treatment of left-ventricular dysfunction involves the use of digoxin, diuretics, and afterload-reducing agents. Long-term therapy with captopril has been shown to improve systolic and diastolic function in some patients. Nifedipine has been shown to improve myocardial perfusion and cold-induced myocardial ischemia, but no information is available on long-term clinical effects. Antiarrhythmic drugs are used to treat symptomatic ventricular arrhythmias. Gastrointestinal Manifestations Antireflux measures (small, frequent meals; avoiding recumbency after eating; elevation of the head of the bed) are recommended for all SSc patients because of the universal occurrence of esophageal reflux. Damage to the esophageal mucosa from acid reflux can be reduced by use of acid lowering drugs such as H2-blockers or proton-pump inhibitors. Cisapride, which increases gastrointestinal motility by enhancing acetylcholine release at the myenteric interface, has improved upper gastrointestinal symptoms in clinical trials. Esophageal strictures are sought by barium studies and respond to dilation. Malabsorption with diarrhea, weight loss and malnutrition, bloating, distention, and abdominal pain may result from hypoperistalsis of the small bowel. A 2-week treatment with antibiot-

CHAPTER 91 / SCLERODERMA AND MORPHEA

ics—e.g., tetracycline, amoxicillin, or metronidazole—may help by diminishing bacterial overgrowth and relieve symptoms of recurrent diarrhea. The occurrence of pseudoobstruction is associated with vomiting and radiographic evidence of dilated bowel. Any precipitating causes (opioid narcotics, electrolyte abnormalities) should be corrected. Surgical exploration should be avoided as the outcome is usually very poor.

LOCALIZED SCLERODERMA (MORPHEA) EPIDEMIOLOGY The incidence and prevalence of localized scleroderma are unknown, but minor types are quite common and may not come to medical attention. It occurs primarily in young adults and children, with a slight predominance in females. All races are affected. There is no known occupational or exposure risk. CLINICAL FEATURES There are several varieties of morphea, including guttate, plaque, generalized, and linear. These can occur either singly or as several types in the same patient. Guttate morphea are small (few millimeter), oval lesions occurring on the neck and chest. Plaque morphea are a few centimeters to a few inches in diameter and can become generalized with widespread involvement. Initially lesions are erythematous or violaceous lesions that evolve into waxy, ivory-colored skin thickening with a surrounding violaceous border. If these lesions become generalized, the appearance can resemble that of systemic sclerosis, but there is no sclerodactyly, Raynaud’s phenomenon, or internal organ involvement. Linear scleroderma occurs most commonly in children and appears as band-like areas of induration along a particular myotome. There is atrophy of the skin, underlying subcutaneous tissue, muscle, and even osseous involvement. When involving the face (en coup de sabre lesion), facial hemiatrophy may result. This lesion is similar to (or may be the same as) the Parry-Romberg syndrome. Linear scleroderma, when involving a limb, often results in shortening, severe contractures, and considerable disability. Histologically, morphea does not differ from the findings in systemic sclerosis, with a chronic inflammatory cell infiltrate in the inflammatory border, and monotonous increase in dermal collagen. The clinical course of most morphea is to spontaneously soften, but to leave areas of atrophy and depigmentation. Occasionally there can be complete resolution. Progression to systemic sclerosis has been reported, but is so extraordinarily rare as to be reportable. Morphea is not associated with Raynaud’s phenomenon, esophageal dysmotility, interstitial lung disease, or renal involvement. Various nonspecific immunological abnormalities have been reported in morphea, including hypergammaglobulinemia, antinuclear antibodies, anti–single-stranded DNA antibodies, antihistone antibodies, and rheumatoid factor. Systemic sclerosis specific antibodies (antitopoisomerase, anticentromere, anti-RNA polymerase) are not found in patients with linear scleroderma. Although there was initial enthusiasm, when many patients with morphea were found to have antibodies to Borrelia burgdorferi, that this signified the cause of morphea, later studies using PCR to search for this organism’s DNA in morphea lesions have been unsuccessful. It is now felt that Borrelia infection plays no role in causing morphea. GENETIC BASIS OF DISEASE The occurrence of multiple family members with localized scleroderma is very unusual. No

837

predisposing genetic lesion or HLA type has been associated with the occurrence of lesions. MOLECULAR PATHOGENESIS The increases in dermal collagen are of the same types seen in systemic sclerosis—I and III. Immunohistochemically determined increases in TGF-β have been described in the active margins of the lesion in close association with fibroblasts shown to be secreting procollagen. Very few other details concerning the molecules involved in pathogenesis have been described. TREATMENT No treatment has been demonstrated to alter the course of localized scleroderma. Often morphea lesions are self-limiting, running a course from initiation to disappearance in several years. D-penicillamine has been applied to more progressive lesions, and in those individuals who may go on to generalized morphea. No controlled study regarding use of this medication has been completed. Deforming linear scleroderma may require plastic surgery. When a limb has become so deformed as to inhibit function, amputation with prosthetic replacement should be considered.

SELECTED REFERENCES Briggs D, Black CM, Welsh K. Genetic factors in scleroderma. Rheum Dis Clin NA 1990;16:31–51. Briggs DC, Stephens C, Vaughan R, Welsh K, Black CM. A molecular and serologic analysis of the major histocompatibility complex and complement component C4 in systemic sclerosis. Arthritis Rheum 1993;36:943–954. Brown JC, Timpl R. The collagen superfamily. Int Arch Allergy Immunol 1995;107:484–490. Cannon PJ, Hassar M, Case DB, Casarella WJ, Sommers SC, LeRoy EC. The relationship of hypertension and renal failure in scleroderma (progressive systemic sclerosis) to structural and functional abnormalities of the renal cortical circulation. Medicine 1974;53:1–46. Donoghue JF. Scleroderma and the kidney. Kidney Int 1992;41:462–477. Feghali CA, Bost KL, Boulware DW, Levy. Mechanisms of pathogenesis in scleroderma: I. Overproduction of Interleukin 6 by fibroblasts cultured from affected skin of patients with scleroderma. J Rheumatol 1992;19:1202–1211. Fleischmajer R. Localized and systemic scleroderma. In: Lapiere CM, Krieg T, eds. Connective Tissue Diseases of the Skin. New York: Marcel Dekker, 1993; pp. 295–314. Hirakata M, Okano Y, Pati U, et al. Identification of autoantibodies to RNA polymerase II: occurrence in systemic sclerosis and association with autoantibodies to RNA polymerases I and III. J Clin Invest 1993;91:2665–2672. Legerton CW III, Smith EA, Silver RM. Systemic sclerosis (scleroderma): clinical management of its major complications. Rheum Dis Clin NA 1995;21:203–216. Ludwicka A, Ohba T, Trojanowska M, et al. Elevated levels of platelet derived growth factor and transforming growth factor-β1 in bronchoalveolar lavage fluid from patients with scleroderma. J Rheum 1995;22:1876–1883. Masi AT, Rodnan GP, Medsger TA Jr, et al. Preliminary criteria for the classification of systemic sclerosis (scleroderma). Arthritis Rheum 1980;23:581–590. Medgser TA Jr. Systemic sclerosis (Scleroderma), localized forms of scleroderma, and calcinosis. In: McCarty DJ, Koopman WJ, eds. Arthritis and Allied Conditions. Philadelphia: Lea and Febiger, 1993; pp. 1253–1292. Okano Y, Steen VD, Medsger TA. Autoantibody reactive with RNA polymerase III in systemic sclerosis. Ann Int Med 1993;119: 1005–1013. Reveille JD, Durban E, MacLeod-St. Clair MJ, et al. Association of amino acid sequences in the HLA-DQB1 first domain with the antitopoisomerase I autoantibody response in scleroderma (progressive systemic sclerosis). J Clin Invest 1992;90:973–980.

838

SECTION IX / DERMATOLOGY

Reveille JD, Owerbach D, Goldstein R, Moreda R, Isern RA, Arnett FC. Association of polar amino acids at position 26 of the HLADQB1 first domain with the anticentromere autoantibody response in systemic sclerosis (scleroderma). J Clin Invest 1992;89: 1208–1213. Seibold JR, Furst DE, Clements PJ. Why everything (or nothing) seems to work in the treatment of scleroderma. J Rheum 1992;19: 673–676. Seibold JR. Connective tissue diseases characterized by fibrosis. In: Kelley WN, Harris ED Jr, Ruddy S, Sledge CB, eds. Textbook of Rheumatology. Philadelphia: Saunders, 1993; pp. 1113–1143.

Sollberg S, Peltonen J, Uitto J, Jimenez SA. Elevated expression of β1 and β2 integrins, intercellular adhesion molecule 1, and endothelial leukocyte adhesion molecule 1 in the skin of patients with systemic sclerosis of recent onset. Arthritis Rheum 1992;35:290–298. Uziel Y, Krafchik BR, Silverman ED, Thorner PS, Laxer RM. Localized scleroderma in childhood: a report of 30 cases. Sem Arthritis Rheum 1994;23:328–340. Yamakage A, Kikuchi K, Smith EA, LeRoy EC, Trojanowska M. Selective upregulation of platelet-derived growth factor a receptors by transforming growth factor β in scleroderma fibroblasts. J Exp Med 1995;175:1227–1234.

839

CHAPTER 92 / MUSCLE DEVELOPMENT

MUSCULOSKELETAL SECTION EDITOR:

LAURENCE KEDES

X

840

SECTION X / MUSCULOSKELETAL

CHAPTER 92 / MUSCLE DEVELOPMENT

92

841

Muscle Development and Differentiation ERIC N. OLSON

INTRODUCTION Vertebrate species contain dozens of different skeletal muscles, each with unique positions, sizes, shapes, contractile properties, and patterns of innervation. In recent years, dramatic progress has been made toward understanding the genetic pathways that control the formation and patterning of skeletal muscle during embryogenesis. The discovery of several families of genes that act at different steps in the developmental pathway leading to skeletal muscle formation has provided a framework for understanding the complexity of this tissue and will undoubtedly yield insight into the causes of several neuromuscular disorders, as well as provide the potential for therapeutic intervention into such diseases.

REGULATION OF MUSCLE DIFFERENTIATION IN VITRO Much of the progress in understanding muscle development can be attributed to the fact that many of the events associated with muscle differentiation in vivo can be recapitulated in tissue culture. Immortalized muscle cell lines or skeletal myoblasts isolated from embryos proliferate in vitro in the presence of fetal bovine serum and other mitogens such as fibroblast growth factor (FGF). Myoblasts are committed to the myogenic lineage, but they do not express muscle structural genes until they are forced to exit the cell cycle. Myoblast differentiation is accompanied by irreversible withdrawal from the cell cycle, fusion to form multinucleate myotubes, and expression of a large array of muscle-specific genes whose products are required for the specialized contractile, excitable, and metabolic properties of the differentiated muscle fiber. Several early studies suggested the existence of myogenic regulatory factors with the potential to activate muscle gene expression. Experiments in which heterokaryons were formed between skeletal muscle cells and various nonmuscle cells demonstrated that muscle genes became activated in nonmuscle cell nuclei and suggested the existence of trans-acting factors in muscle cells that could activate muscle gene expression. Genetic evidence for such myogenic factors was provided by experiments in which the fibroblast cell line 10T1/2 was shown to form skeletal muscle following treatment with the demethylating agent 5'-azacytidine. The frequency of conversion of 10T1/2 cells to muscle (up to 50%) suggested that one gene was responsible for establishing the myogenic phenotype following its activation by demethylation. This hypothFrom: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

esis was confirmed by DNA transfection experiments in which genomic DNA from myoblasts or 5'-azacytidine-treated 10T1/2 cells was shown to convert 10T1/2 cells to myoblasts with a frequency consistent with a single regulatory gene.

THE MyoD FAMILY OF MUSCLE-SPECIFIC TRANSCRIPTION FACTORS The first myogenic regulatory gene to be cloned was myoD, which was isolated by cloning mRNAs that were expressed in myoblasts, but not 10T1/2 fibroblasts. When a cDNA clone encoding MyoD was expressed ectopically in fibroblasts, it was found to activate skeletal muscle gene expression (Fig. 92-1). Subsequently, three related genes, myogenin, Myf5, and MRF4 (also called Myf6 and herculin) were discovered. These genes, which are referred to as the MyoD family, are expressed exclusively in skeletal muscle; there is no detectable expression in cardiac or smooth muscle, despite the fact that these three muscle cell types express many of the same muscle-specific genes. Members of the MyoD family exhibit distinct expression patterns during differentiation of muscle cells in culture. Most established muscle cell lines express either MyoD or Myf5 when they are proliferating as undifferentiated myoblasts. When induced to differentiate by exposure to growth factor-deficient medium, myogenin is rapidly upregulated immediately prior to activation of muscle structural genes. MyoD and Myf5 also continue to be expressed in differentiated myotubes. MRF4 is not expressed until late in the differentiation program. Members of the MyoD family can induce skeletal muscle gene expression when they are introduced into a wide variety of cell types including fibroblasts, melanoblasts, chondroblasts, smooth and cardiac muscle cells, and osteoblasts. In some cell types, these factors activate the entire myogenic program, resulting in formation of multinucleate myotubes and expression of the full array of muscle-specific genes, whereas in other cell types, only a subset of muscle genes is expressed. There are other cell types, such as adipocytes, in which forced expression of the myogenic factors represses the endogenous program of cell-specific gene expression, while in other cell types the skeletal muscle program is coexpressed with the endogenous program. There are also cell types, such as hepatocytes and several types of transformed cells, in which the myogenic factors are unable to activate muscle gene expression. The failure of the myogenic factors to activate muscle gene expression in these cell types suggests either that they lack certain cofactors required by the myogenic factors or they contain inhibitors of muscle gene expression or both.

841

842

SECTION X / MUSCULOSKELETAL

Figure 92-1 Activation of MHC expression in fibroblasts expressing exogenous MyoD. 10T1/2 fibroblasts were transfected with a MyoD expression vector. Cells were then stained using antibodies specific for MyoD (red) and myosin heavy chain (green). MyoD can been seen within the nuclei of transfected cells and myosin is expressed only in the cells that express MyoD. Figure provided by S. Tapscott, Fred Hutchinson Cancer Research Center. (See color insert following p. 684.)

In addition to activating the expression of muscle structural genes, members of the MyoD family regulate one anothers expression. For example, if MyoD is introduced into fibroblasts, it activates the endogenous MyoD gene as well as the myogenin and MRF4 genes. These autoregulatory interactions have been proposed as a mechanism whereby these factors amplify and maintain their expression and thereby commit cells to a myogenic fate.

THE BASIC HELIX-LOOP-HELIX FAMILY OF TRANSCRIPTION FACTORS Members of the MyoD family share about 80% amino acid identity within a 70-amino acid segment that encompasses a region rich in basic amino acids, followed by a region postulated to adopt a helix-loop-helix (HLH) conformation in which two amphipathic α-helices are separated by an intervening loop (Fig. 92-2). Related, but more divergent, basic-HLH (bHLH) motifs are found in members of a superfamily of proteins that regulate cell proliferation and differentiation in species ranging from yeast to humans. Among these are members of the c-myc family of oncoproteins and the products of several Drosophila genes that regulate embryonic cell fates, including the achaete-scute gene products, which regulate neurogenesis, and twist, which regulates mesoderm formation. The HLH motif mediates protein dimerization and brings together the basic regions of bHLH proteins to form a bipartite

DNA binding domain that recognizes the consensus sequence CANNTG (N = any nucleotide), known as an E-box (Fig. 92-3). This DNA sequence motif is found in the control regions of numerous muscle-specific genes, where it binds the myogenic factors and mediates transcriptional activation. The myogenic bHLH factors homodimerize inefficiently, but they readily form heterodimers with members of a family of ubiquitous bHLH proteins, known as E-proteins. There are also HLH proteins that lack basic regions and inhibit the activity of bHLH proteins by forming heterodimers incapable of binding DNA. Among this class of HLH proteins are members of the Id family. Mutagenesis studies have revealed several functional domains in the myogenic bHLH proteins (Fig. 92-4). These factors contain transcription activation domains in their amino- and carboxyl-termini that are important for efficient activation of muscle-specific transcription. The basic regions of the myogenic factors play a dual role in the control of muscle gene expression by mediating DNA binding and by conferring muscle specificity to transcriptional activation. Fine mapping of individual amino acids in the DNA-binding domains of the myogenic factors has shown that two clusters of basic amino acids are required for binding to the E-box consensus sequence (Fig. 92-4). In addition, two adjacent amino acids, alanine-threonine, in the center of the DNA binding domains, are required for activation of muscle gene expression. These residues are conserved in the basic regions of all known myogenic bHLH proteins in species ranging from Drosophila and sea urchins to humans and they are absent from the basic regions of the more than 40 other bHLH proteins described to date. The specificity and conservation of these residues suggest that they represent part of an ancient mechanism for activation of muscle gene expression. If the alanine-threonine residues in the basic region of a myogenic bHLH protein are replaced with the amino acids found at that position in the basic regions of E proteins, the resulting mutant protein will bind DNA normally, but is unable to activate muscle gene expression. Conversely, if these residues along with a third amino acid between the basic region and helix-1 of the myogenic factors are introduced in the corresponding positions in E proteins, they confer on the E proteins the ability to activate muscle gene expression. The exact mechanism whereby amino acids in the basic region direct muscle-specific transcription is unclear. However, one possibility is that these amino acids induce an allosteric change in the myogenic factors following DNA binding, which might expose a protein interface that can interact with accessory factors required for the activation of muscle gene expression. In this regard, the crystal structure of a MyoD homodimer binding to an E-box DNA sequence has been deduced. The structure of the complex confirms that the basic region and first helix of the HLH region form a continuous ahelix. Alanine-threonine in the MyoD basic region lie within the major groove of the DNA binding site and are therefore inaccessible to other proteins. However, the interaction of these residues with the DNA appears to induce a conformational change in the protein, which may influence the proteins with which it can interact.

GROWTH FACTOR CONTROL OF MYOGENESIS As in many cell types, differentiation of myoblasts is coupled to withdrawal from the cell cycle. Whereas many cell types can reenter the cell cycle when stimulated with mitogens, skeletal muscle cells lose the ability to reinitiate DNA synthesis when they differentiate and therefore become irreversibly committed to the postmitotic state. Because myoblasts express MyoD or Myf5

CHAPTER 92 / MUSCLE DEVELOPMENT

843

Figure 92-2 Diagramatic representation of the four myogenic bHLH proteins. A linear representation of each myogenic regulatory factor is shown. The basic (+++) and HLH regions are indicated. The number of amino acids in each protein is shown at the end of each box.

Figure 92-3 A Schematic representation of a MyoD/E12 heterodimer is shown. The basic region and helix-1 form a contiguous α-helix separated from helix-2 by an intervening loop. Each basic region recognizes half of the palindromic sequence CANNTG. E12 is represented in blue and MyoD in red. (See color insert following p. 684.)

(these factors are unable to induce muscle gene expression when cells are exposed to growth factors, however), there must be posttranslational mechanisms that inhibit their activities. There appear to be multiple mechanisms whereby growth factors prevent myogenic bHLH proteins from activating muscle-specific genes. For example, Id proteins are expressed at high levels in proliferating undifferentiated myoblasts and are downregulated when myoblasts are induced to differentiate. Because Id proteins dimerize preferentially with E proteins, they sequester the dimerization partners for myogenic bHLH proteins in myoblasts. The downregulation of Id has been proposed to release E proteins so that they can dimerize with myogenic bHLH proteins and activate musclespecific genes. Consistent with this model, forced expression of Id from an expression plasmid is sufficient to prevent myoblast differentiation under conditions that would normally promote differentiation. In the embryo, Id expression is also downregulated in regions of skeletal muscle differentiation. The nuclear oncogene products Fos and Jun, which are upregulated by mitogens, can also block muscle differentiation by interfering with the activities of myogenic bHLH proteins. The inhibitory activities of Fos and Jun appear to be mediated by competition with the myogenic factors for interaction with a common cofactor required for myogenesis and by direct interactions with the myogenic factors. Members of the MyoD family have also been shown to be targets for several protein kinases. Protein kinase C, which is activated by growth factors, can substitute for growth factors and block myoblast differentiation. The threonine residue in the center of the DNAbinding domains of the myogenic factors, which mediates musclespecific gene activation, has been shown to be efficiently

844

SECTION X / MUSCULOSKELETAL

Figure 92-4 Schematic representation of the myogenin protein. Myogenin is a 224-amino acid protein with the bHLH motif near the center. The 12 amino acids within the basic region are required for sequence-specific DNA binding in conjunction with bHLH motif. The alaninethreonine in the center of the basic region are conserved in and specific to all members of the MyoD family and are important for activation of muscle-specific transcription. This threonine is also a protein kinase C phosphorylation site, which inhibits DNA binding when phosphorylated. S, S/T, serine- and serine/threonine-rich regions, respectively.

phosphorylated by PKC in vitro and in vivo. In the case of myogenin, this phosphorylation prevents DNA binding, presumably because it introduces a negative charge into the center of the DNA-binding domain, resulting in electrostatic repulsion from the DNA. FGF can induce phosphorylation of this threonine in cultured cells, resulting in repression of myogenin’s ability to activate muscle gene expression. The ability of this threonine to serve as a target for growth factor-dependent phosphorylation pathways that block myogenesis demonstrates that this single amino acid plays a dual role in the control of muscle development; it is structurally important for transcriptional activation, and it serves as a target for growth factor signal transduction pathways under control of PKC. Type-β transforming growth factor (TGF-β) also inhibits muscle differentiation by blocking the activities of the myogenic factors. In myoblasts exposed to TGF-β, myogenic bHLH proteins are expressed, and they can bind to their target DNA sequences associated with muscle-specific genes but are unable to activate expression of those genes. It has been proposed that TGF-β interferes with the expression or activity of a cofactor required by the myogenic factors to activate myogenesis. The linkage of muscle differentiation to cell cycle withdrawal suggests that the cell cycle machinery interfaces with the mechanism for muscle gene activation. The retinoblastoma protein (Rb) has been shown to be an important regulator of cell cycle progression. Rb is a nuclear phosphoprotein that is phosphorylated in a cell cycle-dependent manner. In cycling cells, Rb is phosphorylated in a specific domain known as the “pocket,” whereas in growth-arrested cells Rb is dephosphorylated. When quiescent cells are stimulated to reenter S phase, Rb becomes phosphorylated during the G1/S phase transition prior to initiation of DNA synthesis. The pocket of Rb interacts with a variety of nuclear proteins to control cell cycle progression. Among these is the transcription factor E2F, which activates the expression of several genes involved in cell proliferation. Binding of E2F to dephosphorylated Rb inactivates E2F and results in a block to proliferation (Fig. 92-5). Conversely, phosphorylation of Rb releases E2F so that it can activate cell growth. The oncoproteins encoded by several DNA tumor viruses, such as SV40 large T antigen and adenovirus E1A, also bind to the pocket of Rb, which prevents interaction between Rb and E2F, resulting in uncontrolled proliferation. These oncogenes are potent inhibitors of myogenesis and when expressed in terminally differentiated myotubes can lead to reinitiation of DNA synthesis.

MyoD has also been shown to bind to the pocket of dephosphorylated Rb. This interaction requires the basic region and first helix of MyoD and is dependent on the alanine-threonine residues in the basic region that are required for activation of muscle gene expression. Because MyoD and E2F bind the same region of Rb, they must interact with separate molecules of Rb. Exactly how MyoD binds DNA and Rb simultaneously is unclear. The role of Rb in the MyoD-DNA complex also remains to be determined. Rb is upregulated during myogenesis and is required for irreversible exit of muscle cells from the cell cycle. Myoblasts lacking Rb are able to form myotubes and express muscle-specific genes, but they do not irreversibly exit the cell cycle and can resynthesize DNA in response to mitogenic stimulation. The pocket of Rb is phosphorylated by cyclin-dependent protein kinases (CDKs). The CDKs are expressed constitutively throughout the cell cycle but are activated by association with regulatory proteins called cyclins. Cyclin D1 is induced during the G1 to S phase transition and activates CDK4, resulting in phosphorylation of Rb. Forced expression of cyclin D prevents activation of muscle gene expression by MyoD, presumably because it leads to inactivation of Rb, although a more direct role of the cyclin D1/CDK4 complex in phosphorylating MyoD has not been ruled out. There is also a family of CDK inhibitors (CKIs) that act as inhibitors of cell growth by blocking CDK activity. One of these CKIs, called p21/WAF1/CIP1, is upregulated during myogenesis and is likely to play a role in locking cells out of the cell cycle. When CKI is expressed in proliferating myoblasts, it can inhibit growth and induce muscle differentiation even in the presence of high concentrations of mitogens (Fig. 92-5); p21 is also expressed in differentiating muscle cells during embryogenesis. Upregulation of p21 expression is also observed in several differentiated cell types in the embryo, suggesting it may be part of a common mechanism linking cell cycle withdrawal to activation of cell differentiation.

EMBRYONIC ORIGINS OF SKELETAL MUSCLE The formation of skeletal muscle during embryogenesis involves the commitment of mesodermal stem cells to the skeletal muscle lineage and the subsequent differentiation of myoblasts to form multinucleate myotubes, which mature into skeletal muscle fibers. Skeletal muscle is derived from the somites, which form in a rostral-to-caudal gradient by segmentation of the paraxial mesoderm along the neural tube (Fig. 92-6). Newly formed somites

CHAPTER 92 / MUSCLE DEVELOPMENT

845

Figure 92-5 Model depicting the interaction between the cell cycle machinery and the myogenic regulatory pathway. MyoD or Myf5 are expressed in myoblasts and induce the expression of the CDK inhibitor p21 in the absence of growth factors. p21 induces growth arrest, which results in myogenin expression. Myogenin in turn activates muscle differentiation. Growth factor signals induce cyclin D1, which activates CDK4, resulting in phosphorylation of the Rb pocket, as well as inhibition of MyoD function. Phosphorylated Rb is unable to bind E2F and is therefore inactive as a growth suppressor. In the absence of growth factor signals, Rb is dephosphorylated and binds to E2F. E2F normally activates expression of genes required for cell proliferation, but it is inactivated when bound to Rb.

Figure 92-6 Diagrammatic representation of somite maturation. Schematic representation of an early chick embryo is shown on the left. Somites form in a rostral-to-caudal progression by segmentation of the segmental plate mesoderm. Right: cross-sections of the embryo at the indicated level. (A) Paraxial mesodermal cells in the segmental plates have not yet formed somites. (B) Newly formed somites appear as epithelial spheres. (C) Immature somites have become compartmentalized to form the dermamyotome and sclerotome. (D) Mature somites have formed the dermatome, myotome, and sclerotome.

846

SECTION X / MUSCULOSKELETAL

patterning of the somites. In the absence of neural tube, the myotome fails to form. In contrast, the formation of limb muscles is independent of the neural tube or notochord. Several of the myogenic bHLH genes have been analyzed for cis-acting DNA sequences that direct expression in the early somites and subsequently during skeletal muscle formation. The best characterized of these genes is myogenin, which is controlled by DNA sequences in the proximal promoter region. When these sequences are linked to the β-galactosidase gene as a marker and are introduced into transgenic mice, they direct the expression of β-galactosidase in the exact pattern as the endogenous myogenin gene (Fig. 92-7). Within the myogenin gene control region is an E-box that binds the myogenic bHLH proteins with high affinity and a binding site for the myocyte enhancer binding factor-2 (MEF2) family of transcription factors. Together, these two families of regulators control the expression of myogenin in the somites and limb buds.

A GENETIC PATHWAY FOR2 MUSCLE DEVELOPMENT REVEALED BY GENE TARGETING IN TRANSGENIC MICE

Figure 92-7 Expression pattern of a myogenin-lacZ transgene in an 11.5-day mouse embryo. The myogenin promoter was linked to a β-galactosidase reporter gene and introduced into transgenic mice. The expression of β-galactosidase activity was detected by a colorimetric assay and is shown in blue. The expression pattern of the transgene is identical to that of the endogenous gene. (See color insert following p. 684.)

appear as an epithelial sphere, which subsequently becomes compartmentalized to form the dermamyotome and the sclerotome. The sclerotome gives rise to the ribs and vertebrae. The dermamyotome gives rise to the myotome and dermatome, which form axial muscles of the back and the dermis, respectively. Myf5 is the first member of the MyoD family to be expressed during mouse embryogenesis, appearing in the somite at embryonic day 8 (E8). Myogenin transcripts appear in the myotome by E8.5 and MRF4 and MyoD are expressed at E9 and E10.5, respectively. Muscle cells in the limbs arise from cells that emigrate from the ventrolateral edge of the dermamyotome. These migrating cells are committed to the myogenic lineage, but they do not express the myogenic bHLH factors until they reach the limbs. The homeobox gene Pax3 is expressed in the migrating myogenic precursor cells. Evidence that Pax3 is involved in migration or commitment of these cells to the myogenic lineage has been provided by analysis of a mouse mutant called Splotch. The splotch mutation maps to the Pax3 gene and results in the absence of limb muscles. Pax3 may mediate activation of MyoD and Myf5 in response to muscleinducing signals from overlaying ectoderm. Interestingly, a chromosomal translocation resulting in structural rearrangement of the human Pax3 gene has been associated with alveolar rhabdomyosarcoma, a malignant tumor of skeletal muscle (see Chapter 3). Whereas limb muscles are derived from myogenic precursor cells that migrate from the ventrolateral edge of the dermamyotome, the axial muscles of the back are derived from the myotomal compartment of the somites. The neural tube and notochord have been shown to play important roles in somite maturation by serving as a source for signaling molecules that induce

To determine the specific roles of the myogenic bHLH genes in muscle development in vivo, these genes have been deleted from the genome of transgenic mice by homologous recombination. Remarkably, mice lacking MyoD are fully viable and show no obvious muscle abnormalities. The only apparent consequence of MyoD inactivation is a twofold increase in expression of Myf5. Paradoxically, mice homozygous for a null mutation of Myf5 also develop normal skeletal muscle, but die at birth because of the absence of ribs, which results in an inability to breathe. It is unclear why Myf5, which is expressed exclusively in skeletal muscle cells, affects the formation of ribs, which are derived from the somite. One possibility is that Myf5 controls the expression of growth factors or extracellular matrix molecules by muscle cells, which are involved in regulating the behavior of adjacent rib precursor cells in the somite. When both MyoD and Myf5 are inactivated in the same animal, there is a complete absence of skeletal muscle and there are no detectable myoblasts. These findings suggest that MyoD and Myf5 play overlapping roles in the generation of myoblasts. This type of genetic redundancy could occur if MyoD and Myf5 were expressed in the same cells or if they were expressed in separate cells, either of which could support normal muscle development if the other cell population were eliminated. When lacZ expression is driven by the Myf5 locus in Myf5 null embryos, muscle progenitor cells migrate aberrantly. Thus, Myf5 protein appears to be necessary for cells to respond correctly to positional cues and to adopt a myogenic fate. The phenotype of mice lacking myogenin suggests that myogenin acts in a genetic pathway downstream of MyoD and Myf5. In myogenin-null mice, myoblasts are normally specified and positioned, but there is a near-complete absence of differentiated skeletal muscle fibers. The undifferentiated myoblasts that populate the presumptive muscle forming regions of myogenin-null mice express MyoD and Myf5, indicating that these myogenic factors do not require myogenin for their expression and suggesting that they cannot direct the formation of normal skeletal muscle in the absence of myogenin. During normal muscle development, myogenin is downregulated at birth as MRF4 becomes upregulated to high levels. MRF4 is not expressed above background levels in myogenin-null mice, which suggests that it lies downstream of myogenin in the

CHAPTER 92 / MUSCLE DEVELOPMENT

847

Figure 92-8 The genetic pathway for skeletal muscle development. According to this model, either MyoD or Myf5 become expressed when mesodermal precursor cells from the somite become committed to the myogenic lineage. Myoblast-specific genes are regulated by MyoD and Myf5. When the extracellular concentration of growth factors is reduced, MyoD or Myf5 activate the expression of myogenin, which in turn activates myotube-specific genes. Myogenin is normally downregulated when MRF4 is upregulated after birth. In MRF4-null mice, myogenin continues to be expressed at a high level, suggesting that MRF4 is normally required for its downregulation.

myogenic pathway. MRF4-null mice have normal skeletal muscle and are fully viable. However, myogenin is overexpressed in these mice, which suggests that it may compensate for the lack of MRF4. This upregulation of myogenin in the absence of MRF4 also suggests that MRF4 is required for the downregulation of myogenin that normally occurs in postnatal skeletal muscle. Together, the results of gene knockout experiments suggest that muscle cell determination and differentiation are controlled by the type of genetic cascade shown in Fig. 92-8.

MEF2 FACTORS AND MYOGENESIS The majority of skeletal muscle genes contain E-boxes in their control regions and are therefore probably directly activated by the myogenic bHLH factors. However, there are also skeletal muscle genes that lack E-boxes. Many of these genes are controlled by the myogenic bHLH factors through a cascade of events in which the myogenic factors induce the expression of intermediate regulators, which in turn activate muscle-specific genes. MEF2 proteins are among these intermediate regulators. MEF2 factors are unable to activate muscle gene expression alone, but they potentiate the transcriptional activity of bHLH proteins. This potentiation appears to be mediated by direct interactions between the DNA-binding domain of these different types of transcription factors. The MEF2 factors belong to the MADS box family of transcription factors, named for MCM1, which regulates mating type-specific genes in yeast, Agamous and Deficiens, which act as homeotic genes to control flower development in plants, and Serum Response Factor (SRF), which controls several serum-inducible and muscle-specific genes. There are four MEF2 genes in vertebrates: MEF2A,-B,-C, and -D (Fig. 92-9). The proteins encoded by these genes are nearly identical within the MADS domain, which is located at their amino-termini. The MADS domain mediates homo- and heterodimerization and DNA binding to the sequence CTA(A/T)4TAG, which is found in the control regions of nearly every skeletal as well as cardiac muscle gene. Immediately adjacent to the MADS domain is a region known as the MEF2 domain, in which the MEF2 factors also share extensive homology. The function of the MEF2 domain has not yet been determined. The carboxyl-termini of the MEF2 factors are required for transcriptional activation and are subject to complex patterns of alternative splicing, which generates enormous complexity of this family.

Figure 92-9 Diagrammatic representation of the proteins encoded by the four MEF2 genes. A linear representation of each MEF2 factor is shown. The MADS and MEF2 domains are indicated and alternative exons are shown.

In addition to regulating muscle structural genes, MEF2 factors have been shown to regulate the myogenic bHLH genes. The promoters of the myogenin, MyoD, and MRF4 genes have been shown to contain MEF2 sites that are essential for transcription of these genes in vivo and in vitro. During mouse embryogenesis, the MEF2 factors are expressed in the somite myotome after Myf5 and myogenin, which suggests that they do not play a role in the initial activation of these genes, but they are likely to be required for amplification and maintenance of the expression of these genes once their expression has been initiated. MEF2 expression can be induced in nonmuscle cells by forced expression of myogenic bHLH proteins, which suggests that the

848

SECTION X / MUSCULOSKELETAL

RELATIONSHIP OF SKELETAL MUSCLE TO OTHER TISSUES There are several reasons to expect that networks of cell typespecific bHLH proteins may control cell determination and differentiation in cell types other than skeletal muscle. The presence of E-boxes in the control regions of numerous tissue-specific genes suggests that these genes are controlled by specific bHLH proteins. The ubiquitous expression of E-proteins also suggests the existence of cell type-specific dimerization partners for these proteins. Id proteins are also expressed in many types of undifferentiated cells. Indeed, in recent years, several cell type-specific bHLH proteins have been identified in Drosophila and vertebrate species. Further investigations into their mechanisms of action seems likely to reveal fundamental mechanisms for the control of tissue-specific gene expression during development.

SELECTED REFERENCES Figure 92-10 Activation of muscle gene expression by myogenic bHLH and MEF2 factors. Members of the MyoD family positively autoregulate their own expression and directly activate transcription of muscle-specific genes containing E-boxes in their control regions. MEF2 factors are induced in response to myogenic bHLH proteins and can in turn activate the expression of muscle-specific genes that lack E-boxes in their control regions. MEF2 proteins also bind the control regions of several myogenic bHLH genes and therefore participate in a positive feedback loop that amplifies and maintains the myogenic program. Myogenic bHLH proteins are shown in black, designated M.

Mef2 genes lie in a genetic pathway downstream of the myogenic bHLH genes. However, forced expression of MEF2 in fibroblasts has been shown to induce the expression of the myogenic bHLH genes and to activate muscle differentiation. This suggests that MEF2 factors and myogenic bHLH proteins function within a complex regulatory network in which members from either family of factors can autoregulate one another and crossregulate the other family members (Fig. 92-10). In this way, these two families of regulators reinforce one another’s expression and drive cells toward differentiation. The four Mef2 genes are also expressed at the earliest stages of heart development, as well as in smooth muscle cells, suggesting that MEF2 factors may regulate muscle gene expression in multiple muscle cell types. The functions of MEF2 have been analyzed genetically in Drosophila, which contains only a single Mef2 gene, called D-mef2, which encodes a protein that is virtually identical to the vertebrate MEF2 proteins in the MADS and MEF2 domains. The Drosophila MEF2 protein also binds the same DNA sequence as the mammalian factors and can activate transcription through the MEF2 site in mammalian cells. Like its vertebrate relatives, D-MEF2 is expressed specifically in myogenic precursors and their descendants during embryogenesis. Drosophila embryos lacking D-mef2 contain myoblasts that are normally specified and positioned, but these cells are unable to differentiate. The severe muscle defect in these mutant flies suggests that Mef2 provides a function required for activation of muscle gene expression in multiple muscle cell types and suggests that the genetic functions of Mef2 have been conserved since flies and humans evolved from a common ancestor over 600 million years ago.

Benezra R, Davis RL, Lockshon D, Turner DL, Weintraub H. The protein Id: a negative regulator of helix-loop-helix DNA binding proteins. Cell 1990;61:49–59. Bengal E, Ransone L, Scharfmann R, et al. Functional antagonism between c-Jun and MyoD proteins: a direct physical association. Cell 1992;68:507–519. Bour BA, O’Brien MA, Lockwood WL, et al. Drosophila MEF2, a transcription factor that is essential for myogenesis. Genes Dev 1995; 9:730–741. Braun T, Rudnicki MA, Arnold HH, Jaenisch R. Targeted inactivation of the muscle regulatory gene Myf-5 results in abnormal rib development and perinatal death. Cell 1992;71:369–382. Brennan TJ, Chakraborty T, Olson EN. Mutagenesis of the myogenin basic region identifies an ancient protein motif critical for activation of myogenesis. Proc Natl Acad Sci USA 1991;88:5675–5679. Cheng T-C, Wallace M, Merlie JP, Olson EN. Separable regulatory elements govern myogenin transcription in embryonic somites and limb buds. Science 1993;261:215–218. Cserjesi P, Olson EN. Myogenin induces muscle-specific enhancer binding factor MEF-2 independently of other muscle-specific gene products. Mol Cell Biol 1991;11:4854–4862. Davis RL, Cheng P-F, Lassar AB, Weintraub H. The MyoD DNA-binding domain contains a recognition code for muscle-specific gene activation. Cell 1990;60:733–746. Davis RL, Weintraub H. Acquisition of myogenic specificity by replacement of three amino acid residues from MyoD into E12. Science 1992;256:1027–1030. Davis RL, Weintraub H, Lassar AB. Expression of a single transfected cDNA converts fibroblasts to myoblasts. Cell 1987;51:987–1000. Emerson CP. Skeletal myogenesis: genetics and embryology to the fore. Curr Opin Genet Dev 1993;3:265–274. Gu W, Schneider JW, Condorelli G, Kaushal S, Mahdavi VJ, NadalGinard B. Interaction of myogenic factors and the retinoblastinoma protein mediates muscle cell commitment and differentiation. Cell 1993;72:309–324. Halevy O, Novitch BG, Spicer DB, et al. Correlation of terminal cell cycle arrest of skeletal muscle with induction of p21 by MyoD. Science 1995;267:1018–1021. Hasty P, Bradley A, Morris JH, et al. Muscle deficiency and neonatal death in mice with a targeted mutation in the myogenin gene. Nature 1993;364:501–506. Kaushal S, Schneider JW, Nadal-Ginard B, Mahdavi V. Activation of the myogenic lineage by MEF2A, a factor that induces and cooperates with MyoD. Science 1994;266:1236–1240. Konieczny SF, Emerson CP, Jr. 5-azacytidine induction of stable mesodermal stem cell lineages from 10T1/2 cells: evidence for regulatory genes controlling determination. Cell 1984;38:791–800. Kothary FT, Surani MA, Halata Z, Grim M. The splotch mutation interferes with muscle development in the limbs. Anat Embryol 1993; 187:153–160.

CHAPTER 92 / MUSCLE DEVELOPMENT

Lassar AB, Paterson BM, Weintraub H. Transfection of DNA locus that mediates the conversion of 10T1/2 fibroblasts to myoblasts. Cell 1986;47:649–656. Li L, Chambard J-C, Karin M, Olson EN. Fos and Jun repress transcriptional activation by myogenin and MyoD: the amino terminus of Jun can mediate repression. Genes Dev 1992;6:676–689. Li L, Zhou J, James G, Heller-Harrison R, Czech M, Olson EN. FGF inactivates myogenic helix-loop-helix proteins through phosphorylation of a conserved protein kinase C site in their DNA binding domains. Cell 1992;71:1181–1194. Lilly B, Galewsky S, Firulli AB, Schulz RA, Olson EN. D-MEF2: A MADS box transcription factor expressed in differentiating mesoderm and muscle cell lineages during Drosophila embryogenesis. Proc Natl Acad Sci USA 1994;91:5662–5666. Lilly B, Zhao B, Ranganayakulu G, Paterson BM, Schulz RA, Olson EN. Requirement of MADS domain transcription factor D-MEF2 for muscle formation in Drosophila. Science 1995;267:688–693. Ma PCM, Rould MA, Weintraub H, Pabo CO. Crystal structure of MyoD bHLH domain-DNA complex: Perspectives on DNA recognition and implications for transcriptional activation. Cell 1994;77:451–459. Maroto M, Reshef R, Munsterberg AE, Koester S, Goulding M, Lassar AB. Ectopic Pax-3 activates MyoD and Myf-5 expression in embryonic mesoderm and neural tissue. Cell 1997;89:139–148. Martin JF, Miano JM, Hustad CM, Copeland NG, Jenkins NA, Olson EN. A Mef2 gene that generates a muscle-specific isoform via alternative mRNA splicing. Mol Cell Biol 1994;14:1647–1656. Molkentin JD, Olson EN. Combinatorial control of muscle development by basic helix-loop-helix and MADS-box transcription factors. Proc Natl Acad Sci USA 1996;93:9366–9373. Morgan DO. Principles of CDK regulation. Nature 1995;374:131–134. Murre C, McCaw PS, Baltimore D. A new DNA binding and dimerization motif in immunoglobulin enhancer binding, daughterless, MyoD, and myc proteins. Cell 1989;56:777–783. Murre C, McCaw PS, Vaessin H, et al. Interactions between heterologous helix-loop-helix proteins generate complexes that bind specifically to a common DNA sequence. Cell 1989;58:537–544. Nabeshima Y, Hanaoka L, Hayasaka M, et al. Myogenin gene disruption results in perinatal lethality because of severe muscle defect. Nature 1993;364:532–535. Nevins JR. E2F: a link between the Rb tumor suppressor protein and viral oncoproteins. Science 1992;258:424–429. Olson EN. The MyoD family, a paradigm for development? Genes Dev 1990;4:1454–1461.

849

Olson EN, Klein WH. bHLH factors in muscle development: dead lines and commitments, what to leave in and what to leave out. Genes Dev 1994;8:1–8. Parker S, Eichele G, Zhang P, et al. p53-independent expression of p21Cip1 in muscle and other terminally differentiating cells. Science 1995; 267:1024–1027. Pollock R, Treisman R. Human SRF-related proteins: DNA-binding properties and potential regulatory targets. Genes Dev 1991;5:2327–2341. Rao SS, Chu C, Kohtz DS. Ectopic expression of cyclin D1 prevents activation of gene transcription by myogenic basic helix-loop-helix regulators. Mol Cell Biol 1994;14:5259–5267. Rudnicki MA, Braun T, Hinuma S, Jaenisch R. Inactivation of MyoD in mice leads to upregulation of the myogenic HLH gene Myf-5 and results in apparently normal muscle development. Cell 1992;71:383–390. Rudnicki MA, Schnegelsberg PNJ, Stead RH, Braun T, Arnold HH, Jaenisch R. MyoD or Myf-5 is required for the formation of skeletal muscle. Cell 1993;75:1351–1359. Schneider JW, Gu W, Zhu L, Mahdavi V, Nadal-Ginard B. Reversal of terminal differentiation mediated by p107 in Rb-/-muscle cells. Science 1994;264:1467–1471. Shapiro DN, Sublett JE, Li B, Downing JR, Naeve CW. Fusion of PAX3 to a member of the forkhead family of transcription factors in human alveolar rhabdomyosarcoma. Cancer Res 1993;53:5108–5112. Tajbakhsh S, Rocancourt D, Buckingham M. Muscle progenitor cells failing to respond to positional cues adopt non-myogenic fates in myf-5 null mice. Nature 1996;384:266–270. Tapscott SJ, Davis RL, Thayer MJ, Cheng P-F, Weintraub H, Lassar AB. MyoD1: A nuclear phosphoprotein required in a myc homology region to convert fibroblasts to myoblasts. Science 1988;242:405–411. Wachtler F, Crist B. The basic embryology of skeletal muscle formation in vertebrates: the avian model. Semin Dev Biol 1992;3:217–227. Weintraub H. The MyoD family and myogenesis: redundancy, networks and thresholds. Cell 1993;75:1241–1244. Weintraub H, Tapscott SJ, Davis RL, et al. Activation of muscle specific genes in pigment, nerve, fat, liver and fibroblast cell lines by forced expression of MyoD. Proc Natl Acad Sci USA 1989;86:5434–5438. Williams BA, Ordahl CP. Pax-3 expression in segmental mesoderm marks early stages in myogenic cell specification. Development 1994; 120:785–796. Zhang W, Behringer RR, Olson EN. Inactivation of the myogenic bHLH gene MRF4 results in upregulation of myogenin and rib anomalies. Genes Dev 1995;9:1388–1399.

CHAPTER 93 / SKELETAL MUSCLE

93

851

Skeletal Muscle Structure and Function HENRY F. EPSTEIN

INTRODUCTION Skeletal muscle provides unparalleled examples of the interrelationships between structure and function in a biological tissue. Moreover, recent advances in molecular genetics of both humans and experimental organisms have provided models of the perturbation of structure and function in inherited disease. Three distinct systems of molecular machinery are organized to interact with one another to produce a normally contracting muscle: 1. The sarcomere with its interdigitating arrays of protein filaments that actually generate tension and perform work. 2. The sarcoplasmic reticulum and transverse-tubule system with their specialized membranes that couple the electrical signals arising from neuronal excitation with the contraction of the sarcomeres. 3. The membrane-cytoskeleton system that anchors the other two systems to the plasma membrane and to the extracellular protein matrix. Remarkably, each of these systems is not only important for understanding muscle physiology but contains molecules whose genetic alteration is responsible for at least one specific human disease.

MYOFIBRIL STRUCTURE AND ASSEMBLY THE SARCOMERE Skeletal muscle is composed of long fibers that were formed by the fusion of committed cells during differentiation. Each multinucleated fiber contains many microfibers or myofibrils that may be considered the units for muscle contraction. Each myofibril contains repeating assemblies of myosin and actin filaments called sarcomeres (Fig. 93-1). The sarcomeres are the physical basis of the alternating dark and light bands seen in polarized light microscopy that give striated muscle its name. The Myosin Filament Myosin is the principal protein of the thick or myosin filaments of skeletal muscle. The sarcomeric myosin molecule of approximately 500,000 mol wt was the first multifunctional protein to be recognized, consisting of two globular head regions of about 115,000 mol wt each and a long rod-like tail region comprising the rest of the molecule. Each myosin molecule contains three pairs of distinctly encoded polypeptide chains: two heavy chains, two regulatory light chains, and two essential light chains. One of each kind of light chain and the globular From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

portion of each heavy chain constitute each myosin head. The myosin rod contains the α-helical regions of both heavy chains supercoiled about each other. About 300 myosin molecules assemble to form each myosin filament. The assembly has several interesting structural features. The filament is bipolar, divided in half along its long axis by a twofold axis of symmetry. In the central region, rods of myosin molecules from each half pack with another in an antiparallel manner. These interactions may represent the first steps in the assembly of the filaments. On each side of this zone, successive myosin molecules pack in parallel, staggered by about 14.5 nm, to form a right-handed helix. Myosin assembly continues until an exact filament length, 1550 nm, is reached, and then stops abruptly. This precise regulation of myosin assembly is not well understood either in molecular or physical–chemical terms. Several additional proteins are capable of complexing with myosin and become associated with the myosin filaments. The proteins C protein and titin are among the best studied examples. The 140-kDa C protein binds to seven sites in each half-filament. These sites are separated by approximately 44 nm, skipping two myosin repeats. This observation implies that different myosins must be in nonequivalent environments within the filament. Titin is a very large (3000-kDa) protein that creates its own filament which serves as a bridge between the myosin filament and the Z line. Titin shares many structural features with C protein, including myosin binding sites, and contains a protein kinase domain with potential specificity toward myosin light chains. Other titin domains appear to form elastic springs and interact with the Z structures. Myosin filaments are separately anchored by M structures (see below). The Actin Filament Actin is the principal protein of the thin or actin filaments of skeletal muscle. The actin polypeptide is 42 kDa, but it can reversibly polymerize to form long right-handed double helical filaments approximately 1.0 nm long. The assembly is coupled to the hydrolysis of ATP. The three-dimensional structure of the actin monomer has been determined, and it is a bilobed structure of 7 nm long, oriented in the direction of the polar filament axis (Fig. 93-2). The pitch of the filament axis is 36 nm with seven actin monomers per turn in each strand. Two additional proteins are major components of actin filaments, tropomyosin and troponin. Tropomyosin is a dimer of 35-kDa polypeptides that may be identical or closely similar isoforms. The two polypeptides are highly α-helical and supercoil about one another in a very similar manner to the coils of myosin rods. Tropomyosin dimers bind to the deep groove on either side

851

852

SECTION X / MUSCULOSKELETAL

Figure 93-1 Sarcomere of skeletal muscle. Note the thick filaments that are crosslinked at the M line and thin filaments that are crosslinked at the Z line. The M and Z lines are anchored by the cytoskeleton. The vertical lines visible on the thick filaments represent the positions of myosin heads (fine lines) and of myosin-associated C-type proteins (heavier lines at every third fine line). (Courtesy of Professor Hugh E. Huxley.)

of the actin helix. Troponin is a complex of three distinct polypeptides: troponin-T, troponin-I, and troponin-C. Troponin-T, 37 kDa, is elongated and interacts with tropomyosin. Troponin-C, 18 kDa, binds to calcium and is highly similar in three-dimensional structure to other small calcium-binding proteins, including calmodulin, parvalbumin, and the regulatory myosin light chains. Troponin-I, 24 kDa, is necessary for the formation of a functional complex with troponins T and C. The M and Z Structures Each sarcomere contains multiple myosin and actin filaments, each kind aligned in a regular array. Within each array, the filaments are in register without any stagger and are parallel to one another. The myosin filament array is central and is flanked on each side by actin filament arrays. The symmetry of the sarcomere is similar to that of the bipolar myosin filament. The precise alignments of these arrays require additional structures to crosslink the filaments. The myosin filaments are crosslinked in their central zones by M structures. Myomesin, a homolog of C protein and titin, is part of the M structure and is a 165-kDa myosin-binding protein. The actin filaments are crosslinked by Z structures at their distal ends. Alpha-actinin, a dimer of 100-kDa polypeptides, is an actin-binding protein, which can crosslink filaments and is a major component of the Z structure. Cap Z protein is also found in the Z structures and interacts with actin. It is likely that more proteins will be discovered in the sarcomere that function in the assembly or alignment of the myosin and actin filaments. THE SARCOPLASMIC RETICULUM AND TRANSVERSE TUBULE SYSTEM Because skeletal muscle fibers are large multinucleated syncytia up to 100 µM in diameter and millimeters to centimeters in length, a specialized membrane system, the transverse tubules and sarcoplasmic reticulum, couples the depolarizing signal originating from motor neurons to each of the myofibrils, even those deep within the fibers. For each of the myofibrils within the fibers and for each sarcomere unit of the myofibrils, there is a sleeve of membrane, the sarcoplasmic reticulum, which is connected via the transverse tubules to the outer plasma membrane of each fiber. The Calcium Pump The calcium pump molecule, a 115-kDa protein, is the principal protein of the sarcoplasmic reticulum (SR). This molecular pump translocates calcium from the myofibril to the lumen of the SR. Two calcium ions are translocated per ATP

Figure 93-2 Structure of a thin filament with actin and tropomyosin present that has been saturated with myosin heads (S1 fragments). Note the helical arrangements of the proteins (From Stryer L. Biochemistry, 4th ed., New York: W H Freeman, after drawing by Drs. Ronald Milligan and Paula Flicker, with permission.)

hydrolyzed by the pump. The active intermediate of the pump molecule is phosphorylated. The functioning of the pump is determined by the myofibrillar calcium concentration, the availability of ATP, and the sequestering capacity of the SR. This latter function relies on the 55-kDa protein calsequestrin, which binds 40–50 calcium ions at the relatively low affinity of 1 mM. The Calcium-Release Channel The calcium-release channel, also known as the ryanodine receptor, because the binding of that drug was critical to the original identification, is a giant oligmeric molecule composed of four 565-kDa subunits (Fig. 93-3). These channel complexes have multiple transmembrane domains that span the junctional sarcoplasmic reticulum membranes and foot domains that extend into the lumen between the transverse tubules and the sarcoplasmic reticulum. Additional proteins, including a 95-kDa junctional protein, may be important for the interaction of the calcium release channels and the voltagesensitive channels of the transverse tubules. Voltage-Sensitive Channel The voltage-sensitive channel of the transverse tubules is a multisubunit protein that was isolated through its interaction with the drug dihydropyridine. It was originally termed the dihydropyridine receptor and is also called the voltage sensor. The α1-subunit has a mass of 212 kDa and may be

CHAPTER 93 / SKELETAL MUSCLE

Figure 93-3 Interaction of the calcium release and voltage sensitive channels. The voltage-sensitive or sensor channel with its multiple subunits spans the transverse tubule (t-tubule) membranes and may interact with both the extracellular and cytoplasmic compartments. The calcium-release channel with its four giant subunits spans the sarcoplasmic reticulum (SR) membranes and may interact with the cytoplasm and the lumen of the SR. The voltage sensor and calciumrelease channels are shown physically in the SR lumen. (Courtesy of Professors Susan L. Hamilton and Wah Chiu.)

proteolytically cleared to a 175-kDa form. The higher molecular weight form is activated by phosphorylation, whereas the smaller form has lost three phosphorylation sites and the related activation. Additional 15-kDa α2, 55-kDa β, 30-kDa γ, and 27-kDa ε subunits complex with the α1 protein. The γ-subunits are glycoproteins. A 95-kDa junctional triadic protein may be necessary for functional interactions between the tubular voltage-sensitive channel and the reticular calcium release channel (Fig. 93-3). An interesting property of the tubular channel is that it permits calcium influx from extracellular fluids. It must be emphasized that this calcium is not directly involved in the release of calcium from the sarcoplasmic reticulum nor the subsequent activation of contraction. THE MEMBRANE CYTOSKELETON SYSTEM The regular organization of myofibrils, sarcoplasmic reticulum, and transverse tubules in skeletal muscles suggest the existence of an underlying framework. Furthermore, the contractile action of the myofibrils necessitates their anchoring to specific sites on the plasma membranes of their muscle fibers so that the force that they produce is transmitted to the tendons and then to bone. Both of these requirements may be met by the network of protein assemblies called the membrane cytoskeleton system. The importance of this system to muscle function is emphasized by the fact that there are at least three independent membrane associated systems that anchor the actin cytoskeleton of muscle. These systems, named after the first of their constituent proteins to be identified, include the dystrophin, spectrin, and integrin-based systems. The Dystrophin System The dystrophin system was discovered by its alteration in Duchenne-Becker muscular dystrophy. Dystrophin is actually a family of protein isoforms that are produced by alternative splicing of the transcript or alternative starts of transcription of the Duchenne genetic locus. The 427-kDa isoform functions as an integral part of a membrane-cytoskeleton system in skeletal muscle. Muscle dystrophin is actually a member of a larger superfamily of proteins that includes α actinin and the spectrins. These proteins

853

share an actin-binding domain near their amino terminals and a long coiled-coil rod domain. Nearer the carboxyl terminal, dystrophin has a special cysteine-rich domain and then a separate carboxyl-terminal domain. The cysteine-rich domain is responsible for anchoring dystrophin to a complex of transmembrane glycoproteins including the 25-kDa dystrophin-associated protein, the 35- and 43-kDa dystrophinassociated glycoproteins, the 50-kDa glycoprotein adhalin, and the 43-kDa β-dystroglycan (Fig. 93-4). This complex in turn binds the extracellular 156-kDa glycoprotein α-dystroglycan, which in turn binds the 40-kDa muscle-specific laminin isoform merosin. As other laminins, merosin is a component of the extracellular basal lamina and binds to lung fibrillar protein assemblies of the extracellular matrix. The carboxyl-terminal domain interacts with a complex of 50- to 60-kDa cytosolic proteins: α, β1, and β2 syntrophins. The syntrophins also bind to utrophin, a dystrophin homolog that may take part in a fourth membrane-cytoskeletal system and may functionally link the two systems. The Spectrin System The earliest membrane-cytoskeletal system to be discovered was the spectrin-based complex, originally in red blood cell membranes washed free of hemoglobin or ghosts (hence spectrin-ghost protein). Spectrin exists as a dimer of an ubiquitous 284-kDa α spectrin isoform and a 270-kDa β spectrin. The β spectrin amino-terminal domain, homologous to those of α actinin and dystrophin, binds actin filaments. Both spectrin subunits contain the characteristic, long, coiled-coil rod domain, and β spectrin contains a specific domain in the rod near the carboxyl terminal that binds the 215-kDa transmembrane protein ankyrin. A muscle-specific isoform is produced by alternative splicing in the carboxyl-terminal domain. Spectrin, ankyrin, and various cytoskeletal proteins are located at specific sites on the plasma membrane including costameres which are aligned with the myofibrils, and triadic junctions (and thus appear striated) and adhesion sites near muscle-tendon junctions. The Integrin System The integrins are a large family of protein isoforms that form tissue-specific complexes. They are transmembrane proteins that bind a variety of extracellular matrix proteins including various collagen isoforms. The integrins do not directly interact with the actin cytoskeleton, but are functionally linked to the protein vinculin and the cytoskeletal α-actinin isoform that binds actin filaments. In the experimental organisms, Caenorhabditis elegans and Drosophila, muscle-specific β integrin isoforms are localized near the sites of apposition of Z-band-like structures and the plasma membrane. Null or “knockout” mutations of this β integrin produces a lethal disruption of membrane and myofibril organization and even of filament assembly in either organism. In C. elegans, nulls for vinculin have the same deleterious effect. Thus, the integrin-based system may be even more fundamental to the organization of muscle structures, especially early in development, then either the dystrophin or spectrin-based systems. Recently developed "knockout" mutations in mice cannot even form mesoderm.

CONTRACTILE FUNCTION AND REGULATION FUNCTION The Sliding Filament Model Contraction in skeletal muscles occurs by the sliding of actin filaments past myosin filaments (Fig. 93-5). Since myosin filaments are bipolar, the heads on each half are oriented 180° from the other half. Therefore, in contraction, the myosin heads bind and pull the polar actin filaments toward their

854

SECTION X / MUSCULOSKELETAL

Figure 93-4 The dystrophin membrane cytoskeleton system. Dystrophin is shown interacting with the actin cytoskeleton, a complex of membrane-spanning proteins, and a complex of cytoplasmic proteins. Indirectly, dystrophin interacts with extracellular proteins and structures. This complex set of interactions is necessary for normal muscle development and function. (From Campbell KP. Cell 1995;80:675–679, with permission.)

Figure 93-5 Sliding of filaments. Note changes in polarity at the Z line and in each half of the sarcomere (From Stryer L. Biochemistry, 4th ed. New York: WH Freeman, with permission.)

center and that of the sarcomere at the M line. In relaxation, the actin filaments move away from the sarcomere center. In this model, the backbones of the myosin and actin filaments do not undergo significant structural changes during contraction. That is, the filaments themselves do not shrink or expand sufficiently to explain contraction and relaxation. The critical interactions and structural changes involve the association of the myosin heads and actin filaments. The Actomyosin ATP Cycle The interactions of myosin heads with actin and ATP are the critical molecular events in muscle contraction. Actin activates myosin to bind ATP and hydrolyze it to ADP and phosphate. This activation occurs by a cycle (Fig. 93-6). Since the binding of actin, ATP, and ADP to

Figure 93-6 The actomyosin-ATP cycle. An outline of the main biochemical events underlying muscle contraction. (From Stryer L. Biochemistry, 4th ed. New York: WH Freeman, with permission.)

myosin are mutually exclusive, the myosin must be placed into a special high-energy conformation (Fig. 93-7) with a finite lifetime following ATP hydrolysis. The highly transient complex of this special myosin state with actin is believed to be when actual force is generated by myosin upon actin. The cycle is completed when ADP dissociates from the actomyosin complex. The actual muscle, the 600 heads on each myosin filament, interact with 12 neighboring actin filaments. Each actin filament can interact with three myosin filaments. With this arrangement, the probability of myosin and actin finding each other is markedly enhanced over a random distribution. At any time during

CHAPTER 93 / SKELETAL MUSCLE

Figure 93-7 Mechanism of force generation. The biochemical events are matched to molecular structure and the power stroke of muscle contraction in this more detailed version of the actomyosinATP cycle. Note that movement occurs. (From Stryer L. Biochemistry, 4th ed. New York: WH Freeman, with permission.)

active contraction, every actin and myosin filament in the sarcomere is very likely to be participating in multiple actomyosin interactions. The Myosin Head as a Protein Motor In order to accomplish the sliding of filaments and the actomyosin-ATP cycle, the myosin head must be a very specifically designed protein motor. In fact, the heads tethered artificially without their rod domains and filament organization can induce motility and tension on free actin filaments. The myosin heads are truly the business ends of thick filaments.

855

The three-dimensional structure of myosin heads has been determined by X-ray diffraction of protein crystals (Fig. 93-8). The head itself is a 15-nm long structure. At one end, the head has two sides, each of which contains a special site. One site binds actin in a large cleft and the opposite site binds and hydrolyzes ATP. The sites are separated by 3–4 nm. Therefore, a structural change in the region between the sites must occur in order for the binding of ATP and actin to affect one another. The opening and closing of the actin binding cleft may be part of this mechanism (Fig. 93-9). The actual distance that a myosin head may move an actin filament has been measured. It can be as large as 10 nm. Where does this distance arise? The carboxyl terminal region of the heavy chain in the myosin head is a long α helix to which the myosin light chains attach (Fig. 93-8). This complex may serve as a lever arm where the fulcrum is either of the rod domain of the myosin molecule assembled into the backbone of a thick filament or an equivalent tether (Fig. 93-7). Evolutionarily, the myosin light chains had regulatory or modulatory functions with respect to motor functions. REGULATION The key steps in the regulation of muscle contraction are the coupling of the excitation of muscle to the release of calcium from the sarcoplasmic reticulum and the action of calcium on the interaction of myosin and actin filaments. The uptake of calcium from the myofibril into the sarcoplasmic reticulum is the mechanism directly controlling relaxation in skeletal muscle. Calcium Release and Uptake The contraction and relaxation of skeletal muscle ultimately is controlled by the action of motor neurons. Their release of acetylcholine presynaptically and the depolarization of the skeletal muscle plasma membrane following the binding of acetylcholine to specific receptors within specialized membrane sites postsynaptically induce muscle contraction. The absence of motor neuron input and the repolarization of the skeletal muscle plasma membrane are associated with muscle relaxation. The depolarization of the plasma membrane directly affects the voltage-sensitive channels of the transverse tubules. Direct molecular interaction between these channels and the foot domains of the calcium-release channels of the sarcoplasmic reticulum occurs in the junctional region (Fig. 93-3). Allosteric changes in this interaction as a result of transverse tubule membrane depolarization induces a change in conformation of the calcium-release channel. This channel switches from its closed to open state, and calcium is released from the sarcoplasmic reticulum into the myofibrils. A potential source of controversy or confusion, depending on one’s viewpoint, is that both the voltage-sensitive channel of the transverse tubules and the calcium-release channel of the sarcoplasmic reticulum are themselves activated by calcium. Theoretically, calcium ions could trigger calcium-release by their action on these two molecules. However, these effects are considered to be secondary to the allosteric mechanism discussed because extracellular and junctional calcium can be eliminated without inhibition. The allosteric interaction between the two channels is modulated by the action of other proteins. For example, activation of the voltage sensitive channels are markedly enhanced by phosphorylation of the major α1-subunits. Junctional proteins such as 95-kDa protein and other proteins of both the tubules and the reticulum may also influence the interactions of the two major channel proteins.

856

SECTION X / MUSCULOSKELETAL

Figure 93-8 Three-dimensional structure of myosin head. Note that the balls represent individual atoms derived from X-ray diffraction of crystals of proteolytically derived S1 or myosin head. RLC is the regulatory light chain; ELC is essential light chain. (From Stryer L. Biochemistry, 4th ed. New York: WH Freeman, adapted from Ivan Rayment, with permission.)

Figure 93-9 Detail of myosin head interacting with actin or thin filament. Note structural changes within myosin head and between myosin and actin. (From Stryer L. Biochemistry, 4th ed. New York: WH Freeman, after Rayment I, Holden HM, Whittaker M, Yohn CB, Lorenz M, Holmes KC, and Milligan RA, Science 1993;261:58–65, with permission.)

The uptake of calcium is controlled primarily by the calcium pump molecules of the sarcoplasmic reticulum. These molecules act as classical transport pumps; ATP is hydrolyzed by the pump in order to transport calcium ions into the reticulum lumen against an electro-chemical gradient. Secondarily, the binding of the transported calcium ions by the lumenal protein calsequestrin aids in the uptake. Essentially these processes are continuous; they are overwhelmed by the activated calcium release processes and predominate when release is deactivated. Regulation by Calcium of Actin and Myosin Interaction The release of calcium from the sarcoplasmic reticulum into the myofibrillar space activates the cycle of actomyosin-ATP interactions (Figs. 93-6 and 93-7) and the subsequent sliding of actin filaments past myosin filaments (Fig. 93-5). The key interaction of calcium is with the troponin complexes of the actin filaments. Each actin filament with its full complement of tropomyosin and troponin may be considered as a single allosteric unit (Fig. 93-2). The binding of calcium to troponin anywhere on the actin filament changes the potential interactions of all the actin monomers.

This remarkable effect of calcium binding by troponin is mediated by two levels of conformational change. First, calcium induces a change in the troponix complex: tropomyosin interaction. Second, the activated tropomyosins rotate about each actin strand. This rotation is propagated along both strands of actin up and down the filament. The net effect of these conformational changes is the opening of sites on actin subunits for interaction with myosin heads. The allosteric interactions of the actin filaments may also be sensitive to the binding of myosin heads. The uptake of calcium into the sarcoplasmic reticulum leads to reversal of actomyosin activation and the conformational changes in the actin filaments.

