Poly(ADP-Ribosyl)ation (Molecular Biology Intelligence Unit) 0387333711, 9780387333717

Citation preview

MOLECULARBIOLOGY INTELLIGENCE UMT

Poly(ADP-Ribosyl)ation Alexander Biirkle, M.D. Molecular Toxicology Group Department of Biology University of Konstanz Konstanz, Germany

L~DESBIOSCIENCE/EUREKAH.COM GEORGETOWNT,EXAS U.S.A.

SPRINGERSCIENCE+BUSINESSMEDIA

NEWYORK,NEWYORK U.S.A.

POLY(ADP-RIBOSa'L)ATION Molecular Biology Intelligence Unit Landes Bioscience / Eurekah.com Springer Science+Business Media, Inc.

ISBN: 0-387-33371-1

Printed on acid-flee paper.

Copyright ©2006 Landes Bioscience and Springer Science+Business Media, Inc. All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher, except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in the publication of trade names, trademarks, service marks and similar terms even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. While the authors, editors and publisher believe that drug selection and dosage and the specifications and usage of equipment and devices, as set forth in this book, are in accord with current recommendations and practice at the time of publication, they make no warranty, expressed or implied, with respect to material described in this book. In view of the ongoing research, equipment development, changes in governmental regulations and the rapid accumulation of information relating to the biomedical sciences, the reader is urged to carefully review and evaluate the information provided herein. Springer Science+Business Media, Inc., 233 Spring Street, New York, New York 10013, U.S.A. http://www.springer.com Please address all inquiries to the Publishers: Landes Bioscience / Eurekah.com, 810 South Church Street, Georgetown, Texas 78626, U.S~. Phone: 512/863 7762; FAX: 512/863 0081 http://www.eurekah.com http://www.landesbioscience.com Printed in the United States of America. 987654321

Library of Congress Cataloging-in-Publication Data A C.I.P. C a t a l o g u e record for this b o o k is available from the Library o f Congress.

This book is dedicated to the memory of Professor Paul Mandel, Strasbourg, France, the founder of the field ofpoly (ADP- ribose) research.

CONTENTS Preface ............................................................................................... xvii 1. Enzymes in Poly(ADP-Ribose) Metabolism ........................................... 1

Ralph G. Meyer, Mirella L. Meyer-Ficca, Elaine L. Jacobson and Myron K Jacobson Overview of the Enzymes of Poly(ADP-Ribose) Metabolism ................. 1 The PARP Family of Proteins ............................................................... 3 Poly(ADP-Ribose) Glycohydrolase (PARG) .......................................... 9 ADP-Ribosyl Protein Lyase ................................................................. 10 2. PARP-2: Structure-Function Relationship ........................................... 13

Va//rie Schreiber, Michelle Ricoul, Jean-Christophe Am/, Fran¢oise Dantzer, V#onique Meder, Catherine Spenlehauer, Patrick Stiegler, Claude Niedergan$ Laure Sabatier, ½"ncentFavaudon, Josiane Mdnissier-de Murcia and Gilbert de Murcia PARP-2, the Closest PARP-1 Relative ................................................. PARP-2, a Novel DNA-Damage Dependent Poly(ADP-Ribose) Polymerase ........................................................ The Crystal Structure of the Mouse PARP-2 Catalytic Domain: Differences and Similarities ............................................................. PARP-2 Localises Broadly across the Centromere during the Prometaphase and Metaphase Stages .............................. Radiosensitivity of PARP-2 Deficient Mice and Cells .......................... The Physiological Role of PARP-2: A View from the Many Partners of PARP-2 .................................... PARP-2 and the Control of G2/M Transition of the Cell Cycle through Functional Kinetochore ..................................................... PAR~-2 in the Surveillance of Telomere Integrity ............................... PARP-2: Another Actor in Base Excision Repair .................................

16 17 17 19 21 21

23 23 26

3. Targeting the Poly(ADP-Ribose) Glycohydrolase (PARG) Gene in Mammals ................................................................................ 32

Ulrich CortesandZhao-Qi Wang PARG Gene Structure and Expression ................................................. 33 Biological Function of PARG .............................................................. 34

PARG Gene Disruption in ES Cells ..................................................... 34 Embryonic Development in PARG~2-3/A23Mice ................................... 36 Proliferation of PARG 110-Depleted Fibroblast Cells .......................... 36 Immortalization Potential of PARG~23/A2-3Cells ................................... 37 4.

DNA Damage Signaling through Poly(ADP-Ribose) ........................... 41

Maria Malanga and Felix R. Althaus The Different Steps of Signaling .......................................................... 41 Protein Targeting by Poly (ADP-Ribose) ............................................ 43 Polymer-Binding 'Reprograms' Domain Functions of Proteins ........... 44

5. Roles of Poly(ADP-Ribose) Metabolism in the Regulation of Centrosome Duplication and in the Maintenance of Neuronal Integrity ........................................................................... 51

Masanao Miwa, Masayuki Kanai, Masahiro Uchida, Kazuhiko Uchida and Shuji Hanai Dynamic Nature of Poly(ADP-Ribore) Metabolism Due to the Interplay between PARP and PARG ..................................... 52 Control of Cellular Function, Including Centrosome Duplication, by Poly(ADP-Ribosyl)ation ............................................................. 53 Poly(ADP-Ribose) Metabolism in Drosophila Melanogaster ................. 55 6. Functional Interactions of PARP- 1 with p53 ....................................... 61

Rafael Alvarez-Gonzalez, Hanswalter Zentgraf, Manj%d Frey and Hilda Mendoza-Alvarez Sequence Analysis and Functional-Structural Features of p53 .............. 61 Post-Translational Regulation of p53 by Enzymatic Poly(ADP-Ribosyl)ation .................................................................. 62 Primary Structure ofPoly(ADP-Ribose) Polymerase-1 (PARP-1) and Its Multiple Functional Domains ............................................. 62 Functional Interactions of PARP-1 with p53 and Genomic Integrity ... 63 Functional Interactions of PARP-1 with p53 in Apoptotic Cells ......... 63 Functional Regulation of p53 by Covalent Poly(ADP-Ribosyl)ation ... 64 7. Dynamic Interaction between PARP-1, PCNA and

p 2 1 ~'at't/cipl . . . . . . . . . . . . .

67

Ennio Prosperi and A. Ivana Scovassi PARP-1 Interacts with DNA Repair/Replication Proteins ................... 67 PCNA: A Protein with Many Partners ................................................ 68 p21 Regulates the Activity of PCNA ................................................... 69 Effect of the Interaction between PARP-1 and PCNA ......................... 69

8. PARP-1 as Novel Coactivator of NF-~B in Inflammatory Disorders ... 75 Paul 0. Hassa and Michael 0. Hot-tiger The Role of PARP- 1 in Inflammatory Disorders ................................. 75 The Switch from Homeostasis to Disorders ......................................... 76 Nuclear Transcription Factor kappa B (NF-g:B) .................................. 76 Dysregulation of NF-~cB in Disease ..................................................... 81 The Role for PARP-1 as a Novel Coactivator of NF-g:B ...................... 81 Poly(ADP-Ribosyl)ation Activity and NF-g:B-Dependent Gene Expression .............................................................................. 83 Model for the Function of the Transcriptional NF-g:B/p300/PARP-1 Complex in Inflammatory Disorders ........... 84 Perspectives ......................................................................................... 84 9. PARP and Epigenetic Regulation ......................................................... 91

Paola Caiafa Several Epigenetic Modifications Work Together in the Regulation of Gene Expression .............................................. 91

Inhibition of Poly(ADP-Ribosyl)ation Induces DNA Hypermethylation ................................................................. 94 Molecular Mechanism(s) Connecting DNA Methylation to Poly(ADP-Ribosyl)ation: A Puzzle to Solve ................................. 98 10. PARP and the Release of Apoptosis-Inducing Factor from Mitochondria ............................................................................. Suk fin Hong, Ted M. Dawson and Valina L. Dawson Introduction to Cell Death ................................................................ PARP-1 and Nuclear Death Signals ................................................... Poly(ADP-Ribose) Glycohydrolase (PARG) ...................................... Mitochondrial Participation in Cell Death Cascades ......................... Apoptosis-Inducing Factor (AIF) ....................................................... AIF-Mediated Cell Death .................................................................. Mitochondrial Release of AIF ............................................................ Poly(ADP-Ribose) (PAR) Polymer-Mediated Cellular Response ....... 11. Genome Degradation by DNAS1L3 Endonudease: A Key PARP-1-Regulated Event in Apoptosis .................................... A. Hamid Boulares, Alexander G. Yakovlev and Mark E. Smulson Introduction: DNA Fragmentation in Apoptosis ............................... Poly(ADP-Ribosyl)ation-Regulated Endonuclease: Not Such a New Observation ........................................................ Identification of DNAS1L3 as the Poly(ADP-Ribosyl)ation Regulated Ca2+-Mg2+-Dependent Endonuclease ............................. Catalytic Properties of DNAS 1L3 ..................................................... Tissue Distribution of Transcripts Encoding the Mouse Homolog of DNAS 1L3 ............................................... Poly(ADP-Ribosyl)ation and Inactivation of Recombinant DNAS 1L3 by PARP- 1 in Vitro ..................................................... DNAS 1L3 Mediates Internucleosomal DNA Fragmentation during Drug or TNF-Induced Apoptosis ....................................... PARP-1 Cleavage by Caspase Is Required for DNAS1L3-Mediated DNA Fragmentation .............................. PARP-1 Cleavage by Caspases Is Required to Avoid a Stable State of Poly(ADP-Ribosyl)ation and Subsequent Persistent Inactivation of DNAS 1L3 during Apoptosis .................................. DNAS 1L3 Is Required for Etoposide-Induced Internucleosomal DNA Fragmentation and Increases Etoposide Cytotoxicity in Transfected Osteosarcoma Cells ................................................ Induction of Internucleosomal DNA Fragmentation by Acetaminophen Treatment Is Associated with DNAS 1L3 Expression ........................................................... Caspase-3 Activation, Cleavage of DFF45, and Degradation of DNA into 50-kb Fragments Are Insufficient for Induction of Internucleosomal DNA Fragmentation in Etoposide-Treated Osteosarcoma Cells .......................................................................

103 103 104 105

106 107 108 109 112 118 118 119 120 120 121 122 122 125

125

126

127

128

12. NAD-Metabolism and Regulatory Functions ..................................... 132 Mathias Ziegler Energetic Functions of NAD(P) ........................................................ Signalling Functions of Pyridine Nucleotides .................................... NAD(P) Biosynthesis ........................................................................ Subcellular Compartmentation of NAD and Its Metabolism ............. Interplay between PARP 1 and NAD Metabolism ..............................

132 133

135 137 137

13. PARP-1 and the Shape of Cell Death ................................................. 141

Ldszl6 l~rdg The Role of PARP-1 in Cell Death Caused by Nongenotoxic Stimuli .............................................................. 142 The Role of PARP-1 in Cell Death Caused by Genotoxic Stimuli ..... 143 Nuclear-Mitochondrial Cross-Talk in DNA Damage-Induced Cell Death ......................................................... 148 14. Poly(ADP-Ribose) Polymerase (PARP) and Excitotoxicity ................. 153

Domenico E. Pellegrini-Giampietro, Alberto Chiarugi and Flavio Moroni Excitotoxicity .................................................................................... PARP-1 and Excitotoxicity: The Suicide Hypothesis ......................... PARP-1 Inhibitors and Post-Ischemic Neuronal Death ..................... Role of PARP-1 in Models of Mild and Intense NMDA Exposure in Vitro ..........................................................................

153 154 155 157

15. Poly(ADP-Ribose) Polymerase and Ischemia-Reperfusion Injury ....... 164

Prabal K Chatterjee and Christoph Thiemermann Ischemia-Reperfusion Injury ............................................................. Renal Ischemia-Reperfusion Injury .................................................... Role of Reactive Oxygen Species ....................................................... Role of Reactive Nitrogen Species ..................................................... Poly(ADP-Ribose) Polymerase and Ischemia-Reperfusion Injury ...... Poly(ADP-Ribose) Polymerase and Renal Ischemia-Reperfusion Injury ......................................................... Beneficial Effects of Inhibitors of Poly(ADP-Ribose) Polymerase Activity .......................................................................

165

167 167 168 169 170

171

16. Role of Poly(ADP-Ribose) Polymerase Activation in the Pathogenesis of Inflammation and Circulatory Shock .............. 184

Csaba Szab6 Introduction: The Poly(ADP-Ribose) Polymerase Activation Suicide Pathway ............................................................................ 184 Triggers of DNA Single Strand Breakage and PARP Activation in Inflammation and Circulatory Shock ........................................ 185 Involvement of the PARP Pathway in Various Forms of Inflammation ............................................................................ 186

Role of PARP Activation in the Pathogenesis of Systemic Inflammatory Response Syndrome and Circulatory Shock ............ 191 Conclusions: Pharmacological Inhibition of PARPm A Powerful Anti-Inflammatory and Anti-Shock Approach ............ 194 17. Role of Poly-ADP-Ribosytadon in Cancer Development ................... 203

Mitsuko Masutani, Akemi Gunji, Masahiro Tsutsumi, Kumiko Ogawa, Nobuo Kamada, Tomoyuki Shirai, Kou-ichiJishage, Hitoshi Nakagama and Takashi Sugimura Mouse Models of Carcinogenesis ....................................................... Effect of PARP Inhibitors on Carcinogenesis ..................................... Tumorigenesis and Differentiation .................................................... DNA Repair and Genomic Instability ............................................... Chromosome Instability and Cell-Cycle Checkpoints Controls ......... Epigenetic Instability and Control of Gene Expression ...................... Cancer Cell Selection through Cell Death ......................................... Role of PARP in Human Carcinogenesis ........................................... 18.

203 206 206 208 210 210 211 211

PARP Inhibitors and Cancer Therapy ................................................ 218

Nicola J. Curtin Development of Novel PARP Inhibitors ........................................... 220 Cell-Based-Studies with Novel PARP Inhibitors ................................ 223 In Vivo Studies with PARP Inhibitors ............................................... 227 19.

Poly(ADP-Ribosyl)ation and Aging ................................................... 234

Sascha Beneke and Alexander Biirkle The Cellular Poly(ADP-Ribosyl)ation System ................................... 235 Poly(ADP-Ribosyl)ation and Mammalian Life Span ......................... 236 Interaction of Poly(ADP-Ribose) Polymerases with Other Proteins Involved in Aging ............................................................ 238 Index .................................................................................................. 243

EDITOR Alexander Biirkle Molecular Toxicology Group Department of Biology University of Konstanz Konstanz, Germany Chapter 19

CONTRIBUTORS Felix tL Althaus Institute of Pharmacology and Toxicology University of Zurich-Tierspital Zurich, Switzerland Chapter 4 Rafael Alvarez-Gonzalez Department of Molecular Biology and Immunology University of North Texas Health Science Center Fort Worth, Texas, U.S.A. Chapter 6 Jean-Christophe Am~ Unit~ 9003 du CNRS Ecole Supdrieure de Biotechnologie de Strasbourg Illkirch, France Chapter 2 Sascha Beneke Molecular Toxicology Group Department of Biology University of Konstanz Konstanz, Germany Chapter 19 A. Hamid Boulares Department of Pharmacology and Experimental Therapeutics Louisiana State University Health Sciences Center New Orleans, Louisiana, U.S.A. Chapter 11

Paola Caiafa Dipartimento di Biotecnologie Cellulari ed Ematologia Sezione di Biochimica Clinica Universit~ 'La Sapienza' Rome, Italy Chapter 9 Prabal K. Chatterjee Department of Pharmacology and Therapeutics School of Pharmacy and Biomolecular Sciences University of Brighton Brighton, U.IC Chapter 15 Alberto Chiarugi Dipartimento di Farmacologia Preclinica e Clinica Universit~ degli Studi di Firenze Firenze, Italy Chapter 14 Ulrich Cortes International Agency for Research on Cancer (IARC) Lyon, France Chapter 3 Nicola J. Curtin Northern Institute for Cancer Research University of Newcasde upon Tyne Newcastle upon Tyne, U.K. Chapter 18

Fran~oise Dantzer Unit~ 9003 du CNRS Ecole Sup~rieure de Biotechnologie de Strasbourg Illkirch, France Chapter 2

Akemi Gunji Biochemistry Division National Cancer Center Research Institute Tokyo, Japan Chapter 17

Ted M. Dawson Institute for Cell Engineering and Departments of Neurology and Neuroscience Johns Hopkins University School of Medicine Baltimore, Maryland, U.Sak. Chapter 10

Shuji Hanai Department of Biochemistry and Molecular Oncology Institute of Basic Medical Sciences University of Tsukuba Tsukuba Science City, Japan Chapter 5

Valina L. Dawson Institute for Cell Engineering and Departments of Neurology, Neuroscience and Physiology Johns Hopkins University School of Medicine Baltimore, Maryland, U.Sak. Chapter I0 Gilbert de Murcia Unit~ 9003 du CNRS Ecole Sup~rieure de Biotechnologie de Strasbourg IUkirch, France Chapter 2 Vincent Favaudon Unit~ 612 INSERM G~notoxicologie, Signalisation et Radioth~rapie Exp~rimentale Institut Curie Orsay, France Chapter 2 Manfred Frey Division of Tumor Virology German Cancer Research Center Heidelberg, Germany Chapter 6

Paul O. Hassa Institute of Veterinary Biochemistry and Molecular Biology University of Zurich-Irchel Zurich, Switzerland Chapter 8 Suk Jin Hong Institute for Cell Engineering and Department of Neurology Johns Hopkins University School of Medicine Baltimore, Maryland, U.S.A. Chapter 10 Michael O. Hottiger Institute of Veterinary Biochemistry and Molecular Biology University of Zurich-Irchel Zurich, Switzerland Chapter 8 Elaine L. J acobson Department of Pharmacology and Toxicology College of Pharmacy Arizona Cancer Center University of Arizona Tucson, Arizona, U.S~. Chapter 1

Myron K. Jacobson Department of Pharmacology and Toxicology College of Pharmacy Arizona Cancer Center University of Arizona Tucson, Arizona, U.S.A. Chapter 1 Kou-ichi Jishage Chugai Research Institute for Medical Science, Inc. Shizuoka, Japan Chapter 17 Nobuo Kamada Chugai Research Institute for Medical Science, Inc. Shizuoka, Japan Chapter 17 Masayuki Kanai Department of Biochemistry and Molecular Oncology Institute of Basic Medical Sciences University of Tsukuba Tsukuba Science City, Japan Chapter 5 Maria Malanga Institute of Pharmacology and Toxicology University of Zurich-Tierspital Zurich, Switzerland Chapter 4 Mitsuko Masutani Biochemistry Division National Cancer Center Research Institute Tokyo, Japan Chapter 17

V6ronique Meder Unit6 9003 du CNRS Ecole Sup&ieure de Biotechnologie de Strasbourg Illkirch, France Chapter 2 Hilda Mendoza-Alvarez Department of Molecular Biology and Immunology University of North Texas Health Science Center Fort Worth, Texas, U.S.A. Chapter 6 Josiane Mdnissier-de Murcia Unitd 9003 du CNRS Ecole Supdrieure de Biotechnologie de Strasbourg Illkirch, France Chapter 2 Ralph G. Meyer Department of Pharmacology and Toxicology College of Pharmacy Arizona Cancer Center University of Arizona Tucson, Arizona, U.S.A. Chapter 1 Mirella L. Meyer-Ficca Department of Pharmacology and Toxicology College of Pharmacy Arizona Cancer Center University of Arizona Tucson, Arizona, U.S~. Chapter 1 Masanao Miwa Department of Biochemistry and Molecular Oncology Institute of Basic Medical Sciences University of Tsukuba Tsukuba Science City, Japan Chapter 5

Flavio Moroni Dipartimento di Farmacologia Preclinica e Clinica Universit~ degli Studi di Firenze Firenze, Italy Chapter 14

Laure Sabatier Laboratoire de Radiobiologie et Oncologie CEA Fontenay-aux-Roses, France Chapter 2

Hitoshi Nakagama Biochemistry Division National Cancer Center Research Institute Tokyo, Japan Chapter 17

Val~rie Schreiber Unit~ 9003 du CNRS Ecole Sup~rieure de Biotechnologie de Strasbourg Illkirch, France Chapter 2

Claude Niedergang Unit~ 9003 du CNRS Ecole Sup6rieure de Biotechnologie de Strasbourg Illkirch, France Chapter 2

A. Ivana Scovassi Istituto di Genetica Molecolare CNR Pavia, Italy Chapter 7

Kumiko Ogawa Department of Experimental Pathology and Tumor Biology Nagoya City University Graduate School of Medical Sciences Nagoya, Japan Chapter 17 Domenico E. Pellegrini-Giampietro Dipartimento di Farmacologia Preclinica e Clinica Universit~ degli Studi di Firenze Firenze, Italy Chapter 14 Ennio Prosperi Istituto di Genetica Molecolare CNR Pavia, Italy Chapter 7 Michelle Ricoul Laboratoire de Radiobiologie et Oncologie CEA Fontenay-aux-Roses, France Chapter 2

Tomoyuki Shirai Department of Experimental Pathology and Tumor Biology Nagoya City University Graduate School of Medical Sciences Nagoya, Japan Chapter 17 Mark E. Smulson Department of Biochemistry and Molecular Biology Georgetown University School of Medicine Washington, D.C., U.S.A. Chapter 11 Catherine Spenlehauer Unit6 9003 du CNRS Ecole Sup~rieure de Biotechnologie de Strasbourg Illkirch, France Chapter 2 Patrick Stiegler Unit~ 9003 du CNRS Ecole Sup~rieure de Biotechnologie de Strasbourg Illkirch, France Chapter 2

Takashi Sugimura Biochemistry Division National Cancer Center Research Institute Tokyo, Japan Chapter 17

L~szl6 Vir~g Department of Medical Chemistry Medical and Health Science Center University of Debrecen Debrecen, Hungary Chapter 13

Csaba Szab6 Inotek Pharmaceuticals Corporation Beverly, Massachusetts, U.S.A. Chapter 16

Zhao-Qi Wang International Agency for Research on Cancer (IARC) Lyon, France Chapter 3

Christoph Thiemermann Centre for Experimental Medicine, Nephrology and Critical Care William Harvey Research Institute St. Bartholomew's and the Royal London School of Medicine and Dentistry London, U.K. Chapter 15 Masahiro Tsutsumi Department of Oncological Pathology, Cancer Center Nara Medical University Nara, Japan Chapter 17 Kazuhiko Uchida Department of Biochemistry and Molecular Oncology Institute of Basic Medical Sciences University of Tsukuba Tsukuba Science City, Japan Chapter 5 Masahiro Uchida Department of Biochemistry and Molecular Oncology Institute of Basic Medical Sciences University of Tsukuba Tsukuba Science City, Japan Chapter 5

Alexander G. Yakovlev Department of Neuroscience Georgetown University School of Medicine Washington, D.C., U.S.A. Chapter 11 Hanswalter Zentgraf Division of Tumor Virology German Cancer Research Center Heidelberg, Germany Chapter 6 Mathias Ziegler Department of Molecular Biology University of Bergen Bergen, Norway Chapter 12

PREFACE he research field of poly(ADP-ribosyl)ation, which originated in the mid-1960s as an extremely narrow focus of interest for polynucleotide biochemists, has remained a rather stable field for a long time in terms of scientific orientation and researchers involved. Starting from the late 1980s, however, the scene has witnessed a tremendous influx of new people, fresh ideas and novel techniques, which l e d ~ especially over the last five years~to a breathtaking expansion of the field with regards to breadth and depth of scientific information available. What began as a peculiar posttranslational modification somehow associated with DNA repair has now invaded almost all sub-disciplines of biology and has also gained substantial interest by medical researchers and pharmaceutical companies, as there is a wide range of pathophysiological conditions linked with poly(ADP-ribosyl)ation. The last comprehensive monograph on this topic was edited by Gilbert de Murcia and Sidney Shall in 2000, and therefore the demand for a fresh summary of the state-of-the-art is obvious, despite the recent creation of very helpful web resources, such as the "PARP Link" (http:// parplink.u-strasbg.fr/index.html). I should like to thank Landes Bioscience for the invitation to edit a monograph on this topic and all the contributors for their enthusiasm and compliance during the writing phase. May the 19 chapters collated in this book, covering many, but by no means all, aspects ofpoly(ADP-ribosyl)ation, be an useful update for the expert and may they provide, to the non-expert, enlightenment about this complex, highly dynamic and fascinating field!

T

Alexander Bark& M.D.

CHAPTER 1

Enzymes in P01y(ADP-Ribose)Metabolism Ralph G. Meyer, Mirella L. Meyer-Ficca, Elaine L. Jacobson and Myron K. Jacobson Abstract tudies over many years have revealed the central importance of poly(ADP-ribose) metabolism in the maintenance of genomic integrity. While the involvement of poly(ADP-ribose) polymerase- 1 (PARP- 1) in this metabolism has been long known, more recent studies have demonstrated the contribution of many different genes coding for PARPs to promoting cellular recovery from genotoxic stress, eliminating badly damaged cells from the organism, and ensuring accurate transmission of genetic information during cell division. Additionally, emerging information suggests the involvement of ADP-ribose polymer metabolism in the regulation of intracellular trafficking, memory formation and other cellular functions. This chapter reviews the chemistry of ADP-ribose polymer metabolism and the enzymes that catalyze the synthesis and turnover of poly(ADP-ribose).

S

Overview of the Enzymes of Poly(ADP-Ribose) Metabolism The metabolism ofADP-ribose (ADPR) polymers represents a reversible protein modification whose basic enzymology is depicted in Figure 1. The oxidized form of nicotinamide adenine dinucleotide (NAD+) is the substrate for polymer synthesis in reactions in which the glycosylic linkage between nicotinamide and ribose is cleaved, nicotinamide and a proton are released, and the ADPR moiety is used for polymer formation. Poly(ADP-ribose) polymerases (PARPs) catalyze three reactions involved in polymer synthesis. 1 Polymer initiation in most cases involves addition of ADPR to a protein carboxylate group, usually a glutamate residue. Polymer elongation involves formation of novel ribosyl-ribosyl linkages that result in both linear and branched polymer residues. The size of polymers in vivo is known primarily for polymers synthesized in response to genotoxic stress where polymer length varies from a few residues in linear structures to more than 100 residues in multibranched polymers. 1The large variation in polymer size can be attributed to the rapid turnover of polymers and to the nature of the protein acceptors. Generally, automodified PARP-1 contains the largest polymers. 1 The functional roles of ADPR polymers can be attributed to modulation of the function of the protein to which the polymers are covalently attached and/or to noncovalent interactions with other cellular components. The ADPR polymers possess structural features found in both polynucleotides and polysaccharides. Polymers have a high density of negative charge that allows the formation of stable ionic interactions and multiple adenine rings capable of both hydrogen bonding and base stacking interactions. The turnover of ADPR polymer residues is effected by poly(ADP-ribose) glycohydrolase(PARG) which catalyses hydrolysis of both linear and branched polymer residues yielding free ADPR.1 A third enzyme, ADP-ribosyl protein lyase, catalyzes the nonhydrolytic cleavage of protein proximal ADPR residues yielding the unique nucleotide ADP-3"-deoxy-pentos-2"-ulose (ADP-DP). 2 In addition to the basic enzymology shown in Figure 1, related enzyme activities

Poly(ADP-Ribosyl)ation,edited by Alexander Bfirkle. ©2006 and Springer Science+Business Media.

Landes Bioscience

2

Poly(ADP-Ribosyl)ation

~

,~.o oi

o

....

