117 82 11MB
English Pages [328] Year 2023
OUR GENES, OUR CHOICES
Disclaimer --- This book was written by David Goldman in his personal capacity. The opinions expressed in this book are the author’s own and do not reflect the view of the National Institutes of Health, the Department of Health and Human Services, or the United States government.
This page intentionally left blank
OUR GENES, OUR CHOICES
How Genotype and Gene Interactions Affect Behavior SECOND EDITION David Goldman
Office of the Clinical Director, Laboratory of Neurogenetics, NIAAA, Bethesda, MD, United States
Academic Press is an imprint of Elsevier 125 London Wall, London EC2Y 5AS, United Kingdom 525 B Street, Suite 1650, San Diego, CA 92101, United States 50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom Copyright © 2024 David Goldman. Published by Elsevier Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein). Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. ISBN: 978-0-443-22161-3 For information on all Academic Press publications visit our website at https://www.elsevier.com/books-and-journals
Publisher: Stacy Masucci Acquisitions Editor: Elizabeth Brown Editorial Project Manager: Tracy I. Tufaga Production Project Manager: Selvaraj Raviraj Cover Designer: Mark Rogers Typeset by STRAIVE, India
Disclaimer The contents of this book do not reflect David Goldman's official role and are not representative of the US government.
v
This page intentionally left blank
To Aaron, Ariel, Evir, Liliana, Alexandra, and Benjamin. Some of my genes, all of their choices.
This page intentionally left blank
Contents About the author xv Preface xvii 1. Introduction: Thou mayest choose 2. The jinn in the genome Fifteen minutes of fame 7 Some famous geneticists and why they are famous 8 The jinns of knowledge and technology 9 Revolutions in culture and evolution of genes 10 Genes, brain, and individuality 15 The neurogenetics of determinism and freedom 15 References 16
3. 2B or not 2B? Anecdata and data 17 A common stop codon causing impulsivity and hyperarousal 21 Validating an impulsivity gene in a mouse model 25 References 27
4. Stephen Mobley and his X chromosome The death of Stephen Mobley 29 The Kallikak effect 30 Mobley demands a genetic test 31 Combining gene and hormone to predict impulsivity 31 Carrying kohl to Italy 32 The state of DNA in prediction of violence 33 References 35
5. Dial multifactorial for murder: The intersection of genes and culture A murder in the lab Missing puzzle pieces, an obstacle to reductionism Why are some societies more violent? Guns or people?
ix
37 38 41 42
x Contents A fierce people 43 Civilizing people 45 Violent youth 46 References 48
6. Distorted capacity: The measure of the impaired will Conscious and unconscious behavior 49 Context appropriate and inappropriate behavior 51 Personality types and choices 53 The inheritance of impulsivity, and what it means 58 Impulsivity differs from person to person and from species to species 60 Zero-trial learning 62 Impulsivity and aggression in context 62 Measuring impulsivity and aggression 63 Integrating measures and genes 68 Measuring the brain 70 The arousal (thymos) of youth 73 Animal models of arousal, impulsivity, and aggression 74 References 76
7. Distorted capacity: Neuropsychiatric diseases and the impaired will Impulsivity, diminished capacity, and neuropsychiatric disease 77 Disorders of impulse control 87 References 93
8. Inheritance of behavior and genes “for” behavior: Gene wars The debate on the heritability of behavior 96 The genome encodes reaction range 97 Choice and reaction range 99 Reaction range and free will 99 Twin studies and controversies they provoked 100 The debate on genes “for” behavior 101 People are not monkeys 102 The politics of behavioral genetics 104 Antipsychiatry: Are psychiatric diagnoses valid? 107 References 114
9. The scientific and historic basis of genethics Standards of science and evidence Ethics of research: Trust, but verify Genes, jobs, and groups The Genetic Information Nondiscrimination Act Gene therapy
118 120 123 124 126
Contents xi
Group consent and individual consent 127 Beyond a pretense of autonomy 128 References 130
10. The world is double helical: DNA, RNA, and proteins, in a few easy pieces DNA recipes 132 Polymorphism 135 Protein polymorphism 139 DNA polymorphism 141 Measured ancestry 143 References 146
11. The stochastic brain: From DNA blueprint to behavior Self-assembly 148 Cell assembly 149 Interactomes 150 Stochasticity 152 Cascades, chaos, and great attractors 154 Brain assembly 154 Fire together, wire together 157 Fractal neurons 157 Stochasticity in higher order brain structure 158 The stochastic basis of individual and group intelligence 160 Rules guiding the chaos of brain evolution and development 162 Sense of self 163 References 166
12. Reintroducing genes and behavior Behavioral prediction, a science imperfect 171 Commercialization of behavioral prediction 172 The future of genetic behavioral prediction 174 A gene causing anemia 176 A gene causing self-mutilation 177 A gene causing cognitive deficiency 179 References 182
13. Warriors and worriers A common genetic variant “for” warriors and worriers 186 Executive cognitive function 187 Cognitive flexibility and free will 189 Perseveration 192 Worriers and warriors 195 References 198
xii Contents 14. How many genes does it take to make a behavior? Single genes 201 Polygenic and epistatic models of behavior 202 Bayesian reasoning—How to use prior probability 205 Behavior and the single gene 206 References 210
15. The genesis and genetics of sexual behavior Gender and sex 211 Biological determinants of gender 214 We are love machines 216 Sneaker males 218 Slaves to sex: The difficulty of turning off the sex drive 221 How people modulate and harness their sex drives 222 Taboos 223 Homosexuality and the “gay gene” 223 Elliot Gershon and the in-depth family paradigm 224 Discovery of the “gay gene” 225 Is homosexuality inherited from one’s mother? 227 Genes for homosexuality 228 References 229
16. Gene-by-environment interaction Variations on the theme of gene-by-environment interaction 231 Ancient environment × genome interaction 233 Nature × nurture 234 What is gene-by-environment interaction? 236 Genes that modulate stress resilience 237 Intermediate phenotype and endophenotype 240 Interactions leading to psychiatric disease 242 Animal models of gene-by-stress interaction 244 Love, in monkeys? 246 References 249
17. The epigenetic revolution: The imprint of the environment on the genome Measuring environmental contingency 251 An imprint of experience in the DNA 253 Types of epigenetic imprint 257 Wiping the epigenetic slate clean… 258 …But not quite clean 258 Measuring epigenetic variation 259 First look at the epigenetic “depth” of the human genome 261 References 262
Contents xiii
18. Time out for free will Temporary and longer lasting impairments of choice 264 Social implementations of the science of choice 266 Ideology in genomics: The example of race and ancestry 270 References 273
19. The top-down neurogenesis of free will Conscious automata 275 A brief manual of parenting 277 Free will and the conundrum of behavioral causality 279 Exorcizing genetic behavioral determinism 282 Neurogenetically influenced behavioral archetypes 285 References 287
20. Neurogenetic origins of free will Index 295
This page intentionally left blank
About the author David Goldman is one of the world's most influential neurogeneticists, with more than 500 publications cited some 80,000 times and many of his scientific studies published in the world’s leading journals. He is at the forefront of revolutions in imaging genetics, gene-by-environment interactions, pharmacogenetics, and DNA sequencing to identify genes that alter human behavior. He founded the Laboratory of Neurogenetics in 1991 at the National Institutes of Health and is a clinical director there. His awards include the NIH Director’s Award, presented by Francis Collins, for elucidating gene-by-environment interactions in Alcohol Use Disorder. He is the past chair of a human research review committee and has often spoken at international conferences about his work. Born and raised in Galveston, Texas, he attended Yale University, where he graduated in three years. He received his MD degree Magna Cum Laude from the University of Texas Medical Branch. He has three children and three grandchildren and is married to Nadia Hejazi, MD, a pediatric neurologist. Since 1985, he has lived in a house of his design in the woods of Potomac, Maryland, and, at least in pre-pandemic times, commuted by bicycle throughout the year.
xv
This page intentionally left blank
Preface This book is a story, and an analysis, of the neurogenetic origins of ehavior and free will in the context of a modern appreciation of neurob genetic determinism. None of us chooses our parents, ancestors, or the past world that shaped our genomes and that in the present constrains our behaviors. From early in life, we hear “whatever will be will be” or later the more elegant formulations of the inexorability of determinism such as that of Soren Kierkegaard, who said that we all come into this world with sealed orders. Conversely, we are taught that life and what we become is what we make it. From an early age, we learn to treat people as if they are free. We reward or punish them as if they choose, according to them the gift of respect as autonomous beings, rather than treating them as things. In part, we treat others as free because we empathize with their thoughts and emotions and understand the conflicting motivations that led to the behavior that we may or may not like. The failure to make such an emotional connection turns other people into objects and, as they say, makes the world a little colder. However, what if a judge is confronted with a genetic finding that a murderer carries the “2B or not 2B” variant, a human polymorphism found by my own laboratory that predisposes some individuals to impulsively murder? Should that information be used to mitigate a sentence or, paradoxically, should the carrier of this gene be viewed as even more dangerous and likely to repeat the act? In writing this book, I was inspired in part by responses to lectures at national judges' courses on the validity and meaning of such evidence. DNA and brain scans are increasingly used in courtrooms. What should be allowed? How should it be weighed? Is there a neurogenetic basis for moral responsibility, or do we merely treat people "as if" they are free? Since the publication of the first edition of Our Genes, Our Choices a decade ago, there have been dramatic advances in DNA technologies and in our understanding of human genetic variation and the influences of genes on behavior. This revised version updates and, in some instances, corrects the original. A revolution in personal genomics includes the new ability to sequence genomes, raising our consciousness about neurogenetic determinism and posing questions about how that information will be used, as illustrated in popular movies (e.g., GATTACA). Millions of people have purchased genetic scans to better understand their ancestry and nature. However, whereas ancestry can be measured
xvii
xviii Preface with remarkable precision, the reports they receive are sparse and subject to misinterpretation. People are intensely interested in what their genes say about their personalities and behaviors. Throughout this book, I try to minimize the jargon and the glossary (or even the internet) may assist people who get stuck on some piece of terminology. I strove to select stories to give a step-by-step account of how our DNA builds a brain capable of choosing whether to read all or any portion of a book like this. Advances in genetics and genomics allowed the genetic code to be read, and although the meaning of the words is still incomplete, the story they are telling is already profound and often surprising. As illustrated by the story of the “2B or not 2B” gene, the mystery of whether people are free has only been deepened by these new perspectives. As a neurogeneticist, it is a pleasure for me to chronicle some of the most interesting and profound ways in which genes influence human behavior, in the rich and strange new landscape unveiled by the genome revolution. Reflecting the evolutionarily linked community of life on Earth, many human behaviors, for example, sexuality, are paralleled in other species, and some are even influenced by genetic variations that act in similar ways and with similar effects. As I like to say, it is my hope that something we learn about the elusive origins of human behavior of people will shed light on the elusive behavior of cats. In people, neurogenetics gives us new insights into the origins of psychiatric diseases and everyday differences in behavior. I try to describe the ways by which genetic variation interacts with environmental stress and even endocrine differences to make some people “warriors” and others “worriers.” We are neurogenetically individual, but does that mean that from conception our DNA is destiny, or, as I will argue, is our DNA heritage paradoxically an essential ingredient in the freedom of the individual? Life often comes full circle or echoes the events of our formative years, and in my case, the nature of the child was the blueprint for the adult. I became a neuroscientist because, as a child, I perceived that the brain contained the essence of self and humanity, and there is no better way to know oneself than to study the brain. Simultaneously, and as became a lifelong interest, I perceived a problem. All life is based on causal relationships in the physical universe. The more I learned about myself and others, the more apparent became the connections within webs of causality. Could freedom exist in a world where things do not happen by magic? Why was it that as one of five children, I was like but also unlike my older sister, my younger brothers, and my parents? My dear brother Paul developed schizophrenia and later enriched the lives of all in our family with letters on diverse topics. What were the factors in his genetic predisposition, and what were the wages of his experience? Why do psychiatric diseases, and as I will describe,