THE MYOFIBRIL IN HEALTH AND DISEASE Three areas of muscle biology are medically significant. First, the actin and myosin filaments of skeletal muscles represent the body’s largest store of amino acids that may be mobilized for metabolic needs. Skeletal muscle, therefore, plays a necessary role

CHAPTER 93 / SKELETAL MUSCLE

in nutrition and metabolism. Second, hormones and neuronal action control the metabolic functions of muscle. Skeletal muscle is a critical tissue in the effects of diabetes. Injury or dysfunction of motor neurons leads to skeletal muscle atrophy. Third, an increasing number of inherited diseases have been traced to specific proteins of the myofibril. Many of these disorders affect both heart and skeletal muscle, consistent with the sharing of both proteins and their functions between these tissues. MUSCLE PROTEINS AND NUTRITION Myosin and actin are the principal proteins of skeletal muscles. Although relatively stable, these proteins are continuously being degraded and synthesized. Since virtually all myosin and actin are assembled into filaments in muscle, the filaments must also be continuously assembled and disassembled. Skeletal muscle can degrade these proteins to free amino acids and, in certain cases, can oxidize the amino acids in order to synthesize ATP. The amino acids may be reutilized by muscle or may be secreted. Nearly every tissue of the body may take up amino acids and utilize them for protein synthesis. The liver takes up alanine and can utilize it for gluconeogenesis. The glucose may be used by the liver itself or be secreted and used by other tissues. A potential cycle exists between liver and muscle. Amino acids from myofibrillar protein breakdown are secreted by skeletal muscle, taken up by liver, and used in synthesizing glucose. The glucose is secreted, taken up by muscle, and used for contraction or stored as glycogen. Lactate from anaerobic muscle contraction may also be used by the liver for gluconeogenesis (Cori cycle). In addition to amino acids and carbohydrates, skeletal muscle oxidizes fatty acids to produce ATP. Well-nourished persons will use fatty acid oxidation for aerobic contraction and glycolysis/glycogenolysis for anaerobic work. The utilization of protein amino acid stores becomes significant under conditions of fasting or in uncontrolled diabetic states in which glucose and fat utilization are both impaired. Prolonged starvation shuts off protein degradation, and oxidation of ketone bodies derived from prior lipid oxidation predominates. HORMONAL AND NEURONAL CONTROL Insulin and Diabetes The major hormone that regulates both energetics and synthetic metabolism in skeletal muscle is insulin. Insulin stimulates uptake of glucose, amino acids, and fatty acids into skeletal muscle. The synthesis of glycogen and of myofibrillar proteins is markedly increased. In diabetes mellitus of either primary β-cell pancreatic origin or a result of insulin resistance, all of these positive effects of insulin can be decreased. In severe diabetic states, they are markedly reduced. Free fatty acids are mobilized to produce ketone bodies, and the oxidation of ketone bodies becomes a major contributor to energetics in muscle and other tissues that ordinarily require insulin. Neuronal Control Skeletal muscles require functional interactions with both motor neurons and specialized proprioceptive neurons of muscle spindles, not only for proper contractile function but also in the development and maintenance of myofibrillar proteins and structures. Muscle fibers are of two major types. Type I fibers are slow twitch, and their contraction is dependent on aerobic metabolism. Type II fibers are fast twitch, and the contraction is primarily dependent on glycogenolytic (anaerobic) metabolism. There are multiple differences in gene expression between the two fiber types, including metabolic enzymes, membrane channel protein isoforms, and myofibrillar protein isoforms. Most skeletal muscles in humans contain mixtures of the two fiber

857

types. Some fibers show intermediate properties and represent minor subtypes. The differentiation of and maintenance of Type I and II muscle fibers are dependent on the appropriate innervation by “slow” and “fast” motor neurons, respectively. Functional, pathological, or physical deinnervation of either type of muscle fiber leads to atrophy. Reinnervation will lead to development of a fiber of type specificity dependent upon the type of motor neuron. The mechanism of this dependence of skeletal muscle on nerve is not fully understood. In part, skeletal muscle may respond developmentally and metabolically to its own functioning. On the other hand, neurons actively secrete various protein factors and peptides in addition to neurotransmitters such as acetylcholine. Such factors include agrin, an extracellular protein that organizes acetylcholine receptors into special membrane sites and ciliary neurotrophic factor that may stimulate skeletal muscle development and growth more generally. Inherited Muscle Diseases Chapter 94 will discuss much of the exciting new research in inherited diseases of skeletal muscle. Also, inherited diseases of the heart may affect proteins shared with skeletal muscle. It is important to note here that there are genetic alterations of specific proteins of the myofibril, sarcoplasmic reticulum/transverse tubules, and membrane cytoskeleton that lead to alterations of skeletal muscle structure and function. Central core disease in which the myofibrillar compartment becomes selectively depleted in affected skeletal muscle fibers can be associated with specific mutations of the β-cardiac myosin heavy-chain isoform that is shared by slow skeletal muscle fibers. The responsible genetic locus is on human chromosome 14q. The skeletal muscle defects may be associated with or independent of the development of hypertrophic cardiomyopathy in these patients. Malignant hyperthermia in which volatile anesthetics produce uncontrollable muscular contraction and generation of heat during surgery is a result of mutations of sarcoplasmic reticulum calcium release channel. The responsible genetic locus for this disorder is on human chromosome 19q. Duchenne and Becker muscular dystrophies are produced by a variety of mutations in the dystrophin 427-kDa isoform (Fig. 93-4). These disorders affect males primarily because the responsible genetic locus is on human chromosome Xp2.1. As discussed, dystrophin participates in one of several protein systems that anchor cytoskeletal actin filaments to the membrane and the extracellular matrix in skeletal muscle. Although primarily thought of as skeletal muscle diseases, Duchenne-Becker dystrophy patients and even female carriers may have significant cardiomyopathies because of the shared expression of dystrophin.

SELECTED REFERENCES Alberts B, Bray D, Lewis J, Raff M, Roberts K, Watson JD. (Chapter 16 and Glossary). The Cytoskeleton. In: Molecular Biology of the Cell, 3rd ed. New York: Garland, 1994; p. 1294. Campbell KP. Membrane organization of the dystrophin-glycoprotein complex. Cell 1995;80:675–675. Catterall WA. Excitation-contraction coupling invertebrate skeletal muscle: a tale of two calcium channels. Cell 1991;64:871–874. Fleischer S, Inui M. Biochemistry and biophysics of excitation-contraction coupling. Ann Rev Biophys Chem 1989;18:333–364. Rayment I, Holden HM, Whittaker M, Yohn CB, Lorenz M, Holmes KC et al. Structure of the actin-myosin complex and its implications for muscle contraction. Science 1993;261:58–65. Stryer L. (Chapter 15). Molecular Motors. In: Biochemistry, 4th ed. New York: WH Freeman, 1995; p. 1064.

CHAPTER 94 / MUSCULAR DYSTROPHIES

94

859

Muscular Dystrophies ERIC P. HOFFMAN

OVERVIEW Mutations of nine different genes have been shown to be causative of specific types of muscular dystrophy, and linkage studies have shown the existence of approximately six additional genes. The most common cause of muscular dystrophy is mutations of the dystrophin gene on the X chromosome: absent dystrophin results in Duchenne muscular dystrophy in males (>95% of childhood onset lethal dystrophy in males), present but abnormal dystrophin in Becker muscular dystrophy (50% adult ambulatory dystrophy in males), and 10% of isolated female dystrophy patients show symptoms as a consequence of carrier status and skewed X-chromosome inactivation. The pathophysiology of dystrophin abnormalities involves an instability of the plasma membrane of myofibers with a chronic deterioration of the muscle leading to weakness. Myotonic dystrophy is the next most frequent dystrophy, and is extremely variable, showing presentations ranging from neonatal lethality, to childhood mental retardation, adult distal weakness with myotonia, or early onset cataracts. It is dominantly inherited, with considerable variation in phenotype within families. This disease is caused by an increase in size of a trinucleotide repeat (CTG) in the 3' untranslated region of a protein kinase. The size of the expansion generally correlates with disease severity. The mechanism by which the expansion mutation causes the observed multisystemic and variable clinical features is not well-understood. Very recently, mutations of six additional genes have been identified. Mutations of calpain III, a muscle-specific membrane-bound protease, may be a relatively common cause of dystrophy in patients with normal dystrophin. Four different genes encoding specific “sarcoglycans” (transmembrane proteins that are components of the dystrophin-based membrane cytoskeleton) have recently been shown to cause relatively rare cases of Duchenne/Beckerlike muscular dystrophy. Approximately 40% of neonatal onset disease, congenital muscular dystrophy, has been shown to be a result of deficiency of α2 laminin (merosin), a protein involved in attachment of myofibers to the extracellular matrix (basal lamina). In addition to profound weakness at birth, these patients show marked abnormalities of white matter by MRI studies, yet are intellectually normal. Finally, a rare X-linked recessive muscular dystrophy distinguished clinically from Duchenne/Becker muscular dystrophy by early prominent contractures and cardiac From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

conduction defects has recently been found to be caused by mutations of a gene of unknown function in Xq28 (emerin). Additional dystrophies for which genes have been genetically mapped but not yet identified include dominantly inherited limb-girdle muscular dystrophy (chromosome 2p), fascioscapulohumeral dystrophy (4q), and oculopharyngeal muscular dystrophy (15p).

INTRODUCTION Of the nine genes for which mutations have been shown to cause subtypes of muscular dystrophy, six result from abnormalities of the plasma membrane and adjacent basal lamina. Myofibers undoubtedly have a very stringent requirement for a durable and resilient plasma membrane based on two features of its normal physiology and function. First, each contraction of the myofibrils leads to a rapid and dramatic redistribution of the cytoplasm, resulting in changes of the shape of the myofiber. These shape changes would be expected to impart considerable stresses on the plasma membrane. Second, the force generated by the myofibrils must be transduced to the extracellular matrix so that the muscle group can contract and generate force as a single coordinated unit. Thus, this force transduction must be transmitted through the plasma membrane, which must be strongly reinforced to withstand these sheer forces. The study of human disease has been critical in both identifying components of the myofiber membrane cytoskeleton and in determining the relative functions of many of the components. To date, six distinct human inherited muscular dystrophies have been identified that are caused by deficiencies of the membrane cytoskeleton: one intracellular component (dystrophin; Duchenne muscular dystrophy), four transmembrane components (α, β, γ, and δ-sarcoglycans), and one extracellular component (α2 laminin). The muscular dystrophies caused by deficiency of dystrophin or any of the four sarcoglycans are quite similar histologically and clinically. This observation suggests that the progressive muscle diseases caused by these proteins may all share a similar pathogenesis, namely chronic membrane instability leading to progressive dysfunction of the muscle, and that all of the sarcoglycans and dystrophin are required for membrane stability. The deficiency of the extracellular component, α2 laminin, results in a more complicated and severe phenotype (congenital muscular dystrophy), with apparent involvement of the immune system in tissue pathology, and also abnormalities of myofiber regeneration.

859

860

SECTION X / MUSCULOSKELETAL

Figure 94-1 Schematic of proteins involved in muscular dystrophies. Shown is a segment of a myofiber plasma membrane, with the intracellular dystrophin-based membrane cytoskeleton, transmembrane sarcoglycans and dystroglycans, and extracellular basal lamina with laminin 2. Male Duchenne/Becker muscular dystrophy and isolated female dystrophinopathy patients are the result of abnormalities of the dystrophin protein, a common X-linked recessive disease. Rare autosomal recessive childhood-onset dystrophies are caused by mutations in any one of the sarcoglycan proteins (α sarcoglycan [adhalin], β sarcoglycan, γ sarcoglycan, and δ sarcoglycan). Many cases of neonatal onset severe congenital muscular dystrophy are caused by deficiency of α2 laminin (merosin), which is the large subunit of laminin in the myofiber basal lamina. (Reprinted with permission from Duggan et al. NEJM, 1997).

DISORDERS OF THE MEMBRANE CYTOSKELETON: DEFICIENCY OF DYSTROPHIN OR THE SARCOGLYCANS Duchenne muscular dystrophy is one of the most common inherited disorders of humans and is the most frequent muscle disease (1/3500 live born males). Its high frequency in all world populations is a result of the high mutation rate of the gene, which in turn is largely a consequence of the enormous size of the gene (2.5 million bp). Affected children are predominantly male, because of the X-linked recessive inheritance pattern, and appear normal until about 4 or 5 years of age, when muscle weakness becomes obvious. Thereafter, there is a relentless progression of weakness and muscle wasting, leading to confinement to a wheelchair by age 14 years, and death due to respiratory failure by age 20 years (unless the patient is mechanically ventilated). The biochemical deficiency causing the disorder remained obscure until 1987, when the Duchenne muscular dystrophy gene was identified by positional cloning techniques. The gene, which has remained 10 times larger than the next largest gene, encodes a 427-kDa protein dubbed “dystrophin,” and dystrophin-deficiency is recognized to define Duchenne muscular dystrophy at the biochemical level. Mutations of the gene are deletions (55%), duplications (5%), and point mutations (40%). Partial dystrophindeficiency, often the result of in-frame deletion mutations compatible with production of dystrophin that lacks segments of amino acids, is diagnostic of the milder and more variable Becker muscular dystrophy. Presentations of Becker dystrophy include proximal weakness, myoglobinuria, myalgias, hyperCKemia, and cardiac failure, and some patients may remain asymptomatic at advanced ages. Female homozygotes do not exist because of the relative rarity of the disorder; females generally act as heterozygous gene carriers. However, approximately 10% of isolated female dystrophy patients express their symptoms as a consequence of heterozygosity for a dystrophinopathy. Most of these females show preferen-

tial inactivation of the X chromosome containing the normal dystrophin gene (skewed X inactivation). Presentations of isolated female carriers are as variable as in Becker muscular dystrophy in males, although the progression is often mitigated in the female patients. Molecular diagnosis, genetic counseling, and prenatal diagnosis are well established in the dystrophinopathies. Dystrophin protein analysis of muscle biopsy, gene deletion mutation analysis, and gene linkage analysis are all routinely used. The large size of the gene complicates molecular genetic counseling and prenatal diagnosis: There is 10% recombination within the gene, and the high mutation rate leads to many “new mutations,” which may effect a single egg or sperm or may cause the gonads of a parent to be a carrier but not the parent him/herself (gonadal mosaicism). The identification of dystrophin as a major component of the membrane cytoskeleton of muscle fibers permitted use of the dystrophin protein as a probe to isolate and characterize other components. Recent research has identified three complexes of proteins that interact with dystrophin near the plasma membrane and appear to provide one type of connection between the intracellular contractile myofibrils and the extracellular matrix (Fig. 94-1). Rare childhood-onset muscular dystrophies with autosomal recessive inheritance patterns have been shown to be caused by mutations of four of the sarcoglycan proteins (Table 94-1; Fig. 94-1). No disorders have been associated with the primary defects in any of the syntrophins or dystroglycan components. The clinical and histopathological consequences of primary disorders of dystrophin and the sarcoglycans are quite similar, and it can be safely assumed that they share a common molecular pathogenesis. Current models suggest that deficiencies of these proteins lead to membrane instability, which initiates both acute responses (hyperCKemia, myofiber necrosis) and chronic responses (progressive fibrosis, failure of myofiber regeneration, and muscle wasting/weakness). The immune system may play a role in both the acute and chronic responses, and immune suppressants such as prednisone have shown some effect in slowing the progression of

Table 94-1 Molecular Basis of the Muscular Dystrophies Protein defect

Disease

Onset of symptoms

Frequency

Dystrophin-deficiency

Duchenne muscular dystrophy

Childhood

1/3500 males (>90% of childhoodonset dystrophy)

Abnormal dystrophin

Becker muscular dystrophy

Childhood-adult

Dystrophin mosaicism

Isolated female dystrophinopathy

Childhood-adult

Childhood

β Sarcoglycan

SCARMD/LGMDa

Childhood

25dB Iris pigmentary abnormality Heterochromia: eyes of different colors; or clearly demarcated segments of different colors in one eye; Isochromia: characteristic brilliant blue eyes White forelock in hair Dystopia canthorum Affected first-degree relative. Several areas of congenitally hypopigmented skin Medial flaring of eyebrows with synophrys Broad high nasal root Hypoplasia of the alae nasi Graying of the hair before age 30

WS2, %

57

77

20 15 45 98

50 25 25 Absent

35 65 75 Common Common

10 Rare Rare Rare 25

For Waardenburg syndrome, the Waardenburg Consortium defined an affected person as someone having two major, or one major and two minor criteria. The criteria are given in Table 112-2, along with estimated penetrances in WS1 and WS2, from the study of Liu et al. The apparently somewhat higher incidence of hearing loss and heterochromia in WS2 is probably an artefact of ascertainment: Without dystopia as a guide, patients are not diagnosed unless they show several other features of WS. DIFFERENTIAL DIAGNOSIS OF TYPE 1 AND TYPE 2 WAARDENBURG SYNDROME Because dystopia is specific and highly penetrant, WS1 can be diagnosed with confidence and shows little or no genetic heterogeneity. The inner canthi are displaced outward so that the palpebral fissures are short, with little of the sclerae showing on the medial side of the pupil; the inferior lacrimal punctae lie in front of the cornea, and there is a broad high nasal root with synophrys (confluent eyebrows). Experience shows that dystopia canthorum needs to be assessed by measuring the inner canthal, interpupillary, and outer canthal distances (a,b,c respectively, in millimeters) with a rigid ruler held in front of the head, and then applying the formula

TYPE 2 WS AND OTHER RELATED SYNDROMES WS2 is not homogeneous. Clinical definition is arbitrary, and molecular studies show heterogeneity. About 20% of families have mutations in the MITF gene (see below). Rare patients with Hirschsprungs disease plus features of WS2 (and thus describable as WS4) have an endothelin defect (see below). Many patients with auditory-pigmentary syndromes do not fall into any of the categories described above, and many other illdefined auditory-pigmentary syndromes are listed in the literature, but little is known of the underlying genetics. In thinking about their clinical classification, it is useful to try to distinguish melanocyte-specific defects from more general neurocristopathies. Neurocristopathies arise early in embryonic development, whereas death of melanocytes can occur at any time. Nevertheless, this distinction is not always clear-cut. WS1 is mainly caused by dysfunction of the embryonic neural crest, but people with WS1 also often suffer premature graying of their hair 15–30 years after their neural crest completed its differentiation.

W = X + Y + a/b where X = (2a – 0.2119c – 3.909)/c and Y = (2a – 0.2479b – 3.909)/b

TYPE 1 WAARDENBURG SYNDROME Linkage to 2q35 and the Sp Mouse Model WS1 is relatively homogeneous genetically. Linkage was first demonstrated to the marker ALPP (alkaline phosphatase, placental isozyme) in 1989. ALPP maps to the distal long arm of chromosome 2. Recent collaborative studies using a microsatellite polymorphism from the PAX3 gene suggest that all WS1 but no WS2 maps to this region. The map position was originally claimed to be 2q37, but more recent results suggest 2q35-q36. This region shows strongly conserved synteny with the proximal part of mouse chromosome 1, and a search of the corresponding mouse map suggested a likely homolog, Splotch. The heterozygous Sp mouse has no hearing loss, but nor do many

This rather abstruse measure (the numbers come from a discriminant analysis) has proved extremely effective for distinguishing WS1 (W > 1.95) from WS2 (W < 1.95). By contrast, clinical impression, even by skilled judges, has proved unreliable. The distinction is important to make because, as shown below, it holds the key to molecular diagnosis.

GENETIC BASIS OF WAARDENBURG SYNDROME

CHAPTER 112 / WAARDENBURG SYNDROME

1017

Figure 112-2 PAX3 mutations seen in Manchester patients with WS1. The eight exons of the PAX3 gene are shown, with the paired box in exons 2–4 and the homeobox in exons 5–6 shaded. Mutations expected to truncate the gene product are above the gene diagram, those expected to produce a full-length product are below. Note that, whereas truncating mutations are distributed across most of the gene, amino acid substitutions are pathogenic only when located in the 5' part of the paired box and the 3' part of the homeobox. These regions are critical for DNA binding by the PAX3 protein.

Figure 112-1 Facial appearance in (A) Type 1 and (B) Type 2 Waardenburg syndrome. Note dystopia canthorum in the patient in Panel A. This patient has a chromosomal deletion del(2)(q34q36.2) and is mentally retarded in addition to showing typical features of Type 1 WS. Neither Type 1 nor Type 2 WS is normally associated with any mental retardation.

humans with WS1, and phenotypes of orthologous mouse and human mutants are often somewhat different. On some genetic backgrounds, Sp mice have a broad snout, probably corresponding to dystopia canthorum in man. The PAX3/pax-3 Gene A candidate gene, pax-3, mapped to this region in the mouse and was expressed in the developing nervous system of the early embryo in a pattern compatible with the pathology, both of Sp and of WS1. The human homolog is PAX3, originally called HuP2. Searches revealed mutations in both Sp and WS1. PAX3 is a very interesting gene. PAX genes, in both mouse and man, form a family of nine closely related genes. All share a paired box, a highly conserved 128 amino acid DNA-binding domain found in organisms from Drosophila to man. As well as the paired box, PAX3, PAX6, and PAX7 have a homeobox. All the PAX genes are believed to encode transcription factors; the expression pattern of PAX3 in the early embryo is described below. PAX3 was the first human homeobox gene shown to carry mutations causing inherited disease. Other PAX Genes in Human Syndromes PAX6 mutations cause aniridia and other abnormalities of the anterior chamber of the eye. Mutations in PAX2 have been seen in a few families with combinations of eye (e.g., optic nerve coloboma) and kidney problems. The molecular pathology of all these conditions appears to involve

haploinsufficiency. In each case there is a good mouse homolog, sey (small eye) and krd (kidney and retinal defects), respectively. TYPE 2 WAARDENBURG SYNDROME In perhaps 20% of WS2 families the mutation maps to the proximal short arm of chromosome 3, and in some of these families mutations have been demonstrated in MITF, the human homolog of the mouse microphthalmia gene. The human MITF and mouse mi genes encode another transcription factor, a basic helix-loop-helix leucine zipper protein. Only about 20% of families with WS2 show evidence (by linkage or mutation screening) of involvement of MITF; the gene(s) responsible for 80% of WS2 have yet to be mapped or identified. TYPE 3 WAARDENBURG SYNDROME In a few cases, PAX3 mutations have been described. PAX3 is expressed in the developing limb muscles, and WS3 is probably a variant presentation of WS1. TYPE 4 WAARDENBURG SYNDROME Mutations in the endothelin 3 or endothelin receptor B genes (on chromosomes 20 or 13, respectively) have recently been described in well-characterized affected patients with a combination of WS2-like pigmentary abnormalities and Hirschsprung disease.

MOLECULAR PATHOPHYSIOLOGY EXPRESSION OF PAX3 In the mouse, pax-3 is expressed in the developing nervous system starting at d 8–8.5 postcoitum. Expression occurs along the length of the neural tube, but only in the posterior part, including the neural crest. Expression is seen in various parts of the developing brain, and pax-3 expressing cells migrate into the limb buds. The patterns of expression are described in more detail in the reviews cited at the end of this chapter. Preliminary studies in human embryos reveal a similar pattern. The expression pattern fits well with the observed features of WS1 and Sp, although as is usual in these cases, only a subset of expressing tissues is affected. MOLECULAR PATHOLOGY OF PAX3 MUTATIONS PAX3 mutations have been found in many families with WS1, but not in families with WS2. Figure 112-2 shows the mutations that

1018

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

MANAGEMENT AND TREATMENT OF WAARDENBURG SYNDROME Apart from possible early graying of the hair, all features of Waardenburg syndrome are nonprogressive, and treatment is symptomatic. Early screening is important to identify hearing loss in infants.

FUTURE DIRECTIONS

Figure 112-3 MITF mutations seen in Manchester patients with WS2. The gene has nine exons and encodes a DNA-binding basic domain (hatched in exon 6) and helix-loop-helix and leucine zipper dimerization domains (shaded, in exons 8–9). Mutations expected to truncate the gene product are above the gene diagram, those expected to produce a full-length product are below.

have been described in Manchester; similar mutations have been found by other groups within the Waardenburg Consortium. It is noteworthy that the mutations are very heterogeneous, including a complete deletion of the gene, truncating (nonsense or frameshifting) changes spread through exons 2–6, splice site mutations and amino acid substitutions. The substitutions are all of amino acids in the 5' part of the paired domain or the third (recognition) helix of the homeodomain. All of the amino acids involved are highly conserved across species and are known to take part in critical proteinDNA contacts in other PAX and homeodomain genes. All of these changes produce the same clinical result, WS1. We conclude that WS1 is caused by a dosage effect in people heterozygous for a loss-of-function mutation. As deaf people frequently marry, homozygotes may occur, and one case has been described. The child had severe depigmentation and limb contractures, but unexpectedly, considering the phenotype of homozygous Splotch mice, no neural tube defect. Nevertheless, it is likely that most PAX3 mutations that cause WS1 in heterozygous people would be lethal in homozygotes. Substitutions of asparagine 47 in the PAX3 paired domain may possibly produce a special phenotype. The mutation N47H was seen in the only known example of familial WS3, whereas a different substitution of the same amino acid, N47K, occurred in a family whose phenotype was described as craniofacial-deafnesshand syndrome (MIM 122880). Conceivably, these mutations represent a gain of function. Unambiguous (but somatic, not inherited) gain of function mutations of PAX3 are seen in the pediatric tumor alveolar rhabdomyosarcoma. In these tumors, a chromosomal translocation t(2;13)(q35;q14) creates a novel chimeric gene out of the 5' parts of PAX3 and the 3' part of another transcription factor gene, FKHR. MOLECULAR PATHOLOGY OF MITF MUTATIONS Figure 112-3 shows human MITF mutations described in patients with WS2. All of these cases are dominant. In the mouse, both dominant and recessive alleles of mi are known. The difference can be explained by dominant negative effects of mutants that sequester the product of the normal allele in inactive dimers. Some of the human WS2 mutations would be predicted to be recessive in mice; however, maybe human homozygotes would be much more severely affected, as in mice. Transcripts and protein products of these mutant genes have not yet been studied.

For the known genes, PAX3 and MITF, the major problem is the inability to predict the severity of symptoms of WS1 or WS2. Both conditions are variable even within families, so that severity must depend on the action of unidentified modifier genes. This makes genetic counseling difficult. Identifying the modifier genes and identifying the genes responsible for the bulk of WS2 are priorities. Although PAX3 and MITF are known to be transcription factors, very little is known about their upstream regulators or downstream targets in the embryo. Knowing these might help identify genes for other auditory-pigmentary syndromes, as well as shedding light on an important part of embryonic development. Gene therapy for WS seems problematic. PAX3 and MITF have roles in basic pattern formation and differentiation in the embryonic neural crest. This makes them biologically interesting genes, but also makes it difficult to imagine effective postnatal gene therapy.

SELECTED REFERENCES Asher JH, Harrison RW, Morell R, Carey ML, Friedman TB. Effects of Pax3 modifier genes on craniofacial morphology, pigmentation, and viability: a murine model of Waardenburg syndrome variation. Genomics 1996;34:285–298. Asher JH, Sommer A, Morell R, Friedman TB. Missense mutation in the paired domain of PAX3 causes craniofacial-deafness-hand syndrome. Hum Mutat 1996;7:30–35. Baldwin CT, Hoth CF, Amos JA, da-Silva EO, Milunsky A. An exonic mutation in the HuP2 paired domain gene causes Waardenburg’s syndrome. Nature 1992;355:637–638. Barsh GS. Pigmentation, pleiotropy and genetic pathways in humans and mice. Am J Hum Genet 1995;57:743–747. Chakravarti A. Endothelin receptor-mediated signaling in Hirschsprung disease. Hum Mol Genet 1996;5:303–307. Epstein DJ, Vekemans M, Gros P. Splotch (Sp2H), a mutation affecting development of the mouse neural tube, shows a deletion within the paired homeodomain of Pax-3. Cell 1991;67:767–774. Farrer LA, Grundfast KM, Amos J, et al. Waardenburg syndrome (WS) type 1 is caused by defects at multiple loci, one of which is near ALPP on chromosome 2: first report of the WS Consortium. Am J Hum Genet 1992;50:902–913. Foy C, Newton VE, Wellesley D, Harris R, Read AP. Assignment of WS1 locus to human 2q37 and possible homology between Waardenburg syndrome and the Splotch mouse. Am J Hum Genet 1990;46:1017–1023. Galili N, Davis RJ, Fredericks WJ, et al. Fusion of a fork head domain gene to PAX3 in the solid tumour alveolar rhabdomyosarcoma. Nat Genet 1993;5:230–235. Hanson I, Fletcher JM, Jordan T, et al. Mutations at the PAX6 locus are found in heterogeneous anterior segment malformations including Peters’ anomaly. Nat Genet 1994;6:168–173. Hemesath TJ, Steingrimsson E, McGill G, et al. Microphthalmia, a critical factor in melanocyte development, defines a discrete transcription factor family. Genes Dev 1994;8:2770–2780. Hofstra RM, Osinga J, Tan-Sindhunata G, et al. A homozygous mutation in the endothelin-3 gene associated with a combined Waardenburg type 2 and Hirschsprung phenotype (Shah-Waardenburg syndrome). Nat Genet 1996;12:445–447. Hoth CF, Milunsky A, Lipsky N, Sheffer R, Clarren SK, Baldwin CT. Mutations in the paired domain of the human PAX3 gene cause KleinWaardenburg syndrome (WS-III) as well as Waardenburg syndrome Type 1 (WS-1). Am J Hum Genet 1993;52:455–462.

CHAPTER 112 / WAARDENBURG SYNDROME

Klein D. Historical background and evidence for dominant inheritance of the Klein-Waardenburg syndrome (Type III). Am J Med Genet 1983;14:231–239. Liu XZ, Newton VE, Read AP. Waardenburg syndrome Type 2: phenotypic findings and diagnostic criteria. Am J Med Genet 1995;55:95–100. Nobukuni Y, Watanabe A, Takeda K, Skarka H, Tachibana M. Analyses of loss-of-function mutations of the MITF gene suggest that haploinsufficiency is a cause of Waardenburg syndrome type 2A. Am J Hum Genet 1996;59:76–83. Pandya A, Xia XJ, Landa BL, et al. Phenotypic variation in Waardenburg syndrome: mutational heterogeneity, modifier genes or polygenic background? Hum Mol Genet 1996;5:497–502. Puffenberger EG, Hosoda K, Washington SS, et al. A missense mutation of the endothelin-B receptor gene in multigenic Hirschsprung’s disease. Cell 1994;79:1257–1266. Sanyanusin P, Schimmenti LA, McNoe LA, et al. Mutation of the PAX2 gene in a family with optic nerve colobomas, renal anomalies and vesicoureteral reflux. Nat Genet 1995;9:358–363. Steel KP, Barkway C. Another role for melanocytes: their importance for normal stria vascularis development in the mammalian inner ear. Development 1989;107:453–463. Steel KP, Bock GR. Hereditary inner-ear abnormalities in animals: relationship with human abnormalities. Arch Otolaryngol 1983;109:22–29.

1019

Steingrimsson E, Moore KJ, Lamoreux ML, et al. Molecular basis of mouse microphthalmia (mi) mutations helps explain their developmental and phenotypic consequences. Nat Genet 1994;8:256–263. Strachan T, Read AP. PAX genes. Curr Opin Genet Dev 1994;4:427–438. Stuart ET, Kioussi C, Gruss P. Mammalian PAX genes. Annu Rev Genet 1994;28:219–236. Tassabehji M, Newton VE, Liu XZ, et al. The mutational spectrum in Waardenburg syndrome. Hum Mol Genet 1995;4:2131–2137. Tassabehji M, Newton VE, Read AP. Waardenburg syndrome type 2 caused by mutations in the human microphthalmia (MITF) gene. Nat Genet 1994;8:251–255. Tassabehji M, Read AP, Newton VE, et al. Waardenburg syndrome patients have mutations in the human homologue of the Pax-3 paired box gene. Nature 1992;355:635,636. Waardenburg PJ. A new syndrome combining developmental anomalies of the eyelids, eyebrows and nose root with pigmentary defects of the iris and head hair and with congenital deafness. Am J Hum Genet 1951;3:195–253. Xu W, Rould MA, Jun S, Desplan C, Pabo CO. Crystal structure of a paired domain-DNA complex at 2.5Å resolution reveals structural basis for Pax developmental mutations. Cell 1995;80:639–650. Zlotogora J, Lerer I, Bar-David S, Ergaz Z, Abielovich D. Homozygosity for Waardenburg syndrome. Am J Hum Genet 1995;56:1173–1178.

CHAPTER 113 / GREIG SYNDROME AND LIMB DISORDERS

113

1021

Greig Cephalopolysyndactyly Syndrome and Limb Disorders KARL-HEINZ GRZESCHIK

INTRODUCTION In the process of dissecting the factors governing the formation of hands and feet from the limb buds, human genetic disorders in which polydactyly or syndactyly occurs can provide clues to the genes involved in separating and defining the outline of limb elements in normal development. The identification by the positional cloning strategy of a gene believed to be responsible for Greig cephylopolysyndactyly syndrome, a rare developmental disorder, characterized by craniofacial abnormalities and malformations of hands and feet, can serve as a paradigm for the value of this approach, since it drew attention to GLI3, a gene not previously expected to be involved in regulation of human development. Greig syndrome appears to result from haploinsufficiency of this gene. GLI3 is a member of a zinc finger gene family that includes Krüppel, a gene that was known to regulate development in Drosophila. Nevertheless, the involvement of a zinc finger gene in mammalian development had not been demonstrated earlier. The identification of a candidate gene involved in human limb development called for the application of the powerful methods of genetic dissection of this gene in the mouse. Such a combination of candidate gene search in human developmental malformations and gene analysis in animal model systems holds the promise to teach how genes develop the human gestalt.

CLINICAL FEATURES OF GREIG CEPHALOPOLYSYNDACTYLY SYNDROME Greig cephalopolysyndactyly syndrome (GCPS, OMIM 175700) is a rare developmental disorder characterized by craniofacial abnormalities and postaxial polydactyly of the hands, preaxial polydactyly of the feet, and syndactyly of the fingers and toes. It shows autosomal dominant inheritance with full penetrance but with marked inter- and intrafamilial variability. Characteristic limb and skull manifestations observed in more than 50 patients are listed in Table 113-1. Hand and foot malformations are mostly bilateral. If unilateral, a hand is more often affected than a foot. Postaxial polydactyly is common in the hands and preaxial polydactyly in the feet: The thumbs are frequently broad with a broad nail and a malformed distal phalanx, sometimes From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

with a bifid tip. Radiographs of the hands can reveal preaxial polydactyly that is not noted on physical examination. Postaxial polydactyly of the hands usually manifests as a pedunculated postminimus, sometimes merely as a small cutaneous tag, with or without hypoplastic nail and a bony fragment as phalanx. The additional finger articulates at the ulnar side of the hand, mostly at the proximal or middle phalanx of the fifth finger; the fifth metacarpal bone is not affected (Fig. 113-1A). In the feet, preaxial polysyndactyly is the most common malformation, including duplication of the distal phalanx of the hallux, of both phalanges, of both phalanges and metatarsus, and rarely of metatarsus alone (Fig. 113-1B). Broad halluces with broad nails are common. The fifth metatarsal bone, if not duplicated, is often broad and markedly altered in shape. Postaxial polydactyly of the feet has been observed occasionally resulting from duplication of the fifth toe (eventually only a single bony fragment is found) and broadening or duplication of the fifth metatarsal bone. Also in the feet, polydactyly sometimes was noted on radiographs only. Syndactyly affects hands and feet as a constant feature but with some variability in its extent. It is mostly cutaneous in the feet, varying from mild webbing between the first two or three toes to complete cutaneous fusion, sometimes also with nail fusion. In the hands, bony fusion was reported on the distal phalanges of third to fourth fingers and of third to fifth fingers. The craniofacial manifestations are variable and consist of a macrocephaly with a broad prominent forehead, a broad nasal root and brachycephaly (Fig. 113-1C). In individuals with large crania, the interpupillary distance will often be increased. The impression of hypertelorism is confirmed only rarely if canthal indices are used for correction. Prognosis for the mental and physical development of affected individuals is good. Advanced bone age seems to be a frequent finding in GCPS, suggesting an influence of the GCPS gene on bone development.

DIFFERENTIAL DIAGNOSIS The wide variability of manifestations raised the question of whether or not other genetic entities thought to be distinct syndromes are not merely variants of the GCPS. Acrocallosal syndrome (ACLS, OMIM 200990) is an autosomal recessive condition that shares manifestations with GCPS. However, neurological findings such as the severe mental retardation, hypotonia, and absence of corpus callosum in ACLS, are very

1021

1022

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Table 113-1 Skull and Limb Manifestations in GCPS Patientsa Malformation Craniofacial Macrocephaly High, broad forehead Frontal bossing Broad nasal bridge Hands Preaxial polydactyly Clinical Radiological Postaxial polydactyly Clinical Radiological Broad thumbs Syndactyly Feet Preaxial polydactyly Clinical Radiological Postaxial polydactyly Clinical Radiological Broad halluces Syndactyly

Number observed 23/44 25/36 21/36 30/38 4/41 8/22 36/46 10/22 37/40 33/40 40/49 21/26 3/39 3/26 26/29 43/47

aData

compiled from Ausems et al. 1994, according to references described therein. When radiographs were performed after surgical treatment in patients with polydactyly, this item is scored clinically (+) and radiologically (–).

rare in GCPS. Also, the type of polydactyly differs and the syndactyly is less severe in ACLS, generally affecting only the proximal portion between the second and third toes. Since no linkage was found on chromosome 7 in patients with acrocallosal syndrome, it is very unlikely that acrocallosal syndrome and Greig syndrome are allelic disorders. Greig syndrome can be so mild as to be indistinguishable from the normal. Therefore, preaxial polydactyly type IV, or uncomplicated polysyndactyly (OMIM 174700), a dominant trait characterized by broad duplicated thumbs that are radially deviated and duplication of the halluces or second toes, with syndactyly of the second and third toes, may be Greig syndrome. Also, crossed polydactyly, a condition with autosomal dominant inheritance, characterized by polydactyly of the hands and feet with discrepancy of the axes of polydactyly between hands and feet, which occasionally shows syndactyly but no associated craniofacial abnormalities might be Greig syndrome.

THE MOLECULAR GENETICS OF GCPS IDENTIFICATION OF A GCPS CANDIDATE GENE The genetic locus of GCPS had been assigned to chromosome 7p13 by three balanced translocations associated with GCPS in different families with one chromosome breakpoint in this band. This assignment was corroborated by the detection of three sporadic cases carrying overlapping deletions in 7p13, as well as by tight linkage of GCPS to the epidermal growth factor receptor gene in 7p13–p12. In a positional cloning effort, the translocation breakpoint region was covered with a contiguous restriction map employing

arbitrary DNA fragments and the cDNA of the gene GLI3 as hybridization probes. GLI3 had been isolated by homology to the zinc-finger gene GLI, which is amplified in gliomas, and mapped to 7p13. It was considered a potential candidate gene for Greig syndrome, both on the basis of its map position and its homology to the Drosophila gene Krüppel, which is involved in the regulation of development. Since the GCPS translocation breakpoints were found to interrupt the genomic segment coding for the distal half of the GLI3 gene, and since the interstitial deletions of 7p observed in three Greig patients also eliminated the GLI3 gene region, it could be postulated that haploinsufficiency of GLI3 might evoke the developmental malformations in at least some of the Greig syndrome patients. The size of the GLI3 gene could be determined as at least 280 kb. The transcribed GLI3 sequence is composed of 15 exons with the most 3' exon spanning about 2.5 kb of continuously transcribed DNA. The gene is flanked by a CpG island that lies on the 5' side and that is in close proximity to the first exon detected by the cloned GLI3 cDNA. The presence of other genes in this region that might be affected by deletions or positional effects resulting from the translocation events, however, still cannot be excluded. Two large introns of 70 kb each in the 5' region of the GLI3 gene would leave ample room for additional genes. PROPERTIES OF GLI3 The GLI3 gene belongs to the human GLI-Krüppel gene family. It encodes a zinc-finger protein of 1596 amino acids and an apparent molecular mass of 190 kDa. Multiple regions of sequence similarity aside from the zinc-finger region suggest that other aspects of function are shared among the members of this gene family of DNA-binding proteins. The gene is expressed in a wide variety of human adult tissues. Because of their ability to bind specific DNA sequences, zinc-finger proteins are generally considered to be transcription factors. The high similarity between the binding specificities of the members of this family suggests, however, that these genes may assume their specific function only through their developmentally regulated expression or interaction of other nonhomologous protein domains with additional cellular factors.

ANIMAL MODELS FOR GCPS Additional evidence for the influence of GLI3 on the embryonal development of limbs and skull is derived from the allelic mouse mutations extra toes (Xt) anterior digit deformity (add), and possibly also from the mouse mutant Polydactyly Nagoya (Pdn). The spontaneous semidominant mouse mutation extra toes affects limb development with almost complete penetrance in heterozygotes but variable expressivity (Fig. 113-2). Several Xt alleles have been reported since and mapped to a region on mouse chromosome 13 homologous to human chromosome band 7p13. The Xt mutation in a (C3Hx101)F2 mouse was shown to be a deletion of at least 80 kb beginning at the 5' side of the GLi3 gene. In a second Xt allele, XtJ, a more downstream deletion was detected extending from the first zinc-finger motif towards the 3' end of the gene spanning at least 30 kb. The Xt heterozygotes show phenotypic peculiarities similar to GCPS patients: predominantly preaxial polydactyly of the hind limbs and a postaxial nubbin on the forelimbs. They also often have an enlarged interfrontal bone and some show hydrocephaly. Brachyphalangy (Xtbph) is a radiation-induced allele that has similar effects to Xt but is distinguishable in both heterozygotes and homozygotes. Unlike Xt heterozygotes, Xtbph heterozygotes usu-

CHAPTER 113 / GREIG SYNDROME AND LIMB DISORDERS

1023

Figure 113-1 (continued) (C) Head from patient with Greig Cephalopolysyndactyly Syndrome. (Reproduced with permission from Gollop and Fontes, Am J Med Genet 1985;22:59–68.)

Figure 113-1 (A) Right hand, (B) foot, from patients with Greig Cephalopolysyndactyly Syndrome. (Reproduced with permission from Gollop and Fontes, Am J Med Genet 1985;22:59–68.)

ally have an abnormal sternum and occasionally show syndactyly. All Xt alleles are homozygous lethal with more extreme limb and craniofacial defects as well as additional abnormalities of the brain, the central nervous system, and sense organs. The Pdn mouse mutation is very likely to be allelic to Xt, but no effects on genomic structure nor expression of GLI3 could be demonstrated yet. The recessive mouse mutant add is an allele of the gene affected in Xt. This mutant was generated by insertional mutagenesis in mice. The add mutation originates from a transgene insertion within the segment deleted in Xt(C3Hx101)F2 at a maximal distance of about 44 kb from the transcription start of Gli3. The add insertion replaced only 1 bp of the wild-type sequence. Since the GLi3

RNA expression level in homozygous add/add mice showed a reduction similar to that observed in heterozygous Xt/+ mice but no alteration of the transcript size, it can be postulated that the insertion of foreign material disturbed a region important for transcriptional control of Gli3. In homozygous add/add mice the forelimb is disorganized. The morphology of the thumbs is always altered, and sometimes the adjacent finger has an extra phalanx. The effect of the Xt and add mutants on the phenotype of the feet is summarized in Table 113-2. The anterior digit duplications in extra-toes mutants are similar to the phenotype of homozygous ciw Drosophila flies, which are caused by a mutation in the Drosophila Gli protein cubitus interruptus (Ci). This homology in function throughout the animal kingdom suggests that Gli proteins are part of a very basic, conserved signaling pathway.

MOLECULAR PATHOPHYSIOLOGY OF GLI3 DEFICIENCY GLI3 AS A TRANSCRIPTIONAL REGULATOR Based on the phenotypic similarities of Xt heterozygotes with GCPS and the molecular evidence that homologous genes are affected in both species, it is possible to analyze the physiological role of GLI3 and the consequences of its deficiency by comparing GLI3 expression in normal embryonic development with the effects in Xt mice in organs not readily available for analysis in Greig patients. The consequence at the molecular level of both Xt deletions as well as the GCPS mutations detected so far is the complete absence of GLI3 expression from the affected locus. This haploinsufficiency results only in a small number of minor malformations. Xt/Xt homozygotes on the other hand, despite the complete lack of expression from both genes, generally survive until birth; however, with a number of severe malformations. These stress the

1024

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Table 113-2 Induction of Digits According to Gli3a Alleles Present in Mice Alleles

Phenotype

+/+ add/+ add/add Xt/+ Xt/add Xt/Xt

Normal digit number Normal digit number One thumb is changed One extra toe Two extra toes Up to four extra toes

aGli3 expression inversely correlates with the number of digits formed.

Figure 113-2 Plantar views of the feet of +/+ and Xt/+ mice. (A) +/+ right forefoot (B–D) Xt/+ right forefeet, (E) +/+ right hind foot, (F–L) Xt/+ right hind feet. Note short, broad, or bifid polluces and postaxial nubbin of tissue on forefeet and preaxial polysyndactyly of hind feet. (Reproduced with permission from Johnson et al., J Embryol Exp Morphol 1967;17:543–581.)

general importance of the Gli3 protein as a determinator of many parts of the body plan. At the same time, it appears that most of the development can occur without any Gli3 protein. This may reflect the presence of other very similar Gli genes. Lethality may result from the lack of a functioning Gli3 gene after birth rather than from developmental defects. The spatial and temporal pattern of high Gli3 expression in mouse embryos correlates well with the Xt phenotype. The highest level of Gli3 expression is found in limb buds, the developing craniofacial mesenchyme, and the brain, the most severely affected structures in Xt/Xt homozygous mice. In addition, Gli3 is highly expressed in the intercerebral disk anlage and in the sternal cartilage, both in regions affected by Xt mutations. Thus, the proper amount of Gli3 product appears crucial in areas where it is particularly highly expressed. Possible mechanisms for such a dosage-sensitive regulatory system have been elucidated by the analysis of other Krüppel-type zinc-finger genes. In Drosophila, the Krüppel product, the prototype, can both activate and repress gene expression through interaction with a single DNA-binding site. The opposite regulatory effects of the Krüppel protein are concentration-dependent, and they require distinct portions of the protein such as the N-terminal region for activation and the C-terminal region for repression. The

Krüppel zinc-finger-type transcription factor is able to form homodimers through sequences located within the C-terminus. The Krüppel monomer is a transcriptional activator. At higher concentration, Krüppel forms a homodimer and becomes a repressor that functions through the same target sequences as the activator. Most important for the understanding of GLI function is the observation that during limb development in the Drosophila Gli protein Ci acts both as a repressor and activator of decapentaplegic (dpp) transcription in a concentration-dependent manner. The expression of Dpp, a homolog of the mammalian bone morphogenetic proteins (BMP), is normally repressed by low levels of Ci. An increase of Ci levels by hedgehog protein (Hh) signaling might overcome this regulation and result in dpp activation. This dual function of a Gli protein during limb development in Drosophila might represent an evolutionarily conserved aspect of patterning by Hh family members. Sonic Hh plays an important role in mammalian limb development by specifying in conjunction with BMP pattern across the anteroposterior axis of the limb. A function for GLI3 in the control of SHH signal production together with the control of SHH signal reception, may be consistent with the clinical observations in Greig Syndrome. GLI3 IN LIMB DEVELOPMENT A wealth of experimental results on the regulation of limb development allows to speculate on the role that Gli3 might play in this context. Limb formation in the vertebrate embryo is initiated by local proliferations of mesenchymal cells forming a bulge under the ectoderm, which is induced to form an apical ectodermal ridge (AER) at the anterior-to-posterior (first-to-fifth digit) rim of the limb bud by maintaining undifferentiated proliferating mesenchyme in the immediately subjacent area (Fig. 113-3). The AER also elaborates pattern along the proximal-distal axis (i.e., humerus to digits). A region in the posterior mesenchyme, the zone of polarizing activity (ZPA), controls patterning in the anterior-posterior axis. Genes involved in the induction and maintenance of this process in vertebrates are indicated in Fig. 113-3. Distal skeletal morphogenesis is a late event in limb development. Mesenchymal condensations occur in a sequence to become adult skeletal elements. Several steps in this developmental cascade seem to be sensitive to the proper amount of Gli3 product: In mouse embryos, Gli3 expression in the period from day 7 to 12 of gestation at first is strong in undifferentiated mesenchyme of the developing limb. Then it becomes confined to the interdigital mesenchyme between the precartilageous rays of the future digits and the interzonal mesenchyme. Later, Gli3 is highly expressed in the perichondrium, a mesenchymal layer of cells surrounding the developing cartilageous intermediate of the bones to be formed.

CHAPTER 113 / GREIG SYNDROME AND LIMB DISORDERS

1025

Figure 113-3 Molecules involved in pattern specification during limb bud development. Dorsal view of the limb bud. AER, apical ectodermal ridge: The ridge releases factors from signal to the underlying cells to maintain a high mitotic state and patterning activities. PZ, progress zone: Cells in the underlying mesenchyme progressively acquire more distal fates in the course of limb outgrowth. ZPA, zone of polarizing activity: Posterior mesenchyme cells express Sonic hedgehog (SHH) and exhibit polarizing activity, i.e., an active role in coordinating pattern formation at the growing tip of the limb by establishing a dynamic system of positional values in three dimensions. SHH induces FGF4 expression, and its own expression is maintained by FGF4 and a dorsalizing factor from the dorsal ectoderm, WNT7a. The full names of the factors are given in Table 113-3.

The level of Xt expression inversely correlates with the number of digits formed (Table 113-2). There is no indication for a polarized Gli3 expression along the anterior-posterior axis. However, it seems that the developmental field on the anterior side of the limb mesenchyme is more affected by a reduction of gene dosage than the posterior. This reduction may directly determine the additional digits. Alternatively, lack of transcriptional control by the Gli3 protein may initially result in an increased size of the limb and AER that would induce formation of additional digits. Various additional consequences of Gli3 reduction, such as bones with aberrant shapes, fused or elongated bones, and syndactylies suggest that also inducing steps further downstream, such as apoptosis separating digital elements, are influenced by the level of Gli3 product.

CANDIDATE GENES FOR HUMAN LIMB DEVELOPMENT Each of the genes organizing size and shape of individual bones and the separation of individual fingers by apoptosis and outgrowth can be altered resulting in oligo- or polydactyly, syndactylies, or ectrodactyly. As outlined for Greig syndrome, the hint for an association of a gene with a function can originate from the study of human malformations or from candidate genes detected in animal models. Only a few of the many human syndromes associated with malformations of hands and feet could be linked to candidate genes on specific human chromosomes. In rare cases, distinct mutations of candidate genes are known to cause malformations. The rich collection of candidate genes originating from the study of limb

development in animal models awaits its exploitation for the study of human development. Table 113-3 compiles the fragmentary information from these three sources.

CONCLUSIONS PERSPECTIVE OF MOLECULAR DIAGNOSTICS GLI3 spans at least 280 kb of genomic DNA. Little more than the cDNA and intron–exon boundaries have been sequenced so far. The upstream gene sequences that are potentially important for gene regulation still have to be analyzed in detail. The known intron– exon structure will stimulate the search for mutations in this gene; however, a clinical routine application is not readily available. FUTURE RESEARCH The recessive mutation in add/add mice seems to reduce the expression of the Gli3 gene and thus produce less of an otherwise normal protein. This reduction induces overgrowth only in the most sensitive developmental fields, mainly the most anterior part of the forelimbs (thumb and, eventually, digits). All other mutations analyzed so far on the molecular level in Greig patients or Xt mice are deletions or translocations, null mutations, probably eliminating transcription completely. Whereas the mouse system provides advantages for studies of embryonal stages, of homozygotes, and gene rescue studies in transgenic animals, analysis of human developmental mutations holds the promise for a larger variety of mutations in individual unrelated Greig patients. These studies will show if only complete loss of one gene copy results in the Greig phenotype. If not, the position of mutations in the GLI3 gene should indicate domains essential for functionality, the analysis of which in turn might

1026

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Table 113-3 Human Chromosome Regions Potentially Governing Limb Development Gene

Name

LMX1 INHBB PAX8 EN1 HOXD HOXD13 EVX2

LIM homeobox transcription factor 1 Inhibin, beta B Paired box homeotic gene 8 Engrailed homolog 1 Homeo box D cluster Homeo box D13 Even-skipped homeo box 2

DLX1 DLX2 INHA IHH GLI2

Distal-less homeo box 1 Distal-less homeo box 2 Inhibin, alpha Indian hedgehog (Drosophila) homolog GLI-Krüppel family member GLI 2

MSX1 FGFR3

msh (Drosophila) Homeo box homolog 1 Fibroblast growth factor receptor 3

FGF2 MSX2

Fibroblast growth factor 2 msh (Drosophila) Homeo box homolog 2

HOXA EVX1 INHBA GLI3

Homeo box A cluster Even-skipped homeo box 1 Iinhibin, beta A GLI-Krüppel family member GLI 3

DSS1 DLX5 DLX6 WNT2 PAX4

Deleted in SHFM1 Distal-less homeo box 5 Distal-less homeo box 6 Wingless-type MMTV integration site 2, homolog Paired box homeotic gene 4

SHH EN2 BMP1 FGFR1

Sonic hedgehog (Drosophila) homolog Engrailed homolog 2 Bone morphogenetic protein 1 Fibroblast growth factor receptor 1

FGF8 FGFR2

Fibroblast growth factor 8 Fibroblast growth factor receptor 2

FGF4 WNT1 GLI INHBC BMP4 HOXB8 WNT3

Fibroblast growth factor 4 Wingless-type MMTV integration site 1, homolog Glioma-associated oncogene homolog Inhibin, beta C Bone morphogenetic protein 4 Homeo box B8 Wingless-type MMTV Integration site 3, homolog

SOX9 BMP2

SRY (sex-determining region Y)-box 9 Bone morphogenetic protein 2

WNT7a

Wingless-type MMTV Integration site 7a, homolog

help to identify other genes on which the GLI3 product acts to regulate development and other factors cooperating with GLI3 in this function.

Chromosomal location 1q22–23 2cen–q13 2q12–q14 2q13–q21 2q31 2q31 2q24.3–q31 2q24–q31 2q32 2q32 2q33–q34 2 2 4p16 4p16.3–p16.1 4p16.3 4q25–q27 5q34–q35 6q21 7p15–21 7p15–p14 7p15–p14 7p15–p13 7p13 7q11–q22 7q22.1 7q22 7q22 7q31 7q32 7q36 7q36 7q36 7q36 8p21 8p12 10q24–q25 10q25–q26 10q25.3–q26

Syndrome affecting limb development

Synpolydactyly syndrome Split hand split foot

Adelaide type craniosynostosis Achondroplasia, Thanatophoric dysplasia, Hypochondroplasia (Boston-type craniosynostosis) Split hand split foot Saethre-Chotzen syndrome

Greig cephalopolysyndactyly syndrome EEC syndrome Split hand split foot (SHFM1)

Complex bilateral polysyndactyly Triphalangeal thumb

Pfeiffer syndrome Split hand split foot (SHFM3) Apert syndrome, Pfeiffer syndrome, Jackson-Weiss syndrome (Crouzon syndrome)

11q13.3 12q13 12q13 12q13.1 14 17q21–q22 17q21–q22 17q23.1–q24.2 17q24.3–q25.1 20p12 Xq26–26.1 —

Hunter-McAlpine craniosynostosis Campomelic dysplasia X-linked split hand split foot (SHFM2)

The genomic sites upstream of the Gli3 transcription start that are disturbed by the add insertion are candidates for promoter regions governing transcription of Gli3 itself.

CHAPTER 113 / GREIG SYNDROME AND LIMB DISORDERS

The variety of effects on different organ systems manifested by homozygous Xt mice encourages one to search systematically for mutations influencing GLI3 expression or functionality in the multitude of human developmental defects other than the Greig syndrome.

SELECTED REFERENCES Ausems MGEM, Ippel PF, Renardel de Lavalette PAWA. Greig cephalopolysyndactyly syndrome in a large family: a comparison of the clinical signs with those described in the literature. Clin Dysmorphol 1994;3:21–30. Buscher D, Bosse B, Heymer J, Ruther U. Evidence for genetic control of sonic hedgehog by Gli3 in mouse limb development. Mech Dev 1997;62:175–182. Cohn MJ, Tickle C. Limbs: a model for pattern formation within the vertebrate body plan. TIG 1996;12:253–257. Domínguez M, Brunner M, Hafen E, Basler K. Sending and receiving the hedgehog signal: control by the Drosophila Gli protein cubitus interruptus. Science 1996;272:1621–1625. Greig DM. Oxycephaly. Edinburgh Med J 1926;33:189–218. Hui C-C, Joyner AL. A mouse model of Greig cephalopolysyndactyly syndrome: the extra-toesJ mutation contains an intragenic deletion of the Gli3 gene. Nat Genet 1993;3:241–246. Johnson DR. EXtra toes: a new mutant gene causing multiple abnormalities in the mouse. J Embryol Exp Morph 1967;17:543–581.

1027

Marigo V, Johnson RL, Vortkamp A, Tabin CJ. Sonic hedgehog differentially regulates expression of GLI and GLI3 during limb development. Dev Bio 1996;180:273–283. Masuya H, Sagai T, Moriwaki K, Shiroishi T. Multigenic control of the localization of the zone of polarizing activity in limb morphogenesis on the mouse. Dev Biol 1997;182:42–51. Mo R, Freer AM, Zinyk DL, et al. Specific and redundant functions of Gli2 and Gli3 zinc finger genes in skeletal patterning and development. Development 1997;124:113–123. Pohl TM, Mattei M-G, Rüther U. Evidence for allelism of the recessive insertional mutation add and the dominant mouse mutation extra-toes (Xt). Development 1990;110:1153–1157. Roberts DJ, Tabin C. The genetics of human limb development. Am J Hum Genet 1994;55:1–6. Ruppert JM, Vogelstein B, Arheden K, Kinzler KW. GLI3 encodes a 190kilodalton protein with multiple regions of GLI similarity. Mol Cell Biol 1990;10:5408–5415. Schimmang T, Lemaistre M, Vortkamp A, Rüther U. Expression of the zinc finger gene Gli3 is affected in the morphogenetic mouse mutant extra-toes (Xt). Development 1992;116:799–804. Tabin C. The initiation of the limb bud: growth factors, Hox genes, and retinoids. Cell 1995;80:671–674. Vortkamp A, Gessler M, Grzeschik K-H. GLI3 zinc-finger gene interrupted by translocations in Greig syndrome families. Nature 1991;352:539,540. Vortkamp A, Heid C, Gessler M, Grzeschik K-H. Isolation and characterization of a cosmid contig for the GCPS gene region. Hum Genet 1995;95:82–88.

CHAPTER 114 / FGFR-RELATED SKELETAL DISORDERS

114

1029

Fibroblast Growth Factor Receptor-Related Skeletal Disorders Craniosynostosis and Dwarfism Syndromes MAXIMILIAN MUENKE, CLAIR A. FRANCOMANO, M. MICHAEL COHEN, JR., AND ETHYLIN WANG JABS

BACKGROUND Unique mutations in three human fibroblast growth factor receptors (FGFR1, FGFR2, and FGFR3) have been identified as causing various skeletal disorders that affect the skull (craniosynostosis syndromes) and the growth of the long bones (dwarfism syndromes). Craniosynostosis is the premature fusion of one or more cranial sutures. It is relatively common, with an estimated birth prevalence of 340 per million. More than 90 syndromes are associated with craniosynostosis, and the majority are inherited in an autosomal dominant manner. The best known of these include Crouzon, Jackson-Weiss, Pfeiffer, Apert, and Saethre-Chotzen syndromes. These syndromes have in common premature synostosis of the coronal sutures of the skull, underdevelopment of the midface, some cases having variable abnormalities of the extremities and other organ systems. All are inherited as autosomal dominant traits, most with complete penetrance and variable expressivity. Although viewed clinically as distinct entities, Crouzon, JacksonWeiss, Pfeiffer, and Apert syndromes have now been shown to be allelic, with alterations in the same gene (FGFR2). Other craniosynostosis syndromes also shown to be caused by FGFR mutations include Beare-Stevenson cutis gyrata syndrome, Crouzon syndrome with acanthosis nigricans, and a craniosynostosis syndrome associated with a unique FGFR3 point mutation. Interestingly, the two former conditions have dermatologic findings, but the first disorder is associated with FGFR2 mutations, whereas the second is secondary to a FGFR3 mutation. In contrast, there are two craniosynostotic conditions with mutations in transcription factors. Craniosynostosis, Boston type, is a result of a mutation in a homeobox gene, MSX2, and Saethre-Chotzen syndrome has been linked to mutations in the TWIST gene, a basic helix-loophelix transcription factor. It has been suggested that the FGFRs are part of the same signaling pathway as TWIST. Three dwarfing conditions—achondroplasia, hypochondroplasia, and thanatophoric dysplasia—have long been considered a family of skeletal dysplasias because of clinical and radiographic similarities. The phenotypic spectrum ranges from neonatal lethal thanatoFrom: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

phoric dysplasia to mild hypochondroplasia, with achondroplasia somewhere in between. Overlap with the craniosynostosis syndromes is seen in thanatophoric dysplasia, in which cloverleaf skull occurs much more frequently in Type 2 than in Type 1. Recent observations have led to the delineation of a newly recognized syndrome in this family of disorders, which is more severe than achondroplasia but less so than thanatophoric dysplasia; the condition is called severe achondroplasia with developmental delay and acanthosis nigricans (SADDAN). It is now recognized that all four of these disorders are caused by mutations in FGFR3. Comparison of specific mutations in FGFRs and their phenotypes is providing new insights into the role of these receptors in normal and abnormal bone development. Furthermore, these mutations provide molecular signposts to complement clinical descriptions of this family of disorders. First, we will introduce the clinical features of these craniosynostosis and dwarfing syndromes and then discuss their genetic basis.

CLINICAL FEATURES CROUZON SYNDROME Crouzon syndrome (OMIM 123500) is among the earliest described and most commonly inherited causes of craniosynostosis. In addition to premature fusion of the skull bones, this syndrome is characterized by shallow orbits, ocular proptosis, and midface hypoplasia (Fig. 1141A). Craniosynostosis commonly involves the coronal suture leading to brachycephaly with shortening of the anteroposterior diameter of the skull. Additional findings include frontal bossing, hypertelorism, strabismus, beaked nose, mandibular prognathism, high arched palate, dental malocclusion, and conductive hearing loss. On occasion, mental deficiency has been observed. The most consistent finding is severe-to-moderate ocular proptosis because of shallow orbits. The hands and feet are not involved, in contrast to other craniosynostosis syndromes. Crouzon syndrome has a birth prevalence of 15.5 per one million newborns and accounts for approximately 4.5% of all cases of craniosynostosis. This syndrome has an autosomal dominant mode of inheritance with high penetrance and moderate variability in phenotypic expression. On occasion, families with extreme variability have been reported. New mutations represent 56% of all cases, and there is evidence of an advanced paternal age effect. Instances of two affected sibs

1029

1030

SECTION XIII / GENETIC BASIS FOR CONGENITAL MALFORMATIONS

Figure 114-1 (A) Crouzon syndrome: Mother with ocular proptosis, divergent strabismus, hypertelorism, and midfacial hypoplasia. Note marked digital impressions on radiograms of the skull. (From Jones, Smith's Recognizable Patterns of Human Malformation, 1997, p. 421.) (B) Jackson-Weiss syndrome: two members of the Amish pedigree, one who is mildly affected with brachycephaly and hypertelorism and the other with turribrachycephaly. Feet anomalies of large great toes, syndactyly, and medial deviation. (From Fig. 1 of Jabs et al. Nature Genet 8:275– 279,1994; Fig. 1C of Li et al. Genomics 22:418–424, 1994.) (C) Apert syndrome: Infant with severe brachycephaly, frontal bossing, ocular proptosis, hypertelorism, and midfacial hypoplasia. Note significant syndactyly. (From Fig. 2 of Park et al. Am J Hum Genet 57:321–328, 1995.) (D) Crouzon syndrome and acanthosis nigricans: female with brachycephaly, ocular proptosis, hypertelorism. Dermatologic manifestations of hyperpigmentation, hyperkeratosis, and melanocytic nevi. (E) Beare-Stevenson cutis gyrata syndrome: Note fine and coarse wrinkling of the skin with facial features of ocular proptosis and midface hypoplasia. (From Fig. 1 of Przylepa et al. Nature Genet 13:492–494, 1996; Fig. 12 of Hall et al. Am J Med Genet 44:82–89, 1992.) (F) SADDAN syndrome: Female with facial features similar to achondroplasia and hyperpigmentation.

born to unrelated parents with normal phenotypes raises the possibility of germline mosaicism. CROUZON SYNDROME WITH ACANTHOSIS NIGRICANS A subgroup of patients with Crouzon syndrome have a dermatologic disorder—acanthosis nigricans and melanocytic nevi (Fig. 114-1D; OMIM 134934.0011). These skin findings with craniosynostosis breed true as an autosomal dominant condition. Acanthosis nigricans is characterized by verrucous hyperplasia of the skin, particularly in flexural areas. Histopathological findings include marked papillomatosis, the epidermis being thin and

hyperpigmented. Acanthosis nigricans is heterogeneous, and many cases are associated with insulin resistance. This is of interest, since insulin binds to the classic insulin receptor and insulin-like growth factor receptors, which are expressed in keratinocytes. Other endocrine abnormalities (including obesity) and certain malignancies are known causes of acanthosis nigricans, and an association with certain congenital disorders, including Crouzon syndrome, is well-recognized. Acanthosis nigricans associated with Crouzon syndrome is rare and atypical in several ways. It usually occurs in females and its onset is often early, apparent in childhood

CHAPTER 114 / FGFR-RELATED SKELETAL DISORDERS

1031

Figure 114-2 (A) Pfeiffer syndrome: Craniosynostosis of the coronal sutures, mild protrusion of her eyes, mildly broad thumbs, and great toes in a sporadic patient with the common FGFR1 mutation. Craniosynostosis with more severe ocular protrusion, very broad thumbs and great toes and a splice mutation in exon 9 (IgIIIc) of FGFR2 in a patient with familial Pfeiffer syndrome. (From Fig. 5 of Schell et al. Hum Mol Genet 4:323–328, 1995.) (B) Cloverleaf skull resulting from synostosis involving multiple sutures in a newborn with sporadic Pfeiffer syndrome and FGFR2 mutation. Note the broad thumbs and broad medially deviated great toes on X-rays. (C) FGFR3-associated coronal synostosis: plagiocephaly resulting from unicoronal synostosis. Note the coned epiphyses and fused metacarpal bones on radiograms of the hands in the same patient with FGFR3 749C→G (Pro252Arg) mutation. (From Gripp et al., J Pediatr, in press.)

and always by puberty. It is distributed in a distinctive pattern, in the perioral, periorbital, and nasolabial areas as well as the neck, axillae, chest, breasts, and abdomen. Cementomas of the jaws have been found in a number of cases. Additional findings may include choanal atresia and hydrocephalus. BEARE-STEVENSON CUTIS GYRATA SYNDROME Beare-Stevenson cutis gyrata syndrome (OMIM 123790) is a rare condition with only nine isolated cases reported to date. It is presumed to be autosomal dominant because of associated advanced paternal age. It is characterized by furrowed skin (cutis gyrata), acanthosis nigricans, craniosynostosis with craniofacial dysmorphism, digital anomalies, umbilical and anogenital anomalies, and early death (Fig. 114-1E). The most striking feature of this syndrome is the cutis gyrata with a corrugated appearance resembling the fine ribbing of corduroy. Large and deep folds are seen on the forehead, palms, soles, and perianal/genital regions. The craniofacial features of Beare-Stevenson syndrome are similar to the other craniosynostosis syndromes with downslanting of the palpebral fissures, ocular hypertelorism and proptosis, strabismus, dysmorphic ears, choanal atresia, and palatal anomalies. The majority have cloverleaf skull deformities with hydrocephalus, but some patients have a Crouzon-like appearance. In addition, a prominent umbilical stump, genital abnormalities (bifid scrotum, prominent scrotal raphe, rugose labia majora), and anal anomalies (anteriorly placed anus, tissue mounding of the anus) have been noted. Other unusual features include natal teeth, bifid or accessory nipples, pyloric stenosis, and a coccygeal tail. JACKSON-WEISS SYNDROME Jackson-Weiss syndrome (OMIM 123150) is characterized by craniosynostosis and foot

anomalies (Fig. 114-1B). This condition was first reported in an Amish kindred with 88 affected individuals examined and another 50 reported to be affected. The condition is inherited in an autosomal dominant pattern with high penetrance and phenotypic expression so variable that the entire spectrum of dominantly inherited craniofacial dysostoses-acrocephalosyndactylies (e.g., Saethre-Chotzen and Pfeiffer syndromes, with the exception of Apert syndrome) appeared to be represented within this kindred. The facial features include proptosis, maxillary hypoplasia, and frontal prominence. Unlike Crouzon syndrome, the degree of craniosynostosis and shallow orbits is mild to moderate, and none of the over 130 affected individuals in the Amish family are known to have undergone surgical intervention. Another distinguishing feature is that patients with Jackson-Weiss syndrome have the consistent manifestation of abnormal clinical and/or radiographic appearance of the feet, which can occur in the absence of any craniofacial features. Affected individuals have mild cutaneous syndactyly (especially of the second and third toes), broad great toes (broad metatarsal and proximal phalanges) with medial deviation, metatarsal and tarsal fusions (calcaneus, cuboid, navicular, and cuneiform), and/or abnormally shaped metatarsal and tarsal bones. Hand abnormalities are rare. The birth prevalence of this disorder is unknown, but several familial cases have been reported in the literature. PFEIFFER SYNDROME Pfeiffer syndrome (OMIM 101600) is characterized by craniosynostosis, broad thumbs, and broad great toes (Fig. 114-2A,B). Craniofacial features are variable and secondary to synostosis, usually involving the coronal sutures.