%" o

/

r'

or

o~,

/

o iiZ

A~

"T

~o

n

\

..--~

o

!o

~

..~ ...~_

~.

u~_60% positive charges (KR) ;Green" >50% polar (STQN) • Pink" >85% cysteines ;Orange" >85% glycines ; Yellow" >85% prolines ; Cyan" >50% Aromatics (FYW). The secondary structure ofthe murine PARP-2 catalytic domain (aa 209-557) according to Oliver et al is shown. Continued on next page.

.

PARP-2: Structure-Function Relationship 4730 hPARP-I

NSiR~ii~

~r,A~.;

""i;c ~

15

440

...... i[. N;ii~ii

.!~i i;.... K

ZN

ii~i~,~R~!iii;',~',

450

460 ~iiil

480

Nii~;i~ii.:~-

~-

i~ii""i~,i - ~

~iN:S D D $

'

E

SK

....

F 490

/

I

G

500

510

-~!!~!~::: :

b

O~!i ~ i ~ ' ~ : : ' :i ~::::i,ii!:i~] ..... ~i...... ~..... ;~ "~:: "i"":'~!i~;;;~!~ ~ ~ i ~ i ! ~ ~i ~ ~ i ~

'

~

'i~iiii:~iii~ii ~ .; m

.

540

I

H

•; . . ~ : : ~ ~ ~ : ~ } D~i ! ~Q - -,Ft SE S i"~'!"K ........i i i i ~ i i ' i i ~ ~H~ i i ! ~! i ~ i :

~

5&q

I

t ....... i .................... I.................. ; ,~i~iiii::i. 'i~iiii~i~ Oii ~:;ili iiiiii~iPiPi :"~iiiPiii'.ii~~ i i : i~..... ~ i i • i: :~.... ~::ili:~l~.~i "i!Q Q:: ~ .:ii i i ~ iii::~i P ii~::iPii! i ! ~ i i" ::i::i~i!,"~i ~ ! ~ ~ l ~ ,::~ ;:

R i~i~! ~! ......~it r " ~.~ -~:. :

rj ~

520

/

a

............. hPARP-I : ~ i i # i F K Q::i~i!~ " " £~.PARP-2NAP !T~i K~ " ~K" F AA S i: ~ R ~RmPARP-2

,~70

-i~ii~i!ii~iiN N A D S ~ i i Q A ~ ~ i l h l ~ i ; ; ; L

-~.... i~!~::"-::p~

- ~::~i~i::i~. ~

~:~e~.-.:~-~-.-.~-::~.-.::.-.:.;:

~

~;i~

I

590

..................

600 [

i ..................................

"~:::i:i~E"ri i~~~i~:!::~il ~.~ i "~:V";: !. ~::~::...... ~i ~i:~ ~..i

~i!~.. ~:./~i ~i

~i~fi!i.~i.ii~i~i " s ~ ~i|i~i~i~ ~i~....iP Ss

~":~: ~..-'. ~ '~ " ~i

~-'.-'.'i

f~

c 6~0

820

J. . . . . .

NAP

: ~ E ::

630

~iC R ~ " ~9i~;; ¢ :~C

":-:'.'~" :'~f" i { ) " ~ " +"i~" -i'~* ,D. . . .

B)~R

~i!~!ils~.:..:..~:-...~.: ~i ': ;'~

$

iNK. D

.............

640

"~i ....... t ~-~i~in ~:i~i~~"~i~r~:::iQ~ii~

.~iiii;-~::iie;IN ~:K ~iE ~ ........::::::iiii:iiiQi:.

!-.::

!::

D

~:.

f

NAP

g 690 ~

r"j

M'i~i::i~iS #i::i~i"li~i|d~-ff::::::::i::N

..................

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

680 I

660

~.

.....:............:........................ ;.;

670 I

L 650

:~ii::~::Ni~ D :~: D ~i:: N ~ii D~i N :::i::::::::~!!iy ~"~D '~ ~iii:~iP :~i. . . . ::~:.~i .~

ii~iNP!E ~-"iiE

e

mPARP-2

d

/

. . . . . .

I

~ ::N T i~iA N 0 : ~ :4~:i":-~~i~C

:__."i~!~t:i.~".R ~i;K::~ii KN b~ARP-2

K

J

...............:

h

lO0

i~i

~

......... .

M 710

i

i

~i~S

in

..................................................................... .................

...........

j .......................

k •::;:::::-~

-helix

N

rn ::.

-sheet

~.

N

:ii iii.... !~ii::-::

::-

.....

n

~

3~o turn

Figure 1, continued. A color version of this figure is available online at www.Eurekah.com.

i

16

Poly(ADP-Ribosyl)ation

derived from PARP-1 deficient mice. 2'3 This chapter summarizes the data obtained from molecular and genetic approaches developed in our laboratory to better understand the structure-function relationship of the mouse PARP-2.

PARP-2, the Closest PARP-1 Relative Five years ago, we isolated a cDNA encoding a 62 kDa protein sharing considerable homology with the catalytic domain of PARP-1. ~ The assumption that we were dealing with a novel PARP enzyme was supported by the discovery in Arabidopsis thaliana of a gene coding for a PARP related polypeptide of a calculated mass of 72 kDa and by the identification of two PARP homologues: one present at telomeres (Tankyrase 1) ~ and a second present in Vault particles (VPARP). ~ Hence, the protein encoded by this new cDNA was named PARP-2. Figure 1 displays an alignment of seven PARP-2 translated cDNAs from plants and mammals. The N-terminal domain of PARP-2 (Fig. 2) comprising 65 residues in mouse supports several functions: it binds to DNA, it contains a nuclear targeting motif ~ and an interacting interface with the telomeric protein TRF2. 7 The PARP-2 DNA binding domain (DBD) displays some homology with the SAP domain 8 found in various nuclear proteins involved in chromosomal organization and in DNA repair such as AP-endonudease and Ku70. The plant PARP-2 orthologues (Arabidopsisthaliana #NP_192148 and Zea ma/s #T03656) contain two SAP domains located at the N-terminus of the protein. A caspase-3 cleavage site located at the sequence 58DNRD,tI defines the border between the DBD and domain E (homologous to the E domain of PARP-1). 9 Intriguingly, this caspase-3 cleavage site is not present in the rat nor in the bovine sequences (Fig. 1). Domain E acts both as an homodimerization interface and an automodification domain as well. 10Among the PARP family members, PARP-2 is the closest relative to PARP-1, their catalY[lic domain (the F domain, Fig. 2) displays 69% similarity (see below). A caspase-8 cleavage site 83 LQMD 186 marks the border between domains E and F; the specific inactivation of PARP-2 at this conserved Bid-like site occurs in mice during middle cerebral artery occlusion. 11 D N A binding domain

C atalytic domain

Automodification domain

Zn t'ingers

NI-m 1

A

FI

NLS

m

,

m

BRCT

c

FI~

5'1 1 ],2 [3 [ 4 i5,,16 ( 7

:"~ 372

active site988

........

N 524

.~;~'.'" 656

,N

;,.r--..'.N ~ 859 908

mPARP..1 1014

8, 19,) 10'"~11i, ,12,I 13 114[15116117118!;19 )20~'~, ,221

23

]3'

D N A binding domain

NLS

534 mPARP-~

1

65

198

402 451

559

S'll]2 131,151slTl 8 i~ ~oi~,i,u113!,,I,5 '~ ......13'

o "

+1

H1 RNA ~

q~'- , 2000

4000

6000

8000

10000

12000

14000bp

113bp

' ~ i ! ~ -" 1 ~ PARP_2exonl H1RNA,.'PARP-2 promoters

Figure 2. Structure of the mouse PARP-2 gene and comparison with the mouse PARP- 1 exon organization. The gene structure is shown with length in bp, as deduced from Southern blot analysis with PARP-2 cDNA oligonucleotide probes. PARP-2 exons (1-16) as well as RNase P RNA (H 1 RNA, GenBank accession number L08802) are represented as dosed boxes. Mouse PARP- 1 protein and its intron-exon organization has been aligned with the mPARP-2 protein.NLS, nuclear localization signal. (Taken from Am~ et al,3with permission).

PARP-2: Structure-Function Relationship

17

The PARP-2 gene located at position 14 q l 1.2 in human and 14 C1 in mouse consists of 16 exons and 15 introns spanning about 13 kilobase pairs (Fig. 2). Surprisingly, the PARP-2 gene lies head to head with the gene encoding the RNase P RNA subunit (H1 RNA). The distance between the transcription start sites of the PARP-2 and RNase P RNA genes is only 114 base pairs, suggesting that their regulation is coordinated through a bi-directional promoter. This unique promoter organization, conserved between human and mouse, allows an alternative gene expression mediated by RNA polymerase II (PARP-2) and by RNA polymerase III (HI RNA) since they share common transcriptional control elements. 12 However, the expression of both genes is clearly independently regulated since PARP-2 is expressed at higher levels in proliferating cells of lymphoid organs, germinal cells, and duodenum epithelium in contrast to RNase P RNA which, as an essential gene, is ubiquitously expressed. Interestingly, ionising radiation strongly induces the expression of PARP-1 and PARP-2 genes in plants, 13 whereas the same genes are not inducible by DNA strand-breaks in mammals.

PARP-2, a Novel DNA-Damage Dependent Poly(ADP-Ribose) Polymerase Affinity purified mouse PARP-2 (mPARP-2) overproduced in the Sf9 insect cells/baculovirus system displays DNA damage-dependent PARP activity. Using DNAse I treated DNA as a coactivator we estimated a Km of 130 l.tM which represents an affinity for NAD +2.6-fold lower than hPARP-1 (50 gM). The kcat/Km ratio (323 s-1M -1) is 18 times lower than that ofhPARP-1, a value in good agreement with the 5-10% residual activity found in PARP-1 -/- cell extracts stimulated by DNA strand breaks. 3 The DNA-binding domain of mPARP-2 (aa 1-65; Fi§. 2) was identified on the basis of its capacity to bind damaged DNA in a Southwestern assay. Using precisely defined DNA ends, we demonstrated that purified mPARP-2 binds specifically to a gap of one nucleotide (Fig. 3A) and protects about 10 nucleotides. In contrast to PARP-1, PARP-2 does not bind to a break (Fig. 3B). The affinity of PARP-2 for the "inside" of the double helix can also be visualized by electron microscopy (Fig. 3C) where the protein accumulates and seems to displace one of the two DNA strands (arrows) when a break is present. In the absence of any protein acceptor, PARP-2 can modify itself in vitro. Most of the radioactive label is found associated with domain E (aa 64-198) that is not only the PARP-2 automodification domain but also the interactive interface with partners like: PARP- 1, XRCC 1, DNA po113, and DNA Ligase III. Domain E is also involved in PARP-2 dimerization. 1° From reconstitution experiments where either purified PARP-1 or PARP-2 were mixed with purified nuclei from PARP-1 deficient cells (Fig. 3D), it can be inferred that PARP-1 preferentially poly(ADP-ribosyl)ates the linker histone H 1, whereas PARP-2 modifies preferentially the core histone H2B. Therefore, PARP-1 and PARP-2 have different targets both in DNA and in chromatin further indicating that, as chromatin modifiers, they play specific functions.

The Crystal Structure of the Mouse PARP-2 Catalytic Domain: Differences and Similarities The mouse PARP-2 catalytic domain (aa 198-559, domain F, Fig. 2) purified to homogeneity on a 3-aminobenzamide affinity column was recently crystallized (Fig. 4) at 2.8 A resolution. 1~i The PARP-2 catalytic domain consists of two main parts: an a-helical domain in N-terminal (aa 207-324) and a mixed 0t/~ C-terminal domain (aa 332-557) containing the active site. As predicted from the homology with the corresponding region of PARP-1, the overall fold of PARP-2 catalytic domain is very similar to that of PARP-1, though with some interesting differences in the vicinity of the acceptor site (Fig. 4). The local environment of the acceptor site in PARP-2 is modified compared to that of PARP-1, mainly due to a 3 aa insertion in the loop connecting the J-strands k and l (in PARP-1). Within this loop particularly intriguing is the side chain ofY528, which has no equivalent in PARP-1 and points directly into the acceptor

18

Poly(ADP-Ribosyl)ation

A

,~

B

,

p

PARP-1

PARP.2

i 38

41

353 3 :;~i:!:ii~!!!iii!i!:i~;~;i~;ii:1]il;ili~37 i~;i;i;~i:;ii 2,

33 34

2(~

ONase I

C

. 4:

purified nuclei from:

kDa MW wt

43

H1A

~° ':!i?ili',ilil i'~ i=.~I i 20 ~ :: ~ H3 :ii! iiii~iiiiiiiii B~ ~ ~~ ; H2 B 14 :!:~ ~ ~H2 A ~:!'~i~~.i!iiiii!i :~i!! ~ "~H4

complemented with: PARP-I

PARP-2

Figure 3.A) DNAse I footprinting of human PARP-1 and murine PARP-2 bound to a 66 bp DNA duplex containing a gap of one nucleotide located at position 33. The protection observed on the 5' end labelled continuous strand is indicated in brackets on the right. B) DNAse I footprinting of human PARP- 1 and murine PARP-2 bound to a 66 bp DNA duplex containing a nick at position 33. No protection by mPARP-2 can be observed is this case. C) EM visualization of DNA strand displacement by mPARP-2 at a nick present on DNA (arrow). D) Auto and heteromodification reactions catalyzed by hPARP-1 and mPARP-2 following their introduction in purified nuclei from liver of wt or PARP-1 deficient mice.

PARP-2: Structure-Function Relationship

19

B

A PARP-2

::.~.. :~: ~,~ '~ .......................... ......7::~.,~,. P . . . . ~ii!i T529

.......... E534

PARP-1 N531

/ ~i~!!:~...~:~ "~.7- Y 5 3 2 " ~

~-

. ~

....

~:

Figure 4. A) Poly(ADP-ribose) acceptor binding site. The ADP-ribose moiety (magenta) is shown bound to the donor site of PARP-1 (PDB:IEFY), along with a superimposition of the equivalent secondary structure elements in PARP-2.AII residues involved in coordinating the substrate are highly conserved between the two PARP molecules. The extended loop (Leu 523- Thr529) unique to PARP-2 is clearly visible. In particular, residue Tyr528 is presented towards the acceptor site, possibly providing additional interactions with the pyrophosphate backbone of the bound substrate. Amino acid residues shown both in parentheses and underlined are in PARP-1. B) Superimposed Ca traces of PARP-1 (yellow) and PARP-2 (grey) catalytic fragments. The extended loop region is indicated in dotted line. (Taken from Oliver et a114 with permission of Oxford University Press). A color version of this figure is available online at www.Eurekah.com. site (Fig./LA). This specific region is supposed to provide a binding site for the elongation of a terminal ADP-ribose of a growing polymer chain, or for the positioning of the acceptor glutamate in an initiation reaction. In view of the specificity of PARP-2 and PARP-1 towards different histones, it will be interesting to swap the loop between the two enzymes or to perform a mutational analysis of Y528, to test the influence of this region on the heteromodification reaction specificity. Finally, the knowledge of the crystal structure of PARP-2 will certainly help in the development of specific inhibitors that may have also clinical applications as radiosensitizers (see below).

PARP-2 Localises Broadly across the Centromere during the Prometaphase and Metaphase Stages Confinement ofbiomolecules within compartments is essential both for the formation and function of the cell. One of the major characteristics of the newly identified members of the PARP family is their various subcellular localizations. Figure 5 displays the subceUular localization of PARP-2 compared to that of PARP-1 in HeLa cells in various phases of the cell cycle. During interphase the accumulation of both PARP-1 and PARP-2 is dearly visible in the nudeoli of HeLa cells (panels 5A, 5F). PARP-1 has been shown to translocate from the nucleolus to the nudeoplasm when RNA synthesis is inhibited suggesting that its nucleolar location is dependent on the transcriptional state of the nucleolar chromatin. 15 It remains to see whether the same occurs with PARP-2. Centromeres are the site of organization of kinetochores on mitotic chromosomes where chromosomes capture the spindle microtubules to ensure faithful chromosomal segregation during mitosis.16-In mammalian cells, they span tens of megabases and are composed of large arrays of tandemly repeated sequences, the 0~-satellite in human and the minor satellite in

20

Poly(ADP-Ribosyl)ation

anti-PARP-1 a

C s*


,, ,,.,..., N

r ~,

~i~i:!!!i!i!i!ii!i{

25-

....................... f

~

iii!::siiiiii::iii::i::i::}i .......................

..,

(nX)

(n X)+ X

(n X)- X

( nX )+der X

B

......ii!i?~,~

Figure 12. (PARP-1+/-;PARP-2"/-) female embryos display X-chromosome instability. A) Distribution of X-chromosome abnormalitiesin maleand femaleembryosfromvariousgenotypes. B) Examplesofmetaphases with various numbers of X-chromosomes in (PARP-1+/;PARP-2-/-) females. A color version of this figure is available online at www.Eurekah.com. defect due to both (PARP-1 ;PARP-2) disruption. Lethality occuring shordy after gastrulation was also observed in mutant embryos lacking other BER factors such as XRCC 33 and APE. 34 PARP-1 haplo-insufflciency in a PARP-2 -/- context caused female-specific embryonic lethality (Fig. 11 D) associated with X-chromosome instability (Fig. 12) indicating that in this setting half of the normal PARP-1 dosage in the absence of PARP-2 at centromeres induces a high frequency of kinetochore defects during X-chromosome segregation, thus increasing its well known lagging character. Alternatively, one could imagine that, as a silenced region of the genome, X-chromosome may be less frequendy repaired than actively transcribed portions and therefore absolutely needs to be repaired during DNA replication to be faithfully transmitted. Depending on the extent of DNA damage, the inactivated X-chromosome might not be captured in time by the microtubules. This delay is most probably accentuated in the PARP-2 -/context, thus increasing its instability.

30

Poly(ADP-Ribosyl)ation

The presence of PARP-2 in regions of the genome containing repetitive D N A sequences like centromeres, telomeres and r D N A and in association with X-chromosomes suggest a role of PARP-2 in the maintenance of both constitutive and facultative heterochromatin integrity, which may become a target for pharmacological intervention.

Acknowledgements We thank Dr. Sugimura for the 10H anti-poly(ADP-ribose) antibody. We are grateful to Didier Hentsch (IGBMC, Illkirch) and Jean-Christophe Laval (Leica Microsystems, France) for their help with the laser microbeam. This work was supported by funds from Centre National de la Recherche Scientiflque, Association pour la Recherche Contre le Cancer, Electricit~ de France, Ligue Nationale Contre le Cancer and Commissariat ~ l'Energie Atomique.

References 1. de Murcia G, Shall S, eds. From DNA Damage and Stress Signalling to Cell Death: Poly (ADP-Ribosylation) Reactions. Oxford, New York: Oxford University Press, 2000:238. 2. Shieh WM, Ame JC, Wilson MV et al. Poly(ADP-ribose) polymerase null mouse cells synthesize ADP-ribose polymers. J Biol Chem 1998; 273(46):30069-72. 3. Ame JC, Rolli V, Schreiber V e t al. PARP-2, A novel mammalian DNA damage-dependent poly(ADP-ribose) polymerase. J Biol Chem 1999; 274(25):17860-8. 4. Lepiniec L, Babiychuk E, Kushnir S et al. Characterization of an Arabidopsis thaliana cDNA homologue to animal poly(ADP-ribose) polymerase. FEBS Lett 1995; 364(2):103-8. 5. Smith S, Giriat I, Schmitt A et al. Tankyrase, a poly(ADP-ribose) polymerase at human telomeres. Science 1998; 282(5393):1484-7. 6. Kickhoefer VA, Siva AC, Kedersha NL et al. The 193-kD vault protein, VPARP, is a novel poly(ADP-ribose) polymerase. J Cell Biol 1999; 146(5):917-28. 7. Dantzer F, Giraud-Panis MJ, Jaco Iet al. Functional Interaction between Poly(ADP-Ribose) Polymerase 2 (PARP-2) and TRF2: PARP Activity Negatively Regulates TRF2. Mol Cell Biol 2004; 24(4):1595-1607. 8. Aravind L, Koonin EV. SAP-a putative DNA-binding motif involved in chromosomal organization. Trends Biochem Sci 2000; 25(3):112-4. 9. Menissier de Murcia J, Ricoul M, Tartier L et al. Functional interaction between PARP-1 and PARP-2 in chromosome stability and embryonic development in mouse. EMBO J 2003; 22(9):2255-63. 10. Schreiber V, Ame JC, Dolle P et al. Poly(ADP-ribose) polymerase-2 (PARP-2) is required for efficient base excision DNA repair in association with PARP-1 and XRCC1. J Biol Chem 2002; 277(25):23028-36. 11. Benchoua A, Couriaud C, Guegan C et al. Active caspase-8 translocates into the nucleus of apoptotic cells to inactivate poly(ADP-ribose) polymerase-2. J Biol Chem 2002; 277(37):34217-22. 12. Ame JC, Schreiber V, Fraulob V et al. A bidirectional promoter connects the poly(ADP-ribose) polymerase 2 (PARP-2) gene to the gene for RNase P RNA. structure and expression of the mouse PARP-2 gene. J Biol Chem 2001; 276(14): 11092-9. 13. Doucet-Chabeaud G, Godon C, Brutesco C et al. Ionising radiation induces the expression of PARP-1 and PARP-2 genes in Arabidopsis. Mol Genet Genomics 2001; 265(6):954-63. 14. Oliver AW, Ame JC, Roe SM et al. Crystal structure of the catalytic fragment of murine poly(ADP-ribose) polymerase-2. Nucleic Acids Res 2004; 32(2):456-464. 15. Desnoyers S, Kaufmann SH, Poirier GG. Alteration of the nucleolar localization of poly(ADP-ribose) polymerase upon treatment with transcription inhibitors. Exp Cell Res 1996; 227(1):146-53. 16. Choo KH. Domain organization at the centromere and neocentromere. Dev Cell 2001; 1(2):165-77. 17. Adams RR, Carmena M, Earnshaw WC. Chromosomal passengers and the (aurora) ABCs of mitosis. Trends Cell Biol 2001; 11(2):49-54. 18. Earle E, Saxena A, MacDonald A et al. Poly(ADP-ribose) polymerase at active centromeres and neocentromeres at metaphase. Hum Mol Genet 2000; 9(2):187-94. 19. Saxena A, Wong LH, Kalitsis P e t al. Poly(ADP-ribose) polymerase 2 localizes to mammalian active centromeres and interacts with PARP-1, Cenpa, Cenpb and Bub3, but not Cenpc. Hum Mol Genet 2002; 11(19):2319-29. 20. Saxena A, Saffery R, Wong LH et al. Centromere proteins Cenpa, Cenpb, and Bub3 interact with poly(ADP-ribose) polymerase-1 protein and are poly(ADP-ribosyl)ated. J Biol Chem 2002; 277(30):26921-6.

PARP-2: Structure-Function Relationship

31

21. Chalmers AJ. Poly(ADP-ribose) polyrnerase-1 and ionizing radiation: sensor, signaller and therapeutic target. Clin Oncol (R Coil Radiol) 2004; 16(1):29-39. 22. Chalmers A, Johnston P, Woodcock M e t al. PARP-1, PARP-2, and the cellular response to low doses of ionizing radiation. Int J Radiat Oncol Biol Phys 2004; 58(2):410-9. 23. Augustin A, Spenlehauer C, Dumond H et al. PARP-3 localizes preferentially to the daughter centriole and interferes with the G1/S cell cycle progression. J Cell Sci 2003; l l6(Pt 8):1551-62. 24. Sbodio JI, Lodish HF, Chi NW. Tankyrase-2 oligomerizes with tankyrase-1 and binds to both TRF1 (telomere-repeat-binding factor 1) and IRAP (insulin-responsive aminopeptidase). Biochem J 2002; 361(Pt 3):451-9. 25. de Lange T. Protection of mammalian telomeres. Oncogene 2002; 21(4):532-40. 26. Ferreira MG, Miller KM, Cooper JP. Indecent exposure: when telomeres become uncapped. Mol Cell 2004; 13(1):7-18. 27. Kaminker PG, Kim SH, Taylor RD et al. TANK2, a new TRFl-associated poly(ADP-ribose) polymerase, causes rapid induction of cell death upon overexpression. J Biol Chem 2001; 276(38):35891-9. 28. Griffith JD, Comeau L, Rosenfield S et al. Mammalian telomeres end in a large duplex loop. Cell 1999; 97(4):503-14. 29. van Steensel B, Smogorzewska A, de Lange T. TRF2 protects human telon-eres from end-to-end fusions. Cell 1998; 92(3):401-13. 30. Masson M, Niedergang C, Schreiber Vet al. XRCC1 is specifically associated with poly(ADP-ribose) polymerase and negatively regulates its activity following DNA damage. Mol Cell Biol 1998; 18(6):3563-71. 31. Okano S, Lan L, Caldecott KW et al. Spatial and temporal cellular responses to single-strand breaks in human cells. Mol Cell Biol 2003; 23(11):3974-81. 32. Rogakou EP, Boon C, Redon C et al. Megabase chromatin domains involved in DNA double-strand breaks in vivo. J Cell Biol 1999; 146(5):905-16. 33. Tebbs RS, Flannery ML, Meneses JJ et al. Requirement for the Xrccl DNA base excision repair gene during early mouse development. Dev Biol 1999; 208(2):513-29. 34. Ludwig DL, Maclnnes MA, Takiguchi Y et al. A murine AP-endonuclease gene-targeted deficiency with post- implantation embryonic progression and ionizing radiation sensitivity. Mutat Res 1998; 409(1):17-29.

CHAPTER3

Targetingthe Poly(ADP-Ribose) Glycohydrolase(PARG)Gene in Mammals Ulrich Cortes and Zhao-Qi Wang Abstract p

oly(ADP-ribosyl)ation is a post-translational modification of proteins. Upon DNA damage, poly(ADP-ribose) polymerase-1 catalyzes the transfer of ADP-ribose moieties from NAD + onto acceptor proteins to form long and branched polymers. Poly(ADP-ribosyl)ation is an extensive but transient modification as polymer chains can reach more than 200 units on protein acceptors and be degraded within a few minutes by poly(ADP-ribose) glycohydrolase. Homeostasis of poly(ADP-ribose) is thought to play an important function in cellular processes. The importance of pADPR synthesis has been established in vitro and in vivo by using chemical inhibitors and genetically engineered mutant mice devoid of the main pADPR synthesizing enzyme, PARP-1. However, the function of PARG in vivo remains elusive. This chapter describes the generation and characterization of PARG knockout mice.

Introduction Poly(ADP-ribosyl)ation is an immediate cellular response to DNA damage generated either exogenously or endogenously. This post-translational modification is mainly catalyzed by poly(ADP-ribose) polymerase (PARP-1, EC 2.4.2.30), using NAD ÷ as a substrate. 1'2 When PARP-1 binds to DNA strand breaks, poly(ADP-ribose) (pADPR) is synthesized and attached mainly to PARP-1 itself, but also to other target proteins, including many DNA-metabolizing/ binding molecules. 3 Several experimental approaches have been used to study the biological significance of poly(ADP-ribosyl)ation, mainly by inhibiting PARP-1 activity through chemical inhibitors, dominant negative mutants, and gene targeting.4--6These studies indicate that PARP-1 and poly(ADP-ribosyl)ation play a multifunctional role in a wide range of cellular processes such as DNA repair, replication, chromosomal stability, cell proliferation and death, transcription, as well as inflammation. In addition to PARP-1, other PARP-related enzymes have also been identified in mammalian cells (see The PARP Link Homepage; http://parplink.u-strasbg.fr). Analysis of PARP-l-deficient mice revealed that other PARPs contribute only slightly to the overall poly(ADP-ribosyl)ation in mammalian cells.7 Despite much effort, limited information on pADPR metabolism is available and results are often contradictory in different experimental systems. This may be partly due to the paucity of studies on pADPR catabolism and due to the lack of knowledge concerning pADPR degrading enzymes. To date, two distinct enzymes involved in pADPR catabolism have been identified, i.e., pADPR glycohydrolase (PARG) and ADP-ribosyl protein lyase. While the latter possesses an exoglycosidic activity and removes only the proximal ADP-ribose unit from acceptor proteins, PARG possesses both exoglycosidic and endoglycosidic activities and can remove larger oligo (ADP-ribose) fragments via endoglycosidic cleavage. Hence, it is of

Poly(ADP-Ribosyl)ation,edited by Alexander Biirkle. ©2006 and Springer Science+ Business Media.