Preface xix
terrible and dissocial acts, strike close to home and blight the lives of so many of our fellow humans, whom we would also spare such travail? The genomic era holds unprecedented promise for the discovery of clues to the origins of common, genetically influenced psychiatric diseases and, hopefully, will contribute to their prevention and treatment and perhaps even interrupt sequences of causality within those who would victimize others, the victimizers themselves being a type of victim. We are only at the beginning of the exhilarating process of discovering how genes encode the developmental program of the brain. The way the mystery is unfolding, and my own ability to make some contribution, is everything I could have hoped for when I first arrived at the National Institutes of Health in 1979 from Texas as a raw and overly optimistic young physician with a burning desire to understand the human brain and make a name for himself. The progress is also frustratingly slow and less than what is needed for families such as my own and for the millions of patients and families fighting their own everyday struggles against psychiatric diseases. Severe psychiatric diseases, schizophrenia, depression, alcohol use disorder, other addictions, obsessive-compulsive disorder, attention deficit hyperactivity disorder, autism, bipolar disorder, phobias, panic disorder, eating disorder, borderline personality disorder and antisocial personality disorder, etc. are common and affect someone in nearly every family. Yet all are stigmatized to one degree or another, and many are frequently denied. A key to why they are stigmatized may be that they impair their capacity to choose. For each of these diseases, therapy is only partially effective. Several have been the particular focus of an antipsychiatry movement that would deny that these problems should be classified as diseases or that people with them can benefit from medical treatment. And, indeed, it is not uncommon that treatment is inappropriately applied. However, even with our crude knowledge of psychiatric diseases, and a correspondingly pitiful number of diagnostic categories, we can be sure that genes play a strong role in causation. This is because of the inheritance of these diseases. Heritability = Genes. This book describes some of the first instances of gene discovery and false discovery and, in this revision, the rise of polygenic scores that are vastly increasing the power of genetic prediction. Those polygenic scores derive from the whole genome views of human behavior that were only achieved in the past decade via the application of high-throughput genotyping of a million or more genetic markers in populations of a million or more individuals, and as is even now being succeeded by whole exome and whole genome sequencing in such samples. The reason for the large samples is that the gene effects are small, but in their hundreds, they are beginning to add up. Although a genome-level understanding of behavior is still nascent and limited by both our molecular tools for linking sequence variation to behavior and
xx Preface limited knowledge of DNA function, the results are already startling, with hundreds of gene loci detected for even a single choice-related behavior such as risk-taking. As determinants of whatever type are identified, it is only logical that people turn to determinism. A decade ago, in the first edition of this book, I forecast that a new generation of scientists would soon be applying tools immeasurably more powerful, for example, using a new ability to measure the combined imprint of genotype and experience on the genome. In the last decade, not only has gene discovery accelerated, but epigenetic markers can now measure the impact of stress, drugs, and aging on the genome, unlocking part of the complexity of gene-by- environment interaction. Others have also begun to speak of the implications of neurogenetics for behavioral determinism (e.g., Chris Wilmott in Biological Determinism, Free Will and Moral Responsibility: Insights from Genetics and Neuroscience, 2016). However, this territory remains relatively unknown in relation to its importance. Further, I weigh the evidence and make the case “just a little bit differently,” proposing that free will arises from the unique ability of humans to guide the paths they take toward choice, via top-down guided plasticity at the individual level and also from the larger choices we make about the nature of the environments in which we live. Other animals, brilliant or brilliantly adapted though may be, do not seem to be capable of that feat—within themselves, within their communities, and across generations, although if and when some such species are identified we should reconsider their status as moral agents. Understanding the effects of genetic determinants on behavior is still nascent; however, the specter of neurogenetic determinism now looms much larger. We can already be certain of the role of environmental interaction, not only in terms of the general role of stress and endocrine factors but also via the specific gene-by-environment interaction stories that I tell. Genotype can itself become a determinant. With tens of millions of people having taken advantage of opportunities for personal, directto-consumer genomic analysis, it is commonplace to see genotypes being used to guide behavior—and, for example, a person who discovers they have genetic roots in Scotland may dutifully take on those cultural trappings, identifying with an ancestry and culture to which they otherwise have no connection. This book describes how we know about the relationship of genes to behavior, which are foundations for theory of behavior or public policy. It describes a most exciting aspect of the story, which is that as we penetrate deeper into the brain's secret functions, we find that actions of genes tend to be progressively stronger. At the same time, many genetic variants altering human behavior are not “disease genes” but are literally genes “for” behavior, having been selected not to alter an obscure brain function but to influence the behavior itself, leading to differences in reaction range
Preface xxi
of personality and cognition. This fact, and the fact that we all carry thousands of genetic variants, some beneficial and some highly deleterious, represents a profound refutation of the eugenic impulse to identify some humans as “fit” and others as “unfit.” Fitness is all about context. Behavioral genetics has a long and illustrious history but was also from its beginnings and throughout its course marred by eugenics, and polemics on gene vs environment and racial differences, as this book discusses. Putting the “gen” in genocide, the Nazis sought to exterminate anyone with a quantum of Jewish ancestry. Today, a drop of Ashkenazy ancestry is measured very easily and accurately, and again, people are placing great weight on ancestry and genetic predisposition, and genomics sees right through attempts to pass as having a different ancestry or genetic predisposition. The person in the top decile for body mass index polygenic score is always recognizable as being predisposed to obesity, even if throughout their life they have never been overweight. Prior to the genome revolution, which happened only in the last decade of the 20th century, behavior geneticists quantified the inheritance of behavior by measuring the resemblance of people at different degrees of genetic relationship. However, with the completion of the draft sequence of the genome, neuropsychiatric genetics has leapt forward to the level of the specific genes and gene variants that alter behavior and how they work. This book tells the story of how this research is being conducted not on a gene-by-gene basis, but on a global, which is to say “genomic” basis. We have an exciting first glimpse at what we are, and why we are, and the results are startling and controversial in many ways. In modern scientific parlance, the old paradigm of understanding human behavior has been broken or “disrupted” and now our field is struggling to replace it with something new and based not on a quantitative appreciation of causality (“people who experience stress tend to become depressed”) but on a more exacting understanding of who that person was before and after the stress. For me, it has been a privilege to be part of that first generation of “behavioral genomicists” who have been joined by new generations that have not known things any other way. They are a varied and fascinating lot and comprise part of the cast of characters in this book. I hope that I have not done any of them a disservice with how they are represented or by omitting them. I have mainly written about fellows, friends, colleagues, and scientific heroes who I personally know. Some were in my own lab, some at laboratories with which I collaborated or competed, and some are just old friends, with whom I have contended and created. It would not be “in me” or “like me” to write from a biographical perspective, for example, via systematic study of letters or interviews. Yet it is important to point out that in mentioning people in connection to one piece of science, shared events, or a memorable apercu, it was not my intention to minimize their lives or contributions. I only know what I know, and
xxii Preface it is from that perspective, as a physician scientist that I wrote this book and more lately revised it. It also might go without saying that this book is not an autobiography. Much more remains unsaid than could be said, even if there was “more space in the margins.” Meanwhile, the stories I selected are directed toward one fascinating question: Are people free to make choices, or do genes determine behavior? Choosing the stories was the main enjoyment of writing this book. Neurogeneticists are spoiled for choice, and yet I have seen that most of these stories are scarcely known to the public. For example, in some fish, a single genetic switch changes female to male and female behavior to male behavior—both sex and gender are changed. Humans are more complicated—for example, sex and gender are not so tightly linked, but all human behaviors emerge from the expression of DNA. Are people free to make choices, or do genes determine behavior, much as a genetic switch reprograms the behavior of the fish? My studies of human behavior convince me that paradoxically the answer to both questions is “yes” because of neurogenetic individuality and self-guided neural plasticity. In Our Genes, Our Choices, the complexity of human behavior and a person's ability to choose are explained as deriving from the ways in which a relatively small number of genes direct a neurodevelopmental sequence. This lifelong process is guided by individual genotype, molecular and physiological principles, by randomness, and by environmental exposures that we choose as well as ones that we do not choose. This theory affirms and provides a mechanism for the origins of free will and the ability to make moral choices, but I suspect that the debate will continue while humans exist.
1 Introduction: Thou mayest choose Timshel—"Thou mayest”–that gives a choice [between good and evil]. John Steinbeck— East of Eden
Is the human will free? Do genes determine behavior? Paradoxically, the answer to both questions is “yes.” A new theory of behavior based on neurogenetic individuality and top-down driven neural plasticity has profound implications for conceptions of self, social expectations, ethics, and justice. This book begins with a challenge to free will from my research. The discovery: “2B or not 2B?” involves a gene “knockout” of the HTR2B gene, which encodes a receptor for the neurotransmitter serotonin. Serotonin is involved in many aspects of behavior. The functionally disruptive receptor variant causes some people to be impulsive and hyperaroused and irritable, even to the extent of committing senseless murders. Remarkably, it is found in at least 100,000 people in the Finnish population, but as a “founder mutation” it has so far only been observed in individuals who are of Finnish ancestry. Yet, while the inherited variant was a “necessary” factor in the impulsive murders that I and my partners in research studied, the gene alone was insufficient to explain the heinous behavior. “2B or not 2B?” was not the only question. The context of the gene, for example male sex and drunkenness, also mattered. All human qualities, including those that are sublime, creative, and adaptive, and those that are seemingly mundane, destructive, and maladaptive, are ultimately emergent from the expression of a “message in a molecule.” That molecule is DNA. DNA is an information molecule—a polymer in which information is encoded. For example, our DNA contains some 25,000 genic protein-coding regions. However, DNA is ultimately only a chemical that is now easy to synthesize in the laboratory. The total DNA of a species of bacteria was recently made in a laboratory, and it is only a matter of time and motivation before someone synthesizes the genome of a more complex creature—even a human. Also, because of
Our Genes, Our Choices https://doi.org/10.1016/B978-0-443-22161-3.00020-X
1
Copyright © 2024 David Goldman. Published by Elsevier Inc. All rights reserved.