1032

SECTION XIII / GENETIC BASIS FOR CONGENITAL MALFORMATIONS

Findings include a turribrachycephalic skull, midfacial hypoplasia with relative mandibular prognathism, beaked nose, low nasal bridge, hypertelorism, downslanting palpebral fissures, ocular proptosis, strabismus, highly arched palate, and crowded teeth. Anomalies of the hands and feet differ from any other craniosynostosis syndrome. Characteristically, the thumbs and great toes are broad and medially deviated (varus deformity). Brachydactyly is often present, as is partial soft tissue syndactyly of the fingers and toes. Radiographic findings of the hands include malformed and fused phalanges of the thumbs, short middle phalanges (brachymesophalangy) to complete absence, symphalangism (complete osseous fusion of proximal, middle, and distal phalanges that occurs over several years), and occasional fusion of the proximal ends of the fourth and fifth metacarpals. The distal phalanx of the great toe is broad, and the proximal phalanx is malformed. Broad and short first metatarsals and fusion of carpal and tarsal bones have been described. Additional manifestations may include cloverleaf skull (Fig. 114-2B), fusions of cervical and lumbar vertebrae, cubitus valgus, synostosis of the radio-humeral and ulnar-humeral joints, shortened humerus, abnormalities of the pelvis, coxa valga, and talipes calcaneovarus. Abnormalities affecting other organ systems are of low frequency and include hearing deficit, optic nerve hypoplasia, choanal atresia, bifid uvula, supernumerary teeth, gingival hypertrophy, widely spaced nipples, anal atresia or malpositioned anus, pyloric stenosis, umbilical hernia, and bifid scrotum. Intelligence is usually normal, although mental deficiency has been observed, particularly in sporadic cases. Rare CNS anomalies include hydrocephaly, Arnold-Chiari malformation, and seizures. The birth prevalence of Pfeiffer syndrome is not known. Its inheritance pattern is autosomal dominant with complete penetrance and variable expressivity. In addition to familial occurrence, sporadic cases resulting in new mutations are well-known. FGFR3-ASSOCIATED CORONAL SYNOSTOSIS SYNDROME This craniosynostosis syndrome has been defined in patients with a unique point mutation in FGFR3 (OMIM 134934.0014). Clinical findings in over 70 patients from more than 30 unrelated families include bi- or unicoronal synostosis, midface hypoplasia, downslanting palpebral fissures, and ptosis (Fig. 114-2C). It is of note that some mutation carriers do not show any signs of craniosynostosis, having only macrocephaly or even normal head size. Sensorineural hearing loss is seen in some, as is developmental delay. Hand and foot anomalies are found in some, but not all, affected individuals and include brachydactyly with characteristic radiographic findings, such as thimble-like middle phalanges, coned epiphyses, and tarsal fusions. Some patients have broad halluces that are not deviated. It is of interest that this specific FGFR3 mutation was identified in several affected individuals with unicoronal synostosis from a large prospective series, some of which were previously thought to be caused by intrauterine constraint. Height is normal, in contrast to FGFR3-associated dwarfism syndromes. The mode of inheritance is autosomal-dominant. Several previously reported families could be demonstrated to have this specific FGFR3 mutation, one of which was called “craniosynostosis-type Adelaide.” Based on results from an ongoing prospective study, it appears that the mutation associated with this syndrome may be common in sporadic or familial patients with uni- or bicoronal synostosis whose findings do not fit into any of the classic craniosynostosis syndromes. APERT SYNDROME Apert syndrome (OMIM 101200) is characterized by craniosynostosis, midface hypoplasia, and sym-

metric syndactyly of the hands and feet, with a variety of other anomalies (Fig. 114-1C). Craniofacial features are characteristic for Apert syndrome and differ from those of other craniosynostosis syndromes. The cranium is disproportionately high, and the brain is megalencephalic. During infancy, a wide midline calvarial defect is present, extending from the root of the nose to the posterior fontanelle. Bony islands form and coalesce until the gap is completely covered by bone during the third year of life. Although the coronal suture is fused at birth, the anterolateral fontanelles are abnormally large, and both the lambdoid and squamosal sutures are patent. Turribrachycephaly is commonly observed with a flattened occiput and a steep forehead. The cranial base is malformed and often asymmetric. The anterior cranial fossa is very short. Additional craniofacial anomalies include hypertelorism, proptosis secondary to shallow orbits, downslanting palpebral fissures, and frequently strabismus. Hyperopia, myopia, or astigmatism can be found frequently. The nasal bridge is low, and the nose is beaked. Midfacial hypoplasia gives the impression of relative mandibular prognathism. Hearing deficit is secondary to otitis media and is related to the high frequency of cleft palate or bifid uvula. Furthermore, congenital fixation of the stapedial footplate has been reported in a number of cases. Dental anomalies include delayed and/or ectopic eruption, shovel-shaped incisors, and malocclusion. Upper-airway compromise is secondary to reduced nasopharyngeal and oropharyngeal space and poses a risk for obstructive sleep apnea, cor pulmonale, and even sudden death. Characteristic abnormalities of the hands and feet are symmetric. The mid-digital hand mass minimally involves the second, third, and fourth fingers. The first and fifth fingers may be joined to the mid-digital hand mass or may be free. Similar patterns of syndactyly occur in the feet. Additional limb anomalies include limited mobility at the glenohumeral joint as a constant feature that progressively worsens with growth. Slight limitation of elbow extension and rotation occurs commonly, but is nonprogressive in nature. The humerus is commonly shortened to some degree. Radiographically, lack of segmentation of some joints in the hands cannot be observed at birth, but progressive ossification is found with time. Vertebral anomalies particularly cervical fusions are common. Variable degrees of fusion may be observed involving the articular facets, the neural arch or transverse processes, or block fusion of the vertebral bodies. Ossification may not always be evident in early radiographs, although signs of impending fusion may include irregularity in vertical orientation of the vertebral bodies and narrowing of the involved intervertebral spaces. In childhood, the growth pattern in Apert syndrome consists of a slowing of linear growth so that most values fall between the 5th and 50th percentiles. From the onset of adolescence to adulthood, slowing becomes more pronounced. This two-step deceleration results in large measure from rhizomelic shortening of the lower limbs. The adolescent growth spurt takes place within the normal time frame. Intelligence in patients with the Apert syndrome varies from normality to mental deficiency. In patients who had craniectomies performed the first year of life, no significant differences in IQ were found between retarded and nonretarded outcome. It has been speculated that the mental deficiency is secondary to CNS abnormalities, such as malformations of the corpus callosum, the limbic structures, gyral abnormalities, hypoplasia of the cerebral white matter, and heterotopic gray matter. Gyral abnormalities, megalencephaly, and distortion ventriculomegaly are very com-

CHAPTER 114 / FGFR-RELATED SKELETAL DISORDERS

1033

Figure 114-3 (A) Achondroplasia: Note macrocephaly, rhizomelic shortening, leg bowing (From Cohen MM, Jr. The Child with Multiple Birth Defects, 2nd ed., New York: Oxford University Press, 1997; p. 213). (B) Hypochondroplasia is associated with normal face and relatively proportional, short stature (From Cohen MM, Jr. The Child with Multiple Birth Defects, 2nd ed., New York: Oxford University Press, 1997; p. 213). (C) Thanatophoric dysplasia type 1 stillborn twins (Courtesy of MM Cohen, Jr.).

mon. Progressive hydrocephalus may occur on occasion but not nearly as frequently as in Pfeiffer or Crouzon syndromes or with cloverleaf skull deformities. Cutaneous manifestations include skin dimples at the shoulders, elbows, and knuckles. Excess sweating is common. Acneiform lesions are particularly prevalent on the face, chest, and back. Additional skin findings include hyperkeratoses of the midplantar and lateral plantar surface, hypopigmentation of the skin, and paronychial infections with the feet being more commonly involved than the hands. Apert syndrome is mostly sporadic, although over one dozen familial instances with autosomal-dominant inheritance pattern are known. Gonadal mosaicism has been postulated in a family with two affected sibs and unaffected parents. The rarity of familial cases can be explained by reduced genetic fitness of affected individuals because of the severe malformations and the presence of mental deficiency in many cases. Increased paternal age at the time of conception has been found in a large series of sporadic cases, and exclusive paternal origin of new mutations has been confirmed on a molecular level. Birth prevalence of Apert syndrome is 15.5 per one million births. The mutation rate is calculated to be 7.8 × 10–6 per gene per generation. Apert syndrome accounts for about 4.5% of all cases of craniosynostosis. ACHONDROPLASIA Achondroplasia (OMIM 100800) is the most common cause of human dwarfism, with an estimated birth prevalence of 25–66.7 per one million live births. It is inherited as an autosomal-dominant condition, but more than 80% of cases represent new mutations. The phenotype predominantly involves the skeletal system, although manifestations are seen in other organ systems as a result of bony compression. Affected persons are short, with rhizomelic (proximal) shortening of the arms and legs and a disproportionately long trunk (Fig. 114-3A). Often the hands exhibit a “trident” configuration, with separation between the fourth and third fingers resulting in a three-pronged appearance. True megalencephaly is seen, with absolute enlargement of the head circumference in more than 75% of cases. Typical facial characteristics include frontal bossing and midface hypoplasia.

Affected infants often have a thoraco-lumbar gibbus that is usually reducible and resolves as the child begins to sit and walk independently. Frequently, the gibbus is replaced by an exaggerated lumbar lordosis in early childhood, although the gibbus may persist into adulthood and may require surgical correction. Hypotonia is frequently observed in infancy and usually becomes less pronounced with age. Although major motor milestones are usually delayed, intelligence is generally normal unless hydrocephalus or other major complications involving the central nervous system intervene. Final adult height typically ranges from 118 to 145 cm for males and 112 to 136 cm for females. The major medical complications of achondroplasia result from bony compression of the neuroaxis and the respiratory system. A small foramen magnum is found in almost all cases, and may lead to compression of the cervico-medullary junction and result in a high cervical myelopathy and/or central respiratory complications in early childhood, if the medulla is compromised. Small airway passages may lead to airway obstruction with potentially severe obstructive apnea. Recurrent otitis media is a frequent complication in early childhood and may require placement of pressureequalizing tubes to minimize the risk of hearing loss. Radiographic features of achondroplasia include short limbs with short, thickened tubular bones and short phalanges and metacarpals. The ribs are short with concave ends, and the growth plates are notched with a V-shaped appearance. Proximal fibular overgrowth is a frequent observation. The sternum is thick and wide. The vertebral bodies are typically short, flat and “bulletshaped” in infancy. There is a decrease in the interpediculate distance as one progresses to the caudal end of the spine, in contrast to the increase in this dimension seen in persons who do not have achondroplasia. The vertebrae have short pedicles, which further compromise the dimensions of the spinal canal and contribute to symptoms of spinal stenosis in adulthood. Neuroradiologic evaluation of the brain and skull typically demonstrates enlarged ventricles and marked reduction in the size of the foramen magnum. Morphologic evaluation of the growth plate in achondroplasia demonstrates normal, well-organized endochondral ossification

1034

SECTION XIII / GENETIC BASIS FOR CONGENITAL MALFORMATIONS

in many growth plates, but disruption of the growth plate in some weight-bearing joints, with short columns of chondrocytes and widening of the septa between the columns. HYPOCHONDROPLASIA Hypochondroplasia (OMIM 146000) has many clinical similarities to achondroplasia, but is generally milder overall and typically has less craniofacial involvement. Short stature may not be recognized until the child is 2–3 years of age and is less pronounced than usually seen in achondroplasia (Fig. 114-3B). The height range may overlap with that of the general population, and the condition may be underdiagnosed. Hands and feet are broad and wide, but the trident appearance seen in achondroplasia is not usually apparent. At least one paper has reported an increased frequency of learning disabilities among this population. The radiographic features of hypochondroplasia are also similar to, but milder than, those seen in achondroplasia. The long bones appear short and wide, with mild flaring of the metaphysealepiphyseal junctions. The skull is relatively normal in appearance, with only mild midface hypoplasia and frontal bossing. The interpediculate distance in the lumbar spine may stay constant or decrease as one progresses from rostral to caudal. The bone and cartilage in hypochondroplasia look relatively normal on morphologic evaluation, with minimal shortening of the chondrocytic columns and thickening of the matrix between the columns. THANATOPHORIC DYSPLASIA Thanatophoric dysplasia (OMIM 187600) takes its name from the Greek “thanos,” meaning death. Children affected with this condition almost invariably die in the perinatal period, with only a few survivors into early childhood documented in the literature. Early death is caused by severe respiratory compromise, secondary to an extremely small chest, perhaps exacerbated by medullary compression secondary to a small foramen magnum. Thanatophoric dysplasia is the most common of the neonatal lethal skeletal dysplasias. Clinical features include severe short-limbed dwarfism, curvature of the long bones, megalencephaly, and disproportionately long trunk with short ribs (Fig. 114-3C). The condition has been separated into two types, based on the presence or absence of curved tibiae and fibulae. The cloverleaf skull deformity is more often seen with straight tibiae and fibulae (Type 2). Type 1 is generally characterized by curved long bones and the absence of cloverleaf skull. Histopathology of the growth plate is typically irregular with few proliferative and hypertrophic chondrocytes. The hypertrophic zone of the growth plate is replaced by fibrous-appearing tissue, particularly at the periphery of the growth plate. SADDAN Three patients have been recognized with a newly defined syndrome named SADDAN, for severe achondroplasia with developmental delay and acanthosis nigricans (Fig. 114-1F). All three patients were initially diagnosed as having achondroplasia, but were recognized as severe cases, with more profound dwarfism and bowing of the limbs than is typical for achondroplasia. All three cases exhibited acanthosis nigricans (an overlap with Crouzon syndrome with acanthosis nigricans and Beare-Stevenson syndrome) and had severe developmental delay, a highly unusual observation in achondroplasia. Radiographic features are intermediate between those of achondroplasia and thanatophoric dysplasia, and to date no chondro-osseous histopathology has been studied. One phenotypic case of thanatophoric dysplasia Type 1 with long-term survival and the R248C mutation associated with this diagnosis, and not SADDAN, has also been associated with acanthosis nigricans.

DIAGNOSIS Patients with craniosynostosis syndromes are brought to medical and surgical attention because of abnormal head shape. Standard radiologic criteria are used to make the diagnosis of craniosynostosis. The three-dimensional CAT scan is the preferred tool for documentation of prematurely fused sutures and facial dysmorphism. Craniosynostosis cannot be diagnosed radiologically in adults because the sutures are normally fused. Associated increased intracranial pressure can be detected by radiological (see radiogram in Fig. 114-1A), ophthalmologic, and neurological examination, and CAT or MRI scan. A dysmorphology examination is required to detect the presence of associated features such as midface hypoplasia, and digital and anal anomalies. Clinical examination and dermatologic consultation with histologic examination of the skin can document such findings as acanthosis nigricans. Molecular diagnosis may involve analysis of the FGFR1, FGFR2, FGFR3, MSX2, or TWIST genes. Diagnosis of dwarfism syndromes is still made predominantly on clinical and radiographic grounds. There are cases of intermediate severity between achondroplasia and hypochondroplasia, for whom molecular analysis may help to clarify the diagnosis. However, in the vast majority of cases, the diagnoses of achondroplasia, hypochondroplasia, and thanatophoric dysplasia can be made based on clinical observations and radiographic analysis. Unusual cases, such as those with acanthosis nigricans, unusually severe dwarfism, or mental deficiency, should be studied for the FGFR3 mutation associated with SADDAN or other possible FGFR3 mutations. The detection of the molecular basis of these disorders allows for prenatal as well as postnatal diagnosis. However, there are several problems in the application of molecular diagnostic testing, including the difficulty in predicting the severity of the phenotypic expression in affected individuals.

GENETIC BASIS OF DISEASE The genetic bases of the craniosynostosis and short-stature syndromes described above have been elucidated very recently. The identification of the genes involved in these disorders has been a prime example of the so-called positional candidate gene approach. As an initial step toward gene identification, linkage analysis in large families aided in the mapping of the respective syndrome loci. Using microsatellite markers from throughout the genome, achondroplasia was mapped to DNA markers located in chromosome band 4p16. Both Crouzon and Jackson-Weiss syndrome were linked to chromosome 10q25–q26. Pfeiffer syndrome was shown to be heterogeneous with one locus on 8cen, a second locus on 10q25–q26. Because of the lack of large families with Apert syndrome, this disease could be excluded from 8cen markers and was shown to be consistent with linkage to 10q25–26. Similarly, a new autosomal-dominant craniosynostosis syndrome was excluded from the 8cen and 10q25–q26 regions, respectively, and was consistent with linkage to 4p16. Based on the chromosomal map position and the expression pattern, the FGFRs located in the same chromosomal regions were examined as candidate genes. Mutations were identified in FGFR2 on chromosome 10q25–q26 as the cause for Crouzon, Jackson-Weiss, Pfeiffer, Apert, and Beare-Stevenson cutis gyrata syndromes. Mutations in FGFR1 gene on 8p11.1 were identified in a subset of families with Pfeiffer syndromes. FGFR3 on 4p16 is known to be involved in achondroplasia, hypochondroplasia, thanatophoric, and SADDAN, and was shown to cause Crouzon syndrome with acanthosis

CHAPTER 114 / FGFR-RELATED SKELETAL DISORDERS

nigricans and, most recently, FGFR3-associated coronal synostosis syndrome. Thus far, no human disorder has been associated with FGFR4, which is located in 5q35. FIBROBLAST GROWTH FACTOR RECEPTORS Four human FGFRs have been described, all coding for a cell-surface protein with an extracellular region with three immunoglobulinlike domains (IgI, IgII, IgIII), a transmembrane segment, and a split cytoplasmic tyrosine kinase domain (Fig. 114-4). The four FGFRs exhibit an amino acid identity of 55–72%, but differ in their ligand affinity and their tissue and temporal distribution. Structural diversity of the FGFRs is generated by alternative splicing (Fig. 114-4A), which leads to transmembrane receptors or secreted receptors consisting of one, two, or three Ig domains and different carboxyterminal halves of IgIII. IgII, the interloop region, and the N-terminus of IgIII are implicated in ligand binding. In addition, C-terminal sequences of IgIII appear to be important for ligand specificity, although they do not directly interact with their ligands, the fibroblast growth factors (FGFs). FGFs regulate cell proliferation, differentiation, and migration in many different tissues through complex signaling pathways. To date, 15 structurally related proteins of the mammalian FGF family have been identified and are associated with a wide spectrum of functions such as angiogenesis, wound healing, embryonic development, and malignant transformation. The ligand-induced activation of the FGFR kinase activity is mediated by receptor dimerization, which results in transphosphorylation of one receptor molecule by the other in the dimer. The autoactivated FGFR can then bind and phosphorylate intracellular signal transduction proteins and thus activate downstream pathways. Xenopus, chicken, and mouse FGFRs encode proteins similar to human FGFR1 with a high amino acid identity of 78–98%. For the Drosophila (DFR1) and the Xenopus (FGFR) homologs, mRNA expression was shown in the mesoderm during embryogenesis. In Xenopus, posterior and ventral mesoderm development was blocked by a dominant-negative FGFR. Distinct spatial and temporal mRNA expression of FGFs and FGFRs, observed in the mouse, further supports roles for this group of ligands and receptors in embryogenesis. Whereas FGFR1 and FGFR2 mRNAs are coexpressed in prebone/precartilage structures, such as during craniofacial bone formation, FGFR3 is expressed in the cartilage growth plates of long bones during endochondral ossification. MUTATIONS IN FGFRs CAUSE CRANIOSYNOSTOSIS AND DWARFISM SYNDROMES Both linkage and mutation analysis demonstrated genetic heterogeneity in Pfeiffer syndrome. Mutations have been identified in sporadic and familial Pfeiffer syndrome in FGFR1 and FGFR2. A common mutation, a C-to-G transversion in exon 5 of FGFR1 has been detected in all affected individuals of more than 20 unrelated families with Pfeiffer syndrome. This missense mutation, 755C→G, predicts a P252R substitution in the IgII-IgIII linker region, which is highly conserved in FGFRs through evolution as well as among the four human FGFRs. In this respect, it is of great interest that the 755C→G transversion predicting a P252R substitution in the IgII-IgIII linker region of FGFR1 causing Pfeiffer syndrome is at an identical position to the 755C→G change in FGFR2 in Apert syndrome and the 749C→G mutation in FGFR3, causing FGFR3-associated coronal synostosis syndrome (Fig. 114-4B,C; see below). The second gene known to be involved in Pfeiffer syndrome is FGFR2. Approximately 18 different mutations have been shown in familial and sporadic cases, the majority of which either (1)

1035

destroy a specific paired cysteine residue (C342) crucial for the formation of IgIII within the ligand binding region of FGFR2 or (2) introduce an additional unpaired cysteine, which likely interferes with the normal protein structure. Aberrant splicing of exon 9 (exon IIIc or exon B) in FGFR2 was proposed to account for a subgroup of cases. The phenotypic consequences of mutations affecting exon 9 of FGFR2 are remarkable for their pleiotropic effects resulting in three distinct craniosynostosis syndromes: Crouzon, Jackson-Weiss, and Pfeiffer syndrome (Fig. 4B; see below). Molecular analyses of over 80 cases of Crouzon syndrome has revealed 30 different FGFR2 mutations. The majority of these mutations occur in IgIII, the ligand binding region of the extracellular domain. Over half of the mutations are in exon 9 (exon IIIc or exon B), which forms the second half of IgIII, with the others occurring primarily in exon 7 (exon IIIa or exon U), which forms the first half of IgIII (Fig. 114-4A,B). The mutations are missense, splicing, and small in-frame deletions and insertions. As in Pfeiffer syndrome, in the majority of cases, these mutations result in the loss or gain of the amino acid cysteine, and the most frequently mutated amino acid is C342. Mutation of this cysteine residue is predicted to affect the formation of the disulfide bond, which is critical to forming the tertiary structure of IgIII and determine its binding specificity and affinity. FGFR2 normally undergoes alternative splicing resulting in isoforms with different spatiotemporal expression (Fig. 114-4A). The BEK or FGFR2 isoform that includes the product from exon 9 (exon IIIc or exon B) is expressed predominantly in primordial bone. The keratinocyte growth factor receptor (KGFR) isoform, which includes the product from exon 8 (exon IIIb or exon K), is expressed primarily in epithelial lined structures. FGFR2 mutations that only affect exon 9 and the specific expression pattern of this isoform (BEK) in primordial bone are consistent with skeletal changes seen in Pfeiffer and Crouzon syndromes. In contrast, mutations in exon 7 present in both isoforms (BEK and KGFR) would be expected, not only to result in a skeletal phenotype but also in anomalies involving epithelial structures. Interestingly, there are documented cases of low imperforate anus and anorectal fistulas in Crouzon syndrome patients with exon 7 mutations. FGFR2 mutations in both exon 7 and 9 have also been identified in Jackson-Weiss syndrome. In fact, Jackson-Weiss, Crouzon, and Pfeiffer syndromes share a common mutation, C342R. The original Amish family in the description of Jackson-Weiss syndrome had a mutation predicting an A344G substitution that was also found in Crouzon syndrome. The extensive phenotypic variability in this large kindred suggests the involvement of modifying gene(s). A number of these FGFR2 mutations, including C342R, have been demonstrated to cause constitutive activation of the receptor by aberrant intermolecular disulfide-bonding. FGFR2 mutations were also detected in Beare-Stevenson cutis gyrata syndrome. In three sporadic cases, two novel missense mutations were identified predicting an amino acid to be replaced by a cysteine. Two patients had the identical Y375C substitution in the transmembrane domain, and one had a S372C change in the carboxyl-terminal end of the linker region between IgIII and the transmembrane domain (Fig. 114-4B,C). In two other patients, neither of these mutations as found suggesting further genetic heterogeneity. The phenotype of this syndrome is consistent with the mutations occurring in regions that would affect both the BEK (exon 9 or IIIc) and KGFR (exon 8 or IIIb) isoforms and their coexpression, because the BEK isoform is expressed normally in

1036

SECTION XIII / GENETIC BASIS FOR CONGENITAL MALFORMATIONS

Figure 114-4 (A) Schematic diagram of the mammalian fibroblast growth factor receptor (FGFR) putative protein structure with signal peptide (hatched box), three immunoglobulin-like (Ig) domains (loops), and an acidic region (open box) between first and second Ig-like domain, transmembrane domain (black box) and a split tyrosine kinase domain (striped box). Alternative splicing with differential exon use, as shown for the human FGFR2 gene, generates receptor variants that differ in the second half of their third Ig-like domain and presumably their ligand-binding properties. Exons U (upstream, exon IIIa, or 7) and D (downstream containing transmembrane domain, exon 10) flank the alternatively used exons K (KGFR, exon 8) or B (BEK, exon IIIc, or 9) in the resulting mRNA, which constitute the IgIIIb or IgIIIc isoforms, respectively (see text for details). (B) Schematic diagram of the genes encoding FGFR1, FGFR2, and FGFR3 with the positions of the mutations reported in craniosynostosis (gray circles), including Pfeiffer, Apert, Crouzon, Crouzon with acanthosis nigricans, Beare-Stevenson cutis gyrata, and Jackson-Weiss syndrome, and short-stature syndromes (white circles) including achondroplasia, hypochondroplasia, thanatophoric, and SADDAN. (C) Nucleotide sequence of the interloop region between IgII and IgIII with three adjacent amino acids recurrently affected by unique mutations identified in Pfeiffer syndrome (FGFR1), Apert syndrome (FGFR2), FGFR3associated craniosynostosis syndrome (FGFR3), and thanatophoric dysplasia (FGFR3). It is of note that the identical C-to-G substitution in FGFR1, FGFR2, and FGFR3 predicting a proline to arginine change causes different craniosynostosis syndromes. The analogous amino acids in FGFR2 for Beare-Stevenson cutis gyrata syndrome and FGFR3 for thanatophoric dysplasia type 1 are mutated to produce an extra cysteine residue.

CHAPTER 114 / FGFR-RELATED SKELETAL DISORDERS

skeletal structures and the KGFR is expressed in tissues involved in the development of the skin, umbilicus, and anogenital organs. In Apert syndrome, two recurring mutations in exon U (or exon 7) of FGFR2 account for the overwhelming majority of several hundred cases studied worldwide. These are a 934C→G transversion resulting in a S252W substitution in the linker region between IgII and IgIII in about 70% and a 937C→G change resulting in a P253R substitution in 30% (Fig. 114-4B,C). These two distinct amino acid substitutions in FGFR2, which result in Apert syndrome, are both bulky substitutions that may alter the relative orientation of IgII and IgIII, thus affecting the binding of FGF ligands or dimerization with other FGFRs. FGFR3-associated coronal synostosis is caused by a recurrent single-point mutation (749C→G) predicting a P252R substitution in the IgII/III interlinker region of FGFR3 in over 30 unrelated families. Interestingly, this common mutation occurs precisely at the analogous position within the FGFR3 protein as the mutations in FGFR1 (P252R) in Pfeiffer syndrome and FGFR2 (P253R) in Apert syndrome (Fig. 114-4B,C). It has been speculated that the FGFR1 Pro252Arg and FGFR3 Pro250Arg mutations influence FGFR2 signaling, thereby causing the craniosynostosis phenotype. The observation that patients with FGFR1 or FGFR3 mutations tend to have a milder phenotype than those seen with FGFR2 mutations and that Apert syndrome, the most severe phenotype in the spectrum of FGFR2 mutations, is caused by the equivalent Pro252Arg mutation in FGFR2, is consistent with this possibility. In Crouzon syndrome with acanthosis nigricans, a recurrent mutation predicting a A391E substitution occurs in the transmembrane domain of FGFR3 (Fig. 114-4B). The phenotypes associated with FGFR3 mutations are consistent with its spatiotemporal expression. FGFR3 is expressed in primordial bone, ectoderm, epithelial-lined structures, and skin. The majority of FGFR3 mutations are reported in dwarfism conditions, achondroplasia, thanatophoric dysplasia, and hypochondroplasia (Fig. 114-4B). The association of nondwarfism and even nonskeletal phenotypes with FGFR3 mutations reveals the potential for a wide range of pleiotropic effects as well as locus heterogeneity in Crouzon syndrome. Thus, the effects of the FGFR3 mutations in Crouzon syndrome with acanthosis nigricans, FGFR3-associated coronal synostosis syndrome, and those for the different dwarfism syndromes must be functionally distinct. More than 95% of cases of achondroplasia are caused by one of two mutations in FGFR3 (1138G→A is most common, 1138G→C is least common), both of which result in the substitution of an arginine residue for the normal glycine at amino acid 380 in the transmembrane domain of the protein. Interestingly, this alteration is one of the most frequent mutation in the human genome, with an estimated mutation frequency between 5.5 × 10–6 and 2.8 × 10–5 per gamete generation. The transmembrane domain affected by this mutation is important, not only as a cell-membrane anchor, but also in signal transduction. Hypochondroplasia is less genetically homogeneous, with approximately 70% caused by a K540L substitution in the first tyrosine kinase domain of the FGFR3 protein. Mutations for the rest have not yet been identified. Several linkage studies have suggested that some cases of hypochondroplasia may be caused by mutations at a locus distant and distinct from FGFR3, but mutations at other gene loci have not yet been recognized. Thanatophoric dysplasia (TD) Type 1 is caused by series of mutations in the FGFR3 gene, one of which results in R248C,

1037

located in the Ig II/II interloop region (Fig. 114-4B,C). In contrast, Type 2 (TD2; associated with straight limbs with or without cloverleaf skull) is associated with a unique recurring mutation predicting a K650N substitution in the second tyrosine kinase domain of FGFR3. The identical amino acid (K650M) is altered in three patients with SADDAN. In vitro studies have demonstrated that the G380R substitution in achondroplasia, the K540L change in hypochondroplasia, and the K650N alteration in TD2 all are associated with constitutive (ligand-independent) activation of FGFR3. The downstream effects of this constitutive activation and the mechanism whereby it is translated into the phenotypic effects of these mutations are not, as yet, understood.

MOLECULAR PATHOPHYSIOLOGY OF DISEASE How these FGFR mutations alter the intracellular signaling pathway to cause the autosomal dominant phenotypes seen in the craniosynostosis and dwarfism syndromes is poorly understood. The abnormal phenotypes in the FGFR-related skeletal disorders are not simply caused by a deletion of one copy of the gene. This is supported by naturally occurring deletions in humans and also by gene knockout experiments in the mouse. Although individuals with a deletion of the short arm of chromosome 4 have one copy of FGFR3 deleted, they do not have any of the characteristic findings of achondroplasia or thanatophoric dysplasia. Thus, the phenotypes in these skeletal disorders cannot be explained by the classic model of a mutation-induced haploinsufficiency. Further evidence for the unique nature of the identified FGFR mutations comes from knockout studies in transgenic mice. Heterozygous animals that are deleted for one copy of Fgfr1, Fgfr2, or Fgfr3, respectively, appear to be developmentally and anatomically normal. Recent studies provide further evidence that the abnormal phenotypes in FGFR-associated craniosynostosis and dwarfism syndromes are not a result of a loss-of-function, but rather from a gain-of-function of the mutated receptor molecule. This evidence comes from mutation-induced receptor activation and targeted Fgfr gene disruption experiments in the mouse. Biochemical studies have demonstrated that an FGFR2 mutation (Cys342Tyr) found in Pfeiffer and Crouzon syndromes and FGFR3 mutations in achondroplasia and types 1 and 2 thanatophoric dysplasia result in ligand-independent, constitutive activation of the receptor. This conclusion is supported by studies of Fgfr3 knockout mice, which exhibit overgrowth of the long bones and axial skeleton suggesting that FGFR3 is a negative regulator of bone growth. However, individuals with FGFR3-associated coronal synostosis and Crouzon syndrome with acanthosis nigricans have normal longbone growth. It is therefore doubtful that this mutation leads to constitutive FGFR3 activation in the same tissue distribution as the achondroplasia, TD-1, and TD-2 mutations.

MANAGEMENT AND FUTURE DIRECTIONS Multidisciplinary team approach is required for the management of the craniofacial, audiologic, ocular, oral, dental, skeletal, hand, foot, and skin abnormalities found in these conditions. Management is symptomatic and psychosocial support is often indicated. Recommendations for follow-up and management are reviewed for chondrodysplasias by Horton and Hecht (1993) and for craniosynostosis by Cohen (1986). Further studies of development pathways, animal models, cell biologic aspects, and biochemical signal transduction components involved in the pathogenesis and in the modification of the phenotype are required. Chemical

1038

SECTION XIII / GENETIC BASIS FOR CONGENITAL MALFORMATIONS

inhibitors of FGFRs have been developed. As yet, no approaches to the management of these conditions have been devised using techniques of gene therapy.

ACKNOWLEDGMENTS This work was supported in part by NIH Grants P50 DE11131 and RO1 DE11441 (EWJ.), by NIH grants HD2873 and HD29862 (M.M.) and the Division of Intramural Research, National Human Genome Research Institute NIH (C.A.F.; M.M.).

SELECTED REFERENCES Baker KM, Olson DS, Harding CO, Pauli RM. Long-term survival in typical Thanatophoric dysplasia type 1. Am J Med Genet 1997;70:427–436. Bellus GA, Bamshad MJ, Przylepa KA, et al. Severe achondroplasia with developmental delay and acanthosis nigricans (SADDAN): phenotypic analysis of a new skeletal dysplasia caused by fibroblast growth factor receptor 3 Lys650Met mutation (submitted). Bellus GA, Gaudenz K, Zackai EH, et al. Identical mutations in three different fibroblast growth factor receptor genes in autosomal dominant craniosynostosis syndromes. Nat Genet 1996;14:174–176. Bellus GA, McIntosh I, Smith EA, et al. A recurrent mutation in the tyrosine kinase domain of fibroblast growth factor receptor 3 causes hypochondroplasia. Nat Genet 1995;10:357–359. Cohen MM Jr. Craniosynostosis: Diagnosis, Evaluation, and Management. New York: Raven, 1986; pp. 413–590. Cohen MM Jr, Kreiborg S. An updated pediatric perspective on the Apert syndrome. Am J Dis Child 1993;147:989–993. Cohen MM Jr, Kreiborg S. Skeletal abnormalities in the Apert syndrome. Am J Med Genet 1993;47:624–632. Cohen MM Jr, Kreiborg S. The hands and feet in Apert syndrome. Am J Med Genet 1995;56:82–96. El Ghouzzi V, Le Merrer M, Perrin-Schmitt F, et al. Mutations of the TWIST gene in Saethre-Chotzen syndrome. Nat Genet 1997;15:42–46. Gripp KW, McDonald-McGinn DM, Gaudenz K, et al. Identification of a genetic cause for isolated unilateral coronal synostotis: A unique mutation in the fibroblast growth factor receptor 3. J Pediatr 1998, in press. Hall BD, Cadle RG, Golabi M, Morris CA, Cohen MM Jr. BeareStevenson cutis gyrata syndrome. Am J Med Genet 1992;44:82–89. Horton WA, Hecht JT. The chondroplasias. In: Royce, PM; Steinmann, B, eds. Connective Tissue and its Heritable Disorders: Molecular, Genetic, and Medical Aspects. New York: Wiley-Liss, 1993; pp. 641–675. Howard TD, Paznekas WA, Green ED, et al. Mutations in TWIST, a basic helix-loop-helix transcription factor, in Saethre-Chotzen syndrome. Nat Genet 1997;15:36–41. Jabs EW, Muller U, Li X, et al. A mutation in the homeodomain of the human MSX2 gene in a family affected with autosomal dominant craniosynostosis. Cell 1993;75:443–450. Jabs EW, Li X, Scott AF, et al. Jackson-Weiss and Crouzon syndromes are allelic with mutations in fibroblast growth factor receptor 2. Nat Genet 1994;8:275–279. Lajeunie E, Ma HW, Bonaventure J, Munnich A, Le Merrer M, Renier D. FGFR2 mutations in Pfeiffer syndrome. Nat Genet 1995;9:108.

Meyers GA, Orlow SJ, Munro IR, Przylepa KA, Jabs EW. Fibroblast growth factor receptor 3 (FGFR3) transmembrane mutation in Crouzon syndrome with acanthosis nigricans. Nat Genet 1995; 11:462–464. Moloney DM, Wall SA, Ashworth GJ, et al. Prevalence of Pro250Arg mutation of fibroblast growth factor receptor 3 in coronal craniosynostosis. Lancet 1997;349:1059–1062. Muenke M, Gripp KW, McDonald-McGinn DM, et al. A common mutation in the fibroblast growth factor receptor 3 (FGFR3) gene defines a new craniosynostosis syndrome. Am J Hum Genet 1997;60:555–564. Muenke M, Schell U. Fibroblast growth factor receptor mutations in human skeletal disorders. Trends Genet 1995;11:308–313. Muenke M, Schell U, Hehr A, et al. A common mutation in the fibroblast growth factor receptor 1 gene in Pfeiffer syndrome. Nat Genet 1994;8:269–274. Oldridge M, Lunt PW, Zackai EH, et al. Genotype-phenotype correlations for nucleotide substitutions in the IgII-IgIII linker of FGFR2. Hum Mol Genet 1997;6:137–143. Online Mendelian Inheritance in Man (OMIM). Center for Medical Genetics, Johns Hopkins University (Baltimore, MD) and National Center for Biotechnology Information, National Library of Medicine (Bethesda, MD), 1997. World Wide Web URL: http://www3.ncbi.nih.gov/omim/ Park W-J, Bellus GA, Jabs EW. Mutations in fibroblast growth factor receptors: phenotypic consequences during eukaryotic development. Am J Hum Genet 1995;57:748–754. Przylepa KA, Paznekas W, Zhang M, et al. Fibroblast growth factor receptor 2 mutations in Beare-Stevenson cutis gyrata syndrome. Nat Genet 1996;13;492–494. Reardon W, Winter RM, Rutland P, Pulleyn LJ, Jones BM, Malcolm S. Mutations in the fibroblast growth factor receptor 2 gene cause Crouzon syndrome. Nat Genet 1994;8:98–103. Rousseau F, Bonaventure J, Legeall-Mallet L, et al. Mutations in the gene encoding fibroblast growth factor receptor-3 in achondroplasia. Nature 1994;371:252–254. Rousseau PL, Saugier P, Le Merrer M, et al. Stop codon FGFR3 mutations in thanataphoric dwarfism type 1. Nature Genet 1995;10:11,12. Rutland P, Pulleyn LJ, Reardon W, et al. Identical mutations in the FGFR2 gene cause both Pfeiffer and Crouzon syndrome phenotypes. Nat Genet 1995;9:173–176. Schell U, Hehr A, Feldman GJ, et al. Mutations in FGFR1 and FGFR2 cause familial and sporadic Pfeiffer syndrome. Hum Mol Genet 1995; 4:323–328. Shiang R, Thompson LM, Zhu YZ, et al. Mutations in the transmembrane domain of FGFR-3 cause the most common genetic form of dwarfism, achondroplasia. Cell 1994;78:335–342. Tavormina PL, Bellus GA, Webster M, et al. A novel skeletal dysplasia with developmental delay and acanthosis nigricans is caused by a Lys650Met mutation in fibroblast growth factor receptor 3. Am J Hum Genet, in press. Tavormina PL, Shiang R, Thompson LM, et al. Thanatophoric dysplasia (types I & II) caused by distinct mutations in fibroblast growth factor receptor 3. Nat Genet 1995;9:321–328. Wilkie AOM, Slaney SF, Oldridge M, et al. Apert syndrome results from localized mutations of FGFR2 and is allelic with Crouzon syndrome. Nat Genet 1995;9:165–172.

CHAPTER 115 / AARSKOG-SCOTT SYNDROME

115

1039

Aarskog-Scott Syndrome JEROME L. GORSKI

INTRODUCTION Aarskog-Scott syndrome, or Faciogenital dysplasia (FGDY, MIM 305400), is an uncommon X-linked recessive developmental disorder that primarily affects skeletal morphogenesis. The condition was first described by Aarskog and Scott in the early 1970s. Mutations in FGD1, the gene responsible for AarskogScott syndrome, result in a developmental disorder affecting specific skeletal structures that include elements of the face, cervical vertebrae, and distal extremities. Genetic and biochemical analyses show that FGD1 encodes a guanine nucleotide exchange factor (GEF), or activator, for Cdc42, a member of the Rho family of Raslike GTPases. Rho proteins comprise a family of at least eight distinct proteins that are involved in the control of a wide variety of cellular functions, including the organization of the actin cytoskeleton, the control of cellular division, and the transcriptional regulation of gene expression. Together, members of the Rho protein family and their activators regulate cell shape, adhesion, and migration, properties that are involved in tissue morphogenesis. The identification that FGD1, the gene responsible for AarskogScott syndrome, is a RhoGEF and a component of the Rho signal transduction pathway suggests that other components of this signaling pathway will be found to be responsible for defects in mammalian morphogenesis.

CLINICAL FEATURES OF AARSKOG-SCOTT SYNDROME PHYSICAL MANIFESTATIONS OF THE DISEASE The Aarskog-Scott syndrome, or Faciogenital dysplasia (FGDY), phenotype consists of a characteristic set of facial and skeletal anomalies, disproportionate short stature, and urogenital malformations. The cardinal features of this disease are summarized in Table 115-1 and illustrated in Fig. 115-1. Facial Features The face is typically round and the forehead broad with ridging of the metopic suture. Facial features typically consist of widely-spaced eyes (hypertelorism), ptosis, down-slanting palpebral fissures, and a short up-turned (anteverted) nose (Fig. 115-1). The philtrum is commonly long and the maxilla is typically hypoplastic. A variety of external ear anomalies have been described, including low-set ears, posteriorly-rotated auricles, and thickened over-folded helixes.

From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

Musculoskeletal Features Impaired growth is another major manifestation of the disease; growth retardation usually becomes apparent during the first few years of life, and affected males rarely exceed 160 cm in height. Stature is disproportionate and the distal extremities are most severely shortened (Fig. 115-1). Hands and feet are broad and short. Interphalangeal joints are typically hypermobile; however, camptodactyly and contractures of the interphalangeal joints have also been observed. The majority of affected males have a pectus excavatum and inguinal hernias. Typically, relative to affected males, the phenotype of obligate female heterozygotes is mild and limited to relative short stature and a subtler form of the characteristic craniofacial anomalies. Radiographic Abnormalities Radiographic findings are usually limited to the cervical spine and distal extremities; these abnormalities are summarized in Table 115-1. About half of the affected males have a cervical spine abnormality such as spina bifida occulta, odontoid hypoplasia, fused cervical vertebrae, and ligamentous laxity with subluxation. Radiographic abnormalities of the hands and feet typically consist of shortened digits, hypoplasia of the terminal phalanges, clinodactyly, fusion of the middle and distal phalanges, and retarded bone maturation. Other radiographic abnormalities include maxillary hypoplasia, additional pairs of ribs and other segmentation anomalies, and calcified intervertebral disks. Urogenital Anomalies The scrotum typically appears bifid. Scrotal folds commonly extend ventrally around the base of the penis to form what resembles a shawl. Cryptorchidism is common; penile hypospadius has also been reported. Other Features Although mild-to-moderate mental retardation has been described in some affected males, it does not appear to be a consistent feature. Affected males have been observed to have a delayed eruption of teeth; some have congenitally missing teeth. Other anomalies occasionally occur including cleft lip and palate and congenital heart defects. Some males have a single palmar crease or distally placed axial triradii. Affected males and obligate carrier females both appear to have normal fertility. Epidemiology Population surveys estimate that FGDY occurs with a recognized frequency of approximately 1 per one million in the general population. However, because it is likely that mildly affected individuals would not be detected, it is probable that the disease frequency is actually higher than this estimate. EMBRYOLOGIC CORRELATIONS Mutations that result in the Aarskog-Scott syndrome alter the size and shape of a limited number of specific cartilaginous and bony elements but leave other skeletal structures unaffected. Similar abnormalities have been observed in the short-ear (se) and brachypodism (bp) mouse

1039

1040

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Table 115-1 Primary Clinical and Radiographic Features of FGDY Anatomical region or system Craniofacial Musculoskeletal Urogenital Miscellaneous Radiographic Features aItems

Clinical features Broad forehead, abnormally formed ears, hypertelorism, down-slanted palpebral fissures, ptosis, maxillary hypoplasia, anteverted nostrils, hypodontia and malocclusion (cleft lip and palate, enamel hypoplasia)a Disproportionate short stature, distally shortened limbs and brachydactyly, pectus excavatum, soft tissue syndactyly of the digits, interphalangeal joint hypermobility, camptodactyly and clinodactyly, broad feet, inguinal hernias (single palmar crease, abnormal dermatoglyphics) Shawl and bifid scrotum, crytorchidism (hypospadius, renal hypoplasia) Mild-to-moderate mental retardation, strabismus, growth hormone deficiency (congenital heart defects) Spina bifida occulta, odontoid hypoplasia, cervical vertebral defects and ligamentous laxity, calcified intervertebral disks, additional ribs, hypoplastic phalanges, retarded bone age (osteochondritis dissecans)

in parentheses are less commonly observed.

sign of skeletal morphogenesis is the aggregation of mesenchymal cells into regions of high cell density termed condensations; these condensations represent the initial outlines of future skeletal elements. The mesenchymal cells within these condensations later elaborate extracellular matrix and differentiate into cartilage and bone cells; final skeletal shapes result from a combination of partial cell loss (apoptosis), the fusion of adjacent condensations, and the coordinated resorption and deposition of bone by osteoclasts and osteoblasts. Se and bp embryos exhibit abnormalities in the size and shape of specific skeletal condensations, anomalies that result in the mutant phenotype. The similarity of these mutations to Aarskog-Scott syndrome suggests that, like bp and se mice, FGDY males may have an altered pattern of skeletogenesis that affects a limited number of specific skeletal progenitors. Additional analyses will be required to verify this hypothesis. DIFFERENTIAL DIAGNOSIS Several other inherited conditions, including Noonan syndrome, Robinow syndrome, and Leopard syndrome, share clinical features with Aarskog-Scott syndrome; shared features include short stature, hypertelorism, and hypogonadism. Pseudohypoparathyroidism and hydantoin embryopathy also share several physical features with FGDY. In combination with the distinctive radiographic abnormalities, the characteristic pattern of craniofacial abnormalities, disproportionate short stature, shortening of the distal extremities, and characteristic urogenital anomalies distinguish Aarskog-Scott syndrome from these other conditions. Unlike these other conditions, FGDY is an X-linked recessive trait; therefore, a demonstrated X-linked pattern of inheritance assists in confirming the diagnosis.

MOLECULAR GENETICS OF AARSKOG-SCOTT SYNDROME

Figure 115-1 Aarskog-Scott syndrome. Facial features are characterized by a broad forehead, a widow’s peak, hypertelorism, bilateral ptosis of the upper eyelids, midface hypoplasia, a depressed nasal bridge, anteverted nose, and low-set ears. Stature is short and disproportionate with shortened distal extremities; hands and feet are short and broad. Bilateral inguinal herniorrhaphy scars are present. Consequential to scrotal folds joining ventrally over the base of the penis, an abnormal penoscrotal configuration is present. (Courtesy R Gorlin, University of Minnesota.)

mutants. During development, se and bp embryos manifest abnormalities in the formation of a characteristic subset of skeletal progenitors. During mammalian embryogenesis, the first observable

AARSKOG-SCOTT SYNDROME GENE, FGD1, MAPS TO REGION Xp11.21 Gene Localization Pedigree analyses of families segregating FGDY strongly suggested an X-linked recessive pattern of inheritance; subsequent genetic linkage studies mapped the responsible locus to the pericentric region of the X chromosome. The observation of a mother and son who both displayed all of the major features of FGDY in association with a reciprocal X;8 chromosome translocation tentatively localized the disease gene to the X-chromosomal breakpoint, region Xp11.21. The observation that, in the affected female, the translocated X chromosome was active in the majority of cells examined, suggested that the translocation breakpoint directly interrupted the disease gene. By isolating the rearranged X chromosome in a rodent somatic cell hybrid, mapping experiments sublocalized the FGDY-specific breakpoint to an estimated 350-kb region within distal Xp11.21.

CHAPTER 115 / AARSKOG-SCOTT SYNDROME

1041

Figure 115-2 RhoGEFs activate Rho proteins by catalyzing the exchange of GDP for GTP. A variety of stimuli lead to the activation of Rho protein family members via RhoGEFs, including the p21 GTPase Ras, receptor protein tyrosine kinases, and G-protein-coupled receptors. Activated Rho leads to modified cell morphology by a reorganization of the actin cytoskeleton, a modulation of gene transcription by the activation of mitogen-activated protein (MAP) kinase cascade, and the sequential activation of other Rho family member proteins. RhoGAP facilitates the hydrolysis of GTP and the Rho protein inactivation.

Positional Cloning The FGDY-specific X;8 breakpoint was used as a molecular signpost to positionally clone the AarskogScott syndrome gene. Using DNA markers flanking the diseasespecific breakpoint, a YAC contig of the region was assembled; DNA clones derived from this contig were used to identify regional transcripts and identify a cDNA clone, designated FGD1, that spanned the disease-specific breakpoint. A number of lines of evidence indicated that FGD1 was the gene responsible for Aarskog-Scott syndrome: 1. FGD1 was mapped to Xp11.21, the region known to contain the Aarskog-Scott syndrome gene. 2. The FGD1 gene was directly disrupted by a disease-specific X;8 translocation breakpoint. 3. FGD1 mRNA was expressed in tissues involved in the disease phenotype including fetal craniofacial bones. The identification of additional FGD1 mutations in affected FGDY males, including an insertional mutation predicted to result in a severely abbreviated and nonfunctional FGD1 protein, confirmed that FGD1 was the gene responsible for the Aarskog-Scott syndrome. Genomic Organization The complete FGD1 cDNA is 4439 nucleotides in length, a result consistent with the 4.4-kb transcript detected by Northern blot analysis. The FGD1 transcript contains a 2883-nucleotide open reading frame (ORF) that encodes a protein of 961 amino acids with a predicted mass of 107 kDa. Comparative sequence analyses show that the FGD1 gene is composed of 18 exons that span 51 kb of genomic DNA. Exons range from 1210 to 31 bp in size; intron sizes range from 23 kb to 106 bp in size. RNase protection and primer extension analyses show that the FGD1 promoter belongs to a GC-rich, TATA-less class of promoters.

Mutation Analyses Genomic analyses have failed to identify a common structural rearrangement among FGDY patients. These results suggest that FGD1 mutations will be highly heterogeneous. Expression studies indicate that FGD1 is not expressed in readily biopsied tissues such as lymphocytes and fibroblasts; these results suggest that DNA-based analyses will be necessary to identify individual FGD1 mutations. A determination of the genomic structure of the FGD1 gene provides a means for designing DNA-based diagnostic analyses to detect FGD1 mutations. For example, primers designed to polymerase chain reaction amplify each of the FGD1 exons may provide a means of detecting FGD1 mutations by performing single-strand conformational polymorphism (SSCP) analysis. FGD1 ENCODES A RHOGEF, A REGULATOR OF CDC42 SIGNAL TRANSDUCTION Comparative Analyses Comparative sequence analysis indicated that FGD1 encoded a Rho guanine nucleotide exchange factor (RhoGEF). RhoGEFs form a family of cytoplasmic proteins that activate the GTPase Ras-like family of Rho proteins by catalyzing the exchange of bound GDP for free GTP (Fig. 115-2). The p21 GTPase superfamily and the biological roles of the family members are summarized in Table 115-2. The RhoGEF family consists of at least 16 distinct members. Like FGD1, all members contain a 200-amino acid RhoGEF motif; in a number of RhoGEF family members, this domain has been shown to be necessary and sufficient for catalyzing the GDP/GTP exchange for Rho protein family members. In addition, all RhoGEF domains are paired with pleckstrin homology (PH) domains; PH domains are thought to be essential for the proper cellular localization of RhoGEF proteins to the cytoskleton. In addition to the RhoGEF and PH domains, most RhoGEF proteins contain other types of functional domains that are commonly found in signaling

1042

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Table 115-2 p21 GTPase Superfamily Members Family Ras Rho Rab/ARF Ran

Family members

Biological function

H-ras, Ki-ras, N-ras, R-ras, rap1A/B, rap2A/B, ralA/B, TC21 RhoA, RhoB, RhoC, RhoG, Cdc42, Rac1, Rac2, TC10

Regulation of cell growth Regulation of cell differentiation Actin cytoskeleton organization Regulation of gene expression Regulation of cell growth Vesicle transport Importation of nuclear proteins

Rab1,…,Rab26, ARF1,…,ARF6 Ran1

Figure 115-3 Schematic representation of the domain structure of the FGD1 protein compared to other RhoGEF proteins. Domains are drawn approximately to scale. FGD1 contains at least four distinct domains, including a RhoGEF domain and a pleckstrin homology (PH) domain, motifs common to all RhoGEF family members. FGD1 also contains a cysteine-rich zinc finger-like (Cys-rich) motif and a putative Srchomology 3-binding (Grb2-binding) region. Vav contains Src-homology 3 (SH3), Src-homology 2 (SH2), and a putative diacylglycerol/ phorbol ester-binding (DAG/PE) zinc butterfly motif; Bcr contains a Rho GTPase activator protein (RhoGAP) domain. RasGRF and mSos1 contain a Ras guanine nucleotide exchange factor (RasGEF) domain.

molecules. For example, although RasGRF and mSos1 both contain RhoGEF domains, they also contain RasGEF motifs (Fig. 115-3), a result suggesting that these proteins may act as two-headed exchange factors by mediating the activation of both the Rho and Ras signaling pathways. Most of the RhoGEFs identified, including Dbl, ect2, and Ost were discovered by virtue of their transforming capability through gene transfer experiments; other RhoGEFs, including Cdc24, Bcr, mSos1, RasGRF, and Vav, were identified by their role in cell growth regulation. As shown in Fig. 115-2, the signaling cascade that couples RhoGEFs to upstream components remains elusive. The specific biological function of many of the RhoGEF family members remains to be determined. Biochemical Studies In vivo and in vitro studies indicate that FGD1 is a RhoGEF for Cdc42, a member of the Rho protein family. When microinjected into cultured cells, the FGD1 RhoGEF domain induces fibroblasts to form filopodia, actin-associated membrane complexes generated by activated Cdc42. The FGD1 RhoGEF domain specifically binds to the Cdc42 protein, and FGD1-dependent filopodia formation is blocked by complexing Cdc42 to other Cdc42-binding proteins. In addition, in FGD1 expressing fibroblasts, stress-activated protein kinase (SAPK)/c-Jun N-terminal kinase (JNK) activity is stimulated in a manner similar to that

obtained with constitutively activated Cdc42. Together, these results indicate that FGD1 is a specific RhoGEF for Cdc42. FGD1 ENCODES ADDITIONAL STRUCTURAL DOMAINS A Zinc Finger Motif A comparative analysis of FGD1 showed that it contained two additional conserved structural motifs (Fig. 115-3). Like the proto-oncogene RhoGEF member Vav, the 3' region encoded a 50-amino acid cysteine-rich zinc finger-like motif that was similar to, but distinct from, the diacylglycerol/ phorbol ester-binding regulatory domain of protein kinase Cγ. The Vav zinc finger motif is functionally significant; mutations in the conserved cysteine residues abolish transforming activity. Similar putative regulatory domains have been identified in a variety of Ras-associated proteins, including the Raf proto-oncogene, the RhoGAP n-chimerin, and diacylglycerol kinase. The observation that the predicted FGD1 sequence contains a zinc finger-like domain suggests that the FGD1 protein may interact with lipid second messenger molecules. Therefore, it is possible that the FGD1 protein may interact with, or be modified by, the components of multiple signal transduction pathways. A Putative Grb2-Binding Region An analysis of the FGD1 protein showed that the 5' region was remarkably proline-rich and that proline constituted 22% of the first 250 amino acid residues.

1043

CHAPTER 115 / AARSKOG-SCOTT SYNDROME

Table 115-3 Actin-Associated Membrane Complexes Complex type

Involved Rho family member

Focal adhesion

Rho

Filopodia Lamellipodia Cortical stress fibers Adherens junction

Cdc42 Rac Rho ?

Biological function Cellular adhesion Cellular morphology Cellular movement Cellular morphology Cellular adhesion

Proteinsa Integrin, vinculin, talin, PTK, FAK, tensin, paxillin, zyxin, ERM, α-actinin Vinculin, fimbrin, talin, PTK, α-actinin, ERM Talin, fimbrin, integrin, PTK, α-actinin, ERM Fodrin-spectrin, ankyrin, PTK, adhesion molecules Cadherin, catenin, vinculin, PTK, filamin, α-actinin, ERM

aKnown to comprise an actin-associated membrane complex; PTK, protein tyrosine kinase; FAK, focal adhesion kinase; ERM, ezrin, radixin, and moesin.

Since proline-rich regions may contain Src-homology 3 (SH3) binding domains, the 5' portion of the derived FGD1 protein sequence was compared to other sequences known to contain SH3 binding sites. An analysis showed that FGD1 contained two putative SH3 binding sites that exhibited strong similarity to the functionally significant regions of several proteins with demonstrated Grb2 binding activity, including the RasGEF, mSos1. Recently, it has been shown that Grb2 selectively binds the proline-rich motifs of the mSos1 protein to form a link in a signal transduction pathway that functionally ties tyrosine kinase receptors to Ras. Among the identified Ras and RhoGEF family members, mSos1 is unique in containing an SH3 (or Grb2)-binding domain. The identification of a putative proline-rich SH3 binding domain in FGD1 implies that, like Sos, the location and/or activity of the FGD1 protein may be modified by Grb2-like proteins.

MOLECULAR PATHOPHYSIOLOGY OF AARSKOG-SCOTT SYNDROME Identified FGD1 mutations included a frame shift mutation and a chromosomal rearrangement, mutations predicted to result in the total absence of FGD1 protein; therefore, a loss of FGD1 appears to result in the FGDY phenotype. In S. cerevisiae, S. pombe, D. melanogaster, and C. elegans, the loss of a RhoGEF activity is functionally equivalent to the loss of the target Rho protein. Although the structural and functional diversity of the RhoGEF protein family indicates that FGD1 is not the only Cdc42 RhoGEF, it is logical to expect that within expressing cells FGD1 mutations will reduce or modify Cdc42 activity. Therefore, it is likely that an examination of the molecular biology of Cdc42 and the other Rho proteins will illuminate the biological role of FGD1. RHO PROTEINS REGULATE THE ORGANIZATION OF THE ACTIN CYTOSKELETON AND GENE EXPRESSION The mammalian Rho protein family (listed in Table 115-2) consists of at least eight distinct proteins and three subfamilies: Rho, Cdc42, and Rac. Each Rho protein has at least 50–55% homology with any of the others and 30% homology to Ras. Regulation of the Actin Cytoskeleton Rho proteins are part of a signal transduction pathway that modulates a cell’s response to external stimuli by reorganizing the actin cytoskeleton to regulate cellular morphology. Regulated changes in the actin cytoskeleton are required for cytokinesis, cell motility, and cell–cell interactions. Therefore, by regulating the actin cytoskeleton, Rho proteins and their activators regulate cell shape, adhesion, and migration, properties critical to tissue morphogenesis. Expressed in cells, Rho proteins elicit a distinctive actin-associated membrane complex; these complexes, their biological function, and the involved Rho proteins are listed in Table 115-3. In fibroblast cells, microinjection analyses show that activated Cdc42 stimulates the

Figure 115-4 Rho protein family members are activated in a hierarchical cascade. The activation of Cdc42 by FGD1 leads to the sequential activation of Rac and Rho. Activated Cdc42 results in the formation of filopodia, whereas activated Rac leads to the formation of lamellipodia; activated Rho results in the formation of actin stress fibers and focal adhesions. Cdc42 and Rac activate the stress-activated protein kinase (SAPK)/c-Jun N-terminal kinase (JNK) mitogen-activated protein kinase (MAPK) cascade via the Ser/Thr kinase PAK; Rho is required for signaling via PAK to modulate gene transcription through the serum response factor (SRF).

formation of microspikes, or filopodia; in S. cerevisiae, Cdc42 is required for polarized cell growth and bud formation. Related studies show that Rac regulates the formation of lamellipodia and cell ruffles, and that Rho stimulates the formation of focal adhesions and cortical stress fibers. These same experiments show that Rho proteins are activated in a hierarchical cascade (Fig. 115-4), and

1044

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

that activated Cdc42 stimulates the sequential activation of Rac and Rho to sequentially elicit filopodia, lamellipodia, and stress fibers. More recently, use of cell-free assays and intact cell systems show that Rho regulates several enzymes involved in phosphoinositol metabolism, including phosphoinositide 3-kinase, phosphatidylinosital 4-phosphate 5-kinase, and phospholipase D. These results suggest that Rho may regulate the reorganization of the actin cytoskeleton through the metabolism of phospholipid metabolites. However, the exact relationship between Rho, phospholipid metabolism, and the actin cytoskeleton remains to be determined. Mitogen-Activated Protein Kinase (MAPK) Cascade Rho proteins are also part of a signal transduction pathway that activates MAPK cascades to modulate gene transcription. Kinases belonging to the MAPK family are used throughout evolution to control cellular responses to external signals, such as growth factors, inductive signals, and nutrient status. MAPKs have received particular attention, because many of these enzymes translocate to the cell nucleus to modulate the action of transcription factors. Recent studies show that Cdc42 and Rac activate the stress-activated protein kinase (SAPK)/c-Jun N-terminal kinase (JNK) MAPK cascade via the Ser/Thr kinase PAK. In contrast, Rho is required for modulating gene transcription through the serum response factor (SRF), results that suggest the existence of a novel signaling pathway (Fig. 115-4). Implications for Morphogenesis In addition to the identification of FGD1 as a RhoGEF, several lines of evidence indicate that Rho proteins are critical to development. In Drosophila, Cdc42 and Rac1 dominant-negative mutations result in specific developmental defects, including anomalies in the outgrowth of peripheral neurons and abnormal myoblast fusion, defects of cellular polarity that are consistent with the role of Rho proteins in regulating cellular morphology. Cdc42 and Rac1 mutations result in distinct phenotypes, results suggesting that different Rho proteins play specific roles during development. Transgenic mice carrying dominant-negative Rac mutations have similar developmental defects in neurogenesis. In addition, Cdc42 mutations in S. cerevisiae and S. pombe result in abnormalities in cellular morphology. FGD1 mutations could alter morphogenesis in at least two ways. Since FGD1 stimulates Cdc42 to activate the SAPK/JNK MAPK cascade, consequential to FGD1 mutations, diminished Cdc42 activity could alter SAPK/JNK activity and critically change patterns of gene transcription. Alternatively, since FGD1 stimulates Cdc42 to form filopodia, as a result of FGD1 mutations, diminished Cdc42 activity could alter cellular morphology. Since the actin cytoskeleton regulates cell shape, adhesion, and migration, altered patterns of cellular morphology are likely to play a critical roles in skeletogenesis. Additional studies will be necessary to determine how FGD1 mutations perturb human morphogenesis.

FUTURE DIRECTIONS With advances in the molecular understanding of disease processes, it has been appreciated that many diseases, such as the proliferative diseases of cancer, atherosclerosis, and psoriasis, result from a malfunction in signal transduction. This recognition has led to intensive research and the development of therapies based on the modification of cellular signaling in diseased cells. The identification of FGD1 as a RhoGEF and a component of Rho signal transduction suggests that other inherited abnormalities of human morphogenesis may also be the result of defects in embry-

onic signal transduction. Recent experiments have shown that cellular signal transduction pathways can be successfully modulated by a variety of reagents, including small molecules, antibodies, DNA encoding dominant-negative proteins, antisense RNA, and target-specific RNA ribozymes. The continued development of these reagents may provide a means for selectively correcting altered patterns of embryonic signal transduction in the future. Several Ras-associated proteins have been found to be responsible for human genetic diseases. Neurofibromin, the protein defective in von Recklinghausen neurofibromatosis, contains a RasGAP domain homologous to the catalytic domains of p120-GAP. Tuberin, the gene responsible for the form of tuberous sclerosis mapped to chromosome 16 (TSC2), has some homology to the GAP3 protein. The gene responsible for choroderemia (CHM), a retinal degeneration syndrome, was found to be similar to a Rab geranylgeranyl transferase. Among identified RhoGEF family members, FGD1 is the first to be directly implicated in causing an inherited human disease. However, the involvement of FGD1 in human morphogenesis implies that other components of the Rho signaling cascade may also be critical to embryonic development and responsible for related and/or unrelated inherited birth defects. An analysis of FGD1, in the context of its normal function and regulation, is likely to provide important new information regarding the roles that RhoGEF proteins and, by inference, Rho proteins play in signal transduction and mammalian development.

SELECTED REFERENCES Aarskog D. A familial syndrome of short stature associated with facial dysplasia and genital anomalies. J Pediatr 1970;77:856–861. Bard J. Morphogenesis: The Cellular and Developmental Processes of Developmental Anatomy. Cambridge, UK: Cambridge University Press, 1990, pp. 120–180. Boguski MS, McCormick F. Proteins regulating Ras and its relatives. Nature 1993;366:643–653. Bray D. Protein molecules as computational elements in living cells. Science 1995;376:307–312. Cerione RA, Zheng Y. The Dbl family of oncogenes. Curr Opin Cell Biol 1996;8:216–222. Chant J, Stowers L. GTPase cascades choreographing cellular behavior: movement, morphogenesis, and more. Cell 1995;81:1–4. Erlebacher A, Filvaroff EH, Gitelman SE, Derynck R. Toward a molecular understanding of skeletal development. Cell 1995;80:371–378. Gebbink MF, Kranenburg O, Poland M, van Horck FP, Houssa B, Moolenaar WH. Identification of a novel, putative Rho-specific exchange factor and a RhoA-binding protein: control of neuronal morphology. J Cell Biol 1997;137:1603–1613. Gorlin RJ, Cohen MM Jr, Levin LS. Syndromes of the Head and Neck, 3rd ed. Oxford: Oxford University Press, 1990; pp. 295–297. Hall A. Small GTP-binding proteins and the regulation of the actin cytoskeleton. Annu Rev Cell Biol 1994;10:31–54. Hall BK, Miyake T. The membranous skeleton: the role of cell condensations in vertebrate skeletogenesis. Anat Embryol 1992;186:107–124. Hart MJ, Eva A, Zangrill D, et al. Cellular transformation and guanine nucleotide exchange activity are catalyzed by a common domain on the Dbl oncogene product. J Biol Chem 1994;269:62–65. Jockusch BM, Bubeck P, Giehl K, et al. The molecular architecture of focal adhesions. Annu Rev Cell Dev Biol 1995;11:379–416. Kingsley DM. What do BMPs do in mammals? Clues from the mouse short-ear mutation. Trends Genet 1994;10:16–21. Lemmon, MA, Furguson KM, Schlessinger J. PH domains: diverse sequences with a common fold recruit signaling molecules to the cell surface. Cell 1996;85:621–624. Levitzki A. Targeting signal transduction for disease therapy. Curr Opin Cell Biol 1996;8:239–244.

CHAPTER 115 / AARSKOG-SCOTT SYNDROME

Nobes CD, Hall A. Rho, Rac, and Cdc42 GTPases regulate the assembly of multimolecular focal complexes associated with actin stress fibers, lamellipodia, and filopodia. Cell 1995;81:53–62. Olson MF, Pasteris NG, Gorski JL, Hall A. Faciogenital dysplasia protein (FGD1) and Vav, two related proteins required for normal embryonic development, are upstream regulators of Rho GTPases. Curr Biol 1996;6:1628–1633. Pasteris NG, Buckler JM, Cadle A, Gorski JL. Genomic organization of the faciogenital dysplasia (FGD1, Aarskog syndrome) gene. Genomics 1997;43:390–394. Pasteris NG, Cadle A, Logie LJ, et al. Isolation and characterization of the faciogenital dysplasia (Aarskog-Scott syndrome) gene: a putative rho/rac guanine nucleotide exchange factor. Cell 1994; 79:669–678.

1045

Pawson T. Protein modules and signalling networks. Nature 1995;373: 573–580. Porteous MEM, Goudie DR. Aarskog syndrome. J Med Genet 1991; 28:44–47. Schwartz MA, Schaller MD, Ginsberg MH. Integrins: emerging paradigms of signal transduction. Annu Rev Cell Dev Biol 1995;11:549–599. Stossel TP. On the crawling of animal cells. Science 1993;260:1086–1094. Treisman R. Regulation of transcription by MAP kinase cascades. Curr Opin Cell Biol 1996;8:205–215. Vojtek AB, Cooper JA. Rho family members: activators of MAP kinase cascades. Cell 1995;82:527–529. Zheng Y, Fischer DJ, Santos MF, et al. The faciogenital dysplasia gene product FGD1 functions as a Cdc42Hs-specific guanine-nucleotide exchange factor. J Biol Chem 1996;271:33,169–33,172.

CHAPTER 116 / BECKWITH-WIEDEMANN SYNDROME

116

1047

Beckwith-Wiedemann Syndrome ELLEN R. ELIAS, MICHAEL R. DEBAUN, AND ANDREW P. FEINBERG

BACKGROUND Beckwith-Wiedemann syndrome (BWS) is one of the most common overgrowth syndromes, with an estimated incidence of 1/14,000 births. Independently described by Beckwith in 1963 and Wiedemann in 1964, it was originally called the EMG syndrome for its three main features: exomphalos, macroglossia, and gigantism. Patients with BWS present with a distinctive craniofacial phenotype and may have multiple congenital anomalies. Macrosomia is present at birth and accelerated growth occurs during early childhood. In addition, asymmetry of the limbs may develop over time. Patients with BWS, particularly those with clinically significant limb asymmetry, have an increased risk of developing intraabdominal malignancies. BWS is caused by a gene in chromosomal band 11p15.5. Autosomal-dominant transmission has been documented, though most cases are sporadic. A greater likelihood of maternal vs paternal transmission and increased penetrance when inherited from the mother have been reported, but have not been evaluated systematically in a large number of families.

CLINICAL FEATURES BWS is a heterogeneous syndrome that can have many different presentations. All or a few of cardinal features of the syndrome can be present at birth. Thus, the diagnosis of BWS can be obvious or difficult in the neonatal period depending on the set of characteristic features observed. BWS can be suspected at birth because of macrosomia, macroglossia, and the presence of abdominal wall defects, which range in spectrum from diastasis recti or umbilical hernia in the milder cases to omphalocele in the most severe case. In addition to these cardinal features, visceromegaly or enlargement of the internal organs (liver, spleen, kidneys, pancreas, and occasionally the heart) may be identified. Adrenal cytomegaly with associated cysts, a rare pathological finding (Table 116-1), is a hallmark of BWS. The classic craniofacial features associated with BWS include facial nevus flammeus, maxillary hypoplasia with a broad nasal bridge, and a prominent occiput (Fig. 116-1). Characteristic ear lobe creases/pits are a distinctive if not pathomnemonic finding in children with other features of BWS. Microcephaly is seen in a small number of cases. Table 116-2 lists the common craniofacial features associated with BWS.

From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

In addition to dysmorphic features, macroglossia, and abdominal wall defects, other congenital anomalies may be present. These include cardiac defects and, occasionally, cardiomegaly and abnormalities of the gastrointestinal tract. Genital anomalies are common, including cryptorchidism and enlargement of the labia and clitoris. Many other anomalies, such as cleft palate, supernumerary nipples, and syndactyly have been reported with variable frequencies in patients with BWS and are listed in Table 116-3.