Landes Bioscience

Targeting the Poly(ADP-Ribose) Glycohydrolase(PARG) Gene in Mammals

Human

Tim23 ~ [ ~ PARG !1 Ill I II

I

l

Ill

234

56

! |

I' ' I

7

8

9

122.3

Mouse

33

Itlt

II II

[

II

10 II 12

13

1415

16

17t8

Kb

Tim23 ~[11~ PARG II Ill el

........1 I 11

I l i I

I I i I

11

I

ill

1

7

9

13 14

15

1617 18

u lm n l 234

5 6

8

.........................................

I0 I I

94.9 Kb

II

I I

12

I

m

.........................................

Figure 1. Schematic presentation of the PARG locus and its association with Em23 in the human 15and in the mouse genome. Exons are indicated by black boxes. Open reading frames are indicated by arrows. great interest to determine the biological functions of PARG in poly(ADP-ribosyl)ation and in cellular processes. In this chapter, we will focus on the study of elucidating the biological function of PARG using gene targeting technology.

PARG G e n e

Structure

and

Expression

Poly(ADP-ribosyl)ation is a transient and dynamic process since pADPR polymers synthesized in response to DNA damage are rapidly degraded in vivo. The presence of PARG activity, degrading pADPR, was first reported in 1971. 8 Phylogenic distribution of PARG reveals its presence in mammals 9-11 and the existence of orthologs in other organisms such as Drosophila melanogaster, Caenorhabditis elegansandArabidopsis. 12The gene encoding PARG has been mapped • 13 Rat and bov" I n e PARG to chromosome 10ql 1.23 in humans and to chromosome 14B in mice. cDNAs have been doned by Jacobson's and Sugimura's groups, respectively.9-1°The PARG gene contains 18 exons in humans and mice (Fig. 1). In several species, including human, mouse, and rat, the PARG gene is head-to-head with 7~rn23, a gene encoding for a mitochondrial inner membrane translocase involved in preprotein translocation. 14It has been shown recently that the human PARG gene shares a common bidirectional promoter region with ~rn23.15 However, the biological relevance of this shared promoter region remains to be determined. To date, two forms of PARG enzymes have been reported: 110 kDa PARG (PARG 110) is in the nucleus and the cytoplasm, and another isoform at about 60 kDa (PARG60) is in the cytosolic fraction. 16However, the origin of these two isoforms has not been defined. Structural analysis of mammalian PARG 110 reveals the presence of a putative regulatory domain in the N-terminal region and a catalytic domain in the C-terminal part (Fig. 2). In contrast, Drosophila PARG lacks the putative regulatory domain and contains only the catalytic domain. In mammals, the PARG 110 isoform contains in the N-terminal region, a strong nuclear localization signal (NLS) targeting this molecule to the nucleus (Meyer-Ficca et al unpublished data, see http://www.niadyne.com/documents.html), a putative bipartite NLS, as well as a nuclear export signal (NES). 1° Ohashi et a117have reported that PARG110 shuttles between the nudeus and the cytoplasm during the cell cycle in human and mouse cells, with a cytoplasmic localization during mitosis. Interestingly, Affar et all8 have shown that PARG 110 was cleaved by caspase-3 during apoptosis in human cells, concomitant with PARP-1 cleavage. This deava§e generates two catalytically active fragments (85 and 74 kDa) relocated to the cytoplasm, 9 indicating that PARG110 activity must be precisely regulated during apoptosis. Therefore, PARG110 appears to be the most critical enzyme for the regulation of nuclear poly(ADP-ribosyl)ation, through a nucleo-cytoplasmic shuttling, and could be involved in several important cellular processes, such as cell proliferation and cell death. 2°

34

Poly(ADP-Ribosyl)ation Putative regulatory domain

Catalytic domain

pNLS Human

pNLS Mouse

pNLS

Drosophila

Figure 2. Structure of the PARG protein in man, mouse and Drosophila.Functional domains are indicated. A nuclear localizationsignal (NLS) is located in the N-terminus encoded by exon 1 in humans (Meyer-Ficca et al see http:l/www.niadyne.com/documents.html) and mice, and the putative bipartite nuclear localization signal (pNLS) is encoded by exons 3-4. It is noted that DrosophilaPARG lacks the N-terminal putative regulatory domain.

Biological Function of PARG Biochemical studies have shown that PARG possesses exoglycosidase and endoglycosidase activities21'22and catalyzes the hydrolysis of glycosidic bonds between ADPR units of polymers. In response to genotoxic stress, pADPR polymers are rapidly degraded by PARG activity in vivo (tl/2 = I min.). 3 However, in the absence of DNA damage constitutive pADPR polymers have a long half-life (several hours), 23 suggesting a selective affinity of PARG for pADPR chains induced by DNA breaks. Hence, pADPR synthesis and PARG activity are closely and precisely coordinated in response to DNA damage. However, only fewer studies have been performed to determine the biological importance of PARG activity. To gain insight into the biological function of PARG, tannin derivatives have been used to inhibit PARG activity in vitro and in vivo. Recently, it has been shown that gallotannin protects astrocytes from oxidative cell death. 24 In addition, PARG inhibition was reported to protect rats from cerebral ischemia. 25 On the other hand, PARG inactivation in Drosophila results in a lethal phenotype in flies and a severe accumulation of pADPR in the central nervous system associated with neurodegeneration, demonstrating that PARG activity is necessary for the development and survival of neuronal cells in Drosophila.26 Taken together, these studies suggest that PARG plays important roles in various cellular processes most likely through its regulatory role in homeostasis of poly(ADP-ribosyl)ation.

PARG Gene Disruption in ES Ceils To gain insight into the biological functions of PARG activity in mammals, we, in collaboration with Dr Jacobson's laboratory in Tucson, Arizona, USA, disrupted the PARG gene in embryonic stem (ES) cells and in mice using the Cre-lox gene targeting technology (Fig. 3). A mouse genomic fragment containing exons 2, 3 and 4 of the PARG gene was used to construct the targeting vector. A neo/tk cassette containing a neomycin resistance gene (neo) and a thymidine kinase gene (tk) flanked by two lox-P sites, was inserted at the end of intron I into the XbaI site. Another lox-P sequence was introduced into intron 3 in the NsiI

Targeting the Poly(ADP-Ribose) Glycohydrolase(PARG) Gene in Mammals

35

A. Construction of the gene targeting vector 1

2

.. .: .:.

3

4

Wild-type allele (WT)

./::

.:

....': .. ./

.. :.

I

..:

.: ..

........

2

3

4

~~/~t kt ki ~} ~ -~--/-B- ~ - -~~ Lox-P

Taraetinq vector Targeting

Lox-P

Lox-P

B. Gene targeting in ES cells 2

1

3

4

Targeted allele (T)

r----1 probe

"..,.... .......... ........... Cre ........ Recombinase

"""'..... ,,...... '"'.....,.. |

"""-....

4

"Deleted" allele (D 2-3) Lox-P

Figure 3. Structure ofthe targeting vector and partial restriction map ofthe PARGlocus before (A) and after (B) homologous recombination and Cre-mediated recombination. Lox-P sites are represented by empty triangles. Exons are indicated by black boxes. site. The 5'-arm of the targeting vector comprised a 2.9 kb fragment of intron 1 and the 3'-arm a 4.1 kb fragment containing exons 2, 3 and 4 with a part of intron 4. Location and orientation of neo/tk cassette and lox-P sites in the targeting vector were verified by restriction enzyme digestion and direct sequencing. The linearized targeting vector was electroporated into ES cells. After G418 selection, the targeted allele was detected in isolated ES colonies by Southern-blot analysis using a 1.2 kb EcoRI-Apalprobe located outside of the targeting vector (Fig. 3). A targeted allele was observed in 18 out of 400 isolated ES colonies, and 3 out of the 18 clones were confirmed to contain the lox-P site in intron 3. After electroporation with a Cre-expressing vector and selection with ganciclovir, ES clones with deletion ofexons 2 and 3 and devoid of the neo/tk cassette (+/A2-3) were identified by Southern-blot analysis. This approach avoided any possible influence of the neo/tk cassette in intron 1 on the neighboring Tim23. Three PARG+/~-3ES clones were injected into blastocysts to generate chimeric mice. After the mating of chimeric mice with wild-type mice in a 129/Sv background, germline transmission was identified by Southern-blot and PCR analysis using tail D NA.

36

Poly(ADP-Ribosyl)ation

Table 1. Genotype distribution of offspring from inter-crosses between PARG+/Az'3mice Genotype PARG +/+ PARG"*/a2-3 PARGae-3/a2-3

N ° of mice analyzed

Embryonic Day 3.5

EmbryonicDay 13.5

Newborns

7 (28%) 13 (52%) 5 (20%)

26 (24.7%) 55 (52.3%) 24 (24%)

251 (28.7%) 414 (47.5%) 207 (23.8%)

25

105

872

Embryonic Development in P A R G z~'3/z~'3 Mice PARG +/ae-3 mice were intercrossed to generate offspring carrying both disrupted alleles (PARG~-3/~-3). Genotyping mice at various stages of development revealed that homozygous

mutant mice were obtained at a mendelian ratio (Table 1), indicating that disruption of exons 2 and 3 does not affect embryonic development. These mice did not show any obvious phenotype within a 2-year observation period. Western-blot analysis showed that 110 kDa isoform of PARG (PARG 110) was absent in PARGz~-3/ae-3 cells and mice, whereas PARG60 expression was apparendy not affected in these mice and cells (data not shown). These results suggest that PARG 110 and PARG60 molecules are expressed independently. In addition, a significant reduction of PARG activity was found in various organs of mutant mice, suggesting that PARG 110 is the major pADPR catabolizing enzyme and PARG60 is most likely responsible for the residual level of PARG activity, consistent with previous findings showing that PARG60 harbors the pADPR degrading activityY Hence, pADPR glycohydrolase activity is shared by at least two independent molecules in murine cells. The viability of PARG 110-deficient mice suggests that PARG 110 is dispensable during development or that the major function of PARG 110 is compensated for by the presence of backup pathways in the mutant cells, such as PARG 60. Thus, it is still possible that some level of PARG activity is essential for cell life and development. In this regard, it is interesting to note that inactivation of PARG in Drosophila results in lethality under normal physiological conditions. 26 However, the discrepancy between mouse and fly may be explained by the fact that the latter possesses only one form of PARG, whereas mammals evolved two isoforms, one of which contains a regulatory domain in addition to its catalytic domain.

Proliferation of PARG110-Depleted Fibroblast Ceils Since PARG ae-3/ae-3 cells contain a greatly reduced pADPR activity, we speculated that impaired homeostasis of poly(ADP-ribosyl)ation in PARG 110-deficient cells would affect cellular functions. Therefore, we studied the proliferation capacity of PARG 110-deficient embryonic fibroblast cells (MEFs). Cells at early passages (< passage 2) were plated at I x 105/well and the cell number was counted every two days for a total of 10 days. We found that 3 out of 5 PARG ae-3/ae-3cell lines exhibited a higher proliferation rate compared to wild-type counterparts derived from littermates (Fig. 4A). In order to test whether PARG plays a role in cell cycle progression control, we treated PARG 110-deficient fibroblasts with 10 Gy y-radiation and found that mutant cells exhibited normal cell death profiles compared to wild-type cells (Fig. 4B). In addition, there is no apparent difference between wild-type and PARG110-depleted cells in cell cycle progression with or without ionizing radiation (Fig. 4B). Taken together, these results indicate that PARG 110 activity is not essential for the repair of ionizing radiation-induced DNA breaks and that the proliferation abnormalities ofPARG 110--deficient fibroblasts are not due to altered DNA damage response in cell cycle regulation. The mechanism of elevated proliferation in PARG~-3/~-3 cells is not dear. One explanation is that the lack of PARG 110 may affect DNA metabolic enzymes via altered homeostasis of poly(ADP-ribosyl)ation. It is possible that the altered proliferation of mutant cells might be caused by chromosomal aberrations present in these cells.

Targeting the Poly(ADP-Ribose) Glycohydrolase (PARG) Gene in Mammals

37

1 500000 ~

A t,.

i

O

.(3 1oooooo

E ~3

•

WT

500OOO

O

PARG WT

0

2

4

6

8

az-.~a~_,

10

Days

B

G1 =..,.

!

!

Wild-type

iI

Sub G1 . . . . . .

4

2/M i

......

"

12.6 % . . . . . . . . . . . . .

J

4

I

1

J

i PARG~a-3/~w--3 A

t

I

4.7%

9.4 % zoo

z~o

Q

SO

Untreated

~00Chamnets ) ~o

~/-10 Gy (48h)

Figure 4. A) Growth curves ofPARG~-3/ae-3 MEFs. Cells from littermates were seeded at a density of 105 cells/well in 6-well plates and counted every 2 days. B) Analysis of cell death and cell cycle progression after T-irradiations. Cells were irradiated at 10 Gy and the sub-G 1 population was analyzed by flow cytometry (FL2H channel).

Immortalization Potential of PARG ~-3/~-3 Cells To examine the immortalization potential, we used early passage MEFs to follow a 3T3 protocol of immortalization. Cells were serially passaged every three days and were counted before each passage. As shown in Figure 5, P A R G ,A2-3/A2.-3 cells entered senescence at passage 12 in contrast to wild-type cells at passage 6. However, after 20 passages mutant cells started to be

38

Poly(ADP-Ribosyl)ation

~

5

Wild-type 3 ,!;i ;i

A

5

10

15

20

25

30

35

Passages t13

PA R G

o e - 3/oe- s

ID

~!I il ~ i!

/iilti ,~ ,

i, ! ! , , , / /!

'i~ f '~li /

~ ,, 'i / !i/i! /l~!/ill

~iit!i',.i!i/:.l f,

/ / .li; /i".,',/ /I/!'~ili

~ I i' ,' 1/i//.-....-\

/_.,'/l

v q ~ ~' ~ 1' i' i ~

L) !

.,,

.:

5

10

I

15

~

20

:

25

|

30

!

35

Passages 5

PA RP- 1 -/-

=

2 ! ,

/ /i i,~ I

.dYtYvtl'It'l'tl, 5

10

15

20

25

Passages Figure 5. Potential o f P A R G '~-3/a2-3 MEFs for immortalization. Cells with indicated genotypes were seeded at 105cells/well and immortalized following a 3T3 protocol. Every 3 days cells were trypsinized, counted and replated. Data represent the mean of two independent counts. These data represent one of three independent experiments. immortalized, whereas wild-type cells were still under crisis at the end of the study. It was noted that the immortalization of P A R G ae-3/ae-3 cells was slower than P A R P - 1 - / - c e l l s , which entered • • • 28 • senescence at pass;agees2-3,/aeconsistent Wldl previous fidings. This profile of the immorraliTation processes ofPARG~ - 3 -3 cells was consistendy observed in more than 3 cell populations tested for each genotype. Hence, the inactivation of the PARG 110 protein confers an advantage for

Targeting the Poly(ADP-Ribose) Glycohydrolase (PARG) Gene in Mammals

39

immortalization, reminiscent of results obtained with PARP1 null MEFs 28 (Fig. 5). The senescence period ofPARG~-3/~-3 cells was shorter compared with that of wild-type counterparts. The accelerated immortalization in PARG ae-3/~-3 cells is probably caused by chromosome instability that facilitates the accumulation of genetic mutations in these cells, which is necessary to overcome the proliferation block and to be immortalized. It is known that the biological clock of a cell is controlled by the function of telomeres, as it was observed that telomere shortening is associated with cell ageing. In this regard, it is interesting to note that PARP-1 null mutation results in the shortening of telomeres and severe chromosome aberrations in cells. 29'3° In addition, it has been proposed that PARP-1 activity may be an important factor in protecting against or slowing ageing. 31-33Therefore, it is possible that the elevated immortalization potential could be due to PARG's function in regulating poly(ADP-ribosyl)ation which affects many nuclear and DNA metabolism enzymes, induding PARP-1.

Conclusions and Perspectives For more than 30 years PARG has been believed to be the major enzyme implicated in the catabolism of pADPIL Together with previous studies, our data show that although PARG 110 is a major isoform in pADPR catabolism, PARG 110 activity is dispensable for mouse embryonic development. However, the viability of PARG 110-deficient mice is likely due to the presence of PARG60. In this regard, future studies are required to determine how PARG 110 and PARGt0 are regulated and localized in mammalian cells. In order to dissect the function of these two isoforms of PARG, it is necessary to generate a new model carrying a specific knockout of PARG60 without disturbing PARG 110 in vivo. In addition, it is of great interest to determine whether complete abrogation of expression of any PARG molecules in mice and cells is compatible with cell viability in mammals. However, because PARG and Tim23 are localized head-to-head closely in the same chromosome and share a common promoter region (see Fig. 1), disruption of both PARG60 and PARG 110, without disturbing Tim23, may represent a challenge in future studies. PARG has been proposed as a potential target for therapy. Generation of mice and cells lacking specific PARG isoforms would provide useful tools for evaluating pharmaceutical strategies in the treatment of inflammation response and in the development of anti-cancer drugs.

Acknowledgements We thank D. Galendo and M.-P. Cros for their excellent technical assistance. We also thank Dr. Wei-Min Tong and Ms. Virginie P&rilli for their reading and discussion of the manuscript. This study was supported by the Association for International Cancer Research, UK, a grant from Region Rh6ne-Alpes, France and a grant from the Association pour la Recherche sur le Cancer (ARC), France. U.C. was supported by a grant from Ligue Nationale contre le Cancer.

References 1. Chambon P, Weil JD, Doly J et al. On the formation of a novel adenylic compound by enzymatic extracts of liver nuclei. Biochem Biophys Res Commun 1966; 25:638-643. 2. Juarez-Salinas H, Sims JL, Jacobson MK. Poly(ADP-ribose) levels in carcinogen-treated cells. Nature 1979; 282:740-1. 3. D'Amours D, Desnoyers S, D'Silva I et al. Poly(ADP-ribosyl)ation reactions in the regulation of nuclear functions. Biochem J 1999; 342:249-68 Review. 4. Wang ZQ, Auer B, Stingl Let al. Mice lacking ADPRT and poly(ADP-ribosyl)ation develop norreally but are susceptible to skin disease. Genes Dew 1995; 9:509-20. 5. de Murcia JM, Niedergang C, Trucco C et al. Requirement of poly(ADP-ribose) polymerase in recovery from DNA damage in mice and in cells. Proc Natl Acad Sci USA 1997; 94:7303-7. 6. Masutani M, Suzuki H, Kamada N e t al. Poly(ADP-ribose) polyTnerase gene disruption conferred mice resistant to streptozotocin-induced diabetes. Proc Natl Acad Sci USA 1999; 96:2301-4. 7. Shieh WM, Ame JC, Wilson MV et al. Poly(ADP-ribose) polymerase null mouse cells synthesize ADP-ribose polymers. J Biol Chem 1998; 273:30069-72.

40

Poly(ADP-Ribosyl)ation

8. Miwa M, Sugimura T. Splitting of the ribose-ribose linkage of poly(adenosine diphosphate-robose) by a calf thymus extract. J Biol Chem 1971; 246:6362-4. 9. Lin W, Ame JC, Aboul-Ela N e t al. Isolation and characterization of the cDNA encoding bovine poly(ADP-ribose) glycohydrolase. J Biol Chem 1997; 272:11895-901. 10. Shimokawa T, Masutani M, Nagasawa S et al. Isolation and cloning of rat poly(ADP-ribose) glycohydrolase: Presence of a potential nuclear export signal conserved in mammalian orthologs. J Biochem (Tokyo) 1999; 126:748-755. 11. Nagasawa S, Shimokawa T, Masutani M et al. Phylogenic distribution of poly(ADP-ribose) glycohydrolase and poly(ADP-ribose)-digesting phosphodiesterase. Proc Japan Acad 2000; 76:41-44. 12. Panda S, Poirier GG, Kay SA. tej defines a role for poly(ADP-ribosyl)ation in establishing period length of the arabidopsis circadian oscillator. Dev Cell 2002; 3:51-61. 13. Ame JC, Apiou F, Jacobson EL et al. Assignment of the poly(ADP-ribose) glycohydrolase gene (PARG) to human chromosome 10q11.23 and mouse chromosome 14B by in situ hybridization. Cytogenet Cell Genet 1999; 85:269-70. 14. Donzeau M, Kaldi K, Adam A e t al. Tim23 links the inner and outer mitochondrial membranes. Cell 2000; 101:401-12. 15. Meyer RG, Meyer-Ficca ML, Jacobson EL et al. Human poly(ADP-ribose) glycohydrolase (PARG) gene and the common promoter sequence it shares with inner mitochondrial membrane translocase 23 (TIM23). Gene 2003; 314:181-90. 16. Di Meglio S, Denegri M, Vallefuoco S et al. Poly(ADPR) polymerase-1 and poly(ADPR) glycohydrolase level and distribution in differentiating rat germinal cells Mol Cell Biochem 2003; 248:85-91. 17. Ohashi S, Kanai M, Hanai S et al. Subcellular localization of poly(ADP-ribose) glycohydrolase in mammalian cells. Biochem Biophys Res Commun 2003; 307:915-21. 18. Affar EB, Germain M, Winstall E et al. Caspase-3-mediated processing of poly(ADP-ribose) glycohydrolase during apoptosis. J Biol Chem 2001; 276:2935-42. 19. Bonicalzi ME, Vodenicharov M, Coulombe M et al. Alteration of poly(ADP-ribose) glycohydrolase nucleocytoplasmic shuttling characteristics upon cleavage by apoptotic proteases. Biol Cell 2003; 95:635-44. 20. Davidovic L, Vodenicharov M, Affar EB et al. Importance of poly(ADP-ribose) glycohydrolase in the control of poly(ADP-ribose) metabolism. Exp Cell Res 2001; 268:7-13 Review. 21. Miwa M, Tanaka M, Matsushima T et al. Purification and properties of glycohydrolase from calf thymus splitting ribose-ribose linkages of poly(adenosine diphosphate ribose). J Biol Chem 1974; 249:3475-82. 22. Ikejima M, Gill DM. Poly(ADP-ribose) degradation by glycohydrolase starts with an endonucleolytic incision. J Biol Chem 1988; 263:11037-40. 23. Alvarez-Gonzalez R, Althaus FR. Poly(ADP-ribose) catabolism in mammalian cells exposed to DNA-damaging agents. Murat Res 1989; 218:67-74. 24. Ying W, Swanson RA. The poly(ADP-ribose) glycohydrolase inhibitor gallotannin blocks oxidative astrocyte death. Neuroreport 2000; 11:1385-8. 25. Lu XC, Massuda E, Lin Q et al. Post-treatment with a novel PARG inhibitor reduces infarct in cerebral ischemia in the rat. Brain Res 2003; 978:99-103. 26. Hanai S, Kanai M, Ohashi S e t al. Loss of poly(ADP-ribose) glycohydrolase causes progressive neurodegeneration in Drosophila melanogaster. Proc Natl Acad Sci USA 2004; 101:82-6. 27. Brochu G, Shah GM, Poirier GG. Purification of poly(ADP-ribose) glycohydrolase and detection of its isoforms by a zymogram following one- or two-dimensional electrophoresis. Anal Biochem1994; 218:265-72. 28. Tong WM, Galendo D, Wang ZQ. Role of DNA break-sensing molecule poly(ADP-ribose) polymerase (PARP) in cellular function and radiation toxicity. Cold Spring Harb Symp Quant Biol 2000; 65:583-91. 29. d'Adda di Fagagna F, Hande MP, Tong WM et al. Functions of poly(ADP-ribose) polymerase in controlling telomere length and chromosomal stability. Nat Genet 1999; 23:76-80. 30. Tong WM, Hande MP, Lansdorp PM et al. DNA strand break-sensing molecule poly(ADP-Ribose) polymerase cooperates with p53 in telomere function, chromosome stability, and tumor suppression. Mol Cell Biol 2001; 21:4046-54. 31. Grube K, Burkle A. Poly(ADP-ribose) polymerase activity in mononuclear leukocytes of 13 mammalian species correlates with species-specific life span. Proc Natl Acad Sci USA 1992; 89:11759-63. 32. Muiras ML, Muller M, Schachter F et al. Increased poly(ADP-ribose) polymerase activity in lymphoblastoid cell lines from centenarians. J Mol Med 1998; 76:346-54. 33. Ly DH, Lockhart DJ, Lerner RA et al. Mitotic misregulation and human aging. Science 2000; 287:2486-92.

CHAPTER4

DNA DamageSignaling through Poly(ADP-Ribose) Maria Malanga and Felix R. Althaus Abstract everal lines of evidence reveal that poly(ADP-ribose)polymerase-1 (PARP-1) operates in a DNA damage signaling network. Poly(ADP-ribose) metabolism induced by DNA damage participates in DNA repair and contributes to downstream mechanisms leading to cell cycle arrest, cell survival, cell death, or cell transformation. An important element of these multiple actions is the recruitment of DNA damage checkpoint proteins coordinating DNA repair with downstream events. The focus of this overview is the mechanism by which poly(ADP-ribose)--attached to the automodified PARP-1--interacts with DNA damage checkpoint proteins and how it may reprogram the functions of specific protein domains. Several proteins of the genome surveillance system, e.g., p53, p21, DNA-PK, NF-g:B, XRCC1, and XPA are targets of such regulation. In all cases studied, a specific 'polymer-binding' sequence motif of 20 to 26 amino acids is targeted by poly(ADP-ribose) and this motif overlaps with important functional domains responsible for protein-protein or protein-DNA interactions, nuclear import or export, enzymatic catalysis, or protein degradation.

S

Introduction In eukaryotic cells, DNA damage may induce a several thousand fold stimulation of poly(ADP-ribose) metabolism. A few restrictions and rules apply: yeast does not express such a response, and in all other eukaryotes tested so far, the most effective types of DNA damages are those that are substrates for the DNA base excision repair pathway. By far the largest amount of ADP-ribose is processed through the catalytic domains of two nuclear enzymes: poly(ADP-ribose)polymerase-1, (PARP-1), and its catabolic counterpart, poly(ADP-ribose)glycohydrolase (PARG). Other members of the growing PARP family may contribute to this metabolism, albeit to a much lesser extent and with mechanisms that await further elucidation (for reviews see refs. 1,2). The poly(ADP-ribose) metabolism arising from the cooperation of PARP-1 and PARG is involved in DNA base excision repair and in DNA damage signaling to cell survival/cell death pathways. 1'2 The present review focuses on a particular aspect of poly(ADP-ribose) signaling: when PARP-1 is activated, it catalyzes poly(ADP-ribose) synthesis on itself ('automodification'). The polymers on PARP-1 can then recruit other proteins into multiprotein complexes and reprogram their domain functions.