2
1. Introduction: Thou mayest choose
the tools now available to study DNA and the ability to study its effects in powerful contexts, including in animal models in which genes are edited, the science of genetics has now achieved what seemed impossible only a few years ago. We have demonstrated the causal connection, and not just the correlation, between single genetic variants and complex human behaviors. As will be shown, this reductionistic explanation of human behavior is only in its infancy, and complicated by many difficulties and some false leads, and has now been vastly augmented, and folded into, polygenic prediction. With some 25,000 protein building blocks and probably an equal number of regulatory RNA molecules, and even allowing for variations in structure, how is it possible that the DNA message can encode a human brain, with its 1015 (one billion x one million = quadrillion) connections? How, based on the DNA code, can a brain build itself? Is it possible that the complexity of human behavior, and even free will, can be derived from the chemistry of DNA? As will be proposed, the answer lies in the way that this relatively small complement of genes directs a developmental sequence that continues throughout life and that is guided by principles, stochastic in countless details and always completely individual. Pathways from the deep evolutionary, developmental, genetic, molecular and cellular levels of explanation to complex behavior, and the ability of “things we cannot control” to shape behavior, are illustrated by sex. Why do males and females behave as if from different planets and, beyond the effects of culture, how are we to understand the origins of variations in sexual behavior of sex-specific behaviors ranging from attachment, to aggression, to homosexuality? Is there a “gay gene”? Why is there genetics of sexual behavior and how can we understand the diversity, or even perversity, of human sexual behavior? A clue is that in other animals there are sex genes, and several have been identified, although in the human the identity of these genes is yet unknown. What are the implications for choice and conceptions of personal freedom that people are born male, female, or gay, or that a switch in the function of a single gene can cause a fish to stop acting like a female and start acting like a male? Despite the complexity and using methods that can find the needles hidden in the genetic haystack, I and other neurogeneticists have identified the first genes that predict cognitive and emotional differences, and sometimes the same gene can have countervailing effects on both. The ability of genes to predict behavior is explained by example, as are the limitations and nuances that include gene-by-environment interaction. The genetic variants—even in large polygenic combinations—identified so far are not strongly predictive of behavior and their value should not be oversold, as has already been done. Yet they are of explanatory value and foreshadow the discovery of additional variants that must account for the heritability of human behavioral characteristics. Rapid advances in
1. Introduction: Thou mayest choose
3
genetic behavioral prediction enhance the case for neurogenetic determinism. The new science of gene-by-environment interaction is explored via genes that can lead to depression, anxiety, impulsivity, and even suicide. For example, genes predispose some people to be resilient “warriors” and others to be less resilient and pain resistant but sometimes cognitively advantaged “worriers.” Furthermore, polygenic scores are accounting for increasingly larger fractions in the genetic variance (variation), and total variance in liability to a variety of psychiatric diseases, as well as everyday behaviors such as risk-taking. The findings in this relatively young science of behavior genetics are not without controversy, as would be expected. Discovery of mechanisms of gene action lag behind gene identification. However, there are solid examples of genetic variations that affect behavior, and that also affect what the brain is doing during behavior, as is now observable with brain imaging, which is a window on the activity of the brain. By imaging the regional structure, activity, and chemistry of the brain, for the first time the basis of the effects of these behaviorally important genes has been understood. Also, their predictive effects on brain function itself are much stronger than on overtly manifested behaviors. Several genes with weak effects on anxiety and emotion have strong effects on brain responses to emotional challenges. Two genes that influence cognition have strong effects on brain activity while people are asked to perform cognitive tasks that challenge specific parts, and neuronal networks, of the brain. By studying the combined effects of many genes, overlaps in genetic causation that had been observed via cross-inheritance of different diseases in twins are now being observed at the gene level. Genes not only partly determine behavior, but the same genes influence multiple behaviors. This book is concerned with the neurogenetic foundations of behavior and with the possibilities and limits of genetic behavioral prediction; however, it is unavoidable that these discoveries would be connected to conceptions of self and freedom. These questions are strongly embedded in popular culture, as I will touch upon in this book. In recent years the ages old debates about self and freedom have partly motivated many fine books, and here to cite a few (Churchland, 2011; Damasio, 2005; Dawkins, 1996; Dennett, 1984, 1996, 2004; Kane, 2005; Pereboom, 1997; Sternberg, 2010; Wilmott, 2016). Is it sufficient to treat people “as if” they have free will? I argue that it is not. The “as if” stance is inherently inconstant, setting the stage for the easy erosion of individual autonomy whenever situational ethics dictate that it is more expedient to treat people as slaves to causality. Compatibilism is the philosophical position that determinism and free will can be reconciled. However, I would also take issue with Daniel Dennett’s compatibilist formulation, which holds a dependency on culture, “morality memes,” and child upbringing, and even the idea that to have free will one it is necessary that one must believe in free will or be
4
1. Introduction: Thou mayest choose
disabled as a chooser. All are metastable. As will be discussed, individual and group autonomy are the vital bases of moral ethics, for example as applied in the conduct of human research, where these are foundational principles. However, are these principles divined from a philosophical or practical calculus or are they inherent to human nature? Are they suppositions, in which case any system of moral ethics may define humans otherwise, or are they parameters based upon observation? I will argue that individual free will, and by extension the autonomy of people, is neurogenetically encoded, to be revealed by ongoing selfguided neural plasticity throughout life, and that while it may be impaired in some, we have it whether we deny it. To begin with, and finally, each person, including an identical twin, is neurogenetically individual, and their brain development unfolds stochastically throughout life in a way that makes them unique and ultimately self-determined. Consequently, people are predictably unpredictable, although it is not this “unpredict–ability” that itself represents free will, many events, and for example the weather a few months from now, being unpredictable. Freedom is bound to individuality, which is partly the product of neurogenetic determinism, which is itself bound to the ways the human genome, and brain, was shaped by evolution and that would include the random events that altered humankind’s evolutionary path. Freedom does not consist of randomness: philosophers such as Robert Kane and Dennett were correct to emphasize that free will does not originate in quantum randomness—we cannot assign agency to randomness. However, our brains capitalize on randomness as raw material for the development of our individuality. Dennett warned that we should not look too closely at mental activities, or we may discover that we have no selves. Coming at the problems of self and free will from a neurogenetic perspective, I counter that the more closely we look the stronger the self emerges. Our deepening understanding of genetic and environmental predictors inevitably improves behavioral prediction. With genotypes, epigenetic markers of environmental exposure and brain scans reading out the connectomics of the brain we can better anticipate responses of any person, be they neuroscientist or philosopher or someone intrigued by their musings. But whether someone may predict our choices, we may choose.
References Churchland, P., 2011. Braintrust: What Neuroscience Tells us about Morality. Damasio, A., 2005. Descartes’ Error: Emotion, Reason and the Human Brain. Dawkins, R., 1996. The Blind Watchmaker. Dennett, D., 1984. Elbow Room: The Varieties of Free Will Worth Wanting. Dennett, D., 1996. Darwin’s Dangerous Idea: Evolution and the Meaning of Life. Dennett, D., 2004. Freedom Evolves. Kane, R., 2005. A Contemporary Introduction to Free Will.
1. Introduction: Thou mayest choose
5
Pereboom, D., 1997. Free Will. Hackett Readings in Philosophy. Sternberg, E.J., 2010. My Brain Made me Do it: The Rise of Neuroscience and the Threat to Moral Responsibility. Wilmott, C., 2016. Biological Determinism, Free Will and Moral Responsibility: Insights from Genetics and Neuroscience.
This page intentionally left blank
2 The jinn in the genome I never saw no miracle of science That didn’t go from a blessing to a curse Sting—“If I Ever Lose My Faith in You”
Fifteen minutes of fame In 2002 I helped conceive “Our Genes/Our Choices,” a Public Broadcasting series that explored ethical and legal choices created by the genome revolution. Topics of these Fred Friendly seminars included genetic reproductive decisions and genetic privacy. Our session, “Genes, Choices and the Law,” was adeptly moderated by Charles Ogletree, a well-known Harvard law professor. The scenario involved a person with Alcohol Use Disorder (then, wrongly called an “alcoholic”) accused of a crime committed while they were intoxicated, and a genetic test result that might be mitigating (Breyer et al., 2002). I was the geneticist with a predictive test. The scenario, more like science fiction in 2002, is close to reality today, with dozens of genes contributing to AUD identified, and polygenic scores accessible to nearly all following direct-to-consumer genetic testing. My molecular geneticist partner on hand was Dean Hamer, already celebrated for having discovered the “gay gene” (more on that later). Little did I know what I was getting into, and as they say, my 15 min went by so fast. Seated on my left was my “boss” Francis Collins. On my right was Justice Stephen Breyer. Other celebrities included journalist Gwen Ifill, who in 2020 was recognized with her own US Forever stamp. Representing the defendant was Johnny Cochran. The case was a puzzle. Can genes predict behavior? Should predictive tests be used and if so, how? What about genetic tests that could stigmatize groups of people? If a judge rules that evidence can be introduced that a gene influences criminal behavior, would identification of this genetic link in a chain of causality influence our willingness to convict or modify the penalty? Several people including Dean Hamer suggested that I write a book about genes and behavior because of the new
Our Genes, Our Choices https://doi.org/10.1016/B978-0-443-22161-3.00019-3
7
Copyright © 2024 David Goldman. Published by Elsevier Inc. All rights reserved.
8
2. The jinn in the genome
implications of neurogenetics for understanding the origins of human behavior. However, at that time (and while Dean was writing another) I was grappling with the puzzles of determinism and free will to which genetic prediction and inborn genetic determinants are bound. If I had not made up my own mind, what business did I have trying to change someone else’s? Eight years later, I attempted a new synthesis on human choice, based on a concept of neurogenetically determined behavioral individuality, and the result was the first edition of Our Genes, Our Choices.
Some famous geneticists and why they are famous As has been well chronicled, the draft sequence of the human genome was published in 2001 by two rival groups, a corporation led by Craig Venter and a government consortium led by Francis Collins, who codiscovered the cystic fibrosis gene and at that time directed the Human Genome Institute (Lander et al., 2001; Venter and Adams, 2001). Later, Collins became NIH director, stepping down only recently. Because the race for the sequence ended in a virtual dead heat, it was appropriate that both were honored by President Clinton in a Rose Garden ceremony. However, as Collins observed, completion of the draft sequence was only the end of the beginning, bringing us to the starting line of a much longer race to understand how the genome works and to prevent and cure diseases. The advances keep coming. A decade later, Venter synthesized the complete genome of a bacterium (Gibson, 2010). This in-laboratory duplication of nature underlined the fact that all life on Earth, and its complex variations, is ultimately based on the expression of complex chemicals: DNA and RNA, and those chemicals are increasingly open to measurement and manipulation. Beyond its biomedical applications, knowledge of the human genome has indeed, as I speculated back in 2010, enabled us to answer some of the most fundamental questions as to what humanity is and to what peaks it might lift itself, by its own bootstraps. If genetics can never provide all the answers, it can at least do what science does best, which is to pose better questions, and new questions. Also, as will be discussed in some detail, genetics properly applied has a nearly unique ability to identify causal connections, and not just correlations, between the molecular level of the DNA code and people’s most complex attributes, including their behavior. Furthermore, and as would have been difficult to predict even a decade ago, this genomic revolution has been melded with revolutions in ability to capture epigenetic changes in the molecules of life and to image dynamic changes in brain activity such that the nature of thought is beginning to be understood. It is harder to deny that a clock is a mechanism if we take the back off and see its working levers, ratchets, and wheels.
The jinns of knowledge and technology
9
James Watson, who with Francis Crick, won the Nobel Prize in Physiology and Medicine for deciphering the structure of DNA (Watson, 2001; Watson and Crick, 1953), conceived the Human Genome Project. Watson recognized the critical importance of genethics, and as director of the Human Genome Institute set aside a fixed percentage of its funding for the ethical and policy implications of genomics research. The ELSI (Ethical, Legal, and Social Implications) program had a lasting and pervasive impact on the thinking of human geneticists, in part because many, including me, participated in symposia it sponsored. From the beginnings of human genomics research, scientists grappled with the societal implications with some important practical results including the passage of GINA, a federal law that makes it more difficult to discriminate based on genetic information, as discussed in more depth in Chapter 9. However, it is also fair to say that the impact of genetic knowledge and the powerful new tools that implement it are constantly presenting new challenges and such that genethics remains a work in progress. Only in the past decade has CRISPR technology been invented and used to gene edit human embryos. Only in the past decade has direct-to-consumer personal genomics and genealogy exploded and led to the ancestry and phenotypic profiling of felons who happened to leave their DNA at a crime scene but who were never genetically tested themselves. As the leading edges of genomic knowledge and technology rapidly advance and science becomes more specialized, it becomes more difficult for generalists to make accurate assessments, and to some extent experts are being asked to predict the future. For example, will stem cells made from adult tissues suffice for transplantation medicine, and in what time frame and at what costs? The important advances in genetics are coming from scientists with very widely varying goals, perspectives, and backgrounds. Some are human geneticists, and some are not. Some are physicians and some are unfamiliar with medicine. Some have had direct involvement in the medicolegal side of genetics and others not. The public and the scientists advancing frontiers of knowledge will repeatedly be presented with new ethical puzzles created by new capabilities. Making wise use of the knowledge will require a continuing reevaluation and readjustment of mindset and the input of many voices.
The jinns of knowledge and technology Everyone is aware there has been a genome revolution, but all but a few visionaries and writers of science fiction underestimate the implications. By studying our own genomes—the genetic blueprints of our lives—humanity picked up a lamp and two powerful forces were released. They wait expectantly, and we should be very careful what we ask of them. Unlike “real” jinns, they are at large in the world and answer to no master.
10
2. The jinn in the genome
Genomics knowledge and tools are available in any nation. We cannot assume that the decisions we make will determine how these tools are used elsewhere, but we can influence those decisions by word and deed. The jinn of knowledge is the linear sequence of the human genome discovered by the Human Genome Project (Lander et al., 2001; Venter and Adams, 2001), and increasingly, as Collins predicted, we have views of the genome’s three- and four-dimensional depth—its variation between individuals and populations, and its complex regulation including the unfolding of the DNA-encoded developmental programs that enable us to develop from a single cell into a complex organism. The jinn of technology is the ability to rapidly, accurately, and cheaply measure any genome and its functional outputs. Applying the new technology and the template of genomic knowledge, we shift our level of evaluation of personhood to the level of molecular predictors. Obviously, this molecular view of the person is particularly potent prenatally, postnatally, and in infancy, when the individual has had little opportunity to establish their own attributes. We cannot predict exactly how the infant will develop but at birth or prenatally we can obtain a blueprint that is increasingly accurate. It will be tempting to use it and for some purposes we must use it. In adolescents and adults, the genome sequence, as well as marks of the environment on the genome (the rapidly growing science of epigenetics which will be introduced), is also informative—it helps us to understand how the individual came to be what they are and can help to predict their future responses. It is the purpose of this book to explore the uses of this predictive power in behavioral genetics, for better or worse.