CLINICAL COURSE The clinical course during the neonatal period is most remarkable for hypoglycemia, often quite severe. This is felt to be caused by insulin hypersecretion from hypertrophied pancreatic islet cells. Polycythemia may also be problematic. Macrosomia is present at birth, and the growth velocity is greater than normal during the early years of life. The musculoskeletal system in children with BWS can be variably affected. Asymmetry of the musculoskeletal system may manifest as an increase in muscle bulk, bone length, or both. In addition, patients may have scoliosis or chest wall asymmetry independent of asymmetry of the long bones. Bone age is often advanced in early childhood but typically subsides by adolescence. After 4–6 years of age, growth velocity decreases, and puberty is usually achieved at the normal time. Asymmetry, seen in 13–18% of patients, may not be present at birth, but often manifests itself in infancy and early childhood. Patients with BWS generally have intellectual function in the normal range, although no study of formal cognitive evaluations has been completed. Central nervous system damage and subsequent intellectual impairment may result from severe neonatal hypoglycemia if not adequately managed. Patients with BWS have an increased risk of cancer. The risk is age-dependent, with most cancers occurring before 5 years of age and virtually all before 10 years. The risk of cancer appears to be higher in children with limb asymmetry, although the magnitude of asymmetry needed to confer a high risk of cancer has not been quantified. Common tumors in patients with BWS include Wilms’ tumor, adrenocortical carcinoma, neuroblastoma, and hepatoblastoma. Extraabdominal malignancies are much less common in patients with BWS, but have been reported, including rhabdomyosarcoma, glioblastoma, and cardiac fibrous harmartoma.

DIAGNOSIS The diagnosis of BWS is suspected on the basis of clinical phenotype. A constellation of abnormalities including macrosomia, macroglossia, abdominal wall defects, mid-face hypoplasia,

1047

1048

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Table 116-1 Pathological Findings in BWS

Table 116-2 Clinical Features in BWS

Adrenal cytomegaly and cysts Visceromegaly Nephromegaly Prominent lobulation Persistent nephrogenesis Medullary dysplasia/sponge kidneys Hepatomegaly Splenomegaly Hyperplasia of islets of Langerhans

Macroglossia Macrosomia Abdominal wall defects Diastasis recti Umbilical hernia Omphalocele Craniofacial dysmorphism Facial nevus flammeus Ear pit/creases Prominent occiput Maxillary hypoplasia Wide and flat nasal bridge High arched palate Frontal ridging Mild microcephaly

Table 116-3 Congenital Anomalies Associated with BWS

Figure 116-1 Beckwith-Wiedemann patient with maxillary hypoplasia, broad nasal bridge, telecanthus, and macroglossia.

and ear pits or creases should prompt the clinician to consider BWS. The diagnosis of BWS within the first days of life is important so that precautionary measures may be taken to prevent hypoglycemia. Hypoglycemia occurs in the majority of newborns within the first days of life and may manifest with seizures. This is in contrast to infants of diabetic mothers who may present with macrosomia but who often are hypoglycemic at birth. Infants of diabetic mothers may also have other congenital anomalies, but do not present with macroglossia, abdominal wall defects, and the typical ear creases seen in an infant with BWS. Macroglossia may also be observed in storage diseases such as Type II glycogen storage disease as well as in neurofibromatosis and hypothyroidism. Neurofibromatosis can be distinguished by its dermatological findings, and congenital hypothyroidism is usually detected on newborn screening. Most patients with storage diseases do not present with impressive macroglossia at birth. Other rare syndromes that should be considered in the differential diagnosis include Simpson-Golabi-Behmel syndrome, Proteus syndrome, and Perlman syndrome. Simpson-Golabi-Behmel syndrome is an X-linked disorder with a phenotype similar to BWS, including macrosomia, macroglossia, and a predisposition to Wilms’ tumor. Patients, in addition, may show cleft lip and/or palate, unusual alveolar grooves, and skeletal anomalies. An abnormality in glypican-3, which binds to IGF2, has recently been reported in Simpson-Golabi-Behmel patients. Proteus syndrome involves gigantism, macrocephaly, hemihypertrophy, enlargement of the limbs, and risk of malignancy, but has distinctive dermatologic manifestations. Perlman syndrome, involving prenatal over-

More common Cardiac defects/cardiomegaly Genitourinary anomalies Gastrointestinal anomalies Less common Supernumerary nipples Inguinal hernia Skeletal anomalies Cleft palate Diaghragmatic defects Chest wall deformities Clinodactyly/polydactyly Neural tube defect Imperforate anus

growth, renal hamartomas and a depressed nasal bridge, may be part of a continuum with BWS, although patients do not usually have macroglossia, midline abdominal defects, or hemihypertrophy. Other common overgrowth syndromes include Sotos syndrome (cerebral gigantism), Weaver syndrome, and Marshall-Smith syndrome. Sotos syndrome is characterized by early overgrowth and advanced bone age, hypotonia, and mild cognitive impairment in most patients. Facial features are notable for macrocephaly, downslanting palpebral fissures, and a prominent chin. Neoplasms of the kidney and liver have been reported in Sotos syndrome patients, although this is less frequent than in BWS. Weaver syndrome patients also show accelerated growth and advanced bone age. Characteristic facies include a broad forehead, large ears, and a small chin. Patients with Marshall-Smith syndrome show accelerated skeletal growth, but often manifest failure to thrive and severe, life-threatening respiratory problems. Characteristic features include prominent eyes, low-set ears, and choanal narrowing in some individuals. Patients with Sotos syndrome, Weaver syndrome, and Marshall-Smith syndrome can be distinguished from patients with BWS by the marked differences in their physical features and clinical course. Prenatal diagnosis of BWS is possible, made on the basis of second trimester ultrasound detection of large fetal size, polyhydramnios, and abdominal wall defects. A positive family history

CHAPTER 116 / BECKWITH-WIEDEMANN SYNDROME

Table 116-4 Clinical Issues and Management Neonatal Hypoglycemia Airway/feeding problems Surgical management of GU/GI anomalies CT and ultrasound for baseline assessment of intra-abdominal cancer Early childhood Audiology exam Musculoskeletal exam Assessment of asymmetry of the limb Cancer surveillance Ultrasound of kidney/liver/adrenal every 3 months from birth until 7 years of age α fetoprotein every 6–12 weeks of age until age 3a aThe efficacy for screening for hepatoblastoma with alphafetal protein has not been proven. However, anecdotal reports have indicated that this tumor biomarker may be of potential benefit.

is helpful in those cases of BWS acquired as a dominant trait. Chromosomal analysis in a few cases has revealed a translocation or duplication involving 11p15.5, but most patients with BWS have normal chromosomes.

MANAGEMENT/TREATMENT The initial management of the newborn with BWS (Table 116-4) must first address the neonatal hypoglycemia, which may be prolonged and quite severe. The macroglossia may cause airway and/ or feeding difficulties. Cardiac evaluation, including chest radiographs, electrocardiogram, and echocardiogram are indicated if congenital heart disease is suspected. Surgical intervention may be necessary for the abdominal wall defects and genitourinary abnormalities. Ultrasound of abdominal organs should be performed to rule out the presence of congenital tumors, which may require surgical excision and/or chemotherapy. Polycythemia is another common neonatal problem, which tends to subside without intervention. Beyond the neonatal period, the most important aspect of management is tumor surveillance. Cancer screening has traditionally consisted of serial abdominal sonography to detect Wilms’ tumor, the most common cancer in this group, from birth until 7 or 8 years of age. Other than anecdotal reports, there have been limited data to support cancer screening. Nevertheless, screening for Wilms’ tumor in this high-risk cancer population appears prudent. If one plans to do such screening, then screening intervals should be less than 6 months because of the doubling time of Wilms’ tumor. Despite the decreasing incidence of Wilms’ tumor with age, the doubling time of this cancer does not allow for longer screening intervals. There have been no published prospective studies demonstrating the benefit of screening for the other cancers, such as hepatoblastoma with α fetoprotein or neuroblastoma with urinary catecholamines. However, hepatoblastoma is a very aggressive tumor in which early diagnosis and complete surgical resection markedly improves survival. In contrast, factors predicting survival in children with neuroblastoma are more biological and demographically determined, such as N-myc copy number or age. Thus, preliminary data suggest screening for neuroblastoma is unlikely to be effective. Other aspects of management may include an early audiology exam to detect hearing loss, orthopedic intervention for leg length

1049

discrepancy, urologic management of complex genitourinary anomalies, and speech and language therapy for the consequences of macroglossia.

MODE OF INHERITANCE BWS is transmitted as an autosomal-dominant trait, with increased penetrance when transmitted from a carrier mother when compared to the father. Discordant monozygotic twins have also been described. These differences in parental origin-specific penetrance are most likely explained by genomic imprinting, a differential modification of the two parental chromosomes in the gamete or zygote, leading to differential function of the two chromosomes in the offspring. Imprinting is by definition reversible, and it usually refers to monoallelic expression of a specific parental allele of a gene. Several genes in mouse and man have been shown to be imprinted. Notably, the insulin-like growth factor-II gene (IGF-2) is expressed exclusively from the paternal allele in most tissues (exceptions are the choroid plexus, leptomeninges, and adult liver). IGF-II is an autocrine growth factor in cancer and has mitogenic effect on cells mediated by signaling at the IGF-I receptor. Approximately 100 kb telomeric to IGF-2 on 11p15 is the gene H19, which acts as an untranslated RNA and is reciprocally imprinted, that is, expressed only from the maternal allele. H19 has shown growth inhibitory effects on cultured tumor cells. IGF2 and H19 are interesting with regard to BWS because of their chromosomal location and potential growth properties. Based on parental origin-specific differential penetrance data alone, BWS could involve an imprinted gene normally expressed only from the paternal allele, which is somehow activated on the maternal allele in carrier mothers. Alternatively, BWS could represent a gene normally only expressed from the maternal allele, and thus a mutant copy is only phenotypically apparent when transmitted from the mother.

CHROMOSOMAL ALTERATIONS IN BWS BWS was localized to 11p15 by genetic linkage analysis. While BWS predisposes to Wilms’ tumor and other embryonal tumors, it is unlinked to the known Wilms’ tumor gene on 11p13 (WT1). Interestingly, a wide variety of embryonal neoplasms, to which BWS patients are susceptible, including rhabdomyosarcoma, hepatoblastoma, and Wilms’ tumor, show loss of allelic heterozygosity (LOH) of 11p15 in tumors. Thus, using polymorphic markers, one can demonstrate that one copy of 11p15 is lost in these tumors. Strikingly, almost all cases of LOH involve the maternal allele, again suggesting that an imprinted gene on 11p15 may be involved in the pathogenesis of BWS or cancer. Rare patients with BWS show cytogenetic abnormalities (Table 116-5) involving chromosome 11p. These include both balanced rearrangements and unbalanced duplications. The balanced germline chromosomal rearrangements (including translocations and inversions) lie near the end of 11p (generally 11p15.4–p15.5), and they involve the maternally inherited chromosome. In contrast, the unbalanced rearrangements are of paternal origin. The unbalanced rearrangement breakpoints are distributed throughout 11p, and what seems to be important is the extra chromosomal material, which always includes 11p15. Furthermore, carriers of the unbalanced rearrangement breakpoints are themselves unaffected, suggesting that the duplications, not the breakpoints themselves, are pathogenic in the chromosome duplication patients.

1050

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Table 116-5 Genetic Abnormalities in BWS Chromosomal rearrangements (~1%) Balanced translocations and inversions Involve 11p15 Maternal chromosome affected Unbalanced translocations Relative trisomy of 11p including 11p15 Paternal chromosome affected Translocation carriers unaffected Involves KvLQT1 gene Uniparental disomy (~10%) Region between β globin and H-ras on 11p15 Paternal duplication with maternal loss Occurs postzygotically (mosaic) Loss of imprinting (~20%) Activation of the normally silent maternal IGF-2 gene Mutation of p57KIP2 (~5%) Occurs in 70% of embryonal tumors, in both BWS and non-BWS patients

The balanced chromosomal rearrangements in BWS are consistent with either mechanistic hypothesis for an imprinted gene. Thus, the rearrangements could interrupt the coding sequence of a gene normally expressed only from the maternal chromosome, or they could epigenetically activate a gene near the breakpoints on the maternal chromosome that is normally silent on that chromosome. The paternal duplications, however, would suggest that the latter model is correct, that the BWS gene is normally only expressed from the paternal chromosome. Thus, one would be affected by BWS if one had two functional copies of the gene. Normal individuals with one paternal and one maternal chromosome would have only one functional copy. Patients with a paternal duplication would have two, and patients with a balanced rearrangement of the maternal allele would somehow activate their normally silent copy of the BWS gene.

GENOMIC IMPRINTING AND BWS GENES Strong molecular evidence for genomic imprinting in BWS derives from the observation that 10% of BWS patients have paternal uniparental disomy (UPD) of 11p15. Thus, their maternal chromosome has been replaced with a paternal chromosome over a region extending from the β-globin gene on 11p15 to the H-ras gene. UPD could be pathogenic either by causing loss of the normally expressed maternal gene or duplication of a normally expressed and imprinted paternal gene. UPD in the original reports was suggested to be associated with an increased risk of malignancy. However, other laboratories have not confirmed that finding, and it is unwise to quote a risk of malignancy based on the presence or absence of UPD. Prenatal diagnosis might be done using UPD as a marker, on single cells by PCR, as it has been performed on single circulating cells in BWS patients. The problem with that approach is that the sensitivity of the test is limited to the frequency of UPD, which is 10% or less. Furthermore, in contrast to UPD in Prader-Willi syndrome, UPD in BWS is mosaic in patients, with no more than 90% of peripheral blood lymphocytes or skin fibroblasts showing UPD. Thus, it is likely that UPD cases arise by mitotic recombination in early embryogenesis. One feature of the BWS germline rearrangements that is inconsistent with a single-gene hypothesis for this disorder is that there

are two distinct regions of BWS balanced rearrangement breakpoints, with one located approximately 250 kb centromeric to IGF-2, and a second cluster located 4 Mb centromeric to the first, on the opposite side of the β-globin gene cluster. Because it seems most unlikely that a single gene could span all of these breakpoints, it is conceivable that the breakpoints themselves could influence expression of a BWS gene at some distance, similar to mechanisms observed in both yeast and Drosophila. Furthermore, some human birth defects are caused by spreading inactivation from an X chromosome to a fused autosome in patients with X-autosome translocations. Genomic imprinting also appears to play an important role in the tumors to which BWS patients are susceptible. Thus, approximately 70% of Wilms’ tumors show loss of imprinting (LOI) affecting the IGF-2 and H19 genes. In tumors, the maternal allele of IGF-2 is abnormally activated, and the maternal allele of H19 is abnormally silenced. Other embryonal tumors also show LOI, including rhabdomyosarcoma and hepatoblastoma, and thus LOI appears to be a general feature of embryonal tumors. In addition, a “CpG island,” a region rich in CG dinucleotides, appears to distinguish maternal and paternal chromosomes, with the paternal chromosome methylated and maternal chromosome unmethylated. DNA methylation is a covalent modification of cytosine that tends to mark nonexpressed genes but in this case marks the paternal chromosome. In tumors with LOI, the maternal allele is abnormally methylated. Thus, the maternal chromosome appears to undergo an epigenetic switch from a maternal to a paternal epigenotype, with activation of IGF-2, loss of expression of H19, and methylation characteristic of the paternal chromosome. Thus, LOI is the functional equivalent in a tumor of UPD in embryogenesis. Normal tissues of BWS patients also show LOI of IGF-2; however, this affects only approximately 10% of patients. BWS patients with LOI also show the switch from an unmethylated to a methylated CpG island on the maternal chromosome, again suggesting a shift from a maternal to a paternal epigenotype. However, IGF-2 cannot be the only gene involved in BWS, because it shows abnormal imprinting in only a minority of patients. Mutations have been described in p57KIP2 in patients with BWS. p57KIP2 is an inhibitor of several G1 cyclin/Cdk complexes (see Chapter 116-6), and it is a negative regulator of cell proliferation. The p57KIP2 gene is imprinted and only the maternal allele is expressed. Mice lacking p57KIP2 develop some but not all abnormalities that resemble BWS. A third genetic change in BWS in rearrangement of the KVLQT1 gene, a voltage-gate potassium channel formerly associated only with the cardiac conduction disorder long QT syndrome. All of the balanced germline chromosomal rearrangements in this region involve this gene. Its role in BWS is supported by the discovery that KVLQT1 is also imprinted, with preferential expression from the maternal allele. However, imprinting is relaxed in some postnatal tissues, as well as prenatal and postnatal heart, accounting for the lack of parent of origin effect in long QT syndrome. A model representing how several different types of genetic alterations could lead to abnormal imprinting in BWS is shown in Fig. 116-2. This model incorporates what is known about LOI of target genes, alterations in DNA methylation, and chromosomal rearrangements in BWS. The model emphasizes the importance of regarding genomic imprinting as a developmental process that may be altered at several different stages in BWS and cancer.

CHAPTER 116 / BECKWITH-WIEDEMANN SYNDROME

Figure 116-2 A model of genomic imprinting in BeckwithWiedemann syndrome. An imprint organizing center (red rectangle) exerts a long-range cis-acting influence on IGF-2 and other imprinted genes (blue ellipses) via alterations in chromatin structure (represented as DNA loops). This imprint-organizing center establishes the imprinting mark as maternal or paternal. This effect is propagated outward during development, similar to the organizing center on the X-chromosome. Imprinting is maintained in part by allele-specific methylation of CpG islands as well as interactions with trans-acting proteins (green circles). According to this model, loss of imprinting could arise by any of several mechanisms (numbered in the figure): (1) deletion or mutation in the imprint-organizing center itself, which would lead to a failure of parental origin-specific switching in the germline; (2) separation of the imprint organizing center from the imprinted target genes, as seen in BWS germline chromosomal rearrangement cases; (3) abnormal methylation of CpG islands; (4) local mutation of regulatory sequences controlling the target imprinted genes themselves; or (5) loss of or mutations in trans-acting proteins that maintain normal imprinting. (Figure modified from Feinberg, et al. 1994, with permission.)

DIAGNOSTIC TESTING Karyotypic abnormalities are seen in fewer than 1% of BWS patients and generally involve the KVLQT1 gene. LOI of IGF2 is observed in 20–60% of patients, depending on the report, although we believe the more conservative figure is correct. UPD occurs in about 10% of patients. p57KIP2 mutations occur in only 5% of patients. Thus, there is no single diagnostic test for BWS, and the sensitivity of any given test is low. At this point, the most useful test for counseling purposes is UPD. That is because, although the frequency of UPD is low, it would indicate a zero recurrence risk when present, because UPD occurs postzygotically. In contrast, diagnosis of an imprinting mutation (LOI), or mutation p57KIP2 or KVLQT1 would warrant investigation of other family members for a similar alteration. It is important to emphasize that diagnostic testing for BWS is at the research stage, and that identification of additional target genes should increase the sensitivity and specificity of such testing.

CONCLUSIONS In summary, BWS involves several 11p15 genes that are imprinted. One of these, IGF2, is normally expresses from the paternal allele and is abnormally activated in patients with LOI. A second gene, p57KIP2, is expressed from the maternal allele patients and is mutated in a small fraction of patients. A third gene, KVLQT1, is expressed from the maternal allele and rearranged in most BWS patients with balanced germline chromosomal rear-

1051

rangements. The phenotype may be explained in part by the variable involvement of these specific genes, as well as additional as yet unidentified imprinted genes. Of all BWS patients, 1% or fewer show rearrangements of KVLQT1, approximately 10% show uniparental disomy of 11p15, at least 20% involve LOI of IGF2, but most are sporadically occurring and without a defines genetic abnormality. Currently, the most useful genetic test in evaluating the family of a BWS patient, as with most genetic disorders, is a good family history. It is particularly important to examine infant photographs and to obtain a neonatal history on first-degree relatives of the affected patient. If a family appears to show dominantly transmitted BWS, then the risk to future offspring is 50%. An interesting and provocative aspect of BWS is that the phenotype appears to abate with age, as does the risk of malignancy. Given that epigenetic factors may play a role in controlling genomic imprinting, it is possible that those factors could be modified epigenetically using novel therapeutic approaches. For example, drugs such as 5-aza-2'-deoxycytidine, which cause hypomethylation of DNA, might eventually present a novel form of gene therapy for disorders of abnormal imprinting such as BWS. Such ideas are at the earliest in vitro experimental stages, but one cannot help but speculate that, because BWS involves, at least in part, epigenetic changes to the chromosome, these changes might be experimentally or therapeutically reversible.

ACKNOWLEDGMENTS This work was supported by NIH Grant CA 65145 (APF).

SELECTED REFERENCES Aleck KA, Hadro TA. Dominant inheritance of Wiedemann-Beckwith syndrome: further evidence for transmission of “unstable premutation” through carrier women. Am J Med Genet 1989;33:155–160. Beckwith JB. Extreme cytomegaly of the adrenal fetal cortex, omphalocele, hyperplasia of the kidneys and pancreas and Leydig-cell hyperplasia; another syndrome? Presented at the Annual Meeting of the Western Society of Pediatric Research, Los Angeles, Nov 1963. Bischoff FZ, Feldman GL, McCaskill C, Subramian S, Hughes MR, Shaffer LG. Single cell analysis demonstrating somatic mosaicism involving 11p in a patient with paternal isodisomy and BeckwithWiedemann syndrome. Hum Mol Genet 1995;4:395–399. Feinberg AP, Kalikin LM, Johnson LA, Thompson JS. Loss of imprinting in human cancer. Cold Spring Harbor Symp Quant Biol 1994; 59:357–364. Hao Y, Crenshaw T, Moulton T, Newcomb E, Tycko B. Tumor-suppressor activity of H19 RNA. Nature 1993;365:764–767. Hatada I, Ohashi H, Fukushima Y, et al. An imprinted gene p57KIP2 is mutated in Beckwith-Wiedemann syndrome. Nat Genet 1996;14:171–173. Henry I, Bonaitijk-Pellie C, Chehensse V, et al. Uniparental paternal disomy in a genetic cancer-predisposing syndrome. Nature 1991;351: 665–667. Hoovers JMN, Kalikin LM, Johnson LA, et al. Multiple genetic loci within 11p15 defined by Beckwith-Wiedemann syndrome rearrangement breakpoints and subchromosomal transferable fragments. Proc Natl Acad Sci USA 1995;92:12,456–12,460. Koufos A, Grundy P, Morgan K, et al. Familial Wiedemann-Beckwith syndrome and a second Wilms tumor locus both map to 11p15.5. Am J Hum Genet 1989;44:711–719. Lee MP, Hu RJ, Johnson LA, Feinberg AP. Human KVLQT1 gene shows tissue-specific imprinting and encompasses Beckwith-Wiedemann syndrome chromosomal rearrangements. Nat Genet 1997;15:181–185. Lee MP, DeBaun M. Randhawa G, Reichard BA, Elledge SJ, Feinberg AP. Low frequency of p57KIP2 mutations in Beckwith-Wiedmann syndrome. Am J Hum Gen 1997; 61:304–309. Litz CE, Taylor KA, Qiu JS, Pescovitz OH, de Martinville B. Absence of detectable chromosomal and molecular abnormalities in monozygotic twins discordant for the Wiedemann-Beckwith syndrome. Am J Med Genet 1988;30:821–833.

1052

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Lubinsky M, Herrmann J, Kosseff AL, Opitz JM. Autosomal-dominant sex-dependent transmission of the Wiedemann-Beckwith syndrome. Lancet 1974;1:932. Mannens M, Hoovers JMN, Redeker E, et al. Parental imprinting of human chromosome region 11p15.3-pter involved in the BeckwithWiedemann Syndrome and various human neoplasia. Eur J Hum Genet 1994;2:3–23. Moulton T, Crenshaw T, Hao Y, et al. Epigenetic lesions at the H19 locus in Wilms tumour patients. Nat Genet 1994;7:440–447. Ogawa O, Eccles MR, Szeto J, et al. Relaxation of insulin-like growth factor-II gene imprinting implicated in Wilms tumour. Nature 1993;362:749–751. Pettenati MJ, Haines JL, Higgins RR, Wappner RS, Palmer CG, Weaver DD. Wiedemann-Beckwith syndrome: presentation of clinical and cytogenetic data on 22 new cases and review of the literature. Hum Genet 1986;74:143–154. Pilia G, Hughes-Benzie RM, MacKenzie, et al. Mutations in GPC3, a glypican gene, cause the Simpson-Golabi-Behmel overgrowth syndrome. Nat Genet 1996;12:241–247. Ping AJ, Reeve AE, Law DJ, Young MR, Boehnke M, Feinberg AP. Genetic linkage of Beckwith-Wiedemann syndrome to 11p15. Am J Hum Genet 1989;44:720–723. Rainier S, Dobry C, Feinberg AP. Loss of imprinting in hepatoblastoma. Cancer Res 1995;55:1836–1838. Rainier S, Johnson LA, Dobry CJ, Ping AJ, Grundy PE, Feinberg AP. Relaxation of imprinted genes in human cancer. Nature 1993; 362:747–749. Reik W, Maher ER. Imprinting in clusters: lessons from BeckwithWiedemann syndrome. Trends Gent 1997;13:330–334.

Schroeder WT, Chao L-Y, Dao DD, et al. Nonrandom loss of maternal chromosome 11 alleles in Wilms tumors. Am J Hum Genet 1987;40:413–420. Shackney SE, McCormack GW, Cuchural GJ. Growth rate patterns of solid tumors and their relation to responsiveness to therapy. Ann Int Med 1978;89:107–121. Sippell WJ, Partsch CJ, Wiedemann HR. Growth, bone maturation and pubertal development in children with the EMG syndrome. Clin Genet 1989;35:20–28. Steenman MJC, Rainier S, Dobry CJ, Grundy P, Horon I, Feinberg AP. Loss of imprinting of IGF2 is linked to reduced expression and abnormal methylation of H19 in Wilms’ tumor. Nat Genet 1994;7:433–439. Treuner J, Schilling FH. Neuroblastoma mass screening: the arguments for and against. Euro J Cancer 1995;31A:565–568. Weksberg R, Shen DR, Fei YL, Song QL, Squire J. Disruption of insulinlike growth factor 2 imprinting in Beckwith-Weidemann syndrome. Nat Genet 1993;5:143–150. Wiedemann HR. Complex malformatif familial avec hernia, ombilicale et macroglossie—un syndrome nouveau? J Genet Hum 1964;13:223–232. Wiedemann HR. Tumors and hemihypertrophy associated with Wiedemann Beckwith syndrome (Letter). Euro J Ped 1983;141:129. Yan Y, Frisen J, Lee MH, Massague J, Barbacid M. Ablation of the CDK inhibitor p57Kip2 results in increased apoptosis and delayed differentiation during mouse development. Genes Dev 1997;11:973–983. Zhan SL, Shapiro DN, Helman LJ. Activation of an imprinted allele of the insulin-like growth factor-II gene implicated in rhabdomyosarcoma. J Clin Invest 1994;94:445–448. Zhang P, Liegeois NJ, Wong C, et al. Altered cell differentiation and proliferation in mice lacking p57KIP2 indicates a role in BeckwithWiedemann syndrome. Nature 1997;387:151–158.

CHAPTER 117 / PRADER-WILLI AND ANGELMAN SYNDROMES

117

1053

Prader-Willi and Angelman Syndromes ROBERT D. NICHOLLS

BACKGROUND Congenital disease in a patient usually involves a mutation in a gene that obeys established rules of Mendelian inheritance. This chapter describes two genetically related syndromes that do not conform to these principles and thereby give rise to totally independent clinical disorders. Prader-Willi syndrome (PWS) was first described by three German physicians, Prader, Labhart, and Willi, in 1956, and represents the most common form of syndromic genetic obesity. PWS was originally called the HHHO syndrome for its main features of hyperphagia, hypotonia, hypogonadism, and obesity. Angelman syndrome (AS) was first described by the British pediatrician, Harry Angelman in 1965, in which he coined the term “Happy Puppet syndrome,” as it described two of the cardinal features of AS, a happy disposition with unprovoked laughter and severe ataxic movements. PWS and AS each occur at a frequency of about 1/10,000 to 1/20,000 births. PWS and AS are each caused by genetic abnormalities in chromosome 15q11–q13, which involve imprinted genes. Imprinting refers to a class of non-Mendelian inheritance, in which specific genes or chromosomal regions are “marked” differently in the male and female germline. Following fertilization, this allows a somatic cell to know which parent the “marked” gene came from, and leads to parent-of-origin specific gene expression during development. We do not yet know why imprinting occurs, though much is being learned about the complexities and role of imprinting in AS and PWS, which has important implications for not only diagnosis of these two syndromes, but also as a means of potential therapeutic intervention in the future.

CLINICAL FEATURES AND NATURAL HISTORY OF PRADER-WILLI SYNDROME DIAGNOSTIC CRITERIA The clinical phenotype of PWS is characterized by neonatal hypotonia and developmental delay, followed by postnatal onset of hyperphagia with subsequent obesity and other features, including short stature, hypogonadism, and mild-to-moderate mental retardation (Fig. 117-1A). These features have been organized into a set of clinical diagnostic scoring criteria for infancy and early childhood or childhood-toadult periods. However, since most of the clinical features are relatively nonspecific, and non-PWS patients with other disorders can meet the score required to ascertain PWS, these criteria From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

are meant to suggest a possible diagnosis of PWS and the need for further molecular tests (see below), not that a given patient has the syndrome. NEONATAL HYPOTONIA AND FAILURE TO THRIVE Decreased fetal activity during the perinatal period has been associated with PWS. This is followed by severe hypotonia in the neonatal period, with feeding difficulties requiring gavage or other feeding methods. There is failure to thrive and developmental delay in the neonatal/infancy period. Physical therapy is important as a therapeutic approach to poor muscle tone and may help prevent scoliosis and improve ability to exercise in later life. HYPERPHAGIA AND OBESITY The onset of hyperphagia in PWS is between 1 and 6 years of age, with consequent severe obesity. Obesity is perhaps the most significant clinical feature of PWS, since it is life-threatening by the third decade without dietary and behavioral intervention. Hyperphagia in PWS appears to be driven by a failure to reach satiety, thought to be hypothalamic in origin. This results in a high energy intake, which coupled with low activity levels, short stature, and a decreased caloric requirement (for fat-free mass), results in severe obesity that has potentially morbid consequences without intervention, most commonly due to cardiopulmonary failure. However, with dietary and behavior modification, including incorporation of an exercise program, significant weight losses can be achieved, with medical, psychological, and social benefits to the patient. The latter occur because severe obesity in PWS is associated with poor self-image and social acceptance. Early diagnosis and intervention is particularly critical for these aspects. Hyperphagia is associated with food-seeking behavior, obsession with food, and behavioral problems when food is restricted (see below). PWS patients also have a thick, viscous saliva and inability to vomit. There is a specific pattern of fat distribution which is predominantly truncal and involves the proximal limbs. PWS individuals tend to have 30–40% body fat even when near ideal weight for height. The role of increased adipose tissue lipoprotein lipase is unknown. Diabetes mellitus is present in 10–20% of patients. HYPOGONADISM PWS patients have hypogonadism, secondary to hypogonadotropism. Hypogonadism is characterized by genital hypoplasia, delayed or incomplete gonadal maturation, delayed puberty, and infertility in males and females. SHORT STATURE AND SMALL HANDS AND FEET PWS patients have short stature, a consequence of secondary growth hormone (GH) deficiency. The basis for GH deficiency is not known. Therapeutic application of GH has been shown to not only

1053

1054

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Figure 117-1 Clinical phenotype of children with (A) Prader-Willi syndrome or with (B) Angelman syndrome. Note the central obesity, short stature, small hands and feet, and almond-shaped eyes/narrow bifrontal diameter in the PWS child, and the happy disposition, wide-spaced mouth, and teeth, and broad stance of the AS child. (Reprinted with permission from Butler MG. Prader-Willi syndrome: current understanding of cause and diagnosis. Am J Med Genet 1990;35:319–332 [A]; and Williams CA, Zori RT, Stone JW, et al. Maternal origin of 15q11–13 deletions in Angelman syndrome suggests a role for genomic imprinting. Am J Med Genet 1990;35:350-353 [B]).

increase height, but also to increase the muscle mass and decrease fat mass. This results in improved muscle tone, with consequent less problems from effects of hypotonia, and decreased obesity with improvements in behavior, self-image, and social acceptance; these may lead to an increase in measurable IQ. PWS patients have small hands and feet, compared to height, particularly evident after midchildhood, suggesting delay in distal limb maturation, but the basis for this is not known. MENTAL RETARDATION AND LEARNING AND BEHAVIORAL DISORDERS PWS patients have mild-to-moderate mental retardation (IQ in the 20–100 range, usually 40–80), but they also have learning disabilities, particularly of speech, and reading disabilities. Interestingly, PWS patients also fulfill criteria for obsessive–compulsive disorder. There are many behavioral and emotional problems, such as temper tantrums, stubbornness, rage, violence, and stealing. Many of these abnormal behaviors may be associated with hunger and food-seeking behavior. PWS patients also have other unique behavioral features, such as skill with jigsaw puzzles.

CRANIOFACIAL There is mild craniofacial dysmorphism in most PWS patients, including a narrow bifrontal diameter and almond-shaped palpebral fissures, which may be upslanting, and down-turned mouth, with thin upper lip. At least the down-turned mouth may be a consequence of hypotonia, but it is not clear if the craniofacial features represent a primary embryonic defect or secondary consequence of abnormal brain development. OTHER FEATURES The hypopigmentation of the hair, skin, and eyes seen in one-half to two-thirds of PWS patients is associated with presence of a cytogenetic deletion and almost certainly results from deletion of the P gene. However, since mutations in the P gene cause the autosomal-recessive condition, oculocutaneous albinism type 2 (OCA2), in humans and mice, it is unclear why deletions lead to hypopigmentation in PWS (and AS) patients. PWS patients also often display skin picking, perhaps indicating a decreased pain sensitivity, and abnormal temperature control. The somnolence often seen in PWS patients appears not to be a direct consequence of the obesity in the syndrome. Therefore, it is possible that there is a specific sleep (and

CHAPTER 117 / PRADER-WILLI AND ANGELMAN SYNDROMES

possibly circadian rhythm) disturbance in PWS, but this requires further research. DIFFERENTIAL DIAGNOSIS FOR PWS Many of the syndromes that require differential diagnosis with PWS also display obesity, endocrine abnormalities (e.g., growth, hypogonadism, diabetes), and mental retardation. These include Bardet-Biedl syndrome, Borjeson-Forssman-Lehmann syndrome, Cohen syndrome, Albright hereditary osteodystrophy (including pseudohypoparathyroidism [PHP1a] and pseudo-PHP [PPHP]), Smith-Magenis syndrome, Fragile X syndrome, and other rare conditions. Congenital hypotonia in infancy can have many causes, and molecular diagnostics may be important here for PWS (see below). “Acquired” PWS-like conditions also exist, arising from tumors or other insults to the brain, predominantly in the region of the hypothalamus.

CLINICAL FEATURES AND NATURAL HISTORY OF ANGELMAN SYNDROME CLINICAL DIAGNOSTIC CRITERIA AS is associated with gait ataxia (jerky, unsteady, stiff with upheld arms), tremulousness, seizures typically beginning after 1–6 years of age, hyperactivity, severe mental retardation, absence of verbal speech, microbrachycephaly, protruding tongue and drooling, widespaced mouth and teeth, and a happy disposition with inappropriate laughter (Fig. 117-1B). AS patients display abnormal electroencephalograms (EEGs), with a characteristic pattern of large amplitude slow-spike wave activity (usually 2–3 Hz), which is facilitated by eye closure. The EEG abnormalities are a useful diagnostic guide for AS, although after the second decade of life, the EEG can appear normal. Mild cerebral atrophy has been seen on CT scan in about 30% of patients. NEONATAL/INFANCY FEATURES AS children appear normal at birth, but soon suffer severe developmental delay. From early childhood, AS patients have developmental delay, with no development of speech, a smiling, happy disposition, ataxia and limb tremulousness, and microbrachycephaly. Seizures usually begin by 3–6 years of age. MOVEMENT DISORDER Gross motor milestones are delayed, although hyperkinetic movements of the trunk and limbs may be seen in early infancy. Tremulousness may begin as early as 6 months. Most AS children are ambulatory, but there is a prancing ataxic gait with tendency to lean forward, pronounced in running, with upheld arms flexed at the elbows and wide-spaced legs. More severely affected children are very stiff, shaky, and jerky when walking, and some individuals remain nonambulatory. SEIZURES Seizures typically begin after 1 year of age but usually before 3 years. They can be severe, episodic, of all types, but mostly major motor type or minor type, requiring anticonvulsant medication. However, no one medication has proven effective for seizures in AS and usually depends on the minor or major motor type. Seizures can be confused with the tremulousness and jerky movements seen in AS. While the abnormal EEG may be associated with seizure behavior, an abnormal EEG can persist even when seizures are controlled. Seizures may decrease in frequency or cease by the teenage years. INAPPROPRIATE LAUGHTER An apparent happy disposition typifies the AS child. In addition, most reactions to mental or physical stimuli are accompanied by laughter or similar facial expressions. This behavior has an onset early in childhood. It does not appear to be related to the seizure disorder. Sometimes irritability and hyperactivity can also be present.

1055

MENTAL RETARDATION AND SPEECH Developmental milestones are greatly delayed, although formal testing is difficult because of absence of formal speech, hyperactivity, motor difficulties, and attention deficit. Nevertheless, IQ scores are in the severely retarded range, although AS children perform better on receptive social skills, and it is believed that cognitive abilities are higher than thought from developmental testing. Most AS children have one to four words, at most, although a few rare cases with milder presentation (see below) have up to 10–20 words. Nevertheless, most AS children are able to follow and understand commands, except for cases with severe seizures and hyperactivity. In higher-functioning individuals, there may be nonverbal language, such as some degree of sign language. CRANIOFACIAL The microcephaly present in some AS patients may be a secondary result of developmental brain abnormalities, associated with decelerated growth by 1 year of age. Many of these patients also have a flattened occiput, some with a distinct occipital groove. Brachycephaly is a common finding in AS. The mandible is normal in size but can appear large because of a forward and upward orientation, with midface hypoplasia, a thin upper lip, and deep-set eyes. A wide mouth and wide-spaced teeth are often present, and prognathia can result from excessive chewing. Facial dysmorphism is not present at birth. OTHER FEATURES Virtually all AS individuals display features of hyperactivity, both constant movement (including running), and handling and mouthing of objects in infancy. This may improve in adults. There is attention deficit, which in severe cases prevents facial and social cues, but in others allows sufficient attention for developmental training programs. Sleep abnormalities consisting of short periods of sleep and abnormal sleep/wake cycles are characteristic of AS. The role of abnormal circadian rhythms and other features of sleep and hyperactivity remain to be assessed. The hypopigmentation of the hair, skin, and eyes seen in one-half to two-thirds of AS patients likely results from deletion of the P gene, as explained above for PWS. It is present only in PWS and AS patients with deletions of 15q11–q13. AS patients also share intriguing behavioral features, such as love of water, of bright, shiny objects, and of music. About 5% of AS patients show obesity. DIFFERENTIAL DIAGNOSIS FOR AS Conditions that need to be considered in the differential diagnosis of AS include Rett syndrome, a neurodegenerative syndrome affecting girls only, distinguished by the history of regression and loss of acquired skills, the alpha-thalassemia-mental retardation syndrome, X-linked (ATRX syndrome), ataxic cerebral palsy with less severe mental retardation, and others.

CLINICAL MANAGEMENT OF THE SYNDROMES For PWS, feeding and behavior management are critical, particularly from an early age. Appetite suppressant agents have not generally been successful in PWS. Growth hormone replacement for height, improved muscle-to-fat mass, and associated behaviors is beneficial during childhood, but is still controversial. Learning disabilities require evaluation and may require speech therapy. Early intervention with physical therapy can significantly help difficulties associated with poor muscle tone. While hormonal treatment can be applied to hypogonadism, the benefits are not clear. For AS, the principle management issues relate to seizure control (in which no one anticonvulsant has been successful) and to hyperactivity, sleep disturbances, and education. It would also be

1056

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

3. A distal region containing genes that are expressed from both alleles and thus show normal Mendelian inheritance, (which includes a cluster of gamma-aminobutyric acid (GABA) class A receptor genes and the P gene involved in the autosomal-recessive disorder OCA2).

Figure 117-2 Molecular classes of Prader-Willi and Angelman syndromes. The chromosome 15 genotypes (and frequency) are shown for the major classes of AS and PWS. UPD, uniparental disomy; P, paternal (black); M, maternal (gray); M(P), maternal inheritance with paternal imprint (or epigenotype); P(M), paternal inheritance with maternal epigenotype; X, structural gene mutation. (See text for details.)

beneficial to treat the ataxia and tremulousness of the limbs in AS, since this may improve the ability to learn sign language, or, potentially, computer skills, both very difficult for AS individuals. Although the biological basis of AS and PWS are unknown, both syndromes are likely to arise predominantly from failure of an aspect of neural development during embryogenesis (particularly the hypothalamus or associated structure for PWS and the cerebellum or other limbic structure in AS). However, some features of both syndromes clearly have postnatal onset, including in AS of seizures (and their resolution) and coarsening of the facial appearance during the first two decades of life, and in PWS of hyperphagia (hypotonia and poor feeding ability may prevent this in infancy and early childhood), behavioral disturbances, and small hand and foot size. No consistent abnormality has been identified in PWS or AS by biochemical, endocrine, or anatomical studies.

GENETIC BASIS OF PRADER-WILLI AND ANGELMAN SYNDROMES CHROMOSOME 15 There are multiple molecular genetic mechanisms that can lead to PWS and AS (Fig. 117-2), but despite this complexity, each leads to a common gene deficit. Each molecular mechanism that leads to PWS and AS abolishes imprinted (parent-of-origin)-specific expression, such that paternal gene expression is abrogated in PWS (one or more genes) and maternal gene expression is silenced in AS (probably one gene). Chromosome 15q11–q13 is now known to have three subregions (Fig. 117-3): 1. A proximal region containing genes that are expressed from the paternally inherited chromosome only (hence each of these genes is a candidate to play a role in PWS). 2. A central region containing a gene, or genes, expressed from the maternally inherited chromosome only (and thus this gene is a candidate for AS).

CYTOGENETIC DELETIONS Cytogenetically visible deletions of 15q11–q13 were first observed in PWS by Ledbetter and colleagues in 1981 and in AS by Latt and Magenis and their colleagues in 1987. Although initially thought that the deletions in PWS and AS would be adjacent, it was first shown in 1989 that the majority of the deletions in PWS and AS are in fact of a very similar size. This conclusion was supported by the cloning of a yeast artificial chromosome (YAC) contig spanning 15q11–q13, which showed that about 4 Mb of DNA was deleted in these PWS and AS patients. It is likely that specific sequences at the proximal and distal breakpoints predispose to recombination, to generate the common deletion. Although the deletion size is generally the same in PWS and AS, from the study of differently sized deletions in rare patients, it was subsequently realized that the PWS and AS genes do in fact map to adjacent chromosome regions (Fig. 117-3). PARENTAL ORIGIN OF DELETIONS Since the majority of deletions in PWS and AS are the same size, it was initially an enigma as to how two different clinical syndromes arise. However, the use of cytogenetic and molecular polymorphisms led to the findings that all deletions in PWS are paternal in origin, and all deletions in AS are maternal in origin (Fig. 117-2). Therefore, the parental origin determines the clinical nature of the syndrome, and both parental contributions of chromosome 15q11–q13 are required for normal development. UNIPARENTAL DISOMY (UPD) UPD was first discovered in PWS in 1989, in which studies with molecular polymorphisms showed that two PWS children had inherited two maternal alleles but no paternal allele for chromosome 15 (Fig. 117-2). Subsequent studies showed that maternal UPD is quite common in PWS, occurring in 25% of cases. There is an association of advanced maternal age with UPD in PWS, consistent with maternal nondisjunction and subsequent formation of a trisomic zygote. Since trisomy 15 is lethal during embryogenesis, there is selection for loss of one chromosome 15, which one-third of the time would lead to maternal UPD and hence PWS. UPD also occurs in AS, in which it is paternal in origin (Fig. 117-2) and may be associated with a milder phenotype, but a larger study population is needed to confirm this. The finding of UPD confirms that the parental origin of 15q11–q13, and hence genomic imprinting, underlies the etiology of PWS and AS. Thus, a second totally normal copy of a maternal chromosome 15 cannot complement the missing paternal chromosome in PWS, indicating that the maternally inherited PWS genes are normally silent, and that the paternally inherited PWS genes are expressed. Likewise, paternal UPD in AS indicates that the paternally inherited AS gene is normally silent, and that it is the maternally inherited allele that is active. Therefore, the absence of expression of the active PWS or AS genes leads to the respective syndrome (by any mechanism in Fig. 117-2). PATERNAL PWS AND MATERNAL AS GENES In AS, there are about 20% of patients who do not have a deletion, UPD, nor imprinting mutation (see below). Some of these cases are familial. These patients inherit a copy of chromosome 15 from each parent (biparental) and likely have a point mutation (or small DNA rear-

CHAPTER 117 / PRADER-WILLI AND ANGELMAN SYNDROMES

1057

Figure 117-3 Human chromosome 15q11–q13 map. Genetic loci are shown as circles and FISH probes as filled circles. Regions containing the AS and PWS structural genes are indicated above the line, and the region affected by microdeletions in AS and PWS imprinting mutation patients in the expanded segment below (the latter define the imprinting center, IC). Two genes are expressed from the IC, the SNRPN gene and IC transcript. The shortest region of microdeletion overlap in AS and PWS is shown by bars; these correspond to elements critical for paternal (pat) to maternal (mat) imprint switching in the female germline or maternal to paternal imprint switching in the male germline, respectively. At the top, the presence (+) and absence (–) of imprinted features are shown. The zigzag lines reflect the end points of the common cytogenetic deletion in AS and PWS. P, paternal-only expression; M, maternal-only expression in brain; cen, centromere; tel, telomere.

rangement) in a single, maternally expressed gene (Fig. 117-2). Recently, the UBE3A gene (formerly E6AP), encoding a ubiquitinprotein ligase thought to be involved in protein degradation pathways, was shown to have mutations in AS patients of this class. Recent studies have shown that the UBE3A gene undergoes spatially (tissue-specific) regulated imprinted expression in human and mouse. In contrast, no PWS patients with the classical phenotype appear to come under this class (Fig. 117-2). This may mean that at least two genes are required for classical PWS (all cases of PWS would then represent a contiguous gene syndrome); mutations in a single gene may therefore only result in a PWS-like syndrome that is presently unrecognized. Consistent with this idea, there are multiple paternally expressed genes in the 15q11–q13 region involved in PWS (Fig. 117-3). Nevertheless, a single major gene has not yet been ruled out. FAMILIAL CASES RESULTING FROM IMPRINTING MUTATIONS About 5% of AS and PWS patients have been shown to have inherited a copy of chromosome 15 from both the mother and the father, but they have abnormal DNA methylation imprints (see below) throughout the imprinted 15q11–q13 region (Fig. 117-2). However, the DNA methylation is typical of the syndrome (uniparental maternal in PWS and uniparental paternal in AS). This suggests that such patients have a mutation in the imprinting process, since the paternally inherited chromosome in PWS patients has maternal DNA methylation, and the maternally inherited chromosome in AS patients has paternal DNA methylation. These AS and PWS patients have the classical clinical phenotype, and about half the cases are familial. The molecular basis of this class of patients is described below. BALANCED TRANSLOCATIONS IN PWS Three PWS patients have been described with a balanced translocation of 15q11–q13. One patient with classical PWS has a breakpoint in the 5' part of the SNRPN gene, implicating this gene as a major PWS gene. Two patients with a PWS-like syndrome (not fulfilling all major clinical criteria of PWS) are associated with a second cluster of breakpoints about 50–150 kb distal of the SNRPN gene, and thus other genes may contribute to the PWS phenotype. However, how the translocations lead to the phenotype in this rare class of patients is presently not understood.

MOLECULAR DIAGNOSTICS OF PRADER-WILLI AND ANGELMAN SYNDROMES FLUORESCENT IN SITU HYBRIDIZATION (FISH) FISH, with unique probes from 15q11–q13 (Fig. 117-3), and a centromeric control probe, is the method of choice of cytogenetic diagnostic laboratories to identify large deletions or unbalanced translocations in suspected PWS and AS cases. Probes within at least the PWS and AS critical regions (Fig. 117-3) should be utilized. MICROSATELLITE ANALYSIS FOR UPD Polymorphic marker analysis, preferably the use of highly informative microsatellites such as (CA)n markers, using parental and affected-child DNA samples, is commonly used to identify UPD. However, both parental DNA samples are often not available. DNA METHYLATION ANALYSIS Several genes and DNA markers in 15q11–q13 (ZNF127, NDN, PW71, SNRPN [Fig. 1173]) have been shown to display differential DNA methylation of the paternal and maternal alleles. These differences arise in the male and female germline or are established in early development. The most reliable methylation probe is at SNRPN, in which the maternal allele is completely methylated, and the paternal allele is completely unmethylated (Fig. 117-4), in all tested tissues. Clearly, AS samples with only the unmethylated band and PWS samples with only the methylated band can be easily distinguished from one another and from normal controls (or patients with another diagnosis). Methylation at SNRPN represents the best single diagnostic tool for these syndromes, since it identifies all PWS and AS patients with a deletion, unbalanced translocation, UPD, or imprinting mutation (it does not detect the biparental class of AS patients), and does not require parental DNA samples. However, DNA methylation is not reliable to distinguish between different classes of PWS and AS patients. Neverthless, DNA methylation detects about 98% of classical PWS patients (deletion, UPD, and imprinting mutations) and 80% of AS patients (deletion, UPD, and imprinting mutations). PWS AND AS IMPRINTING MUTATIONS Imprinting mutation cases are diagnosed by the absence of a deletion by FISH (with the exception of a SNRPN deletion only) and the absence of UPD (tested by polymorphic markers), but with the presence of abnormal DNA methylation using any of the probes that detect differential DNA methylation. About one-half of cases, including

1058

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

IMPRINTING MUTATIONS IN PWS AND AS In those families in which a microdeletion in the IC has been identified, or in which multiple sibs of PWS or AS have already occurred, “at risk” or prenatal diagnosis are options, based on detection of the microdeletion. Since imprinting mutations can be transmitted silently for multiple generations (see below), relatives of the same sex as the transmitting parent have up to a 50% risk for transmitting the mutation silently, whereas offspring of the opposite sex have a risk of up to 50% for a child with the syndrome. In such cases, referral for molecular testing and genetic counseling are strongly recommended (see Saitoh et al., 1997, for further discussion of recurrence risks). In cases with no detectable microdeletion or IC mutation, no known recurrence has occurred, and such cases may be sporadic (unpublished data). OTHER FAMILIAL AS In families with multiple affected AS cases, once an imprinting mutation has been ruled out, it is likely that there is an inherited mutation in the AS gene. Paternal inheritance of this gene will be phenotypically silent, since the paternal AS gene is normally silent, and the AS phenotype will only appear after maternal inheritance. Now that the UBE3A gene has been identified as the AS gene, the specific familial mutation can be identified, and “at risk” individuals can be screened. Females carrying the mutation inherited from their father have a 50% recurrence risk. Prenatal testing and counseling are recommended in such cases. Figure 117-4 Molecular diagnosis by DNA methylation. DNA from peripheral blood leukocytes is digested with XbaI/NotI and probed with a SNRPN 5' probe. NotI does not cut if DNA is methylated (maternal band; PWS UPD patient) but does cut if DNA is unmethylated (paternal; AS UPD patient). PWS and AS patients are easily distinguished from each other and from normal individuals, who show a normal “biparental” pattern.

all familial cases, have a microdeletion overlapping the first exon (PWS) or upstream of the first exon of the SNRPN gene (AS). This region is termed the imprinting center (IC), and microdeletions or mutations in the IC are detected by molecular techniques in a research laboratory (Fig. 117-3). STRUCTURAL GENE MUTATION IN AS Since the UBE3A gene has been identified as the AS gene, it is now possible to perform mutation analysis. However, since most cases arise de novo, it is likely that most patients will have different mutations in the AS gene. The ease of performing repeated mutation screens on each new case will depend on whether mutations cluster in functional domains of the encoded protein or throughout the fairly large coding region (or in regulatory regions) of the UBE3A gene.

GENETIC COUNSELING IN PRADER-WILLI AND ANGELMAN SYNDROMES DELETIONS AND UPD IN PWS AND AS Essentially, the recurrence risk of a deletion or UPD in extended families is the population risk (1/10,000 to 1/20,000). Nevertheless, in instances of advanced maternal age, the risk for having a child with UPD is significantly larger, though still small, but prenatal screening is possible where there is concern for the risks. The presence of a familial, phenotypically silent balanced translocation in chromosome 15q11–q13 has a risk of meiotic rearrangement or nondisjunction, leading to a deletion of 15q11–q13 or UPD, respectively, and prenatal diagnosis and genetic counseling are recommended in such cases.

MOLECULAR GENETICS OF PRADER-WILLI AND ANGELMAN SYNDROMES FEATURES OF IMPRINTING As is evident from the preceding discussion, imprinted genes are characterized by monoallelic expression (paternal or maternal allele only) and differential DNA methylation imprints, whereby the two alleles differ in methylation status. In addition, imprinted chromosome regions show asynchronous DNA replication, in which the two parental alleles appear to replicate at different times in S phase of the cell cycle, in comparison to nonimprinted chromosome regions in which replication is synchronous for the two alleles. Nevertheless, the basis for replication timing asynchrony is unknown, and it may be a consequence rather than a cause of imprinting. Replication timing has been used to specifically identify PWS and AS UPD patients by cytological techniques. The potential role of other factors in imprinting, such as chromatin structure, histone acetylation, and DNA binding proteins, is presently unknown. The mechanism of imprinting is complex and not completely understood, but the following picture has emerged. Imprinting is set in the germ line, and must be erased and reset (i.e., is reversible) at each generation, specific for the sex of that individual. The first step in imprint erasure and resetting is controlled by the IC, since mutations in this region block resetting of the imprint (see below). The IC signal is then transferred bidirectionally by an unknown mechanism to each imprinted gene in 15q11–q13 (Fig. 117-3). This allows each imprinted gene to set its own germ line (gametic) imprint, based on DNA methylation specific for the male or female germ line. Subsequent to fertilization, some imprinted genes carry the gametic imprint in the body of the gene and undergo modification of the imprint at the promoter, which then locks in the imprinted state and/or determines the allele-specific expression. It is well-known that the presence of DNA methylation blocks transcription, whereas an unmethylated promoter is capable of transcription, depending on tissue-specific factors.

CHAPTER 117 / PRADER-WILLI AND ANGELMAN SYNDROMES

1059

OCULOCUTANEOUS ALBINISM The only clinical feature known to be shared between PWS and AS, in deletion patients, is hypopigmentation of the skin, hair, and eyes, and is thus a result of a nonimprinted gene (the P gene). Therefore, at least in deletion patients, AS and PWS represent contiguous gene syndromes in which the clinical features arise because of more than one gene. Some PWS (and AS) patients have albinism (OCA2), at the frequency expected for carriers of OCA2, which arises from inheritance from one parent of a P gene mutation coupled with a large deletion of the other allele. MOUSE MODELS All the identified genes from human chromosome 15q11–q13, including imprinted genes and nonimprinted genes, have homologs in the same respective genetic order in the central part of mouse chromosome 7. The imprint status of these genes is also conserved between humans and mice. These findings suggest that mouse models should be an excellent model of PWS and AS. Mice with radiation-induced deletions of the nonimprinted genes allow assessment of the contribution of such genes to recessive phenotypes in this chromosomal region. Mice with maternal UPD specifically for the PWS/AS homologous region, or with paternal UPD, albeit for a much larger chromosome region that may contain other imprinted genes, have been generated by clever breeding techniques, and these animals have imprinted phenotypes that may provide models of the imprinted PWS and AS phenotypes. For example, the maternal UPD mice die at birth, probably because of the inability to feed, which may mimic the failure to thrive and gavage feeding necessary for newborn PWS infants. Nevertheless, such animals are difficult and expensive to generate, and models based on targeted mutation of the mouse homologs of the specific PWS and AS genes, or the IC elements, may provide more workable models in the future. IMPRINTING MUTATIONS Perhaps one of the most remarkable findings to come from the study of PWS and AS is that of patients with mutations in the imprinting process. These patients have microdeletions that define an IC, overlapping and upstream from the SNRPN gene (Fig. 117-3), as previously discussed. Familial mutations of the IC can transmit silently through many generations, as long as the transmitting sex does not change (Fig. 117-5). Imprinting is a reversible process in which the germline imprint must be erased and reset at each generation; thus, in the male germline, the maternally inherited imprint must be erased and reset as a male imprint, or epigenotype (referring to information that is heritable and alters the phenotype of offspring but is not encoded specifically in the genetic code of DNA), and in the female germline, the paternally inherited imprint must be erased and reset (switched). IC mutations block this resetting of the imprint in the opposite germline, leading to a failure to reset the imprint specific for the sex of that individual. Thus, when a maternally derived imprinting mutation [represented by M(M)] is transmitted through a male, the maternal epigenotype cannot be reset in his germline to the normal paternal epigenotype (Fig. 117-5A). He therefore transmits a maternal imprint on his paternal chromosome [P(M)] to 50% of his gametes; resulting offspring inheriting the abnormal gametic epigenotype [P(M)] also inherit a normal maternal epigenotype (M) from their mother, and thus they are homozygous for a maternal epigenotype (Fig. 117-5A). This is equivalent to maternal UPD, and hence PWS develops. Likewise, transmission of a paternally derived imprinting mutation [P(P)] through a female results in failure to switch the imprint and inheritance of the abnormal epigenotype [M(P)], along

Figure 117-5 Inheritance of imprinting mutations. The inherited chromosome 15 imprint (epigenotype) in somatic cells is shown inside the female (circles) and male (squares) symbols, whereas the germline imprint is shown below these symbols. When an imprinting mutation in the IC arises during maternal meiosis (I, left), a maternal epigenotype (i.e., imprint) is fixed into this chromosome. Females transmit the maternal epigenotype silently (I and II, left), but the IC mutation blocks resetting of the imprint in a male (III, left). Thus, he transmits a maternal epigenotype to 50% of his offspring, who have a “uniparental” maternal epigenotype and hence will have PWS. An analogous pattern occurs for origin and transmission through males of an IC mutation generating a fixed paternal epigenotype, but maternal transmission now gives rise to AS at 50% risk (right).

with a normal paternal imprint (P), and thus the resulting offspring are homozygous for a paternal imprint and develop AS (Fig. 1175B). In such PWS and AS patients, the mutation has no direct effect in the patient; rather, the mutation exerts its effect in the parental germ line, and it is a consequence of this effect that genetic disease arises in the offspring. This clearly represents a new paradigm as a mechanism of inherited genetic disease.

FUTURE STUDIES IDENTIFICATION OF THE PWS AND AS GENES The most important immediate research is to identify the gene(s) responsible for clinical features in PWS and to explain the clinical pleiotropy (range of features) of AS in terms of the molecular functions of the UBE3A gene. Accurate molecular diagnosis can then follow in those cases in which a gene mutation is a possible mechanism for the syndrome. Identification of the etiological gene will allow subsequent studies of:

1060

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

1. Developmental expression of the gene, which may lead to insight as to the embryological or postnatal stages at which cellular and hence phenotypic abnormalities arise. 2. Characterization of the protein product of the gene. 3. Development of mouse models in which the function of the mouse homolog of the PWS or AS gene is ablated. 4. Ultimately, the derivation and testing of potential therapeutic approaches to clinical features of the two syndromes. MECHANISM OF IMPRINTING AND RELATIONSHIP TO OTHER DISORDERS INVOLVING IMPRINTED GENES Imprinting also plays a role in several cancers and other genetic disorders, such as Beckwith-Wiedemann syndrome (BWS; see Chapter 116). In particular, many of the same mechanisms, including UPD, specific parental origin chromosome rearrangements and imprinting mutations occur in BWS, suggesting that an understanding of the imprinting and disease mechanisms in PWS and AS may lead to insight in other conditions. THERAPEUTIC APPROACHES TO PWS AND AS Additional advances may result from clinically based applications (previously discussed), such as melatonin treatment of the sleep disorder in AS and PWS. Recent advances in understanding the genes involved in feeding behavior and obesity, such as the Ob gene, may lead to better understanding of the pathway of hyperphagia and obesity in PWS and hence to therapeutic approaches, even though specific mutations in such genes are not present. More important will be the identification of the specific gene products missing in PWS patients, and the protein substrates of the UBE3A gene product that are deregulated in AS patients, which may lead to important therapeutic approaches. Finally, an understanding of the mechanism of imprinting, by which genes are silenced or activated as a consequence of their parental origin (set during gametogenesis), may lead to therapeutic approaches based on reactivation of silent imprinted genes or silencing of abnormally expressed imprinted genes in somatic cells.

ACKNOWLEDGMENTS This chapter is dedicated to the memory of Dr. Harry Angelman. I sincerely thank the AS and PWS families who make this work possible, my friends who work with me as collaborators, as well as Jim Amos-Landgraf and Nancy Rebert for illustrations. Work from the author’s laboratory is supported by the NIH #HD31491, March of Dimes Birth Defects Foundation, and the International Human Frontiers of Science Project.

SELECTED REFERENCES Albrecht U, Sutcliffe JS, Cattanach BM, et al. Imprinted expression of the murine Angelman syndrome gene, Ube3a, in hippocampal and Purkinje neurons. Nat Genet 1997;17:75–78. Angelman H. “Puppet children”: a report on three cases. Develop Med Child Neurol 1965;7:681–688. Boyd SG, Harden A, Patton MA. The EEG in early diagnosis of the Angelman (Happy Puppet) syndrome. Eur J Pediatr 1988;147:508 –513. Butler MG. Prader-Willi syndrome: current understanding of cause and diagnosis. Am J Med Genet 1990;35:319–332. Cassidy SB. Prader-Willi syndrome. Curr Prob Pediat 1984;14:1–55. Cassidy SB. Prader-Willi syndrome. . 1996. Cassidy SB. Uniparental disomy and genomic imprinting as causes of human genetic disease. Environ Mol Mutag 1995;25(Suppl 26):13–20. Cassidy SB, Beaudet AL, Knoll JMH, et al. Diagnostic testing for PraderWilli and Angelman syndromes. Report of the ASHG/ACMG Test and Technology Transfer Committee. Am J Hum Genet 1996;58:1085–1088.

Cassidy SB, Ledbetter DH. Prader-Willi syndrome. Neurol Clin 1989; 7:37–54. Cattanach BM, Barr JA, Evans EP, et al. A candidate mouse model for Prader-Willi syndrome which shows an absence of Snrpn expression. Nat Genet 1992;2:270–274. Christian SL, Bhatt NK, Martin SA, et al. Integrated YAC contig map of the Prader-Willi/Angelman region on chromosome 15q11-q13 with average STS spacing of 35 kb. Genome Res 1998;8:146–157. Christian SL, Robinson WP, Huang B, et al. Characterization of two proximal deletion breakpoint regions in both Prader-Willi and Angelman syndrome patients. Am J Hum Genet 1995;57:40–48. Clayton-Smith J, Pembrey ME. Angelman syndrome. J Med Genet 1992;29:412–415. Conroy JM, Grebe TA, Becker LA, et al. Balanced translocation 46,XY,t(2;15)(q37.2;q11.2) associated with atypical Prader-Willi syndrome. Am J Hum Genet 1997;61:388–394. Dittrich B, Buiting K, Korn B, et al. Imprint switching on human chromosome 15 may involve alternative transcripts of the SNRPN gene. Nat Genet 1996;14:163–170. Driscoll DJ. Genomic imprinting in humans. In: Friedman T, ed. Molecular Genetic Medicine, ch. 4. New York: Academic, 1994; pp. 37–77. Dykens EM, Leckman JF, Cassidy SB. Obsessions and compulsions in Prader-Willi syndrome. J Child Psychol 1996;37:995–1002. Fryburg JS, Breg WR, Lindgren V. Diagnosis of Angelman syndrome in infants. Am J Med Genet 1991;38:58–64. Glenn CC, Driscoll DJ, Yang TP, Nicholls RD. Genomic imprinting: potential function and mechanisms revealed by Prader-Willi and Angelman syndromes. Molec Hum Reprod 1997;3:321–332. Glenn CC, Saitoh S, Jong MTC, et al. Gene structure, DNA methylation and imprinted expression of the human SNRPN gene. Am J Hum Genet 1996;58:335–346. Greenberg F, Elder FFB, Ledbetter DH. Neonatal diagnosis of PraderWilli syndrome and its implications. Am J Med Genet 1987; 28:845–856. Gunay M, Cassidy SB, Nicholls RD. Prader-Willi and other syndromes associated with mental retardation and obesity. Behaviour Genet (Special issue, Allison D., Faith MS, ed.) 1997;27:307–324. Holm VA, Cassidy SB, Butler MG, et al. Prader-Willi syndrome: consensus diagnostic criteria. Pediatrics 1993;91:398–402. Horsthemke B, Maat-Kievit A, Sleegers E, et al. Familial translocations involving 15q11–q13 can give rise to interstitial deletions causing Prader-Willi or Angelman syndrome. J Med Genet 1996;65:133–136. Kishino T, Lalande M, Wagstaff J. UBE3A/E6-AP mutations cause Angelman syndrome. Nat Genet 1997;15:70–73. Knoll JHM, Nicholls RD, Magenis E, Graham JM, Jr, Lalande M, Latt SA. Angelman and Prader-Willi syndromes share a common chromosome 15 deletion but differ in parental origin of the deletion. Am J Med Genet 1989;32:285–290. Knoll JHM, Rogan PK, Nicholls RD, Wu B, Korf B, White LM. Allelespecific replication of 15q11-q13 loci: a diagnostic test for detection of uniparental disomy. Am J Hum Genet 1996;59:423–430. Lee S-T, Nicholls RD, Bundey S, Laxova R, Musarella M, Spritz RA. Mutations of the P gene in type II oculocutaneous albinism, PraderWilli syndrome plus albinism, and “autosomal recessive albinism.” New Engl J Med 1994;330:529–534. MacDonald HR, Wevrick R. The necdin gene is deleted in Prader-Willi syndrome and is imprinted in human and mouse. Hum Molec Genet 1997;6:1873–1878. Matsuura T, Sutcliffe JS, Fang P, et al. De novo truncating mutations in E6-AP ubiquitin-protein ligase gene (UBE3A) in Angelman syndrome. Nat Genet 1997;15:74–77. Mitchell J, Schinzel A, Langlois S, et al. Comparison of phenotype in uniparental disomy and deletion Prader-Willi syndrome: sex specific differences. Am J Med Genet 1996;65:133–136. Mutirangura A, Jayakumar A, Sutcliffe JS, et al. A complete YAC contig of the Prader-Willi/Angelman chromosome region (15q11– q13) and refined localization of the SNRPN gene. Genomics 1993; 18:546–552. Nakao M, Sutcliffe JS, Durtschi B, Mutirangura A, Ledbetter DH, Beaudet AL. Imprinting analysis of three genes in the Prader-Willi/Angelman

CHAPTER 117 / PRADER-WILLI AND ANGELMAN SYNDROMES

region: SNRPN, E6-associated protein, and PAR-2 (D15S225E). Hum Mol Genet 1994;3:309–315. Nicholls RD, Saitoh S, Horsthemke B. Imprinting in Prader-Willi and Angelman syndromes. Trends Genet 1998;14:194–200. Nicholls RD, Glenn CC, Jong MTC, Saitoh S, Mascari MJ, Driscoll DJ. Molecular pathogenesis of Prader-Willi syndrome. In: Bray GA, ed. The Genetics and Molecular Biology of Obesity. Pennington Center Nutrition Series, vol. V. Baton Rouge: Louisiana State University Press, 1996; pp. 560–577. Nicholls RD, Gottlieb W, Russell LB, Davda M, Horsthemke B, Rinchik EM. Evaluation of potential models for imprinted and nonimprinted components of human 15q11–q13 syndromes by fine structure homology mapping in the mouse. Proc Natl Acad Sci USA 1993;90: 2050–2054. Nicholls RD, Knoll JHM, Butler MG, Karam S, Lalande M. Genetic imprinting suggested by maternal heterodisomy in non-deletion Prader-Willi syndrome. Nature 1989;342:281–285. Prader A, Labhart A, Willi H. Ein syndrom von adipositas, kleinwuchs, kryptorchismus and oligophrenie nach myotonicartigem zustand in neugeborenalter. Schweiz Med Wochenschr 1956;86: 1260,1261. Rinchik EM, Bultman SJ, Horsthemke B, et al. A gene for the mouse pinkeyed dilution locus and for human type II oculocutaneous albinism. Nature 1993;361:72–76. Rougeulle C. Glatt H, Lalande M. The Angelman syndrome candidate gene, UBE3A/E6-AP, is imprinted in brain. Nat Genet 1997; 17:14,15.

1061

Saitoh S, Buiting K, Rogan PK, et al. Minimal definition of the imprinting center and fixation of a chromosome 15q11–q13 epigenotype by imprinting mutations. Proc Natl Acad Sci USA 1996;93:7811–7815. Saitoh S, Cassidy SB, Conroy JM, et al. Clinical spectrum and molecular diagnosis of Angelman and Prader-Willi syndrome imprinting mutation patients. Am J Med Genet 1997;68:195–206. Saitoh S, Kubota T, Ohta T, et al. Familial Angelman syndrome caused by imprinted submicroscopic deletion encompassing GABAA receptor β3-subunit gene. Lancet 1992;339:366,367. Sapienza C. Parental imprinting of genes. Sci Am 1990;263:52–60. Smeets DFCM, Hamel BCJ, Nelen MR, et al. Prader-Willi syndrome and Angelman syndrome in cousins from a family with a translocation between chromosomes 6 and 15. N Engl J Med 1992;326:807–811. Spritz RA, Bailin T, Nicholls RD, et al. Hypopigmentation in the PraderWilli syndrome correlates with P gene deletion but not with haplotype of the hemizygous P allele. Am J Med Genet 1997;71:57–62. Vu TH, Hoffman AR. Imprinting of the Angelman syndrome gene, UBE3A, is restricted to brain. Nat Genet 1997;17:12,13. Wevrick R, Francke U. Diagnostic tests for the Prader-Willi syndrome by SNRPN expression in blood. Lancet 1996;348:1068,1069. Williams CA, Angelman H, Clayton-Smith J, et al. Angelman syndrome: consensus for diagnostic criteria. Am J Med Genet 1995a;56:237,238. Williams CA, Zori RT, Hendickson J, et al. Angelman syndrome. Curr Prob Pediatr 1995b;25:216–231. Williams CA, Zori RT, Stone JW, et al. Maternal origin of 15q11–13 deletions in Angelman syndrome suggests a role for genomic imprinting. Am J Med Genet 1990;35:350–353.