The Different Steps of Signaling A reasonable understanding on how poly(ADP-ribose) may exert its signal functions has been achieved. A growing number of DNA damage checkpoint proteins containing a poly(ADP-ribose)-binding sequence motif of 20 to 26 amino acids have been identified. 3'4

Poly(ADP-Ribosyl)ation,edited by Alexander Btirkle. ©2006 Landes Bioscience and Springer Science+Business Media.

42

Poly(ADP-Ribosyl)ation

Sensor(s) PARP-I/PARP-2 Termination Poly(ADP-ribose)~ degradation

/••ff• .,~~9~~

Amplification _Poly(ADP-ribose) ~~.~ synthesis

poly(ADP-ribose)-bindingproteins

DNA repair Cell cycle Apoptosis Figure 1. The major steps of the DNA strand break signaling model. See text for details. PARP-bound polymers bind strongly, but noncovalently to these sequences and reprogram the resident domain functions. How do the polymers get in touch with these proteins ? The PARP-1/ PARG system has all the hallmarks of a DNA strand break based signal transduction mechanism, with poly(ADP-ribose) playing the effector role. 4 The model features the following steps (Fig. 1): PARP-1 has a high binding affinity for DNA ends and acts as a DNA damage sensor. 5 Binding leads to a more than 500-fold activation of PARP-l's catalytic activity6 and the original signal (i.e., the lesion on DNA) is transduced into protein-bound ADP-ribose polymers of various size (ranging from few to over 200 units) and structural complexity (linear or branched), clustered at the site of damage. PARP-1 acts as a homodimer 7'8 and in DNA damaged cells it serves as a major acceptor of poly(ADP-ribose), i.e., it catalyzes its automodification at multiple sites simultaneously, Vthus leading to signal amplification. The automodified PARP-1 stays in the vicinity of the DNA strand break. 1011 ' Anot h e r n u cl ea r member of the PARP famil" y, PARP - 2 , is also able to catalyze DNA damage-dependent automodification 12 and can act as a catalytic homodimer, or together with PARP-1 as a heterodimer,13 in the response to genotoxic stress. The effector step of signaling consists in the selective recruitment of poly(ADP-ribose)-binding proteins to the vicinity of DNA breaks. 3'4'1°'11These proteins may either directly participate in DNA base excision repair or coordinate repair with chromatin unfolding, cell cycle progression, and cell survival or cell death pathways. The relative aflqnity and local availability ofpoly(ADP-ribose) binding partners, associated with the extent of DNA damage, may determine the type of response and the signaling outcome. For instance, PARP-1 automodification may allow the rapid recruitment of the BER complex as a primary response to the DNA strand breaks, l°'rl and then recruit p53 and modulate its multiple signal functions leading to cell cycle arrest or cell death. 14

DNA Damage Signaling through Poly(ADP-Ribose)

43

The amount of XRCC1, the scaffold protein on which the BER complex is assembled 15 and also a poly(ADP-ribose) binder, 3 might determine the number of repair foci and the threshold above which survival/death programs are activated. Finally, signal termination is achieved after poly(ADP-ribose) degradation by PARG. This enzyme disengages poly(ADP-ribose)-bound proteins and reverses the automodification status of PARP-1/PARP-2, which are now ready for a new round of DNA strand break binding. DNA damage induced polymers are degraded several hundred fold faster than constitutive polymers of undamaged cells. 16By virtue of dynamic and reversible automodifications, PARP-1 and PARP-2 can rapidly change the spectrum of partner proteins for recruitment into multiprotein complexes. PARPs may acquire increased binding affinity and/or new partner specificity upon automodification (vide infra). The following sections summarize some of the evidence leading to this model. A focal question was: which proteins become targets of poly(ADP-ribose)-binding and what are the consequences of polymer-binding on specific domain functions of these proteins?

Protein Targeting by Poly (ADP-Ribose) How can poly(ADP-ribose) target proteins in a chromatin environment? (ADPribose) polymers are variously sized acidic molecules, some of them containing branches. 17 The ribose-phosphate-phosphate-ribose backbone of poly(ADP-ribose) has a higher negative charge density than DNA and therefore may attract basic proteins from DNA. 18'19 The helical conformation 2° and the branched structure of long polymers, might also be involved in conferring some binding specificity. 21 The first evidence that poly(ADP-ribose) might play a role beyond that of a posttranslational protein modification was presented by Ohashi et al reporting that the activity of DNA ligase in reconstituted chromatin could be stimulated by either polymer addition or by PARP-1-bound polymers, synthesized in the course of the incubation in vitro. 22 It turned out that histones bind directly to poly(ADP-ribose) or PARP-l-bound polymers and this can cause the release of DNA from nucleosomal core particles. 19'21'23-25The observation that poly(ADP-ribose) degradation by PARG restores the nucleohistone structure led to the mechanistic model of a histone shuttle. 24Thus, the PARP-1/PARG system has the capacity to target histones for reversible dissociation from DNA. The potential biological relevance of this phenomenon lies in the fact that poly(ADP-ribose), by virtue of its affinity for histones, might act in vivo as a catalyst ofnucleosomal unfolding, by transiently displacing histones from DNA and hence facilitating DNA access to repair proteins in localized areas of the chromatin. 24 Indeed, histones exhibit a high preference for poly(ADP-ribose)-binding in the presence of DNA; a polymer of 40 ADP-ribose residues is sufficient to dissociate the entire histone complement of a chromatosome. Histones display different affinity for poly(ADP-ribose), the hierarchy of binding being HI>H2A>H2B=H3>H4. For all of them, however, interaction is far stronger and more specific than would be expected on the basis of electrostatic interactions. For instance, poly(ADP-ribose)-bound histones resist phenol partitioning, strong acids, detergents, and high salt concentrations. 21 An additional element of specificity is represented by the fact that protein basicity and/or DNA binding ability are not sufficient to confer affinity for poly(ADP-ribose). On the other hand, size and branching of (ADP-ribose) polymers are important determinants of bindin~ as branched polymers are a highly preferred target, followed by long linear moleoAes.21These classesof polymers are also synthesized in vivo and their levels increase in response to DNA damage.26-29The am~ingly high specificity ofhistone-poly(ADP-ribose) interactions could be explained by the discovery that binding only occurs at specific histone domains (C-terminus of histone H 1 and N-terminal tails of core histones) 3° and within such domains, only at distinct sequences that define, over a stretch of 20-26 amino acids, a highly homologous binding motif. 3 Amino acid conservation within these sequences entails the physicochemical properties of specific residues rather than their identity; the binding motif comprises at its C-terminal part a block of regularly spaced hydrophobic and basic residues, that by mutational analysis have been found to be critical for binding; 3 the interaction with poly(ADP-ribose) is

44

Poly(ADP-Ribosyl)ation

further strengthened by, but not absolutely dependent on, flanking arginines or lysines and a cluster of basic amino acids at the N-terminus. These rules initially defined in the authors' laboratory, 3'4 have recendy been confirmed by Poirier's group. 31 Screening of protein sequence databases with the polymer-binding consensus motif has led to the identification of other potential poly(ADP-ribose) interaction partners (Table 1), many of which are direcdy involved in the cellular response to DNA damage, at the level of damage recognition and processing (i.e., XPA, XRCC-1, MSH6, DNA ligase III, DNA polymerase E) and/or in later events responsible for cell cycle regulation/apoptosis (i.e., p53, p21, NF-kB, iNOS, DNA-PK, caspase activated DNase). In view of the widespread occurrence of PARPs in different cellular compartments, 1'2it is noteworthy that MARCKS (Myristoylated-Alanine-Rich-C-kinase-Substrate) and MRP (MARCKS-related protein), proteins regulating rearrangements of the actin cytoskeleton, also carry a poly(ADP-ribose) binding motif in their effector domain. 33 For most ofthe proteins listed in (Table 1), actual poly(ADP-ribose) binding has been biochemically confirmed by photoaffinity labeling and/or by a polymer blot binding assay using synthetic peptides, covering the putative interaction sequences, and/or on full length proteins, immobilized on nitrocellulose. Recendy, a combination of liquid-phase isoelectric focusing for protein extract fractionation, poly(ADP-ribose) binding testing by the blot assay, and MALDI-TOF mass spectrometry for protein identification, has allowed large scale screening and characterization of various heterogeneous nuclear ribonucleoproteins (hnRNPs) as poly(ADP-ribose) interaction partners. 31 These hnRNPs share a highly homologous poly(ADP-ribose) binding motif in the RNA recognition domain. In addition to the proteins listed in (Table 1), for which the binding sites have been elucidated, the following polymer-binding proteins have been identified: caspase 7, 34the 20S proteasome, 35 the telomere binding protein TRF-2, 36 lamins 31 and several other nuclear and nuclear matrix proteins, whose identity has yet to be established. 29'37It is predictable that the in silico approach combined with biochemical testings, 3'4 and protein fractionation techniques in conjunction with mass spectrometry31 will reveal a much larger family of polymer-binding proteins. The biological relevance of poly(ADP-ribose) interaction with the identified protein targets should, however, be tested with appropriate functional assays.

Polymer-Binding 'Reprograms' Domain Functions of Proteins How does polymer-binding affect domain functions in proteins? - Firstly, the binding of poly(ADP-ribose) to the polymer-binding consensus motif is very strong. 3'21 Secondly, in almost all cases studied so far, thepolymer-binding sequence overlaps with strategic functional domains in the target protein. 3'4,14 . . . 31 . . 33 36 38 Thirdly, ~polymer binding may enhance and inhibit separate domain functions in the same protein. 3~The tumor suppressor protein p53 is a particularly well studied example: the polymer-binding sites colocalize with the sequence specific DNA binding domain (residues 153-181 and 231-256) and with the C-terminal domain (amino acids 326-351), containing nuclear localization signal, nuclear export signal and tetramerization funtion. 14p53 plays a key role in transduction pathways induced by several types of cellular stress, by regulating the expression of gene products that can either lead to cell cycle arrest in G 1, thereby preventing the replication of DNA before the damage is repaired, or cause cell death by apoptosis. 39In vitro studies have demonstrated that poly(ADP-ribose) binding at the target sites is able to block (or reverse) p53 association both with ssDNA and, at higher concentrations, with a ds-oligonucleotide containing a p53 consensus sequence. 14Thus, p53 may differentially respond to DNA damage-induced poly(ADP-ribose): at low levels of DNA damage, a few polymers clustered on PARP-1/PARP-2 could block the ssDNA binding function and favour the transcriptional activity of p53. Conversely, high amounts of poly(ADP-ribose), associated with excessive DNA damage and massive NAD + consumption, could inhibit p53 activities completely, and thus contribute to directing cells towards caspase-independent programmed cell death 40'41 or necrosis.42 Hence, poly(ADP-ribose) metabolism might operate as a dual mechanism that activates p53-dependent and p53-independent

DNA DamageSignalingthroughPoly(ADP-Ribose)

45

0@ °..

0 0

k

oO

o

Z

N"

0

o.Ouc.,

C

.o_ c

~'~ L

~'~ ~o= =

•

&l.

•

6~

~z

n'~o

0@

~ii',',~i~i!~ii~

v rr v~

~

~

~

=~= ~ ~ ~

~< ~ Z ?~

rr W -I-

~

rr n'n-

~

~iiii~',',~~'~i~,!~~

=o'~ •~

k

v v Z v v u_

rr v v x- 0 F-rrW

~,~

:~ ~.~.~

I(j ~

~ -CO ~

" ~',~ii~',~::@:iii~!ili:~:i',i~',',~::~ i~','~ i~i '~

v _Joo ~¢Z w

~i~iiii~i~',~i~i'J@: !~i~

C

........ - ~ Q O 0 ( O @0 °C ~

C °..

7rowrToD-->-

W r T Y C ~ > ~ :

cO

•. . J O - J w w--o~

O~l--_JWn" ~n'~OCC_j

Y.-Jw~c~

O_

>->-~W

:~-r

u.

~

~

w>..'~'~

rr 3¢

0
_

> > v . J c C O ~ ~ o r r ~ o rr

~oZ u-~ o. rr

F-Z

6 0¢ ._ no .I ,L ,,+--,, t" t~ C"

t~

E £ ',4--

C

-r'.E ~-5

O .-t--

,..0

.E

~o

.,9o

°_

< ,.'-

"7

0 r-

~-t~

•-

N'I"

._ tt.i

"7


~

-~'-

c ~.E

~

-~

,.,_. ,.-

-~
350). Bars represent mean + SEM of at least four experiments. Intense NMDA exposure elicited the formation of a high intensity PAR fluorescence peak in the neuronal population, indicative of the neosynthesis of PAR, that was prevented by preincubation with DPQ. ** p < 0.01 vs. control; # p < 0.01 vs, control and p < 0.05 vs. NMDA (ANOVA + Tukey's w-test). to 300 }aM N M D A for 10 min. The PARP- 1 inhibitor was unable to modify caspase-3 activity in cortical cells following mild or intense N M D A insults, except for a slight but significant transient increase in the enzyme activity that was observed 6 h after exposure to 2 m M N M D A . Finally, when we examined the effects of D P Q on the neuronal levels of ATP, the PARP-1 inhibitor had little effect following exposure to 300 gM N M D A or 6 h after exposure to 2 m M N M D A but restored the levels of ATP almost completely (from 4 + 4% to 72 + 5% of control levels) 24 h after the intense N M D A exposure.

160

Poly(ADP-Ribosyl)ation

These results are in line with previous literature data showing that necrotic cell death is prevented by targeted deletion of the PARP-1 gene or by pharmacological inhibition of the enzyme, while apoptosis is unaffected. 19'32 PARP-1 deletion or inhibition have been shown to attenuate cell injury in models in which the type of death is predominantly necrotic, including cerebral 25'26 and myocardial ischemia, 56 and streptozotocin-induced diabetes. 57'58 In contrast, reduction of PARP-1 activity does not protect against killing of hepatocytes by TNF-cz with actinomycin D or the death of thymocytes elicited by ceramide, dexamethasone, CD-95, or ionomycin, which are forms of apoptotic cell death. 59 Although PARP-1 activity is also known to affect cell death and survival in a DNA-independent manner via the regulation of transcription factors, 12'6° the finding that D P Q spares ATP in surviving neurons exposed to severe NMDA incubation, as previously demonstrated in other cell types, 61 suggests that in our necrotic model PARP-1 is overactivated and leads to neuronal death through the depletion of NAD and ATP cellular stores (the so-called "suicidal role"). In the milder model, PARP-1 cleavage by caspase-3 activation facilitates apoptosis indirectly by preventing the energy failure induced by PARP-1 overactivation, thereby preserving cellular ATP that is essential for the apoptotic process. It is interesting to note that 6 h after intense NMDA exposure, DPQ-pretreated cells exhibited a significant increase in caspase-3-1ike activity that was associated with an increase in the number of apoptotic cells. These data suggest that, as a result of PARP-1 inhibition and the subsequent recovery of ATP levels, there could be a shift in the type of cell death, allowing some cells that would have otherwise died by necrosis to die by apoptosis. In accordance with this view, previous studies have reported that the maintenance of cellular energy levels induced by PARP-1 inhibition after injury permits caspase activation and switches the type of cell death from necrosis to apoptosis. 19'36'-62'63The recovery of energy levels induced by PARP-1 inhibition can also allow the survival of some cells destined to die by apoptosis, as reported in PC 12 cells exposed to oxidative damage. 64 PARP-1 activity increases at early time points after neuronal injury. The formation of PAR in the rat neocortex peaks at 30 min and 2 h after experimental traumatic brain injury before returning to baseline levels. 65 In cortical neurons exposed to NMDA, PAR immunoreactivity can be revealed as early as 15 min following NMDA exposure, with a maximum 1 h later. 16Therefore, we examined the extent of poly(ADP-ribosyl)ation 1 h after mild or intense NMDA exposure in this study and we observed that the neuronal formation of PAR was increased in a DPQ-sensitive manner in the intense (Fig. 3B) but also in the mild model. These findings confirm that PARP-1 may be transiently activated in the early phases of apoptosis. 51'66'67 Soon afterwards, however, the protein is cleaved and inactivated by caspase-3, thus preventing ATP depletion by PARP-1 overactivation and affording the energy required for the apoptotic active process. This idea is supported by the early but transient reduction in ATP levels following mild NMDA exposure, that was soon restored to levels similar to those observed following prolonged incubation with the pro-apoptotic PKC inhibitor staurosporine.

Concluding Remarks Excitotoxicity (glutamate-mediated neuronal death) is responsible for numerous neurological and psychiatric conditions that have a high social and financial impact on the society. In several animal models of neurodegenerative diseases, PARP-1 genetic or pharmacological inhibition provides impressive and unparalleled protection, suggesting an importat role for poly(ADP-ribosyl)ation in excitotoxicity. The identification of pathways through which PAR formation fuels excitotoxic neuronal death may increase our understanding of neurodegenerative processes paving the way to innovative therapeutic approaches.

Poly(ADP-Ribose) Polymerase (PARP) and Excitotoxicity

161

References 1. Olney JW. Excitotoxic amino acids and neuropsychiatric disorders. Annu Rev Pharmacol Toxicol 1990; 30:47-71. 2. Lipton SA, Rosenberg PA. Excitatory amino acids as a final common pathway for neurological disorders. N Engl J Med 1994; 330:613-622. 3. Doble A. The role of excitotoxicity in neurodegenerative disease: Implications for therapy. Pharmacol Ther 1999; 81:163-221. 4. Lipton P. Ischemic cell death in brain neurons. Physiol Rev 1999; 79:1431-1568. 5. Meldrum BS. Glutamate as a neurotransmitter in the brain: Review of physiology and pathology. J Nutr 2000; 130:1007S- 1015S. 6. Lee J-M, Zipfel GJ, Choi DW. The changing landscape of ischaemic brain injury mechanisms. Nature 1999; 399(supp.):A7-A14. 7. Dirnagl U, Iadecola C, Moskowitz MA. Pathobiology of ischaemic stroke: An integrated view. Trends Neurosci 1999; 22:391-397. 8. Gladstone DJ, Black SE, Hakim AM. Toward wisdom from failure: Lessons from neuroprotective stroke trials and new therapeutic directions. Stroke 2002; 33:2123-2136. 9. Grotta J. Neuroprotection is unlikely to be effective in humans using current trial designs. Stroke 2002; 33:306-307. 10. Aarts MM, Tymianski M. Novel treatment of excitotoxicity: Targeted disruption of intracellular signalling from glutamate receptors. Biochem Pharmacol 2003; 66:877-886. 11. Berger NA. Poly(ADP-ribose) in the cellular response to DNA damage. Radiat Res 1985; 101:4-15. 12. Chiarugi A. Poly(ADP-ribose) polymerase: Killer or conspirator? The 'suicide hypothesis' revisited. Trends Pharmacol Sci 2002; 23:122-129. 13. Shall S, de Murcia G. Poly(ADP-ribose) polymerase-l: What have we learned from the deficient mouse model? Mutat Res 2000; 460:1-15. 14. Szab6 C, Dawson VL. Role of poly(ADP-ribose) synthetase in inflammation and ischaemia-reperfusion. Trends Pharmacol Sci 1998; 19:287-298. 15. Pieper AA, Verma A, Zhang J et al. Poly(ADP-ribose) polymerase, nitric oxide and cell death. Trends Pharmacol Sci 1999; 20:171-181. 16. Yu S-W, Wang H, Poitras MF et al. Mediation of poly(ADP-ribose) polymerase-l-dependent cell death by apoptosis-inducing factor. Science 2002; 297:259-263. 17. Cosi C, Suzuki H, Milani D et al. Poly(ADP-ribose) polymerase: Early involvement in glutamate-induced neurotoxicity in cultured cerebellar granule cells. J Neurosci Res 1994; 39:38-46. 18. Zhang J, Dawson VL, Dawson TM et al. Nitric oxide activation of poly(ADP-ribose) synthetase in neurotoxicity. Science 1994; 263:687-689. 19. Ha HC, Snyder SH. Poly(ADP-ribose) polymerase is a mediator of necrotic cell death by ATP depletion. Proc Nad Acad Sci USA 1999; 96:13978-13982. 20. Mandir AS, Poitras MF, Berliner AR et al. NMDA but not nonNMDA excitotoxicity is mediated by poly(ADP-ribose) polymerase. J Neurosci 2000; 20:8005-8011. 21. Choi DW. Excitotoxic cell death. J Neurobiol 1992; 23:1261-1276. 22. Meldrum B, Garthwaite J. Excitatory amino acid neurotoxicity and neurodegenerative disease. Trends Pharmacol Sci 1990; 11:379-387. 23. Pellegrini-Giampietro DE, Peruginelli F, Meli E et al. Protection with metabotropic glutamate 1 receptor antagonists in models of ischemic neuronal death: Time-course and mechanisms. Neuropharmacology 1999; 38:1607-1619. 24. Nicoletti F, Bruno V, Copani A et al. Metabotropic glutamate receptors: A new target for the therapy of neurodegenerative disorders? Trends Neurosci 1996; 19:267-271. 25. Eliasson MJL, Sampei K, Mandir AS et al. Poly(ADP-ribose) polymerase gene disruption renders mice resistant to cerebral ischemia. Nat Med 1997; 3:1089-1095. 26. Endres M, Wang Z-Q, Namura S et al. Ischemic brain injury is mediated by the activation of poly(ADP-ribose) polymerase. J Cereb Blood Flow Metab 1997; 17:1143-1151. 27. Goto S, Xue R, Sugo N e t al. Poly(ADP-ribose) polymerase impairs early and long-term experimental stroke recovery. Stroke 2002; 33:1101-1106. 28. Takahashi K, Greenberg JH, Jackson P et al. Neuroprotective effects of inhibiting poly(ADP-ribose) synthetase on focal cerebral ischemia in rats. J Cereb Blood Flow Metab 1997; 17:1137-1142. 29. Meli E, Pangallo M, Baronti R et al. Poly(ADP-ribose) polymerase as a key player in excitotoxicity and post-ischemic brain damage. Toxicol Lett 2003; 193:153-162.

162

Poly(ADP-Ribosyl)ation

30. Chiarugi A, Meli E, Calvani M e t al. Novel isoquinolinone-derived inhibitors of poly(ADP-ribose) polymerase-l: Pharmacological characterization and neuroprotective effects in an in vitro model of cerebral ischemia. J Pharmacol Exp Ther 2003; 305:943-949. 31. Giovannelli L, Cozzi A, Guarnieri I e t al. Comet assay as a novel approach for studying DNA damage in focal cerebral ischemia: Differential effects of NMDA receptor antagonists and poly(ADP-ribose) polymerase inhibitors. J Cereb Blood Flow Metab 2002; 22:697-704. 32. Moroni F, Meli E, Peruginelli F et al. Poly(ADP-ribose) polymerase inhibitors attenuate necrotic but not apoptotic neuronal death in experimental models of cerebral ischemia. Cell Death Differ 2001; 8:921-932. 33. Nicotera P, Lipton SA. Excitotoxins in neuronal apoptosis and necrosis. J Cereb Blood Flow Metab 1999; 19:583-591. 34. Wang KKW. Calpain and caspase: Can you tell the difference? Trends Neurosci 2000; 23:20-26. 35. Bowes J, Thiemermann C. Effects of inhibitors of the activity of poly (ADP-ribose) synthetase on the liver injury caused by ischaemia-reperfusion: A comparison with radical scavengers. Brit J Pharmacol 1998; 124:1254-1260. 36. Leist M, Single B, Castoldi AF et al. Intracellular adenosine triphosphate (ATP) concentration: A switch in the decision between apoptosis and necrosis. J Exp Med 1997; 185:1481-1486. 37. Gwag BJ, Lobner D, Koh J-Y et al. Blockade of glutamate receptors unmasks neuronal apoptosis after oxygen-glucose deprivation in vitro. Neuroscience 1995; 68:615-619. 38. Martin LJ, AI-Abdulla NA, Brambrink AM et al. Neurodegeneration in excitotoxicity, global cerebral ischemia, and target deprivation: A perspective on the contributions of apoptosis and necrosis. Brain Res Bull 1998; 46:281-309. 39. Charriaut-Marlangue C, Aggoun-Zouaoui D, Represa A et al. Apoptotic features of selective neuronal death in ischemia, epilepsy and gpl20 toxicity. Trends Neurosci 1996; 19:109-114. 40. Kaminski Schierle GS, Hansson O, Ferrando-May E et al. Neuronal death in nigral grafts in the absence of poly (ADP-ribose) polymerase activation. NeuroReport 1999; 10:3347-3351. 41. van Lookeren Campagne M, Gill R. Ultrastructural morphological changes are not characteristic of apoptotic cell death following focal cerebral ischemia in the rat. Neurosci Lett 1996; 213:111-114. 42. Rosenblum WI. Histopathological clues to the pathways of neuronal death following ischemia/ hypoxia. J Neurotrau 1997; 14:313-326. 43. MacManus JP, Hill IE, Preston E et al. Differences in DNA fragmentation following transient cerebral or decapitation ischemia in rats. J Cereb Blood Flow Metab 1995; 15:728-737. 44. Martinou J-C, Dubois-Dauphin M, Staple JK et al. Overexpression of BCL-2 in transgenic mice protects neurons from naturally occurring cell death and experimental ischemia. Neuron 1994; 13:1017-1030. 45. Hara H, Friedlander RM, Gagliardini V e t al. Inhibition of interleukin 113 converting enzyme family proteases reduces ischemic and excitotoxic neuronal damage. Proc Nad Acad Sci USA 1997; 94:2007-2012. 46. Namura S, Zhu J, Fink K et al. Activation and cleavage of caspase-3 in apoptosis induced by experimental cerebral ischemia. J Neurosci 1998; 18:3659-3668. 47. Gottron FJ, Ying HS, Choi DW. Caspase inhibition selectively reduces the apoptotic component of oxygen-glucose deprivation-induced cortical neuronal cell death. Mol Cell Neurosci 1997; 9:159-169. 48. Okamoto M, Matsumoto M, Ohtsuki T et al. Internucleosomal DNA cleavage involved in ischemia-induced neuronal death. Biochem Biophys Res Commun 1993; 196:1356-1362. 49. Nitatori T, Sato N, Waguri S et al. Delayed neuronal death in the CA1 pyramidal cell layer of the gerbil hippocampus following transient ischemia is apoptosis. J Neurosci 1995; 15:1001-1011. 50. Colbourne F, Sutherland GR, Auer RN. Electron microscopic evidence against apoptosis as the mechanism of neuronal death in global ischemia. J Neurosci 1999; 19:4200-4210. 51. Simbulan-Rosenthal CM, Rosenthal DS, Iyer Set al. Transient poly(ADP-ribosyl)ation of nuclear proteins and role of poly(ADP-ribose) polymerase in the early stages of apoptosis. J Biol Chem 1998; 273:13703-13712. 52. Nicholson DW, Thornberry NA. Caspases: Killer proteases. Trends Biochem Sci 1997; 22:299-306. 53. Cosi C, Colpaert F, Koek W e t al. Poly(ADPribose) polymerase inhibitors protect against MPTP-induced depletions of striatal dopamine and cortical noradrenaline in C57B1/6 mice. Brain Res 1996; 729:264-269. 54. Meli E, Pangallo M, Picca R et al. Differential role of poly(ADP-ribose) polymerase-1 in apoptotic and necrotic neuronal death induced by mild or intense NMDA exposure in vitro. Mol Cell Neurosci 2004; 25:172-180.