Revolutions in culture and evolution of genes We live in a progress-oriented society, and indeed civilizations that fail to embrace change and thereby languish technologically are likely to be consigned to the sidelines of history, if not physically or culturally obliterated by invaders with better guns, productivity or, in the Jared Diamond sense, sustainable systems. Recently, challenges of a global nature, namely global warming (aka climate change), emergent pandemics, and financial crises have emphasized the unity of humankind and challenged the idea that any society can remain neutral or uninvolved. The genomic revolution is of that type. The problem of genome technologies and other powerful and transformative tools is a dilemma of change. Other societies may embrace what one forgoes. It is therefore unsurprising that in more successful modern societies in which books such as this are written on electronic media there is collective agreement on an ideology of progress (consider the pompous corporate slogan, “Progress is our most important product”). The need to
Revolutions in culture and evolution of genes
11
embrace technology explains much of the behavior of contemporary nations, and the need to embrace technology explains much of the behavior of contemporary nations and their relative success. Well before an atomic bomb obliterated Hiroshima and a beeping Sputnik awoke Americans to the fact that they lagged in the space race, it was widely recognized that the most important natural resource of a nation was not oil or uranium, but its populace. How large was it, how well educated, and how well enabled to compete? The space race launched by Sputnik ultimately helped to end the Soviet Union, but as a nation rich in intellect, Russia is again rising, and one possibly beneficial side effect is that some of its bright intellects migrated around the world. The economic engine of China was unleashed by leaders raised under communism but savvy enough to comprehend that it was not working. As exemplified by their suppression of the internet and massive surveillance, the need of the Chinese to join in the global interchange of information, so well exemplified by the diaspora of Chinese people, will inevitably erode authoritarian limits on speech, activity, and—arguably—thought, in China but also abroad. Overall, scientific advances make us better off, but whereas change is inevitable the devil lurks within the details of application and laws of unintended effect. Things usually do not unfold exactly according to plan. However, even when they fear the consequences, scientists are drawn to experimentation, invention, and mastery. In Kurt Vonnegut’s Cat’s Cradle, ice-nine (a water crystal of high melting point) froze the world’s oceans (Vonnegut, 1963). In the real world, the first atomic bomb was detonated in a desert called Jornada del Muerto. Observing the unmatched force, Robert Oppenheimer said, “Now we have become Death, Destroyer of Worlds” (Bird and Sherwin, 2005). But does it really improve matters to destroy a desert or a world while making comments memorable only if anyone survives to hear them? Were those “just pretty words”? As the composer John Adams asked in Doctor Atomic, what were the physicists really thinking? For better or worse, Edward Teller, father of the superbomb, the hydrogen weapon, was certain of the correctness of his actions, but Oppenheimer was tortured by the potential for future catastrophe created by an unparalleled power in human hands. Fortunately, the first fission and fusion blasts did not trigger a thermonuclear chain reaction in Earth’s atmosphere, as some physicists feared. It led to death and suffering of thousands of civilians in two cities and to a new and terribly dangerous era that continues to threaten life on Earth. Oppenheimer was a tragic figure precisely because he foresaw some of the things to come, and even the possibility of Armageddon. Working in the realm of imagination, in Slaughterhouse 5, Vonnegut dreamed up the Tralfamadorians, an alien race gifted with transcendent foresight into the mists of the future (Vonnegut, 1969). However, the Tralfamadorians also had a fascination with the nature of matter and the design of an “ultimate
12
2. The jinn in the genome
weapon.” At a point in the future—now imminent—the project approaches completion but beyond that point they cannot see, possibly because their test destroyed the universe and there is nothing to see. Faced with an ultimate choice they do what it is their nature to do: build the weapon. Are humans like Tralfamadorians in their inability to resist an impulse to open the next and deadly dangerous, “black box”? It is fair to say that scientists spend their lives opening as many black boxes as they can get our hands on and are usually either not thinking about the consequences or rationalizing. There is a little Tralfamadorian in all of us and many of the activities of science and technology expose humanity and the world to dangers that are quite unexpected. We are not prescient. When fire was captured by humans, no one would have thought that the burning of hydrocarbons would 1 day alter the climate of the whole planet. When lead was added to ceramics, gasoline, and paints, no one expected that humanity would 1 day pay a price for its toxicity, and it took generations of exposure and observation to unravel the causal connection. Yet, scientific knowledge can be our greatest shield against consequences. Tralfamadorians, Oppenheimers, Tellers and their physics experiments aside, the danger is rarely the experiment itself, it is the technical application of the knowledge. It is science that enables us—to the limited extent and ability humans have—to peer into the future and understand the consequences of our actions. It is because of science that we understand the effects of heavy metals and toxins in the environment and have used that knowledge to save lives and improve lives. The developing brains of children are shielded from lead that would otherwise diminish their capacities. Similarly, those of us who are fortunate enough to live in societies where the knowledge is applied are routinely shielded from a multitude of hidden dangers ranging from the inorganic such as asbestos to the organic such as vitamin deficiencies, viruses, and bacteria, and when an emergent virus causes a pandemic, vaccines and treatments developed using new technologies and exhaustively tested in clinical trials can save millions of lives, as is happening in the COVID-19 pandemic, which is still unfolding as I write this second edition of OGOC. Science gave us the crystal ball that predicted the consequences of the release of fluorinated hydrocarbons into the atmosphere, and that knowledge enabled a successful international effort that has preserved the ozone layer, which is now recovering from its depletion. The crystal ball is often a bit cloudy, but it is because of science that we know that the massive release of carbon dioxide into the atmosphere is causing global warming. Many individuals who know—and who do not deny—are constantly working in large ways and small ways to at least mitigate the potentially catastrophic effects. As for me, I thought I would at least commute to work on a bicycle until some motorist pried my cold, dead fingers from
Revolutions in culture and evolution of genes
13
the handlebar, but then the pandemic pretty much ended commuting to work for three years. Without science, global warming, pandemics, and other catastrophes are threats that are invisible as they are potent, but with science we can see that they are in our possible futures and can make rational choices to avoid them. However, there is a mismatch between progress in genetic and other scientific knowledge and technological progress. This lag between technological progress and our knowledge, or the ability to integrate and use that knowledge—which is the essence of wisdom—endangers us. We become aware that knowledge is unequally distributed, and it is frustrating when others seem to be unaware of important facts. Then we become aware that the problem is not just these “other people,” with whom we disagree politically. We have begun to lag the knowledge curve in many important areas. A parent may appreciate this when watching the facility with which their children grasp the latest technical modality. The new generation is always better able to adapt, but ultimately as a species, how well do our genomes match up to the new challenges? Can our brains and scientifically based understanding enable us to properly use, balance, and adapt to the inventions and consequences of our science? Perhaps not. Our handheld electronics and fiber-optic cables serve us better than stone flints and steel. However, the demands that the world we call modern places on humans are arguably not only different but more difficult than the demands that were presented in previous millennia. Humankind’s first steps up the cultural ladder required genetic changes enabling language and complex social behavior, a process that took millions of years. In the past 60,000 years, which is almost the last instant of time in human evolution, technological progress has advanced exponentially and become decoupled from genotype. The technological revolutions—agricultural, industrial, atomic, computer, communications, internet, genetic, nanotech, and biotech—erupt at shorter and shorter intervals, as shown in Fig. 2.1. As the pace of cultural evolution quickens such that many people have experienced seven technological revolutions within their lifetimes and, to a considerable extent, as the nature of the technological “race” changes, many fall behind. Evolutionarily, the genome is slower to respond. Email was a wonderful tool until we discovered that it turns “work” (whatever that is) into a 24-hour-a-day proposition. Cell phones are handy and continuous videoconferencing is now more the rule than the exception, but are not there sometimes when each of our lives would be improved by an interval of silence, for perhaps the space of half an hour, or again seeing others face to face and not behind a mask? Life is a continuing education. The Red Queen, a role my technologically advanced niece Hannah performed, summed up the challenge,
14
2. The jinn in the genome
FIG. 2.1 Genetic evolution and cultural revolutions.
“In Looking Glass World you have to run as fast as you can, just to stay in the same place” (Carroll, 1871). Our genomes cannot keep up. However, our brains and our networked brainpower can. By knowing ourselves and the world better, and integrating that knowledge, humanity can make wiser choices. “Wiser” means decisions informed by science and knowledge of the world and humanity itself. “Choice” means free decisions made by individuals and individuals acting in groups. This book, as a tour of behavioral genetics—where has it gotten us and where is it going?—is an attempt to contribute to dialogue on the question of what we are and how can we choose, and indeed whether humans can choose. At a deeper level it advances a new concept of neurogenetic individuality, top-down guided neural plasticity, and free will. It is my belief that an understanding of the origins of behavior—our individuality—opens a passage to a better relationship between persons and society. The message undercurrent is that people can maximize their own potential and more adroitly surf new technological waves, and perhaps avoid a few of the most dangerous, using the genomes we have and not necessarily the genomes we might want. As we learn more about our inborn genetic determinants, we can use that knowledge rather than feeling constrained by it, or even, doomed. The message is that we are not motes in a stream or interchangeable parts in a social machine, but individuals with capability, and therefore responsibility, to choose the type of world we would like to live in and pass on to our children.
The neurogenetics of determinism and freedom
15
Genes, brain, and individuality This book mainly explores the phenomenon of genetic variation and its effects on the brain. The sequencing of the human genome and dramatic technological innovations are naturally leading to the identification of genes that encode variation in behavior and cognition. These advances include the first keys to the inherited origins of psychiatric diseases; other behaviors that may or may not be defined as clinical disorders, depending on which generation of diagnostic categories is in use; and predictors of behavior. Even the politics of racial identity has been battered as people have come to understand that they have a unique ancestry and combination of culture and experience. This genetically and neurodevelopmentally informed perspective leads to a new perception of personal identity that transcends simplistic profiling. The new conception of individual identity, so strongly exemplified by modern politicians, entertainers, athletes, artists, businesspeople, and scientists who transcend crudely formed and often cruelly applied classifications, is powerfully based on new findings on genetic similarities. We have learned of the close genetic affinity each person has to all other humans anywhere, with only a small fraction of our heritage of genetic variation being assignable to differences between populations or races. However, our genetic variation, including at least 22 million relatively common genetic variants (polymorphisms), combines in myriad ways to make each newborn something the world has never seen before. Even genetically identical twins, despite their resemblance for physical and behavioral characteristics, are truly unique at all levels from molecules to cell, body, and behavior, and this is true because of the stochastic (random) nature of the processes by which development unfolds.
The neurogenetics of determinism and freedom The identification of neurogenetic determinants of behavior suggests a new opportunity to interpret the role of volition or “choice”: the key question being, “Do we have free will?” versus the more modestly framed and sometimes, but not always, practical, “Does it make sense to treat us as if we are free?” We will take on the question of whether genes can predict criminal behavior, and the answer is yes but of course it is more complicated. We will differentiate the ability to predict a thing from the origins of the thing. We will discuss the genetic origins of sexual behavior. We will see that genes can strongly influence resilience, emotionality, and cognition, and that when views of the activity and chemistry of the brain are obtained these gene effects are more profound. Although the genetic revolution is
16
2. The jinn in the genome
changing our understanding of all aspects of the human condition, the focus of this book is the intersection of genetic individuality and the origins and individuality of behavior, a concept that I label “neurogenetic individuality.” Nearly every complex human behavior: personality, intelligence, criminal and antisocial behaviors, sexual behavior, suicide, addiction, depression, and anxiety disorders.
References Bird, K., Sherwin, M.J., 2005. American Prometheus: The Triumph and Tragedy of J. Robert Oppenheimer. Participants: Breyer, S., Ifill, G., Cochran Jr., J., McGowan, A., King, P., Goldman, D., Collins, F., Hamer, D., deGenova, J., Allred, G., 2002. Genes on Trial. Genetics, Behavior, and the Law. A Fred Friendly Seminar, PBS. Moderated by Charles Ogletree,. Carroll, L., 1871. Through the Looking Glass. Macmillan. Gibson, D.G., et al., 2010. Creation of a bacterial cell controlled by a chemically synthesized genome. Science 329, 52–56. https://doi.org/10.1126/science.1190719. Lander, E.S., Linton, L.M., Birren, B., et al., 2001. International Human Genome Sequencing Consortium. Initial sequencing and analysis of the human genome. Nature 409, 860–921. Venter, J.C., Adams, M.D., et al., 2001. The sequence of the human genome. Science 291, 1304–1351. Vonnegut, K., 1963. Cat’s Cradle. Vonnegut, K., 1969. Slaughterhouse-Five or The Children’s Crusade. Watson, J.D., 2001. The Double Helix: A Personal Account of the Discovery of the Structure of DNA. Watson, J.D., Crick, F., 1953. The molecular structure of nucleic acids: a structure for deoxyribose nucleic acid. Nature 171, 737–738.