CHAPTER 118 / FRAGILE X SYNDROME

118

1063

Fragile X Syndrome DAVID L. NELSON

INTRODUCTION Fragile X syndrome (MIM 30955) is among the most common of human single-gene disorders and is the leading cause of inherited mental retardation, with an estimated frequency of 1/2000 to 1/4000 individuals. It is inherited as an X-linked dominant disorder with reduced penetrance. Males and females can be affected; however, the degree of mental retardation is usually more severe in males. Fragile X syndrome derives its name from the observation of a folate-sensitive fragile site at Xq27.3 in chromosomes at metaphase prepared from afflicted individuals (Fig. 118-1). This site was initially identified in 1969, although more complete characterization of the parameters of its appearance was not completed until the mid-to-late 1970s. Fragile X syndrome had previously been recognized clinically as Martin-Bell syndrome, after a report of X-linked mental retardation in an English family in 1943. The defective gene in fragile X syndrome (FMR1) was identified by positional cloning in 1991 following efforts to locate the chromosomal anomaly. Both the fragile site and the gene defect derive from expansion and methylation of a trinucleotide repeat (CGG) located in the 5' untranslated region of the FMR1 mRNA. This was the first of a large number of unstable trinucleotide repeats found to cause human genetic disorders and remains among the best understood for the mutation’s effect on gene function.

CLINICAL FEATURES The association of the fragile X site with families exhibiting X-linked mental retardation allowed the clinical description of the syndrome to be improved. This description is being refined with the advent of DNA-based diagnosis; however, it remains a highly variable phenotype. The syndrome is difficult to diagnose in newborns and the physical features gradually accumulate with age. However, even in fully affected males, the facies are subtle and generally unremarkable to those unfamiliar with the syndrome, especially those in younger patients (Fig. 118-2). Adult male patients generally exhibit a long and narrow face with moderately increased head circumference (>50th percentile). Prominence of the jaw and forehead with particularly large and mildly dysmorphic ears are typical. Some of the phenotype is reminiscent of a mild connective-tissue disorder, exhibiting hyperextensible joints, high arched palate, pes planus, pectus excavatum, and mitral valve prolapse. Macroorchidism (enlarged testicular volume) is a common finding in postpubescent affected From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

males. Almost 90% of such males exhibit testicular volumes in excess of 25 mL as measured by an orchidometer. Mental retardation and developmental delay are the most significant clinical features of the fragile X syndrome. Prepubescent males have delayed developmental milestones, and some may display avoidance behavior similar to autism, as well as hyperactivity and attention deficit. The latter two are frequent presenting complaints in boys with fragile X syndrome. Development of speech and language is almost always involved but to variable degrees. While absence of speech is rare, milder communication difficulties are common, including a characteristic jocular, litany-like speech. Mental retardation ranges from profound to borderline with an average IQ in the moderately retarded range. It has been estimated that fragile X syndrome accounts for as much as 20% of all boys with IQ levels between 30 and 55. Females are usually much less involved compared with affected males. Somatic signs may be absent or mild, although the facies of older females may tend to resemble affected males. The mental retardation, in particular, is less severe with most female patients falling in the mild-to-borderline range. A number of studies suggest increased emotional lability in both affected and carrier females. These manifest as both behavioral and psychiatric abnormalities. While these are highly variable, several studies have identified schizotypal features, depression, social avoidance, anxiety, and shyness among girls and women with IQs in both the normal and affected ranges.

DIAGNOSIS Patients are typically referred for developmental delay, and because of the relatively mild phenotype in young patients, diagnosis is rarely achieved on clinical grounds alone. Any family with an X-linked pattern of mental retardation should be considered for fragile X testing. Cytogenetic testing for the presence of the fragile site was the only diagnostic test available until the discovery of the gene and mutation. Since the CGG repeat expansion mutation is found in virtually all fragile X patients, the use of DNA testing for diagnosis is highly reliable and has largely supplanted the use of cytogenetic testing for the fragile site. However, because of the prevalence of mental retardation from other cytogenetic abnormalities, it is important to consider a routine karyotype in the evaluation of any case of developmental delay.

MOLECULAR GENETICS MUTATIONS The vast majority (>99.5%) of fragile X patients carry the same mutation, which is a massive expansion of

1063

1064

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Figure 118-1 Partial karyotype of Geimsa-stained human chromosomes showing the fragile X site (arrow). (Reprinted with permission from JAMA 1994;271:536–542.)

a trinucleotide repeat (CGG) located in the 5' untranslated region of the FMR1 gene. This repeat is polymorphic in the human population and outside of fragile X families ranges in length from 5 to as many as 50 triplets. The most common alleles carry either 29 or 30 repeats. Approximately 70% of females are heterozygous at this locus, and normal-sized alleles are transmitted with high fidelity to the next generation. Affected individuals are found to carry more than 200 repeats and typically exhibit repeats in excess of 500. These can range to as many as 2000 and are often mosaic in length, with many different lengths observed in a single sample from an individual. These large numbers of repeats are termed “full mutations” and are usually found to be methylated at C residues within the repeat sequence as well as in the nearby CpG island that marks the gene’s promoter. Methylation has been found to correlate with loss of mRNA production, presumably through diminished transcriptional initiation. This results in loss of the product of the FMR1 gene (FMRP) and the disease. Thus, the expansion mutation in fragile X syndrome leads to loss of function of the FMR1 gene. Additional loss of function mutations have been identified in the FMR1 gene, and these are found to confer the same phenotype, reducing the likelihood that other genes are affected by methylation of the region and play a significant role in the phenotype. Mosaicism in the pattern of methylation is seen in some patients, and it is methylation status that most closely correlates with disease severity, although the extent of variation of the mutation among tissues in a single individual complicates such correlation studies, which typically rely on blood leukocytes for DNA testing. There is evidence that translation of mRNAs carrying long (>200) CGG repeats is also diminished, suggesting that large, unmethylated alleles may also confer the disease. Alleles of a size intermediate between those found in the general population and those in affected individuals have been termed

Figure 118-2 Mentally retarded adolescent male with fragile X syndrome. Note long facies with prominent forehead and ears. Typical of most patients, there is no major dysmorphia associated with this syndrome, confounding the clinical diagnosis. (Reprinted with permission from JAMA 1994;271:536–542.)

“premutations.” These are found in unaffected carriers (both male and female) of the disorder in fragile X families. No instance of expansion has been identified to arise from a normal allele, and premutations appear to be maintained for many generations in fragile X lineages without selective disadvantage. Premutations can range in length from 44 to ~200 repeats and are typically found to be between 70 and 100 repeats. They are not methylated in males, although they, like normal alleles, are subject to methylation associated with X inactivation in females. Premutations are found to change in size in nearly all transmissions from parent to offspring, yet are usually somatically stable. Thus they exhibit an extremely high mutation frequency. Mutations can take the form of changes within the premutation size range or they can involve massive expansions (transitions) to the disease, causing full mutation (Fig. 118-3). However, these latter events are found exclusively to occur on transmission from mothers to their children, never from fathers to their daughters. This sexspecificity had been recognized in empiric pedigree studies of fragile X families along, with the tendency for the probability of this event to increase in subsequent generations of a fragile X family. The molecular genetic basis for the female transition specificity is as yet unknown. However, as shown in Table 118-1, the increasing likelihood of the disease in a family can be accounted for by the observation of increasing risk of transition to the full mutation with escalating size of premutation in a female carrier, coupled with the tendency for the repeats to increase in size, while transmitted in the premutation size. This unstable character to the CGG repeat explains most of the peculiar (non-Mendelian) inheritance patterns found in fragile X syndrome. Similar phenomena in other human genetic disorders caused by unstable triplet repeats can also explain the deviations from Mendelian inheritance in those diseases as well.

CHAPTER 118 / FRAGILE X SYNDROME

1065

Figure 118-3 Representative fragile X syndrome pedigree. Partially filled circles (female) and squares (male), nonpenetrant carriers; closed squares, penetrant males; open squares, normal males. Numbers by each symbol indicate the number of FMR1 trinucleotide repeats (bold numbers signify the abnormal allele). Note that the premutation changes when transmitted as evidenced by siblings having unique premutation sizes. (Reprinted with permission from JAMA 1994;271:536–542.) Table 118-1 Incomplete Penetrance Based on Repeat Length—Riska of Full Expansion on Transmission Repeat numberb 91

Maternal carrier, % 99

Paternal carrier, % 0 0 0 0 0

Table reprinted with permission from JAMA 1994;271:536–542. aRisk of having penetrant offspring for fragile X syndrome based on premutation size. bParental alleles. Risk of the premutation expanding to a full mutation is indicated for maternal or paternal transmission. Actual risk for a penetrant offspring of a carrier female is half the indicated value to reflect segregation of the normal X chromosome, which half the time will be included in the mature ovum. These data from Fu et al., which illustrate the molecular basis of the Sherman paradox, require further investigation before being used as definitive risk assessments during genetic counseling.

Instability of the FMR1 CGG repeat has been found to increase as the number of repeats is enlarged; however, interrupting AGG triplets have been found to play an important role in maintaining stability of the repeat, both in human pedigrees and in evolutionary time. A threshold of approximately 35 uninterrupted CGG repeats appears to mark the transition from reasonably stable to completely unstable transmissions and explains the observation of stable alleles with longer repeat tracts than some unstable alleles. The FMR1 gene has conserved the CGG repeats throughout mammalian evolution, suggesting that these have a specific function in the gene. A report of specific binding proteins suggests one possibility for function. PREVALENCE OF THE PREMUTATION The high incidence of fragile X syndrome, coupled with the rather unusual inheritance pattern of premutation alleles, suggests a very high prevalence of premutations in the general population. Small-scale studies have suggested allele frequencies of 1/500 to 1/600. A recent large study found a carrier frequency for premutations (defined as >55 repeats) of 1/259 women (approximately 1/500 X chromosomes). This finding underscores the impact of this very common genetic disorder.

GENE AND GENE PRODUCTS The FMR1 gene has been completely sequenced (GenBank accession L29074), and spans 38 kb divided into 17 exons. To date, no other genes have been described in the immediate vicinity. A variety of alternatively spliced transcripts has been observed in human and mouse. These lead to a number of protein isoforms. The protein product of the FMR1 gene, FMRP has been found to be present in many adult and fetal tissues, but is concentrated in neurons, particularly those of the hippocampus and in Purkinje cells of the cerebellum. High levels of FMRP are also found in cells of the testis, primarily spermatogonia. These two sites of high-level expression are consistent with the major phenotypic aspects of the disorder. Within the cell, FMRP appears predominantly cytoplasmic; however, isoforms lacking exon 14 appear to be limited to the nucleus. In the cytoplasm, the protein appears to interact with ribosomes and may form complexes with two or more recently identified proteins of similar amino acid sequence, FXR1 and FXR2. Homologs of all three members of this gene family have been identified from a variety of vertebrates. Each of these three proteins is found to have features of RNA-binding proteins, and RNA binding has been demonstrated in vitro for each. FMRP appears to have a selectivity for certain mRNAs. These features of FMRP begin to suggest a protein involved in mRNA metabolism, perhaps with a role in the regulation of translation. It remains to be discovered how defects in this function might lead to mental retardation. NEARBY FRAGILE SITES Two additional rare, folate-sensitive fragile sites are found distal to FRAXA, the site in the FMR1 gene responsible for fragile X syndrome (Fig. 118-4). These are FRAXE and FRAXF, which were both assumed to be FRAXA by cytogenetic testing, and were only distinguished after the development of DNA probes at FRAXA. Each of these sites also results from expansion and methylation of CGG repeats, and FRAXE is clearly associated with a relatively rare form of mental retardation distinguishable from fragile X syndrome. FRAXF appears to cause no abnormalities in individuals carrying expanded repeats. A gene (FMR2) whose expression is reduced by FRAXE expansion has been recently identified. This gene is quite large (approximately 500 kb) and is transcribed from the FRAXE CpG island. The gene

1066

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Figure 118-4 Map of the Xq27.3–q28 region. Three fragile sites—FRAXA, FRAXE, and FRAXF—are shown relative to the X chromosome at Xq27.3–q28. The two genes associated with FRAXA and FRAXE are indicated as FMR1 and FMR2, respectively, with their lengths. IDS indicates iduronate sulfatase, the gene defective in Hunter’s syndrome. Distances indicated are approximate.

product bears no resemblance to the FMR1 gene, and its function is not suggested by its sequence, although it shares similarity with transcription factors. MOLECULAR GENETICS OF CHROMOSOMAL FRAGILE SITES Two additional autosomal folate-sensitive fragile sites have been characterized at the molecular level. These are FRA16A and FRA11B. Each results from an expansion of a polymorphic CGG repeat sequence, with strong similarity to the expansions seen at FRAXA, FRAXE, and FRAXF. Each shows significant methylation in the fully expanded form, suggesting that methylation of expanded CGGs does not require that the repeat be located on the X chromosome. FRA16A, like FRAXF, is not known to be associated with pathology in the expanded form. FRA11B, however, appears to predispose to the loss of the terminal portion of 11q in the gametes of carriers of the fragile site. The resulting offspring develop 11q- or Jacobsen syndrome, a disorder resulting from haploidy for this region of chromosome 11 (11q23–qter). Jacobsen syndrome patients exhibit severe mental retardation and specific dysmorphic features. FRA11B results from expansion of a CGG repeat found in the 5' untranslated region of the protooncogene CBL2. No relationship between this fragile site and elevated risk of oncogenesis has been found. Each of the five fragile sites identified at the molecular level is characterized by expanded and methylated CGG repeats, and each is rare and folate-sensitive. Fragile sites are classified as either rare or common depending on their frequency in the human population, with common sites being homozygous in nearly every person. Rare fragile sites are typically found in less than 1% of individuals. Fragile sites are visualized in metaphase preparations of chromosomes after “induction” or treatment of the cells with a variety of reagents that elicit the sites. In addition to folate, fragile sites can be induced by distamycin-A, BrdU, aphidicolin, and 5azacytidine. The folate-sensitive sites are the most numerous of the rare fragile sites, but aphidicolin-inducible sites are plentiful among the common sites. There are 21 known folate-sensitive sites, and these can be elicited by a variety of conditions. A culture medium lacking folic acid and thymidine is the simplest method, although inhibitors of folate metabolism such as methotrexate can also be used. High thymidine can also induce these fragile sites at concentrations below those that result in mitotic arrest. These treatments all impinge on the pyrimidine biosynthetic pathway and result in diminished dTTP pools in the cell. The mechanism by which the depletion of dTTP results in a chromosomal abnormality at a CGG repeat is unknown, as is the precise molecular nature of the fragile site itself. It is likely that all folate-sensitive fragile sites will be found to result from CGG repeat expansion. Recent work has described a common aphidicolin-sensitive site on chro-

mosome 3 (FRA3B) as a broad region without specific repetitive sequence elements. However, a gene associated with renal carcinoma, FHIT, has been found to span the region of fragility. It remains to be seen whether specific sequence elements will be found associated with common fragile sites.

MOLECULAR PATHOPHYSIOLOGY The emerging evidence that FMR1 plays a role in cytoplasmic mRNA metabolism, possibly in control of translation, is somewhat at odds with the pathophysiology of fragile X syndrome. The disorder has profound consequences on intellect, yet few structural or pathological abnormalities are found. It might be expected that loss of a protein involved in a fundamental process such as translational control would lead to much more widespread anomalies. It may therefore be the case that for most tissues, sufficient redundancy of function is provided by other similar proteins (FXR1 and FXR2, for example), and that it is only in the neurons that absence of FMR1 is problematic. A knockout mouse model of fragile X syndrome is similarly mildly affected, with very subtle learning defects and enlarged testes. Thus, the very highly conserved FMR1 protein can be regarded as dispensable for normal growth and structural development (if not normal intellectual development).

MANAGEMENT/TREATMENT Currently, a variety of educational interventions are employed to treat fragile X children. These can have significant benefits. Pharmacological treatment has yet to be standardized, with trials ongoing of a variety of psychoactive substances, particularly serotonin reuptake inhibitors. Additional understanding of the normal function of the FMR1 gene may provide suggestions for effective drug intervention. The potential for treating the disease through gene therapy is tempered by the observation of affected females, which suggests strongly that the gene product acts in a cell-autonomous fashion. In females, the random pattern of X-inactivation found in the critical tissue (presumably neurons) affects the intellectual potential of women carrying full mutations. This has significant bearing on the potential for treatment by gene therapy, since this modality would necessitate targeting of neurons and would require a rather high efficiency (50% or higher) to exceed the average number of expressing neurons found in female patients.

FUTURE DIRECTIONS Of the large number of human genetic disorders known to be caused by expanding trinucleotide repeats, fragile X syndrome is the best characterized with respect to the effect of the mutation on the gene. The challenges that remain, however, are large. They include development of a better understanding of the natural his-

CHAPTER 118 / FRAGILE X SYNDROME

tory of the fragile X mutation and improved methods for its detection. With improved tests, the potential for population-based screening for this common mutation will be a reality; however, along with such a scheme comes a significant challenge in providing meaningful information to those at risk. The very dynamic nature of this mutation is confusing to professionals, and this can be quite difficult to convey to individuals at risk. Given the high frequency of the disease and its profound cost to families and society, the demand for population-based screening will be high. It will be vital to develop inexpensive and accurate tests in support of such efforts, along with effective vehicles to convey the information to the public. Clearly, a large focus on the function of the FMR1 gene product is required for developing rational treatment strategies. What is the role of FMR1 in the development of neuronal circuits? What are the consequences of its absence, and how can these be modulated? Despite 5 years of attention to this interesting protein, these investigations remain in their infancy. The added benefit of such studies may be an improved understanding of fundamental cell biological processes, possibly ones integral to learning and memory.

ACKNOWLEDGMENTS The author wishes to acknowledge his many friends and colleagues for advice and assistance in preparation of this manuscript and to extend thanks to the many fragile X families whose samples and support have aided in the studies described. Figures 118-1 through 118-4 and Table 118-1 are reprinted by permission from JAMA, vol. 271, pp. 536–542, copyright 1994, American Medical Association. Work described in this manuscript was supported in part by a grant from NIH #R01-HD29256.

SELECTED REFERENCES Ashley CT Jr, Wilkinson KD, Reines D, Warren ST. FMR1 protein contains conserved RNP-family domains and demonstrates selective RNA binding. Science 1993;262:563–566. Bakker CE, Verheij C, Willemsen R, et al. Fmr1 knockout mice: a model to study fragile X mental retardation. Cell 1994;78:23–33. Eberhart DE, Malter HE, Feng Y, Warren ST. The fragile X mental retardation protein is a ribonucleoprotein containing both nuclear localization and nuclear export signals. Hum Mol Genet 1996;5:1083–1091. Eichler EE, Holden JJA, Popovich BW, et al. Length of uninterrupted CGG repeats determines instability in the FMR1 gene. Nat Genet 1994;8:88–94. Feng Y, Zhang F, Lokey LK, et al. Translational suppression by trinucleotide repeat expansion at FMR1. Science 1995;268:731–734. Fu YH, Kuhl DPA, Pizzuti A, et al. Variation of the CGG repeat at the fragile X site results in genetic instability: resolution of the Sherman paradox. Cell 1991;67:1047–1058. Gecz J, Gedeon AK, Sutherland GR, Mulley JC. Identification of the gene FMR2, associated with FRAXE mental retardation. Nat Genet 1996;13:105–108. Gu Y, Shen Y, Gibbs RA, Nelson DL. Identification of FMR2, a novel gene associated with the FRAXE CCG repeat and CpG island. Nat Genet 1996;13:109–113. Hansen RS, Canfield TK, Fjeld AD, Mumm S, Laird CD, Gartler SM. A variable domain of delayed replication in FRAXA fragile X chromosomes: X inactivation-like spread of late replication. Proc Natl Sci USA 1997;94:4587–4592. Hirst MC, Barnicoat A, Flynn G, et al. The identification of a third fragile site, FRAXF in Xq27–28 distal to both FRAXA and FRAXE. Hum Mol Genet 1993;2:197–200. Jones C, Slijepcevic P, Marsh S, et al. Physical linkage of the fragile site FRA11B and a Jacobsen syndrome chromosome deletion breakpoint in 11q23.3. Hum Mol Genet 1994;3:2123–2130.

1067

Khandjian EW, Corbin F, Woerly S, Rousseau F. The fragile X mental retardation protein is associated with ribosomes. Nat Genet 1996;12:91–93. Knight SJL, Flannery AV, Hirst MC, et al. Trinucleotide repeat amplification and hypermethylation of a CpG island in FRAXE mental retardation. Cell 1993;74:127–134. Kooy RF, D’Hooge R, Reyniers E, et al. Transgenic mouse model for the fragile X syndrome. Am J Med Genet 1996;64:241–245. Lachiewicz AM. Females with fragile X syndrome: a review of the effects of an abnormal FMR1 gene. Mental Retardation and Developmental Disabilities Res Rev 1995;1:292–297. Lubs HA. A marker X chromosome. Am J Hum Genet 1969;21:231–244. Lugenbeel KA, Peier AM, Carson NL, Chudley AE, Nelson DL. Intragenic loss of function mutations demonstrate the primary role of FMR1 in fragile X syndrome. Nat Genet 1995;10:483–485. Martin JP, Bell J. A pedigree of mental defect showing sex-linkage. J Neurol Psychiatry 1943;6:154–157. Nancarrow JK, Kremer E, Holman K, et al. Implications of FRA16A structure for the mechanism of chromosomal fragile site genesis. Science 1994;264:1938–1941. Oberlé I, Rousseau F, Heitz D, et al. Instability of a 550-base pair DNA segment and abnormal methylation in fragile X syndrome. Science 1991;252:1097–1102. Ohta M, Inoue H, Cotticelli MG, et al. The FHIT gene, spanning the chromosome 3p14.2 fragile site and renal carcinoma-associate t(3; 8) breakpoint, is abnormal in digestive tract cancers. Cell 1996;84:587–597. Parrish JE, Oostra BA, Verkerk AJMH, et al. Isolation of a GCC repeat showing expansion in FRAXF, a fragile site distal to FRAXA and FRAXE. Nat Genet 1994;8:229–235. Pieretti M, Zhang F, Fu YH, et al. Absence of expression of the FMR-1 gene in fragile X syndrome. Cell 1991;66:817–822. Richards RI, Holman K, Yu S, Sutherland GR. Fragile X syndrome unstable element, p(CCG)n, and other simple tandem repeat sequences are binding sites for specific nuclear proteins. Hum Mol Genet 1993;2:1429–1435. Rousseau F, Rouillard P, Morel M-L, Khandjian EW, Morgan K. Prevalence of carriers of premutation-size alleles of the FMR1 gene-and implications for the population genetics of the fragile X syndrome. Am J Hum Genet 1995;57:1006–1018. Schwemmle S, de Graffe E, Deissler H, et al. Characterization of FMR1 promoter elements by in vivo-footprinting analysis. Am J Hum Genet 1997;60:1354–1362. Sherman SL, Jacobs PA, Morton NE, et al. Further segregation analysis of the fragile X syndrome with special reference to transmitting males. Hum Genet 1985;69:289–299. Siomi H, Siomi MC, Nussbaum RL, Dreyfuss G. The protein product of the fragile X gene, FMR1, has characteristics of an RNA-binding protein. Cell 1993;74:291–298. Sittler A, Devys D, Weber C, Mandel J-L. Alternative splicing of exon 14 determines nuclear or cytoplasmic localisation of FMR1 protein isoforms. Hum Mol Genet 1996;5:95–102. Sutherland GR. Fragile sites on human chromosomes: demonstration of their dependence on the type of tissue culture medium. Science 1977;197:265,266. Verkerk AJMH, Pieretti M, Sutcliffe JS, et al. Identification of a gene (FMR-1) containing a CGG repeat coincident with a breakpoint cluster region exhibiting length variation in fragile X syndrome. Cell 1991;65:905–914. Warren ST. The expanding world of trinucleotide repeats. Science 1996;271:1374,1375. Wilke CM, Hall BK, Hoge A, Paradee W, Smith DI, Glover TW. FRA3B extends over a broad region and contains a spontaneous HPV16 integration site: direct evidence for the coincidence of viral integration sites and fragile sites. Hum Mol Genet 1996;5:187–195. Willemsen R, Smits A, Mohkamsing S, et al. Rapid antibody test for diagnosing fragile X syndrome: a validation of the technique. Hum Genet 1997;99:308–311. Yu S, Pritchard M, Kremer E, et al. Fragile X genotype characterized by an unstable region of DNA. Science 1991;252:1179–1181. Zhang Y, O’Connor P, Siomi M, et al. The fragile X mental retardation syndrome protein interacts with novel homologs FXR1 and FXR2. EMBO J 1995;14:5358–5366.

CHAPTER 119 / DOWN SYNDROME

119

1069

Down Syndrome STYLIANOS E. ANTONARAKIS

BRIEF HISTORICAL NOTE Down syndrome was first described by Dr. Langdon Down in 1866. The advanced mean maternal age at the time of birth of an individual with Down syndrome was first emphasized in a study by Shuttleworth in 1909. Lejeune, Gautier, and Turpin showed in 1959 that individuals with Down syndrome had a supernumerary small acrocentric chromosome that was subsequently recognized as being chromosome 21. Soon after (1960–1961), cases of trisomy 21 resulting from translocations were described by Polani and Fraccaro. Mosaicism for trisomy 21 was first published by Clarke in 1961. In the same year, some patients were described with partial trisomy 21 and phenotypic characteristics of Down syndrome (Ilbery et al., 1961; Dent et al., 1963; Hall, 1963). The parental and meiotic origin of the extra chromosome 21 was first studied (1970–1980) by cytogenetic heteromorphisms (e.g., de Grouchy 1970) and more recently by DNA polymorphisms (e.g., Antonarakis et al., 1991). The association of recombination rate and abnormal chromosomal segregation was first observed in 1987 (Warren et al.). The human genome project and the identification of genes on chromosome 21 are beginning to provide the tools for understanding the pathophysiology of the diverse phenotypes in Down syndrome. Important landmarks in this process are the descriptions of the first linkage map of chromosome 21 (Warren et al., 1989) and the first physical map using yeast artificial chromosomes (Chumakov et al., 1992).

the phenotype. The pathological, metabolic, and neurochemical changes of Alzheimer disease are present after the third decade in the brains of almost all individuals with DS, who also show a progressive loss of cognitive functions. Congenital heart disease, mainly atrioventricular (AV) canal and ventricular septal defect (VSD) is present in 30–40% of living individuals with DS, while in affected fetuses the estimated presence of AV canal is about 70%. Duodenal stenosis or atresia, imperforated anus, and Hirschsprung disease are present in 1.0–2.5% of DS individuals, frequencies much higher than in the general population. Furthermore, DS patients have a 10–20 times greater risk of developing acute leukemia. Both acute nonlymphoblastic and acute lymphoblastic leukemia may be observed. The most characteristic leukemia of DS is the acute megakaryocytic leukemia (AMKL); it has been estimated that AMKL is 200–400 times more frequent in DS than in the normal population. Another form of abnormal clonal cellular proliferation of white blood cells that is observed in some newborns with DS is the “transient” acute leukemia or “leukemoid reaction.” There is spontaneous, complete remission of this blood abnormality. Additional phenotypic findings in trisomy 21 of note include immunologic defects, hematologic alterations, hyperuricemia, thyroid dysfunction and autoimmunity, growth retardation, and sterility in males. The life expectancy of DS patients is reduced, and the causes of death are the congenital malformations (mainly heart defects), infections, and leukemia.

CHROMOSOMAL ABNORMALITIES IN DOWN SYNDROME

THE CLINICAL PHENOTYPES OF DOWN SYNDROME The clinical diagnosis of Down syndrome (DS) (MIM 190685) depends on the presence of a number of different abnormal features that affect many tissues and organs. Table 119-1 contains a partial list of physical characteristics and abnormalities found in DS, and Fig. 119-1 shows an individual with the syndrome. No individual DS patient has all of these features. Furthermore, these features may be present in other syndromes or in otherwise normal individuals. Diagnostic indexes and scoring systems have been developed for accurate diagnosis in patients suspected to have DS. Mental retardation is always present. The mean intelligence quotient (IQ) typically drops from about 80 the first year to approximately 30 the second decade of life. Newborns and infants invariably have hypotonia, which is the most common feature of From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

Since the discovery of Lejeune in 1959, the phenotype of DS has been associated with trisomy for chromosome 21. Trisomy 21 is the most common autosomal aneuploidy in live births; it occurs with a frequency of 1.03–1.30 per 1000 live births in all ethnic groups. Table 119-2 shows the frequencies at birth of specific chromosomal disorders, including trisomy 21. The majority of trisomy 21 cases are because of a free supernumerary chromosome 21. For every 1000 individuals with DS, 925 are because of free trisomy 21, 46 to translocation of the supernumerary, 21 to another acrocentric chromosome, and two to other reciprocal translocations (Fig. 119-2). Recognizable mosaicism for a normal and a trisomic cell line accounts for 27 of 1000 individuals with DS.

MATERNAL AGE AND DOWN SYNDROME It has long been recognized that maternal age is an important determinant of the incidence of DS in newborns. Figure 119-3 shows the increase of the incidence of trisomy 21 in mothers of

1069

1070

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Table 119-1 Phenotypic Features of Down Syndromea % Physical characteristics Oblique (upslanting) palpebral fissures Loose skin on nape or neck Narrow palate Brachycephaly Hyperflexibility Flat nasal bridge Large gap between 1st and 2nd toes Short, broad hands Short neck Abnormal teeth Epicanthic folds Short 5th finder Open mouth Brushfield spots Furrowed tongue Transverse palmar crease Folded or dysplastic ear Protruding tongue Abnormal dermatoglyphics Skeletal abnormalities Incurved 5th finder Pelvic dysplasia (newborns) Atlantoaxial or atlantoccipital instability Mental retardation Hypothyroidism Neurological features Hypotonia Delayed dissolution of early reflexes Alzheimer-like brain pathology (over 35 years of age) Congenital anomalies Congenital heat defects (CHD) (of living patients) AV canal (of total CHD) VSD (of total CHD) ASD (of total CHD) Fallot’s tetralogy (of total CHD) PDA (of total CHD) Gastrointestinal abnormalities Duodenal atresia Imperforated anus Hirschsprung disease Infertility in males Leukemia AMKL (acute megakaryocytic) AML/ALL Leukemoid reaction (newborn)

80 80 75 75 75 70 70 65 60 60 60 60 60 55 55 55 50 45 85 55 70 15–20 100 7–17 100 80 100 30–40 40 31 9 6 9 2.5 1.0 0.6 100 1.1 200–400 × normal 10–20 × normal 0.1

aThe frequencies of phenotypic features were obtained from numerous sources, and most have been rounded to the nearest 5%. (Majority of data are from Epstein CJ, 1995.)

different ages. The risk remains low, at about 1 in 1500 to 1 in 1000 live births, until age 30 and then increases substantially. For mothers of 35 years of age, the risk is about 1 in 380 live births, at age 40 it is 1 in 110, and at 45 years about 1 in 30. The reason for this strong maternal age effect is presently unknown, and many theories have been formulated to explain this well-documented phenomenon. There is no convincing evidence for a paternal age effect. Molecular analysis has clearly shown that this advanced maternal age effect is associated with errors of maternal meiotic origin and not paternal or mitotic errors; in these latter two categories of errors, the mean maternal age is not different from that of the mean

maternal age in Western societies. This finding is against the theory of relaxation of selection in utero that seeks to explain the increased incidence of DS with maternal age by a decrease (in older mothers) in the rate of a spontaneous abortion of aneuploid embryos and fetuses. In addition, there is no maternal age effect in trisomy 21 resulting from translocations. The theory of aging of sperm has not been supported, as no paternal age effect has been shown in paternally derived trisomy 21; however, more data are needed for this category of meiotic errors. Interestingly, more male trisomy 21 individuals than expected have been observed in DS originating from paternal meiotic errors.

CHAPTER 119 / DOWN SYNDROME

1071

including the pericentromeric markers, is interpreted as the result of a mitotic, postzygotic error. Table 119-3 shows the results of two large independent studies of more than 500 families. The conclusions are as follows:

Figure 119-1 Facial photograph of a 2-year-old male with trisomy 21 (Down syndrome).

Many conspectuses with trisomy 21 do not survive to birth. Data from amniocentesis without any further intervention during pregnancy revealed that over 20% of the DS fetuses die in utero. Furthermore, data from chorionic villus sampling at approximately 9–10 weeks of gestation indicate that about 20% of DS fetuses die between the 9th to 10th and 16th to 18th week of gestation. The estimate of survival of a DS fetus to term is about 30%. The production-line hypothesis was advanced by Henderson and Edwards in 1968. Based on mouse studies, they proposed that oocytes with greater number of chiasmata in their chromosomes are formed earlier in embryogenesis and are ovulated earlier in adult life than oocytes with fewer chiasmata. Warren et al. in 1987 first provided evidence that there is a reduced number of crossingover events in chromosomes 21 that undergo nondisjunction. Further studies from our laboratory and that of Sherman/Hassold have shown that, in maternal meiosis I errors, the number of cases with no detectable crossover is excessive, and the total length of the observed linkage maps of these nondisjoined chromosomes 21 is about half of normal. Thus, it is now clear that there is abnormal (reduced) recombination rate in maternal meiosis I nondisjunction; this abnormality in recombination may have some causative effect in nondisjunction. In maternal meiosis II errors, Lamb et al. (1996) observed increased recombination that was largely restricted to the maximal long arm of chromosome 21.

ORIGIN OF THE EXTRA CHROMOSOME IN TRISOMY 21 FREE TRISOMY 21 The availability of highly informative DNA polymorphic markers has allowed the determination of the parental origin of the extra chromosome 21 in free trisomy 21. Furthermore, DNA markers close to the centromeric region can indicate the stage of meiosis in which the segregation error had occurred. Finally, homozygosity for all markers throughout 21q,

1. Errors in meiosis that result in trisomy 21 were overwhelmingly of maternal origin, as only about 6.5% occurred during spermatogenesis. 2. The majority of errors in maternal meiosis occurred in meiosis I, and the mean maternal age associated with these was 31.2 years. The meiosis I errors accounted for 77.5% of maternal meiotic errors and 68% of all instances of free trisomy 21. 3. Maternal meiosis II errors comprised 22.5% of maternal origin errors and 20% of all cases of free trisomy 21. The mean maternal age in these errors was 32.5 years. 4. In a minority (6.6%) of the total number of families in which there was paternal nondisjunction, there are more meiosis II than meiosis I errors (62% and 38% of paternal meiosis I and II errors, respectively). The mean maternal and paternal ages in this category were similar to the mean reproductive age in Western societies. 5. In about 5.5% of free trisomy 21, the extra chromosome 21 appeared to result from an error in mitosis. In these families, the mean maternal age was around 28 years, which is similar to the mean maternal reproductive age. As expected, there was no preference for which chromosome 21 was duplicated in mitotic errors, and the extra chromosome was equally likely to be derived from either parent. TRANSLOCATION TRISOMY 21 In all 24 cases examined of de novo translocation t(14;21) trisomy 21, the extra chromosome 21 appears to be maternal in origin. In these cases, the mean maternal age was 29.2 years. The most likely mechanism of these events is that the translocation occurs before crossing over in meiosis I and is followed by normal segregation in meiosis I and II. In the majority (14 of 17 cases) of de novo translocation t(21;21) trisomy 21, the abnormal chromosome is an isochromosome (dup21q), rather than the result of a translocation caused by a fusion between two heterologous chromatids. About half of the isochromosomes studied were of paternal and half of maternal origin. The dup21p was probably formed by failure of separation, either of the chromatids in meiosis II or of sister chromatids in early mitosis. Finally, all of the t(21;21) true de novo translocations are of maternal origin similar to the t(14;21). ORIGIN OF THE EXTRA CHROMOSOME IN OTHER TRISOMIES For comparison with trisomy 21, the parental and meiotic origin of other human trisomies are as follows: 1. In trisomy 18, in which more than 150 families were studied, the origin of the extra chromosome 18 is maternal in about 90%. Postzygotic, mitotic errors probably account for about 8% of the cases. Among the maternal meiotic errors, about 34% occur in meiosis I and 66% in meiosis II. The excess of maternal meiosis II errors is in contrast with the data from all the other aneuploidies studied. Increased maternal age is associated with both maternal meiotic errors. 2. In trisomy 16, the most common trisomy at conception but not at birth, because virtually all such fetuses are spontaneously aborted, the origin of the extra chromosome 16 is always maternal. Furthermore, in all cases the error is found

1072

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Table 119-2 Chromosomal Abnormalities in 1000 Live Birthsa All abnormalities Autosomal trisomies Trisomy 21 Trisomy 18 Trisomy 13 Balanced autosomal rearrangements Unbalanced autosomal abnormalities Sex chromosomal abnormalities in phenotypic males XXY XYY Sex chromosomal abnormalities in phenotypic females 45,X XXX

5.71–6.25 1.23–1.42 1.03–1.30 0.09–0.12 0.05–0.07 1.85–1.93 0.48–0.59 2.55–2.59 0.92–1.04 0.90–0.92 1.33–1.51 0.09–0.13 0.86–1.04

aNumbers have been derived after karyotyping of more than 70,000 consecutive newborns. (Most of the data are from Hook and Hamerton, 1977.)

Figure 119-2 Histogram of frequencies of the various chromosomal abnormalities found in Down syndrome. The number of abnormalities per 1000 individuals with Down syndrome is shown. The data are averages from various sources.

in maternal meiosis I. There is increased maternal age associated with trisomy 16. 3. In trisomy 13, the data from about 50 families show that the origin of the extra chromosome 13 was maternal in about 90% of cases. Among maternal errors, the great majority (approximately 90%) occur in maternal meiosis I. 4. The data on a limited number of families with spontaneous abortions with trisomies 14, 15, and 22 indicate that, again, the majority of the errors (87%) occur in maternal meiosis. 5. In XXY (Klinefelter) syndrome, about half of the cases result from paternal and half from maternal meiotic errors. Among the paternal errors, all occur in meiosis I; among the maternal errors, 75% occur in meiosis I and 25% in meiosis II. About 5% of the maternal cases can be attributed to mitotic error. In XXX females, only 10% have received two copies of the paternal X, the remaining 90% have inherited two copies of the maternal X chromosome. Among the maternal origin XXX females, 68% are because of meiosis I and 32% to meiosis II errors.

PRENATAL DIAGNOSIS AND COUNSELING Programs for prenatal detection of trisomy 21 have been introduced by many public and private health care providers. Chromosomal analysis after amniocentesis or chorionic villus sampling have been done in hundreds of thousands of pregnancies. Amniocentesis is usually carried out at approximately 16 weeks, while chorionic villus sampling is performed at 8–10 weeks of pregnancy. In most health systems, both of these procedures are usu-

ally offered to women over 35 years of age because the risk of abortion from the procedure almost equals the risk of trisomy 21. Prenatal detection of trisomy by cytogenetic analysis of the fetus has slightly reduced the number of liveborn with DS, since in the general population only a minority of trisomy 21 individuals are born to mothers older than 35 years of age. Rapid DNA-based methods (either fluorescent in situ hybridization [FISH] or DNA polymorphisms or dosage analysis) have been developed to detect trisomy 21 in fetal tissues (after amniocentesis or chorion villus sampling); however, these methods are not in wide use, because they only detect trisomy 21, whereas a karyotype detects all of the other visible chromosomal abnormalities. Monitoring of all pregnancies using maternal serum alpha fetoprotein (AFP) in combination with chorionic gonadotropin (UCG), inhibin A, and unconjugated estrol (uE3) have provided a way to estimate a woman’s risk of having an affected fetus. If the risk exceeds a specified cutoff level (usually 1 in 250), then a karyotypic analysis of fetal cells is indicated. Although this “triple test” is not trisomy 21-specific and has a high false positive rate, it is a valuable tool for the in utero detection of trisomy 21. The introduction of this screening test in the United Kingdom and elsewhere had resulted in the in utero detection of more than half of all cases of trisomy 21. Many efforts are now being directed toward the development of methods to detect aneuploidies in fetal nucleated erythrocytes or other cells in the maternal circulation. No such method has yet been introduced in large-scale clinical trials, because their technology is still in the development stage.

1073

CHAPTER 119 / DOWN SYNDROME

Figure 119-3

Incidence of Down syndrome cases per 1000 live births as a function of maternal age.

Table 119-3 Analysis of the Origin of the Supernumerary Chromosome 21 in 510 Families with Free Trisomy 21 Using DNA Polymorphismsa Error Maternal meiosis I Maternal meiosis II Paternal meiosis I Paternal meiosis II Mitosis Maternal chrom duplication Paternal chrom duplication

N

%

Mean maternal age (years)

347 101 13 21 28 17 11

68.04 19.80 2.55 4.12 5.49

31.2 32.5 26.4 27.9 28.4 27.2 28.4

aData

are from two major studies: Johns Hopkins University (Antonarakis, et al., Nature Genet 1993) and Emory University (Sherman/Hassold) (Lamb et al., 1996).

The counseling of pregnant couples includes the explanation of the incidence of trisomy 21 in different-age mothers, the phenotypes of DS, and the presentation of risks of available diagnostic and screening procedures. Postdiagnostic counseling (after the discovery of a chromosomal abnormality) is necessary. The risk of a couple with one member being a carrier for a Robertsonian or reciprocal chromosome 21 translocation differs according to the translocations and the sex of the carrier parent. This empirically observed risk differs from the theoretical risk based on the meiotic products. Empirical data show that if the mother is a Robertsonian translocation carrier of chromosome 21 and either 13, 14, 15, or 22, the risk for an unbalanced, trisomy 21-affected offspring is about 15%; if the father is the carrier of such a translocation the risk is about 2%. For reciprocal translocations involving chromosome 21, the empirical risk is about 10% for both translocation carriers. The risk of trisomy 21 among the viable offspring of t(21;21) carriers is the expected 100%. The maternal age-independent risk for recurrent trisomy 21 for a couple who already has one child with free trisomy 21 is slightly higher than that of the general population of the same maternal age. This is probably because of germline and/or somatic mosaicism for a trisomy 21 cell line in one of the parents. In a recent study of 22 families with two free trisomy 21 offspring, parental mosaicism has been detected in five cases. It is not known if a predisposition for nondisjunction (other than mosaicism) exists in some families with two or more affected offspring.

CHROMOSOME 21: GENOME MAPPING In order to understand the pathophysiology of the different phenotypes of trisomy 21, knowledge of the genes that map in chromosome 21 is necessary. The identification and characteriza-

tion of these genes and the subsequent elucidation of their function may permit the assignment of specific phenotypes to specific gene products. The elucidation of the role of the additional copy of certain genes in specific clinical abnormalities may then open new possibilities for intervention in order to modify the resulting phenotype. The effort for mapping, cloning, sequencing, and initial characterization of chromosome 21 genes is part of the international human genome project initiative, which is already greatly enhancing the understanding of the molecular pathophysiology of human disorders. Chromosome 21 is the smallest human chromosome. The long arm (21q) is approximately 40 Mb (40 million nucleotides) and makes up about 1.2% of the human genome; the short arm (21p), which is highly homologous to those of the other four acrocentric chromosomes, is around 10–15 Mb. Because complete absence of 21p is not associated with clinical phenotypes, and trisomy 21 resulting from t(21;21), in which 21p is also deleted from the translocated chromosomes, is not different from the free trisomy 21, we do not consider 21p to be as important or contributory to the phenotypes of trisomy 21. The estimated number of genes on 21q is approximately 600–1000. Figure 119-3 shows schematically the different maps that have been developed for chromosome 21. The information from these different maps is integrated in a unifying system, the final goal of which is to decipher the nucleotide sequence of the entire 21q. The different maps are briefly described below. CHROMOSOMAL BREAKPOINT MAP The breakpoints of naturally occurring chromosomal abnormalities serve as the initial landmarks in the building of maps. A large collection of such abnormal chromosomes has been introduced into rodent-human somatic cell hybrids and used to divide the chromosome in intervals of a few megabases each.

1074

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

LONG-RANGE RESTRICTION SITE MAP A relatively accurate estimate of the distance in kilobases between landmark cleavage sites of selected restriction endonucleases is provided after restriction analysis. Infrequent cutting enzymes studies, such as NotI, are used for the initial “framework” map that provides information about the size of the chromosome. An approximately 37-Mb-long range map of NotI restriction sites has been produced for 21q. LINKAGE MAP At present, the linkage map is perhaps the most medically useful of all chromosomal maps. It consists of polymorphic DNA markers ordered, by recombination frequencies to each other, throughout the entire length of the chromosome. The majority of the polymorphic markers used are because of the normal variation of short sequence repeats and can be detected after PCR amplification. The current linkage map of 21q is the most dense map of all human chromosomes. More than 130 DNA markers have been analyzed and placed on the linkage map; the average genetic distance between adjacent markers is less than 1.5% recombination. The linkage map constructed from male meioses is shorter than that made from female meioses, probably because the meiotic process is different in the two sexes. The female map is usually longer in the proximal half of 21q; however, in the most distal region of 21q, there is an excess of male recombination and therefore a longer male linkage map. The total lengths of the male and female maps are about 55 cM, and 85 cM respectively; the sex-averaged map is 67 cM long. The linkage map of chromosome 21 is an outstanding tool for mapping both Mendelian and complex polygenic traits for the identification of specific chromosomes involved in aneuploidies or malignancies, for the determination of the origin of nondisjunction, and for the determination of the extent of partial trisomies 21. RADIATION HYBRID MAP When somatic cell hybrids that contain a single chromosome 21 are irradiated and then fused with a rodent cell line, radiation hybrids are produced that carry only small fragments of the human chromosome. By screening a large number of such hybrids for the presence or absence of particular loci, another type of a statistical map can be built up. Neighboring loci are most likely to be present together in these hybrids, while distant loci are separated. The major advantage of radiation hybrid mapping is that there is no need for polymorphic markers; any piece of DNA can be assigned to this map. The radiation hybrid map of chromosome 21 is becoming increasingly important with the addition of expressed sequence tags (ESTs), the short sequences of human cDNA clones. CLONING OF CHROMOSOME 21 IN OVERLAPPING CLONES The cloning of chromosomal 21 DNA and the identification of a continuous array of overlapping clones (contigs) is an essential step in the characterization of disease-related genes, sequencing of chromosomal DNA, precise mapping of genes, and study of the importance of partial trisomy 21 to the DS phenotypes. Various vectors have been used to clone large fragments of DNA; the most widely used are cosmids, yeast artificial chromosomes (YACs), P1 bacteriophage, P1 artificial chromosomes (PACs), bacterial artificial chromosomes (BACs). An almost continuous array of chromosome 21 YACs has been produced for 21q and now serves as the backbone for the mapping efforts. This contig was composed of 810 YACs and extends from the 21cen to 21qter. There are several areas of noncoverage, with YACs and regions with erroneous order of the mapping objects; this map is therefore in constant improvement. A cosmid contig of a 4-Mb region in the

so-called DS critical region (see below) has been recently made. Additional smaller cosmid contigs for different regions of 21q have been constructed. All of these maps of overlapping clones are necessary for the identification and characterization of chromosome 21 genes that are involved in various disease phenotypes, including those of DS. DOWN SYNDROME CRITICAL REGION (DSCR) It has long been recognized that some patients with a clinically recognized DS have only a partial trisomy 21. Many laboratories have therefore collected DNA from patients with (1) phenotypic characteristics of DS and partial trisomy 21, (2) partial trisomy 21 without the phenotypes of DS, and (3) phenotypes of DS without any obvious chromosomal abnormality. The latter group of patients is intriguing, because even after careful and extensive molecular analysis, no chromosome 21 triplication has been found, and yet the phenotypic analysis of these patients fulfills the minimal criteria for the diagnosis of DS. It is possible that these patients have a small triplicated region of chromosome 21 that has yet escaped detection or that abnormalities of few genes on chromosome 21, or elsewhere, result in a phenotype similar to that of DS. The study of patients with partial trisomy 21 and several phenotypes of DS suggested that there is a region of approximately 4 Mb between DNA markers D21S17 and ETS2 (Fig. 119-4) that, if triplicated, is associated with numerous features of DS, including flat nasal bridge, protruding tongue, high arched or narrow palate, folded ears, short and broad hands, clinodactyly of fifth finger, large gap between first and second toes, joint hyperlaxity, muscular hypotonia, short stature, and mental retardation. This region is termed Down syndrome critical region (DSCR). Extension of the triplication distally to DNA marker BCEI includes additional features of DS, such as oblique eye fissure, epicanthus, Brushfield spots, transverse palmar crease, and several dermatoglyphic signs. In all of these studies, the presence of a phenotypic feature is of importance; its absence is not taken into account, because patients with the full trisomy 21 do not show all of the phenotypic characteristics. The DSCR, which is about 10% of 21q, may contain between 50 and 100 genes (assuming that the gene distribution is the same throughout 21q). There are, however, three known patients with proximal triplication of chromosome 21 (that does not extend to the DSCR) and several phenotypes of DS, including facial features, microcephaly, short stature, hypotonia, abnormal dermatoglyphics, and mental retardation. This observation argues against a single chromosomal region being responsible for most of the DS characteristics. On the other hand, the conflicting data can be explained by a DSCR as described on distal 21q and an additional partial trisomy 21 syndrome on proximal 21q. More cases are necessary to clarify the contribution of several 21q regions to the phenotypes of DS and the precise mapping of these regions. For the remaining discussion of this chapter, the DSCR definition is for the region between D21S17 to ETS2. The atrioventricular canal, which is a characteristic heart defect in DS, has been mapped to a large 5- to 6-Mb region between D21S267 and MX1. GENES AND DISORDERS The number of genes on chromosome 21 is estimated to be about 600–1000. The gene density varies according to the chromosomal band. There are, therefore, “gene-rich” and “gene-poor” regions on 21q. Chromosomal bands 21q22.3 and 21q22.1, which do not stain with Giemsa, probably contain the majority of genes of this chromosome. In contrast, the large Giemsa-positive 21q21 band is apparently gene-poor.

CHAPTER 119 / DOWN SYNDROME

1075

Figure 119-4 Schematic representation of chromosome 21 and several of its maps. (A) Human chromosome 21 homologous regions of mouse chromosomes. (B) Mapping of the known cloned genes on chromosome 21. (C) Down syndrome critical region (DSCR). (D) Chromosome 21 breakpoint map. (E) Pulsed field gel electrophoresis map of NotI restriction fragments. (F) Linkage map using highly informative, short sequence repeats, polymorphic markers. (G) Radiation hybrid map. (H) sequence tagged site (STS) map and yeast artificial chromosome (YAC) contig. (I) Cosmid contigs of the DSCR and a portion of chromosomal band 21q22.3.

A total of 54 genes thus far have been described that map on 21q. These represent only 5.4–9% of the total number of genes on 21q, and more information on them can be obtained from the pub-

licly available databases (Genome Database, OMIM, Genbank). The function of some of these genes is known; their protein products belong to various categories, such as proteins involved in cell

1076

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

division, cell signaling and communication, cell structure and motility, metabolism, cell defense, gene/protein expression, and development of the organism. Some genes have been associated with specific disorders such as the APP with a rare form of early onset Alzheimer’s disease (MIM 104760), CBS with homocystinuria (MIM 236200), ITGB2 with a leukocyte adhesion deficiency syndrome (MIM 600065), AML1 with translocations involved in acute myeloid leukemia (MIM 151385), PFKL with a form of hemolytic anemia caused by phosphofructokinase deficiency (MIM 171860), SOD1 with one form of amyotrophic lateral sclerosis (MIM 147450), HCS with biotin-responsive multiple carboxylase deficiency (MIM 253270), ENTC with enterokinase deficiency (MIM 226200), and CST6 (cystatin B) with one type of progressive myoclonus epilepsy (MIM 254800). The loci for several monogenic hereditary disorders have been mapped to chromosome 21, such as the autoimmune polyglandular disease syndrome (MIM 240300), one form of manic-depressive illness (MIM 125480), one form of hereditary deafness (MIM 601072), and one form of a platelet disorder. The genes for all of these disorders map outside of the DS critical region, and all but one map in the gene-rich 21q22.3 region. The following genes map in the so-called DSCR (from D21S17 to ETS2): 1. Carbonyl reductase (CBR) is one of several monomeric, NADPH-dependent oxidoreductases with wide specificity for carbonyl compounds. Lymphoblasts with increased copy number of CBR show increased aldo-, keto-, and quinone reductase activity. 2. SIM2 is homologous to the drosophila single-minded gene. Its protein product is a transcription factor with basic helixloop-helix motif and two PAS domains. The drosophila sim protein is a master developmental regulator of the fly nervous system midline lineage. The role of three copies of SIM2 in trisomy 21 is under investigation. 3. The gene for the p60 subunit of the human chromatin assembly factor I (CAF1p60) encodes a protein that belongs to the WD-motif family and interacts with other polypeptide subunits to promote assembly of histones to replicating DNA. Its potential role in some phenotypes of DS is also under investigation. 4. The holocarboxylase synthetase (HCS) gene encodes an enzyme that catalyses biotin incorporation in various carboxylases involved in fatty acid synthesis, gluconeogenesis, or amino acid catabolism. 5. The ERG oncogene is related to the ets family of oncogenes. It encodes a nuclear phosphoprotein that binds to purinerich sequences, and perhaps it is required for maintenance or differentiation of early hematopoietic cells. Translocation of chromosome 21 associated with some forms of myeloid leukemias or Ewing sarcomas show breakpoints within the ERG gene, resulting in abnormal fusion of mRNA transcripts. The role of ERG in some phenotypes of DS is unknown. 6. The gene for G-protein-coupled inwardly rectifying potassium channel (GIRK2) has been also mapped within the DSCR. Its mouse homolog gene is responsible for the weaver mutant mouse. The GIRK2 belongs to a family of proteins that participate in the formation of heteromultimeric channels whose activation is regulated via a G-proteincoupled receptor. The role of GIRK2 in DS is unknown.

7. The gene for the small polypeptide PEP19 that accumulates in the developing cerebellum encodes a cytoplasmic protein containing several motifs common in proteins involved in calcium-dependent signal transduction. It is predominantly expressed in cerebellar Purkinje cells and granule cells of the olfactory bulb. 8. The gene for TPRD that encodes a protein with tetratricopeptide repeat motifs; these domains were found in proteins involved in the regulation of RNA synthesis or mitosis. The involvement of this gene in DS is unknown. Short sequences of additional genes map in the DSCR and have been obtained as parts of the exon trapping and cDNA selection experiments or the EST collections from various laboratories. The full content of genes in the DSCR and other parts of 21q will soon permit the exploration of the contribution of each gene to the trisomy 21 phenotypes. THE TRANSCRIPTION/GENE MAP A medically and biologically important map is being developed, in which the mapping objects are sequences from transcripts of chromosome 21 genes. Two main methods, namely exon trapping and cDNA selection, using the cloned material from the physical maps, are now contributing to the development of such a map. More than 1000 partial sequences from chromosome 21 genes have already been identified; these sequences are probably obtained from at least 50% of the genes of this chromosome. In addition, large international efforts to partially sequence hundreds of thousands of human cDNAs from different tissues (ESTs) will soon result in the mapping of many new transcripts to chromosome 21. Finally, future efforts to determine the sequence of the entire 21q will provide the basis for understanding the contribution of chromosome 21 genes in human disease and DS in particular.

MOUSE MODELS FOR TRISOMY 21 The laboratory mouse has attracted attention as a potential model for trisomy 21. Because three copies of many genes are contributing to DS transgenic mice with three or more copies of the mouse homologs of chromosome 21, genes can serve as models for overexpression of selected genes. Transgenic mice with overexpression of genes such as SOD1 or APP or other genes have been made in order to study the contribution of each gene to the phenotype. Furthermore, breeding of transgenic mice with different transgenes will produce strains with more than one overdosed gene. The use of YACs for microinjected material into mouse oocytes will also permit the study of overexpression of several neighboring genes (perhaps all of the DSCR genes) in one transgenic strain. Human chromosome 21 is homologous to segments of three mouse chromosomes, namely mouse chromosomes 16, 17, and 10, as shown in Fig. 119-4A. The entire DSCR is included in the mouse chromosome 16. A mouse with trisomy 16 (Ts16) was created that exhibited some characteristics of DS; however, its value was limited because Ts16 mice die in utero, and mouse chromosome 16 contains a number of genes found on human chromosomes other than 21. A more recent mouse model has been generated using a reciprocal chromosomal translocation. This model, Ts65Dn, contains a partial trisomy 16 that corresponds to human chromosome 21 from marker APP to MX1, which includes all of the DSCR. However, in addition, the Ts65Dn mouse contains a trisomy for the pericentromeric region of mouse chromosome 17 that is not homologous to human chromosome 21. The Ts65Dn mice display a variety of phenotypic abnormalities, in-

CHAPTER 119 / DOWN SYNDROME

1077

cluding early developmental delay, reduced birth weight, muscular trembling, male sterility, abnormal facies, and impaired performance in the Morris water maze test. The heart defect characteristic of DS has never been observed in liveborn Ts65Dn mice, nor do these mice show the pathology of early-onset Alzheimer’s disease. More extensive evaluation of the phenotype of the Ts65Dn mice is needed to further assess the similarity of this mouse model to the human trisomy 21.

TRISOMY: MOLECULAR PATHOPHYSIOLOGY OF INCREASED GENE DOSAGE Trisomy 21 and all other partial or full trisomies are fascinating disorders in which the phenotypes do not result from abnormal gene products but result from increased (three instead of two) copies of normal genes. The different phenotypes of trisomy 21 can therefore be viewed as the result of the increased concentration of gene products of a number of genes that map on chromosome 21. Gene dosage effects for autosomal loci have been described in Drosophila. In addition, overexpression of transgenes in mice cause distinctive phenotypes. For example, mice with an additional copy of the human erythropoietin gene develop polycythemia, a condition of increased number of red blood cells resulting from the stimulatory hematopoietic effect of the increased levels of erythropoietin in mouse. A duplication of a relatively small region (about 1.5 Mb) on chromosome 17 that includes the PLP22 gene causes CharcotMarie-Tooth disease type IA, which is a common autosomal-dominant peripheral neuropathy. Overexpression of bmi-1 gene (a member of the polycomb-group gene family) in transgenic mice results in abnormalities of the axial skeleton of the animals. Furthermore, transgenic mice with overexpression or ectopic expression of hox transgenes develop skeletal abnormalities and other malformations. It is therefore conceivable that the increase dose of some chromosome 21 genes trigger directly or indirectly (through products of genes throughout the genome) a pathway that results in a physical abnormality, homeostatic disturbance, altered developmental process, and abnormal cognitive function. It is also conceivable that some but not all genes, when in three copies, result in an abnormal phenotype. There are perhaps hundreds of genes on chromosome 21, a trisomy of which does not have any effect on the phenotype. In contrast, the trisomy 21 of some genes may always be detrimental for the organism and always results in an abnormality (Fig. 119-5A). Furthermore, some phenotypes may be allelespecific (i.e., only the triplication of certain alleles is associated with a phenotype), whereas extra copies of other alleles are compatible with normal development (Fig. 119-5B). Of course, as with many of the Mendelian, monogenic disorders, the phenotype can be modified by the contribution of the protein products of all of the other human genes, as well as by environmental factors. Gene dosage may be important in a number of gene products. Some illustrative examples from genes in mammals include the following: 1. Receptors and ligands. Increased amounts of ligands or receptors may modify the receptor-mediated cascades. For example, as mentioned above, increased levels of erythropoietin (a ligand for the erythropoietin receptor) lead to increased volume of red blood cells because of stimulation of erythropoiesis. Overexpression of the low-density lipoprotein receptor in transgenic mice prevents diet-induced hypercholesterolemia.

Figure 119-5 (A) Schematic representation of the effect of different chromosome 21 genes on the pathophysiology of Down syndrome. Filled squares represent genes, three copies of which contribute to the DS syndrome. Empty squares represent genes that, if trisomic, do not contribute to the DS phenotype. (B) Possible effects of trisomy of genes in the pathophysiology of Down syndrome. Examples of allele-independent and allele-dependent dosage effects are shown. In the alleledependent example, the combination of two stripped and one gray allele contributes to the phenotype; in contrast, the combination of two gray and one stripped allele does not have an effect on the phenotype.

2. Genes important in pattern formation in development. The overexpression or ectopic expression of hox and other development-controlling genes in transgenic mice results in abnormalities of the skeleton and other organs. 3. Enzymes. The overproduction of such proteins involved in key steps of metabolic pathways may result in a recognizable phenotype. A mutant form of alpha1-antitrypsin, which acts as antithrombin III, causes a fatal bleeding disorder from an effective increase of antithrombin III activity by 50%. However, the enzyme levels are usually not tightly regulated, and in general, three copies of genes encoding enzymes are probably not associated with phenotypes of trisomy. 4. Molecules involved in cell–cell interactions. The rate of cellular adhesivity is modified by differences in concentration of the neural cell adhesion molecule (NCAM). A 50% increase in concentration produces a fourfold increase in cell adhesion. This particular alteration may be important for some aspects of morphogenesis and anomalies of trisomy. 5. Subunits of multimeric proteins. Many proteins are composed of a number of subunits, each being a product of a different gene. The number of molecules of each subunit in the multimeric protein is constant and well-regulated. Abnormalities in the function of the multimeric protein can therefore arise by a change in the number of molecules of a particular subunit in the whole complex. This abnormal stoichiometry may result from the extra copy of a certain gene in a trisomy. Imbalance of α- and β-subunits of human hemoglobin leads to different forms of thalassemias (hereditary anemias).

1078

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

6. Transcription factors. Gene products that regulate the expression of “downstream” genes are perhaps the best candidates for gene-dosage imbalance. Increased concentration of transcriptional regulators may either increase or decrease the expression of genes or alter the temporal or tissue distribution of expression of these genes. The examples and possibilities listed above are only some alternatives to understanding the relationship of increased gene product and a resulting phenotype. The list can be enriched by other examples of different classes of gene products; furthermore, additional examples come from other model organisms, cellular biology, and transgenic animals.

THE FUTURE Patients with Down syndrome will greatly benefit from the results of the human genome project, in the identification of all chromosome 21 genes and the elucidation of their function. Products of these genes will be associated with specific phenotypes of the syndrome. Therapeutic interventions based on the molecular pathophysiology of the syndrome will be introduced that may ameliorate or prevent the appearance of certain phenotypes. Better, faster, and more accurate diagnostic methods will be introduced. Further understanding of the molecular and cellular bases of cognitive functions may also lead to prevention or treatment of the mental impairment of this common syndrome.

ACKNOWLEDGMENTS Research in the author’s laboratory on human chromosome 21 and Down syndrome is supported by grants from the Swiss FNRS, the European Union, and funds from the University and Cantonal Hospital of Geneva. I thank all members of the laboratory (past and present) for their ideas, discussion, debates, and planning and execution of experiments, and all the clinical colleagues for their excellent care of patients and families enrolled in our studies. Dr. H. Scott is particularly acknowledged for the critical reading of this manuscript. This chapter is dedicated to all families with Down syndrome members.

SELECTED REFERENCES Aitken DA, Wallace EM, Crossley JA, et al. Dimeric inhibin A as a marker for Down’s syndrome in early pregnancy. N Engl J Med 1996;334:1231–1236. Antonarakis SE. Human chromosome 21: genome mapping and exploration, circa 1993. Trends Genet 1993;9:142–148. Antonarakis SE, Avramopoulos D, Blouin JL, Talbot CC, Schinzel AA. Mitotic errors in somatic cells cause trisomy 21 in about 4.5% of cases

and are not associated with advanced maternal age. Nat Genet 1993;3:146–150. Antonarakis SE and the Down Syndrome Collaborative Group. Parental origin of the extra chromosome in trisomy 21 as indicated by analysis of DNA polymorphisms. N Engl J Med 1991;324:872–876. Chen H, Chrast R, Rossier C, et al.. Cloning of 559 potential exons of genes on human chromosome 21 by exon trapping. Genome Res 1996;6:747–760. Chumakov I, et al. A continuum of overlapping clones spanning the entire human chromosome 21. Nature 1992;359:380–386. Delabar JM, Theophile D, Rahmani Z, et al. Molecular mapping of 24 features of Down syndrome on chromosome 21. Eur J Hum Genet 1993;1:114–124. De la Cruz F, Gerald PS. Trisomy 21 (Down Syndrome): Research Perspectives. Univ Parc Press, 1981. Epstein CJ. Down syndrome (trisomy 21). In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The Metabolic and Molecular Bases of Inherited Disease. New York: McGraw-Hill, 1995; pp. 749–794. Epstein CJ. Mechanisms of the effects of aneuploidy in mammals. Ann Rev Genet 1988;22:51–75. Hook EB, Hamerton JL. The frequency of chromosome abnormalities detected in consecutive newborn studies, differences between studies, results by sex and severity of the phenotypic involvement. In: Hook EB, Porter IH, eds. Population Cytogenetics: Studies in Humans. New York: Academic, 1977; pp. 63–79. Jacobs PA , Hassold TJ. The origin of numerical chromosomal abnormalities. Adv Genet 1995;33:101–133. Korenberg JR, et al. Down syndrome phenotypes: the consequences of chromosomal imbalance. Proc Natl Acad Sci USA 1994;91: 4997–5001. Lamb NE, et al. Susceptible chiasmate configurations of chromosome 21 predispose to non-disjunction in both maternal meiosis I and meiosis II. Nature Genet 1996;14:400–405. Lucente D, Chen HM, Shea D, et al. Localization of 102 exons to a 2.5 Mb region involved in Down syndrome. Hum Mol Genet 1995;4: 1305–1311. McInnis MG, Chakravarti A, Blaschak J, et al. A linkage map of human chromosome 21: 43 PCR markers at average interval of 2.5 cM. Genomics 1993;16:562–571. Peterson A, Patil N, Robbins C, Warg L, Cox DR, Myers RM. A transcript map of the Down syndrome critical region on chromosome 21. Hum Mol Genet 1994;3:1735–1742. Reeves RH, Irving NG, Moran TH, et al. A mouse model for Down syndrome exhibits learning and behaviour deficits. Nat Genet 1995;11: 177–183. Sherman SL, Petersen MB, Freeman SB, et al. Non-disjunction of chromosome 21 in maternal meiosis I : evidence for a maternal age-dependent mechanism involving reduced recombination. Hum Mol Genet 1994;9:1529–1535. Warren AC, Chakravarti A, Wong C, et al. Evidence for reduced recombination of a non-disjoined chromosome 21 in Down syndrome. Science 1987;237:652–654.

CHAPTER 120 / THE 22q11 DELETION

120

1079

The 22q11 Deletion DiGeorge and Velocardiofacial Syndrome DEBORAH A. DRISCOLL AND BEVERLY S. EMANUEL

INTRODUCTION The association of thymic aplasia with congenital hypoparathyroidism was initially noted by Lobdell in 1959 but was not recognized as a syndrome until 1965, when Dr. Angelo DiGeorge described a group of infants with congenital absence of the thymus and parathyroid glands. Subsequently, facial dysmorphia and cardiac defects, specifically conotruncal malformations, were included in the spectrum of DiGeorge syndrome (DGS) (MIM 188400). Acknowledgment of the phenotypic overlap with velocardiofacial syndrome (VCFS) (MIM 192430) and identification of a common genetic etiology for these disorders has led to further expansion of the phenotype and a better understanding of the physical, cognitive, neurological, and psychiatric disorders that DiGeorge and velocardiofacial patients may develop. DGS was presumed to be a heterogeneous disorder. However, cytogenetic and molecular studies have shown that deletion of chromosomal region 22q11 is the leading cause of DGS and VCFS. This may result from an unbalanced translocation with monosomy 22pter→q11.2, a cytogenetically visible interstitial deletion of 22q11 [del(22)(q11.21q11.23)] or a submicroscopic deletion. In rare instances, DGS has been associated with other chromosomal abnormalities, in particular, deletions of the short arm of chromosome 10 [del(10)(p13)]. Individual case reports of other chromosomal rearrangements seen in association with DGS include monosomy and trisomy 18q, monosomy 18p, monosomy 12p with trisomy 1q, monosomy 5p, partial trisomy 1q, and duplication 9q. DGS has been associated with exposure to teratogens such as retinoic acid and alcohol, and with maternal diabetes. There have been several reports of cytogenetically normal infants with DGS and renal agenesis born to women with insulin-dependent diabetes. Molecular studies of two of these patients failed to detect a deletion within 22q11. DGS is believed to arise as a result of abnormal cephalic neural crest cell migration into the third and fourth pharyngeal arches. The neural crest cells populate the pharyngeal pouches that contribute to the development of the bones of the skull, mesenchyme of the face and palate, thymus, and thyroid, and the neuronal constituents of the head and neck. Kirby (1983) demonstrated that removal of the premigratory cardiac neural crest cells in the chick From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

embryo results in cardiac outflow tract anomalies similar to those seen in DGS. The cardiac neural crest cells have also been shown to be important in supporting the development of the glandular derivatives of the pharyngeal arches, lending further support to the hypothesis that this is a disorder of the neural crest cells (Kirby and Bockman, 1984). DGS was initially considered a rare disorder. However, recent studies suggest that the 22q11 deletion, seen in approximately 90% of DGS patients, may occur as frequently as 1 in 3000–4000 live births. The majority of 22q11 deletions occur as new mutations in the affected individual. However, a deletion may be transmitted in an autosomal-dominant fashion from parent to offspring. The clinical features are highly variable between individuals and within families with the 22q11 deletion, ranging from classic DGS to individuals with mild facial dysmorphia and a history of learning and speech difficulties. Although previously described as two distinct disorders, DGS and VCFS are manifestations of a single disorder, the 22q11 deletion.