Poly(ADP-Ribose) Polymerase (PARP) and Excitotoxicity

163

55. Bonfoco E, Krainc D, Ankarcrona M et al. Apoptosis and necrosis: Two distinct events induced, respectively, by mild and intense insults with N-methyl-D-aspartate or nitric oxid/superoxide in cortical cell cultures. Proc Natl Acad Sci USA 1995; 92:7162-7166. 56. Pieper AA, Walles T, Wei G et al. Myocardial postischemic injury is reduced by polyADPripose polymerase-1 gene disruption. Mol Med 2000; 6:271-282. 57. Burkart V, Wang ZQ, Radons J et al. Mice lacking the poly(ADP-ribose) polymerase gene are resistant to pancreatic beta-cell destruction and diabetes development induced by streptozocin. Nat Med 1999; 5:314-319. 58. Pieper AA, Brat DJ, Krug DK et al. Poly(ADP-ribose) polymerase-deficient mice are protected from streptozotocin-induced diabetes. Proc Natl Acad Sci USA 1999; 96:3059-3064. 59. Leist M, Single B, Kunstle G et al. Apoptosis in the absence of poly-(ADP-ribose) polymerase. Biochem Biophys Res Commun 1997; 17:518-522. 60. Ziegler M, Oei SL. A cellular survival switch: Poly(ADP-ribosyl)ation stimulates DNA repair and silences transcription. Bioessays 2001; 23:543-548. 61. Lee YJ, Shacter E. Oxidative stress inhibits apoptosis in human lymphoma cells. J Biol Chem

1999; 274:19792-19798. 62. Eguchi Y, Shimizu S, Tsujimoto Y. Intracellular ATP levels determine cell death fate by apoptosis or necrosis. Cancer Res 1997; 57:1835-1840. 63. Walisser JA, Thies RL. Poly(ADP-ribose) polymerase inhibition in oxidant-stressed endothelial cells prevents oncosis and permits caspase activation and apoptosis. Exp Cell Res 1999; 251:401-413. 64. Cole KK, Perez-Polo JR. Poly(ADP-ribose) polymerase inhibition prevents both apoptotic-like delayed neuronal death and necrosis after H(2)O(2) injury. J Neurochem 2002; 82:19-29. 65. LaPlaca M, Raghupathi R, Verma A et al. Temporal patterns of poly(ADP-ribose) polymerase activation in the cortex following experimental brain injury in the rat. J Neurochem 1999; 73:205-213. 66. Scovassi AI, Poirier GG. Poly(ADP-ribosylation) and apoptosis. Mol Cell Biochem 1999; 199:125-137. 67. Boulares AH, Yakovlev AG, Ivanova V e t al. Role of poly(ADP-ribose) polymerase (PARP) cleavage in apoptosis. Caspase 3-resistant PARP mutant increases rates of apoptosis in transfected cells. J Biol Chem 1999; 274:22932-22940.

CHAPTER 15

Poly(ADP-Ribose)Polymerase and Ischemia-Reperfusi0nInjury Prabal K. Chatterjee and Christoph Thiemermann

Abstract p

oly (adenosine 5'-diphosphate ribose) polymerase-1 (PARP-1) is an abundant chromatin-bound enzyme which is present in the nuclei of most cells. The physiological role of PARP-1 involves its activation by single strand breaks in DNA after which it transfers ADP-ribose moieties from nicotinamide adenine dinucleotide (NAD ÷) to various nuclear proteins including histones and even to PARP-1 itself (automodification) forming extended chains of ADP-ribose. This reaction leads to the generation of nicotinamide, which, via negative feedback, inhibits PARP-1 activity. However, continuous or excessive activation of PARP-1 (and perhaps other, less well characterized members of the family of poly[ADP-ribose] polymerases, collectively termed "PARP") during ischemia-reperfusion (I-R) leads to excessive PARP activation resulting in a substantial depletion in intracellular levels of NAD ÷. As NAD ÷ functions as an electron carrier in the mitochondrial respiratory chain, its depletion rapidly leads to a fall in intracellular levels of adenosine triphosphate (ATP). Moreover, nicotinamide can be recycled back to NAD + in a reaction which consumes ATP and thus continuous/excessive activation of PARP results in a fall in ATP via two different mechanisms ultimately leading to cell deathma pathophysiological process commonly referred to as The PARP Suicide Hypothesis. Oxygen-derived radicals such as superoxide anions and hydroxyl radicals cause strand breaks in DNA, activation of PARP and depletion ofNAD ÷and ATP. Peroxynitrite, which is generated when equimolar amounts of nitric oxide react with superoxide anions, also causes strand breaks in DNA, activation of PARP and ultimately cell death. Inhibitors of PARP activity can reduce the renal dysfunction caused by I-R of the kidney in vivo and also reduce cellular injury and death caused by oxidative stress to renal cells in culture. In this chapter we review the current evidence that PARP activation plays an important role of the pathophysiology of renal I-R injury and how PARP inhibitors could provide a new therapeutic strategy for patients suffering I-R injury of the kidney.

Introduction Poly (adenosine 5'-diphosphate ribose) polymerase-1 (PARP-1) [also known as poly (ADP-ribose) synthetase, PARS; E.C. 2.4.2.30] is a chromatin-bound enzyme, which is abundandy present in the nuclei of numerous cell types. Single strand breaks in DNA trigger the activation of PARP-1, which transfers ADP-ribose moieties from nicotinamide adenine dinucleotide (NAD ÷) to various nuclear proteins including histones and PARP (automodification domain) itself. This reaction leads to the generation of nicotinamide, which is an inhibitor (negative feedback) of PARP-1 activity. Continuous or excessive activation of PARP-1 (and perhaps other, less well characterized members of the PARP enzyme family) produces extended

Poly(ADP-Ribosyl)ation, edited by Alexander Biirkle. 02006 and Springer Science+Business Media.

Landes Bioscience

Poly(ADP-Ribose)Polymeraseand Ischemia-Repe~CusionInjury

165

chains of ADP-ribose on nuclear proteins and results in a substantial depletion of intracellular NAD +. As NAD ÷functions as an electron carrier in the mitochondrial respiratory chain, NAD ÷ depletion rapidly leads to a fall in intracellular levels of adenosine triphosphate (ATP). Moreover, nicotinamide can be recycled to NAD +in a reaction that consumes ATE Thus, excessive activation of PARP leads to a fall in ATP (by two different mechanisms), which may ultimately cause cell death. 1'2 Oxygen-derived radicals such as superoxide anions (02"-) and hydroxyl radicals (OH') which are formed during ischemia-reperfusion (I-R), cause strand breaks in DNA, activation of PARP and depletion of NAD + and ATE Furthermore, peroxynitrite (ONOO-), which is also generated during I-R when equimolar amounts of nitric oxide (NO') react with O2"', also causes strand breaks in DNA, activation of PARP and ultimately cell death. 3 In the following chapter, we will review the role of PARP in the development of I-R injury with a particular focus on the kidney. Finally, we will review the beneficial actions of PARP inhibitors against I-R injury and how PARP inhibitors could provide a new therapeutic strategy for patients suffering I-R injury of the kidney which is a major contributor to the development of acute renal failure (ARF).

Ischemia-ReperfusionInjury It is now accepted that I-R injury is a major contributor to the pathophysiology of many life-threatening diseases. 4-6 Tissues require a continuous supply of molecular oxygen (O2) in order to survive and maintain normal physiological function and this is severely compromised during ischemia. Following a period of ischemia, reperfusion of ischaemic tissue with oxygenated blood is essential for reducing or preventing the cell death associated with ischemia. However, although beneficial, reperfusion of ischaemic tissue also contributes to the overall injury caused by I-R--a phenomenon known as "reperfusion-injury". Thus while ischaemic-injury is precipitated by the lack of 02, reperfusion-injury is associated with the return of 02 (Fig. 1). During the ischaemic phase, hypoxia per se is considered to be the key factor in the development of ischaemic-injury. The series of disturbances which occur during ischemia are precipitated by a lack of 02, leading to impaired cellular function and ultimately, to cell death. Decreased oxygen supply to tissues results in impairment of mitochondrial oxidative phosphorylation, with anaerobic glycolysis becoming the only available energy source. Consequently, cellular levels of ATP will decrease. Thus, a lack of 02 and ATP depletion play a central role in the development of ischaemic-injury (Fig. 1). Furthermore, following prolonged ischemia and despite the restoration of oxygenated blood flow, the ability of tissues to regenerate ATP may also be impaired. During ischemia, ATP, ADP and adenosine monophosphate are degraded by the action of the enzyme 5'-nudeotidase, resulting in nucleoside adenosine which diffuses freely through the cell membrane. During reperfusion, these purine metabolites are washed out and thus the substrates required for ATP generation are lost from the post-ischaemic cells. Ischemia also produces characteristic changes to cell structure. Early changes include reversible cell swelling due to lack of ATP which disrupts the normal function of the Na+-K+-ATPase pump which normally extrudes sodium ions from cells. The resulting electrolyte imbalance and influx of water into cells causes swelling of the cells and its organelles. During prolonged ischemia, cellular swelling is followed by a loss of mitochondrial matrix and finally by disintegration, with expansion and formation of vesicles in the endoplasmic reticulum and cytoplasm. The final steps preceding cell death involve lysosomal rupture with accompanying enzyme release. The damage caused by ischemia does not become apparent until the organ is reperfused with oxygenated blood and it is now dear that reperfusion exacerbates the injury which occurs during ischemia. This indicates that reperfusion-injury is caused as a consequence of return of oxygenated blood to ischaemic tissue and is supported by the fact that anoxic reperfusion of ischaemic tissues results in litde additional injury. It is therefore apparent that a major contributor to the development of I-R injury is the formation of cytotoxic oxidants derived from 02, as discussed in greater detail below. Ischemia leads to intracellular accumulation of hypoxanthine and conversion of xanthine dehydrogenase (which normally reduces NAD +) to xanthine oxidase. During reperfusion, the reaction of 02 with hypoxanthine and xanthine oxidase

166

Poly(ADP-Ribosyl)at~on o u

.o ,.Q °,.q

¢j

0 ~j

O~ O= O~

]

~_~,,=, ._

~

=

I~1 I01

i ~I~

=_ g

,,--.-. I

c:

O i

~g h- .J 0 l(J

iOe';

I-'-

v

r-----l

~

Uj

~"- -.~

~:

,~ ~ , , , -

(~

Z

"r

~

Z

uJ

(,,,) < { ! @ "

o.~

.~..-~

w

O. ---~

o~

o~

Poly(ADP-Ribose) Polymeraseand Ischemia-Repe~CusionInjury

167

leads to the production of reactive oxygen species (ROS) such as O2"-, hydrogen peroxide (H202) and OH'. Ischemia is essential for this process, as reduction in PO2 is essential for the activation ofxanthine oxidase and the substrates used by xanthine oxidase are generated by the ischemia-induced breakdown of adenine nucleotides to hypoxanthine. However, it should be noted that reperfusion-injury only exacerbates ischaemic-injury within a limited time subsequent to ischemia as prolonged ischemia alone results in cell death after which restoration of blood flow will not affect the viability of the tissue.

Renal Ischemia-Reperfusion Injury The primary factor in ischemia of the kidney is a reduction in renal blood flow, which may occur in association with large vessel renal vascular disease involving renal artery thrombosis, embolism and atherosclerosis. 7 Significant renal ischemia is also a common problem during aortic surgery, renal transplantation or subsequent to cardiovascular anaesthesia, leading to renal dysfunction and injury. 8-1° For example, surgical procedures which involve clamping of the aorta and/or renal arteries (e.g., surgery for supra- and juxtarenal abdominal aortic aneurysms and renal transplantation) are particularly liable to produce renal ischemia, leading to significant post-operative complications including ARF.9'11;12The incidence of renal dysfunction in high-risk patients undergoing this type of surgery has been reported to be as high as 50 . . . . . . %. 1 .u Thus,. renal. ischemia is . now recognised as. the major cause of. ischaemic ARF. ~'-2 2 Renal epithelial proximal tubular (PT) cells are highly specialised both in terms of morphology and function, which allows for efficient transport of ions, water and macromolecules across cell layers. These transport mechanisms are highly selective in nature and are governed by intracellular energy. However, during renal ischemia these processes are severely perturbed 13'19'2° leading to cellular injury, dysfunction and death via a combination of necrosis and apoptosis. 21-24At the onset ofischemia, there is either a partial reduction or total cessation of blood flow through one or more arteries that feed specific vascular beds in the kidney, thus severely limiting the supply of 02. Such insult is sufficient to trigger a large decrease in ATP levels and the release of catecholamines. Intracellular accumulation ofNa + ions occurs which is accompanied by an influx of water leading to swelling of cells (especially endothelial cells). Concurrendy, loss of fluid from the intravascular space increases blood viscosity and increases haemoconcentration. Adhesion of blood cells, particularly polymorphonudear leukocytes (PMNs), further increases blood viscosity. Together, these changes interfere with the restoration of microcirculation during reperfusion thus leading to capillary obstruction ('no reflow' phenomenon). This scenario is complicated by the fact that reperfusion, although essential for the survival ofischaemic tissue, causes additional cellular injury (reperfusion-injury) 25-28and is also a major contributor to early allograft rejection subsequent to renal transplantation, and adversely affects the long-term allograft survival. 29Thus, early therapeutic intervention is likely to reduce the incidence of ARF and associated mortality following surgical manipulations which involve renal I-R.

Role of Reactive Oxygen Species During the process of normal cellular metabolism, 02 undergoes a series of univalent reductions leading sequentially to the production of O2"-, H202 and OH" 30 (Fig. 1). Several enzymes can generate ROS including components of the mitochondrial electron transport chain, xanthine oxidase, cytochrome P450 monooxygenases, lipoxygenase, nitric oxide synthase and NADPH oxidase. 31'32 Oxidative stress is caused by an imbalance in the ratio of oxidant and antioxidant, with an excess of oxidant relative to antioxidant capacity. There is now good evidence that oxidant stress contributes to development of I-R injury and many biochemical and immunohistochemical studies have demonstrated an important role for ROS. 6 '28 '30 '33 -39 Collectively, ROS are instrumental in impairing overall renal function 4°-42 (Fig. 1). ROS are widely implicated in the development of the tissue injury and dysfunction associated with I-R and cause endothelial cell damage and increased vascular permeability, PMN activation and infiltration into tissues, lipid peroxidation and oxidation, direct inhibition ofmitochondrial chain enzymes, inactivation

168

Poly(ADP-Ribosyl)ation

ofglyceraldehyde-3-phosphate dehydrogenase (GAPDH), inhibition of membrane Na÷-K+-ATPase activity, inactivation of membrane sodium channels, formation of chemotactic factors such as leukotriene B4 and other oxidative modifications 35'37'43-48(Fig. 1). Furthermore, during ischemia, degradation of ATP results in increased production of hypoxanthine, and xanthine and xanthine dehydrogenase is converted into xanthine oxidase35'44'49;5°(Fig. 1). In renal ischemia, tissue levels ofhypoxanthine can increase between 10 and 300-fold greater than normal. 51'51Reintroduction of O2 upon reperfusion leads to metabolism of purine by xanthine oxidase and production of O2"- 34,35,50,53,54(Fig. 1). O2"- is the one-electron reduction product of O2 and can act either as a reductant or as an oxidant; thus it readily reduces cytochrome c, tetranitromethane and certain metal chelates, but oxidises sulphite, catecholamines, tetrahydrobiopterins and leucoflavins. 53-55 In the intracellular environment, which is rich in reductants, O2" acts primarily as an oxidant. I-R is also a stimulus for PMN activation, PMN-endothelial interaction and infiltration of PMNs into ischaemic tissues where they act as another source of 02 .-.35 Several aspects of the injury caused by I-R can be attributed to the formation of O2"- including endothelial damage and increased microvascular permeability,56 formation ofchemotactic factors such as leukotriene B4,57 release of histamine from mast cells58 and DNA damage. 59 Catecholamines such as noradrenaline are deactivated by 02"- leading to the hypotension observed in septic shock. 6° Two reaction products can be formed from 02"-; (i) the hydroxyl radical (OH'), which is a powerful oxidant formed from 02"- via the Fenton and Haber-Weiss reactions and (ii) ONOO-(peroxynitrite), which is generated when equimolar amounts of 02"- react with NO" which can then decompose into OH" and NO2. Both OH" and ONOO- can cause DNA strand breakage with subsequent excessive activation of PARP, which contributes to the cellular dysfunction, injury and death observed during I-tL 61-69 Generation of ROS contributes significantly to the renal dysfunction and injury associated with I-R of the kidney and consequent . . ARE• 6 ,33,34 ,yo-74 In particular, the . susce p tibili ty of the. PT to renal I-R-injury leads to acute tubular necrosis, which plays a pivotal part in the pathogenesis ofARF, y2,r5Traditionally, ROS have been considered to exert their effects via a direct toxic action on target cells3y e.g., PARP activation during renal I-R subsequent to DNA damage by ROS leading to depletion ofNAD + and ATP levels and ultimately renal cell death. 68 However, recent studies have also suggested a contributory role for ROS in gene induction and may act as signal transduction molecules for transcription factors including nuclear factor-~cB (NF-~cB) and activator protein-1 AP-1. 76,ry The role of ROS in the pathogenesis of I-R-induced renal injury has been demonstrated in several studies in which the beneficial effects of various interventions which either reduce the generation of, or the effects of, ROS have been investigated. 19 These therapeutic strategies include administration of agents which prevent the generation of ROS such as deferoxamine or N-acetylcysteine, y8-81 inhibition of enzymes responsible for production of ROS, e.g., inhibition of xanthine oxidase by allopurinol, 8°'81 administration of antioxidant enzymes which degrade ROS, such as superoxide dismutase (SOD), catalase or SOD/catalase mimetics such as EUK- 134, 74'83'~ scavengers of ROS such as mannitol, 79 sodium benzoate, 84 uric acid, 79Tempol and Tempone • 71,73Additionall y , attenuation of ROS eneration will also reduce DNA strand breakage and consequently reduce PARP activation. 65 '~ '68 '85 Although these interventions have shown promise, the potential benefits of the systemic administration of these agents are limited due to confounding factors, e.g., the large chemical structure of SOD does not allow it to be effectively1f~ered at the glomerulus of the kidney and thus may not reach the PT in adequate amounts.

9,

Role of Reactive Nitrogen Species Production of NO" and its contribution to the development of renal I-R injury and associated ARF has been investigated and reviewed. 87-89 NO', produced by nitric oxide synthase (NOS) plays an important role in renal function, both under normal and pathophysiological conditions. 90-92 The three isoforms of NOS have been located in the kidney; the endothelial and neuronal (constitutive) isoforms have been identified in the renal vasculature and macula densa, respectively 93 and inducible NOS (iNOS) can be induced in the kidney by cytokines

Poly(ADP-Ribose) Polymeraseand Ischemia-Reper[usion Injury

169

and bacterial lipopolysaccharide (LPS), during I-R and under inflammatory conditions leading to renal toxicity. 87-89'94-1°3 Generally iNOS is only expressed after induction by appropriate stimuli, however, several tissues including the PT constitutively express iNOS mRNA. I°4 In situ hybridisation of normal rat kidneys using iNOS cDNA revealed that the $3 segment of the PT is the most intensely labelled nephron segment, whereas labelling of the $1 and $2 segments was much weaker.I°4 Furthermore, two isomers ofiNOS have been identified in the rat kidney; restriction cleavage analysis and DNA sequencing has been used to identify two structurally distinct iNOS mRNAs identified as vascular smooth muscle-type iNOS (vsm-iNOS) and macrophage type iNOS (mac-iNOS) 1°5 Many in vivo and in vitro investigations have demonstrated that inhibition or the absence of iNOS activity reduces renal I-R injury. 98"103'106-110 Conversely, it has also been shown that inhibition of constitutively expressed endothelial NOS (eNOS) can exacerbate renal I-R injury by promoting renal vasoconstriction and microvascular thrombosis 111,112and it has now been suggested that an imbalance between the expression and activity of iNOS and eNOS is an important contributor to the pathophysiology of ARE. 113 Together, these results suggest that NO" generated by NOS significantly modulates renal I-R injury. Furthermore, the reaction of NO" with O2"- to form ONOO- 114-119causes injury via direct oxidant injury and protein tyrosine nitration 12°-125 (Fig. 1). There is now substantial evidence that ONOO- accounts for a significant proportion of the tissue injury associated with I-R and inflammation. 126 Formation of ONOO- during I-R has been demonstrated in many organs induding the brain, heart, lung and gu t127"130as well as in the kidney. 106,131,132Furthermore, there is now good evidence that ONOO- can nitrate and deactivate antioxidant enzymes such as SOD 133' 134-and causes the inactivation of NO" 135

Poly(ADP-Ribose) Polymeraseand Ischemia-ReperfusionInjury The concept that the activation of PARP contributes to the pathophysiology of I-R injury has evolved over the last decade. In 1994, Zhang and colleagues reported that benzamide (and other inhibitors of PARP activity) could reduce the neurotoxic effects of N-methyl-D-aspartate and NO" in brain slices of the rat: a finding which suggested that the activation of PARP may contribute to the tissue injury associated with stroke. 136 In January 1997, we reported that the administration, prior to reperfusion, of several chemically distinct inhibitors of PARP activity including 1,5-dihydroxyisoquinoline (ISO), 3-aminobenzamide (3-AB), benzamide and nicotinamide reduces the infarct size caused by regional myocardial I-R in the anaesthetised rabbit. 137 We, therefore, proposed that (i) the activation of PARP contributes to the pathophysiology of I-R injury, and (ii) that inhibitors of this enzyme may be useful in the therapy of myocardial infarction and other conditions associated with I-R injury. 137 Our concept that the ability of PARP inhibitors to protect tissues/organs against I-R was not limited to the heart was based on the finding that these agents also reduced by 50% the degree of skeletal muscle necrosis caused by prolonged periods of hind limb I-IL 137 In an independent investigation published in November 1997, Zingarelli and colleagues also reported that the PARP-inhibitor 3-AB reduces the myocardial injury caused by regional myocardial I-R in the anaesthetised rat. 138Today, we know that the ability of PARP inhibitors to reduce myocardial infarct size is not limited to rodents, as the administration of 3-AB prior to reperfusion also causes the infarct size caused by occlusion and reperfusion of the left anterior coronary artery in the anaesthetised pig. 139 The protective effects of PARP inhibitors in I-R are likely to be due to the inhibition of PARP activity (rather than nonspecific effects), as the tissue injury caused by I-R of the heart, 140 brain 141 and gut 142 is also significantly reduced in PARP knock-out mice. Given that inhibitors of PARP activity cause a substantial reduction in infarct size in many organs including the heart and brain, it is not surprising that several pharmaceutical companies have selected "PARP" as a target for drug development. However, there is a great need for potent, water-soluble PARP-inhibitors for the following reasons: In contrast to certain isoquinolinone derivatives, such as ISO and DPQ, [3,4-dihydro-5- [4-(piperidin- 1-yl)butoxy] isoquinolin- 1(2H)-one], 143'1443-AB is a weak inhibitor of PARP activity that does not readily cross cell membranes. 145'146Although ISO and D P Q are more potent inhibitors of PARP activity than is 3-AB, both have to be dissolved in

170

Poly(ADP-Ribosyl)ation

dimethylsulphoxide (DMSO, 10% w/v). DMSO itself is a potent scavenger of OH" and can inhibit PARP activity.147,148 Thus, it is not surprising that DMSO itself reduces the organ injury in conditions associated with organ I-R. 149,150 Thus, there is still a great need for the development of potent, water-soluble inhibitors of PARP activity. In 1991, Suto and colleagues described 5-aminoisoquinolin-l(2H)-one (5-AIQ) as a water-soluble and potent inhibitor of PARP-1 activity in a cell-free preparation (enzyme purified 900-fold from calf thymus). 143As 5-AIQ was not commercially available, we optimised the synthetic route (previously described by Wenkert et al 1964, ref. 151) for 5-AIQ and recently reported that 5-AIQ causes a concentration-dependent inhibition of PARP activity (IC50:--0.01 mM) in human cardiac myoblasts. 152In addition, small amounts (0.3.mg/k~l. 52°f5-AIQ abolished the multiple organ m}ury" " caused by severe haemorrh e and resuscltauon. We have also recend reported that 5-AIQ reduces the tissue injury ca~ed by regional myocardial I-R, 153liver I-R 114 and lung I-R. 155

Poly(ADP-Ribose) Polymeraseand Renal Ischemia-ReperfusionInjury As discussed above, the kidney is sensitive to different forms of oxidant-mediated injury. An early study suggested by Schnellmann and colleagues suggested that in PT cell suspensions, PARP activation did not plaYl5~ role of acute renal PT cell injury and death caused by agents which cause oxidative stress. Specifically,suspensions of rabbit PT cells were exposed to antimycin A (a mitochondrial inhibitor), carbonyl cyanide p-(trifluoromethoxy)phenylhydrazone (FCCP, a protonophore) and tert-butyl hydroperoxide (t-BHP, an oxidant). No evidence of DNA fragmentation was observed with any of these agents during cell death. 156 However, in early 1999, we hypothesized that oxidative stress would result in DNA damage to renal cells, produce excessive activation of PARP and thus lead to cellular injury and death. We subsequendy demonstrated that exposure of primary cultures of rat PT cells to oxidative stress in the form of H202 resulted in a significant increase in DNA strand breaks, activation of PARP, increased cellular injury and cell death, and demonstrated a positive correlation between PARP activation and cytotoxicity.65 In a subsequent and similar study using a porcine renal cell-line (LLC-PK1 cells), exposure of cultures to similar concentrations of H202 increased PARP activity and produced decreases in intracellular NAD ÷ and ATP levels.157 Jung and colleagues have investigated and compared the effects of H202 and the organic hydroperoxide, tert-butyl hydroperoxide (t-BHP) on PARP activation and cell death using cultures of an opossum kidney (OK) cell-line. 158 Both H202 and t-BHP were found to reduce cellular ATP levels and produce significant cellular death but only H202 was found to increase PARP activity and it was concluded from these studies that H202 and t-BHP may have different mechanisms for promoting cell death. 158 Subsequently, it has been demonstrated that Na+-dependent phosphate uptake in these cells is inhibited in a dose-dependent manner by H202 and that this effect ofH202 on membrane transport is associated with PARP activation but not with lipid peroxidation, whereas the effect of t-BHP is associated with lipid peroxidation159--an interesting finding which appears to explain why Schnellmann and colleagues did not observe renal PARP activation in the presence of t-BHP (see above and ref. 156). A recent study has also shown that unlike in OK cell cultures, rabbit cortical slices cell death mediated by H202 and t-BHP occurs via a lipid peroxidation-dependent mechanism and is independent of PARP activation. 160 The role of PARP activation in the development of the renal dysfunction and injury caused by I-R of the kidney has also been investigated. In the earliest studies, the role of PARP in renal ischaemic injury (produced by blood depletion) was investigated by Direnfeldt and colleagues in 1987.161 In that study, mouse kidneys were exposed to ischemic insults both in vitro by warm ischemia (37°C, to assess the effects of blood loss at normal body temperature) and in vivo after depletion of blood supply by arterial clamping. Kidneys treated using both methods exhibited decreased levels of PARP activity.161 However, kidneys exposed to cold ischemia in vitro (0°C, to assess the effects of organ storage as utilised for transplantation) exhibited elevated levels of PARP activity.161 Furthermore, DNA isolated from ischaemic kidneys had a stimulatory effect upon exogenous PARP-1 extracted from calf thymus. In the same study,