3 2B or not 2B? Anecdata and data The foundations of science are monuments and rubble of past achievement. When the science of today is viewed from the perspective of the future it will be seen as small—“little did they know”—and ramshackle, with critical gaps and errors in architecture. One goal of modern genomics is to move from the gapped and piecemeal to the complete by studying the whole of the structure, variation, and expression of the genome. This approach has paid great dividends, but also grapples with ultimate limitations because the levels of human genetic and epigenetic variation, and their interactions, are nearly unlimited. Science advances by serendipity and leaps of insight, but most discoveries are constructed piece by piece from many dependable building blocks. How narratives of discovery are told is a more a matter of style and strategy. Step-by-step approaches to story-reading, as well as storytelling, are perfectly good for they who are patient to wait, but as we will discuss, not everyone is. Innate predisposition becomes cognitive style, some people flipping through books to the ending, to see if the exercise leads anywhere interesting. I assay books via quick read of random pages, avoiding the ending because it is more fun to guess the murderer if it then takes Holmes a few hundred pages to deduce the snake in the victim’s bed was put there by the Indian snake handler, not the parlor maid. If a gun hangs over the fireplace in scene one there is no doubt as to whether it will be used, but by whom and why? Like a detective procedural, laboratory science is a step-by-step performance of well-known recipes. Whereas serendipitous discoveries are occasionally made when something is done out of sequence or backward, most deviations are punished. We can list many obstacles to the discovery of genes “for” behavior, but lately the off-the-shelf procedures, including genome sequencing and genome wide association, have transformed the ability to find genes altering behavior. Hundreds of gene locations have been mapped by the brute force casting a genomic net across very large datasets. The identification of genes for complex behaviors has mooted the discussion from whether we can
Our Genes, Our Choices https://doi.org/10.1016/B978-0-443-22161-3.00025-9
17
Copyright © 2024 David Goldman. Published by Elsevier Inc. All rights reserved.
18
3. 2B or not 2B?
discover genes “for” behavior. The remaining questions are many and diverse—how many genes, their molecular mechanisms, natural roles, specificities for behaviors vs pleiotropy, gene-by-environment interactions, developmental trajectories, and gene-by-gene interactions. Much of this book concerns the implications of genes for behavior that have already been detected in these early hours of a scientific revolution, discoveries that enhance the case for genetic determinism. Although most of the inheritance of behavior remains unassigned to any specific functional locus, it is remarkable to see the way knowledge and expectations have changed since the first edition of this book. At that time, and before GWAS, Kenneth Kendler, who laid foundations of our understanding of the inheritance of psychiatric diseases, stated a set of principles which implied that we should not push forward with naïve strategies or, more seriously, that we face ultimate obstacles in identifying genes for behavior (Kendler, 2012). Prominent neuropsychiatrist Danny Weinberger once remarked to me that after listening to Ken’s ten “thou shalt nots” he felt like an Israelite admonished for having worshipped a Golden Calf. The problems Ken divined plausibly formed an intractably Gordian knot. However, as Orwell sardonically said, “ignorance is strength.” Confounding many experts, GWAS, the analysis of crudely measured traits in very large datasets, worked. Originally attempted only by the wisest or most foolish, GWAS proved over and again to be a powerful tool to connect the most complex and crudely measured behavioral traits to individual genetic markers, and—if GWAS fulfills its promise—it will lead to identification of functional gene variants that drive traits and representing a triumph of genetic reductionism. This chapter introduces a gene that can predispose people to commit murder. It is only weakly predictive, but increasingly, genes of weak effect are being combined into polygenic scores that are much more strongly predictive. Polygenic scores, although still nascent, can already be computed for traits such as risk-taking and addiction, casting a spell of genetic determinism or at least guilt-mitigating predisposition on the dialogue. Much of this book is devoted to how it can be true that there are genes and polygenic scores (PGS) “for” behavior and wherein locus of responsibility is preserved beyond genes and environmental factors “beyond our control.” The dilemmas posed by genetic prediction deepen as predictors become more informative, and technologies including prenatal testing make them more actionable. As demonstrated by recent studies on the use of polygenic scores for phenotypes such as height and body mass index, genetic markers can predict as much as half of human variability for traits such as height and general cognitive ability. As discovered in the UK Biobank study, a person in the top decile for body mass index has about a 7% chance of morbid obesity (BMI >40) but persons in the bottom
Anecdata and data
19
decile for BMI PGS seldom become morbidly obese. Regardless of their BMI PGS, people make choices about diet and lifestyle, but even for a trait as causally complex as body mass, genotype clearly alters their metabolic reaction range and their choices. The predictive value of genes can be combined with other measures, and for example studies have shown that people with genetic liability for breast and ovarian cancer, prostate cancer, and colon cancer, as indicated by different PGS, can benefit by being more intensively screened with other methods. In this way a person’s genes, if they know them, can guide choices in medical care. For example, in 2013, actress Angelina Jolie underwent prophylactic double mastectomy upon learning that she was at elevated risk for breast cancer due to a mutation in BRCA1, a gene critical in DNA repair. Yearly, and due to the “Angelina Jolie effect,” thousands of women, many with a family history of breast cancer, undergo genetic testing that can include not only BRCA1 and BRCA2 (a cousin of BRCA1) but also the CHEK2, PALB2, PTEN, and TP53 genes, as well as many others that can be combined into gene scores and PGS. As PGS and genetic prediction are increasingly used, it has also been learned that PGS developed in one population are usually not as predictive for another. This is because some causal genetic variants differ strongly in frequency across populations, as is known for breast cancer. One in 40 Ashkenazi Jews carry one of three founder mutations in BRCA1 and BRCA2, and women carrying one of these variants have a lifetime risk of breast cancer of >70% and are also predisposed to other cancers, including of the ovaries. In formal terms, genetic variants that are far more abundant in certain populations are said to have high FST values (a statistic measuring cross-population variance). Among the 25,000 genes, each population carries genetic risk variants but at higher and lower abundances. In sum total, the impact on health may be similar, refuting the myth of racial superiority, but the profiles of vulnerability vary for one disease or another, and for example fair-skinned people are at higher risk for melanoma, people of Northern Europeans ancestry are also at higher risk for cystic fibrosis and phenylketonuria, and people of West African ancestry are at higher risk for sickle cell anemia. Correspondingly, the genetic tools necessary to assess vulnerability can differ widely across people with different ancestry. Equal access to tools for making health care choices would demand that these gaps in knowledge of genetic vulnerability factors be addressed, and that ancestry (a genetic construct capturing genes) as well as ethnicity (a social construct capturing culture and environment) be evaluated. What happens if a gene for impulsivity, enhanced arousal, and irritability is discovered, one that can cause people to commit murder, and is present in more than 100,000 people of a particular ancestry? Although that answer may change in coming years, so far not much. My laboratory
20
3. 2B or not 2B?
found such a gene: a functional knockout of HTR2B, a receptor for the neurotransmitter serotonin. As will be seen, this dramatic discovery had many precursors, including common genetic variants with more modest effects on molecular function and behavior. From the genetic association and in vivo (in the body) functional studies on those common variants, we anticipated that much inherited behavioral variation was also attributable to variants that are uncommon or population specific. There must be many rare variants that affect behavior and that are found only in individual patients and families, and two will be discussed. However, until 2009, it was usually impractical to find the rare variants that might exist in only one person or in their immediate family. Only a decade earlier, the first human genomes were sequenced at a cost of hundreds of millions of dollars. At the time, genomes were sequenced one cloned DNA fragment at a time, and with the help of banks of sequencing machines and robots for cloning DNA fragments. Today, and using technology for so-called massively parallel DNA sequencing, millions or billions of DNA fragments can be simultaneously sequenced, a person’s complete sequence delivered within a day or two. The sequencing can even be performed on a handheld device using unprocessed DNA. The problem of reading a very long DNA “book” is overcome by cutting it into many small pieces that can be read very quickly, and at the same time, and—more recently—by reading the sequences of very long pieces of DNA. When unordered fragments of DNA are sequenced, the reads are stitched together (assembled) or mapped back to the canonical human genome, which is about three billion DNA bases in length. The sequencing is often performed with high redundancy, with each DNA base being sequenced as many as 50–100 times so that the sequence variations that a person carries can be reliably detected. The technology is so powerful that a typical student produces more data in their first few days than I did in years as a postdoctoral fellow. Furthermore—and as may in the long run damage the field—most geneticists are now reliant on data generated and assembled by others, just as few readers have assembled the printed or electronic copy of a book they might consult. For the relatively modest cost of sequencing a genome, what might an accused murderer interested in shifting locus of responsibility get? Potentially, their life. People have 25,000 genes, which provides ample opportunity to discover DNA findings that at least look powerful. For example, recent sequencing studies have revealed that the average person of European ancestry carries up to 100 “stop codons,” but most are quite rare and their effects on disease risk are unknown. A stop codon can completely block the function of a gene, and in addition there are many other types of sequence variation that can be strongly predicted to alter molecular function and thus contribute to disease risk. In evaluating the effects of these gene variants, we must look beyond their strong effects on
A common stop codon causing impulsivity and hyperarousal
21
molecular function to their effects on behavior, which are usually much weaker and modified by other factors.
A common stop codon causing impulsivity and hyperarousal My laboratory applied the power of massively parallel sequencing to convicted violent felons who were inpatients at a forensic psychiatry unit headed by Matti Virkkunen at the University of Helsinki, Virkkunen having retired about a decade ago (Virkkunen et al., 1994). In Finland, many convicted offenders are assessed psychiatrically, and this work had yielded important findings about biological determinants of impulsive behavior. For example, Virkkunen and Markku Linnoila had found that elevated testosterone and low serotonin levels played important roles. The prisoners studied by Virkkunen were, as a group, not typical of convicted offenders in the USA. A larger number of them were impulsive, hyperaroused, and irritable and who for the most part committed assaults and arsons for no discernible purpose. Many senselessly killed friends and acquaintances, and often following some minor irritation. This type of behavior can be classified as intermittent explosive disorder (IED), except that in current diagnostic classifications the diagnosis of IED is excluded if in the context of alcoholism or antisocial personality disorder (ASPD). Almost all also had alcohol use disorder, and many had attempted suicide or even completed suicide later (Virkkunen et al., 1994). We teamed up to conduct the first systematic genetic study of people with this type of extreme dyscontrol behavior. We also worked with an American psychiatrist, LaVonne Brown, a world expert in psychiatric assessment and impulsive behavior, and whose work on how to define impulsivity will be discussed later. Virkkunen collected the DNA, extensive biochemical data, and psychiatric and behavioral histories from the criminals, their relatives, and “normal control” men who lived in the Helsinki area, where a large fraction of the Finnish population is found. Science really is too slow, and the “2B or not 2B” discovery was made more than 15 years after Markku’s life was taken by renal cancer, and furthermore the geneticist Leena Peltonen was a posthumous coauthor on the paper in Nature, Peltonen having died just as the study was completed. Peltonen, Virkkunen, and Linnoila were all Finnish. It is frowned upon to stereotype a people by ethnicity, but for some reason it seems to work, which is a continuing rationale to measure ethnicity as well as ancestry, and much as it is important to measure gender as well as sex. Jaakko Lappalainen, who led some of the early work in my lab on genes and impulsivity, introduced me to the Finnish mentality with a story about a man taken out drinking the night before his wedding. He got stone drunk. Following the night of revelry—perhaps the happiest night of his life—his friends rolled him up in a carpet, as if he was Cleopatra,
22
3. 2B or not 2B?