CLINICAL FEATURES DiGEORGE SYNDROME The three classic features of DGS include congenital cardiac defects, immune deficiencies secondary to aplasia or hypoplasia of the thymus, and hypocalcemia resulting from small or absent parathyroid glands. The cardiovascular malformations involve the conotruncal region and include interrupted aortic arch type B, truncus arteriosus, and tetralogy of Fallot. The immunologic deficits are variable. Although some patients may have a profound immunodeficiency requiring thymic transplantation, many patients can maintain reasonable lymphocyte counts and normal proliferative responses to mitogens. Recent studies suggest that patients are at risk for a spectrum of parathyroid gland abnormalities, including latent hypoparathyroidism which may evolve into hypocalcemic hypoparathyroidism during adolescence or early adulthood. Patients with DGS often have dysmorphic facial features, such as hypertelorism, low-set prominent ears, and micrognathia. Since 1965, the spectrum of clinical features associated with DGS has expanded to include cleft palate, cleft lip, renal agenesis, neural tube defects, and hypospadias. Advances in medical and surgical management of children with complex congenital cardiac disease have led to increased survival of newborns and children with DGS. Recent neuropsychological evaluations indicate that these children are at risk for cognitive impairment and neurological and psychiatric problems.

1079

1080

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Figure 120-1 Frontal and side views of two patients with the 22q11 deletion syndrome. (A) Note prominent nasal root, bulbous nasal tip, protruberant right ear with underdeveloped superior helix in 9-month-old Hispanic male. (B) Note prominent nasal root, bulbous nasal tip, malar hypoplasia, long facies, thick superior helices, and protruberant ears in 11-year-old Caucasian male. (Kindly provided by Dr. EH Zackai and DM McDonald-McGinn, MS.)

VELOCARDIOFACIAL SYNDROME The major clinical features in VCFS include palatal abnormalities (cleft palate, submucous cleft palate and velo-pharyngeal insufficiency [VPI]), cardiac defects, learning disabilities, and a characteristic facial appearance. Other features seen less frequently include microcephaly, short stature, slender hands and digits, inguinal and umbilical hernias, and scoliosis. Parents may report a history of nasal regurgitation, feeding difficulties, and failure to thrive. In contrast to DGS, ventricular septal defects are the most frequent cardiac defects in VCFS patients. Characteristic facial features have been described that include a long face, a prominent nose with a bulbous nasal tip and narrow alar base, almond-shaped or narrow palpebral fissures, malar flattening, recessed chin, and malformed prominent ears (Fig. 120-1). However, these features are not always apparent during infancy and childhood. An early study of clinically diagnosed VCFS children reported a history of developmental delay and mean verbal IQ scores in the

dull normal intelligence range to mildly retarded range. More recently, an analysis of the psychoeducational and neurodevelopmental status of a cohort of patients, ages 4–20, with the 22q11 deletion, found similar full-scale IQ scores (mean 73.1 ± 10.1). In contrast, there was a striking difference between verbal and performance IQ scores with 9 of 10 patients demonstrating higher verbal IQ scores. These studies suggest that both males and females with the 22q11 deletion are at risk for nonverbal learning disabilities. In addition to developmental delay and hypotonia, neural tube defects have been described in a small number of patients. Magnetic resonance imaging (MRI) of the brains of VCFS patients has identified several structural brain abnormalities including a small vermis, small posterior fossa, and small cysts adjacent to the anterior horns, suggesting that neurological findings may not be uncommon. However, these findings did not appear to correlate with developmental, cognitive, or personality disorders. In contrast,

CHAPTER 120 / THE 22q11 DELETION

Lynch et al. (1995) described a symptomatic 34-year-old man with VCFS and a 22q11 deletion who presented with progressive gait difficulties, muscle stiffness, dysarthria, and dysmetria suggestive of a neurodegenerative disorder. Cerebellar atrophy was diagnosed by MRI. The significance of these findings will require prospective neurological assessments of deletion-positive patients to determine their risk for developing neurological problems. Recent studies suggest that individuals with VCFS are at a higher risk than the general population to develop schizophrenia during late adolescence and early adulthood. Additional studies that include a larger sample size will be necessary to determine the relative risk for developing schizophrenia in this population of patients. Molecular studies of families with schizophrenia have suggested linkage to the long arm of chromosome 22, and it has been hypothesized that there may be a schizophrenia-susceptibility gene in this region. However, it remains to be proven whether these two findings indicate the presence of a single locus associated with schizophrenia within the DiGeorge chromosomal region (DGCR) on 22q. PHENOTYPIC VARIABILITY ASSOCIATED WITH THE 22q11 DELETION SYNDROME There is a wide range of intraand interfamilial phenotypic variability associated with the 22q11 deletion. While some individuals present with classic findings of DGS, others, including the parents of some individuals, have relatively subtle features such as minor dysmorphic facial features and mild cognitive impairment. Furthermore, the observed differences between patients with DGS and VCFS probably reflects an ascertainment bias. For example, the majority of DGS patients are identified in the neonatal period with a major congenital heart defect, whereas many patients with VCFS are diagnosed in a cleft palate or craniofacial center and remain unrecognized until they reach school age when speech and learning difficulties are evident. DGS and VCFS were previously recognized as two distinct disorders. However, there is significant phenotypic overlap and they share a common etiology. Thus, we and others consider deletion-positive DGS and VCFS patients as having the same disorder, the 22q11 deletion syndrome. OTHER DISORDERS ASSOCIATED WITH THE 22q11 DELETION The DiGeorge anomaly has been observed in several other syndromes, including CHARGE association, asymmetric crying face syndrome, Kallman syndrome, and Noonan syndrome. However, the 22q11 deletion has not been found to be causally related to these disorders in a significant number of cases. In contrast, deletions of 22q11 have been detected in patients with conotruncal cardiac malformations, conotruncal anomaly face syndrome (CTAFS), and Opitz syndrome. Knowledge of the genetic basis for these disorders has also expanded the spectrum of features associated with DGS and, more specifically, with the 22q11 deletion. CONOTRUNCAL CARDIAC MALFORMATIONS Several studies have detected 22q11 deletions in patients with isolated and familial forms of congenital cardiac defects, suggesting that the genes in this region play a major role in cardiac development. Goldmuntz et al. (1993) demonstrated that 20–30% of newborns and children ascertained through the cardiology clinic with one of the three most common types of cardiac defects seen in DGS (interrupted aortic arch type B, truncus arteriosus, and tetralogy of Fallot) have deletions within 22q11. Although patients with hypocalcemia and immunodeficiencies were excluded from the study, prospective follow-up of these patients has demonstrated

1081

that they eventually manifest subtle dysmorphic facial features and report a history of nasal regurgitation suggestive of velopharyngeal insufficiency or unrecognized palatal problems. Similar findings were noted in the reported familial studies. CONOTRUNCAL ANOMALY FACE SYNDROME Conotruncal anomaly face syndrome (CTAFS) has striking similarities to DGS and VCFS and is characterized by the presence of conotruncal cardiac defects in association with a characteristic facial appearance. The facial features include ocular hypertelorism, lateral displacement of the inner canthi, flat nasal bridge, small mouth, narrow palpebral fissures, bloated eyelids, and malformed ears. This syndrome has been well characterized in a population of Japanese patients. The finding of 22q11 deletions in CTAFS patients confirmed an earlier suggestion that CTAFS and VCFS are the same entity. OPITZ/GBBB SYNDROME Recent cytogenetic and molecular studies suggest that Opitz/GBBB syndrome is genetically heterogeneous. However, there appears to be at least one autosomaldominant locus on chromosome 22q11.2. Opitz/GBBB syndrome is characterized by hypospadias, laryngotracheal anomalies, and hypertelorism. Patients may have other features, including cleft lip and palate, cardiac defects, umbilical and inguinal hernias, cryptorchidism, imperforate anus, and facial dysmorphia such as telecanthus, prominent nasal bridge, or depressed nasal root with anteverted nares. Four patients, including a father and son, with Opitz syndrome with 22q11 deletions have been reported, suggesting that, in some cases, deletions within 22q11 are causally related to Opitz syndrome. Linkage analysis of families with Opitz syndrome provide additional evidence that there is a genetic locus within 22q11 responsible for this disorder as well as a locus on the X chromosome (Xp22). While there are some phenotypic similarities between these disorders, additional studies will be necessary to determine if Opitz syndrome is a genetically distinct disorder from DGS and VCFS.

CYTOGENETIC AND MOLECULAR STUDIES Cytogenetic studies of patients with DGS provided the initial evidence linking chromosome 22 to DGS. Early reports of DGS patients with unbalanced translocations resulting in the loss of 22pter→q11 and two patients with interstitial deletions of 22q11 suggested that this region of 22 was important in the etiology of DGS. Subsequently, restriction fragment-length polymorphism analysis and DNA dosage studies demonstrated that approximately 90% of cytogenetically normal DGS patients have microdeletions within 22q11. The presence of features common to DGS in patients diagnosed with VCFS prompted investigators to study VCFS patients for evidence of a 22q11 deletion. Cytogenetic studies using high-resolution banding techniques detected interstitial deletions (del[22][q11.21q11.23]) in 20% of VCFS patients. Molecular studies with chromosome 22 probes previously shown to be deleted in patients with DGS demonstrated that the majority of individuals with VCFS have deletions of 22q11. Furthermore, these studies coincided with the increasing utilization of fluorescence in situ hybridization (FISH) for the detection of microdeletions and subtle translocations. FISH of metaphase chromosomes using DNA probes from the DiGeorge chromosomal region (DGCR) led to a dramatic increase in the detection of 22q11 deletions in patients with features of DGS, VCFS, CTAFS, and conotruncal cardiac malformations (Fig. 120-2).

1082

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Figure 120-2 FISH. Metaphase chromosomes from a patient with 22q11.2 deletion syndrome hybridized with the probe N25 (D22S75) from the DGCR and control probe cos 82 that marks the distal long arm of 22. A hybridization signal for N25 is seen on only one chromosome 22 (arrow). This is consistent with a deletion on the other 22 (arrowhead).

THE DiGEORGE CHROMOSOMAL REGION The majority of patients have large deletions (1–2Mb) within 22q11; this commonly deleted region is referred to as the DiGeorge chromosomal region (DGCR). Delineation of this region was initially based on the finding of a consistently deleted region within 22q11 in patients with DGS; the minimal region of overlap in the first 14 patients studied spanned the region from D22S75 (N25) through D22S259 (R32). Since then, efforts in several laboratories have been directed toward determining the “minimal critical region” or the smallest region of overlap between patients with unbalanced translocations and interstitial deletions. YAC and cosmid contigs have been constructed across this region to facilitate the identification of candidate genes, and pulsed field gel electrophoresis (PFGE) and translocation breakpoint mapping have been utilized to develop a physical map of the region. The positioning of DGS/VCFS-associated translocations in this region of 22 has been used to refine the localization of DNA markers within the region and to narrow the “minimal critical region.” Many of these translocations serve as landmarks within 22q11 (Fig. 120-3). Initially, the critical region was defined by positioning a balanced 10;22 translocation (GM05878) established from an unaffected father of a child with DGS. Subsequently, an unbalanced 11;22 translocation (GM00980) from a VCFS patient was positioned centromeric to the 10;22 translocation as the distal boundary of the minimal critical region. More recently, we have narrowed the critical region to a 200- to 300-kb segment by positioning the 22q11 breakpoint of an unbalanced 15;22 translocation from a patient with features of DGS and VCFS between locus D22S75(N25) and GM00980. The proximal boundary of the critical region and the commonly deleted region lies between loci D22S427 and D22S36.

Within this “minimal critical region” lies a balanced 2;22 translocation that was initially described by Augusseau in 1986. The patient (ADU) had telecanthus, micrognathia, severe aortic coarctation with a hypoplastic left aortic arch, decreased E rosettes, and mild neonatal hypocalcemia consistent with a diagnosis of partial DGS. Her mother (VDU), who is also a balanced carrier of this translocation, reportedly has features suggestive of VCFS including hypernasal speech, micrognathia, and an inverted T4/T8 ratio. This is the only balanced translocation in the region that is associated with a classical DGS/VCFS phenotype. Budarf et al. (1995) recently cloned this translocation and identified a candidate gene, DGCR3, which is disrupted by the chromosomal rearrangement. Numerous other DGS/VCFS-associated translocations have been localized within the commonly deleted region, distal to the minimal critical region (Fig. 120-3). PARENTAL ORIGIN OF THE 22q11 DELETION Sporadic 22q11 deletions may be of either maternal or paternal origin. Several investigators have suggested that there may be an excess of maternally-derived de novo 22q11 deletions; however, these studies are limited by their small sample size. Furthermore, both maternally and paternally derived DGS/VCFS-associated translocations occur with approximately equal frequency, suggesting that there is not a preferential parental origin associated with DGS and VCFS. Hence, it is unlikely that parental origin accounts for the phenotypic variability seen in association with the 22q11 deletion. This is in contrast to Prader-Willi and Angelman syndromes, where paternal and maternal deletions of chromosome 15q11–13, respectively, are the rule.

CLINICAL AND PRENATAL DIAGNOSIS Deletions of 22q11 may be detected using high-resolution banding techniques; however, several studies have clearly shown that FISH of metaphase chromosomes using DNA probes from this region of chromosome 22 is the most efficient method for detecting 22q11 deletions. Routine cytogenetic analysis continues to be important for the detection of other chromosomal rearrangements, such as translocations, which may involve chromosomes other than 22. Hence, it is recommended that FISH be used as an adjunct to routine cytogenetic analysis. At the present time, many commercial and hospital-based laboratories are utilizing FISH with the probes from the DGCR for the clinical and prenatal detection of the 22q11 deletion (Fig. 120-2). Antenatal detection of the 22q11 deletion by FISH has been successfully performed on cultured amniocytes and chorionic villi obtained from at-risk pregnancies by FISH. Pregnancies considered at high risk include those of deletion-positive parents and those in which ultrasound or echocardiography demonstrates a fetal conotruncal cardiac malformation. Although the recurrence risk for normal parents with a previously affected child is low, parents often request prenatal testing for the deletion to exclude the small possibility of germline mosaicism. Since the size and extent of the deletion do not correlate with the phenotype, the phenotypic outcome cannot be predicted based solely on the presence of a deletion. Fetal imaging techniques, such as ultrasonography and fetal echocardiography, may be used to evaluate the

CHAPTER 120 / THE 22q11 DELETION

1083

Figure 120-3 Diagram of the DGCR demonstrating the relative positions of several loci and translocations, including ADU/VDU and the t(15;22), GM00980, and GM05878, which have been used to define the minimal critical region. The location of the breakpoints associated with the interstitial deletions or microdeletions and the distal translocations are included. The commonly deleted region or DGCR includes the deletion boundaries and the minimal critical region.

fetus for the presence of a cleft palate and/or cardiac malformation. Subtle congenital anomalies, such as a submucous cleft palate and dysmorphic features, cannot be appreciated by sonography. Fetal sonography and echocardiography are the only available prenatal tests that are helpful when evaluating at-risk pregnancies that are not causally related to an abnormality of 22q11. MANAGEMENT ISSUES Patients with the 22q11 deletion are ascertained through a variety of medical specialty clinics on the basis of their presenting feature(s). Recognition of which patients are at risk for the deletion is important so that the medical community and the schools can meet their needs. This has been clearly demonstrated in the cardiology clinic where a significant number of deletion-positive patients were ascertained solely by the presence of a conotruncal malformation. Because of the wide range of congenital anomalies, immune, endocrine, speech, neurologic, cognitive, and psychiatric problems that 22q11 deletion

patients may develop, we urge clinicians to consider comprehensive evaluations by a team of medical professionals in genetics, cardiology, immunology, endocrine, plastic surgery, speech and hearing, and developmental pediatrics. Because DGS, VCFS, CTAFS, and (in some instances) Opitz/GBBB syndrome share a common etiology, we should consider deletion-positive patients with these disorders as having the 22q11 deletion syndrome. Prospective studies of 22q11 deletion patients will further delineate the range and severity of physical, cognitive, and neuropsychological problems for which these patients are at risk. Furthermore, identification of a 22q11 deletion enables the clinician to provide the affected patients and their families with a more accurate assessment of their recurrence risk. Although the majority of the 22q11 deletions occur de novo only in the affected individual, a deletion may be transmitted in an autosomal-dominant fashion. Hence, we recommend deletion testing of parents of

1084

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

deletion-positive individuals. Because the phenotype is so variable, parents with only minor features and mild cognitive impairment are often undiagnosed until the birth of a more severely affected child. Features that are highly suggestive of a 22q11 deletion in one of the parents include mild facial dysmorphia, history of poor school performance, hypernasal speech, and nasal regurgitation. Deletion-positive parents and children have a 50% chance of transmitting the deletion-bearing chromosome in a subsequent pregnancy. Presently, it remains difficult to counsel families of nondeleted children with DGS and VCFS unless an etiology such as a translocation or maternal diabetes has been identified. It remains a possibility that these individuals may have a smaller deletion that cannot be detected with the probes currently available or a mutation within one of the genes in the DGCR. Alternatively, there may be more than one chromosomal locus associated with DGS and VCFS.

IDENTIFICATION OF GENES IN THE DiGEORGE CHROMOSOMAL REGION (DGCR) Two genes, catechol-O-methyltransferase (COMT) and glycoprotein Ibβ (GpIbb), have been mapped to the DGCR, but distal to the minimal critical region. COMT is important in the metabolism of catecholamines, such as noradrenaline, adrenaline, and dopamine. It has been suggested that it may play a role in the development of the psychiatric disorders associated with the 22q11 deletion syndrome. GpIbb is a component of the major platelet receptor for von Willebrand factor. Defects in the receptor result in a rare autosomal-recessive bleeding disorder, Bernard-Soulier syndrome (BSS), which is characterized by excessive bleeding, thrombocytopenia, and very large platelets. A single patient with features of BSS and VCFS with a 22q11 deletion has been described. Haploinsufficiency for this region of 22 unmasked a mutation in the remaining GpIbb allele, resulting in manifestations of BSS. Although presumably an infrequent occurrence, this mechanism, unmasking a recessive allele, might explain some of the phenotypic variability seen among patients with 22q11 deletions. Several genes have been isolated from the DGCR that may play a role in the development of some of the phenotypic features observed in association with deletions of 22q11. The most promising candidate genes lie within the minimal critical region. One of the first novel genes described in the region, N25 cDNA clone, was identified by screening fetal liver and brain libraries with a chromosome specific NotI linking clone from the commonly deleted region. The N25 cDNA is expressed in adult skeletal muscle, and studies are in progress to determine its function and developmental expression pattern. DGCR3 was identified using a positional cloning approach. It was hypothesized that the only balanced translocation associated with a DGS/VCFS phenotype in the DGCR, ADU/VDU [t(2;22)], interrupts a locus critical to DGS. The 2;22 translocation disrupts an open reading frame whose predicted protein product is weakly homologous to rat and mouse androgen receptor sequences at the N-terminal transactivation domain. In addition, it contains a leucine zipper motif, suggesting that it might be a DNA binding protein. Another candidate gene, DGCR2, maps 10 kb distal to the ADU/VDU breakpoint, within the minimal critical region, and is deleted in known-deletion patients. This gene is expressed in a variety of human adult and fetal tissues and codes for a potential adhesion receptor, which might mediate specific adhesive interactions resulting in abnormal migration of the

neural crest cells or interaction with the branchial arches. By sequence analysis, this gene appears to be homologous to the genes LAN and IDD. Although these genes are not disrupted by the ADU/ VDU translocation, the translocation may separate a locus control region from the coding sequence. The human homolog of the rodent citrate transport protein (CTP) gene has been cloned and mapped to the minimal critical region in 22q11. CTP is located on the mitochondrial inner membrane and functions as an anion transport protein. However, its role in the 22q11 deletion syndrome remains to be determined. Several groups have isolated a gene (CLTCL/CLTD/CLH-22) with significant homology to clatharin heavy chain, a major structural component of coated pits and vesicles. CLTCL lies within the minimal critical region and is expressed predominately in adult skeletal muscle. Further studies will be necessary to determine the possible role this gene plays in the 22q11 deletion syndrome. Several other genes have been identified distal to the minimal critical region, suggesting that they may not be as crucial to the phenotype. Functional and genetic analyses of these genes may determine whether they modify or influence the phenotypic outcome. The novel gene, N41 cDNA, is expressed in human tissue, including heart, liver, brain, and placenta. A zinc finger gene, ZNF74, is expressed during mouse embryogenesis as well as in human fetal tissues. A cDNA clone, T10, isolated from a mouse embryo library, encodes a serine and threonine-rich protein with no strong homologs or known function. This gene is expressed during early mouse embryogenesis as well as in human fetal tissue, but is probably not the major gene involved in DGS. Another cDNA, TUPLE1, was isolated from a human fetal brain library and a 10.5-day mouse embryo library. TUPLE1, also referred to as HIRA, encodes a putative transcriptional regulator. Although it has been shown to be deleted in patients with known 22q11 deletions, it does not appear to be deleted nor rearranged in nondeleted patients with DGS, VCFS, or conotruncal cardiac defects. Finally, the gene LZTR-1, was isolated from a fetal brain cDNA library. Sequence analysis demonstrates homology to a leucine-zipperlike transcriptional regulator protein. LZTR-1 is expressed in fetal brain, heart, liver, kidney, and lung. However, the LZTR-1 gene is not consistently deleted in DGS patients and does not appear to be rearranged in the nondeleted patients. Thus, it is unlikely to be a good candidate gene for the major features of DGS/VCFS. Further studies are necessary to understand the role that each of these genes might play in the development of the complex and variable phenotype of the 22q11 deletion syndrome. Additional studies are in progress to determine whether nondeleted patients with DGS/VCFS have mutations within these or other genes in the minimal critical region. To understand the effects of haploinsufficiency of these genes, functional studies and the creation of transgenic or knockout mice will be required. These studies may enable us to determine if these disorders are caused by deficiency of a single gene or loss of several genes within the DGCR.

CONCLUSION The 22q11 deletion associated with DiGeorge and velocardiofacial syndrome is one example of a microdeletion syndrome. Such syndromes were well-defined clinical entities prior to the identification of a specific chromosomal deletion. The deletion is rarely identified by routine cytogenetic analysis, often requiring high-resolution banding and/or molecular cytogenetic techniques such as FISH. These techniques have led to the detection of dele-

CHAPTER 120 / THE 22q11 DELETION

tions in a proportion of patients with the following clinically recognizable syndromes: Prader-Willi and Angelman syndromes (15q), Miller Dieker and Smith Magenis syndromes (17p), LangerGideon (8q), WAGR (Wilm’s tumor, aniridia, genital abnormalities, and retardation) (11p), Williams syndrome (7q), Alagille syndrome (20p), and Rubenstein-Taybi syndrome (16p). These disorders have been referred to as “contiguous gene deletion syndromes” or, more recently, “microdeletion syndromes.” It has been hypothesized that the clinical features result from the deletion of several different genes in close proximity within the same chromosomal segment. The deleted region presumably contains numerous genes, and it is unlikely that all of these genes play a major role in the development of the phenotype associated with these disorders. Some of these disorders may result from haploinsufficiency of a single gene, whereas others may result from the loss of several genes within the deleted segment. As the genes are identified within the segment lost in each of the microdeletion syndromes, additional studies will be necessary to understand their function and role in the pathogenesis of each disorder.

SELECTED REFERENCES Budarf ML, Collins J, Gong W, et al. Cloning a balanced translocation associated with DiGeorge syndrome and identification of a disrupted candidate gene. Nat Genet 1995;10:269–278. Burn J, Takao A, Wilson D, et al. Conotruncal anomaly face syndrome is associated with a deletion within chromosome 22. J Med Genet 1993;30:822–824. Conley ME, Beckwith JB, Mancer JFK, Tenckhoff L. The spectrum of the DiGeorge syndrome. J Pediatr 1979;94:883–890. Daw SC, Taylor C, Kraman M, et al. A common region of 10p deleted in DiGeorge and velocardiofacial syndromes. Nat Genet 1996;13:458–460. Demczuk S, Aledo R, Zucman J, et al. Cloning of a balanced translocation breakpoint in the DiGeorge syndrome critical region and isolation of a novel potential adhesion receptor gene in its vicinity. Hum Mol Genet 1995;4:551–558. Driscoll DA, Budarf ML, Emanuel BS. A genetic etiology for DiGeorge syndrome: consistent deletions and microdeletions of 22q11. Am J Hum Genet 1992;50:924–933. Driscoll DA, Salvin J, Sellinger B, et al. Prevalence of 22q11 microdeletions in DiGeorge and velocardiofacial syndromes: implications for genetic counselling and prenatal diagnosis. J Med Genet 1993; 30:813–817. Driscoll DA, Spinner NB, Budarf ML, et al. Deletions and microdeletions of 22q11.2 in velo-cardio-facial syndrome. Am J Med Genet 1992; 44:261–268. Emanuel BS, Driscoll D, Goldmuntz E, et al. Molecular and phenotypic analysis of the chromosome 22 microdeletion syndromes. In: The Phenotypic Mapping of Down Syndrome and Other Aneuploid Conditions. New York: Wiley-Liss, 1993; pp. 207–224. Goldmuntz E, Driscoll D, Budarf ML, et al. Microdeletions of chromosomal region 22q11 in patients with congenital conotruncal cardiac defects. J Med Genet 1993;30:807–812. Gong W, Emanuel BS, Galili N, et al. Structural and mutational analysis of a conserved gene (DGSI) from the minimal DiGeorge syndrome critical region. Hum Mol Genet 1997;6:267–276. Gottlieb S, Emanuel BS, Driscoll DA, et al. The DiGeorge syndrome minimal critical region contains a goosecoid-like (GSCL) homeobox

1085

gene that is expressed early in human development. Am J Hum Genet 1997;60:1194–1201. Greenberg F. DiGeorge syndrome: an historical review of clinical and cytogenetic features. J Med Genet 1993;30:803–806. Greenberg F, Elder FFB, Haffner P, Northrup M, Ledbetter DH. Cytogenetic findings in a prospective series of patients with DiGeorge anomaly. Am J Hum Genet 1988;43:605–611. Halford S, Wadey R, Roberts C, et al. Isolation of a putative transcriptional regulator from the region of 22q11 deleted in DiGeorge syndrome, Shprintzen syndrome and familial congenital heart disease. Hum Mol Genet 1993;2:2099–2107. Holmes SE, Riazi MA, Gong W, et al. Disruption of the clathrin heavy chain-like gene (CLTCL) associated with features of DGS/VCFS: a balanced (21;22)(p12;q11) translocation. Hum Mol Genet 1997;6: 357–367. Jaquez M, Driscoll DA, Li M, et al. Unbalanced 15;22 translocation in a patient with manifestations of DiGeorge and velocardiofacial syndrome. Am J Med Genet 1997;70:6–10. Karayiorgou M, Morris MA, Morrow B, et al. Schizophrenia susceptibility associated with interstitial deletions of chromosome 22q11. Proc Natl Acad Sci USA 1995;92:7612–7616. Kirby ML, Bockman DL. Neural crest and normal development: a new perspective. Anat Rev 1984;209:1–6. Kirby ML, Gale TF, Stewart DE. Neural crest cells contribute to aorticopulmonary septation. Science 1983;220:1059–1061. Lammer EJ, Opitz JM. The DiGeorge anomaly as a developmental field defect. Am J Med Genet 1986;29:113–127. Lindsay EA, Greenberg F, Shaffer LG, Shapira SK, Scambler PJ, Baldini A. Submicroscopic deletions at 22q11.2: variability of the clinical picture and delineation of a commonly deleted region. Am J Med Genet 1995;56:191–197. Lindsay EA, Halford S, Wadey R, Scambler PJ, Baldini A. Molecular cytogenetic characterization of the DiGeorge syndrome region using fluorescence in situ hybridization. Genomics 1993;17:403–407. Lynch DR, McDonald-McGinn DM, Zackai EH, et al. Cerebellar atrophy in a patient with velocardiofacial syndrome. J Med Genet 1995; 32:562,563. McDonald-McGinn DM, Driscoll DA, Bason L, et al. Autosomal dominant “Optiz” GBBB syndrome due to a 22q11.2 deletion. Am J Med Genet 1995;59:103–113. Pulver AE, Nestadt G, Goldberg R, et al. Psychotic illness in patients diagnosed with velo-cardio-facial syndrome and their relatives. J Nerv Ment Dis 1994;182:476–478. Ravnan JB, Chen E, Golabi M, Lebo RV. Chromosome 22q11.2 microdeletions in velocardiofacial syndrome patients with widely variable manifestations. Am J Med Genet 1996;66:250–256. Scambler PJ, Carey AH, Wyse RKH, et al. Microdeletions within 22q11 associated with sporadic and familial DiGeorge syndrome. Genomics 1991;20:201–206. Shprintzen RJ, Goldberg RB, Young D, Wolford L. The velo-cardio-facial syndrome: a clinical and genetic analysis. Pediatrics 1981;67:167–172. Sirotkin H, Morrow B, DasGupta R, et al. Isolation of a new clathrin heavy chain gene with muscle-specific expression from the region commonly deleted in velo-cardio-facial syndrome. Hum Mol Genet 1996;5:617–624. Wadey R, Daw S, Taylor C, et al. Isolation of a gene encoding an integral membrane protein from the vicinity of a balanced translocation breakpoint associated with DiGeorge syndrome. Hum Mol Genet 1995;4:1027–1033. Wilson DI, Cross IE, Goodship JA, et al. A prospective cytogenetic study of 36 cases of DiGeorge syndrome. Am J Hum Genet 1992;51:957–963.

CHAPTER 121 / OROFACIAL CLEFTING

121

1087

Orofacial Clefting JACQUELINE T. HECHT AND SUSAN H. BLANTON

INTRODUCTION Clefts of the orofacial region are a group of common birth defects that may range from those with little clinical consequence, such as a bifid uvula, to those requiring extensive surgical intervention, such as bilateral cleft lip with palatal involvement. While orofacial clefts often occur as isolated defects, they are also part of over 300 recognized genetic syndromes. In most cases, the underlying etiology is unknown, but it is widely accepted that genetic and nongenetic factors are important. The most common orofacial clefts involve the lips and palate. As with the larger group of orofacial clefts, clefts of these structures most often occur as isolated defects. However, McKusick currently lists cleft lip with or without associated cleft palate (CLP) or only cleft palate (CP) in nearly 100 syndromes each. These do not include the nearly 100 recognized chromosome syndromes. Whereas the causes of clefting are poorly understood, nonsyndromic cleft lip with or without cleft palate (CL/P) and nonsyndromic cleft palate (CPO) are generally regarded to be distinct conditions. This distinction is based on both population and embryologic studies. Epidemiological and observational studies have demonstrated that CL/P and CPO do not usually segregate together in the same families. Animal studies have shown that neural crest cells migrate into the region that eventually becomes the facial structures, and these cells play an important role in the skeletal, connective, and dental tissues of facial morphogenesis. The anterior neuropore closes at the fourth gestational week, and there is active growth during the fifth and sixth weeks. Maxillary swellings arise and enlarge through ectomesenchymal proliferation, and these swellings become the anterior portion of the first pharyngeal arch. The medial nasal prominences merge with each other and the bilateral maxillary processes. The upper lip is formed laterally by the maxillary prominences and medially by fused nasal prominences by the end of the sixth week. A defect in any of the developmental steps could lead to a cleft lip. The primary palate consists of two merged medial nasal prominences called the intermaxillary segment. This segment is made up of two portions, the labial component, which becomes the philtrum and the triangular palatal component of bone that includes four maxillary incisor teeth. The secondary palate includes 90% of the hard and soft palate but does not include the anterior portion

From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

that holds the incisor teeth. The palatal shelves originate from swellings on the medial surface of the maxillary prominences that appear during the sixth week and grow downward and lateral to the tongue. Elevation of the palatal processes to a horizontal plane begins during the seventh week, and closure occurs by the ninth week. Closure occurs by fusion and is slightly later in males than females. Clefts of the secondary palate may result from hypoplasia of the shelves, delayed timing of the shelf fusion, interruption of normal fusion, or a problem in programmed cell death. CL/P and CPO are complex malformations that require a multidisciplinary approach with evaluations by plastic surgery, otolaryngology, genetics, speech pathology, dental, and dietary disciplines. Practitioners representing these disciplines usually comprise the craniofacial teams that provide care for these children. Multiple surgeries are often required for lip and palate repair. Recurrent otitis media is a common complication, often requiring placement of PE tubes, especially when there is palatal involvement. Dental, especially orthodontic, intervention, is required for most of the children, and speech therapy is commonly necessary because of nasal speech patterns. Finally, genetic evaluation and counseling provides information pertaining to correct diagnosis and recurrence risks for future children. The following discussion will focus on some common causes and research pertaining to the etiology of clefting.

NONSYNDROMIC CLEFTING NONSYNDROMIC CLEFT LIP AND PALATE (CL/P) CL/P occurs in approximately 1/1000 live births, and males are affected approximately twice as frequently as females. CL/P is unilateral in 80% of cases and bilateral in the remaining 20% (Fig. 121-1). The left side is affected in 70% of the unilateral cases; 85% of the bilateral and 70% of the unilateral cases are associated with cleft of the anterior palate. Etiologic studies of CL/P are complicated by the difficulties in sorting out the inheritance patterns. Familial recurrences are welldocumented but do not follow a Mendelian pattern of inheritance. The multifactorial model was developed to explain the familial recurrences, and the recurrence risk estimates used in genetic counseling practice are based on this model. For isolated cases, a recurrence risk of 3–5% for subsequent pregnancies is given, but higher risks of 10% are quoted if there are other affected first-degree relatives. While the multifactorial model has been useful, it has not provided a completely satisfactory explanation for the observed familial cases. Indeed, complex segregation analyses of different populations have suggested a Mendelian contribution with auto-

1087

1088

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Figure 121-1

Infant with bilateral CL/P.

somal-dominant and/or recessive genes that are influenced by other genes, sex, and environmental factors (Table 121-1). Other studies suggest that oligogenic inheritance with as many as six to eight major genes may play a role in the development of CL/P. Growth factors are an example of multiple interacting genes that may play a role in orofacial clefting and are discussed below. Thus CL/P appears to be a genetically heterogeneous condition with multiple factors impacting the final developmental outcome. Identification of these genetic factors has proven to be a challenge. Two approaches, association and linkage studies, using candidates genes and random mapped polymorphic markers, have been undertaken to study the genetic basis of CL/P. These approaches have been successfully used to delineate the molecular causes of other multifactorial conditions such as Hirschsprung disease and diabetes. The two statistical methods of studying clefting, linkage and association studies, each has strengths and weaknesses. Linkage testing requires multiplex families, generally with at least three affected offspring, and some assumptions regarding the mode of inheritance must be made. With these criteria fulfilled, it is possible to perform genome-wide screening to look for linkage. In some cases, a single family may be sufficient to establish linkage, although this has not been the case with clefting, as there are generally only a few affected individuals in a pedigree. Association studies, on the other hand, make no assumptions about the mode of inheritance. They require only a single affected individual from a family, but large numbers are required. A control population is also needed, and special attention must be paid to ensuring that the allele frequencies being used are correct. Association studies are looking for a specific allele of a gene to be associated with the

disease, so that random testing is not practical, but testing of candidate genes is. Finally, association studies are generally better at detecting modifier loci than a linkage study. Given the apparent complex nature of clefting, it is unlikely that a single method will succeed in elucidating the etiologies of clefting. Candidate genes have been identified in human and/or murine embryology studies. For example, transforming growth factor-α (TGF-α) is expressed in the pharyngeal pouch and activates epidermal growth factor (EGF/TGFA) receptors. It has been suggested that cell–cell contacts during embryogenesis may be discrete, and sequential events are mediated by growth factor receptor transduction. For this reason, all of the growth factors— such as retinoic acid receptor-α (RAR-α), TGFs, EGF, and so on— are excellent candidates and many of these have been tested. The results of both association and linkage studies are summarized in Table 121-2. Association studies from different populations have suggested that TGF-α plays a role in the development of CL/ P. However, two linkage studies have demonstrated that TGF-α does not segregate with the disease phenotype in 20 multigenerational families. TGF-α is now considered to have a modifying effect on the development of CL/P. Two candidate genes for CL/P, RARA-α and MSX1 (Hox 7), have been identified in murine models. MSX1 knockout mice exhibit craniofacial and tooth abnormalities in addition to cleft palate, and a CL/P gene has been mapped to the mouse chromosomal region containing RAR-α. Association of RAR-α with CL/ P has been demonstrated in two population studies but not in a third, and linkage was excluded to RAR-α in ten multigenerational CL/P families. These results suggest that RAR-α does not play a major role in CL/P. MSX1 was excluded in a CLP linkage study, but the results of a recent association study suggest this locus may play a role in CPO. Linkage studies have suggested there may be CL/P loci on chromosomes 6p, 4p, and 19q. A locus on chromosome 6p was first suggested by Eiberg et al. in 1987, when evidence for linkage to Factor 13A (F13A) at 6p23 was found. Additionally, linkage to F13A was demonstrated in 14 of 21 families from Italy, and chromosomal rearrangements involving 6p in individuals with CL/P have been reported. However, linkage to F13A was not found in 20 US and UK CL/P families, and a subsequent study expanded to include 33 Caucasian CL/P families was not able to confirm linkage to the 6p chromosomal area that spans from F13A to D6S89. A LOD (Log odds) score suggestive of linkage to an anonymous marker, D4S192, was reported in a single large family, and an association study provided evidence for linkage disequilibrium with the same marker. These findings could not be substantiated in 33 multigenerational CL/P families. Finally, linkage has recently been reported to BCL3 on chromosome 19q. Seventeen of thirty-nine families were found to have a posteriori probability of greater than 50% of being linked to this chromosomal region. BCL3 is a proto-oncogene, which was the candidate gene in the immediate chromosomal vicinity. Additional studies need to be performed to determine whether mutations can be identified in BCL3 or whether there is another candidate gene in close proximity to BCL3. The families linked to BCL3 have been excluded from the chromosome 4 and 6 markers, providing additional evidence that CL/P is genetically heterogeneous. A great deal of work remains to be done to identify the genetic loci that contribute to human CL/P. This will entail large numbers of families and populations for linkage and

1089

CHAPTER 121 / OROFACIAL CLEFTING

Table 121-1 Segregation Analyses Author(s)/year Chung et al., 1974 Chung et al., 1984 Demanis et al., 1984 Chung et al., 1986 Marazita et al., 1986 Hecht et al., 1991 Marazita et al., 1992 Nemana et al., 1992 Ray et al., 1993

Mode inheritance/model

Population

Multifactorial/AR/major gene Multifactorial/major gene Multifactorial/mixed AR/major gene Multifactorial AR/major gene Major gene AD/major gene AR/major gene Major gene AD/coD/major gene

Hawaiian Chinese Caucasian Caucasian Japanese Caucasian Chinese Caucasian Chinese Indian Indian

Table 121-2 Loci Implicated in Development of CL/P Loci Locus

Method

Results

TGFA

Association Linkage Association Linkage Linkage Linkage Association Linkage Linkage

± – + – ± ± + ± +

RARA F13A D6S89 D4S192 BCL3

association studies, respectively, in order to have sufficient power to identify the different CL/P loci. NONSYNDROMIC CLEFT PALATE (CPO) CPO occurs in 0.4/1000 live births, and females are affected more often than males. CPO most commonly occurs sporadically; however, isolated families with multiple affected family members have been reported. A multifactorial model has also been used to describe recurrence of CPO in families. However, segregation analysis has suggested that a major gene explains half the familial recurrence of CPO. It has also been suggested that an oligogenic model with six loci would explain the data. Interestingly, in a large breeding colony of Brittany spaniels, CPO was found to be segregating as an autosomal-recessive disorder. A chromosomal locus on chromosome 6p has also been implicated in CPO, but has not been confirmed by linkage studies in multigenerational CPO families. Association of TGFA with CPO has been reported in one study. A systematic study of CPO by both association and linkage methods remains to be undertaken. SYNDROMIC CLEFT LIP AND PALATE (CLP) AND CLEFT PALATE (CP) Some of the several 100 syndromes that feature orofacial clefts have been mapped, and the genes responsible are known in some cases (Table 121-3). Here we discuss a few of the syndromes in which orofacial clefts are a prominent finding. Van der Woude syndrome (VWS) is a well-described autosomal-dominant condition associated with CLP and/or CP, cleft uvula, hypodontia, and paramedian lower-lip pits. Lip pits are observed in 65–80% of cases, and surgical removal is recommended as these fistulas may communicate with underlying sali-

vary glands, which can produce watery discharge. Not all gene carriers exhibit lip pits or CLP/CP, and family studies suggest that the penetrance is 75–80%. One to three percent of CLP cases are estimated to be VWS. VWS was mapped to chromosome 1q in 1990 with tight linkage to the renin gene. The chromosomal region containing the VWS gene has since been refined to a 4.1-cM region between microsatellite markers D1S245 and D1S414. Linkage of nonsyndromic CL/P and CPO to VWS was excluded using tightly linked markers. Stickler syndrome (Hereditary Arthro-ophthalmopathy) is an autosomal-dominant condition characterized by flat facies, CP, myopia, retinal detachment, deafness, and epiphyseal abnormalities. Pierre Robin sequence (see Sporadic Conditions discussion) may be the presenting diagnosis in infancy. Stickler syndrome demonstrates marked variability, and individuals may inadvertently be misdiagnosed as CPO. Retinal detachments, myopia, and sensorineural deafness are common complications, and annual ophthalmologic and hearing evaluations are suggested. Arthritis occurs in adulthood and may be progressive with advancing age. Mutations in COL2A1 (12q14.3) and COL11A2 (6p21.2) have been identified in individuals with Stickler syndrome. The mutations in COL2A1 produce stop codons that lead to truncated proteins, whereas the characterized COL11A2 mutation causes a change in the splice donor site, resulting in an in-frame exon skipping. The associated phenotype appears to be milder than that observed for COL2A1 mutations; however, only one case has been reported. An autosomal-recessive mutation in COL11A2 has been reported, and the associated phenotype appears to be more severe than the autosomal-dominant form. Other type II collagenopathies, Kniest dysplasia and spondyloepiphyseal dysplasia, also have CP. An X-linked type of CP (CPX) associated with ankyloglossia (tongue-tie) was first described and characterized in an Icelandic native family and later in a Mennonite native family. Males are affected with CP and tongue-tie, and only rarely will carrier females be similarly affected. Submucous CP and bifid or absent uvula are also found in this condition. The CPX gene was mapped to Xq in a large Icelandic family in 1987 and the map location has been further refined to a 5-cM region within the interval of Xq21.1–q21.31. CHROMOSOMAL CAUSES OF CLEFTING Velocardial facial syndrome (VCF) or Shprintzen syndrome was first described by Shprintzen et al. in 1978 and is characterized by craniofacial,

1090

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Table 121-3 Mapped Clefting Syndromes MIM Cleft lip/palate 109400 129900 305400 161200 263520 312870 313850 119300 194190 193500 Cleft palate 101400 301590 222600 211970 215150 154500 311300 182290 309583 183900 108300 184840 303400 214100 170995 170993

Syndrome

Location

Basal cell nevus syndrome EEC syndrome Faciogenital dysplasia Nail-Patella syndrome Polydactyly with neonatal chondrodystrophy Simpson dysmorphia syndrome Thoracoabdominal syndrome Van der Woude syndrome Wolf-Hirschorn syndrome Waardenburg syndrome, Type I

9q22.3–q31 7q11.2–q21.3 Xp11.21 9q34.1 4q13 Xcen–q21.3 Xq26.1 1q32–q41 4p13.7 2q235

Acrocephalosyndactyly type III Anophthalmos Diastrophic dysplasia Camptomelic dwarfism Chondrodystrophy with sensorineural deafness Mandibulofacial dysostosis Otopalatodigital syndrome Smith-Magenis syndrome Snyder-Robinson syndrome Spondyloepiphyseal dysplasia, congenita Stickler syndrome, type 1 Stickler syndrome, type 2 X-linked cleft palate Zellweger syndrome 1 Zellweger syndrome-2 Zellweger syndrome-3

7p21.3–p21.2 Xq27–q28 5q31–17q24.2q33 17q24.2 6p22–p21.3 5q31.3–q32 Xq28 17p11.2 Xp21 12q13.11 12q13.11 6p22– Xq13–q21 7q11 1p22–p21 8q21.1

cardiac, and palatal abnormalities. It is estimated to be the most common syndrome featuring CP, having a prevalence of 1/5000 live births. The facies are distinctive with malar hypoplasia, prominent nose with a squared nasal root and narrow alar base and CP. Hypotonia, failure to thrive, and short stature are frequent in infancy and childhood. Cardiac anomalies include ventricular septal defect, right aortic arch, tetralogy of fallot, and aberrant left subclavian artery. Robin sequence is occasionally observed. Learning disabilities and impairment are consistent findings. Speech development is delayed and hypernasality is common. Phenotypic findings can overlap with the DiGeorge syndrome, which also includes thymic aplasia/hypoplasia and hypocalcemia. Most cases of VCF syndrome are sporadic, although familial cases demonstrating autosomal-dominant inheritance have been reported. Cytogenetic abnormalities involving 22q11 are detected in approximately 20% of VCF syndrome patients. Deletion of molecular markers from this chromosomal region have recently been identified in 82% of VCF patients. No difference in the sex of the parent transmitting the chromosomal deletion was found among the patients with deletions and on whom DNA was available suggesting that imprinting does not play a major role in the etiology of this syndrome. Further discussion of VCF can be found in Chapter 120. Trisomy 13 is a common chromosomal abnormality, with a prevalence of 1/5000 live births. Microcephaly, secondary to the holoprosencephaly spectrum of brain abnormalities, microphthalmia, retinal dysplasia, and polydactyly of the hands and sometimes in the feet are common. CLP is present in about 60–80% of cases. Cardiac abnormalities are identified in 80% of cases and

Locus

FGD1

PAX3

DTD SOX9 COL11A2

COL2A1 COL2A1 COL11A2 PXMP1 PAF1

include ventricular septal defect, patent ductus arteriosus, and atrial septal defect. The constellation of phenotypic findings is caused by the presence of an extra chromosome 13, either as a free trisomy or, rarely, as a translocation. The risk of recurrence for trisomy 13 is less than 1% except where maternal age is a factor, which elevates the risk. Translocations are associated with an empiric risk of recurrence of 5–10%, depending on the sex of the parent who is the translocation carrier. Trisomy 13 is generally lethal by 1 year of age with only a few long-term survivors reported. TERATOGENIC CAUSES OF CLEFTING Fetal Hydantoin syndrome is a constellation of phenotypic findings associated with in utero exposure to Dilantin or phenytoin. These findings include unusual facies, digit and nail hypoplasia, and growth and mental deficiencies. CLP is found approximately 11 times more frequently in exposed offspring compared with those not exposed. Deficiency of folic acid during pregnancy may play an etiologic role in some cases of CL/P and CPO. A recent study has demonstrated that supplementation during pregnancy decreased the birth prevalence of orofacial clefting. SPORADIC CONDITIONS WITH CLEFTING Pierre Robin sequence is characterized by micrognathia, glossoptosis, and a U-shaped CP. Posterior airway obstruction may occur secondary to the small mandible and tongue obstruction. Hypoplasia of the mandibular area prior to 9 weeks’ gestation is postulated to be an initiating defect in this condition. The tongue becomes posteriorly displaced and interferes with the normal movement and fusion of the palatal shelves, causing the large U-shaped cleft compared with the V-shaped palate defect observed in CPO. Deformational processes have also been implicated in some cases.

CHAPTER 121 / OROFACIAL CLEFTING

The etiology and natural history depends on whether there are other associated anomalies or whether it is part of a syndrome as observed in the Stickler syndrome (see above). Facio-auriculo-vertebral spectrum (Goldenhar/Hemifacial microsomia) is a nonrandom association of anomalies involving the first and second branchial arches. The prevalence is estimated to be 1/3000–5000 live births, and males are affected more often than females. The spectrum includes unilateral hypoplasia malar, maxillary and mandibular regions facial musculature, vertebral and ocular abnormalities, with epibulbar dermoid being the most common eye finding. The clefts most often associated with this spectrum are lateral-like extensions from the corner of the mouth, although CLP and CP have been observed. Unilateral microtia with accessory pits and tags are frequent and may be associated with deafness. The soft palate may show poor resilience and lead to speech problems. Hemivertebrae, most commonly in the cervical spine, are often present, and cardiac and renal anomalies may occur. The etiology is unknown, and there is a very low recurrence risk for other family members.

CONCLUSION As illustrated, CLP and CP are associated with a wide range of syndromes whose underlying etiologies are diverse. Although not all of the known genes for syndromic forms of CLP and CP have been tested for linkage to the nonsyndromic forms, most of them have been excluded from playing a major role in the development of nonsyndromic CL/P and CPO. Given the complex nature of craniofacial embryology, it is not surprising that many different genes may exert an effect.

SELECTED REFERENCES Ahmad NN, Ala-Kokko L, Knowlton RG, et al. Stop codon in the procollagen II gene (COL2A1) in a family with the Stickler syndrome (arthro-ophthalmopathy). Proc Natl Acad Sci USA 1991;88: 6624–6627. Ardinger HH, Buetow KH, Bell GI, Bardach J, VanDemark DR, Murray JC. Association of genetic variation of the transforming growth factor alpha gene with cleft lip and palate. Am J Hum Genet 1989;45: 348–359. Asada H, Kawamura Y, Maruyama K, et al. Cleft palate and decreased brain gamma-aminobutyric acid in mice lacking the 67-kDa isoform of glutamic acid decarboxylase. Proc Natl Acad Sci USA 1997;94: 6496–6499. Beiraghi S, Foroud T, Diouhy S, et al. Possible location of a major gene for cleft lip and palate to 4q. Clin Genet 1994;46:255,256. Blanton SH, Crowder E, Malcolm S, et al. Exclusion of linkage between cleft lip with or without cleft palate and markers on chromosomes 4 and 6. Am J Hum Genet 1996;58:239–241. Carinci F, Pezzetti F, Scapoli L, et al. Nonsyndromic cleft lip and palate: evidence of linkage to a microsatellite marker on 6p23. Am J Hum Genet 1995;56:337–339. Carter CO. Genetics of common disorders. Br Med Bull 1969;25:52–57. Carter CO, Evans K, Coffey R, Fraser Roberts JA, Buck A, Fraser Roberts M. A three generation family study of cleft lip with or without cleft palate. J Med Genet 1982;19:245–261. Chenevix-Trench G, Jones K, Green A, Martin N. Further evidence for an association between genetic variation in transforming growth factor alpha and cleft lip and palate. Am J Hum Genet 1991;48:1012,1013. Chenevix-Trench G, Jones K, Green AC, Duffy DL, Martin NG. Cleft lip with or without cleft palate: associations with transforming growth factor alpha and retinoic acid receptor loci. Am J Hum Genet 1992;51:1377–1385. Chung CS, Ching GHS, Morton NE. A genetic study of cleft lip and palate in Hawaii. II Complex segregation analysis and genetic risks. Am J Human Genet 1974;26:177–188.

1091

Chung CS, Bixler D, Watanabe T, Koguchi H, Fogh-Anderson P. Segregation analysis of cleft lip with or without cleft palate: a comparison of Danish and Japanese data. Am J Hum Genet 1986;39:603–611. Davies AF, Stephens RJ, Olavesen MG, et al. Evidence of a locus for orofacial clefting on human chromosome 6p24 and STS content map of the region. Hum Mol Genet 1995;4:121–128. Demenais F, Bonaiti-Pellie C, Briard ML, Feingold J. An epidemiological and genetic study of facial clefting in France. II Segregation analysis. J Med Genet 1984;21:436–440. Eiberg H, Bixler D, Nielsen LS, Conneally PM, Mohr J. Suggestion of linkage of a major locus for nonsyndromic orofacial cleft with F13A and tentative assignment to chromosome 6. Clin Genet 1987;32: 129–132. Farrall M, Holder S. Familial recurrence-pattern analysis of cleft lip with or without cleft palate. Am J Hum Genet 1992;50:270–277. Feng H, Sassani R, Bartlett SP, et al. Evidence from family studies for linkage disequilibrium between TGFA and a gene for nonsyndromic cleft lip with or without cleft palate. Am J Hum Genet 1994;55: 932–936. Ferguson MW. Palate development: mechanisms and malformations. Isr J Med Sci 1987;23:309–315. Field LL, Ray AK, Marazita ML. Transforming growth factor alpha: A modifying locus for nonsyndromic cleft lip with or without cleft palate? Eur J Hum Genet 1994;2:159–165. Fitzpatrick D, Farrell M. An estimation of the number of susceptibility loci for isolated cleft palate. J Craniofac Genet Dev Biol 1993;13: 230–235. Gorlin RJ. Syndromes of the Head and Neck. New York: Oxford University Press, 1990. Gorski SM, Adams KJ, Birch PH, Chodirker BN, Greenberg CR, Goodfellow PJ. Linkage analysis of X-linked cleft palate and ankyloglossia in Manitoba Mennonite and British Columbia native kindreds. Hum Genet 1994;94:141–148. Hanson JW. Risk of the offspring of women treated with hydantoin anticonvulsant with emphasis on the fetal hydantoin syndrome. J Pediat 1989;4:662–668. Healey SC, Mitchell LE, Chenevix-Trench G. Evidence for an association between non-syndromic cleft lip with or without cleft plate and a gene located on the long arm of chromosome 4. Am J Hum Genet 1994;55:A47. Hecht JT, Annegers JF. Familial Component of Epilepsy in Cleft Lip and Palate. Ann Arbor, MI: Univer Microfilms Int, 1988. Hecht JT, Wang Y, Blanton SH, Michels VV, Daiger SP. Cleft lip and palate: no evidence of linkage to transforming growth factor alpha. Am J Hum Genet 1991;49:682–686. Hecht JT, Wang Y, Connor B, Blanton SH, Daiger SP. Nonsyndromic cleft lip and palate: no evidence of linkage to HLA or factor 13A. Am J Hum Genet 1993;52:1230–1233. Hecht JT, Yang P, Michels VV, Buetow KH. Complex segregation analysis of nonsyndromic cleft lip and palate. Am J Hum Genet 1992a;49:674–681. Hecht JT, Wang Y, Blanton SH, Michels VV, Daiger SP. Cleft lip and palate: no evidence of linkage to transforming growth factor alpha. Am J Hum Genet 1991;49:682–686. Holder SE, Vintiner GM, Farren B, Malcolm S, Winter RM. Confirmation of an association between RFLPs at the transforming growth factor-alpha locus and non-syndromic cleft lip and palate. J Med Genet 1992;29:390–392. Jara L, Blanco R, Chiffelle I, Palomino H, Carreno H. Evidence for an association between RFLPs at the transforming growth factor alpha (locus) and nonsyndromic cleft lip/palate in a South American population. Am J Hum Genet 1995;56:339–341. Jones KL. Smith’s Recognizable Patterns of Human Malformations, 4th ed. Philadelphia: WB Saunders, 1988. Juriloff DM, Mah DG. The major locus for multifactorial nonsyndromic cleft lip maps to mouse chromosome 11. Mammalian Genome 1995, 6:63–69. Lidral A, Basart A, Romitti P, et al. Candidate gene analysis of nonsyndromic cleft lip with or without cleft palate (NS-CL/P) and cleft palate (NS-CPO) in humans. Am J Hum Genet 1995;57(Abstract 58).

1092

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Marazita ML, Hu D-N, Spence MA, Liu Y-E, Melnick M. Cleft lip with or without cleft palate in Shanghai, China: evidence for an autosomal major locus. Am J Hum Genet 1992;51:648–653. Marazita ML, Spence MA, Melnick M. Major gene determinationof liability to cleft lip with or without cleft palate: a multiracial view. J Craniofacial Genet Dev Biol 1986;2:89–97. Mitchell LE. Transforming growth factor alpha locus and nonsyndromic cleft lip with or without cleft palate: a reappraisal. Genet Epidemiol 1997;14:231–240. Mitchell LE, Risch N. Mode of inheritance of nonsyndromic cleft lip with or without cleft palate: a reanalysis. Am J Hum Genet 1992;51:323–332. Moore GE, Ivens A, Chambers J, et al. Linkage of an X-chromosome cleft palate gene. Nature 1987;326:91,92. Moore K. The Developing Human: Clinically Oriented Embryology, 4th ed. Philadelphia: WB Saunders, 1988. Morrow B, Goldberg R, Carlson C, et al. Molecular definition of the 22q11 deletions in velo-cardio-facial syndrome. Am J Hum Genet 1995;56:1391–1403. Murray JC, Nishimura DY, Buetow KH, et al. Linkage of an autosomal dominant clefting syndrome (Van der Woude) to loci on chromosome 1q. Am J Hum Genet 1990;46:486–491. Nemana LJ, Marazita ML, Melnick M. Genetic analysis of cleft lip with or without cleft papate in madras, India. Am J Med Genet 1992;42:5–9. Oliver-Padilla, Martinez-Gonzalez. Cleft lip and palate in Puerto Rico: a thirty-three year study. Cleft Palate J 1986;23:48–57. Online Mendelian Inheritance in Man, OMIM (TM). The Human Genome Data Base Project, Johns Hopkins University, Baltimore, MD. World Wide Web , 1995. POSSUM, V4.0. Computer Power Group and The Murdoch Institure for Research into Birth Defects. Melbourne, Australia. Ray AK, Field LL, Marazita ML. Nonsyndromic cleft lip and palate with or without cleft palate in west Bengal, India: evidence for an autosomal major locus. Am J Hum Genet 1993;52:1006–1011. Richtsmeier JT, Sack GH JR, Grausz HM, Cork LC. Cleft palate with autosomal recessive transmission in Brittany spaniels. Cleft PalateCraniofacial J 1994;31:364–371. Sanford LP, Ormsby I, Gittenberger-de Groot AC, et al. TGFbeta2 knockout mice have multiple developmental defects that are non-overlap-

ping with other TGFbeta knockout phenotypes. Development 1997; 124:2659–2670. Sassani R, Bartlett SP, Feng H, et al. Association between alleles of the transforming growth factor-alpha locus and the occurrence of cleft lip. Am J Med Genet 1993;45:565–569. Satokata I, Maas R. MSX1 deficient mice exhibit cleft palate and abnormalities of craniofacial and tooth development. Nat Genet 1994; 6:348–355. Shiang R, Lidral AC, Ardinger HH, et al. Association of transforming growth factor alpha gene polymorphisms with isolated non-syndromic cleft palate (CPO). Am J Hum Genet 1993;53:836–843. Shprintzen RJ, Goldberg RJ, Young D, Wolford L. The velo-cardio-facial syndrome: a clinical and genetic analysis. Pediatr 1981;67:167–172. Spranger J, Winterpacht A, Zabel B. The type II collagenopathies: a spectrum of chondrodysplasias. Eur J Pediatr 1994;153:56–65. Stein J, Hecht JT, Blanton SH. Exclusion of retinoic acid receptor and a cartilage matrix protein in non-syndromic CL(P) families. J Med Genet 1995a;32:78. Stein J, Mulliken JB, Stal S, et al. Nonsyndromic cleft lip with or without cleft palate: evidence of linkage to BCL3 in 17 multigenerational families. Am J Hum Genet 1995b;57:257–272. Stoll C, Qian JF, Feingold J, Sauvage P, May E. Genetic variation in transforming growth factor alpha: possible association of BamH1 polymorphism with bilateral sporadic cleft lip and palate. Am J Hum Genet 1992;50:870,871. Tolarov M, Harris J. Reduced recurrence of orofacial clefts after periconceptional supplementation with high-dose folic acid and vitamins. Teratology 1995;51:71–78. Vikkula M, Mariman EC, Lui VCH, et al. Autosomal dominant and recessive osteochondrodysplasias associated with the COL11A2 locus. Cell 1995;80:431–437. Vintiner GM, Holder SE, Winter RM, Malcolm S. No evidence of linkage between the transforming growth factor-alpha gene in families with apparently autosomal dominant inheritance of cleft lip and palate. J Med Genet 1992;29:393–397. Wyszynski DF, Mestri N,. McIntosh I, et al. Evidence for an association between markers on chromosome 19q and non-syndromic cleft lip with or without cleft palate in two groups of multiplex families. Hum Genet 1997;1:22–26.

CHAPTER 122 / MOLECULAR GENETICS OF HEARING DISORDERS

122

1093

Molecular Genetics of Hearing Disorders WILLIAM J. KIMBERLING

BACKGROUND Hearing is the result of a mechanical transduction process that translates sound waves into neural signals, which are then interpreted by our minds as language, music, signals of danger, or just plain noise. Hearing is a critical sense for communication, and whenever hearing is impaired, oral communication suffers, and one’s quality of life is considerably diminished. The cochlea is the sensory organ responsible for hearing; it is complex and hundreds of genes must be responsible for its development and normal homeostasis. Thus, it is no surprise that hearing loss occurs in a great many different genetic disorders. The three major parts of the ear are the outer, middle and inner ear. The inner ear develops from the otic placode, whereas the middle and outer ear develop primarily as a result of differentiation of the first and second branchial arches. The inner ear is divided into two parts: the membranous labyrinth (balance) and osseous labyrinth or cochlea (hearing). While both structures share certain features, they are quite different in function. Consequently, each part of the ear must be controlled by different sets of genes, but their effects must be well coordinated in order to produce a single effective hearing organ. As expected, some genes are expressed in more than one part of the ear, whereas others are specific. For example, some hearing loss disorders have vestibular symptoms and others do not; some disorders involve neurosensory hearing loss in combination with conductive loss resulting from external and/or middle ear abnormalities, and others have only a neurosensory component. The clinician must be cognizant of the fact that the ear comprises a set of embryologically related but distinct structures, and that patterns of involvement of the parts of the hearing organ are often important clues in the differential diagnosis of hearing disorders. Hearing loss disorders are divided into syndromic and nonsyndromic categories, depending on whether there are any associated nonotologic symptoms accompanying the hearing deficit. While most childhood deafness is nonsyndromic, advances in the molecular genetics of the auditory system have occurred more with syndromic disorders because they can be recognized as distinct groups and are more amenable to positional cloning strategies. Epidemiological studies to determine the magnitude of hereditary vs nongenetic causes of hearing loss are not common, especially when adult-onset hearing impairment is being considered.

From: Principles of Molecular Medicine (J. L. Jameson, ed.), ©1998 Humana Press Inc., Totowa, NJ.

The role of heredity in causing hearing impairment in children is better defined, and most studies indicate that the overall frequency of deafness in children is about 1/1000. Over one-half of all causes of childhood deafness can be identified as genetic in origin. Information about the role of heredity in adult-onset hearing loss is virtually nonexistent, although one would predict that genes play a major role; the fact that several adult-onset progressive hearing loss genes have been localized supports this hypothesis.

DIAGNOSIS OF HEARING IMPAIRMENT The diagnosis of hearing impairment is aided by the consideration of a variety of symptoms that permit the subcatergorization of hearing loss disorders into several different groups. The list below gives some of the clinical parameters and tests that are frequently helpful in diagnosing the different forms of hearing impairment: 1. Mode of inheritance (dominant, recessive, x-linked, mitochondrial, not inherited). 2. Age of onset (congenital, early childhood, young adult, and so on). 3. Severity (mild to profound). 4. Ears involved (bilateral vs unilateral or symmetric vs asymmetric). 5. Type of hearing loss (sensorineural, conductive, or mixed). 6. Frequencies involved and audiologic profile (e.g., high frequency, low frequency, U-shaped, and so on). 7. Presence and degree of vestibular deficit. 8. Presence of external-, middle-, or inner-ear malformation. 9. Evaluation for common syndromes: electroretinogram (Usher syndrome), thyroid function and perchlorate discharge test (Pendred syndrome), electrocardiogram (Jervell and Lange-Nielsen syndrome), renal function tests (Alport syndrome). The audiogram defines the severity of the hearing loss and is an indispensable component to the clinical evaluation of patients and families with hereditary types of hearing impairment. Some families have distinct audiometric patterns, whereas other families show considerable variation in hearing acuity. Examples of selected different audiologic profiles are presented in Figure 122-1. In some disorders, such as Usher Type II, the audiologic profile is quite consistent across families, whereas in other disorders, like Branchio-oto-renal syndrome, the audiological phenotype can vary considerably even between ears of the same person. Hearing can also be assessed using auditory brain-stem responses (ABR)

1093

1094

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Figure 122-1 This presents examples of different types of audiograms seen in a few hereditary disorders. The circles and triangles indicate the hearing loss observed in right and left ears, respectively. (A) Sloping audiogram characteristic of Usher Type II; the hearing loss is mild in the low frequencies and tapers to a severe-to-profound loss in the high frequencies. (B) Two audiograms at different ages and illustrates a configuration and progression typical of dominantly inherited high-frequency hearing loss. (C) U-shaped audiogram where the loss of hearing acuity is mostly in the middle frequencies. (D) Lowfrequency type of hearing impairment. Such audiometric configurations are often helpful in diagnosing specific inherited forms of deafness.

and oto-acoustic emissions (OAE). The ABR is useful for separating a central from a peripheral (i.e., cochlear) hearing problem, and the OAE assesses hair cell function. The age at onset and nature of the progression of the hearing loss are also important variables in characterizing a family’s phenotype. Some dominant progressive hearing losses involve a specific range of frequencies early in life, and only at older ages are all the frequencies involved. It is frequently necessary to ask for older audiograms to document the type hearing loss at an earlier age. The role of the inner ear for balance is often forgotten by geneticists, and the presence or absence of vestibular symptoms can often be important for the differential diagnosis. For example, Usher syndrome type I has vestibular areflexia, whereas the milder type II does not. Inner-ear malformations are associated with some syndromes and not others. For example, Pendred syndrome often has a Mondini deformity. The audiogram, vestibular studies, and CT of the temporal bone are essential for giving a full description of the hearing loss phenotype.

CLINICAL FEATURES OF DISORDERS WITH HEARING LOSS NONSYNDROMIC HEARING LOSS Dominant Nonsyndromic Hearing Loss The identification of the genes responsible for nonsyndromic hearing loss has been difficult and slow.