Poly(ADP-Ribose) Polymeraseand Ischemia-ReperfiusionInjury

171

analysis of DNA extracted from 'cold-storage' kidneys using electron microscopy revealed large (-500 base-pair) single-stranded regions. It was thus suggested that PARP activity was related to the nature of DNA damage resulting from an ischaemic insult. 161 One study last year by Mangino and colleagues investigated if PARP activation contributes to the development of hypothermic I-R injury which occurs in kidneys which are cold stored for transplantation. 162 Specifically, in cortical slices of canine kidney, prolonged cold preservation after reperfusion increased PARP enzyme activity. However, this increase in PARP activity could not be reduced by attenuation of the presence of ROS using catalase, Trolox or DMSO. Furthermore, reperfusion-injury in PTs isolated from canine kidney was increased by inhibitors of PARP activity rather than decreased. Taken together, these findings prompted the conclusion that increased PARP activity during cold storage may provide a protective function rather than an injurious role in hypothermic preservation for transplantation and that ROS are sufficient but not necessary for PARP activation under these conditions. 162 We have recently demonstrated PARP activation in rat models of I-R injury in which the role of ROS and oxidative stress have been investigated using the deferoxamine and the ROS-scavenger Tempol, both of which inhibited PARP activation in the reperfused kidney66'71 suggesting the existence of a PARP-suicide pathway within the kidney subjected to I-R and its significant role in the development of ischaemic ARF.68 Induction of poly(ADP-ribosyl)ation was reported in the kidney after in vivo application of renal carcinogens such as trichloroethene and dichloroacetylene subsequent to DNA double-strand breaks. 163 Potassium bromate and ferric nitrilotriacetate, whose nephrotoxicity is thought to result from ROS formation, both induced poly(ADP-ribosyl)ation with the concomitant formation of DNA double-strand breaks. ~64Recently, PARP activation has been associated with both gentamicin and cisplatin-induced nephrotoxicity. 165-167 Both types of nephrotoxicity involve the generation of ROS 168-171and the beneficial effects of different therapeutic approaches aimed at reducing ROS formation have been evaluated. 165'172-175

Beneficial Effects of Inhlbitors of Poly(ADP-Ribose) Polymerase Activity Various inhibitors of PARP activity have been used in an attempt to elucidate the role(s) of PARP activation in the development and progression of I-R injury and inflammation and to reduce the organ injury associated with PARP activation during these pathophysiological processes. 176-179 Over the last two decades, two major classes of inhibitors of PARP activity have been developed: analogues of benzamide and the isoquinolinones. Inhibitors of both types have been used extensively in in vitro and in vivo models of I-R injury and inflammation. 63 One of the earliest studies investigating the role of inhibitors of PARP activity in renal tissues reported that 3-AB reduced DNA damage and necrosis caused byl,2-dibromo-3-chloropropane in rat kidney and testicular tissues.18° In 1999, we investigated whether the traditional inhibitors of PARP activity, 3-AB, ISO or the endogenous inhibitor of PARP activity, nicotinamide, could reduce oxidative stress in primary cultures of rat PT cells.65We observed that all three inhibitors of PARP activity could provide significant protection against H202-mediated PT cell injury and death (Table 1).65 In contrast, 3-aminobenzoic acid (3-ABA) and nicotinic acid, respective structural analogues of 3-AB and nicotinamide lacking PARP-inhibitory action, did not provide any protection against H202-mediated cellular injury or cell death (Table 1).65 Furthermore, both 3-AB and ISO did not reduce the incidence of DNA strand breaks caused by exposure of PT cells to H202. 65 In the same study, we demonstrated that the cellular injury and cell death was mediated by OH" radicals derived from H202 via the Haber-Weiss and Fenton reactions as the iron-chelating agent deferoxamine also provided significant protection. 65 Incubation of PT cell cultures with catalase, which effectively reduces H202 to H20, also provided significant protection against H202-mediated cellular dysfunction and necrosis and, in contrast to the PARP inhibitors, both deferoxamine and catalase significantly reduced DNA damage caused by H202. 65

172

Poly(ADP-Ribosyl)ation

Table 1. Effect of PARP inhibitors and inactive analogues on PARP activation, cellular injury and ceil death caused by Hz02. The effects of a ROS scavenger (Tempol), deferoxamine and catalase were also assessed

Treatment Untreated H202 only (1 mM) H202 (1 mM) +

3-Aminobenzamide (3 mM) 1,5-Dihydroxyisoquinolinone (0.3 mM) Nicotinamide (3 mM) 5-Aminoisoquinolinone (0.3 mM) 3-Aminobenzoic acid (3 mM) Nicotinic acid (3 mM) Tempol (3 mM) Deferoxamine (3 raM) Catalase (3 U/ml)

PARP Activation (pmol/well/min)

Cellular Injury (% viability)

Cell Death (% cytotoxicity)

10.02 + 1.04* 44.31 + 2.48

100" 20.70 + 1.74

21.07 + 2.45* 61.65 + 3.82

24.86 + 3.07*

46.66 + 2.23*

33.85 + 2.93*

21.99 + 1.95* 25.34 + 2.96* 15.09 + 2.45*

56.52 + 3.61" 45.31 + 2.79* 57.17 + 4.09*

35.69 + 4.02* 43.78 + 2.19" 26.95 + 3.14"

40.67 + 4.04

17.84 + 1.49

56.13 + 4.28

43.36 + 6.87 ND ND ND

19.56 37.35 48.88 90.83

53.75 44.11 39.34 40.87

+ + + +

2.34 2.90* 2.88* 11.50*

+ + + +

4.18 2.82* 2.67* 2.50*

* P 4NQO treatment Squamous cell carcinoma (oral, esophagus) --> IQ treatment Hepatocellular carcinoma BHP treatment

Adenoma (lung)

SCIDParp-1 -/- mice PS3-/-Parp-1 -/- mice PS3-/-Parp- I -/- mice Ku8Ct/-Parp-1 -/- mice PARP-DBD pS3/-mice

Papilloma (forestomach)

-~

Thymic lymphoma Carcinomas (colon & breast) Medulloblastoma Thymic lymphoma Hepatocellular carcinoma T-cell lymphoma

1" 1" 1" ,I, 1" 1"

References

Nozaki et al 20039 Tong et al 20027 Masutani et al unpublished Tsutsumi et al 20008 Tsutsumi et al 20008 Nozaki et al 20039 Nozaki et al 20039 Gunji et al unpublished Ogawa et al unpublished Ogawa et al unpublished Ogawa et al unpublished Morrison et al 199711 Tong et al 200113 Tong et al 20031s Conde et al 200117 Tong et al 20027 Beneke et al 200114

*(-->) no change, (1") elevated, (,I,) reduced incidence, compared to wild-type mice, respectively

frequency of HCC development was increased in 21-23-month-old Parp-1 -/- mice, whereas no tumors were observed in other tissues (Masutani et al unpublished). Administration of N-nitrosobis(2-hydroxypropyl)amine (BHP) resulted in the development of liver hemangioma and hemangiosarcoma at significantly higher frequencies in Parp-1 -/- than in Parp-1 ÷/+ mice. 8 Another alkylating agent, azoxymethane also enhanced the frequency of tumor development both in the colon and liver in Parp-1 -/- compared with that in Parp-1 ÷/+mice. 9 In addition to the differences in tumor incidence, the size of tumors, mainly adenocarcinoma, was larger in Parp-1 -/- than in Parp-1 ÷/+ mice in the colon, suggesting that loss of Parp-1 affects tumor growth. In sharp contrast, the frequency of hepato- and pulmonary carcino/ +/ genesis was not different among Parp-1--, Parp-1 - and Parp-1 ÷/+ mice administered IQ (2-amino-3-methylimidazo[4,5-fquinoline), a cooked food-borne heterocyclic amine that produces bulky adducts on DNA (Ogawa et al unpublished). 4-Nitrosoquinoline 1-oxide (4NQO) mimics ultraviolet (UV)-induced damage and generates a DNA adduct, which is removed mainly by nucleotide excision repair involving XPA (xeroderma pigmentosum group A). It was reported that development of oral tumors in mice given 4NQO was markedly higher in XPA -/- than in wild-type mice. 1° In contrast, there was no difference in tumor incidence between Parp-1 -/- and Parp-1 ÷/÷mice after 4NQO administration in drinking water (Gunji et al unpublished). These experiments strongly suggest that susceptibility to carcinogenesis under Parp-1 deficiency depends substantially on the type of DNA damage and implies the significant contribution of Parp-1 in BER and/or DNA strand break repair pathways to prevent carcinogenesis.

Role of Poly-ADP-Ribosylation in CancerDevelopment

205

The impact of the combination of deficiency of Parp-1 and DNA-PKorp53 was also studied in mice. SCID mice harbor a mutation in the gene encoding a catalytic subunit of the DNA-PK complex and show immunodeficiency due to the lack of V[D]J recombination to produce mature T and B cells. In Parp-1-/-SCID mice, a marked increase of the frequency of T-cell lymphoma was observed from an early age compared to SCID mice, although the frequency of B-cell lymphoma was not increased. 11 P53 is a major genome guardian and is involved in the regulation of both proper cell cycle and apoptosis after DNA damage. Mice lacking p53 (p53 -/-) show a high incidence of spontaneous tumors as well as an increase in various types of genomic instabilities after DNA damage. 12 Tong et al reported that the deficiency of p53 in Parp-1-/- mice, harboring exon 2 disruption of Parp-1, significandy promotes the development of thymic lymphoma, colon and breast carcinomas. 13 A transgenic mouse, which overexpresses the DNA binding domain of Parp-1, as a dominant negative mutant (PARP-DBD), also showed increased incidence of T-cell lymphomagenesis in p53 -/-mice with a significandy shorter tumor latency period. 14 It is possible that the suppressive effect of PARP-DBD on DNA repair promotes the accumulation of genomic instability and contributes to lymphomagenesis. Another intriguing finding is the spontaneous development of medulloblastoma in the cerebellum of 8-week-old p53-/-Parp-1 -/- mice and that nearly half of these mice harbor medulloblastoma by 6 months of age.15 Lee et al reported the spontaneous development of medulloblastoma in the knockout mice harboring both DNA ligase IV and p53 disruption. 16 Since DNA ligase IV is a key enzyme in nonhomologous end joining (NHEJ) repair and accumulating evidence implies that Parp-1 also participates in NHEJ repair, it is possible that the NHEJ-dependent DSB repair is important in prevention of medulloblastoma formation in the cerebellum. Different consequences of p53 and Parp-1 deficiencies were reported by Conde et al in p53-/-Parp-1 -/- mice, harboring exon 4 disruption in Parp-1 gene. These mice show a lower frequency ofthymic lymphoma compared to p53-/- mice. 17In relation to this, H-ras-transformed fibroblasts derived from p53-/-Parp-1-/- mice showed reduction of inducible nitric oxide synthase (iNOS) expression, nitric oxide release and decreased potential of cell growth and tumorigenesis compared to those derived from p53 -/- mice. 17 The diminished cell growth is likely to be related to the decreased level of nitric oxide, since the stimulatory role of nitric oxide in proliferation and its inhibitory role in apoptosis have been reported. 18 We suggest that the role of Parp-1 in cell proliferation significantly affects tumorigenesis under certain conditions Notabl Y, B-cell IYm p homa develo P ment was not increased either in SCIDPa rp-14 11 - or in p53-/-Parp-1 -/- mice13 whereas T-cell lymphoma development was augmented in these cases. Since impairment of S-phase entry of the B-cell population in splenocytes was also / 19 observed in Parp-1--mice, the evidence indicates the possibility that Parp-1 is required for B-cell lymphoma development by supporting its cell proliferation potential. The genetic background of mice may also affect the extent of contribution of Parp-1 in cell proliferation, since we observed that Parp-1-/- mice of the C57BL/6 congenic strain show partial lethality during late embryogenesis 19and reduced body-weight gain (Ogawa et al unpublished), whereas those of ICR/129Sv mixed genetic background do not show such phenotypes. The elucidation of the relationship of other Parp family members to carcinogenesis awaits the results of further experiments including those using the transgenic animal models. Recent studies reported the involvement of Parp-2 in BER processes,20 of Parp-3 in centrosome regulation 21 and of tankyrase in telomere length regulation. 22 Hence, dysfunction of these molecules may cause genomic instability and is expected to have certain impacts on carcinogenesis. Compared to the poly-ADP-ribosylation reaction, the involvement of poly(ADP-ribose) degradation by Parg on carcinogenesis has not been elucidated yet. Parg-deficient embryonic stem (ES) cells)3 and Drosophila24 have become available recently. Since Parg-deficient ES cells show increased sensitivity to alkylating agents and ](-irradiation and undergo early apoptosis (Fujihhara et al unpublished), dysfunction of the Panggene may also be involved in carcinogenesis possibly through regulation of recovery from DNA damage. "

206

Poly(ADP-Ribosyl)ation

Effect of PARP Inhlbitors on Carcinogenesis Several carcinogenesis experiments have been carried out using Parp inhibitors or modulating NAD level, summarized in Table 2. Parp inhibitors including 3-aminobenzamide and 3-methoxybenzamide augment 25-29 or decrease 3°-32 the incidence of tumors depending on the treatment protocol of inhibitors, carcinogens, tissue, or animals used in the studies. 33 The exact explanation for this phenomenon is still not available. Since Parp inhibitors also block activity of Parp family members other than Parp-1, the inhibition of various Parp family proteins should have a substantial influence on the susceptibility to carcinogens. The experimental model of Boyonoski et al manipulated NAD levels in vivo and showed that NAD deficiency increases the incidence of tumors and leukemia, 34 whereas supplementation with the NAD precursor niacin delayed the onset of HCC. 35These results suggest that maintenance of cellular NAD levels prevents or delays carcinogenesis. In vitro transformation systems using cultured cells in combination with various types of carcinogens have been also described. The effects of Parp inhibitors exhibit a wide spectrum as depicted in Table 3. Parp inhibitors suppressed transformation induced by various type of carcinogens, 36 not only methylating agents but also one inducing bulky adducts, benzo[a]pyrene, 36 as well as ionizing irradiation (IR) 37'38'42 and UV. 37'39 In contrast, against ethylating agents, such as ethylnitrosourea and ethylmethanesulfonate, 4°-42 Parp inhibitors enhanced such transformation. The enhancement of in vitro transformation was observed in a time-specific manner. Simultaneous treatment with carcinogen and Parp-inhibitor was effective whereas Parp inhibitor treatment 24 hr after carcinogen exposure was not. 41 These in vitro transformation systems use immortalized cells and the cell proliferation potential may be the major factor that influences the overall frequency of transformation. It is thus conceivable that Parp inhibitors may reduce the proliferation capacity of cells, leading to a low transformation frequency. However, no plausible explanation is currendy available as to why Parp inhibitors exert the opposite effects on transformation induced by methylating and ethylating agents. Parp inhibitors also increase the transformation of NII-I3T3 cells by transfection of SV40 DNA and this was further proven to be due to the increase in integration frequency of SV40 DNA into the genome. 43

Tumorigenesis and Differentiation Functional loss of Parp modulates tumorigenesis and differentiation of malignant cells as summarized in Table 4. Early studies showed that Parp inhibitors induced differentiation of various types of tumor cell lines.44-4y In the case of HL-60 cells48 and H-ras transformed NIH3T3 cells,47 loss of amplified oncogenes, c-myc and H-ras genes, was induced after treatment with Parp inhibitors and these changes could reduce the proliferation potential of tumor cells. Teratocarcinoma EC-AI cells also showed differentiation to endodermal cells during tumorigenesis in nude mice.44 During cell differentiation, an increase in poly-ADP-ribosylation activity generally occurs at the commitment stage, which is followed by decrease in its activity. Reversion of the tumorigenic phenotype was also observed in vivo with a different type of Parp inhibitor, 5-iodo-(>-amino-1, 2-benzopyrone, which is thought to interfere with zinc-finger fimcxion of Parp- 1.49 When Parp-1 -/- mouse ES cells were subcutaneously injected into nude mice, the recipient mice developed teratocarcinoma-like tumors, similar to the case with Parp-1 +/+ ES cells.5° In Parp-1 -/- tumors, the trophoblast lineage cells, including trophoblast giant cells and spongiotrophoblasts, are preferentially induced and large blood lacuna structures were secondarily induced, probably by the action of trophoblasts. Parp-1 -/- ES cells in culture showed elevated expression levels of trophoblast marker genes, 51 suggesting that loss of Parp-1 promotes commitment to a trophoblast lineage. Trophoblast giant cells emerged after a repeated endoreduplication process and showed up to 1,000N ploidy. Loss of Parp-1 function may, therefore, also enhance endoreduplication. Overexpression of PARP-DBD in HeLa cells also interfered with tumorigenesis in nude mice, accompanied by an increased frequency ofapoptosis. 52 These examples indicate that Parp dysfunction can markedly influence tumor phenotype.

Role of Poly-ADP-Ribosylation in CancerDevelopment

207

Table 2. Effect of PARP inhibitors on carcinogenesis Carcinogen

Inhibitor/ Treatment

Species Tissue

Streptozotocin

Nicotinamide

Rat ,1

Nicotinamide

Rat*l

Renalcell tumor Pancreas Insulinoma

3-Aminobenzamide

Rat*2

Pancreas insulinoma

$

Benzamide

Rat ,2

Pancreas Insulinoma

1"

3-Aminobenzamide

Rat*2

Pancreas lnsulinoma

1"

Benzamide

Rat ,2

Pancreas Insulinoma

1"

Liver

1"

AIIoxan

Diethylnitrosamine 3-Aminobenzamide Rat ,2 Nicotinamide Ethylnitrosourea

Rat ,3

Supplementation Rat ,4 with niacin or nicotinamide 3-Aminobenzamide Rat ,8

Tumor

Kidney

Kidney Liver

y-GTP positive foci .7 Renaltubular cell tumor HCC

7~,80c-Dihydroxy-90c,100cepoxy-7,8,9,10tetrahydrobenzopyrene Methylnitrosourea 3-Aminobenzamide Rat "1

Liver

y-GTP positive foci

Liver

1,2-Dimethylhydrazine

3-Aminobenzamide Rat*2

Liver

3-Aminobenzamide Rat ,3

Liver

3-Aminobenzamide Rat ,2

Liver

y-GTP positive foci y-GTP positive foci y-GTP positive foci y-GTP positive foci

Niacin deficiency

Rat ,4

Liver

Diethylnitrosamine 3-Aminobenzamide Rat ,3 + phenobarbital

Liver

Luminol

Liver

N-Nitrosobis(2hydroxypropyi) amine Ethylnitrosourea

Methylazoxymethanol

Rat ,3

Incidence Refs.

3-Aminobenzamide Rat ,3

Colon

3-Aminobenzamide Medaka

Liver

$ ,6

1" $

Rakieten, 19713o Rakieten, 19713o Yamagami, 198525 Yamagami, 198525 Yamagami, 198525 Yamagami, 198525 Takahashi, 198226 Rosenberg, 198527 Boyonoski, 200235 Denda, 198833

_.),9 1"

Denda, 198833 Denda, 198833 Denda, 198833 Denda, 198833

GST-P positive foci .1° GST-P positive foci

Boyonoski, 200234 Tsujiuchi, 199031

GST-P positive foci Adenocarcinoma Hepatoma

Tsujiuchi, 199031 Nakagawa, 198832 Miwa, 198528

Oral squamous cell carcinoma

Miller, 198929

( Oryzias latipes) Dimethylbenz[a] anthracene

3-Methoxybenzamide

"1 Holtzman rat *2 Wistar rat *3 Fischer rat

*4 Long-Evans rat *7 y-Giutamyl transpeptidase "10 Glutathione S-transferase *5 Suppression *8 Wistar & Fischer rats placental form *6 Elevation *9 No change ,

Hamster Cheek pouch

,l, , 5

,,,

208

Poly(ADP-Ribosyl)ation

Table 3. Effect of PARP inhibitors on in vitro transformation

Transformation Frequency Refs.

Carcinogen

Inhibitor

Species

Cells

1,1-Dimethyl hydrazine

Benzamide

Human

Fibroblast

~,1,2

Benzo[a]pyrene

Benzamide

Human

Fibroblast

$

[~-Propiolactone

Benzamide

Human

Fibroblast

,1,

Methylazoxymethanol

Benzamide

Human

Fibroblast

,I,

MNNG

Benzamide

Human

Fibroblast

,I,

3-Hydroxy-1 -propanesulfonic acid y-sulfone Ionizing radiation (IR)

Benzamide

Human

Fibroblast

,I,

Benzamide 3-Aminobenzamide Benzamide 3-Aminobenzamide Benzamide 3-Aminobenzamide 3-Aminobenzamide 3-Aminobenzamide 3-Aminobenzamide 3-Aminobenzamide

Mouse

C3HIOTIII2

,I,

Hamster Mouse

Embryo cells C3H10T/1/2

,!, ,l,

Hamster Mouse

Embryo cells C3H10T/1/2

,1, $

Hamster Mouse

Embryo cells BALB/c3T3A31-1

,1, -4 *3

Mouse

BALB/c3T3A31-1

-4

Mouse

C3H10T1/2

1`*4

Mouse

BALB/c3T3A31-1 C3H10T1/2

1"

Mouse

C3H10T1/2

,l,

Mouse

NIH3T3

1"

UV

MNNG

Methylcholanthrene Aflatoxin B1 Ethylnitrosourea Ethylmethanesulfonate

IR + 12-0-tetradecanoylphorbol-13-acetate SV40 DNA

3-Aminobenzamide 3-Methoxybenzamide

"1 S-phase treatment was most effective *2 Suppression

Kun, 198336 Kun, 198336 Kun, 198336 Kun, 198336 Kun, 198336 Kun, 198336 Borek, 198437,42 Borek, 198437,39 Borek, 198437,42 Lubet, 19844° Lubet, 198440 Borek, 198442 Lubet, 19844°, 198641 Borek, 19863? Strain, 198543

*3 No change *4 Elevation

ParcJ- ES cells produced tumors as in the case of wild-type ES cells (Fujihara et al unpublished). The differentiation potential of ES cells was not different among Parg genotypes. Induction of trophoblast lineage was not observed in Parg-/- tumors, suggesting that Parg deficiency and the resulting impairment of poly(ADP-ribose) degradation is not related to trophoblast induction.

DNA Repair and Genomic Instability As mentioned above, the spectrum of susceptibility ofParp-14- mice to carcinogens implied a significant contribution of Parp-1 in BER and DNA strand break repair. In BER, after removal of damaged bases, such as 8-hydroxy-dG, or alkylated bases by glycosidases, single

Role of Poly-ADP-Ribosylation in Cancer Development

209

Table 4. Consequence of Parp inhibition for tumorigenesis and differentiation o f cancer cells Methodology

Cell or Tissue

Outcome

Gene disruption

Parp-1 -/ mouse ES cells

Induction of trophoblast Nozaki et al, giant cells in teratocarcinoma- 199950 like tumor (in nude mice) Decreased tumorigenesis Conde et al, 200117 Decreased tumorigenesis Hans et al, 199952

H-ras transformed Parp-1 -/ MEFs HeLa cells

Dominant-negative mutant expression Parp inhibitor 3-Aminobenzamide

Mouse teratocarcioma EC-AI cells Mouse Friend erythroleukemic cells Mouse Friend erythroleukemic cells H-ras transformed NIH3T3 cells Human HL-60 cells

Benzamide Nicotinamide Benzamide

Benzamide 4-Hydroxyquinazoline 5-1odo-6-aminoH-ras transformed 1,2-benzopyrone endothelial cells Prostate carcinoma cells

Differentiation into endodermal epitheloid Differentiation into erythrocytes Differentiation into erythrocytes*l Loss of transformed phenotype Differentiation into granulocytes Reversion of tumorigenicity Reversion of tumorigenicity

References

Ohashi et al, 198444 Terada et al, 197945 Brac et al, 198746 Nakayasu et al, 198847 Shima et al, 198948 Bauer et al, 199549 Bauer et al, 199549

,1 Differentiation is inhibited depending on the concentration of the inhibitor. .

.

.

.

.

.

.

.

.

strand scission is introduced by AP-endonuclease. The recruitment of a molecular scaffold XRCC-1 (X-ray repair cross-complementing factor-I) to the repair site is a critical step in BER because XRCC1 further recruits DNA ligase IIIcz, DNA polymerase I], and polynucleotide kinase. 53 Association of certain polymorphisms in XRCC1 gene with lung and other cancers was reported by Divine et al.54 It was demonstrated that Parp- 1 is necessary for the assembly or stability of XRCC-1 nuclear foci at the site of DNA damage. 55 Dantzer et al showed that Parp-1 acts in the strand displacement step of the DNA fill-in reaction by DNA polymerase 13 and FEN-1 (flap endonuclease-1) and that long-patch BER is substantially delayed, whereas short-patch repair is only slighdy affected in extracts from Parp-1 -/- MEE56 Recendy, defective poly-ADP-ribosylation in cells from Werner syndrome (WS) was reported following DNA damages introduced by an oxidative or an alkylating agent. 57 WS protein (WRN) interacts with proteins acting in BER, including polymerase/5, PCNA, FEN-I, replication factor A as well as Parp-l.58 WRN may facilitate Parp-1 activation in the BER process. WS is characterized by the early onset of cancer, which may be pardy explained by the defective poly-ADP-ribosylation activity in the BER process. Malanga et al also reported the repair of stalled DNA topoisomerase I (topoI)-DNA covalent complex through reactivation of topoi by Parp-1. 59 In addition, Parp-2 was also shown to be involved in BER by Schreiber et al. 2° Parp-1 is also activated by double-strand break (DSB) and participates in NHEJ catalyzed by DNA-PK complex. Parp-1 activates auto-phosphorylation activity of DNA-PK and Ku70/ l 80 complex, 60 whereas Parp-1 act'vity is suppressed b y DNA-PK. 61 In SCIDParp-1 -/T-lymphocytes, V[D]J recombination, which is carried out by NHEJ, is partially restored, ll

210

Poly(ADP-Ribosyl)ation

Furthermore, in a recombination-inducible SCID cell line, poly(ADP-ribose) formation was shown to occur during the resolution stage of V[D]J recombination where nascent opened coding ends are generated. Poly(ADP-ribose) formation colocalized with foci positive for the recombination protein Mrel I and facilitated coding-end resolution. In contrast, this response was not observed in wild-type cells possessing a functional catalytic subunit of DNA-dependent protein kinase. 62 WRN protein physically interacts also with Parp-1,63 and Ku70/80-induced stimulation ofWRN exonucleolytic activity was interfered with poly-ADP-ribosylation of Ku70/ 80 by Parp-1. 64 Parp-1 may thus regulate the exonudeolytic activity of WRN and prevent accidental recombination reaction during NHEJ. Treatment of Parp-1 -/- mice with BHP, an alkylating agent, did not enhance the frecjuency of point mutation, but rather increased the deletion frequency compared with Parp-1 +'+ mice (Shibata et al unpublished). This finding supports the current evidence that Parp-1 is involved in the NHEJ process. Lack of elevation in point mutation under Parp-1 deficiency also suggests that Parp-1 is probably not required in BER, at least until the removal of the damaged base, but may function after DNA strand break introduction by preventing further conversion of single strand breaks into DSBs, which will be predominantly repaired by NHEJ. Thus, Parp-1 is possibly involved in the repair of DSBs that occur during the process of the BER reaction.