and delivered him to his bride’s doorstep, where in the morning she found him, dead. Very sad, but in a weird way amusing. Maybe the point is not so easily explained to someone not raised Finnish. Another Finnish postdoc, Jyrkki Vanakoski, who did his dissertation on the sauna, introduced me to an incredible story of soldierly perseverance, Väinö Linna’s Unknown Soldiers (Linna, 2016). Kaija Valkonen tested limits of equipment and patience by twisting control knobs further than mechanically possible. A tenant/landlord conflict ensued because cold-adapted Kaija left her bedroom windows wide open throughout the winter. Markku Koulu and Ulla-Marie Pesonen did great work in my lab and then greater work when they returned to the University of Turku and discovered a role for the neuropeptide Y gene in obesity. NPY is a crucial mediator of stress resilience, and, as discussed later, its genetic variation alters anxiety and emotion, and via a gene by environment interaction with stress. Clearly, gene-by-environment interaction has been at work over generations in Finland. Winter light, lichened granite, silent snow, cold waters, and solemn forest are distilled into Sibelius’ music and perhaps in some way explain the sauna, if not how so many Finns came to possess the HTR2B stop codon. Two friends are ice-fishing. After a week of silence except for creaking boots and cries of unknown forest things, one remarks, “It is cold today.” Replies the other, “Did we come to fish, or talk?” Stranger still, when Finns decide to break the rules, they tend to break them in unpredictable ways. Linnoila was a child television star, motorcyclist, and classical guitarist before he ever became a world-renowned biological psychiatrist. Such multitalented individuals can never be classified. We all know that ethnic stereotypes don’t work! Hidden in the genome of some of the Finns we began studying in the late 1980s was a different mystery: an ancient stop codon mutation in the serotonin receptor, HTR2B. It was found by Laura Bevilacqua, an Italian psychiatrist, now faculty at Mount Sinai School of Medicine. Helping train young postdoctoral fellows from around the world is the best part of my job—and the saddest when they leave. For example, I admit a special fondness for young neuroscientists from Italy: Chiara Mazzanti, Alessandro Rotondo, Francesca Ducci, Silvia Castrogiovanni, Nicoletta Galeotti, and not least Stefano Michelini the erstwhile soccer star, soccer coach, magnetic man, and author of an unusual novel: Sauna (Michelini, 2013). Sauna contains pages with the following mysteriously minimalistic text “ … … … … … … … … … … … … … … … … ” “ … … … … … … … … … … … … … … … ” “…………………………………………”. When I opened Stefano’s book at a random place somewhere near the middle, I was fascinated to see one of these “saunatas.” I can’t translate Italian, but here was something that resonated: pay attention to unstated thoughts and motives and enjoy the silence. Then measure it with brain
A common stop codon causing impulsivity and hyperarousal
23
scanners and DNA sequencers, neurons and DNA never being silent. Otherwise, all too often all we really know is “…” and what our imaginations leave us. Another lesson seemed to be to write the way one wanted, and perhaps within the artifact that is a book in which someone might find a purpose, even if not the one intended. Laura Bevilacqua’s DNA sequencing that led to the HTR2B stop codon required the talents and efforts of many people with advanced skills and special ability to solve molecular and computational problems. In such projects there is always a scientist, usually young, taking the risks and driving things forward but as is also usual, Laura had the help of a multinational team, this one including Qiaoping Yuan, who earned his PhD in Forestry in China and became an expert in bioinformatics while assembling the genome sequence of rice; Zhifeng Zhou, a brilliant molecular biologist from China with a wry sense of humor; and Longina Akhtar, who grew cells for functional studies. At Inserm, Stephane Doly and Luc Maroteaux did critical behavioral studies on the Htr2b knockout mouse. Genetics is an unusual domain of science because from time to time and now on an everyday basis via GWAS an incredible complexity of data can be boiled down to a reductionistic, explanatory result. In this case, the reductionistic finding was a stop codon. The location of the stop codon in the 5HT2B serotonin receptor is shown in Fig. 3.1 (Cravchik and Goldman, 2000). Each circle (or large dot) represents an amino acid building block of the receptor protein chain, which is folded so that it crosses the cell membrane seven times. When the stop codon is present this elegant architecture is guillotined after only 20 amino acids. Obviously, this small protein fragment is incapable of performing the normal functions of this serotonin receptor to bind serotonin at the cell membrane and initiate an intracellular signaling cascade. What we can learn from the picture is that the HTR2B mutation was likely to be severe, equivalent to a “knockout” of the gene. Neurotransmitter receptors of this type have an elegant structure honed by millions of years of evolution and consisting of an extracellular portion (at top), seven regions that span the cell membrane of the neuron (middle), and some intracellular regions that interact with signaling proteins (bottom). However, the stop codon terminates the protein after only a small fragment, corresponding to only part of the extracellular domain and omitting the rest of the receptor, is made. In males, the stop codon was more common in violent offenders and others with alcohol use disorder and several other psychiatric diagnoses marked by hyperarousal and irritability, namely borderline personality disorder, intermittent explosive disorder, and antisocial personality disorder. Most of the male stop codon carriers who we had identified in the criminal population had also attempted suicide. Characteristically, they committed their crimes while intoxicated, expressing their irritability, arousal, and impulsivity in a context in which top-down control of
24
3. 2B or not 2B?
Q20*
Truncated protein
FIG. 3.1 A population-specific HTR2B stop codon (allele frequency 1.5%) truncating and disabling the Serotonin 2B receptor, a seven-transmembrane domain G-protein coupled neurotransmitter receptor. Source: Bevilacqua, L., Doly, S., Kaprio, J., et al., 2010. Populationspecific HTR2B stop codon predisposes to severe impulsivity. Nature 468,1061–1066. Figure drawn after Cravchik, A., Goldman, D., 2000. Neurochemical individuality: genetic diversity among human dopamine and serotonin receptors and transporters. Arch. Gen. Psychiatry 57, 1105–1114.
ehavior is impaired and engaging in senseless violence. For example, they b might become angered and immediately tried to set fire to a nightclub full of people or choke someone to death with bare hands. The behavior of some of the men who carried the stop codon was as chilling as it was pointless. While drunk, one was choking a woman who had irritated him but instead strangled a man who intervened. Remorseful, he was imprisoned for several years, but following his release throttled to death another man. A remarkable result, which is probably due to a combination of causation, chance, and the unique genetic and social context of Finland, is that the three double murderers in our sample were carriers of the HTR2B stop codon. In the whole country of Finland, double murderers are rare. Yet, here were three unrelated double murderers who all carried the same severe genetic variant. Counterbalancing this picture developed by studying severe violent offenders and comparing them to controls, we directly detected the stop codon in 174 Finns (and since that time more than 1000), and learned
Validating an impulsivity gene in a mouse model
25
that people carrying this genetic variant are cognitively within the normal range (although there may be nuances) and most are nonimpulsive. By extrapolation, the stop codon is found in over 100,000 Finns, but almost all are “normal” and ultimately it is likely to be found that the stop codon influences arousal, rather than impulsivity or risk taking per se. In some Finnish criminals the stop codon was a necessary factor, but it was not sufficient to account for the behavior of any of them. In women, the effect of the HTR2B stop codon is different. Several had alcohol use disorder, but others were apparently normal, and none was known to have committed murder. In families in which the stop codon was transmitted transgenerationally it was associated with behavioral control problems, and again the exceptions were mainly women. Unlike the monoamine oxidase A stop codon mutation, which was found in only one Dutch family and will be discussed in a later chapter, we were immediately able to find the HTR2B stop codon in a series of families. The high frequency of the stop codon, and its restriction to Finns, reflects the unique nature of the Finnish population, which has genetic characteristics differentiating it from other European populations, and which has several disease-causing mutations that are either rare or completely absent in other populations. In fact, it was difficult to find the HTR2B stop codon in anyone whose DNA we did not collect in Finland. Laura Bevilacqua genotyped the DNA from several thousand other individuals from various populations distributed all around the world. Finally, she announced she had found an American who carried it. Perhaps it was not, after all, unique to Finns. Our collaborator, Emil Coccaro, at the University of Chicago, had more information. She had alcohol use disorder, and her ancestry was Finnish. Sequencing of hundreds of thousands of people worldwide has verified that *20 is a Finnish variant. South Asians have a different, but rarer HTR2B stop codon, but other rare instances among Europeans are attributable to the few individuals of Finnish ancestry in these populations.
Validating an impulsivity gene in a mouse model An obvious problem with human behavioral genetics, and the human genetic study that enabled us to discover the HTR2B stop codon, is that scientists cannot readily perform controlled experiments. Human geneticists study the individuals, family constellations, and populations that exist and usually cannot manipulate environmental exposures. People who have the stop codon and who as a result are impulsive might for several reasons be underrepresented in epidemiological samples. The criminals could have declined to take part in our study, but fortunately few did. As we will discuss at more length, the Finnish men we studied choose their environments or have them thrust upon them, and their e nvironments d iffer in important
26
3. 2B or not 2B?
ways. Also, if we are interested in the action of a genetic variant we must be concerned with the effects of genetic background. What are the other 25,000 genes doing? On the other hand, what we care most deeply about may be the human behavior, and the genetic variant that alters behavior in the human might be rare or nonexistent in an animal model. Fortunately for the HTR2B stop codon story, we were able to evaluate a relevant animal model, and thereby show that the discovery we had made in people had predictive validity in another species where we could control both environment and genetic background. Luc Maroteaux, a scientist who has long studied Htr2b in rodents, had used recombinant DNA methods to target and disrupt the gene in mice. In these mice, other genetic and environmental factors are held constant, but the Htr2b gene has been knocked out. Key to the study of such an animal model is that impulsive behavior, novelty seeking, and responses to novelty are measurable and correspond to human behavior. Another advantage is that we did not have to worry about under ascertainment of mice that happened to be impulsive, because the mice, in contrast to people, could not wander off or decline to take part. We will discuss the definition and measurement of impulsivity and arousal in people in more depth later. For now, it is sufficient to say that impulsivity of these mice was measured experimentally and, in several ways, as illustrated by the “2B” and “not 2B” mice in the yin-yang clock face shown in Fig. 3.2, and credit for which goes to Luc and his lab. The patient
FIG. 3.2 Impulsivity in Htr2b +/+ vs Htr2b −/− mice. Figure courtesy of Arnauld Belmer
References 27
“2B” mouse waits, biding its time in exploring a new environment, or to go to a more dangerous open area, even to seek food when it is hungry. The Htr2b +/+ (normal) mouse takes time to investigate a new thing placed in its box. Although famished, the “2B” mouse cautiously bides its time at the walls and in the corners before venturing to the center where it knows it is likely to find a food pellet, but where it is also more exposed. In contrast, the impulsive “not 2B” Htr2b −/− mouse is less inhibited by fear, intolerant to delay. It is a risk-taker. When placed in a novel environment it rapidly explores it. When a novel object is placed in its box it more readily touches it. Perhaps it is something tasty! If it is hungry, it does not wait long before sallying into open spaces that mice fear and, at least this time, it finds a food pellet waiting. Later we will discuss that the “not 2B” mouse is different in how it discounts reward vs the amount of time it has to wait for a larger payoff. The “not 2B” mouse and the people who carry the stop codon are not stupid but too impatient to wait a long time even when the delayed reward would ultimately be greater. What does the discovery of genetic variations with a role in impulsive behaviors mean, and what does it mean that we can even create a controlled model in the mouse that validates genetic effects observed in people? First, that the question of whether impulsivity and its resulting deleterious behaviors are in part inherited is settled. In retrospect, it could not be otherwise: behavior is the product of brain, and the building blocks of the brain are subject to genetic variation. We are all neurochemically individual. We have learned something about the nature of the inherited variations that alter impulsivity. With the genes discovered we can state that the genes that make some people susceptible to behavioral dyscontrol work in different ways. Looking forward, we will see that some variants, such as Han Brunner’s MAOA stop codon, are rare—the vulnerability genotype being found in only one Dutch family. Others, such as a different variant at MAOA, are common—someone in every gathering of people probably has the vulnerability genotype. Some, such as the HTR2B stop codon, are found in multiple families, but are restricted or relatively restricted to only one population. Several of the variants, and perhaps all, are strongly dependent in their action on contexts, for example being male, having a high testosterone level, and having experienced an early life stress. The genetic variant may be a necessary part of the behavioral syndrome, but alone it is not sufficient.
References Cravchik, A., Goldman, D., 2000. Neurochemical individuality: genetic diversity among human dopamine and serotonin receptors and transporters. Arch. Gen. Psychiatry 57, 1105–1114.
28
3. 2B or not 2B?
Kendler, K.S., 2012. Levels of explanation in psychiatric and substance use disorders: implications for the development of an etiologically based nosology. Mol. Psychiatry 17 (1), 11–21. https://doi.org/10.1038/mp.2011.70. Linna, V., 2016. Unknown Soldiers – International Edition. Penguin Press. Michelini, S., 2013. Sauna. Stefano Michelini. Virkkunen, M., Rawlings, R., Tokola, R., et al., 1994. CSF biochemistries, glucose metabolism, and diurnal activity rhythms in alcoholic, violent offenders, fire setters, and healthy volunteers. Arch. Gen. Psychiatry 51, 20–27.