The traditional approach to positional cloning for dominantly inherited disorders relies on having large, multigenerational kindreds. When families are large enough, evidence of linkage can be found without having to pool information from several different families, thus avoiding the problem of heterogeneity. Unfortunately, this research strategy is difficult to apply to dominant congenital profound deafness. Congenitally deaf people belong to a separate subculture and will tend to marry other deaf people, thus complicating attempts to follow gene transmission. Because of this phenomenon, the initial focus of localization of nonsyndromic dominant genes has been directed toward dominantly inherited progressive hearing loss. Because of their later age of onset, members of families with progressive hearing loss seldom see themselves as part of the deaf community and do not marry within it. At least 10 dominant hearing loss genes exist. The first gene for nonsyndromic hearing loss, called DFNA1, was localized to chromosome 5q31 (DFNA is used as code for dominant hearing impairment genes and DFNB for recessive genes; the numerals 1, 2, … are assigned in order of discovery). The first linkage was observed in a large Costa Rican kindred whose members had a rapidly progressing hearing loss with adult onset that initially involved the low frequencies. Replication of this linkage in a second family has not been reported, and so it must be presumed that the gene is uncommon. A second gene was localized to chromosome 13q at or near a nonsyndromic recessive deafness gene (DFNA2). This family had a hearing loss that was evident by 4 years of age. Hearing impairment ranged from 15 to 90 dBHL, and the higher frequencies were more involved than the lower frequencies. The HL was not progressive in most gene carriers. Another dominant progressive hearing loss gene was localized to chromosome 1p32 (DFNA3) in two families, one Indonesian and the other American, both of which showed a progressive loss with onset in the high frequencies. A gene for a progressive hearing loss involving all frequencies was localized to chromosome 19, and a family with high-frequency hearing loss has been linked chromosome 7p15, whereas another with a low-frequency loss was localized to 4p16.3. An up-to-date listing of the new localizations of dominant, recessive, X-linked, and mitochondrial hearing loss genes can be found on the Internet by accessing the Hereditary Hearing Loss Homepage at www.uia.ac.be/u/dnalab/hhh.html . This web site also gives the flanking markers for each linkage as well as the appropriate references citing the original linkage observation. Recessive Nonsyndrome Hearing Loss Autosomal recessive nonsyndromic hearing impairment is responsible for 80% of all the cases of childhood deafness. Most deaf-by-deaf marriages result in hearing offspring, a phenomenon that underscores the fact that several different loci are involved. The number of genes responsible has been variously estimated as at between 7 and 150. Since the recessive nonsyndromic families cannot be differentiated from each other, few useful pools of families are generated for gene localization research. Despite this problem, significant advances have been made into locating recessive genes responsible for nonsyndromic hearing loss, notably by focusing research efforts on population isolates. Three loci for autosomal-recessive congenital profound nonsyndromic hearing loss have been localized. Two are from consanguineous kindreds from Tunisia: One at 13q13 (DFNB1) and a second at 11q13 (DFNB2) in the region homologous to the mouse shaker-1 gene and close to myosin VIIa, the gene for Usher Type Ib. A third gene (DFNB3) for nonsyndromic recessive congenital deafness has been localized to chro-

1095

CHAPTER 122 / MOLECULAR GENETICS OF HEARING DISORDERS

Table 122-1 Usher Syndrome Genotype/Phenotype Relationship Subtype Type I (MIM 276900)

Type II (MIM 276901) Type III aAll

Location

1a 1b

14q 11q

1c 1d 2a 2b 3

11p Unlinked 1q Unlinked 3q

Gene Myosin VIIa

Phenotypea Profound sensory neural hearing loss Vestibular areflexia

Sloping mild-to-profound hearing loss Normal vestibular responses Progressive hearing loss Variable vestibular deficit

have retinitis pigmentosa.

mosome 17p in an Indonesian population. A congenital-recessive severe-to-profound deafness (DFNB4) found in the Druze population has been mapped to 7q13, and consanguineous families from India were used to identify a locus on chromosome 14q. X-Linked Hearing Loss X-linked deafness is a special category of deafness because its pattern of inheritance is distinctive. As a specific cause of hearing impairment, X-linkage is not common, accounting for no more that 2% of all cases. The most common form is X-linked deafness with stapes fixation and perilymphatic gusher (DFN3), which has been localized to Xq13–q21.1. The structural defect is an open connection between the cerebrospinal fluid and the perilymph that causes the “gusher” seen in inner-ear surgery. The human gene, Pou3F4, was found to carry mutations that correlate with the gusher phenotype, suggesting that it and DFN3 are the same. Two other X-linked nonsyndromic genes are know to be located at Xq21(15) and Xp21.2. Mitochondrial Inheritance Defects in the mitochondria affect the cellular respiratory chain and lead to a wide variety of progressive clinical manifestations, namely ataxia, epilepsy, dementia, myopathy, polyneuropathy, retinal pigment anomalies, and cardiomyopathy with conduction anomalies (see Chapter 103). Hearing loss is a regular feature and is often the first clinical symptom of many of the mitochondrial disorders. Some mitochondrial disorders involve a full spectrum of disease, whereas others involve only hearing. For example, one family with sensorineural hearing loss of varying severity but with no other pathological feature had a mutation at nucleotide position 7445, which converted the 3' terminal T residue of tRNA-ser to a C and caused a silent alteration to the stop codon. Another mutation close to the same position was reported in association with Leber’s hereditary optic neuropathy, pointing to the fact that the mitochondrial mutations do not yet sort into neat diagnostic categories. Summary Progress has been made in the localization of genes involved in nonsyndromic hearing loss. However, studies of associated symptoms are still lacking. Presence or absence of vestibular defect, existence of inner-ear structural abnormality, brain stem responses, and the results of oto-acoustic emission studies are infrequently reported. As a consequence, it is not yet possible to determine if different hearing loss genes produce slightly different phenotypes. Undoubtedly, as research progresses, clinical studies will be found that will aid in the differentiation of one nonsyndromic type from another. SYNDROMIC HEARING LOSS There are two syndromes that, because of their high frequency among the congenitally deaf, are especially important to consider when evaluating the deaf

patient. These are Usher syndrome and Waardenburg syndrome. Together, these syndromes comprise the largest proportion of identifiable hereditary deafness. It is important to realize that, because of the late diagnosis of the RP for Usher syndrome and the reduced penetrance for Waardenburg syndrome, many patients with either syndrome may initially be considered to have a nonsyndromic hearing loss. Usher Syndrome Usher syndrome (US) is defined as hearing impairment with retinitis pigmentosa and is estimated to be responsible for about 5% of all of the cases of childhood deafness. It is both clinically and genetically heterogeneous. There are three phenotypic types (Table 122-1). Type I has a congenital profound hearing loss, vestibular areflexia, and retinitis pigmentosa. It has a frequency of 4.4/100,000 people. Type II is milder, showing a stable sloping audiogram (see Fig. 122-1A); patients with US II have reasonably good speech and are often helped by hearing aids. Significantly, they have a normal vestibular response. Type III displays a progressive hearing loss and variable vestibular responses. There are at least five, possibly seven, genes responsible for the different types of Usher syndrome. Table 122-1 outlines the relationship between the five known genes and the three phenotypes as they are currently recognized. The most common types of US are Usher IIa (on chromosome 1q) and Usher Ib (on chromosome 11q). Usher Type III is linked to markers on chromosome 3q and is more common in Finland. In the United States, it was estimated that 88% of all Usher II result from mutations at the USH2a locus on chromosome 1, and the rest were either type III or unlinked to either 1q or 3q markers. Thus, it is likely that many families failing to show linkage to 1q are actually Type III Usher syndrome. However, a very small proportion, between 4 and 6%, are not linked to either the USH2a locus nor to the USH3 locus and may represent a novel type of Usher syndrome Type II. Usher type I also shows considerable genetic heterogeneity. The first localization was to chromosome 14q. This gene was subsequently found to constitute approximately 20–30% of Usher I patients in a worldwide series of patients. Two other Usher I linkages were found, one to 11q (USH1b) and one on 11p (USH1c). The Usher Ic variant appears to be limited mostly to the French Acadian population of southern Louisiana and has not yet been observed outside of descendants of that population. Usher Ib accounts for 70–80% of the Usher cases seen in a large series from the United States and Europe. Myosin VIIa on human chromosome 11q has been recently been identified as the gene responsible for Usher Ib. Myosin VIIa is coded by a large gene of over 110 kb

1096

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

Table 122-2 Locations of Nonsyndromic Hearing Loss Genes Gene IDa Dominant DFNA1 DFNA2 DFNA3 DFNA4 DFNA5 DFNB6 Recessive DFNB1 DFNB2 DFNB3 DFNB4 DFNB5 X-linked DFN1 DFN2 DFN3 aThere

Location

Phenotype

5q31 1p32 13q 19q13 7p15 4p13.6

Low-frequency HL beginning in the second decade progressing to a profound loss involving all frequencies Variably late-onset high-frequency HL with variable rates of progression to a severe-to-profound loss Juvenile-onset high-frequency HL Mild HL across all frequencies progressing to a severe-to-profound loss Adult-onset high-frequency HL Adult-onset low-frequency HL

13q 11q 17p 7q13 14q Xq21 Xp21.2 Xq21

Profound congenital HL Profound, childhood-onset HL Profound congenital HL Severe-to-profound congenital HL Profound congenital HL Congenital Profound HL in males; mild-to-moderate HL in female heterozygotes Mixed conductive and HL; perilymphatic gusher with stapes fixation

is not yet a common consensus for locus nomenclature, and the gene ID given here may differ slightly with some reports in the literature.

Table 122-3 Identified Genes Causing Hearing Impairment in Humans Disorder (MIM no.) Neural crest/pigmentary Waardenburg syndrome Type I (193500) Waardenburg syndrome Type II (193510) Craniofacial Treacher Collins syndrome (154500) Crouzon syndrome (123500) Collagen disorders Alport syndrome (203780) Alport syndrome (301050) Osteogenesis imperfecta (166200) Stickler syndrome (108300) Stickler syndrome (108300) Other disorders Norries disease (310600) Neurofibromatosis Type II (101000) Stapes fixation with gusher (304400)

of genomic sequence whose coding region of about 7.5 kb has now been completely sequenced. Forty-nine exons have been identified, and several mutations occurring primarily in the first 14 exons have been observed. The role of Myosin VIIa in the inner ear and the retina has yet to be elucidated. Waardenburg Syndrome Waardenburg syndrome (WS) is an autosomal-dominant disorder that is recognized by its characteristic signs of pigmentary abnormality in the form of a white forelock, premature graying, and heterochromia iridis (two different colored eyes). About 2% of all children in schools for the deaf have Waardenburg syndrome. Deafness is variable; sometimes it is bilateral, sometimes unilateral, sometimes there is no hearing impairment at all. The syndrome is divided into two types, WS1 and WS2, based on the presence or absence of the dystopia canthorum. WS1 has been shown to be a result of mutations in the PAX3 gene on chromosome 2. The PAX3 gene is a transcription factor and controls the migration of melanocytes from the neural crest. A second gene for WS2 has been localized to chromosome 3 and

Inheritance

Linkage

Gene

AD AD

2q35 3p12

PAX3 MITF

AD AD

5q13 10q25

TCOF1 FGFR2

AR XL AD AD AD

2q35 Xq 17q21 12q12 6p21.3

COL4A3/4 COL4A5 COL1A1 COL2A1 COL11A2

XL AD XL

Xp11.4 22q12.2 Xq13

NDP NF2 POU3F4

has been shown to be homologous to microphthalmia in the mouse. This gene, MITF, is also a transcription factor (see Chapter 112).

COLLAGEN DISORDERS ASSOCIATED WITH HEARING LOSS ALPORT SYNDROME Alport syndrome (AS) is glomerular nephritis with high-frequency progressive sensorineural hearing loss (see Chapter 68). Most cases of AS are inherited in an X-linked pattern, although both autosomal-dominant and recessive patterns of inheritance have been noted in a few families. Typical of X-linked disorders, males are more severely affected than females. X-linked AS is a result of mutations in the collagen 4α5 gene (COL4A5). Collagen 4α5 is a component of the basement membrane of the glomerulus and the cochlea. Mutations in COL4A5 result in a defective collagen that renders the basement membrane susceptible to lamelation and splitting. Presumably, this process leads to the gradual deterioration of the glomerulus and the cochlea, which accounts for the gradual nature of the disease.

CHAPTER 122 / MOLECULAR GENETICS OF HEARING DISORDERS

The range and severity of expression of the Alport gene is more variable between families than within, suggesting that different mutations have different clinical effects. For example, some X-linked Alport cases do not have an associated hearing deficit but others do. No correlation between the site of mutation and the presence of hearing deficit has yet been made, but it is tempting to speculate that some parts of the COL4A5 molecule are more critical for hearing impairment than other regions. A recessive form of Alport syndrome with mutations in the COL3A3 and COL3A4 collagen genes has recently been described. STICKLER SYNDROME Stickler syndrome is a clinically variable disorder associated with arthropathy, ocular symptoms (vitreoretinal degeneration, severe myopia, retinal detachments, and cataracts), hearing impairment, Pierre-Robin sequence, and midface hypoplasia. There is great phenotypic variation both between and within families. Whereas the hearing loss is variable, it seems to involve the high frequencies more often, and no correlation between the hearing loss and degree of orofacial involvement has been noted. Stickler syndrome was found to result from mutations in the COL2A1 gene. The type of COL2A1 mutation may be related to the severity of the Stickler phenotype. Mutations causing premature termination of transcription of the COL2A1 gene are associated with the “typical” Stickler phenotype, which includes hearing loss and facial hypoplasia, while mutations replace glycine codons with codons for bulkier amino acid appear to produce phenotypes of lethal chondrodysplasia or with only ocular problems. About 50% of Stickler syndrome families fail to show linkage to COL2A1, indicating the existence of a second type of Stickler syndrome from a mutation in another gene that may be located on chromosome 6p22–6p21.3. OSTEOGENESIS IMPERFECTA Osteogenesis Imperfecta refers to several different heritable disorders characterized by bone fragility. Deafness is a frequent feature of both Types I and IV. Type I is a result of mutation in the COL1A1 gene on chromosome 17q21, and Type IV is a result of mutations in the COL1A2 gene on chromosome 7. About 50% of all Type I patients have hearing loss. Type IV (COL1A2) is also associated with hearing loss but only in about 30% of the cases.

CRANIOFACIAL DISORDERS BRANCHIO-OTO-RENAL SYNDROME Branchio-otorenal syndrome (BOR) is characterized by renal anomaly, external ear anomaly, and mixed hearing impairment. The hearing impairment can range from none to severe. It may be sensorineural (20%), conductive (30%), or mixed (50%). The hearing loss is quite variable between individuals and even between the ears of the same person. The inner-ear anomalies include malformations of the middle ear, cochlear hypoplasia, a wide internal auditory canal, and malformation of the vestibular system. The gene for BOR has been localized to chromosome 8q, but has not yet been identified. CROUZON SYNDROME Crouzon syndrome is one of the most distinctive and frequent forms of dominantly inherited craniosynostosis. Fifty-five percent of Crouzon patients have a conductive hearing loss and 13% have atresia of the external auditory canals. Crouzon syndrome has been mapped to chromosome 10q25–26 and found to result from mutations in the fibroblast growth factor receptor-2 gene (see Chapter 114). TREACHER COLLINS SYNDROME (MANDIBULOFACIAL DYSOSTOSIS) Treacher Collins syndrome shows significant craniofacial dysmorphology. The phenotype is characterized by down-slanted palpebral fissures with a coloboma on the outer

1097

supraorbital rims, and zygomas give the face a narrow appearance. Cleft palate is seen in about 35% of the cases. The external auricle is often malformed. The ossicles as well as the cochlea and labyrinth have been observed to be absent or hypoplastic. Bilateral hearing loss is seen in over half of the cases. The gene (PTX-2) has been mapped 5q13–5q33.3 near the gene for DFNA1.

OTHER DISORDERS ASSOCIATED WITH HEARING LOSS PENDRED SYNDROME Pendred Syndrome is an association between deafness and euthyroid goiter (see Chapter 50). The hearing loss may be progressive. The cause of the hearing defect is a congenital bilateral malformation of the cochlea of the Mondini type. A patient with multiple congenital anomalies and Pendred syndrome was observed to have a chromosomal abnormality, dup(10p)del(8q), suggesting that a gene for Pendred was on one or the other of these chromosomes. NEUROFIBROMATOSIS 2 Neurofibromatosis Type 2 (NF2) is an autosomal-dominant disorder characterized by the development of bilateral schwannomas from the vestibular nerves (see Chapter 106). Gene carriers often show eighth nerve dysfunction starting in early childhood and is manifest by tinnitus, bilateral hearing loss, and vestibular dysfunction. NF2 is always important to consider in the differential diagnosis of patients with sudden onset hearing loss, especially when unilateral. Gene linkage analysis has localized the NF2 gene to chromosome band 22q2. The NF2 gene has recently been identified and encodes a protein that has a high degree of homology with moesin, ezrin, and radixin, a family of proteins probably responsible for linking the cell membrane with the cytoskeleton. The mechanism by which mutations of this gene act to cause tumor growth is not known. NORRIES DISEASE Norries Disease (ND) is an X-linked disorder characterized by progressive atrophy of the eyes, mental disturbances, and deafness. It has been mapped to chromosome Xp11.4, and its gene spans 28 kb and consists of three exons. Detailed molecular analyses of genomic deletions in Norrie patients shows that they are heterogeneous, both in size of material deleted and in position of the deletions. It has been hypothesized that, because of homology with known proteins, that norin codes for a protein that regulates neural cell differentiation and proliferation.

MANAGEMENT AND TREATMENT OF HEREDITARY HEARING IMPAIRMENTS From a genetic perspective, diagnosis is the main problem encountered when dealing with hereditary hearing loss disorders and must be solved before other issues of management can be effectively addressed. Advances in molecular diagnostics will undoubtedly prove to be valuable in this area. Even though a limited number of syndromic genes have been identified so far, it is possible to verify diagnosis for Waardenburg syndrome, one type of Usher syndrome, Alport syndrome, X-linked gusher, and Norries disease. Together, these syndromes constitute a large proportion of patients seen for hearing impairments. With regard to nonsyndrome hearing losses, several linkages have been found, but none of the specific genes responsible have been identified. Improved efficiency in diagnostic screening is certainly to be expected as more genes are identified, and it is reasonable to expect that such screening will become increasingly important in other aspects of management. Early diagnosis is critical for the development of communication skills in children with hearing losses. As

1098

SECTION XIII / GENETIC BASIS OF CONGENITAL MALFORMATIONS

more mutations are identified, infants in at-risk families can be diagnosed as early as desired. Early diagnosis may be important with regard to other strategies of intervention. For example, Usher I patients seem to do well with cochlear implants and early diagnosis would be useful in planning patient care that could lead to implant surgery done at a stage of development most effective for enhancing the language skill of Usher patients. Thus, it is expected that, as molecular genetics improves the diagnostic skills for the detection of specific genetic mutations, novel approaches to the treatment of the hearing-impaired will emerge.

SELECTED REFERENCES Barker DF, Hostikka SL, Zhou J, et al. Identification of mutations in the COL4A5 collagen gene in Alport syndrome. Science 1990;248:1224–1227. Chen ZY, Battinelli EM, Hendriks RW, et al. Norrie disease gene: characterization of deletions and possible function. Genomics 1993;16: 533–535. de Kok YJ, van der Maarel SM, Bitner-Glindzicz M, et al. Association between x-linked mixed deafness and mutations in the POU domain gene POU3F4. Science 1995;267:685–688. Fontaine B, Rouleau GA, Seizinger BR, et al. Molecular genetics of neurofibromatosis 2 and related tumors (acoustic neuroma and meningioma). Ann NY Acad Sci 1991;615:338–343.

Fraser FC, Ling D, Clogg D, Nogrady B. Genetic aspects of the BOR syndrome—branchial fistulas, ear pits, hearing loss, and renal anomalies. Am J Med Genet 1978;2:241–252. Garretsen TJ, Cremers CW. Clinical and genetic aspects in autosomal dominant inherited osteogenesis imperfecta type I. Ann NY Acad Sci 1991;630:240–248. Gorlin RJ, Toriello HV, Cohen MM. Hereditary Hearing Loss and Its Syndromes. New York: Oxford University Press, 1995. Kelsell DP, Dunlop J, Stevens HP, et al. Connexin 26 mutations in hereditary non-syndromic sensorineural deafness. Nature 1997;387:80–83. Liu XZ, Walsh J, Mburu P, et al. Mutations in the myosin VIIA gene cause non-syndromic recessive deafness. Nat Genet 1997;16:188–190. Moller CG, Kimberling WJ, Davenport SL, et al. Usher syndrome: an otoneurologic study. Laryngoscope 1989;99:73–79. Morton NE. Genetic epidemiology of hearing impairment. Ann NY Acad Sci 1991;630:16–31. Petit C. Genes responsible for human hereditary deafness: a symphony of a thousand. Nat Genet 1996;14:385–391. Weil D, Blanchard S, Kaplan J, et al. Defective myosin VIIA gene responsible for Usher syndrome type 1B. Nature 1995;374:60,61. Weil D, Kussel P, Blanchard S, et al. The autosomal recessive isolated deafness, DFNB2, and the Usher 1B syndrome are allelic defects of the myosin-VIIA gene. Nat Genet 1997;16:191–193. Zlotogora J, Sagi M, Schuper A, Leiba H, Merin S. Variability of Stickler syndrome. Am J Med Genet 1992;42:337–339.

INDEX

1099

Index

A Aarskog-Scott syndrome, clinical features, 1039, 1040 differential diagnosis, 1040 FDG1, Grb2-binding region in product, 1042, 1043 GTPase product, 1039, 1041, 1042 mapping, 1040, 1041 mutation analysis, 1041 structure, 1041 molecular pathophysiology, 1043, 1044 Abetalipoproteinemia, clinical features, 414 diagnosis, 414 genetics and molecular pathophysiology, 414 management, 414 Abetalipoproteinemia, phenotype and gene, 880 ACE, see Angiotensin-converting enzyme Acetaminophen, hepatotoxicity, 372 Achondroplasia, clinical features, 1033, 1034 diagnosis, 1034 genetic basis of disease, 1034, 1035, 1037 management, 1037, 1038 Acidosis, acute renal failure role, 653 ACLS, see Acrocallosal syndrome Acquired immunodeficiency syndrome (AIDS), see Human immunodeficiency virus Acrocallosal syndrome (ACLS), differential diagnosis, 1021, 1022 Acromegaly, clinical manifestations, 445 growth hormone, hypersecretion, 444, 445 ACTH, see Adrenocorticotropic hormone Actin, see Cytoskeleton; Muscle Activin, follicle-stimulating hormone, regulation of release, 521

receptors, 522, 523 synthesis sites, 522 transgenic mouse studies, 522, 523 types, 521 Acute intermittent porphyria, see Porphyria Acute lymphoblastic leukemia (ALL), chromosomal translocations, 235–237 classification characteristics, 233, 234 epidemiology, 233 molecular analysis, 233, 234, 238 tumor suppressor gene mutations, 238 Acute myeloid leukemia (AML), chromosomal translocations, 237, 238 classification characteristics, 233, 234 epidemiology, 233 molecular analysis, 233, 234, 238 Acute pancreatitis, clinical features, 401 diagnosis, 401 genetic predisposition, 402, 403 management/treatment, 403, 404 molecular pathophysiology, 401, 402 Acute renal failure (ARF), apoptosis, 654 classification, 651 etiology, 651 hypoxia effects on gene expression, 654, 655 leukocyte adherence role, 653, 654 recovery, gene expression during, growth factors, 656 immediate-early genes, 655, 656 proto-oncogenes, 655, 656 tubular component of ischemic disease, acidosis role, 653 calcium role, 652, 653 phospholipase role, 653 purine depletion, 653 reactive oxygen species role, 653 structural changes, 652 volume regulation, 652 vascular component of ischemic disease, 651, 652

1099

AD, see Alzheimer’s disease; Atopic dermatitis ADA, see Adenosine deaminase ADCAII, see Autosomal dominant cerebellar ataxia Adeno-associated virus, gene therapy application, 776, 777 Adenosine deaminase (ADA), gene therapy in severe-combined immunodeficiency, 286, 779 deficiency, molecular basis of immunodeficiency, 287, 288 Adenovirus, gene therapy application, 776, 777 ADH, see Antidiuretic hormone ADPK, see Autosomal dominant polycystic kidney disease Adrenal cortex neoplasm, clinical features, 502 genetic basis of disease, 502 management, 502 Adrenal gland, see Adrenal cortex neoplasm; Adrenal hypoplasia congenita; Adrenoleukodystrophy; Carney complex; Congenital adrenal hyperplasia; Hereditary isolated glucocorticoid deficiency; Primary familial glucocorticoid resistance; Pseudohypoaldosteronism; Syndrome of apparent mineralocorticoid excess Adrenal hypoplasia congenita (AHC), clinical manifestations, 498, 604 forms, 498, 604 genetic basis of disease, 498, 499, 604 management, 498 Adrenocorticotropic hormone (ACTH), Cushing’s syndrome, hypersecretion, 443, 444 pituitary cell secretion, 443, 444 regulation, 443 tumor secretion, 443 Adrenoleukodystrophy, clinical manifestations, 499 genetic basis of disease, 499 management, 499

1100 phenotype and gene, 880 Affective disorder, age at onset, 996 comorbidity, 997 diagnosis, 997, 998 drug response, 996, 997 epidemiology, 997 genes in bipolar disorder, 1001, 1002 heredity, adoption studies, 1000, 1001 associative findings, 1002 family studies, 998–1000 linkage studies, 1001, 1002 twin studies, 1000 history of study, 995 Kraepelin’s concept, 996 management/treatment, 1003 molecular pathophysiology, 1002, 1003 nosology, DSM-IV classification, 995, 996 ICD-10 classification, 995, 996 unipolar versus bipolar disorder, 995 AHC, see Adrenal hypoplasia congenita AIDS, see Acquired immunodeficiency syndrome Alagille syndrome, molecular basis of disease, 368 Albinism, see Oculocutaneous albinism Albumin, defects in thyroid deficiency, 467 Alcoholism, alcohol hepatotoxicity and liver disease, 371, 1006 developmental genetic modeling, 1009–1011 genes, aldehyde dehydrogenase, 1009 dopamine D2 receptor, 1008, 1009 monoamine oxidase B, 1009 heredity, adoption studies, 1007 heritability estimates, 1008 history of study, 1005 twin studies, 1006 natural history, 1005, 1006 Aldehyde dehydrogenase, defects in alcoholism, 1009 Aldolase, red cell enzymopathy, 199, 200 Aldosterone, see also Pseudohypoaldosteronism, synthesis, 481, 482 ALL, see Acute lymphoblastic leukemia Allergy, see Asthma; Atopic dermatitis Alport syndrome, clinical features and diagnosis, 665 glomerular filtration barrier structure, 665, 666 hearing loss, 1096, 1097 heredity, 665 therapy, 667, 668

INDEX

type IV collagen gene and mutations, 665, 666 ALS, see Amyotrophic lateral sclerosis Alström syndrome, clinical features and genetics, 592 Alzheimer’s disease (AD), β-amyloid precursor protein, gene mutations, 902, 903, 905 isoforms, 901 posttranslational processing, 901– 903 animal models, 905 apolipoprotein E, ε4 allele in disease, 903–905 hereditary risk factors, 901 mouse models, 105, 106 pathogenesis, 901 phenotypes and genes, 878 presenelin genes, mutation in disease, 904, 905 AMH, see Anti-Müllerian hormone AML, see Acute myeloid leukemia Amyloidosis, liver, 370, 371 β-Amyloid precursor protein (APP), gene mutations in Alzheimer’s disease, 902, 903, 905 isoforms, 901 knockout mice studies, 106 overexpression in transgenic mice, 105, 106 posttranslational processing, 901–903 Amyotrophic lateral sclerosis (ALS), animal models, 104, 105, 910, 911 clinical features and differential diagnosis, 907, 908 pathogenesis and genetics, 909, 910 pathology, 908, 909 phenotypes and genes, 881 treatment, 911 Androgen receptor (AR), insensitivity syndromes, clinical features, 550, 551, 581, 582 diagnosis, 550, 551, 609 gene mutations, 430, 582–584 genetic and molecular pathophysiology, 551 ligands, 429, 430, 581 structure and properties, 428, 429, 581 Angelman syndrome (AS), albinism association, 1059 clinical features and course, 1055 diagnosis, differential diagnosis, 1055 molecular diagnosis, 1057, 1058 genetic basis of disease, 1056, 1057 genetic counseling, 1058 genomic imprinting, 1058–1060 incidence, 1053 management/treatment, 1055, 1056, 1060 mouse model, 1059 Angina, see Coronary atherosclerosis

Angiotensin-converting enzyme (ACE), hypertension role, 146, 151, 153 Angiotensinogen, hypertension role, 146, 151, 153 Animal models, see also Transgenic mouse, Alzheimer’s disease, 105, 106, 905 amyotrophic lateral sclerosis, 104, 105, 910, 911 Angelman syndrome, 1059 behavior, 981–985 cardiac arrhythmia, 159 congenital heart disease, 117, 121, 122, 124 Down syndrome, 1076, 1077 Gerstmann-Straüssler-Scheinker disease, 936 Greig cephalopolysyndactyly syndrome, 1022, 1023 neurofibromatosis, 968 Prader-Willi syndrome, 1059 xeroderma pigmentosum, 108, 109, 757 Ankylosing spondylitis, human leukocyte antigen association, 276 ANP, see Atrial natriuretic peptide Anthralin, psoriasis treatment, 799 Anti-Müllerian hormone (AMH), disorders of production, clinical features and diagnosis, 551 genetic and molecular pathophysiology, 551, 552 genital development role, 527, 528 puberty role, 570 receptor disorders, 552 Antibody, antigenic determinants in autoimmune disease, 256 B-cell superantigens, antibody interactions, 254 antibody production, 254 B-cell stimulation, 254, 255 immunodeficiency, hyper-immunoglobulin M, 289 immunoglobulin subclass deficiency, 289 X-linked agammaglobulinemia, 289 immunoglobulin E, heritability of serum levels, 805 role in atopic skin inflammation, 803, 804 specificity, molecular basis, 252, 253 variable region restriction, 253, 254 VH segments, classification, 252, 253 gene mapping, 253 germline repertoire analysis, 253 Antidiuretic hormone (ADH), knockout mice studies, 102, 103 Antithrombin (AT), deficiency, incidence, 221

1101

INDEX

type I deficiency, 222 type II deficiency, 222 function with heparin sulfate in anticoaglation, 220 gene structure, 220, 221 structure, 220 α-1-Antitrypsin deficiency, clinical manifestations, 379 diagnosis, 379 emphysema, 344, 345 genetic basis of disease, 379, 380 management/treatment, 380 molecular pathophysiology, 380 APC, mutations in cancer, 78, 79, 81 Apert syndrome, clinical features, 1032, 1033 diagnosis, 1034 genetic basis of disease, 1034, 1035, 1037 management, 1037, 1038 Apolipoprotein A-I, transgenic mouse overexpression, 103 Apolipoprotein B, see Abetalipoproteinemia Apolipoprotein E, ε4 allele in Alzheimer’s disease, 903– 905 transgenic mouse overexpression, 103 Apoptosis, acute renal failure, 654 inducers, 72 pathways, 72 APP, see β-Amyloid precursor protein AQP2, see Aquaporin 2 Aquaporin 2 (AQP2), mutations in nephrogenic diabetes insipidus, 671, 672 AR, see Androgen receptor ARF, see Acute renal failure Aromatase (P450arom) deficiency, clinical features and diagnosis, 555, 556 genetic and molecular pathophysiology, 556 AS, see Angelman syndrome Asthma, adhesion molecule recruitment of immune cells, 321, 322 clinical features, 309, 310, 319 diagnosis, 310 DNA markers and molecular genetics, 326 eosinophils in inflammation, 322 epidemiology, 309, 319 gene mapping, 310–312 heredity, 324, 326, 805, 806 histamine releasing factors in disease, clinical relevance, 314 purification of immunoglobulin Edependent histamine releasing factor, 313, 314 types, 312, 313 late-phase reaction, 309, 310, 312

management/treatment, 314, 315 molecular pathophysiology, 312–314 nitric oxide in inflammation, 322, 323 T-helper-1 response, interferon-γ, generation, 323 recruitment of immune cells, 323, 324 overview, 319, 320, 323 T-helper-2 response, interleukin-4 generation, 320 recruitment of immune cells, 321 overview, 319, 320 Th2 cell response, 804, 805 Astrocytoma, see Brain tumor AT, see Antithrombin; Ataxia-telangiectasia Ataxia-telangiectasia (AT), clinical features, 759, 760 complementation groups, 759 diagnosis, 760, 761 gene, see ATM, genetic basis of disease, 761, 762 molecular basis of immunodeficiency, 288 molecular pathophysiology, 762, 763 phenotype and gene, 878 treatment/management, 763 Ataxin 1, characterization in spinocerebellar ataxia, 916 ATM, discovery, 761 locus, 762 molecular pathophysiology of ataxiatelangiectasia, 762, 763 mutations in cancer, 79 protein function, 762 Atopic dermatitis (AD), allergens, role in disease, 801 candidate genes, positional cloning, 806 types, 806, 807 cutaneous infection association, 801, 802 cytokine expression patterns, 802, 803 diagnosis, 801, 802 familial predisposition, 805 histologic findings, 802 immunoglobulin E, heritability of serum levels, 805 role in skin inflammation, 803, 804 management, 807, 808 skin reaction patterns, 801 Th2 cell response, 804, 805 Atopy, 271 Atrial natriuretic peptide (ANP), knockout mice studies, 103 Autoimmune disease, see also Autoimmune hepatitis; Dermatomyositis; Diabetes mellitus, type 1; Grave’s

disease; Hashimoto’s thyroiditis; Idiopathic thrombocytopenia; Insulindependent diabetes mellitus; Lupus erythematosus, specific skin disease; Multiple sclerosis; Pemphigoid; Pemphigus; Polymyositis; Primary biliary cirrhosis; Psoriasis; Rheumatoid arthritis; Sjögren’s syndrome; Systemic lupus erythematosus; Systemic sclerosis, classification, 299 genetic basis of autoimmunity, 303, 304 molecular pathophysiology, 304, 305 Autoimmune hepatitis, clinical features, 369 pathogenesis, 369, 370 treatment, 370 Autosomal dominant cerebellar ataxia (ADCAII), phenotype and gene, 878 Autosomal dominant polycystic kidney disease (ADPK), clinical features and variability, 675– 677, 682 diagnosis, 677 PKD1, expression, 680 mutations, 637, 678–680 polycystin product, 678–680 structure, 678 PKD2, mutations, 637 protein product, 680–682 structure, 680, 681 prognosis, 675 treatment, 682, 683 Autosomal dominant polycystic liver disease, see Polycystic liver disease AZF, see Azoospermia factor Azoospermia factor (AZF), deletion in germinal cell failure, 607, 608

B BAC, see Bacterial artificial chromosome Bacterial artificial chromosome (BAC), DNA cloning, 13, 61 Bare lymphocyte syndrome, molecular basis of immunodeficiency, 286, 287 Basal cell carcinoma (BCC), see also Basal cell nevus syndrome, molecular pathogenesis, 786, 787 risk factors, immune suppression, 785, 786 inherited cancer syndromes, 785 ultraviolet radiation, 785 Basal cell nevus syndrome (BCNS), clinical features, jaw, 745 skeletal, 745, 746 skin, 745

1102 visceral tumors, 746 diagnosis, 746 genetic basis of disease, 747 incidence, 745 treatment, 747 Base excision repair, genes, 759 pathway, 754, 755 Basement membrane zone (BMZ), cutaneous complexity, 729 Basic helix-loop-helix (bHLH) proteins, cell differentiation and determination, 848 myogenesis role, 842, 843 BCC, see Basal cell carcinoma B-cell, activation, 255 events in immune response, 284 immune deficiency, carrier state and prenatal diagnosis, 285, 286 combined T-cell diseases, adenosine deaminase deficiency, 287, 288 ataxia-telangiectasia, 288 bare lymphocyte syndrome, 286, 287 purine nucleoside phosphorylase deficiency, 287 Wiscott-Aldrich syndrome, 287 X-linked severe combined immunodeficiency, 286 defects in disorders, 285 hyper-immunoglobulin M, 289 immunoglobulin subclass deficiency, 289 laboratory assessment, 283, 285 surface markers, 285 symptoms, 283 therapy, 286 X-linked agammaglobulinemia, 289 lymphoma, see Lymphoma, regulation of immune response, 257 signaling pathways, 255, 256 superantigens, antibody interactions, 254 antibody production, 254 B-cell stimulation, 254, 255 BCL-1, translocation in mantle cell lymphoma, 245 Bcl-2 kidney development role, 637 mutations in cancer, 80 translocation in lymphoma, 243–245 BCL-6, translocation in diffuse large cell lymphoma, 245 BCNS, see Basal cell nevus syndrome Beare-Stevenson cutis gyrata syndrome, clinical features, 1031 diagnosis, 1034 genetic basis of disease, 1034, 1035, 1037

INDEX

management, 1037, 1038 Becker muscular dystrophy (BMD), incidence, 859, 861 phenotype and gene, 882 Becker myotonia, phenotype and gene, 882 Beckwith-Wiedemann syndrome (BWS), chromosomal alterations, 1049, 1050 course, 1047 diagnosis, clinical features, 1047, 1048 differential diagnosis, 1048 molecular testing, 1050, 1051 prenatal diagnosis, 1048, 1049 genomic imprinting, 1050, 1051 incidence, 1047 management/treatment, 1049, 1051 mode of inheritance, 1049 Behavior, see also Affective disorder; Alcoholism; Schizophrenia, bird singing, 983 genetic factors, 979, 980, 985, 986 invertebrate models, Drosophila sex behaviors, 982, 983 sensitization in marine mollusk, 981, 982 molecular biology techniques in analysis, 979, 980 mouse models, addiction, 983, 984 aggression, 983 depression, 984, 985 nongenetic factors, 980, 981 Berlin breakage syndrome, 763, 764 bHLH, see Basic helix-loop-helix Bilateral anorchia, 593 Bile acid, metabolism, 366 secretion, 366, 367 Bipolar disorder, see Affective disorder Blaschko’s lines, see Epidermal nevi; McCune-Albright syndrome, 713 BLM, see Bloom syndrome Bloom syndrome, BLM gene mutations, 696, 764–766 clinical features, 764, 765 diagnosis, 765 genetic basis of disease, 765, 766 management/treatment, 766, 767 BMD, see Becker muscular dystrophy BMP-7, see Bone morphogenetic protein-7 BMZ, see Basement membrane zone Bone marrow, see also Hematopoiesis, cellular composition, 171, 172 transplantation, cancer therapy treatment, 176, 177 immune cell deficiency therapy, 286 paroxysmal nocturnal hemoglobinuria, 231 sickle cell anemia, 188 Bone morphogenetic protein-7 (BMP-7), knockout mice studies, 104

BOR, see Branchio-oto-renal syndrome Bovine spongiform encephalopathy (BSE), epidemiology, 933, 934 transmission to other animals, 934 BPAG2, mutations in junctional epidermolysis bullosa, 725 Brain tumor, clinical evaluation, 972 clinical presentation, 971 genetic predisposition, 972, 973 growth factor receptor expression, 974 incidence, 971 molecular pathology, astrocytoma, 973, 974 medulloblastoma, 974 meningioma, 974 oligodendroglioma, 974 pathology and classification, 971, 972 therapy, 974, 975 Branchio-oto-renal syndrome (BOR), hearing loss, 1097 BRCA1, mutations in cancer, 79, 630 BRCA2, mutations in cancer, 79, 630 Breast cancer, DNA content, 626 epidermal growth factor receptor, 627 ERBB2 activation, 626, 627 estrogen receptor expression, 625, 631, 632 evolution, 625 growth factor expression, insulin-like growth factor 1, 628 transforming growth factor-α, 627 transforming growth factor-β, 628 hereditary forms, 629 incidence, 625 loss of heterozygosity, 630, 631 metastasis, 631 myc activation, 627 progesterone receptor expression, 626, 631 prognostic factors, 631 proliferation rate, 626 ras mutations, 627 treatment, 632 tumor suppressor genes, 628, 629 BSE, see Bovine spongiform encephalopathy Bullous pemphigoid, autoantibodies, 818–820 blister formation mechanisms, 819 clinical features, 817 diagnosis, 817 management, 819 BWS, see Beckwith-Wiedemann syndrome,

C CAH, see Congenital adrenal hyperplasia Calcium, acute renal failure role, 652, 653

INDEX

cyclic AMP response element-binding protein, induced phosphorylation, 427 muscle contraction regulation, 852, 853, 855, 856 parathyroid hormone secretion regulation, 475–477 signal transduction, slow inositol pathway, 426 voltage-dependent calcium channels, 426 Calcium/calmodulin-dependent protein kinase II (CaMKII), deficiency and aggression, 983 learning role, 982 Calpain III, deficiency in muscular dystrophy, 862 CaMKII, see Calcium/calmodulin-dependent protein kinase II Cancer, see also Adrenal cortex neoplasm; Basal cell carcinoma; Brain tumor; Breast cancer; Carney complex; Lung cancer; Melanoma; Multiple endocrine neoplasia; Oncogene; Parathyroid gland neoplasia; Pituitary tumor; Renal cell carcinoma; Retinoblastoma; Rhabdomyosarcoma; Squamous cell carcinoma; Thyroid cancer; Tumor suppressor gene; Wilms’ tumor, hepatitis viruses in oncogenesis, 396, 397 inherited syndrome, 80 multistep pathogenesis, 80–82 Cardiac arrhythmia, animal models, 159 clinical syndromes, 158, 159 long QT syndrome, candidate genes, 157, 158 clinical features, 157 management/treatment, 158 molecular basis, 159 Cardiac myosin binding protein C, mutation in hypertrophic cardiomyopathy, 128, 129 β Cardiac myosin heavy chain, mutation in hypertrophic cardiomyopathy, 128 Cardiac troponin T, mutation in hypertrophic cardiomyopathy, 128, 129 Cardiomyopathy, see Dilated cardiomyopathy; Hypertrophic cardiomyopathy Carney complex, clinical features, 501 genetic basis of disease, 501, 502 management, 501 CAT, see Chloramphenicol acetyltransferase CBAVD, see Congenital bilateral absence of the vas deferens CCD, see Central core disease CCSP, see Clara cell secretory protein Celiac sprue, clinical features, 411

diagnosis, 411, 412 epidemiology, 411 genetic basis, 412 management, 412 molecular pathophysiology, 412 Cell cycle, cell proliferation and differentiation, overview of control, 70–72 checkpoint pathways, 69, 70 cyclin/cyclin-dependent kinase control, 65–72, 427, 428 duration of phases, 65 G1/S, restriction point control, 67 G2/M transition, 69 knockout mice studies, 98, 99 M/G1, proteolysis pathways, 69 meiosis overview, 44, 65 mitosis overview, 44, 65 retinoblastoma protein regulation, 958, 959 S phase, defects in disease, 68 mechanisms controlling DNA replication, 68, 69 Cell-mediated immunity, see T-cell Central core disease (CCD), phenotype and gene, 857, 882, 951 CF, see Cystic fibrosis cGMP, see Cyclic GMP Charcot-Marie-Tooth disease (CMT), clinical features and pathogenesis, 921, 922 CMT1A duplication, 921, 922, 925, 926 molecular diagnosis, 925 peripheral myelin protein-22 in pathogenesis, 922–925 phenotypes and genes, 884, 885, 921, 922, 924, 925 prevalence, 921 reciprocal recombination, 923, 924 Chemokine, 267–270 Chloramphenicol acetyltransferase (CAT), reporter gene, 23, 24 Christmas disease, see Hemophilia B Chromosome, aberrations, 5, 7, 49, 50 karyotypes and terminology, 5, 6 staining, 5, 7, 871, 872 structure, 3, 5, 26, 43 Chronic lymphocytic leukemia (CLL), chromosomal translocations, 238 classification characteristics, 233, 234 epidemiology, 233 molecular analysis, 233, 234, 238 Chronic myelogenous leukemia (CML), chromosomal translocations, 74, 235, 238 classification characteristics, 233, 234 epidemiology, 233 molecular analysis, 233, 234, 238 Chronic obstructive pulmonary disease (COPD), see also Emphysema,

1103 definition, 339 epidemiology, 339 pathology, 339 Chronic progressive external ophthalmoplegia (CPEO), 943 Cicatricial pemphigoid, clinical features, 817 diagnosis, 817 management, 819 Clara cell secretory protein (CCSP), knockout mice studies, 102 Cleft lip, see Orofacial clefting Cleft palate, see Orofacial clefting CLL, see Chronic lymphocytic leukemia Cloning DNA, bacterial artificial chromosomes, 13, 61 contribution to elucidation of neurological disease mechanisms, 876 cosmid vectors, 13, 61 Escherichia coli plasmids, 12 expression cloning, 875 functional cloning, 872 gene families, 875 human analogs of genes causing disorders in other species, 875 library construction and screening, 13, 14, 48, 60, 61, 874, 875 phage vectors, 12, 13, 61 positional cloning, 872 random cloning, 875 yeast artificial chromosomes, 13, 61 CML, see Chronic myelogenous leukemia CMT, see Charcot-Marie-Tooth disease Coagulation factor deficiency, factor V, 216 factor VIII, see Hemophilia A, factor IX, see Hemophilia B, factor X, 216 factor XI, 216 management/treatment, 216, 217 von Willebrands factor, see von Willebrands disease Codon, genetic code, 43, 44 Collagen, type IV gene and mutations in Alport syndrome, 665, 666 type VII gene defects in dystrophic epidermolysis bullosa, 730–732 Congenital adrenal hyperplasia (CAH), 11β-hydroxylase deficiency, CYP11B1 mutations, 485, 486, 488, 555 CYP11B2 mutations, 487, 488, 555 epidemiology, 485, 554 prenatal diagnosis, 488 treatment, 488, 492 17α-hydroxylase deficiency, 490 21-hydroxylase deficiency, classic salt-wasting, 481, 482, 552 classic simple virilizing, 481, 552, 576

1104 cryptic, 552 epidemiology, 482 gene defects, 483–486 genetic and molecular pathophysiology, 553, 554 hormonal diagnosis, 485, 552, 553 nonclassic, 482, 552 prenatal diagnosis, 485, 558 treatment, 485, 491, 492 3β-hydroxysteroid dehydrogenase deficiency, 488, 555 lipoid congenital adrenal hyperplasia, 490, 491, 544, 594, 595 treatment, 491, 492, 558 Congenital bilateral absence of the vas deferens (CBAVD), clinical features, 606 diagnosis, 606 genetic basis of disease, 606 molecular pathophysiology, 6066 therapy, 606 Congenital heart disease, clinical features, 117, 121 diagnosis, 121 genes in cardiac morphogenesis, animal models in mutational analysis, 117, 121, 122, 124 channels, 121 extracellular proteins, 119 receptors, 119, 120 secreted proteins, 119 transcription factors, 118, 119 genetic basis of disease, 121, 122 management/treatment, 124 molecular pathophysiology, 123, 124 Congenital muscular dystrophy, phenotype and gene, 882 Connexin 32 gene (Cx32), mutation in Charcot-Marie-Tooth disease, 924 Conotruncal anomaly face syndrome (CTAFS), chromosome 22q11 deletion, 1081 diagnosis, 1082, 1083 management, 1083, 1084 Contraception, 524 COPD, see Chronic obstructive pulmonary disease Coronal synostosis syndrome, clinical features, 1032 diagnosis, 1034 genetic basis of disease, 1034, 1035, 1037 management, 1037, 1038 Coronary atherosclerosis, clinical features, 133 gene therapy, 161, 162, 165 management/treatment, 138, 139 molecular pathophysiology, 135 role, lipids, 135–137 monocyte chemoattractant protein-1, 138

INDEX

mononuclear cells, 137 shear stress, 138 smooth muscle cells, 137 vascular cell adhesion molecule-1, 137, 138 stable angina, diagnosis, clinical presentation, 134, laboratory evaluation, 134, 135 physical examination, 134 unstable coronary syndromes, diagnosis, clinical presentation, 133, 134 laboratory evaluation, 134 physical examination, 134 Corticotropin releasing factor (CRF), expression in depression, 984 Cortisol, synthesis, 481, 482 CPEO, see Chronic progressive external ophthalmoplegia CREB, see Cyclic AMP response element-binding protein CREM, see Cyclic AMP response element modulator Creutzfeldt-Jakob disease, see also Prion, clinical manifestations, 928, 929 diagnosis, 929, 930 etiology, 927, 928, 930, 931 familial disease, 931–933 genetic linkage, 931–933 iatrogenic disease, 930, 931, 933 phenotype and gene, 878 CRF, see Corticotropin releasing factor Crigler-Najjar syndrome, clinical features, 367 molecular basis of disease, 367 Crohn’s disease, complications, 408 diagnosis, 408, 409 extraintestinal manifestations, 408 location of disease, 407 medical therapy, 410, 411 molecular pathophysiology, 409, 410 signs and symptoms, 407, 408 surgical treatment, 411 Crouzon syndrome, acanthosis nigricans association, 1030, 1031 clinical features, 1029, 1030 diagnosis, 1034 genetic basis of disease, 1034, 1035, 1037 hearing loss, 1097 management, 1037, 1038 Cryptorchidism, clinical features, 608 etiology, 608 treatment, 609 CTAFS, see Conotruncal anomaly face syndrome Cushing’s syndrome, adrenocorticotropic hormone hypersecretion, 443, 444

clinical features, 443, 444 Cutaneous lupus erythematosus, see Lupus erythematosus, specific skin disease Cx32, see Connexin 32 gene Cyclic AMP response element modulator (CREM), cyclic AMP response element-binding protein regulation, 427 gene autoregulation, 37, 38 Cyclic AMP response element-binding protein (CREB), behavior role, 981 binding protein, 36 cell proliferation and differentiation, overview of control, 72 modular functional domains, 32, 33 phosphorylation in signal transduction, 427 Cyclic GMP (cGMP), signal transduction, 423 Cyclin/cyclin-dependent kinase, control of cell cycle, 65–72, 427, 428, 844 Cyclosporine A, psoriasis treatment, 799 CYP11A, see P450 side-chain cleavage enzyme CYP11B, see 11β-Hydroxylase deficiency CYP17, see 17α-Hydroxylase deficiency CYP21, see 21-Hydroxylase deficiency Cystic fibrosis (CF), cystic fibrosis transmembrane conductance regulator, gene, discovery, 329 genotype-phenotype correlation, 334 point mutation analysis, 85, 86, 333, 334 structure, 329 localization of expression, 331 regulation, 330, 331 structure, 329, 330 incidence, 329, 380 liver disease, clinical manifestations, 381 diagnosis, 381 genetic basis of disease, 381 management/treatment, 381 molecular pathophysiology, 381 pancreas disease, 331, 404, 405 related lung disease pathogenesis, airway inflammation, chronic response, 333 initiation, 331 bacterial pathogens, 331–333 risk assessment, 92 therapy, correction of protein trafficking, 336 corticosteroids, 335, 336 deoxyribonuclease I, 335 gene therapy, 336, 337

INDEX

modulation of ion and water transport, 336 pentoxifylline, 336 protease inhibitors, 336 transgenic mouse model, 102 Cytochrome b5 reductase, red cell enzymopathy, 200, 204, 205 Cytochrome P450, see also Aromatase deficiency; 11β-Hydroxylase deficiency; 17α-Hydroxylase deficiency; 21-Hydroxylase deficiency; P450 side-chain cleavage enzyme, drug metabolism and liver damage, 372 Cytokine, 267–271 Cytoskeleton, muscle membrane cytoskeleton system, dystrophin, 853 integrin, 853 spectrin, 853 Rho regulation of actin cytoskeleton, 1043, 1044

D Darier’s disease (DD), clinical features, 719 diagnosis, 719 genetic basis of disease, 719 management/treatment, 720 molecular pathophysiology, 719, 720 DAX-1, mutations and clinical features in sex determination disorders, 538, 539, 564 DCC, mutations in cancer, 78, 81 DD, see Darier’s disease Deafness, see Hearing loss DEB test, see Diepoxybutane test Dehalogenase, defects in thyroid deficiency, 466 Dejerine-Sottas disease (DSD), phenotypes and genes, 885, 921 Delayed puberty, see Puberty Dentatorubropallidoluysian atrophy (DRPLA), phenotype and gene, 879, 898, 914, 917 Denys-Drash syndrome, WT1 mutations and clinical features, 539–541 Deoxyribonuclease I, cystic fibrosis therapy, 335 Depression, see Affective disorder Dermatomyositis, clinical features, 301 diagnosis, 302 management/treatment, 306 DGS, see DiGeorge syndrome Diabetes insipidus, see Nephrogenic diabetes insipidus Diabetes mellitus, type 1 autoantibodies, 433 autoantibody, 251 heredity, 433, 434 human leukocyte antigen association, 276, 277

prevalence, 433 screening, 435 susceptibility loci, 434, 435 therapy, 435 transgenic mouse studies, 106, 107 Diabetes mellitus, type 2 heredity, 433, 436 insulin gene mutations, insulin, 438 receptor, 438, 439 late inset disease, candidate genes, 439–441 clinical implications of molecular genetics, 442 susceptibility loci, 441, 442 maturity onset diabetes of the young, gene mutations, 437, 438 phenotypes, 437, 438 mitochondrial diabetes mellitus, clinical features, 435–437 gene mutations, 436 polygenicity, 436 prevalence, 433 Diagnostic and Statistical Manual of Mental Disorders, 4th ed. (DSM-IV), affective disorder classification, 995, 996, 998 Diamond blackfan anemia, treatment, 178 Diclofenac, hepatotoxicity, 373 Diepoxybutane (DEB) test, fanconi anemia diagnosis, 767–769 Diffuse large cell lymphoma, see Lymphoma DiGeorge syndrome (DGS), chromosome 22q11 deletion, 1079, 1081, 1082 clinical features, 1079, 1081 cytogenetic analysis, 1081, 1082 diagnosis, 1082, 1083 gene identification in DiGeorge chromosomal region, 1084 gene mutations in cardiac malformation, 122 hypoparathyroidism, 477 incidence, 1079 management, 1083, 1084 molecular basis of immunodeficiency, 288 Dilated cardiomyopathy, clinical features, 130 gene loci in familial disease, 130, 131 Diphosphoglycerate mutase, red cell enzymopathy, 199, 200 DMD, see Duchenne muscular dystrophy DNA, see also Chromosome; Gene; Genome, cloning, see Cloning DNA, gel electrophoresis, 15, 16 genomic DNA isolation, 11 mutation, detection, 19, 20, 49

1105 types, 19, 48, 49 polymorphism detection, microsatellites, 19 restriction fragment-length polymorphisms, 18 variable number tandem repeats, 19 protein interaction analysis, DNA affinity chromatography, 22 DNA footprinting, 21 electrophoretic mobility shift assay, 21, 22 sequencing, 17, 18, 20, 45, 46, 61 Southern blotting, 16, 17 structure, 43 DNA affinity chromatography, DNAbinding proteins, 22 DNA cloning, see Cloning DNA DNA footprinting, principle, 21 promoter studies, 27, 28 DNA ligase, molecular biology applications, 9 DNA polymerase, molecular biology applications, 10 DNA repair, see also Base excision repair; Nucleotide excision repair, tumor suppressor gene mutation in cancer, 79, 80 DNA replication, mechanisms controlling, 68, 69 Dopamine receptor, addiction role, 984 D2 defects in alcoholism, 1008, 1009 Dowling-Meara syndrome, see Epidermolysis bullosa, simplex form Down syndrome, chromosome 21, genome mapping and Down syndrome critical region, 1073, 1076, 1078 clinical phenotypes, 1069, 1070 genetic counseling, 1072, 1073 history of study, 1069 incidence, 1069 maternal age risk, 1069–1071, 1073 molecular pathophysiology of increased gene dosage, 1077, 1078 mouse models, 1076, 1077 origin of trisomy 21 free trisomy 21, 1071 origin in other trisomies, 1071, 1072 translocation trisomy 21, 1071 prenatal diagnosis, 1072 DRPLA, see Dentatorubropallidoluysian atrophy DSD, see Dejerine-Sottas disease Dsg proteins, autoantibodies in pemphigus, 814–816 DSM-IV, see Diagnostic and Statistical Manual of Mental Disorders, 4th ed. Dubin-Johnson syndrome, molecular basis of disease, 368 Duchenne muscular dystrophy (DMD),

1106 diagnostic testing, multiplex polymerase chain reaction, 83 Southern blot analysis, 84, 85 dystrophin mutation , 43, 83 gene therapy, 101, 161, 862 incidence, 859–861 linkage analysis, 87 phenotype and gene, 882 transgenic mice studies, 101 Dwarfism, see Achondroplasia; Hypochondroplasia; Thanatophoric dysplasia Dystrophic epidermolysis bullosa, see Epidermolysis bullosa Dystrophin, diagnostic testing, multiplex polymerase chain reaction, 83 Southern blot analysis, 84, 85 gene therapy, 101 membrane cytoskeleton system, 853 mutation and disease, 43, 83, 860

E EB, see Epidermolysis bullosa EBV, see Epstein-Barr virus EGFR, see Epidermal growth factor receptor EHK, see Epidermolytic hyperkeratosis Elastin, degradation role in emphysema, 345, 346 Electrophoretic mobility shift assay (EMSA), principle, 21, 22 promoter studies, 28 Emery-Dreifuss muscular dystrophy, phenotype and gene, 882 Emphysema, α1-antitrypsin deficiency, 344, 345 definition, 339, 340 elastin degradation role, 345, 346 epidemiology, 339 macrophage proteinases, 342, 343 matrix metalloproteinases, emphysema expression patterns, 344 expression by inflammatory cells, 342, 344 monocyte proteinases, 342 neutrophil, cathepsin G, 342 elastase, 342 metalloproteinases, 342 proteinase 3, 342 serine proteinases, 341, 342 smoking-induced retention in lung, 341 pathogenesis, inflammation–destruction–repair hypothesis, 340, 341 proteinase–antiproteinase hypothesis, 340

INDEX

EMSA, see Electrophoretic mobility shift assay Endothelin, vasoactivity, 142 Epidermal growth factor receptor (EGFR), expression in breast cancer, 627 Epidermal nevi, clinical features, 713 genetic basis of disease, 713, 714 management/treatment, 714, 716 molecular pathophysiology, 714 mosaicism and Blaschko’s lines, 713 Proteus syndrome association, 715, 716 Epidermolysis bullosa (EB), dystrophic form, clinical variability, 730 gene therapy, 733 genetic counseling, 732 genotype/phenotype correlations, 731 prenatal diagnosis, 732, 733 type VII collagen defects, 730–732 incidence, 699 junctional form, gene therapy, 727 hemidesmosome-anchoring filament complex structure, 723 molecular basis, generalized atrophic benign form, 725 hemidesmosomal subtypes, 724 Herlitz form, 724 heterogeneity, 696, 723, 724, 726, 729, 730 integrin mutations, 725 non-Herlitz form, 724 pathogenesis similarity with muscular dystrophy, 725 prenatal diagnosis, 725, 727 simplex form, clinical features, 699 diagnosis, 699, 704 genetic basis of disease, 696, 699, 700, 729 management/treatment, 704 molecular pathophysiology, 700–703 types, 699 types, 699, 723 Epidermolytic hyperkeratosis (EHK), clinical features and heterogeneity, 707 diagnosis, 707 genotype/phenotype correlation, 709 incidence, 707 keratin mutations, 707–710 management/treatment, 710 molecular pathophysiology, 710 screening, 710, 711 Epilepsy, phenotypes and genes, 881 Epimerase deficiency, heredity and diagnosis, 385 Episodic ataxia, phenotypes and genes, 879 EPO, see Erythropoietin

Epstein-Barr virus (EBV), lymphoma association, 241, 246, 247 ER, see Estrogen receptor ERBB2, activation in breast cancer, 626, 627 Erythrocyte, see Red blood cell Erythropoietin (EPO), receptor, 174 therapy, premature neonates, 178 renal failure, 177, 178 Estrogen, ovarian cycle control, 613, 614 puberty role, 570 resistance, 570 synthesis, 481, 482, 614 Estrogen receptor (ER), expression in breast cancer, 625, 626, 631, 632 mutations in disease, 430 structure and properties, 428, 429 Ethics, genetic disease testing, diseases with available predictive testing, 88 factors influencing, 57 genetic counselors, 93, 94 Human Genome Project, 62, 63 nonpaternity detection, 57, 58 protocols, 87, 88

F FAA, see Fanconi anemia FAC, see Fanconi anemia Faciogenital dysplasia, see AarskogScott syndrome Facioscapulohumeral dystrophy, phenotype and gene, 883 Factor deficiency, see Coagulation factor deficiency Familial amyloidotic polyneuropathy (FAP), phenotype and gene, 885 Familial hypercholesterolemia (FH), gene therapy, 161 molecular basis, 161 Familial hypocalciuric hypercalcemia (FHH), gene mutations, 476, 477 hyperparathyroidism, 475–477 Fanconi anemia, cell cycle anomalies, 770, 771 clinical features, 767, 768 complementation groups, 769 diagnosis, 768, 769 genetic basis of disease, 769, 770 incidence, 767 molecular pathophysiology, 770, 771 treatment/management, 771 FAP, see Familial amyloidotic polyneuropathy Farwestern blot, protein–protein interaction analysis, 22 Fatal familial insomnia, see also Prion,

INDEX

clinical manifestations, 928, 929 diagnosis, 929, 930 etiology, 927, 928 phenotype and gene, 878 FCMD, see Fukuyama-type congenital muscular dystrophy FDG1, see Aarskog-Scott syndrome FGFR, see Fibroblast growth factor receptor FGR, see Primary familial glucocorticoid resistance FH, see Familial hypercholesterolemia FHH, see Familial hypocalciuric hypercalcemia Fibroblast growth factor receptor (FGFR), gene mutations in skeletal disorders, clinical features, 1029–1034 genetic basis of disease, 1034, 1035, 1037 management, 1037, 1038 molecular diagnosis, 1034 molecular pathophysiology, 1037 gene types, 1029 types, 1035 FISH, see Fluorescence in situ hybridization Fluorescence in situ hybridization (FISH), marker detection, 7 mutation detection, 19, 46 screening for genetic disease, 57 FMR1, see Fragile X syndrome Follicle, see Ovary Follicle-stimulating hormone (FSH), isolated deficiency, 578 modulation, activin, 521–523 follistatin, 523 inhibin, 521, 522 ovarian cycle control, 613, 614 ovary actions, 521 pituitary cell secretion, 443, 444 pituitary tumor secretion, 445, 446, 519, 520 puberty role, 571 receptor mutations, 578 regulation of secretion, 445, 446 testes actions, 521 Follistatin, follicle-stimulating hormone, regulation of release, 522 transgenic mouse studies, 523 Fragile X syndrome, clinical features, 1063 diagnosis, 1063 incidence, 1063 management/treatment, 1066 molecular pathophysiology, 1066 phenotypes and genes, 886, 914 screening, 918, 919, 1066, 1067 trinucleotide repeats in FMR1

analysis, 86, 87 CGG repeats, 918, 1063–1066 FMRP functions, 1065 fragile sites, loci, 1065, 1066 prevalence, 1065 Friedreich’s ataxia, phenotype and gene, 879, 919 Fructose intolerance, hereditary, 385 FSH, see Follicle-stimulating hormone Fukuyama-type congenital muscular dystrophy (FCMD), phenotype and gene, 883

G Galactosemia, clinical manifestations, 384 diagnosis, 384 genetic basis of disease, 384, 385 management/treatment, 385 molecular pathophysiology, 385 GCPS, see Greig cephalopolysyndactyly syndrome G-CSF, see Granulocyte colony-stimulating factor Gene, see also DNA, candidate gene approach, 872 cis and trans regulation, 25 cloning, see Cloning DNA, clusters, 43, 44 distribution by function, 25, 26 functional cloning, 872 linkage analysis, 47, 48, 872 mapping, 46, 47, 59, 60, 871–874 positional candidate approach in disease gene identification, 872, 873 positional cloning, 872 promoter structure, 26, 27 random cloning, 875 size range, 871 structure and organization, 4, 5, 43 transcription assays, nuclear run-on assay, 23 reporter gene expression, 23, 24 yeast two-hybrid system in identification, 875 Gene therapy, Alport syndrome, 668 comparison to classical drug therapy, 775, 776 coronary atherosclerosis, 161, 162, 165 cystic fibrosis, 336, 337 dystrophin, 101, 161 ex vivo gene therapy, 163, 775, 776 expression levels, 777 hemophilia B, 217 hypertension, 161, 162, 166 immunodeficiency patients, 286 in vivo gene therapy, 163, 164, 775, 776 junctional epidermolysis bullosa, 727 skin disorders,

1107 acquired disorders, 778 dystrophic epidermolysis bullosa, 733 gene transfer for systemic disorders, metabolic processing of circulating substrate, 778, 779 systemic delivery, 778 vaccination, 778 inherited disorders, 777, 778 thrombosis, 162, 166, 167 transfection, 776 vectors, 164 viral transduction, 776, 777 Genetic counseling, Angelman syndrome, 1058 diagnosis, types of testing, 91, 92 diseases with avalailable predictive testing, 88 Down syndrome, 1072, 1073 education and training, 94, 95 ethics, 93, 94 Huntington’s disease, 898, 899 information gathering process, 90, 91 Prader-Willi syndrome, 1058 primary care provider partnership, 93 process overview, 89, 90 profession overview, 89 provision of information, 92 psychological assessment and counseling, 92, 93 resources for referral, 89, 90, 93 risk assessment, 92 schizophrenia, 993 support, ongoing and decision-making, 93 Genetic disease, autosomal dominant disorders, 52, 53 autosomal recessive disorders, 53, 54 chromosomal disorders, 50 classification, 49, 50 ethical considerations, 57, 58 genomic imprinting in disease, 54, 55 Mendelian disorders, phenotypic variability, 51, 52 transmission, 50, 51 mitochondrial inheritance, 56 mosaicism in disease, 56 multifactorial disease, 54 screening, 56, 57 somatic mutations, 56 trinucleotide repeats in disease, 55, 56 X-linked disorders, 54 Y-linked disorders, 54 Genome, isolation of genomic DNA, 11 organization in humans, 3, 26, 43, 44, 871 repetitive DNA, 4 scanning in disease gene identification, 875 Genomic imprinting, role in disease, 54, 55, 1050, 1051, 1058–1060 Genotype, definition, 44

1108 Gerstmann-Straüssler-Scheinker disease, see also Prion, clinical manifestations, 928, 929 diagnosis, 929, 930 etiology, 927, 928 phenotype and gene, 878 transgenic mouse model, 936 GH, see Growth hormone Gilbert’s syndrome, molecular basis of disease, 368 Gli1, upregulation in skin cancer, 786, 787 GLI3, see Greig cephalopolysyndactyly syndrome Globin, see Hemoglobin; Thalassemia Glucocorticoid, deficiency, see Hereditary isolated glucocorticoid deficiency, resistance, see Primary familial glucocorticoid resistance Glucose 6-phosphate dehydrogenase, red cell enzymopathy, clinical features, 199, 202 diagnosis, 202 epidemiology, 199 genetic basis, 203 management, 203, 204 molecular pathophysiology, 203 Glucose 6-phosphate isomerase, red cell enzymopathy, 198, 200 Glucose–galactose malabsorption, clinical features, 413 genetics and molecular pathophysiology, 413, 414 management, 414 Glutathione synthetase, red cell enzymopathy, 200, 204 GM-CSF, see Granulocyte–macrophage colony stimulating factor GM1 gangliosidosis, phenotype and gene, 880 GM2 gangliosidosis, diagnosis, 908 GnRH, see Gonadotropin-releasing hormone Gonadotropin-releasing hormone (GnRH), analogs, contraception, 524 precocious puberty treatment, 573 hypogonadotropic hypogonadism, 576, 577 infertility treatment, 524 puberty role, 570 structure and function, 519, 520 G protein, mutations, 424 pituitary tumor, 447 precocious puberty, 574 thyroid-stimulating hormone receptor, 464 seven transmembrane domain receptors, 423, 424

INDEX

Granulocyte colony-stimulating factor (G-CSF), receptor, 174 therapy, neonatal sepsis, 177 neutropenia, chemotherapy-induced disease, 176, 177 congenital disease, 177 Granulocyte–macrophage colony stimulating factor (GM-CSF), receptor, 174, 175 Grave’s disease, autoantibodies, 469, 470 prevalence, 469 Greig cephalopolysyndactyly syndrome (GCPS), animal models, 1022, 1023 clinical features, 1021 differential diagnosis, 1021, 1022 GLI3 gene, identification, 1021, 1022 limb development role, 1024, 1025 mutation in other diseases, 1025– 1027 properties, 1022 transcription regulation by protein, 1023, 1024 molecular diagnosis, 1025 Growth factors, see also specific factors, cell proliferation and differentiation, overview of control, 71 oncogenes, 74–77 Growth hormone (GH), acromegaly, hypersecretion, 444, 445 familial combined pituitary hormone deficiency, 455, 456 gene cluster, 452 insulin-like growth factor I in action, 451, 457 isolated growth hormone deficiency, biodefective growth hormone, 454, 455 clinical features, 451 diagnosis, 451 growth hormone-releasing hormone receptor mutations, 455 heredity, 451, 452 IGHD IA, gene defects, 452 IGHD IB, gene defects, 452, 453 IGHD II, gene defects, 453, 454 IGHD III, gene defects, 454 prevalence, 451 pituitary cell secretion, 443, 444 puberty role, 569 receptor, 452 regulation of secretion, 444 replacement therapy, 456, 457

H Hailey-Hailey disease, clinical features, 720, 721

diagnosis, 721 genetic basis of disease, 721 management/treatment, 721 molecular pathophysiology, 721 Hashimoto’s thyroiditis, autoantibodies, 470 prevalence, 469 HAV, see Hepatitis A virus Haw River syndrome, see Dentatorubropallidoluysian atrophy HBV, see Hepatitis B virus HCP, see Hematopoietic cell phosphatase HCV, see Hepatitis C virus HD, see Huntington’s disease HDV, see Hepatitis D virus HE, see Hereditary elliptocytosis Hearing loss, anatomy, 1093 classification of disorders, 1093 collagen disorders, Alport syndrome, 1096, 1097 osteogenesis imperfecta, 1097 Stickler syndrome, 1097 craniofacial disorders, 1097 diagnosis, 1093, 1094 management/treatment, 1097, 1098 melanocyte role, 1015 neurofibromatosis type 2, 1097 nonsyndromic hearing loss, dominant, 1094 mitochondrial, 1095 recessive, 1094, 1095 X-linked, 1095 Norries disease, 1097 Pendred syndrome, 1097 syndromic, Usher syndrome, 1095, 1096 Waardenburg syndrome, 1096 Heart disease, see Cardiac arrhythmia; Congenital heart disease; Coronary atherosclerosis; Dilated cardiomyopathy; Hypertrophic cardiomyopathy Hematopoiesis, bone marrow, cellular composition, 171, 172 growth factors, clinical applications, 176–178 receptors, 173–175 regulation of production, 173 signaling mechanisms, 175, 176 types, 172, 173 Hematopoietic cell phosphatase (HCP), activation in signal transduction, 176 Hemidesmosome-anchoring filament complex, defects in junctional epidermolysis bullosa, 723–725 structure, 723 Hemochromatosis, clinical manifestations, 378 diagnosis, 378

INDEX

genetic basis of disease, 378, 379 management/treatment, 379 molecular pathophysiology, 379 Hemoglobin, deficiencies, see Thalassemia, globin genes, 179 regulation of production, 179 structural variants, congenital Heinz body hemolytic anemia, 189 Hb C, 188 Hb D, 188, 189 Hb E, 188, 189 Hb M, 189 oxygen binding abnormalities, 189 sickle cell anemia, anti-sickling therapy, 188 bone marrow transplant, 188 clinical features, 186, 187 diagnosis, 187 gene mutation, 185–187 geographic distribution, 186 management in pregnancy, 188 sickle/β-thalassemia, 188 structure, 179 Hemophilia A, factor VIII, activation, 209 gene, defects, 209–211 structure, 209 genetic screening, 211, 212 management/treatment, 216, 217 molecular pathophysiology, 212 Hemophilia B, factor IX, activation, 212 gene, defects, 212 structure, 212 genetic screening, 212, 213 incidence, 212 management/treatment, 216, 217 molecular pathophysiology, 213, 214 Hepatitis A virus (HAV), discovery, 387 family classification, 387, 388 genome organization, 388 molecular pathophysiology, hepatitis, 394–396 oncogenesis, 396, 397 mutations, 397 structure, 388 vaccination, 399 Hepatitis B virus (HBV), diagnosis of infection, 391 discovery, 387 family classification, 387, 388 genome organization, 389 molecular pathophysiology, 394 mutations, 397, 398 proteins, 389, 390

replication, 390, 394–396 structure, 389 therapy, 398, 399 transmission, 388 vaccination, 399 Hepatitis C virus (HCV), diagnosis of infection, 392 discovery, 387 family classification, 387, 388 genome organization, 391, 392 molecular pathophysiology, hepatitis, 394–396 oncogenesis, 396 mutations, 398 replication, 391, 392, 396 risk factors, 391 structure, 391 therapy, 398, 399 vaccination, 399 Hepatitis D virus (HDV), diagnosis of infection, 393 discovery, 387 expression in tissues, 392, 393 family classification, 387, 388 genome organization, 392 molecular pathophysiology, hepatitis, 394–396 oncogenesis, 397 mutations, 398 replication, 393, 395, 396 structure, 392 transmission, 392 Hepatitis E virus (HEV), diagnosis of infection, 393, 394 discovery, 387 family classification, 387, 388 genome organization, 393 molecular pathophysiology, 394 structure, 393 transmission, 393 Hepatitis G virus (HGV), discovery, 387, 394 family classification, 387, 388 hepatitis GB virus homology, 394 transmission, 394 Hepatocyte, functions, 365 promoters, 365, 366 Hepatoerythropoietic porphyria, see Porphyria Hereditary coproporphyria, see Porphyria Hereditary elliptocytosis (HE), clinical features, 191 diagnosis, 191 gene mutations, 192, 193, 195 treatment, 195 Hereditary isolated glucocorticoid deficiency, clinical features, 497 diagnosis, 497 genetic basis of disease, 497, 498