Chromosome Instability and Cell-Cycle Checkpoints Controls Hallmarks of chromosome instability in cancer cells include aneuploidy, hyperploidy, gene amplification, loss ofheterozygosiw/ (LOH) and gene rearrangement. Tong et al7 reported that Ku80 haploinsufflciency in Parp-1-- mice increased the incidence of HCC and the presence of chromosome instability in those tumors, such as chromatid/chromosome breaks, end-to-end fusions and recurrent nonreciprocal translocations. Hyperploidy was observed in spontaneously immortalized Parp-14- MEFs.65-67 Comparative genomic hybridization analysis revealed that chromosome gain and loss were enhanced in Parp-1 -/- compared with Parp-1 ÷/÷MEFs. 65 Parp-1 -/- cells showed increased sister-chromatid exchange frequencies (SCEs) and micronuclei formation. 6,68 One of the possible mechanisms for these gross chromosome instabilities is deregulation of cell-cycle checkpoints. Parp-1 direcdy interacts with p5369 and was shown to be involved in p53-mediated G1 arrest after DNA damage. 7°-72 Parp-1 was further found to complex with PCNA and p21 after DNA damage introduced by N-methyl-N'nitro-N-nitrosoguanidine (MNNG) treatment. 73 Moreover, Kanai et al demonstrated that Parp-1 is located in the centrosome and interacts with p53 to regulate centrosome replication and function. 74 Halappanavar et al showed a defective mitotic checkpoint arrest accompanying down-regulation ofcyclinB 1/cdk-1 kinase activity in Parp-1 -/- MEFs. 75They also found that Parp-1 -/- MEFs with higher ploidy were resistant to apoptosis in the G 1 phase compared to wild-type cells, indicating defective post-mitotic checkpoints under Parp-1 deficiency. Despite the accumulated evidence, it remains to be clarified whether Parp-1 -/- tumors show higher chromosome instability and a more malignant phenotype compared to Parp-1 +/÷tumors. Another important issue is to understand whether Parp-1 or poly-ADP-ribosylation is involved in the formation of LOH during tumorigenesis.

Epigenetic Instability and Control of Gene Expression Epigenetic changes in the gene are defined as nongenetic changes inheritable to daughter cells during cell growth. It has been demonstrated that epigenetic alteration of gene expression by hypo- or hyper-methylation substantially contributes to carcinogenesis. 76 Previous studies reported global hypomethylation in cancer cells and that genome hypomethylation induced by DNA methyltransferase mutation increases tumor incidence in mice. 77 In contrast, local hypermethylation of CpG islands in promoter regions in various tumor suppressor genes, including p16 tN¢4, was observed frequently in cancer cells.76 Using the PARP inhibitor

Role of Poly-ADP-Ribosylation in CancerDevelopment

211

3-aminobenzamide, Zardo et a178reported the presence of a genome-wide negative correlation between DNA methylation and poly-ADP-ribosylation. The promoter region of the H0O gene also displayed this negative correlation. 79 Further studies are needed to investigate whether Parp-1 dysfunction leads to hypermethylation of cancer-related genes, and whether it promotes carcinogenesis. On the other hand, poly-ADP-ribosylation can be involved in the control of gene expression independent ofDNA methylation. In Drosophila, engineered Parp-deficient flies displayed attenuation of the expression of genes located in puff loci, accompanied by the lack of puff formation as well as marked decrease of the induction of immune related genes such as Diptericine. 8° Gene expression was also reported to be altered under Parp-1 deficiency,81 including those ofiNOS, 82 and histone acetyltransferase. 83A function ofParp- 1 as a coactivator could be involved in these phenomena. These possible functions of Parp-1 in the regulation of gene expression may lead to the alteration of differentiation potential and may ultimately affect tumor phenotypes.

Cancer Cell Selection through Cell Death During cancer development, cancer cells may encounter various forms of cell-death pressure, depending on the site of their growth and surrounding microenvironment. During rapid proliferation of cancer cells in limited tissue space, hypoxic/anoxic conditions may prevail, which in turn enhance p53-dependent apoptosis. Such conditions may preferentially select p53-deficient cancer cells and cells overexpressing bcl-2, an apoptosis inhibitory protein, as described by Graeber et al. 84 In this regard, p53 gene alteration is detected in more than one-half of human tumors and bcl-2 overexpression is also frequently observed in B-cell lymphoma, prostate cancer and colorectal cancer in humans. On the other hand, oxidative cell death may be also induced in inflammatory conditions during carcinogenesis. Reactive oxygen species and reactive nitrogen species, including nitric oxide, produced by macrophages, induce rapid activation of Parp-1, leading to NAD depletion and apoptosis-inducing factor (AIF)-dependent cell death. Yu et al demonstrated that Parp-1 is necessary for this process. 85 It may thus be speculated that Parp-l-deficient cells may be selected out under such oxidative stress conditions (Fig. 1). Previous studies showed that neuronal cell death 86 and streptozotocin-induced pancreatic ~-cell death were inhibited by either the Parp inhibitor, 3-aminobenzamide, or Parp-I deficiency. '87 '88 The experiments by Yamagami et al, in which the development of insulinoma in rats treated with streptozotocin was markedly enhanced by treatment with the Parp inhibitor 3-aminobenzamide,25adds further support for this scenario. -

4

Role of PARP in Human Carcinogenesis Molecular and biochemical studies as well as animal model studies suggest that PARP is involved in carcinogenesis, although the relation of the functional loss of PARP to human carcinogenesis is largely undetermined yet. Several pioneering studies investigated the changes in PARP-1 gene expression and gene structure in human cancers. In a series of studies, Bhatia et al89-91 demonstrated that a PARP-1 pseudogene on chromosome 13q33-qter presents a two allele (A/B) polymorphism and that the frequency of the B allele is higher in African Americans and is associated with endemic Burkitt lymphomas (1.7-fold), multiple myeloma and prostate cancers in the African American population. Enhanced activity and expression of PARP-1 in Ewing's sarcoma cell lines were reported by Prasad et al.92 The same group later reported that enhancement of PARP-1 gene expression is due to activation of transcription factors ets-1 in Ewing's sarcoma cells. 93 Bieche et a194 showed that weak expression of the PARP-1 gene is associated with higher genomic instability in breast cancer. They also showed that chromosome lq41-42, where the PARP-1 gene is located, is frequently amplified in cancers that overexpress PARP-1. Other studies reported low formation of poly(ADP-ribose) induced by bleomycin treatment in peripheral lymphocytes from laryngeal cancer patients, 95 suggesting

212

Poly(ADP-Ribosyl)ation

Appearance of tumor - - - < . . _ j -

Hypoxic/anoxic condition Apoptosis tll

Oxidative condition

AIF-dependent cell death (Cell death accompanying NAD depletion)

¢

.o.

.002ocro0oge

Survival of: pS3 m u t a n t Clonal expansion

cells

~

b¢1-2 overexpressing cells

Parp-l-deficient

cells

~

Tumor growth Figure 1. A possible model for selection of Parp-l-deficient cells during carcinogenesis. Hypoxic or anoxic conditions often prevail in tumors, which may lead to thepreferential selection ofp53-deficient cancer cells and cells overexpressingthe anti-apoptotic protein bcl-2.~4We speculate that, in contrast, Parp-l-deficient cells may be selected under oxidative or nitrosative stress conditions, which also may prevail during cancer formation. For details, see text. that low PARP activity correlates with a higher risk of laryngeal cancer. We found that the gastric cancer cell line MKN28 harbored a structural alteration in PARP-1 gene, 96 although it is not known yet whether this affects the function of PARP-1. Recent biochemical studies suggest that dysfunctions of PARP-2 and PARP-3 are also closely related to carcinogenesis. In this regard, Augustin et al21 indicated that the PARP-3 gene is located at chromosome 3p21.1-3p21.31, where LOH is frequently observed in the early stages of lung cancer. Extensive investigation of genetic alterations of these PARP family genes may facilitate our understanding of the role of PARP in human carcinogenesis.

Concluding Remarks Carcinogenesis in humans generally increases dramatically with age and it is considered that five or more genetic or epigenetic events may be necessary for development of cancer. 1,2 Each event leads to evolution of a certain tumor cell population from the selective pressure given from the microenvironment. Several lines of evidence obtained from research over the years imply that Parp-1 is involved in epithelial carcinogenesis and lymphomagenesis as a tumor suppressor factor. On the other hand, Parp-1 seems to be required for carcinogenesis and lymphomagenesis through its function in promotion of cell proliferation and inflammatory responses. Moolgavkar and Luebeck proposed that intervention strategies aimed at reducing the rate of clonal expansion of initiated/premalignant cells should be more effective than those designed to decrease the rate of early mutational events in the multistage process of human carcinogenesis. 97 In this context, a better understanding of the roles of poly-ADP-ribosylation

Role of Poly-ADP-Ribosylation in CancerDevelopment

213

in transcription, cell-cycle-control, cell proliferation and modulation of immune responses is also important, especially for cancer development at an advanced age. It was reported that haploinsufflciency of caretaker genes, such as histone H 2 A 98,99 and NBS 1°° genes significandy enhances the susceptibility to carcinogenesis. Since Parp-1 functions as both a caretaker and gate-keeper of the genome, haploinsufficiency of Parp-1 may also enhance carcinogenesis. In this context, susceptibility of Parp-1 +/- animals should be further investigated over extended time periods after application of various stimuli. The combination of haploinsufficiency in either caretaker genes or gate-keeper genes may further enhance the carcinogenic process during the long lifespan of human beings. The identification of various Parp family members as well as Parg evoked intriguing questions on their functions, including whether these proteins are related to carcinogenesis, and further studies using animal models should clarify the impact of their deficiency on susceptibility to carcinogenesis.

References 1. Armitage P, Doll R. The age distribution of cancer and a multi-stage theory of carcinogenesis. Br J Cancer 1954; 8(1):1-12. 2. Sugimura T. Multistep carcinogenesis: A 1992 perspective. Science 1992; 258(5082):603-607. 3. Masutani M, Nakagama H, Sugimura T. Poly(ADP-ribose) and carcinogenesis. Genes Chromosomes Cancer 2003; 38(4):339-348. 4. Masutani M, Suzuki H, Kamada N e t al. Poly(ADP-ribose) polymerase gene disruption conferred mice resistant to streptozotocin-induced diabetes. Proc Natl Acad Sci USA 1999; 96(5):2301-2304. 5. Wang ZQ, Auer B, Stingl Let al. Mice lacking ADPRT and poly(ADP-ribosyl)ation develop normally but are susceptible to skin disease. Genes Dev 1995; 9(5):509-520. 6. de Murcia JM, Niedergang C, Trucco C et al. Requirement of poly(ADP-ribose) polymerase in recovery from DNA damage in mice and in cells. Proc Nad Acad Sci USA 1997; 94(14):7303-7307. 7. Tong WM, Cortes U, Hande MP et al. Synergistic role of Ku80 and poly(ADP-ribose) polymerase in suppressing chromosomal aberrations and liver cancer formation. Cancer Res 2002; 62(23):6990-6996. 8. Tsutsumi M, Masutani M, Nozaki T et al. Increased susceptibility of poly(ADP-ribose) polymerase-1 knockout mice to nitrosamine carcinogenicity. Carcinogenesis 2001; 22(1):1-3. 9. Nozaki T, Fujihara H, Watanabe M et al. Parp-1 deficiency implicated in colon and liver tumorigenesis induced by azoxymethane. Cancer Sci 2003; 94(6):497-500. 10. Ide F, Oda H, Nakatsuru Yet al. Xeroderma pigmentosum group A gene action as a protection factor against 4-nitroquinoline 1-oxide-induced tongue carcinogenesis. Carcinogenesis 2001; 22(4):567-572. 11. Morrison C, Smith GC, Stingl L et al. Genetic interaction between PARP and DNA-PK in V(D)J recombination and tumorigenesis. Nat Genet 1997; 17(4):479-482. 12. Donehower LA, Harvey M, Slagle BL et al. Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumours. Nature 1992; 356(6366):215-221. 13. Tong WM, Hande MP, Lansdorp PM et al. DNA strand break-sensing molecule poly(ADP-ribose) polymerase cooperates with p53 in telomere function, chromosome stability, and tumor suppression. Mol Cell Biol 2001; 21(12):4046-4054. 14. Beneke R, Moroy T. Inhibition of poly(ADP-ribose) polymerase activity accelerates T-cell lymphomagenesis in p53 deficient mice. Oncogene 2001; 20(56):8136-8141. 15. Tong WM, Ohgaki H, Huang H et al. Null mutation of DNA strand break-binding molecule poly(ADP-ribose) polymerase causes medulloblastomas in p53(-/-) Mice. Am J Pathol 2003; 162(1):343-352. 16. Lee Y, McKinnon PJ. DNA ligase IV suppresses medulloblastoma formation. Cancer Res 2002; 62(22):6395-6399. 17. Conde C, Mark M, Oliver FJ et al. Loss of poly(ADP-ribose) polymerase-1 causes increased tumour latency in p53-deficient mice. EMBO J 2001; 20(13):3535-3543. 18. Kim PK, Zamora R, Petrosko P et al. The regulatory role of nitric oxide in apoptosis. Int Immunopharmacol 2001; 1(8):1421-1441. 19. Watanabe F, Masutani M, Kamada N e t al. Impairment in S-phase entry of splenocytes of Parp-1 knockout mice. Proc Japan Acad 2003; 79 Ser B(8):248-251. 20. Schreiber V, Ame JC, Dolle P e t al. Poly(ADP-ribose) polymerase-2 (PARP-2) is required for efficient base excision DNA repair in association with PARP-1 and XRCC1. J Biol Chem 2002; 277(25):23028-23036. 21. Augustin A, Spenlehauer C, Dumond H et al. PARP-3 localizes preferentially to the daughter centriole and interferes with the G1/S cell cycle progression. J Cell Sci 2003; l l6(Pt 8):1551-1562.

214

Poly(ADP-Ribosyl)ation

22. Smith S, Giriat I, Schmitt A et al. Tankyrase, a poly(ADP-ribose) polymerase at human telomeres. Science 1998; 282(5393):1484-1487. 23. Gunji A, Fujihara H, Kamada N et al. Lack of altered frequency of sister-chromatid exchanges in poly(ADP-ribose) glycohydrolase-deficient mouse ES cells treated with methylmethanesulfonate. Proc Japan Acad 2003; 79 Ser B(10):305-307. 24. Hanai S, Kanai M, Ohashi S et al. Loss of poly(ADP-ribose) glycohydrolase causes progressive neurodegeneration in Drosophila melanogaster. Proc Nad Acad Sci USA 2004; 101(1):82-86. 25. Yamagami T, Miwa A, Takasawa S e t al. Induction of rat pancreatic B-cell tumors by the combined administration of streptozotocin or aUoxan and poly(adenosine diphosphate ribose) synthetase inhibitors. Cancer Res 1985; 45(4):1845-1849. 26. Takahashi S, Ohnishi T, Denda A et al. Enhancing effect of 3-aminobenzamide on induction of gamma-glutamyl transpeptidase positive foci in rat liver. Chem Biol Interact 1982; 39(3):363-368. 27. Rosenberg MR, Novicki DL, Jirde RL et al. Promoting effect of nicotinamide on the development of renal tubular cell tumors in rats initiated with diethylnitrosamine. Cancer Res 1985; 45(2):809-814. 28. Miwa M, Ishikawa T, Kondo T et al. Enhancement by 3-aminobenzamide of methylazoxymethanol acetate-induced hepatoma of the small fish "Medaka" (Oryzias latipes)'. In: Althaus FR, Hilz H, Shall S, eds. ADP-Ribosylation of Proteins. Berlin-Heidelberg-New York-Tokyo: Springer-Verag0 1985:480-483. 29. Miller EG, Rivera-Hidalgo F, Binnie WH. 3-Methoxybenzamide, a possible initiator for DMBA-induced carcinogenesis. In: Jacobson MK, Jacobson EL, eds. ADP-Ribose Transfer Reactions. Mechanisms and Biological Significance. New York: Springer-Verlag,1989:287-290. 30. Rakieten N, Gordon BS, Beaty A et al. Pancreatic islet cell tumors produced by the combined action of streptozotocin and nicotinamide. Proc Soc Exp Biol Med 1971; 137(1):280-283. 31. Tsujiuchi T, Tsutsumi M, Denda A et al. Possible involvement of poly ADP-ribosylation in phenobarbital promotion of rat hepatocarcinogenesis. Carcinogenesis 1990; 11 (10): 1783-1787. 32. Nakagawa K, Utsunomiya J, Ishikawa T. Inhibition of methylazoxymethanol acetate initiation of colon carcinogenesis in rats by treatment with the poly(ADP-ribose)polymerase inhibitor 3-aminobenzamide. Carcinogenesis 1988; 9(7):1167-1171. 33. Denda A, Tsutsumi M, Yokose Y e t al. Effects of 3-aminobenzamide on the induction of gamma-glutamyl-transpeptidase-positive foci by various chemicals in rat liver. Cancer Lett 1988; 39(1):29-36. 34. Boyonoski AC, Spronck JC, Gallacher LM et al. Niacin deficiency decreases bone marrow poly(ADP-ribose) and the latency of ethylnitrosourea-induced carcinogenesis in rats. J Nutr 2002; 132(1):108-114. 35. Boyonoski AC, Spronck JC, Jacobs RM et al. Pharmacological intakes of niacin increase bone marrow poly(ADP-ribose) and the latency of ethylnitrosourea-induced carcinogenesis in rats. J Nutr 2002; 132(1):115-120. 36. Kun E, Kirsten E, Milo GE et al. Cell cycle-dependent intervention by benzamide of carcinogen-induced neoplastic transformation and in vitro poly(ADP-ribosyl)ation of nuclear proteins in human fibroblasts. Proc Natl Acad Sci USA 1983; 80(23):7219-7223. 37. Borek C, Ong A, Morgan WF et al. Inhibition of X-ray- and ultraviolet light-induced transformation in vitro by modifiers of poly(ADP-ribose) synthesis. Radiat Res 1984; 99(2):219-227. 38. Borek C, Cleaver JE. Antagonistic action of a tumor promoter and a poly(adenosine diphosphoribose) synthesis inhibitor in radiation-induced transformation in vitro. Biochem Biophys Res Commun 1986; 134(3):1334-1341. 39. Borek C, Morgan WF, Ong A et al. Inhibition of malignant transformation in vitro by inhibitors of poly(ADP-ribose) synthesis. Proc Natl Acad Sci USA 1984; 81(1):243-247. 40. Lubet RA, McCarvill JT, Putman DL et al. Effect of 3-aminobenzamide on the induction of toxicity and transformation by ethyl methanesulfonate and methylcholanthrene in BALB/3T3 cells. Carcinogenesis 1984; 5(4):459-462. 41. Lubet RA, McCarvill JT, Schwartz JL et al. Effects of 3-aminobenzamide on the induction of morphologic transformation by diverse compounds in Balb/3T3 cells in vitro. Carcinogenesis 1986; 7(1):71-75. 42. Borek C, Ong A, Cleaver JE. Methylating and ethylating carcinogens have different requirements for poly(ADP-ribose) synthesis during malignant transformation. Carcinogenesis 1984; 5(12):1573-1576. 43. Strain AJ. Inhibitors of ADP-ribosyl transferase enhance the transformation of NIH3T3 cells following transfection with SV40 DNA. Exp Cell Res 1985; 159(2):531-535. 44. Ohashi Y, Ueda K, Hayaishi O et al. Induction of murine teratocarcinoma cell differentiation by suppression of poly(ADP-ribose) synthesis. Proc Natl Acad Sci USA 1984; 81(22):7132-7136.

Role of Poly-ADP-Ribosylation in Cancer Development

215

45. Terada M, Fujiki H, Marks PA et al. Induction of erythroid differentiation of murine erythroleukemia cells by nicotinamide and related compounds. Proc Natl Acad Sci USA 1979; 76(12):6411-6414. 46. Brac T, Ebisuzaki K. Inhibitors of poly(ADP-ribose) polymerase prevent Friend cell differentiation. In: Althaus FR, Hilz H, Shall S, eds. ADP-Ribosylation of Proteins. Berlin-Heidelberg-New York-Tokyo: Springer, 1985:446-452. 47. Nakayasu M, Shima H, Aonuma S e t al. Deletion of transfected oncogenes from NIH 3T3 transformants by inhibitors of poly(ADP-ribose) polymerase. Proc Natl Acad Sci USA 1988; 85(23):9066-9070. 48. Shima H, Nakayasu M, Aonuma Set al. Loss of the MYC gene amplified in human HL-60 cells after treatment with inhibitors of poly(ADP-ribose) polymerase or with dimethyl sulfoxide. Proc Nad Acad Sci USA 1989; 86(19):7442-7445. 49. Bauer PI, Kirsten E, Varadi G et al. Reversion of malignant phenotype by 5-iodo-6amino-1,2-benzopyrone a noncovalendy binding ligand of poly(ADP-ribose) polymerase. Biochimie 1995; 77(5):374-377. 50. Nozaki T, Masutani M, Watanabe M et al. Syncytiotrophoblastic giant cells in teratocarcinoma-like tumors derived from Parp-disrupted mouse embryonic stem cells. Proc Natl Acad Sci USA 1999; 96(23): 13345-13350. 51. Hemberger M, Nozaki T, Winterhager E et al. Parp 1-deficiency induces differentiation of ES cells into trophoblast derivatives. Dev Biol 2003; 257(2):371-381. 52. Hans MA, Muller M, Meyer-Ficca M et al. Overexpression of dominant negative PARP interferes with tumor formation of HeLa cells in nude mice: Evidence for increased tumor cell apoptosis in vivo. Oncogene 1999; 18(50):7010-7015. 53. Caldecott KW. Protein-protein interactions during mammalian DNA single-strand break repair. Biochem Soc Trans 2003; 31(Pt 1):247-251. 54. Divine KK, Gilliland FD, Crowell RE et al. The XRCC1 399 glutamine allele is a risk factor for adenocarcinoma of the lung. Mutat Res 2001; 461(4):273-278. 55. EI-Khamisy SF, Masutani M, Suzuki H et al. A requirement for PARP-1 for the assembly or stability of XRCC1 nuclear foci at sites of oxidative DNA damage. Nucleic Acids Res 2003; 31(19):5526-5533. 56. Dantzer F, Schreiber V, Niedergang C et al. Involvement of poly(ADP-ribose) polymerase in base excision repair. Biochimie 1999; 81(1-2):69-75. 57. von Kobbe C, Harrigan JA, May A et al. Central role for the Werner syndrome protein/ poly(ADP-ribose) polymerase 1 complex in the poly(ADP-ribosyl)ation pathway after DNA damage. Mol Cell Biol 2003; 23(23):8601-8613. 58. Opresko PL, Cheng WH, von Kobbe C et al. Werner syndrome and the function of the Werner protein; what they can teach us about the molecular aging process. Carcinogenesis 2003; 24(5):791-802. 59. Malagna M, Althaus FR. Poly(ADP-ribose) reactivates stalled DNA topoisomerase I and induces DNA strand break resealing. J Biol Chem 2004; 279(7):5244-5248. 60. Ruscetti T, Lehnert BE, Halbrook J et al. Stimulation of the DNA-dependent protein kinase by poly(ADP-ribose) polymerase. J Biol Chem 1998; 273(23): 14461-14467. 61. Ariumi Y, Masutani M, Copeland TD et al. Suppression of the poly(ADP-ribose) polymerase activity by DNA-dependent protein kinase in vitro. Oncogene 1999; 18(32):4616-4625. 62. Brown ML, Franco D, Burkle A et al. Role of poly(ADP-ribosyl)ation in DNA-PKcs-independent V(D)J recombination. Proc Natl Acad Sci USA 2002; 99(7):4532-4537. 63. Adelfalk C, Kontou M, Hirsch-Kauffmann M et al. Physical and functional interaction of the Werner syndrome protein with poly-ADP ribosyl transferase. FEBS Lett 2003; 554(1-2):55-58. 64. Li B, Nacarro S, Kasahara N et al. Identification and biochemical characterization of a Werner syndrome protein complex with Ku70/80 and PARP-1. J Biol Chem 2004; 279(14):13659-13667. 65. Simbulan-Rosenthal CM, Haddad BR, Rosenthal DS et al. Chromosomal aberrations in PARP(-/-) mice: Genome stabilization in immortalized cells by reintroduction of poly(ADP-ribose) polymerase cDNA. Proc Natl Acad Sci USA 1999; 96(23):13191-13196. 66. Nozaki T, Fujihara H, Kamada N et al. Hyperploidy of embryonic fibroblasts derived from Parp-1 knockout mouse. Proc Japan Acad 2001; 77 Ser B(6):121-124. 67. Simbulan-Rosenthal CM, Rosenthal DS, Luo R et al. Inhibition of poly(ADP-ribose) polymerase activity is insufficient to induce tetraploidy. Nucleic Acids Res 2001; 29(3):841-849. 68. Wang ZQ, Stingl L, Morrison C et al. PARP is important for genomic stability but dispensable in apoptosis. Genes Dev 1997; 11(18):2347-2358. 69. Vaziri H, West MD, Allsopp RC et al. ATM-dependent telomere loss in aging human diploid fibroblasts and DNA damage lead to the post-translational activation of p53 protein involving poly(ADP-ribose) polymerase. EMBO J 1997; 16(19):6018-6033.

216

Poly(ADP-Ribosyl)ation

70. Nozaki T, Masutani M, Akagawa T et al. Suppression of G 1 arrest and enhancement of G2 arrest by inhibitors of poly(ADP-ribose) polymerase: Possible involvement of poly(ADP-ribosyl)ation in cell cycle arrest following gamma-irradiation. Jpn j Cancer Res 1994; 85(11):1094-1098. 71. Wieler S, Gagne JP, Vaziri H et al. Poly(ADP-ribose) polymerase-1 is a positive regulator of the p53-mediated G1 arrest response following ionizing radiation. J Biol Chem 2003; 278(21 ): 18914-18921. 72. Agarwal ML, Agarwal A, Taylor WR et al. Defective induction but normal activation and function of p53 in mouse cells lacking poly-ADP-ribose polymerase. Oncogene 1997; 15(9):1035-1041. 73. Frouin I, Maga G, Denegri M e t al. Human proliferating cell nuclear antigen, poly(ADP-ribose) polymerase-1, and p21wafl/cipl. A dynamic exchange of partners. J Biol Chem 2003; 278(41):39265-39268. 74. Kanai M, Tong WM, Sugihara E et al. Involvement of poly(ADP-Ribose) polymerase 1 and poly(ADP-Ribosyl)ation in regulation of centrosome function. Mol Cell Biol 2003; 23(7):2451-2462. 75. Halappanavar SS, Shah GM. Defective control of mitotic and post-mitotic checkpoints in poly(ADP-ribose) polymerase-l(-/-)fibroblasts after mitotic spindle disruption. Cell cycle 2004; 3(3):335-342. 76. Sugimura T, Ushijima T. Genetic and epigenetic alterations in carcinogenesis. Mutat Res 2000; 462(2-3):235-246. 77. Gaudet F, Hodgson JG, Eden A et al. Induction of tumors in mice by genomic hypomethylation. Science 2003; 300(5618):489-492. 78. Zardo G, D'Erme M, Reale A et al. Does poly(ADP-ribosyl)ation regulate the DNA methylation pattern? Biochemistry 1997; 36(26):7937-7943. 79. de Capoa A, Febbo FR, Giovannelli F et al. Reduced levels of poly(ADP-ribosyl)ation result in chromatin compaction and hypermethylation as shown by cell-by-cell computer-assisted quantitative analysis. FASEB J 1999; 13(1):89-93. 80. Tulin A, Spradling A. Chromatin loosening by poly(ADP)-ribose polymerase (PARP) at Drosophila puff loci. Science 2003; 299(5606):560-562. 81. Simbulan-Rosenthal CM, Ly DH, Rosenthal DS et al. Misregulation of gene expression in primary fibroblasts lacking poly(ADP-ribose) polymerase. Proc Natl Acad Sci USA 2000; 97(21):11274-11279. 82. Hassa PO, Hottiger MO. A role of poly(ADP-ribose) polymerase in NF-kappaB transcriptional activation. Biol Chem 1999; 380(7-8):953-959. 83. Ota K, Kameoka M, Tanaka Yet al. Expression of histone acetyltransferases was down-regulated in poly(ADP-ribose) polymerase-l-deficient murine cells. Biochem Biophys Res Commun 2003; 310(2):312-317. 84. Graeber TG, Osmanian C, Jacks T et al. Hypoxia-mediated selection of cells with diminished apoptotic potential in solid tumours. Nature 1996; 379(6560):88-91. 85. Yu SW, Wang H, Poitras MF et al. Mediation of poly(ADP-ribose) polymerase-l-dependent cell death by apoptosis-inducing factor. Science 2002; 297(5579):259-263. 86. Eliasson MJ, Sampei K, Mandir AS et al. Poly(ADP-ribose) polymerase gene disruption renders mice resistant to cerebral ischemia. Nat Med 1997; 3(10): 1089-1095. 87. Burkart V, Wang ZQ, Radons J et al. Mice lacking the poly(ADP-ribose) polymerase gene are resistant to pancreatic beta-cell destruction and diabetes development induced by streptozotocin. Nat Med 1999; 5(3):314-319. 88. Pieper AA, Brat DJ, Krug DK et al. Poly(ADP-ribose) polymerase-deficient mice are protected from streptozotocin-induced diabetes. Proc Natl Acad Sci USA 1999; 96(6):3059-3064. 89. Bhatia KG, Cherney BW, Huppi K et al. A deletion linked to a poly(ADP-ribose) polymerase gene on chromosome 13q33-qter occurs frequently in the normal black population as well as in multiple tumor DNA. Cancer Res 1990; 50(17):5406-5413. 90. Bhatia K, Huppi K, Cherney Bet al. Relative predispositional effect of a PADPRP marker allele in B-ceU and some non B-cell malignancies. Curr Top Microbiol Immunol 1990; 166:347-357. 91. Lyn D, Cherney BW, Lalande M e t al. A duplicated region is responsible for the poly(ADP-ribose) polymerase polymorphism, on chromosome 13, associated with a predisposition to cancer. Am J Hum Genet 1993; 52(1):124-134. 92. Prasad SC, Thraves PJ, Bhatia KG et al. Enhanced poly(adenosine diphosphate ribose) polymerase activity and gene expression in Ewing's sarcoma cells. Cancer Res 1990; 50(1):38-43. 93. Soldatenkov VA, Albor A, Patel BK et al. Regulation of the human poly(ADP-ribose) polymerase promoter by the ETS transcription factor. Oncogene 1999; 18(27):3954-3962. 94. Bieche I, de Murcia G, Lidereau R. Poly(ADP-ribose) polymerase gene expression status and genomic instability in human breast cancer. Clin Cancer Res 1996; 2(7):1163-1167.