4 Stephen Mobley and his X chromosome The line separating good and evil passes not through states, nor between political parties either—but right through every human heart. Alexander Solzhenitsyn
The death of Stephen Mobley We have seen that genes that affect behaviors in general and impulsivity and arousal in particular are unlikely to work in isolation. As discussed, the HTR2B stop codon does not have that type of unilateral effect on behavior. Despite occasional overenthusiastic reporting of gene effects on behavior, there is consensus that behavioral causation is multidimensional. However, it has occurred to defense attorneys that juries ought to hear about any gene that could predispose their client to commit a crime. It has also occurred to prosecuting attorneys that they should do everything possible to exclude that evidence. A courtroom is a laboratory where we can see what happens when forces collide, and within the constraints of precedent and legality. This debate is likely to intensify as we move from the single gene era to polygenic scores and gene scores informative for traits such as risk-taking. Stephen Mobley died when he was 39 years old, at 8 pm, on March 1, 2005, at a state prison located in Butts County, Georgia, south of Atlanta. He had been injected with a lethal intravenous dose 8 minutes earlier. Minutes before that, he had been granted a stay of execution, but it was soon withdrawn. Mobley was a paradoxical figure and was made into a symbol even though he was not well suited to the role. In his final statement, he expressed thanks for the opportunity to atone for his sins. Then his executioners in an adjacent room injected Mobley with a combination of sodium pentothal, to put him to sleep; pancuronium, to stop his breathing; and potassium chloride, to stop his heart. Outside the prison, a small group protested, and vigils were held statewide. Laura Moye of Amnesty International was one of the few to attend the execution. Afterward she said, “We think the state should not be in the
Our Genes, Our Choices https://doi.org/10.1016/B978-0-443-22161-3.00002-8
29
Copyright © 2024 David Goldman. Published by Elsevier Inc. All rights reserved.
30
4. Stephen Mobley and his X chromosome
business of killing people to show that killing people is wrong.” Friends and relatives of the victim did not witness Mobley’s death and were not reported to have held vigils. However, the victim’s family apparently supported commutation of Mobley’s penalty to a life sentence. Mobley had committed murder 14 years earlier while during a threeweek spree of armed robberies of restaurants and dry cleaners. While robbing a Domino’s pizza about an hour’s drive northeast of Atlanta, he shot the night manager in the head. This man, who Mobley found alone in the restaurant, was a 24-year-old college student, John Collins. A few weeks later Mobley was caught following a high-speed chase through Atlanta after another robbery. He was still armed with the same gun. Mobley is said to have confessed to a prison guard that he ordered Collins to his knees, made him beg for his life and, after his victim began crying, killed him execution style with a shot to the back of the head. As described by a journalist, Mobley was funny, charming, and affable, even in the minutes prior to his execution.
The Kallikak effect Mobley was born into a prosperous family and was represented by a former Attorney General of Georgia and a former county District Attorney. His lawyers pointed to the long history of criminal behavior in the Mobley family, claiming that his genes made it inevitable that he would do the same. The Mobley family allegedly resembled the bad side of the legendary Kallikak family, whose pedigree was featured in genetic textbooks of the first half of the twentieth century. The Kallikak family tree had two branches, the founding father being a Revolutionary War soldier who dallied with a barmaid but later married a Puritan girl. Supposedly, the descendants of the barmaid were uniformly morally degenerate and cognitively defective drunkards and thieves. That side of the family tree was depicted with dark twisted limbs and toadstools at the roots. On the other hand, the descendants of the Puritan wife were chaste and upstanding. That side of the tree had smooth limbs and flowers in bloom at the base. Was Mobley’s Kallikak-like family history mitigating? As Debbie Denno, a professor at Fordham, pointed out in her analysis of the Mobley case (Denno, 1996), being a regular churchgoer or loved by one’s family may be admissible evidence. However, it has also been reported that Mobley had “Domino” tattooed to his back and hung a Domino pizza box in his prison cell. Should inherited predisposition be off the table, whether as a mitigating factor or perhaps even to make the case for guilt or irredeemability? In a case where genetic testing for Arizona death row inmate Jeffrey Landrigan
Combining gene and hormone to predict impulsivity
31
was rejected, a lower court judge wrote, “The potential for future dangerousness inherent in Landrigan’s alleged genetic predisposition to violence would have negated its mitigating capacity for evoking compassion.” Also, an Idaho judge accepted evidence for the defendant’s “propensity to commit murder” and used it to help justify the death sentence.
Mobley demands a genetic test Mobley’s attorneys petitioned that Mobley be tested—at state e xpense—for a predictive variant in the monoamine oxidase A (MAOA) gene. Like the 5-hydroxytryptamine (serotonin) receptor 2B (HTR2B) stop codon we discovered in Finns, and which is found in over 100,000 of them, the MAOA stop codon is a severe genetic variation which blocks the function of a key protein involved in neurotransmitter function. It had recently been discovered by Han Brunner, in one Dutch family. Unlike the HTR2B stop codon, Brunner’s MAOA stop codon was exceedingly rare, if not absent, outside this one family. Various geneticists, including me, were contacted by defense attorneys to test their clients for the stop codon in MAOA (I declined). Meanwhile, my lab and others were investigating other impulsive individuals and families, including some of the same ones in whom we eventually discovered the HTR2B stop codon, for the presence of Brunner’s genetic variant, but not finding it. The judge rejected having Mobley’s MAOA genotype accepted as evidence. This decision made good genetic sense. The MAOA gene is on the X chromosome, which a male inherits from his mother, and the son has a 50% chance of inheriting the stop codon variant from his mother if she is carrying one copy of the variant. The MAOA stop codon variant appears to be highly penetrant: the males in the Dutch family who inherited the variant had a behavioral dyscontrol (impulsivity) syndrome. This leads to an X-linked recessive pattern of transmission in the family: carrier mothers, 50% of male offspring affected, other affected males on the maternal side of the pedigree. Instead of this pattern, Stephen Mobley’s family had multigenerational impulsivity with transmission from father to son and son to grandson, as well as violent females (Fig. 4.1). An X-linked stop codon was therefore one of the least likely causes of his behavior.
Combining gene and hormone to predict impulsivity The story of functional genetic variants at MAOA did not end with Brunner’s rare stop codon. Because MAOA will probably follow the same pattern of variation as other genes, it could have been predicted that eventually hundreds or even thousands of variants will eventually be found at
32
4. Stephen Mobley and his X chromosome .
.
. C C
T
C
.
.
C C T
T
T T C CC T
. CT
C
C C CT C C
Borderline mental retardation Dyscontrol behaviors: • Aggressive outbursts • Arson • Attempted rape • Exhibitionism
C
No fibroblast MAOA activity Abnormal monoamine metabolism: urinary HIAA, HVA, VMA urinary normetanephrine & tyramine Stephen Mobley, Convicted February 17, 1991 Executed March 1, 2005
FIG. 4.1 Brunner syndrome: X-linked dyscontrol in a Dutch family due to the monoamine oxidase A (MAOA) C936T stop codon (Brunner et al., 1993).
or near this same gene. Of course, most of these will be rare. Several years ago, a common functional polymorphism was discovered in the same MAOA gene that Brunner studied. The alteration in function was discovered by Dean Hamer, who will appear again later (in Chapter 15) as the discoverer of the “gay gene.” This new MAOA variant altered the expression of MAOA at the regulatory level: at the level of DNA transcription, the process by which the MAOA gene is expressed as an RNA transcript that will be translated into the enzyme protein, and altered brain neurocircuitry (Buckholtz et al., 2008). The polymorphism is of the variable number of tandem repeat (VNTR) type involving varying numbers of copies of a short DNA sequence found consecutively, or in tandem, like boxcars in line on a train. The MAOA VNTR has several common variants that differ in length owing to the number of repeats that are present. At least two of these length variants are associated with a lower level of MAOA expression and MAOA enzyme activity, and it is these relatively common lower activity MAOA variants that have been associated with impulsive behavior.
Carrying kohl to Italy Several years ago, the Italian Court of Appeals in Trieste reduced a convicted murderer’s sentence based upon MAOA VNTR genotype and some other genetic and brain imaging evidence of a type we will discuss later. Abdelmayout Bayout, a citizen of Algeria, had admitted stabbing and
The state of DNA in prediction of violence
33
illing Walter Felipe Novoa Perez in 2007 because the victim had insulted k Bayout over his kohl eye makeup. Although Bayout had already had his sentence reduced based on psychiatric illness, the appellate judge, Pier Valerio Reinotti, requested an independent evaluation. Two neuroscientists at the University of Padova evaluated Bayout and informed the judge that Mr. Bayout’s genotypes at MAOA and four other genes made him prone to violence following provocation. In his decision, Judge Reinotti stated that he was strongly compelled by the genotype evidence and reduced Bayout’s sentence by an additional year, to a total of eight years and two months. According to the judge, Mr. Bayout’s genes “would make him particularly aggressive in stressful situations.” As has been discussed, the genotypes had little predictive value. Furthermore, based on the earlier examples from Montana and Arizona, it is equally possible that another judge could have come to the same conclusion about the effect of MAOA on Mr. Bayout’s behavior, but therefore increased the sentence.
The state of DNA in prediction of violence The power of MAOA genotype, HTR2B genotype, or multiple other genotypes recently implicated in risky behaviors, and suchlike, to predict violence of an individual is low. There is no case law on polygenic predictors based on GWAS traits such as risk-taking, although police are already using genotype to construct physiognomic profiles of suspects, including height, weight, and color of eyes and hair. However, the number of cases in which DNA evidence has been introduced to demonstrate that a defendant was genetically predisposed to violence, depression, or addiction is rapidly accelerating. According to Nita Farahany, a bioethicist at Vanderbilt University, by a decade ago there were already more than 200 attempts to introduce mitigating DNA evidence in cases tried in the USA and a few were successful at the penalty phase of these trials. For many reasons that are a main preoccupation of this book, there is strong resistance to the use of such evidence, but as Bayout’s lawyer said, “My client is clearly an ill person and everything that allows the Judge to better evaluate the case and to decide the right sentence should be investigated.” Therefore it is worthwhile to discuss just how the common MAOA VNTR might be most accurately used in individual cases where behavioral prediction and explanation is so difficult, as contrasted with their use in large studies identifying group differences. The same procedure could be applied to polygenic scores. If one genotype is not highly predictive on its own, why not combine it with other information? To identify circumstances where MAOA genotype could help predict or understand one person’s behavior we should understand the circumstances where group behavioral predictions are powerful. The MAOA variants have
34
4. Stephen Mobley and his X chromosome
been repeatedly associated with impulsivity, but not in a general context. The first indications of behavioral prediction with MAOA have emerged only when the effect of the gene was measured in combination with other powerful predictors of behavioral problems. For impulsivity and aggression, two powerful predictors are testosterone and trauma, and we are still early in the process of understanding how such combined information can predict the behavior of an individual, as opposed to a group of individuals. The relationship of testosterone to impulsivity and aggression to testosterone is clear. Worldwide, and across cultures, men are far more likely than women to commit violent mayhem. The ratio of males to females on death row approaches 100 to 1. As a predictor of impulsivity and violence, the Y chromosome is an informative genetic marker, or to put it a better way, the absence of a Y chromosome is compelling. Women are statistically unlikely to commit violent crimes and when they do it is more likely that there was a specific motive, a psychosis, or a man involved. In most nonhuman primates, including our closest relatives such as chimpanzees, gorillas, and orangutans but also numerous Old World monkey species, males are also more aggressive. It appears that one important reason is the much higher testosterone levels of males. As can be seen on the left side of Fig. 4.2 (Sjöberg et al., 2008), higher testosterone in men correlates with Testosterone predicts aggression but MAOA genotype is permissive
Lifetime Aggression
Low acvity MAOA-VNTR
High acvity MAOA-VNTR
3
3
2
2
1
1
0
0 10
20
30
40
50
60
10
20
30
40
50
60
Cerebrospinal fluid testosterone (pg/mL) ASPD + AUD
AUD, no ASPD
Control
FIG. 4.2 In males with alcohol use disorder, antisocial personality disorder or neither (“controls”), higher testosterone levels in cerebrospinal fluid predict increased lifetime aggression (Brown–Goodwin Scale), but the lower activity (transcriptionally less active) MAOA VNTR genotype is permissive for this effect (left panel of figure). After Sjöberg, R., Ducci, F., Barr, C., Newman, T.K., Liliana Dell’Osso, L., Virkkunen, M., Goldman, D, 2008. A non- additive interaction of a functional MAO-A VNTR and testosterone predicts antisocial b ehavior. Neuropsychopharmacology 33, 425–430.