1109 management, 498 Hereditary nephritis, see Alport syndrome Hereditary neuropathy with pressure palsies (HNPP), clinical features and pathogenesis, 922 molecular diagnosis, 925 peripheral myelin protein-22 gene deletion, 923 phenotype and gene, 885, 921, 925 reciprocal recombination, 923, 924 Hereditary nonpolyposis colorectal cancer (HNPCC), tumor suppressor gene mutations, 79, 80 Hereditary progressive dystonia (HPD), phenotype and gene, 879 Hereditary pyropoikilocytosis (HPP), clinical features, 191 diagnosis, 191 gene mutations, 192, 193 treatment, 195 Hereditary spherocytosis (HS), clinical features, 191 diagnosis, 191 gene mutations, 191, 192 treatment, 195 Hermansky-Pudlak syndrome (HPS), see Oculocutaneous albinism, type VIA HEV, see Hepatitis E virus Hexokinase, red cell enzymopathy, 198, 200 HGP, see Human Genome Project HGV, see Hepatitis G virus Hirschsprung’s disease (HSCR), phenotypes and genes, 886 Ret receptor gene mutations, 509 Histamine releasing factor (HRF), asthma role, clinical relevance, 314 purification of immunoglobulin Edependent histamine releasing factor, 313, 314 types, 312, 313 Histone, structure and function, 3, 26 HIV, see Human immunodeficiency virus HLA, see Human leukocyte antigen HNPP, see Hereditary neuropathy with pressure palsies HOCM, see Hypertrophic obstructive cardiomyopathy HOKPP, see Hypokalemic periodic paralysis Holt-Oram syndrome, gene mutation, 122 Hormone, see also specific hormones, classification, 419 feedback regulation, 419, 420 receptors, classification, 419 nuclear receptors, mechanism of action, 428

1110 physiologic effects, 429, 430 structure and classification, 428, 429 seven transmembrane domain receptors, 423, 424 signal transduction, calcium, 426, 427 cell cycle regulation, 427, 428 cyclic AMP, 427 mitogen-activated protein kinases, 425, 426 p21ras, 425, 426 single transmembrane domain receptors, cytokine receptor family, 422 guanylyl cyclase receptor family, 423 phosphotyrosine phosphatase family, 423 serine-threonine kinase receptor family, 423 tyrosine kinase receptor family, 421, 422 Hox, transgenic mouse studies, 100, 101, 121 HOX11, ectopic expression in lymphoma, 246 HPD, see Hereditary progressive dystonia HPP, see Hereditary pyropoikilocytosis HPS, see Hermansky-Pudlak syndrome HPV, see Human papilloma virus H-ras, see p21ras HRF, see Histamine releasing factor HS, see Hereditary spherocytosis HSAS, see Hydrocephalus due to stenosis of the aqueduct of Sylvius HSCR, see Hirschsprung’s disease HTLV1, see Human T-cell lymphotrophic virus-1 Human Genome Project (HGP), databases, 61, 62 DNA sequencing, 61 ethics, 62, 63 gene mapping, 59, 60 genomic cloning, 60, 61 goals, 59, 62 initiation, 59 scope of project, 59 Human immunodeficiency virus (HIV), clinical manifestations of infection, 294 genetic variability, 296, 297 life cycle, 295, 296 lymphoma association, 246, 247 mortality, 293 natural history of infection, 293, 294 protease inhibitors in therapy, 297 retrovirus classification, 293 reverse transcriptase inhibitors in therapy, 296 structure, 294, 295 vaccination, 251, 296 Human leukocyte antigen (HLA),

INDEX

bare lymphocyte syndrome defects, 286, 287 class I molecules, function, 274 nomenclature, 273–275 peptide binding, 274 structure, 273, 274 tissue distribution, 274, 275 class II molecules, function, 274 nomenclature, 273–275 peptide binding, 274, 275 structure, 273, 274 tissue distribution, 274, 275 disease associations, ankylosing spondylitis, 276 insulin-dependent diabetes mellitus, 276, 277 molecular mechanisms, 279, 280, 303 multiple sclerosis, 277 psoriasis, 795, 796 relative risk determinations, 276 rheumatoid arthritis, 277–280, 304 systemic lupus erythematosus, 304 systemic sclerosis, 304 genetic organization, 275 typing, 275, 276 Human papilloma virus (HPV), infection as cancer risk factor, 786 Human T-cell lymphotrophic virus-1 (HTLV1), lymphoma association, 247 Humoral immunity, see Antibody; B-cell Huntington’s disease (HD), clinical features, 891, 892 diagnosis, 893 epidemiology, 891 HD gene, huntingtin product expression, 896, 897 mutation, CAG repeats, 894, 895 mapping, 893, 894 molecular pathogenesis, 897 repeat length and phenotypic variability, 895, 896 linkage analysis, 893 pathology, 892, 893 phenotype and gene, 878, 914 screening and genetic counseling, 898, 899 transgenic mice studies, 105 Hydrocephalus due to stenosis of the aqueduct of Sylvius (HSAS), phenotype and gene, 886 Hydrogen peroxide, generation defects in thyroid deficiency, 465 11β-Hydroxylase deficiency, CYP11B1 mutations, 485, 486, 488, 555 CYP11B2 mutations, 487, 488, 555 epidemiology, 485, 554 prenatal diagnosis, 488

treatment, 488, 492 17α-Hydroxylase deficiency, 490 clinical features and diagnosis, 548, 596, 597 complete combined 17,20-lyase deficiency, 597 gene, 547 genetic and molecular pathophysiology, 548, 597 isolated 17,20-lyase deficiency, 598 partial combined 17,20-lyase deficiency, 597, 598 treatment, 598 17β-Hydroxylase deficiency, clinical features and diagnosis, 549, 598, 599 genes, 548, 549 genetic and molecular pathophysiology, 549, 599 isoenzymes, 598 treatment, 599 21-Hydroxylase deficiency, classic salt-wasting, 481, 482, 552 classic simple virilizing, 481, 552, 576 cryptic, 552 epidemiology, 482 gene defects, 483–486 genetic and molecular pathophysiology, 553, 554 hormonal diagnosis, 485, 552, 553 nonclassic, 482, 552 prenatal diagnosis, 485 treatment, 485, 491, 492 3β-Hydroxysteroid dehydrogenase deficiency, 488, 555 clinical features, 547, 595 diagnosis, 547, 595 genetic and molecular pathophysiology, 547, 595, 596 therapy, 596 Hyperkalemic periodic paralysis (HYPP), phenotype and gene, 883 Hypertension, frequency distribution of blood pressure, 146 gene therapy, 161, 162, 166 genetic basis, determination, allele-sharing method, 148 association study, 148, 149 candidate gene approach, 146, 147, 150 genome scanning, 147, 150 linkage analysis, 147, 148 evidence, 145, 146 nitric oxide, pulmonary hypertension treatment, 143 rat models, advantages in study, 149 gender effect, 151, 152 human clinical implications, 149, 150, 153

1111

INDEX

inbred strains, 149 phenotyping, 152, 153 quantitative trait loci, 150, 151 threshold in definition, 145, 146 Hypertrophic cardiomyopathy, childhood disease, epidemiology, 129 fatty acid oxidation defects, 129, 130 clinical features, 127 gene loci in familial disease, 128, 129 prevalence, 128 Hypertrophic obstructive cardiomyopathy (HOCM), phenotypic variability, 51 Hypobetalipoproteinemia, phenotype and gene, 881 Hypochondroplasia, clinical features, 1034 diagnosis, 1034 genetic basis of disease, 1034, 1035, 1037 management, 1037, 1038 Hypoglycemia, see also Persistent hyperinsulinemic hypoglycemia of infancy, differential diagnosis in children, 513, 514 Hypokalemic periodic paralysis (HOKPP), phenotype and gene, 883 Hypothalamic–pituitary hormone axis, feedback regulation of hormone release, 419, 420, 459 Hypothalamic–pituitary–gonadal axis, maturation in puberty, 570, 571 modulation in reproduction, 519, 524 Hypoxia, see Acute renal failure HYPP, see Hyperkalemic periodic paralysis

I IBD, see Inflammatory bowel disease IDDM, see Insulin-dependent diabetes mellitus Idiopathic thrombocytopenia (ITP), autoantibody, 251 IFN-γ, see Interferon-γ IGF-I, see Insulin-like growth factor I IGF-II, see Insulin-like growth factor II IGHD, see Isolated growth hormone deficiency Immunoglobulin, see Antibody Immunoprecipitation, protein–protein interaction analysis, 22, 23 Immunotherapy, rheumatoid arthritis, 305, 306 Inflammation, 267, 268 Inflammatory bowel disease (IBD), see also Crohn’s disease; Ulcerative colitis, clinical features, 407 complications, 408 diagnosis, 408, 409

epidemiology, 407 extraintestinal manifestations, 408 medical therapy, 410, 411 molecular pathophysiology, 409, 410 surgical treatment, 411 transgenic mouse studies, 101, 102 Inhibin, follicle-stimulating hormone, regulation of release, 521, 522 knockout mice studies, 99 puberty role, 570 synthesis sites, 522 transgenic mouse studies, 522 Insulin, see also Persistent hyperinsulinemic hypoglycemia of infancy, gene mutations in diabetes mellitus, type 2 insulin, 438 receptor, 438, 439 muscle effects, 857 receptor mutation in disease, 421, 422 resistance, see Diabetes mellitus, type 1 Insulin-dependent diabetes mellitus (IDDM), see Diabetes mellitus, type 1 Insulin-like growth factor I (IGF-I), expression in breast cancer, 628 growth hormone action, 451, 457 Insulin-like growth factor II (IGF-II), gene mutations in Wilms’ tumor, 689 rhabdomyosarcoma defects, 866, 867 Integrin, blocking antibodies in anti-inflammatory therapy, 648 ligand-binding domain, 647, 648 mediated phagocyte adhesion, 645 membrane cytoskeleton system, 853 mutations in junctional epidermolysis bullosa, 725 Interferon (IFN), 267, 268 Interferon-γ (IFN-γ), T-helper-1 response in asthma, generation, 323 recruitment of immune cells, 323, 324 Interleukins, 267, 268 Interleukin-2, knockout mice studies, 102 Interleukin-4, T-helper-2 response in asthma, generation, 320 recruitment of immune cells, 321 Interleukin-10, knockout mice studies, 102 International classification of diseases (ICD-10), affective disorder classification, 995, 996 Intravenous γ-globulin (IVIG), B-cell deficiency therapy, 286 Ischemia, see Acute renal failure Isolated growth hormone deficiency (IGHD), biodefective growth hormone, 454, 455

clinical features, 451 diagnosis, 451 growth hormone-releasing hormone receptor mutations, 455 heredity, 451, 452 IGHD IA, gene defects, 452 IGHD IB, gene defects, 452, 453 IGHD II, gene defects, 453, 454 IGHD III, gene defects, 454 prevalence, 451 ITP, see Idiopathic thrombocytopenia IVIG, see Intravenous γ-globulin

J Jackson-Weiss syndrome, clinical features, 1031 diagnosis, 1034 genetic basis of disease, 1034, 1035, 1037 management, 1037, 1038 JAK, see Janus kinase Janus kinase (JAK), cytokine receptors, 422 signal transduction, 175 Junctional epidermolysis bullosa, see Epidermolysis bullosa

K Kallmann’s syndrome, clinical features, 577, 601, 602 diagnosis, 602 genetic basis of disease, 577, 602 incidence, 577, 600 molecular pathophysiology, 602 phenotype and gene, 886 treatment, 577 Kennedy’s syndrome, see Spinobulbar muscular atrophy Keratin, assembly, 701, 702 function of keratin network, 703, 704 intermediate filaments, 701–703 knockout mice studies, 108 mutations, epidermal nevi, 713, 714 epidermolysis bullosa simplex, 696, 700, 701, 703 epidermolytic hyperkeratosis, 696, 707–710 intermediate filament diseases, 704, 705 screening, 710, 711 Kidney, see also Acute renal failure; Alport syndrome; Nephrogenic diabetes insipidus; Polycystic kidney disease, development, genes, 636–638 genetic cascades, 636 methodologies in study, 635, 636 processes in development, 636 stages, 635

1112 glomerular filtration barrier structure, 665, 666 neoplasms, see Renal cell carcinoma; Von Hippel Lindau syndrome; Wilms’ tumor, potassium channels, ATP-sensitive channels, 659–661, 664 classification, 659 inward-rectifying channels, 661–664 ROMK channels, 661–664 systemic sclerosis, involvement, 831, 836 Kit receptor, mutation studies in animal sex determination, 542 Klinefelter’s syndrome, diagnosis, 590 features, 564, 565, 578, 589, 590 genetic basis of disease, 590, 591 therapy, 591 46,XX male variant, clinical features, 591 diagnosis, 592 genetic basis of disease, 592 Koebner syndrome, see Epidermolysis bullosa simplex, simplex form K-ras, see p21ras Kuru, see also Prion, clinical manifestations, 928, 929 diagnosis, 929, 930 etiology, 927, 928, 930

L Lactase deficiency, clinical features, 412, 413 diagnosis, 413 epidemiology, 412, 413 genetic basis, 413 management, 413 molecular pathophysiology, 413 α2 Laminin, see Merosin Laminin, 5, defects in junctional epidermolysis bullosa, 724 Laurence-Moon-Biedl syndrome, hypogonadotropic hypogonadism, 603, 604 LDL, see Low-density lipoprotein Leber’s hereditary optic neuropathy (LHON), 944 Leigh’s disease, 943, 944 Leptin, 267, 268, 270 puberty role, 570 LESSD, see Lupus erythematosus, specific skin disease Leukemia, see Acute lymphoblastic leukemia; Acute myeloid leukemia; Chronic lymphocytic leukemia; Chronic myelogenous leukemia Leukocyte extravasation, acute renal failure role, 653, 654 homing, activated lymphocytes, 642, 643

INDEX

effector/memory cells, 645, 646 naive lymphocytes, 641, 642, 645 integrins, blocking antibodies in anti-inflammatory therapy, 648 ligand-binding domain, 647, 648 mediated phagocyte adhesion, 645 selectins, blocking antibodies in anti-inflammatory therapy, 648 effector/memory cell emigration, 645, 646 ligand-binding domain, 647 ligands, 644 lymphocyte homing, 646, 647 neutrophil influx contribution, 643 regulated expression, 645 structural and physical factors contributing to neutrophil adhesion, 643 stages in transendothelial migration of phagocytes, 641 Leydig cell, dysfunction, 588, 589, 591 hypoplasia, 542, 543, 593, 594 LH, see Luteinizing hormone LHON, see Leber’s hereditary optic neuropathy LHRH, see Luteinizing hormone-releasing hormone Li-Fraumeni syndrome, p53 mutations and breast cancer, 630 Lim-1, knockout mice studies, 100 Limb development, candidate genes, 1025, 1026 Limb-girdle muscular dystrophy, phenotypes and genes, 883 LINES, see Long interspersed repetitive elements Linkage analysis, Duchenne muscular dystrophy, 87 multiple endocrine neoplasia type 1, 47, 48 Lipoid congenital adrenal hyperplasia, see Congenital adrenal hyperplasia Liver, see also specific diseases, alcoholic liver disease, 371 amyloidosis, 370, 371 cell types and functions, 365–367 drug-induced liver disease, 371–373 embryogenesis, 365 hepatitis viruses in oncogenesis, 396, 397 Localized scleroderma, clinical features, 837 epidemiology, 837 genetic basis of disease, 837 molecular pathogenesis, 837 treatment, 837 LOH, see Loss of heterozygosity Long interspersed repetitive elements (LINES), 4, 44

Long QT syndrome, candidate genes, 157, 158 clinical features, 157 management/treatment, 158 Loss of heterozygosity (LOH), breast cancer, 630, 631 rhabdomyosarcoma, 866, 867 Louis-Barr syndrome, see Ataxia-telangiectasia Low-density lipoprotein (LDL), oxidation in coronary atherosclerosis, 136, 137 receptor, knockout mice studies, 103 Luciferase, reporter gene, 23, 24 Lung cancer, classification, 357 p53 mutations, 359, 360 retinoblastoma protein–E2F pathway inactivation, 357–359 Lupus erythematosus, specific skin disease (LESSD), classification, 821 clinical features, 821, 822 diagnosis, 822 genetic basis of disease, 823 immunopathology, 823 management/treatment, 826 molecular pathophysiology, 823–826 pathology, dermal changes, 823 dermal–epidermal junction changes, 822, 823 epidermal changes, 822 subcutaneous changes, 823 ultrastructural changes, 823 ultraviolet radiation effects, 824–826 Lupus, see Lupus erythematosus, specific skin disease; Systemic lupus erythematosus Luteinizing hormone (LH), gene mutations and hypogonadism, 602, 603 ovarian cycle control, 613, 614 ovary actions, 521 pituitary cell secretion, 443, 444, 519, 520 pituitary tumor secretion, 445, 446 puberty role, 571 receptor mutations, 578 regulation of secretion, 445, 446 testes actions, 521 Luteinizing hormone-releasing hormone (LHRH), embryology of secreting neurons, 600 gene mutation and hypogonadism, 602 17,20-Lyase, see 17α-Hydroxylase deficiency Lymphoma, B-cell lymphoma, BCL-1 translocation in mantle cell lymphoma, 245 BCL-2 translocation, 243–245

INDEX

BCL-6 translocation in diffuse large cell lymphoma, 245 molecular pathology, 241, 242 MYC gene deregulation, 242, 243 classification, 241 diagnosis, 247 epidemiology, 241 T-cell lymphoma, chimeric gene formation, 246 HOX11 ectopic expression, 246 molecular pathology, 245, 246 MYC gene deregulation, 246 t(2;5)(p23;q35) translocation, 246 virus association, Epstein-Barr virus, 241, 246, 247 human immunodeficiency virus, 246, 247 human T-cell lymphotrophic virus-1, 247

M M structure, see Muscle Machado-Joseph disease (MJD), phenotype and gene, 879, 897, 898, 914, 917 Mad cow disease, see Bovine spongiform encephalopathy Major histocompatibility complex, see Human leukocyte antigen Malaria, vaccination, 251 Malignant hyperthermia (MH), clinical features, 949 diagnosis, 952 diagnostic testing, 950 epidemiology, 949 gene, 857 incidence, 950 linkage analysis, 951 management/treatment, 952, 953 phenotypes and genes, 883 physiological basis of disease, 951, 952 RYR1 gene mutations, humans, 950, 951, 953 pigs, 950 Mantle cell lymphoma, see Lymphoma MAO-B, see Monoamine oxidase B MAPKs, see Mitogen-activated protein kinases MAS, see McCune-Albright syndrome Maturity onset diabetes of the young (MODY), see Diabetes mellitus, type 2 Mayer-Rokitansky-Küster-Hauser (MRKH) syndrome, 556 McCune-Albright syndrome (MAS), clinical features, 573, 574 gene mutations, 574 ovarian effects, 618 treatment, 575, 576 MCP-1, see Monocyte chemoattractant protein-1

Mdm2, knockout mice studies, 98, 99 Medulloblastoma, see Brain tumor MEF2, myogenesis role, 847, 848 Meiosis, see Cell cycle Melanin, biosynthesis, 737, 738, 743 deficiency, see Oculocutaneous albinism Melanocyte, hearing role, 1015 Melanoma, chromosomal aberrations, 789, 790 ras mutations, 790 MELAS, see Mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes MEN-1, see Multiple endocrine neoplasia type 1 MEN-2, see Multiple endocrine neoplasia type 2 Meningioma, see Brain tumor Menke’s syndrome, phenotype and gene, 881 Menopause, 618, 620 Merosin, deficiency in congenital muscular dystrophy, 862 MERRF, see Myoclonic epilepsy with ragged red fibers syndrome Methotrexate, psoriasis treatment, 799 MH, see Malignant hyperthermia β2-Microglobulin, knockout mice studies, 101 Microsatellite, detection, 19, 871 Miller-Dieker lissencephaly syndrome, phenotype and gene, 887 MITF, mutations in Waardenburg syndrome, 1017, 1018 Mitochondrial diabetes mellitus, see Diabetes mellitus, type 2 Mitochondrial disease, aging role, 946 autosomally transmitted multiple mitochondrial deletions, 943 chronic progressive external ophthalmoplegia, 943 inheritance of genetic diseases, 56 laboratory diagnosis, 943 Leber’s hereditary optic neuropathy, 944 Leigh’s disease, 943, 944 mitochondrial DNA, pathogenetic point mutations, 944 pathophysiology of mutations, 944, 945 structure and function, 941–943 mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes, 943 myoclonic epilepsy with ragged red fibers syndrome, 943 neurological manifestations, 942 systemic manifestations, 942, 945 treatment, 946

1113 Mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes (MELAS), 943 Mitogen-activated protein kinases (MAPKs), Rho regulation, 1044 signal transduction, 425, 426 Mitosis, see Cell cycle MJD, see Machado-Joseph disease MMNCB, see Multifocal motor neuropathy with conduction block Möbious syndrome, hypogonadotropic hypogonadism, 604 MODY, see Maturity onset diabetes of the young Monoamine oxidase B (MAO-B), defects in alcoholism, 1009 Monocyte chemoattractant protein-1 (MCP-1), role in coronary atherosclerosis, 138 Mood disorder, see Affective disorder Morphea, see Localized scleroderma MPZ, see P0 gene MRF4, transgenic mouse studies of myogenesis, 846, 847 MRKH, see Mayer-Rokitansky-KüsterHauser MS, see Multiple sclerosis Müllerian-inhibiting substance, see AntiMüllerian hormone Multifocal motor neuropathy with conduction block (MMNCB), diagnosis, 907 Multiple endocrine neoplasia type 1 (MEN-1), gene mutations in pituitary tumors, 447, 448 linkage analysis, 47, 48 oncogene, 80 phenotypic variability, 52 Multiple endocrine neoplasia type 2 (MEN-2), clinical features, 505 diagnosis, 505, 506 hyperparathyroidism management, 510 medullary thyroid cancer, 471, 472, 507–510 pheochromocytoma management, 510 Ret receptor gene, functions, 506, 507 mapping, 506 mutations, 77, 80, 480, 507–509 Multiple sclerosis (MS), human leukocyte antigen association, 277 Muscle, cancer, see Rhabdomyosarcoma, contraction, actomyosin ATP cycle, 854, 855 calcium regulation, 855, 856 myosin head as protein motor, 855 sliding filament model, 853, 854 differentiation, regulation in vitro, 841

1114 disorders, see specific diseases, insulin effects, 857 membrane cytoskeleton system, dystrophin, 853 integrin, 853 spectrin, 853 myogenesis, embryogenesis of skeletal muscle, 844–846 growth factor control, 842–844 MEF2 role, 847, 848 MyoD role, 841–844, 846 transgenic mouse studies, 846, 847 neuronal fibers, 857 proteins and nutrition, 856, 857 sarcomere, actin filament, 851, 852 M structure, 852 myosin filament, 851 Z structure, 852 sarcoplasmic reticulum, calcium pump, 852 calcium-release channel, 852 voltage-sensitive calcium channel, 852, 853 Muscular dystrophy, see also Becker muscular dystrophy; Congenital muscular dystrophy; Duchenne muscular dystrophy; Emery-Dreifuss muscular dystrophy; Fukuyama-type congenital muscular dystrophy; Limb-girdle muscular dystrophy; Myotonic dystrophy; Oculopharyngeal muscular dystrophy, calpain III deficiency, 862 classification and incidence, 859, 861 dominantly inherited dystrophies, 862 merosin deficiency in congenital disease, 862 Myc, activation in breast cancer, 627 deregulation in lymphoma, 242, 243, 246 N-myc, kidney development role, 637 Myf5, transgenic mouse studies of myogenesis, 846 Myocardial infarction, see Coronary atherosclerosis Myoclonic epilepsy with ragged red fibers syndrome (MERRF), 943 MyoD, basic helix-loop-helix, 842 knockout mice studies, 101 myogenesis role, 841–844, 846 phosphorylation, 843, 844 rhabdomyosarcoma expression, 865– 867 transgenic mouse studies of myogenesis, 846 Myoglobinuria, phenotype and gene in hereditary disease, 883 Myosin, see Muscle

INDEX

Myotonia congenita, phenotype and gene, 884 Myotonic dystrophy, CTG repeats in DM kinase gene, 859, 862 hypogonadism, 593 incidence, 859, 861 phenotype and gene, 884, 914, 917, 918

N ND, see Norrie disease Nemaline myopathy, phenotype and gene, 884 Nephrogenic diabetes insipidus, acquired disease, 672 autosomal recessive disease and aquaporin 2 mutations, 671, 672 congenital disease and vasopressin V2 receptor mutations, 669–671 Neurofibromatosis, animal models, 968 clinical features, 963 diagnosis, 963 hearing loss in type 2, 1097 incidence, 963 management/treatment, 968, 969 neurofibromin, functions, 967, 968 mutations in type 1, 964, 965 phenotypes and genes, 885, 963 schwannomin, functions, 968 mutations in type 2, 965–967 types, 963, 964 Neurofibromin gene (NF1), functions of protein, 967, 968 mutations in neurofibromatosis type 1, 964, 965 Neutrophil, emphysema role, cathepsin G, 342 elastase, 342 metalloproteinases, 342 proteinase 3, 342 serine proteinases, 341, 342 smoking-induced retention in lung, 341 NF1, see Neurofibromin gene NF2, see Schwannomin gene NIDDM, see Noninsulin-dependent diabetes mellitus Nijmegen breakage syndrome, 763, 764 Nitric oxide (NO), airway inflammation in asthma, 322, 323 cardiovascular therapy, 141 cell types in synthesis, 141, 142 guanylate cyclase activation, 141 pulmonary hypertension treatment, 143 septic shock response, 143, 144 vascular tone control, 142, 143 Nitric oxide synthase (NOS), catalysis, 141, 142 inhibitors, 142, 144

isoforms, 141 knockout mice studies, 142 regulation, 141–143 nM23, mutations in pituitary tumors, 448 NO, see Nitric oxide Nobel prizes, molecular genetics, 9, 10 Noninsulin-dependent diabetes mellitus (NIDDM), see Diabetes mellitus, type 2 Noonan syndrome, clinical features, 592, 593 diagnosis, 593 genetic basis of disease, 593 treatment, 593 Norrie disease (ND), hearing loss, 1097 phenotype and gene, 887 Northern blot, principle, 17 NOs, see Nitric oxide synthase NPHP1, defects in polycystic kidney disease, 677, 678 N-ras, see p21ras Nuclear run-on assay, transcription assay, 23 Nucleosome, structure and function, 3, 26 Nucleotide excision repair, defects in xeroderma pigmentosum, 756, 757 genes, 759 pathway, 754, 755

O OCA, see Oculocutaneous albinism Oculocutaneous albinism (OCA), candidate genes, overview, 697 diagnosis, 739, 740 management/treatment, 742 melanin biosynthesis, 737, 738 molecular pathophysiology, 742 type I, clinical features, 737 genetic basis, 740, 741 type II, clinical features, 737, 738 genetic basis, 741 type IV, clinical features, 738, 739 genetic basis, 741 type V, clinical features, 738 type VIA, clinical features, 738 genetic basis, 741, 742 Oculopharyngeal muscular dystrophy (OPMD), phenotype and gene, 884 Oligodendroglioma, see Brain tumor Oligonucleotide, radiolabeling, 11 synthesis, 14, 15 Oncogene, see also specific genes, classes and functions, growth factors, 74–77

INDEX

receptor tyrosine kinases, 77 signal transduction molecules, 77 transcription factors, 78 discovery, 73 viral oncogenes, 73 OP-1, kidney development role, 637 Opitz/GBBB syndrome, chromosome 22q11 deletion, 1081 diagnosis, 1082, 1083 management, 1083, 1084 OPMD, see Oculopharyngeal muscular dystrophy Orofacial clefting, candidate genes, 1088, 1089 chromosomal causes, 1089, 1090 developmental biology, 1087 nonsyndromic cleft lip and palate, 1087–1089 nonsyndromic cleft palate, 1089 sporadic conditions with clefting, 1090, 1091 syndromic cleft lip and palate, 1089 teratogenic causes, 1090 treatment, 1087 Osteogenesis imperfecta, hearing loss, 1097 Ovary, embryogenesis and dysgenesis, 527– 530, 611, 612, 619, 620 failure, 620, 621 follicle, atresia, 615–617 folliculogenesis, 612 selection and dominance, 614, 615 steroidogenesis, 614 gonadotropin actions, 521 granulosa cell tumors, 618 hormonal control of ovarian cycle, 613, 614 luteal regression, 618 luteal transition, 617, 618 McCune-Albright syndrome, 618 multifollicular ovulation, 618, 619 ovarian senescence, 618 ovulation, 617 polycystic ovary syndrome, 621–623

P P, mutations in oculocutaneous albinism, 741, 742 P0 gene (MPZ), mutation in CharcotMarie-Tooth disease, 924 p21ras, associated proteins in disease, 1039– 1044 gene types, 447, 626, 627 mutation, melanoma, 790 squamous cell carcinoma, 787 mutations in breast cancer, 627 mutations, 447 oncogene, 77, 81, 82

signal transduction, 175, 176 signal transduction, 425, 426 P450scc, see P450 side-chain cleavage enzyme P450 side-chain cleavage enzyme (P450scc), deficiency, clinical features, 594, 595 diagnosis, 595 genetic basis of disease, 595 therapy, 595 steroidogenesis role, 544, 547 P450arom, see Aromatase p53, apoptosis role, 72 cell cycle control, 67, 70, 359, 360 knockout mice studies, 98 mutations in cancer, 79–82, 359, 360, 471 brain tumors, 973, 974 breast cancer, 628, 629 melanoma, 790 rhabdomyosarcoma, 867 squamous cell carcinoma, 787, 788 WT-1 interactions with protein, 688, 689 PAM, see Potassium-aggravated myotonia Pancreas, see Acute pancreatitis; Pancreatic insufficiency Pancreatic insufficiency, clinical features, 404 cystic fibrosis association, 331, 404, 405 diagnosis, 404 genetic basis of disease, 404 incidence, 404 management/treatment, 405 molecular pathophysiology, 404, 405 Paramyotonia congenita (PC), phenotype and gene, 884 Parathyroid gland neoplasia, clonality, 478, 479 incidence, 478 molecular pathogenesis, 480 oncogenes, 479, 480 parathyroid hormone secretion, 480 tumor suppressor genes, 479 Parathyroid hormone (PTH), familial hypocalciuric hypercalcemia and hyperparathyroidism, 475–477 gene, 475 hyperparathyroidism management in multiple endocrine neoplasia type 2, 510 hypoparathyroidism, DiGeorge syndrome, 477 familial isolated hypoparathyroidism, 477, 478 polyglandular failure syndrome, 477 renal dysgenesis and sensorineural hearing loss, 477

1115 regulation of synthesis and secretion, 475 tumor secretion, 480 Paroxysmal nocturnal hemoglobinuria (PNH), clinical features, 227 diagnosis, 227 flow cytometry in diagnosis, 229 history of study, 227, 228 PIG-A, clonal expansion of mutated cells, 229–231 gene, locus, 227, 228 mutations, 228–230 glycosyl-phosphatidyl inositol anchor synthesis, 227–229 treatment, 231 Pax-2, kidney development role, 637 PAX3, mutations in Waardenburg syndrome, expression in embryogenesis, 1017 molecular pathology, 1017, 1018 type 1 disease, 1016, 1017 type 3 disease, 1017 myogenesis role, 846 rhabdomyosarcoma defects, 866 Pax-8, defects in thyroid deficiency, 469 PBC, see Primary biliary cirrhosis PC, see Paramyotonia congenita PCOS, see Polycystic ovary syndrome PCR, see Polymerase chain reaction PDGF, see Platelet-derived growth factor Pedigree, notation by genetic counselor, 90, 91 risk assessment, 92 Pelizaeus-Merzbacher disease, phenotype and gene, 881 Pemphigoid, see Bullous pemphigoid; Cicatricial pemphigoid; Pemphigoid gestationis Pemphigoid gestationis, clinical features, 817 diagnosis, 817 management, 819 Pemphigus foliaceus (PF), autoantibodies to Dsg proteins, 814– 816 clinical features, 811, 812 diagnosis, 812, 813 endemic form, 811 genetic basis of disease, 813, 814 management/treatment, 815, 816 molecular pathophysiology, 814, 815 Pemphigus vulgaris (PV), autoantibodies to Dsg proteins, 814– 816 clinical features, 811, 812 diagnosis, 812, 813 genetic basis of disease, 813, 814 molecular pathophysiology, 814, 815

1116 Pendred syndrome, clinical features and genetics, 465 hearing loss, 1097 Pentoxifylline, cystic fibrosis therapy, 336 Percutaneous translumenal angioplasty (PTCA), effectiveness in coronary atherosclerosis treatment, 139, 165 Peripheral myelin protein-22 (PMP22), Charcot-Marie-Tooth disease pathogenesis role, 922–925 deletion in hereditary neuropathy with pressure palsies, 923 Peroxisome proliferator-activating receptor (PPAR), structure and properties, 428, 429 Persistent hyperinsulinemic hypoglycemia of infancy (PHHI), clinical features, 513 diagnosis, 513 genetic basis of disease, 513–515 incidence, 513 management/treatment, 516 molecular pathophysiology, 515, 516 PF, see Pemphigus foliaceus Pfeiffer syndrome, clinical features, 1031, 1032 diagnosis, 1034 genetic basis of disease, 1034, 1035, 1037 management, 1037, 1038 PFIC, see Progressive familial intrahepatic cholestasis Phenotype, definition, 44 Phenylketonuria (PKU), gene mutations, 985 PHHI, see Persistent hyperinsulinemic hypoglycemia of infancy Phosphatidyl inositol 3 kinase, signal transduction, 176 Phosphofructokinase, red cell enzymopathy, 198–200 Phosphoglycerate kinase, red cell enzymopathy, 199, 200 Phospholipase A2 (PLA2), acute renal failure role, 653 Phosphotyrosine phosphatase receptor, 423 PIG-A, clonal expansion of mutated cells, 229– 231 gene, locus, 227, 228 mutations in paroxysmal nocturnal hemoglobinuria, 228–230 glycosyl-phosphatidyl inositol anchor synthesis, 227–229 Pit-1, autoregulation, 37 mutation and disease, 38 mutation in anterior pituitary failure, 446, 455, 456, 460, 462

INDEX

Pituitary, anterior pituitary failure, Pit-1 mutations, 446, 455, 456, 460, 462 Prop-1 mutations, 456 cell types and hormone secretion, 443, 444 Pituitary tumor, adrenocorticotropic hormone secretion, 443, 444 clonal origin of adenomas, 446, 447 G protein mutations, 447 growth factors in pathogenesis, 448, 449 growth hormone secretion, 444, 445 hypothalamic origin of adenomas, 447 p21ras mutations, 447 prolactinomas, 445 thyroid-stimulating hormone secretion and clinical manifestations, 445, 460 tumor suppressor gene mutations, 447, 448 PKD1, expression, 680 mutations in autosomal dominant polycystic kidney disease, 637, 678–680 polycystin product, 678–680 structure, 678 PKD2, mutations in autosomal dominant polycystic kidney disease, 637 protein product, 680–682 structure, 680, 681 PKU, see Phenylketonuria PLA2, see Phospholipase A2 Platelet-derived growth factor (PDGF), kidney development role, 637 knockout mice studies, 104 PMP22, see Peripheral myelin protein-22 PNH, see Paroxysmal nocturnal hemoglobinuria Po, knockout mice studies, 105 POF, see Premature ovarian failure Polycystic kidney disease, see also Autosomal dominant polycystic kidney disease, clinical features, 675–677 diagnosis, 677 genetic basis of disease, 677, 678 treatment, 682, 683 types, 675 Polycystic liver disease, autosomal dominant polycystic liver disease, gene loci, 370 cyst features, 370 treatment, 370 Polycystic ovary syndrome (PCOS), diagnostic criteria, 621

genetic studies, 622, 623 hyperinsulinemia association, 621 Polymerase chain reaction (PCR), dystrophin mutation testing, 83 lymphoma diagnosis, 247 neurological disease diagnosis, 888, 889 overview and applications, 15, 45 reverse transcriptase-PCR in mutation detection, 19, 45, 49 screening for genetic disease, 57, 871 Polymorphism, detection, microsatellites, 19 restriction fragment-length polymorphisms, 18 variable number tandem repeats, 19 role in human disease, 45 Polymyositis, clinical features, 301 diagnosis, 302 management/treatment, 306 POMC, see Pro-opiomelanocortin Porphyria, acute intermittent porphyria, clinical manifestations, 382 diagnosis, 382, 383 genetic basis of disease, 383 management/treatment, 383 molecular pathophysiology, 383 δ-aminolevulinic acid synthase induction, 382 hepatoerythropoietic porphyria, clinical manifestations, 383 diagnosis, 384 genetic basis of disease, 384 management/treatment, 384 molecular pathophysiology, 384 hereditary coproporphyria, 383 heredity, 381, 382 porphyria cutanea tarda, 383 variegate porphyria, 383 Potassium-aggravated myotonia (PAM), phenotype and gene, 884 Potassium channels, kidney, ATP-sensitive channels, 659–661, 664 classification, 659 inward-rectifying channels, 661–664 ROMK channels, 661–664 PPAR, see Peroxisome proliferatoractivating receptor Prader-Willi syndrome (PWS), albinism association, 1059 clinical features and course, 1053– 1055 diagnosis, differential diagnosis, 1055 molecular diagnosis, 1057, 1058 genetic basis of disease, 1056, 1057 genetic counseling, 1058 genomic imprinting, 1058–1060

1117

INDEX

hypogonadotropic hypogonadism, 603 incidence, 1053 management/treatment, 1055, 1056, 1060 mouse model, 1059 Precocious puberty, see Puberty Premature ovarian failure (POF), 620, 621 Presenelin genes, mutations in Alzheimer’s disease, 904, 905 Primary biliary cirrhosis (PBC), molecular basis of disease, 368, 369 treatment, 369 Primary familial glucocorticoid resistance (FGR), clinical features, 495 clinical management, 497 diagnosis, 495 genetic basis of disease, 495–497 Primary sclerosing cholangitis (PSC), clinical features, 369 molecular basis of disease, 369 Prion (PrP), diseases, see Bovine spongiform encephalopathy; Creutzfeldt-Jakob disease; Fatal familial insomnia; Gerstmann-Straüssler-Scheinker disease; Kuru, gene, mutations and genetic linkage, 931– 933 polymorphisms, 933 structure and organization, 928 prenatal screening for mutations, 937 protein X binding, 934, 936, 938 PrPSC, formation, 934–936 prevention of formation, 937, 938 purification, 928 scrapie, prion discovery, 927, 928 resistant breeds of sheep, 938 PRL, see Prolactin Progesterone, ovarian cycle control, 613, 614 receptor expression in breast cancer, 626, 631 Programmed cell death, see Apoptosis Progressive familial intrahepatic cholestasis (PFIC), classification, 368 molecular basis of disease, 368 Prolactin (PRL), pituitary cell secretion, 443, 444 prolactinomas, 445 receptor, 445 regulation of secretion, 445 Promoter, binding protein analysis, 27, 28 enhancers, 27 TATA box, 26–29

transcription start site determination, 26, 27 Pro-opiomelanocortin (POMC), pituitary cell secretion, 443, 444 Prop-1, mutation in anterior pituitary failure, 456 Prostacyclin, vasoactivity, 142 Protein C, deficiency, 222–224 function, 222, 225 structure, 222 Protein S, deficiency, 224 function, 222 structure, 222 Protein truncation test (PTT), mutation detection, 19, 20, 49 Protein X, prion binding, 934, 936, 938 Protein–protein interaction, farwestern blotting, 22 immunoprecipitation, 22, 23 yeast two-hybrid system, 23 PrP, see Prion PSC, see Primary sclerosing cholangitis Pseudohermaphroditism, 46,XX pseudohermaphroditism, aromatase deficiency, 555, 556 clinical features of fetal androgenization, 552 11β-hydroxylase deficiency, 554, 555 21-hydroxylase deficiency, 552–554 Mayer-Rokitansky-Küster-Hauser syndrome, 556 46,XY pseudohermaphroditism, see also specific enzyme deficiencies, androgen insensitivity syndromes, 550, 551 anti-Müllerian hormone defects, 551, 552 causes, 542 clinical features, general, 542 Leydig cell hypoplasia, 542, 543 testosterone biosynthesis defects, 543–549, 594–599 Pseudohypoaldosteronism, clinical manifestations, 499, 500 genetic basis of disease, 500 management, 500 Psoriasis, clinical features, arthritis, 793, 794 course, 794 nails, 793 skin, 793 definition, 793 diagnosis, 794 genetic basis of disease, epidemiology, 795 human leukocyte antigen associations, 795, 796 juvenile-onset psoriasis, 795

non-HLA genes, 796 molecular pathophysiology, lesions, 796 T-cell role, 797, 798, 800 transgenic mouse studies, 108 PTCA, see Percutaneous translumenal angioplasty PTH, see Parathyroid hormone PTT, see Protein truncation test Puberty, contrasexual development, 576 delayed puberty, causes, 576 constitutional delay, 576 hypergonadotropic hypogonadism, gonadotropin receptor mutations, 578 Klinefelter’s syndrome, 578 Turner’s syndrome, 578, 579 hypogonadotropic hypogonadism, central nervous system disorders, 578 functional gonadotropin deficiencies, 578 isolated follicle-stimulating hormone deficiency, 578 isolated gonadotropin deficiency, 578 Kallmann’s syndrome, 577, 600– 602 overview, 576, 577 hormonal control, 569, 570 hypothalamic–pituitary–gonadal axis maturation, 570, 571 normal development, 569, 570 precocious puberty, classification, 572 features, 571 general evaluation, 572 gonadotropin-dependent, pathophysiology, 572, 573 treatment, 573 gonadotropin-independent, ectopic hormone production, 575 familial male precocious puberty, 575 McCune-Albright syndrome, 573– 575 treatment, 575, 576 Pulmonary hypertension, systemic sclerosis, 831, 836 Purine nucleoside phosphorylase, deficiency, 287 PV, see Pemphigus vulgaris PWS, see Prader-Willi syndrome Pyruvate kinase, red cell enzymopathy, 199, 200

R RA, see Rheumatoid arthritis Radiolabeling, DNA, 11 RAR, see Retinoic acid receptor

1118 Ras, see p21ras Raynaud’s phenomenon, systemic sclerosis, 830, 831 Rb mutations, breast cancer, 629 cancer pathogenesis role, 956, 957, 960 lung cancer, 357–359 pituitary tumors, 448 protein, see Retinoblastoma protein, retinoblastoma, 73, 74, 79 cell cycle regulation, 958, 959 discovery, 956, 957 effects on protein expression and function, 958 terminal differentiation role, 959 transcription regulation, 958 types, 957, 958 RCC, see Renal cell carcinoma Reactive oxygen species (ROS), acute renal failure role, 653 Receptor tyrosine kinases, oncogenes, 77 Recombinant protein expression, adenovirus vectors, 20, 21 baculovirus expression system, 20 Escherichia coli expression system, 20 mammalian cell line expression, 20 translation, in vitro, 20 Red blood cell, enzymopathies, consequences, 197 cytochrome β5 reductase, 200, 204, 205 glucose 6-phosphate dehydrogenase, clinical features, 199, 202 diagnosis, 202 epidemiology, 199 genetic basis, 203 management, 203, 204 molecular pathophysiology, 203 glutathione synthetase, 200, 204 glycolytic enzymopathies, aldolase, 199, 200 clinical features, 197 diagnosis, 197, 198 diphosphoglycerate mutase, 199, 200 genetic basis, 198 glucose 6-phosphate isomerase, 198, 200 hexokinase, 198, 200 management, 198 molecular pathophysiology, 198 phosphofructokinase, 198–200 phosphoglycerate kinase, 199, 200 prevalence, 197, 200 pyruvate kinase, 199, 200 triosephosphate isomerase, 199, 200 membrane,

INDEX

disorders, see Hereditary elliptocytosis; Hereditary pyropoikilocytosis; Hereditary spherocytosis, proteins and genes, 191, 193 structure, 191 5α-Reductase deficiency, clinical features and diagnosis, 549, 550, 584 enzyme structure and mechanism, 584, 585 genetic and molecular pathophysiology, 550, 585, 586 isoenzymes, 584, 585, 599 Renal cell carcinoma (RCC), gene locus, 689 hereditary disease, see Von Hippel Lindau syndrome Renin, hypertension role, 146, 151, 153 Reporter gene, transcription assays, 23, 24 types, 23, 24 Reproduction, regulation, gonads, 520–523 hormones, see specific hormones, hypothalamic–pituitary–gonadal axis modulation, 524 hypothalamus, 519 integration of signals, 523 pituitary, 519, 520 Restriction endonuclease, molecular biology applications, 9, 45 Restriction fragment-length polymorphism (RFLP), detection, 18, 45 Ret, functions, 506, 507 gene, mapping, 506 mutations in Hirschsprung disease, 509 mutations in multiple endocrine neoplasia type 2, 77, 80, 480, 507–509 kidney development role, 637 mutations in cancer, 77 Retinitis pigmentosa, phenotypes and genes, 880 Retinoblastoma, clinical features, 955, 956 diagnosis, 956 incidence, 955 management/treatment, 960, 961 phenotype and gene, 885 Rb mutations, 73, 74, 79 cell cycle regulation, 958, 959 discovery, 956, 957 effects on protein expression and function, 958 terminal differentiation role, 959 transcription regulation, 958 types, 957, 958 screening, 961, 962 two-hit hypothesis in pathogenesis, 955

Retinoblastoma protein (Rb), apoptosis role, 72 cell proliferation and differentiation, overview of control, 71, 72, 74, 357, 358 domains, 357 gene, see RB, knockout mice studies, 99 myogenesis role, 844 phosphorylation, 71, 72, 74, 358, 359, 844 protein–protein interactions, 358, 359 Retinoic acid receptor (RAR), mutations in disease, 430 structure and properties, 428, 429 Retinoid X receptor (RXR), structure and properties, 428, 429 Retinoids, psoriasis treatment, 798, 799 Retrovirus, gene therapy application, 776, 777 Reverse transcriptase, molecular biology applications, 11 Reye’s syndrome, epidemiology and histopathology, 372 RFLP, see Restriction fragment-length polymorphism Rhabdomyosarcoma, histology, 865 incidence, 865 molecular genetics, alveolar rhabdomyosarcoma, 866 embryonal rhabdomyosarcoma, 866, 867 myogenic factor expression, 865–867 origin, 867 p53 mutations, 867 treatment, 867 Rheumatoid arthritis (RA), 269, 270 autoantibody, 251, 256 clinical features, 299 diagnosis, 301 human leukocyte antigen association, 277–280, 304 management/treatment, 305, 306 Rho, actin cytoskeleton regulation, 1043, 1044 mitogen-activated protein kinase cascade role, 1044 morphogenesis role, 1044 mutation in Aarskog-Scott syndrome, 1039–1043 Ring X syndrome, 566, 567 RNA, gel electrophoresis, 15, 16 isolation, messenger RNA, 11 total cellular RNA, 11 Northern blotting, 17 reverse transcriptase-PCR in mutation detection, 19 RNA polymerase,

INDEX

transcription factors, 28 types, 28 ROS, see Reactive oxygen species RXR, see Retinoid X receptor Ryanodine receptor gene (RYR1), mutations, central core disease, 951 malignant hyperthermia, humans, 950, 951, 953 pigs, 950 RYR1, see Ryanodine receptor gene

S Saddan dysplasia, clinical features, 1034 diagnosis, 1034 genetic basis of disease, 1034, 1035, 1037 management, 1037, 1038 Sarcoglycan, deficiency in disease, 860 Sarcomere, see Muscle Sarcoplasmic reticulum, see Muscle SCF, see Stem cell factor Schizophrenia, candidate genes, 991 diagnosis, 993 genetic counseling, 993 heredity, adoption studies, 990 family studies, 989 twin studies, 989, 990 neurobiologic spectrum disorders, 992, 993 psychiatric spectrum disorders, 991, 992 quantitative models of genetic transmission, 990, 991 treatment, 993 Schwannomin gene (NF2), functions of protein, 968, 969 mutations in neurofibromatosis type 2, 965–967 SCID, see Severe combined immunodeficiency Scleroderma, see Localized scleroderma; Systemic sclerosis SCS, see Slow channel syndrome Selectins, blocking antibodies in anti-inflammatory therapy, 648 effector/memory cell emigration, 645, 646 ligand-binding domain, 647 ligands, 644 lymphocyte homing, 646, 647 neutrophil influx contribution, 643 regulated expression, 645 structural and physical factors contributing to neutrophil adhesion, 643 Sepsis, 269, 270 Septic shock, nitric oxide response and inhibitor therapy, 143, 144

Serine-threonine kinase receptor, 423 Severe combined immunodeficiency (SCID), 270, 271 see X-linked severe combined immunodeficiency Sex determination, definition, 527 disorders, clinical features, 535, 536 DAX-1 mutations and clinical features, 538, 539, 564 diagnosis, 556–558 Kit receptor/steel factor mutation studies in animals, 542 management, gonadectomy, 558 hormone replacement therapy, 558 sex assignment, 558 overview, 533, 534 sex reversal, 534, 535 SF-1 mutation studies in animals, 541, 542 SOX9 mutations and clinical features, 539, 564 SRY mutations, 536, 538, 563, 564 true hermaphroditism, 534, 535 WT1 mutations and clinical features, 539–541 genes, 530, 532 testes determination, 530, 532 Sex differentiation, definition, 527 diagnosis of disorders, 556–558 genes, 532, 533 management of disorders, gonadectomy, 558 hormone replacement therapy, 558 sex assignment, 558 overview of disorders, 533, 534 46,XX pseudohermaphroditism, aromatase deficiency, 555, 556 clinical features of fetal androgenization, 552 11β-hydroxylase deficiency, 554, 555 21-hydroxylase deficiency, 552–554 Mayer-Rokitansky-Küster-Hauser syndrome, 556 46,XY pseudohermaphroditism, see also specific enzyme deficiencies, androgen insensitivity syndromes, 550, 551 anti-Müllerian hormone defects, 551, 552 causes, 542 clinical features, general, 542 Leydig cell hypoplasia, 542, 543 testosterone biosynthesis defects, 543–549, 594–599 SF-1, see Steroidogenic factor-1 SH2 domain, tyrosine kinase receptors, 421

1119 Shear stress, role in coronary atherosclerosis, 138 Short interspersed repetitive elements (SINES), 4, 44 SINES, see Short interspersed repetitive elements Site-directed mutagenesis, overview, 15 Sjögren’s syndrome (SS), clinical features, 301 diagnosis, 302, 303 management/treatment, 306 Skin, see also Keratin, barrier function, 775 culture models, 775 disorders, see specific diseases, gene therapy, acquired disorders, 778 dystrophic epidermolysis bullosa, 733 gene transfer for systemic disorders, metabolic processing of circulating substrate, 778, 779 systemic delivery, 778 vaccination, 778 inherited disorders, 777, 778 renewal, 775 SLE, see Systemic lupus erythematosus Sliding filament model, muscle contraction, 853, 854 Slow channel syndrome (SCS), phenotype and genes, 882 SMA, see Spinal muscular atrophy SMC, see Smooth muscle cell Smooth muscle cell (SMC), role in coronary atherosclerosis, 137 Somatic Crossover Point Mapping, candidate gene identification, 765, 766 Southern blot, dystrophin mutation testing, 84, 85 lymphoma diagnosis, 247 mutation detection, 49 principle, 16, 17 SOX9, mutations and clinical features, 539, 564 Spastic paraplegia, familial, pathogenesis and genetics, 910 pathology, 909 phenotypes and genes, 880 Spectrin, membrane cytoskeleton system, 853 Spermatogenesis, final phase of maturation, 588 hormonal control, 587 meiotic maturation, 588 spermiogenesis, 588 stem-cell renewal by mitosis, 587 Spinal muscular atrophy (SMA), clinical features, 908 pathogenesis and genetics, 910, 914 phenotypes and genes, 882 Spinobulbar muscular atrophy, diagnosis, 907

1120 pathogenesis and genetics, 910, 914, 917 phenotype and gene, 882, 897 Spinocerebellar ataxia, ataxin 1 characterization, 916 CAG repeat expansion in type 1, 914– 916 clinical findings, 914 pathogenesis of neurodegeneration, 916, 917 phenotypes and genes, 879, 897, 914 type 3 disease, 917, 919 type 6 disease, 919 Sporadic generalized glucocorticoid resistance, see Primary familial glucocorticoid resistance Squamous cell carcinoma, molecular pathogenesis, 787, 788 risk factors, human papilloma virus infection, 786 immune suppression, 785, 786 inherited cancer syndromes, 785 ultraviolet radiation, 785 SRY, mutations in sex determination disorders, 536, 538, 563, 564 sex determination, 532, 536, 563 SS, see Sjögren’s syndrome StAR, see Steroidogenic acute regulatory protein Startle disease, phenotype and gene, 884 STAT, phosphorylation by Janus kinases, 422, 425 signal transduction, 176 Steel factor, see Stem cell factor Stem cell, characteristics in bone marrow, 171, 172 growth factors, clinical applications, 176–178 receptors, 173–175 regulation of production, 173 signaling mechanisms, 175, 176 types, 172, 173 Stem cell factor (SCF), mutation studies in animal sex determination, 542 ovarian ontogeny role, 611 receptor, 175 Steroidogenic acute regulatory protein (StAR), defects and clinical features, 544 Steroidogenic factor-1 (SF-1), developmental role, 429, 503 knockout mice studies, 107 sex determination, 530, 541, 542 Stickler syndrome, hearing loss, 1097 orofacial clefting, 1089 Streptokinase, myocardial infarction treatment, 139

INDEX

Sulfasalazine, psoriasis treatment, 799 Sulfonylurea receptor (SUR1), gene mutation in persistent hyperinsulinemic hypoglycemia of infancy, 513–515 Superoxide dismutase, mutations in amyotrophic lateral sclerosis, 909, 910 Supravalvar aortic stenosis (SVAS), gene mutation, 122 SUR1, see Sulfonylurea receptor Surfactant protein, knockout mice studies, 102 SVAS, see Supravalvar aortic stenosis Swyer syndrome, 46,XY pure gonadal dysgenesis, 592 Syndrome of apparent mineralocorticoid excess, clinical manifestations, 500 genetic basis of disease, 500, 501 management, 501 Systemic lupus erythematosus (SLE), autoantibody, 251 clinical features, 300 diagnosis, 301, 302 human leukocyte antigen association, 304 management/treatment, 306 Systemic sclerosis, cardiac involvement, 831, 836 clinical features, 300, 301, 829–831 diagnosis, 302, 831 epidemiology, 829 fibrotic manifestations, 831 genetic basis of disease, 831, 832 human leukocyte antigen association, 304 management/treatment, 306, 835–837 molecular pathophysiology, cytokine roles, 835 extracellular matrix abnormalities, 833, 835 growth factor roles, 835 immunologic aspects, 833 inflammatory aspects, 833 vascular lesions, 832, 833 pulmonary hypertension, 831, 836 Raynaud’s phenomenon, 830, 831 renal involvement, 831, 836 types, 829 vascular manifestations, 830, 831

T TATA box, binding proteins, 26–29 T cell, antigen presentation, see Human leukocyte antigen, events in immune response, 284 immune deficiency, carrier state and prenatal diagnosis, 285, 286 combined B-cell diseases,

adenosine deaminase deficiency, 287, 288 ataxia-telangiectasia, 288 bare lymphocyte syndrome, 286, 287 purine nucleoside phosphorylase deficiency, 287 Wiscott-Aldrich syndrome, 287 X-linked severe combined immunodeficiency, 286 defects in disorders, 284 DiGeorge anomaly, 288 laboratory assessment, 283, 285 signal transduction defects, 289 surface markers, 285 symptoms, 283 T-cell receptor defect, 288, 289 therapy, 286 ZAP-70 deficiency, 288 lymphoma, see Lymphoma, management/treatment, 798–800 psoriasis role, 797, 798, 800 regulation of immune response, 257 T-helper responses, see also Asthma, Th2 cell response in allergic diseases, 804, 805 tolerance in autoimmunity prevention, 304, 305 tek, knockout mice studies, 103 Testes, bilateral anorchia, 593 cryptorchidism, 608, 609 determination, 530, 532 embryogenesis, 527–530, 587 eugonadotropic germinal cell dysfunction, 605, 606 gonadotropin actions, 521 histology, 587 hypergonadotropic hypogonadism, 578, 579, 589–602 hypogonadism, 588 hypogonadotropic hypogonadism, 576, 578, 602–604 idiopathic germinal cell failure, 606–608 Leydig cell, dysfunction, 588, 589, 591 hypoplasia, 542, 543, 593, 594 spermatogenesis, final phase of maturation, 588 hormonal control, 587 meiotic maturation, 588 spermiogenesis, 588 stem-cell renewal by mitosis, 587 systemic disorders in infertility, 605 Testicular feminization, see Androgen receptor, insensitivity syndromes Testosterone, conversion to dehydrotestosterone, 429, 430 synthesis, 481, 482, 543, 549 testes development role, 529 TGF-α, see Transforming growth factor-α

INDEX

TGF-β, see Transforming growth factor-β Thalassemia, acquired Hb H disease, 185 classification, 179, 180 geographic distribution, 179, 180 hemoglobinopathies, 189, 190 management, 185 α-thalassemias, clinical features, 183 diagnosis, 183, 184 genetic basis, 184 mental retardation syndromes, 184, 185 pathophysiology, 184 β-thalassemias, clinical features, 180, 181 δβ-thalassemia, 183 diagnosis, 181 γδβ-thalassemia, 183 genetic basis, 181 Hb E/β-thalassemia, 182, 183 hereditary persistence of fetal hemoglobin syndromes, 183 intermediate forms, 183 pathophysiology, 182 Thanatophoric dysplasia, clinical features, 1034 diagnosis, 1034 genetic basis of disease, 1034, 1035, 1037 management, 1037, 1038 Thrombosis, antithrombin, deficiency, incidence, 221 type I deficiency, 222 type II deficiency, 222 function with heparin sulfate in anticoaglation, 220 gene structure, 220, 221 structure, 220 clinical presentation of venous thrombosis, 219 epidemiology, 219 factor V mutation, 224, 225 gene therapy, 162, 166, 167 laboratory assessment, 220 management, acute venous thromboembolism, 224, 225 asymptomatic individuals, 225 pregnancy, 225 protein C, deficiency, 222–224 function, 222, 225 structure, 222 protein S, deficiency, 224 function, 222 structure, 222 Thyroglobulin, defects in thyroid deficiency, 465–466

Thyroid cancer, clonal analysis, 470 growth factors and oncogenes, 470, 471 incidence, 470 multiple endocrine neoplasia and medullary thyroid cancer, gene mutations, 471, 472, 507, 510 management, 509, 510 multistep pathogenesis, 81 radiation induction, 470 recombinant thyroid-stimulating hormone in diagnosis, 472 transgenic models, 472 Thyroid hormone, classification, 459 Grave’s disease, 469, 470 Hashimoto’s thyroiditis, 470 hypothalamic, pituitary hormone axis in regulation, 420, 445, 459 mutations in thyroid hormone resistance, 430, 467, 468 peripheral monodeiodination, 467 serum-binding protein defects in deficiency, albumin, 467 thyroxine-binding globulin, 466, 467 transthyretin, 467 signs and symptoms of dysfunction, 459, 462 structure and properties, 428, 429 synthesis and deficiency, dehalogenase defects, 466 hydrogen peroxide generation defects, 465 iodide transport defects, 465 overview, 464 Pendred’s syndrome, 465 thyroglobulin defects, 465, 466 thyroperoxidase defects, 465 thyroid-stimulating hormone secretion and clinical manifestations, 445, 460 thyrotropin-releasing hormone deficiency, 459, 460 transcription factor defects in deficiency, Pax-8, 469 thyroid transcription factor-1, 468 thyroid transcription factor-2, 468, 469 Thyroid-stimulating hormone (TSH), b chain mutations in deficiency, 462 Pit-1 mutations in deficiency, 460, 462 pituitary tumor secretion and clinical manifestations, 445, 460 receptor mutations, 452, 464 Thyroperoxidase, defects in thyroid deficiency, 465 Thyroxine-binding globulin, defects in thyroid deficiency, 466, 467

1121 Tissue plasminogen activator (tPA), myocardial infarction treatment, 139 Torsion dystonia, phenotype and gene, 879 TP53, see p53 tPA, see Tissue plasminogen activator Transcription factors, coactivators and corepressors, 29, 30 combinations for tissue-specific expression, 35, 36 dimerization, 34, 35 DNA binding and coactivator/corepressor recruitment, 36, 37 DNA sequence-specific transcription factors, 29 general transcription factors, 28, 29 modular functional domains, 32, 33 mutations and rearrangements in human disease, 38, 40 oncogenes, 78 Transcriptional activation, chromatin reorganization, 30 gene autoregulation, 37 phosphorylation and dephosphorylation, 30, 31 repressor protein displacement, 32 Transcriptional repression, gene autoregulation, 37, 38 mechanisms, 32 Transforming growth factor-α (TGF-α), expression in breast cancer, 627 overexpression in transgenic mice, 104 Transforming growth factor-β (TGF-β), expression in breast cancer, 628 knockout mice studies, 101 myogenesis role, 844 Transgenic mouse, applications, cardiovascular biology, 102, 103 cell cycle control, 98, 99 dermatopathology, 108, 109 developmental biology, 99, 101 endocrinology, 106, 107 hematology/immunology, 101, 102 metabolic disorders, 107 muscle biology, 101, 846, 847 neurobiology, 104, 106 pulmonary biology, 102 renal biology, 103, 104 reproductive biology, 107 production, 97, 98 Transthyretin, defects in thyroid deficiency, 467 Transthyretin, mutations in amyloidosis, 370, 371 Treacher Collins syndrome, hearing loss, 1097 Trinucleotide repeat, analysis in fragile X syndrome, 86, 87

1122 CAG repeat disorders, 894, 898 classification of diseases, 877, 888 neurological disorder overview, 877, 888, 913 role in disease, 55, 56, 105 sequences in human genome, 913, 919 Triosephosphate isomerase, red cell enzymopathy, 199, 200 Triple A syndrome, see Hereditary isolated glucocorticoid deficiency Trisomy, see Down syndrome α Tropomyosin, mutation in hypertrophic cardiomyopathy, 128, 129 TRP-1, mutations in oculocutaneous albinism, 741 True hermaphroditism, see Sex determination TSH, see Thyroid-stimulating hormone Tuberous sclerosis, phenotypes and genes, 887 Tumor necrosis factor, 267, 268 Tumor suppressor gene, see also specific genes, breast cancer, 628, 629 classes and functions, apoptosis-related proteins, 80 cell-cycle control proteins, 79 cell-surface proteins, 78, 79 DNA repair proteins, 79, 80 transcription factors, 79 discovery, 73, 74 mutations in pituitary tumors, 447, 448 Turner’s syndrome, incidence, 565 male syndrome, see Noonan syndrome, pathophysiology, 565, 566 phenotype, 565, 578, 579 Twinning, genetic control of ovulation, 618, 619 Tyrosinase, mutations in oculocutaneous albinism, 740, 742 Tyrosine kinase receptor, dimerization, 421

U UDP-glucuronosyltransferase, 1 deficiency, see Crigler-Najjar syndrome; Gilbert’s syndrome Ulcerative colitis, complications, 408 diagnosis, 408, 409 extraintestinal manifestations, 408 medical therapy, 410, 411 molecular pathophysiology, 409, 410 signs and symptoms, 408 surgical treatment, 411 Ultraviolet radiation (UV), effects in lupus erythematosus, specific skin disease, 824, 826 psoriasis treatment, 799

INDEX

skin cancer risk factor, 785 Uniparental disomy, see Genomic imprinting Urate oxidase, knockout mice studies, 107 Usher syndrome, hearing loss, 1095, 1096 UV, see Ultraviolet radiation

V V2R, see Vasopressin V2 receptor VCAM-1, see Vascular cell adhesion molecule-1 Van der Woude syndrome (VWS), orofacial clefting, 1089 Variable number tandem repeat (VNTR), detection, 19 Variegate porphyria, see Porphyria Vas deferens, see Congenital bilateral absence of the vas deferens Vascular cell adhesion molecule-1 (VCAM-1), role in coronary atherosclerosis, 137, 138 Vasopressin V2 receptor (V2R), intracellular trafficking, 671 mutations in congenital nephrogenic diabetes insipidus, 669, 671 structure, 669 Vav, signal transduction, 176 VCF, see Velocardiofacial syndrome VDR, see Vitamin D receptor Velocardiofacial syndrome (VCFS), chromosome 22q11 deletion, 1079, 1081 clinical features, 1080, 1089, 1090 diagnosis, 1082, 1083 management, 1083, 1084 Venous thrombosis, see Thrombosis VHL, see Von Hippel Lindau syndrome Vitamin D, psoriasis treatment, 798, 799 Vitamin D receptor (VDR), mutations in disease, 430 structure and properties, 428, 429 VNTR, see Variable number tandem repeat Von Hippel Lindau syndrome (VHL), mutations in VHL, 79, 690 phenotypes and genes, 885 von Willebrands disease (VWD), diagnosis, 214 management/treatment, 216, 217 molecular pathology, type 1, 216 type 2, 215 type 2A, 215, 216 type 2B, 216 type 2M, 216 type 2N, 216 type 3, 216 von Willebrands factor gene structure, 214, 215 VW, see Van der Woude syndrome

VWD, see von Willebrands disease W Waardenburg syndrome, clinical features of types, 1015 diagnosis, criteria, 1015, 1016 differential diagnosis, 1016 hearing loss, 1096 management/treatment, 1018 MITF mutations, molecular pathology, 1018 type 2 disease, 1017 PAX3 mutations, expression in embryogenesis, 1017 molecular pathology, 1017, 1018 type 1 disease, 1016, 1017 type 3 disease, 1017 phenotypes and genes, 887 Walker-Warburg syndrome (WWS), phenotype and gene, 887 Weber-Cockayne syndrome, see Epidermolysis bullosa simplex, simplex form Wilms’ tumor, familial disease, 689 incidence, 685 insulin-like growth factor 2 gene mutations, 689 tumor suppressor gene, see WT-1 two-hit model of tumorigenesis, 685 Wilson disease, clinical manifestations, 375 diagnosis, 375 genetic basis of disease, 375, 377 management/treatment, 377 molecular pathophysiology, 377 phenotype and gene, 880 Wiscott, Aldrich syndrome, molecular basis of immunodeficiency, 287 Wnt-4, kidney development role, 637 WT-1, inactivation in renal tumors, 78, 79, 688 isolation, 685, 686 kidney development role, 636, 685, 687, 688 knockout mice studies, 100, 104, 107 mutations and clinical features in sex determination disorders, 539, 541 nephrogenic rests and timing of inactivation, 688 p53 interactions with protein, 688, 689 structure, 686 transcriptional regulation and growth suppression, 686, 687 WWS, see Walker-Warburg syndrome

X X chromosome, aneuploidy, 564, 565 inactivation and dosage compensation, 561, 563

1123

INDEX

inactivation, 64 monosomy, see Turner’s syndrome, ring X syndrome, 566, 567 structure and function, 561 X-linked disorders, hearing loss, 1095 overview, 54 Xeroderma pigmentosum (XP), clinical findings, 750, 753 complementation groups, 749, 753, 754 diagnosis with unscheduled DNA synthesis assay, 752, 753, 755, 756 genetic basis of disease, 753, 756 history of study, 749, 750 incidence, 750 management/treatment, 757, 758

molecular pathophysiology of disease, 696, 697, 756, 757 transgenic mouse model, 108, 109, 757 X-linked myotubular myopathy, phenotype and gene, 884 X-linked severe combined immunodeficiency, molecular basis, 286 X-linked spastic paraplegia (XLSP), phenotype and gene, 887 XLSP, see X-linked spastic paraplegia XP, see Xeroderma pigmentosum

structure and function, 561 Yq deletions, 567, 568 YAC, see Yeast artificial chromosome Yeast artificial chromosome (YAC), DNA cloning, 13, 61 gene mapping, 46, 47 Yeast two-hybrid system, cDNA library screening, 875 protein–protein interaction analysis, 23 transcriptional coactivator identification, 30

Y

Z

Y chromosome, aneuploidy, 565, 566 azoospermia factor, 607, 608

Z structure, see Muscle ZAP-70 deficiency, molecular basis of immunodeficiency, 288