Role of Poly-ADP-Ribosylation in Cancer Development

217

95. Rajaee-Behbahani N, Schmezer P, Ramroth H et al. Reduced poly(ADP-ribosyl)ation in lymphocytes of laryngeal cancer patients: Results of a case-control study. Int J Cancer 2002; 98(5):780-784. 96. Masutani M, Nozaki T, Sasaki H et al. Aberration of poly(ADP-ribose) polymerase-1 gene in human tumor cell lines: Its expression and structural alterations. Proc Japan Acad 2004; 80 Ser B(2):114-118. 97. Moolgavkar SH, Luebeck EG. Multistage carcinogenesis and the incidence of human cancer. Genes Chromosomes Cancer 2003; 38(4):302-306. 98. Bassing CH, Suh H, Ferguson DO et al. Histone H2AX: A dosage-dependent suppressor of oncogenic translocations and tumors. Cell 2003; 114(3):359-370. 99. Celeste A, Difilippantonio S, Difilippantonio MJ et al. H2AX haploinsufficiency modifies genomic stability and tumor susceptibility. Cell 2003; 114(3):371-383. 100. Dumon-Jones V, Frappart PO, Tong WM et al. Nbn heterozygosity renders mice susceptible to tumor formation and ionizing radiation-induced tumorigenesis. Cancer Res 2003; 63(21):7263-7269.

CHAPTER 18

PARP Inhibitorsand CancerTherapy Nicola J. Curtin

Abstract he compelling evidence for the role of poly(ADP-ribose) polymerase(s) (PARP) in the cellular reaction to genotoxic stress was the stimulus to develop inhibitors as therapeutic agents to potentiate DNA-damaging anticancer therapies. The earliest inhibitors, the benzamides, developed in the 1980s provided "proof of principle" evidence that such an approach was feasible but they lacked the potency and specificity for advanced preclinical evaluation. Over the last two decades potent PARP inhibitors have been developed using structure activity relationships (SAR) and crystal structure analysis. These approaches have identified key desirable features for potent inhibitor-enzyme interactions. The resulting PARP inhibitors are up to 1,000 times more potent than the dassical benzamides. These novel potent inhibitors have helped define the therapeutic potential of PARP inhibition. They significantly enhance the in vitro cytotoxicity of DNA monofunctional alkylating agents e.g., temozolomide, topoisomerase I poisons and ionising radiation. PARP inhibitors increase the antitumour activity of these three classes of anticancer agents in vivo, in some cases resulting in complete tumour regression. On the basis of these extremely promising preclinical data, clinical trials with a PARP inhibitor, in combination with temozolomide, commenced in June 2003 in the UK. This trial will allow the evaluation of PARP inhibition as a therapeutic manoeuvre in cancer for the first time.

T

Introduction The involvement of PARP in the cellular response to cancer therapy has been under investigation for almost half a century. The observation that stimulated these enquiries was that DNA alkylatinglagents, the first cancer chemotherapeutic agents, depleted the NAD + content of tumour cells. The ADP-ribose polymer product and NAD ÷ consuming enzyme responsible, which we now know to be PARP- 1, was discovered later.2 Continuing research has shown that PARP is rapidly activated by alkylating agent-induced DNA strand breaks and that PARP activation precedes the resealing of DNA lesions. This process imposes a high energy cost on the cell underlining the importance of poly(ADP-ribose) synthesis in the immediate DNA damage response. The compelling evidence for the role of PARP in the cellular reaction to genotoxic stress has provided an impetus for the development ofinhibitors both as a tool to determine the function of PARP and as potential therapeutic agents for use in combination with DNA-damaging therapies. Knowledge of the catalytic mechanism of the enzyme is important for the development ofinhibitors. PARP catalyses the cleavage of its substrate, NAD; at the nicotinamide-ribose bond with polymerisation of the ADP-ribose moiety and release of nicotinamide (Fig. 1). Nicotinamide is itself a weak inhibitor of the reaction and most PARP inhibitors contain the nicotinamide pharmacophore. The earliest inhibitors were simple analogues of nicotinamide-the benzamides (Fig. 2)

Poly(ADP-Ribosyl)ation, edited by Alexander Btirkle. @2006 Landes Bioscience and Springer Science+Business Media.

PARP Inhibitors and Cancer Therapy

219

* Attachment point (:z Linear chain Branched Chain 6 fl

Nicotinamide

rX

0 //~P~ I/~p~C]~

'~hl

O I .r--T ~ N / H I II I I

N ~ ~'~"N

,:,

I

Bond undergoing Cleavage

~ ~ ' * OC OH OH

NH2 Figure 1. NAD cleavagemechanism of PARP.

Benzamide was first shown to inhibit PARP activity by Shall,3 and the development of 3-substituted benzamide PARP inhibitors by PurneU and Whish 4 enabled the investigation of the effects of PARP inhibition in cells exposed to DNA damaging agents. In the first of such studies, the PARP inhibitor 3-aminobenzamide (3AB) at 1-3 mM, but not the inactive analogue 3-aminobenzoic acid, inhibited NAD ÷ depletion, delayed the rejoining of DNA strand breaks and increased the cytotoxic effects induced by the DNA alkylating ~ent, dimethyl sulphate, detected as reduced colony formation of proliferating cells in culture. The specificity of the benzamides for PARP has been questioned, in particular, at the concentrations needed for chemo- and radiopotentiation (3-10 mM) they can inhibit de novo purine biosynthesis and glucose metabolism, 6'7 which could potentially complicate the interpretation of the data. Studies of the role of PARP in response to some cytotoxic agents e.g., cisplatin, 1,3-bis(2-chloroethyl)-l-nitrosourea (BCNU) and etoposide based on the use of b e ~ i d e s show a degree of inconsistency: For example, enhancement of cisplatin cytotoxicity by 3AB has been reported in some cells,8 but not others. 9These apparendy conflicting data could be due to cell specific differences in repair mechanisms and/or the nonspeciflc effects of the benzamides. Nevertheless, several studies have consistendy demonstrated potentiation of DNA monofunctional alkylating agents and ionising radiation by 3AB leading to the conclusion that PARP participated in DNA base excision repair (BER) (reviewed in refs. 10,11). These conclusions have been vindicated by molecular studies using antisense strategies, 12 trans-dominant inhibition of PARP13'14 and disruption of the gene encoding PARP-1,15'16 which also render cells hypersensitive to alkylating agents and ionising radiation and are associated with delayed DNA repair. 17Additionally, several studies have demonstrated the interaction of PARP-1 with other components of the BER pathway. 18-21

o

o

o

NH2 Nicotinamide IC5o = 210 FM

Figure 2. Classical PARP inhibitors.

Benzamide IC5o = 22 FM

3-aminobenzamide IC so = 33 p,,M

(3AB)

220

Poly(ADP-Ribosyl)ation

The essential nature of PARP in the repair of DNA strand breaks has been disputed, with some authors reporting that DNA strand break rejoining is not compromised in the absence of PARP in cell-free systems,22'23 although other studies show DNA repair is inhibited in the absence of PARP both in cell free21 and intact cell systems. 17 Nevertheless, the use of PARP inhibitors to obstruct DNA repair and hence increase the cytotoxicity of certain classes of anticancer agents remains a viable strategy because inhibited PARP and a lack of PARP are not exactly equivalent. The difference between inhibited PARP and PARP deficiency has been most clearly illustrated in studies demonstrating that cell extracts depleted of PARP could repair nicked plasmid DNA efficiently but DNA repair was inhibited in PARP containing cell extracts in the presence of 3AB.22 Inhibition of DNA repair by prevention of poly(ADP-ribosylation) may result from a failure of PARP to dissociate from the DNA nick, thereby obstructing the access of the repair machinery. Moreover, PARP inhibitors may have a more profound effect on DNA repair than studies with PARP-1 deficient cells would suggest because ADP-ribose polymer formation in response to N-methyl-N'nitro-N-nitrosoguanidine (MNNG), and its inhibition by benzamide, was demonstrated in cells from PARP-1 null mice. 24 Subsequently a second PARP, PARP-2, which is activated by damaged DNA, despite lacking zinc fingers, and is sensitive to 3AB and other PARP inhibitors was discovered.25 PARP-2 knockout mice are also viable and fertile26 but, critically, deletion of both PARP-1 and PARP-2 confers embryonic lethality27 demonstrating the importance of at least one functional DNA damage activated PARP. The high degree of sequence homology in the catalytic domains between PARP-1 and PARP-2 and the demonstration that 3-AB inhibits PARP-2 suggests that much of the chemo- and radiosensitization by PARP inhibitors results from inhibition of both PAR~-I and PARP-2.

Development of Novel PARP Inhlbitors The limited potency and specificity of the benzamide inhibitors has been the incentive to develop better inhibitors to elucidate the consequences of PARP inhibition in cells and animals. To this end PARP inhibitors with increasing potency have been developed over the last two decades through a combination of structure activity relationships (SAR), determined by screening of a variety of existing compounds and novel chemical entities, and crystal structure-based drug design. These inhibitors have been extensively reviewed recently references 11 and 28-30, and the following section will refer only to significant and/or recent developments in relation to anticancer therapy. A major contribution to the development of novel inhibitors has been the work of Banasik et a131 who screened 170 compounds for their inhibitory potency against PARE. This study identified several compounds with potent PARP inhibitory activity including 4-amino-l,8-naphthalimide, 2-nitro-6- [5H]-phenanthridinone, 1,5-dihydroxyisoquinolineand 2-methylquinazolin-4-[3 H] -one (Fig. 3), which have been used as lead compounds for subsequent drug development by various groups. Around the same time structure activity studies of novel chemical entities identified the dihydroisoquinolines including 3,4-dihydro- 5-methylisoquinolin- 1(2H)-one, PD 128763, 32,33and the quinazxslinones including 8-hydroxy-2-methylquinazolin-4-[3H]-one, NU1025 ~l as being around 50-fold more potent PARP inhibitors inhibitors than 3AB (Fig. 4). Cell based studies demonstrated that PD128763 and NU1025 at concentrations of 50-100 gM gave superior chemopotentiation to 1-5 mM 3AB. The conclusion from these studies was that certain features were required for potent PARP inhibition viz., an electron-rich aromatic or polyaromatic heterocyclic system, a carboxamide group, with at least one amide proton for putative hydrogen bonding, and a non-cleavable bond at the position corresponding to the 3-position of the benzamides. The restriction of the carbonyl group into the anti conformation with respect to the to the 2,3-bond of the heteroaromatic ring was crucial for PARP inhibition. In the potent compounds illustrated in Figures 3 and 4 this is achieved by anchoring the carboxamide covalently. The alternative approach of restricting the carboxamide in the favourable orientation through intramolecular hydrogen bonding, also yielded potent PARP inhibitors. The 2-substituted benzimidazole-4-carboxamides were found to exhibit extremely potent PARP inhibitory activity in the nanomolar concentration range, e.g.,

PARP Inhibitors and Cancer Therapy

221

0

0

~~CH3

~ N H OH 1,5-dihydroisoquinoline IC5o= 0.39 ,aM

2-methylquinazolin-4-[3H}one IC5o= 5.6 t~M

0

0 NH

H2

0

~

N

H

L ~ ~ NO2

4--amino-1,8-naphlhalimide ICso = O.18 .M

2-nilro-6[5/-/~enanl~ dinone IC5o= 0.35 ~tM

Figure 3. Compounds identified by Banosite et al, 1992. 2-(4-hydroxyphenyl)benzimidazole-4-carboxamide, NU1085 (Fig. 5) which demonstrated good chemosensitizing activity in cell cultures at a concentration of 10 gtM.39 Another important development in the development of PARP inhibitors was solving the crystal structure of the PARP catalytic domain. 4° Crystalization of PD128763, 4-amino-naphthalimide and NU1025 in the NAD + binding site confirmed the prediction from the SAR study that the carboxamide group made several important hydrogen bonds with the protein and that restriction of the carboxamide within a heteroring rather than allowing free rotation, as in 3AB, improved the interaction. 41 This study and investigations at Agouron Pharmaceuticals 42'43 demonstrated that potent PARP-1 inhibition was dependent on the orientation of the inhibitor carboxamide oxygen to form two hydrogen bonds with Ser904-OG and the Gly863-N, while the adjacent inhibitor amide nitrogen donates a hydrogen bond to Gly863-O. Analysis of the binding of the benzimidazole, NU1085, demonstrated that the pendant 2-phenyl rin4~ extends into a larger pocket which can accommodate a number of substitutions (Fig. 6). Using a rational drug design approach, based on crystallographic data, 0

0 NH H

CH 3 PD 128763 IC so = 420 nM

Figure 4. ConformationaUyrestricted PARP inhibitors.

N~CH3 OH NU 1025 IC 50 = 400 nM K i = 48 nM

222

Poly(ADP-Ribosyl)ation

0

~~N

0

NH2

~

NH2 N

CH3

( ~

OH NU1064 IC so = 1000 rlVl Ki = 99 nM

NU1085 IC so = 80 nM Ki = 6 nM

Figure 5. Benzimidazole PARP inhibitors. several tricyclic lactam indoles and benzamidazoles were developed in which the carboxamide group was maintained in the favourable orientation by incorporation into a 7-membered ring (Fig. 7). 43-46These tricyclic compounds make critical hydrogen bonds between the lactam ring and the Gly863 and Ser904 residues with the nonplanar conformation of the seven-membered

9:

'

A

,:~:.....

............ ~ ......

Figure 6. NU 1085 located in the PARP-1 catalytic domain.

7

//

::

PARP Inhibitors and Cancer Therapy

0

223

0

0

0

N/CH3 \

CH3 Benzamidazoles

Tricyclic Lactam Indoles

Tricyclic Benzimidazoles T13 R =4' F Ki = 5 n M

AG14361 Ki < 5 nM

Figure 7. Benzimidazoles, tricydic lactam indoles and benzimidazoles. lactam probably contributing to this ideal interaction by allowing the lactam carbonyl and-NH to adopt a position closer to the protein residues involved in the hydrogen bonding. AdditionaUy, the cocrystal structure indicates that nearby tyrosine residues (Tyr907, Tyr889, and Tyr896) in the protein apparendy participate in/r-n-type interactions with the benzimidazole/ indole core and the C-2 substituent. Analysis of both structures reveal an ordered water molecule that participates in hydrogen bonding (2.8-3.1 A) between the free benzimidazole/indole NH and the important catalytic Glu988 residue. 47 The critical role of the free NH of the carboxamide moiety was confirmed using a SAR approach asprevention of hydrogen bonding by methylation at this position abrogated PARP inhibition. 45

Cell-Based-Studies with Novel PARP Inhibitors These more potent PARP inhibitors have helped resolve the discrepancies resulting from studies with the b e ~ i d e s . For example, an investigation using some of the more potent inhibitors identified by Banasik, 31 i.e., 4-amino-l,8-naphthalimide, 2-nitro-6-[5H]-phenanthridinone and 1,5-dihydroxyisoquinoline as well as 4-hydroxyquinazoline (Fig. 3) and 3AB (Fig. 2) investigated chemosensitization in a panel of ovarian cancer cell lines.48These studies demonstrated that all of the novel PARP inhibitors potentiated the cytotoxicity of the monofunctional agent MNNG, none of them potentiated cisplatin, and only 3AB potentiated the cytotoxicity of the bifunctional agent 1,3-bis(2-chloroethyl)- 1-nitrosourea (BCNU). Studies with novel PARP inhibitors not only confirmed the potentiation of ionising radiation and monofunctional alkylating agents, but also demonstrated the potentiation of topoisomerase I poisons, as described below.

lonisin g Radiation Approximately 40% of cancer patients receive radiotherapy as some part of their treatment (http://www.cancerhelp.org.uk/help) and radiosensitization may represent the most significant anticancer therapeutic utility of PARP inhibitors. Radioresistance is a major clinical problem and there is good evidence that it is due to a radiation-resistant, growth-arrested cell fraction within a tumour that can reenter the cell proliferation cycle and thereby repopulate the tumour after radiotherapy.49'5° In some in vitro radiopotentiation studies attempts have been made to mimic the clinical situation by measuring recovery from potentially lethal damage (PLD). In in vitro models of PLD recovery, the increased survival of growth-arrested cells is assessed after a recovery period, in comparison with the survival of cells without the recovery period. In other studies with exponentially growing cells irradiation dose-response curves in the presence and absence of PARP inhibitor have been compared. Alternatively, recovery from

224

Poly(ADP-Ribosyl)ation

sub-lethal damage has been measured, in this case the dose of radiation that would cause PLD is divided into two doses separated by a recovery period with or without a PARP inhibitor. The dihydroisoquinolinone, PD128763, (Fig. 4) at a concentration of 500 gM blocked recovery from both sub-lethal and potentially lethal doses of X-rays and approximately doubled X-ray-induced cell kill in both proliferating and stationary cultures. In contrast, a concentration of 10 mM 3AB was required to block PLD recovery and this concentration only partially blocked recovery from sub-lethal damage. 33 Similar studies showed that the quinazoline, NU1025 (Fig. 4) at a concentration of 200 gtM significantly enhanced the cytotoxicity of ~'-irradiation by 1.4-fold compared to a (non significant) 1.2-fold sensitization by 10 mM 3AB in exponentially growing L1210 cells. Both inhibitors were capable of blocking recovery from PLD. In these studies the radiosensitization by NU1025 was accompanied by a significant retardation in DNA strand break rejoining. 37 Radiosensitization studies have been conducted using some of the potent inhibitors identified by Banasik et al.31 Weltin et al showed that 6-[5H]-phenanthridinone (Fig. 3), at a concentration that inhibited PARP by >90% (100 ~tM), 2 hours before and after ~,-irradiation significantly increased radiation-induced cytostasis and apoptosis. 51 Comparable radiosensitization was achieved with 3AB but at concentrations 200x higher. Schlicker et al used 4-amino-l,8-naphthalimide (Fig. 3) to demonstrate a concentration-dependent radiosensitization in a panel of human and rodent cell lines. PARP activity was undetectable at concentrations greater than 30 gtM and following exposure to the nontoxic concentration of 20 gtM 4-amino-l,8-naphthalimide caused sensitizer enhancement ratios of 1.3 to 1.5.52 More recently the very potent tricyclic benzimidazole PARP inhibitor AG 14361 (Fig. 7) has shown potent radiosensitizing activity and, at a concentration of only 0.4 ~tM, AG 14361 inhibited recovery from PLD by 73%. 46 Ionising radiation causes a plethora of DNA damage, base modifications, single and double-strand breaks, with DNA double-strand breaks (DSB) considered the most cytotoxic. PARP has largely been associated with the repair of DNA single strand breaks and the general assumption has been that radiosensitization by PARP inhibitors is caused by inhibition of BElL However, in 1980 Bejamin and Gill 53 found that blunt-ended DNA DSBs were the most powerful activators of PARP in an in vitro system. Studies with NU1025 (Fig. 4) have shown that inhibition of PARP retards rejoining of ionising radiation-induced DNA DSBs. 54 Moreover, PARP has been shown to stimulate DNA-PK, 55 an important component of the nonhomologous end-joining pathway of DNA DSB repair.56 Recent studies demonstrated that PARP inhibition by AG 14361 caused a 40% increase in T-irradiation-induced cytotoxicity in proliferating cells, a 70% reduction in PLD recovery and a 40% reduction in DNA DSB repair. 57 Combination of AG 14361 with a novel DNA-PK inhibitor caused a >90% reduction in DNA DSB repair and completely abolished PLD recovery.57 Interestingly, these studies showed that the combination of inhibitors had a more profound effect in PARP and DNA-PK proficient cells than in their deficient counterparts suggesting that inhibition of either of these enzymes is more detrimental than their absence. This could be attributed to a loss of mutual stimulation or by the inhibited enzyme obstructing the access of the other to the DNA break. In these studies AG 14361 caused a modest but, nevertheless, significant radiosensitization and inhibition of DNA DSB repair in PARP-/" cells, suggesting PARP-2 may also contribute to repair of and recovery from IR-induced DNA DSB. The effect of PARP inhibitors on the survival of cells exposed to short pulses of ionising radiation separated by a short time interval (1-60 seconds) has been investigated. 58 Cell survival fluctuated with the duration of the interval between doses of radiation. PARP-1-/- cells did not exhibit this oscillatory response and it was largely abolished in PARP proficient cells by 3AB (10 mM), 4-amino-l,8-naphthalimide (30 gtM) and 2-nitro-6[5H]phenanthridinone (100 gtM) (Fig. 3). In contrast, the DNA-PK inhibitor, wortmannin, did not affect the oscillatory response, suggesting that PARP, but not DNA-PK, plays a major role in the early radiation response. These authors proposed that the rapid poly(ADP-ribosylation) of target proteins, or

PARPInhibitorsand CancerTherapy

225

recruitment of repair proteins to sites of initial DNA damage affected the induction of, or response to, DNA damage by the second dose of irradiation.

Monofunctional DNA Alkylating Agents Since the first study of PARP inhibitors combined with DNA-methylating agents in 19805 there has been significant interest in the development of PARP inhibitors as modulators of resistance to andcancer DNA methylating agents. Agents of current clinical interest indude 5-(3,3-dimethyl-l-triazeno)imidazole-4-carboxamide (DTIC) and, in particular, temozolomide, an orally available drug recently licensed for the treatment of malignant gliomas, astrocytomas and melanomas (http://virtualtrials.com/temodar). The ultimate alkylating species from both these drugs is methyltriazenoimidazole-4-carboxamide (MTIC) which methylates the DNA at the Ot-position of guanine, the N 3 position of adenine and the N 7 position of guanine. 59 In the majority of cells the most cytotoxic lesion is Ot-methyl guanine, even though it only accounts for 5% ofthe total lesions. This lesion can be repaired by methylguanine methyltransferase (MGMT) but this protein is rapidly saturated and the persistence of Ot-methyl guanine leads to mispairing. The other, much more numerous, N-methylpurines are targets for BER. There have been several studies investigating temozolomide chemosensitization by PARP inhibitors (reviewed in ref. 60). The first of such studies compared the ability of PD 128763 and NU1025 (Fig. 4) with that of the much weaker classical PARP inhibitors; benzamide and 3AB (Fig. 2) to sensitize L1210 cells to temozolomide. 61 This study revealed an excellent correlation between the potency of the compounds as PARP inhibitors and their ability to enhance temozolomide-induced DNA strand breakage and cytotoxicity. Temozolomide cytotoxicity was increased 4- to 7-fold by coincubation with only 50-100 ~tM NU1025 or PD128763, whereas concentrations of 1 mM benzamide and 5 mM 3-aminobenzamide were needed for a similar level of potentiation. Subsequently 200 gtM NU1025 was shown to enhance the cytotoxicity of MTIC by 3.5-fold when it was administered simultaneously with the MTIC or only added after the MTIC had been removed. 37 These studies confirmed that PARP inhibition during the repair phase alone was sufficient for potentiation of MTIC cytotoxicity. In the same study exposure to 200 gtM of the benzimidazole PARP inhibitor, NU 1064 (Fig. 5), caused a 3-fold enhancement of temozolomide cytotoxicity. These studies were extended to an investigation of temozolomide potentiation by NU1025 and the more potent benzimidazole PARP inhibitor NU1085 (Fig. 5) in a panel of 12 human tumour cell lines representative of the commonest human malignancies: lung, colon, breast and ovarian. 39 In this study, a concentration of only 10 gtM NU1085 was required to potentiate temozolomide up to 6-fold. Potentiation was not dependent on tissue of origin or p53 status of the cell line. The novel and very potent PARP inhibitors have been used at very low micromolar or sub-micromolar concentrations to potentiate temozolomide cytotoxicity in a variety of cell lines. A series of potent (Ki all < 10 nM) benzimidazoles and tricyclic lactam indoles (Fig. 7), at a concentration of only 0.4 gtM, potentiated temozolomide-induced growth inhibition of LoVo (human colon carcinoma) by up to 5.3-fold. 45 In this study the analogue that was inactive as a PARP inhibitor, by virtue of methylation of the carboxamide nitrogen, did not potentiate temozolomide. In general, although having similar PARP inhibitory Kis the tricyclic compounds were the most potent chemosensitizers. Other very potent novel PARP inhibitors, CEP 6800 (Fig. 8) and GPI 15427 (structure not reported: ICs0 for PARP inhibition = 31 nM) have also been investigated in combination with temozolomide. CEP 6800 used at a concentration of 1 gtM markedly increased temozolomide-induced DNA damage and cytotoxicity in U251MG human glioblastoma cells,62 and GPI 15427 increased temozolomide growth inhibition by 9-fold, 2.5-fold and 2.6-fold in murine melanoma (B 16) and lymphoma (L5178Y) and human glioblastoma (SJGBM2) cells, respectively. 63 Ot-methylguanine is the most cytotoxic lesion caused by temozolomide because failure to repair it results in a post-replicative lesion that is a substrate for the mismatch repair (MMR)

•~!h) ssoa!q!qu! d~ffVd jo s~!sas alopu ! mea3r 1 3!i3L-~!saptre alOZ~!tu!zua q auaaod aqj= 6~'sau!l lla3 atla jo srueas ff

m L

0