References 35
increased aggression, as measured by a lifetime aggression score called the Brown–Goodwin scale, a scale measuring impulsivity and aggression discussed more elsewhere. It should also be mentioned that one reason the relationship between testosterone and aggressive behavior is so strong in this figure is that the sample included highly impulsive and aggressive men. The interaction of stress with MAOA and other genes to lead to impulsive behavior and related traits such as hyperarousal and irritability is less obvious, but very powerful. As seen before, Francesca Ducci, Rickard Sjoberg, and I found that in Virkkunen’s impulsive criminal offenders, higher testosterone levels indeed predict higher levels of aggression, but this happened only in the men who have the common lower activity MAOA variant that is itself associated with aggression. The combination of high testosterone level and low MAOA activity genotype represents a “double whammy” gene-by-environment interaction. This gene-by-endocrine interaction is so far one of the few known, but as will be seen the discovery of interactions of genes with other factors, especially stress, represents one of the most salient accomplishments of genetics of complex behavior. Gene-by-stress interactions are the topic of a later chapter. Where will gene-by-sex, gene-by-endocrine, and geneby-stress interactions lead us? Fast-forwarding to the present, a defense attorney defending a murder case might make much better use of the MAOA gene, but he would have to work a little harder at it, and more than genotype might have to be measured. Furthermore, new molecular tools enable the measurement of the epigenetic impact of stress on the genome. It should be emphasized that these are group effects; however, early life stress exposure leads to genome-wide changes in DNA methylation, potentially indexing stress as experienced rather than an individual’s recall of stressful events. Seldom will it be sufficient to consider one factor. Finally, many diseases are influenced and defined by social context. Such diseases, ranging from addictions to obesity, are likely to not only be multifactorial in nature but in circular fashion are only fully comprehensible in social context.
References Brunner, H.G., Nelen, M., Breakefield, X.O., Ropers, H.H., van Oost, B.A., 1993. Abnormal behavior associated with a point mutation in the structural gene for monoamine oxidase A. Science 262, 578–580. Buckholtz, J.W., Callicott, J.H., Kolachana, B., Hariri, A.R., Goldberg, T.E., Genderson, M., Egan, M.F., Mattay, V.S., Weinberger, D.R., Meyer-Lindenberg, A., 2008. Genetic variation in MAOA modulates ventromedial prefrontal circuitry mediating individual differences in human personality. Mol. Psychiatry 13 (3), 313–324. Denno, D.W., 1996. Legal implications of genetics and crime research. In: Bock, G., Goode, J. (Eds.), Genetics of Criminal and Antisocial Behavior. John Wiley & Sons, pp. 248–264. Fordham Law Legal Studies Research Paper, 1996. Sjöberg, R.L., Ducci, F., Barr, C.S., Newman, T.K., Liliana Dell’Osso, L., Virkkunen, M., Goldman, D., 2008. A non-additive interaction of a functional MAO-A VNTR and testosterone predicts antisocial behavior. Neuropsychopharmacology 33, 425–430.
This page intentionally left blank
5 Dial multifactorial for murder: The intersection of genes and culture I recognize in thieves, traitors and murderers, in the ruthless and the cunning, a deep beauty—a sunken beauty. Jean Genet
A murder in the lab One way of introducing an idea is with a quotation. On the other hand, National Public Radio accompanies its features with snippets of music to suggest a mood. Another way to engage the reader is to begin a chapter with a story or parable. Instead of explaining quotations and in general using thousands of words to express what could be said with hundreds, perhaps they should be more self-revelatory, at the expense of adding self-referential paragraphs such as this one, to at least give the reader some better idea of what has gone on “behind the curtain.” I begin this chapter on the complex origins of murder and violent behavior with a brutal murder. Who is not at once repulsed and attracted by a diabolical act? However, there are many murders to choose between, providing the opportunity to select a case proving any theory of behavior. Perhaps the murderer was an irredeemable criminal unwisely paroled, perhaps he was an impulsive murderer with an unknown stop codon or unfavorable polygenic score, who then became a double murderer, or a victim of child abuse, or uneducated, or impoverished, or a racist, or a gun nut, or a religious fanatic or of the wrong political orientation. Scientists often claim that the plural of someone else’s anecdote is their data. The choice of the “anecdata”: the vignette, the quotation, or the mood music sets the stage and often overwhelms the impact of reasoning based on larger trends, statistics, or systematic thinking. Of course, any study of groups can also be biased, in multiple ways, but that is another story. About murder,
Our Genes, Our Choices https://doi.org/10.1016/B978-0-443-22161-3.00011-9
37
Copyright © 2024 David Goldman. Published by Elsevier Inc. All rights reserved.
38
5. Murder is multifactorial
Mark Kleiman wrote, in When Brute Force Fails: How to Have Less Crime and Less Punishment, that we go forward based on “big cases” to make “bad laws” (Kleiman, 2009). We are haunted by the face of a child who should not have died and who should not have died in some horrible way. That’s why laws are named after real victims, for example “Megan’s Law” and the “Brady Act.” A double murderer took the life of my close colleague, Dr. Michelle Filling-Katz. I hope that the telling of her story is taken as respectful, nongratuitous, and nonexploitative of her memory, memory being a legacy of Michelle and her husband Norman. Michelle and Norman were shot dead in their kitchen on August 11, 1992. Michelle, a pediatric neurologist, was a brilliant, dynamic scientist who let no obstacle stand in the way of her care for patients or her research. Michelle herself endured the ravages of autoimmune disease and giving her special insight into the suffering of her patients, many of whom had von Hippel–Lindau disease, a genetic disease whose manifestations include cancer of the kidney. At the time of her death, she was 36 years old and had joined my lab 2 years before. Michelle and Norman Katz had both been army physicians. Colonel Norman Katz was a Green Beret who served with the 82nd Airborne Division in Vietnam. He received two Bronze Stars for bravery and retired only in 1989 after having been Chief of Pediatric Ophthalmology at Walter Reed Army Medical Center. The murderer was Michelle’s stepson, Jayant Katz, then a 20-year-old architecture student who I briefly encountered a week or two before the murders. The only thing I noticed about him was that he was very concerned about his car. Jayant Katz had Manic-Depressive Illness, a disease which more frequently leads people to impulsively harm themselves, but as I will discuss in Chapter 7, sometimes leads people to harm others, even ones closest to them. What do we learn from such murders? Would Michelle and Norman be alive today if Jayant had not had access to a gun? Or should we take away lessons in family dynamics? Perhaps Jayant Katz was exposed to violent cultural influences. Perhaps the schools were at fault. Or should we understand that Jayant’s schizophrenia—a heritable disease—played the most important explanatory role? Surely his psychiatric illness played some major role. In the Maryland Circuit Court located in the county where I live, Jayant Katz pled guilty to two counts of second-degree murder. He said to the judge: “Yes. I killed my mom—it’s my stepmom actually—and my dad. Bang. The parent dead. A thousand voices fill my head.” Katz received a 15-year sentence and is now 50 yrs. old, the same age as his father.
Missing puzzle pieces, an obstacle to reductionism Psychiatric diseases can play a role in violence, and genes can do so either by contributing to such diseases or through their independent effects that cut across diagnoses or cocontribute to various psychiatric diagnoses.
Missing puzzle pieces, an obstacle to reductionism
39
Thus far, we have shown that functional variants of genes, monoamine oxidase A MAOA and serotonin receptor 2B HTR2B, can predict violence. The Y chromosome predicts violence. The interaction of testosterone with MAOA genotype predicts violence. Such particulate genetic findings can be useful for understanding individual behavioral differences and predisposition, but they do not constitute a reductionistic understanding of the problem of violence in societies. Psychiatric diseases can play a role in aggression, and genes can contribute to aggression both by predisposing to psychiatric diseases and through effects that occur independently of whether someone meets criteria for a psychiatric diagnosis, as Jayant Katz may have, and by effects that cut across diagnoses. For example, two personality disorders that are associated with hyperarousal and aggression are antisocial personality disorder and borderline personality disorder, and a large twin study conducted in Norway showed that these disorders are strongly cross-transmitted. Epidemiologically, both ASPD and BPD are strong risk factors for alcohol use disorder and substance use disorder, thereby further increasing the odds of aggressive behavior, including the self-directed aggression of suicide. Genomic studies are beginning to identify the polygenic origins of hyperarousal and aggression, and the specific inherited alleles that contribute to these disorders and that interact with critical environmental exposures. Presently (2022), Open Targets Genetics, a data resource for genes discovered in genome-wide association studies (GWAS), lists 71 genes that may contribute to “risky behavior.” At the time of the first edition of this book Open Targets Genetics did not exist and there were no genome-wide significant loci for risky behavior. However, most of the genetic puzzle pieces are still missing, and the functional (causal) variants are unknown at almost all the genes that have been implicated on a statistical basis. Murder is multifactorial. Even the most lethal genes are not highly predictive. Although they might be predictive if we fully understood the contexts in which they lead to violence, it is fair to say that we have only begun the process of understanding the interactions. It is unsurprising that we do not understand these interactions because we had identified enough of the genes with which environment interacts. Scientists with a primarily environmental perspective on the origins of behavior often demand of geneticists that they measure the environment. If only we knew how. Increasingly, geneticists are being shown how, and increasingly geneticists are showing scientists performing longitudinal and family studies how to incorporate genotype into their work. Also, and as will be seen later, when we take up gene-by-environment interaction, it is often not enough to measure the environment. One must study people who have experienced the relevant environmental exposures, for example measuring the epigenetic imprint of stress on the genome and not just the history of trauma itself.
40
5. Murder is multifactorial
A main problem for solving complex jigsaw puzzles of behavioral causation is that it is very much more difficult without all the pieces in hand. As a young man with a psychiatric illness, Jayant Katz killed his parents, but few psychiatrically ill people commit murder. Three of our HTR2B stop codon carriers became double murderers but most lived their lives with no indication that this genetic variation was present (Bevilacqua et al., 2010). Clearly, there are some pieces of the puzzle that elude our science. On several occasions, my daughter Evir (who is now a teacher) instructed our family in that simple lesson by holding back a piece from the jigsaw puzzles we assembled over winter break. As the piles of pieces shrank, her brothers would begin to suspect that there was again a piece missing and would know full well who was the culprit. Aaron and Ariel would shout as children do: “Mommy, Daddy, bad Evir has done it again! She hid the piece!” But no matter what methods of interrogation were applied Evir would not admit the crime. The children would do what scientists must do, which is solve the puzzle with a missing piece after which she would triumphantly place the last one—which was usually something rather essential, like the eye of the whale. Fortunately, scientists are clever people who enjoy the challenge of solving puzzles with many missing pieces, and pieces that may belong to other puzzles. Now that we have a few of the genetic pieces of what is a really complicated multidimensional puzzle of behavior it does appear that some progress is finally being made. Via GWAS, we are learning general information about the overall puzzle and the nature of its pieces—colors. For example, are the puzzle pieces (genes) ones that tend to be expressed in certain cells (such as different types of neurons and cells located in different brain regions) or at times of development? Are they involved in certain processes and for example neurotransmission or neuronal plasticity? Are genes involved in one disease also implicated in others, their “puzzle pieces” in part being shared? Combined, it is increasingly possible to predict complex phenotypes via polygenic scores. Nevertheless, the use of PGS in behavior and psychiatry is still nascent. Much work is needed to understand the stability and generalizability of PGS, and overall, more genes need to be identified to improve predictive power. This will require still larger GWAS datasets, but also the use of paradigms, such as DNA sequencing, that go beyond GWAS to capture effects of uncommon and rare genetic variants. It will require measurement of intermediate phenotypes including molecules and brain processes, to trace the identities of functional loci and their mechanisms. At this point, and using genotypes only as isolated predictors, the most informative genotypes and polygenic scores are simply not very predictive. As discussed earlier, the HTR2B stop codon is found in a large fraction of Finnish murderers, especially ones convicted of double homicide. However, it has a rather high allele frequency: greater than 1% and about one in 50 Finns carry this variant (Bevilacqua et al., 2010).
Why are some societies more violent?
41
As is worth saying again, we know that almost all of them are normal, and very few will ever murder anyone.
Why are some societies more violent? Just as we can understand the effects of genes without claiming to have found the gene predictor, we also do not have to wait for the full decoding of gene-by-environment interactions to find effects of environments that predict the behaviors of groups of people. Like the predictive values of individual genes, the predictive values of environmental factors are expected to be small and difficult to trace, and they often are. Why does one society have more murder and violent crime than another? Why, within societies, do murder rates substantially decrease, and occasionally increase, across generations? Genotype can play only a minor or nonexistent role in determining these transnational and secular variations. Although allele frequencies have not changed, poorly understood social changes can double or halve the murder rate. Countries whose peoples are of very similar genetic makeup have dramatically different murder rates. Volumes are written about the origins of violence, but they usually seem directed toward proving some theory or another and underestimate the role of biological factors such as the importance of being male, young, high in testosterone, and psychiatrically ill, and they do not consider the problem of genotype (Pinker, 2003). For murder, it is also important to come to terms with the fact that murder is rare on a national basis even if tragically common in some locales and demographic groups, and for example young men. The US has one of the world’s highest murder rates, but the rate is only about five per 100,000. However, all the predictors of murder are much more common than in many other countries. Guns are a predictor of murder but there are more than enough guns in the USA to arm every man, woman, child, infant, and household pet. Yet, few of these guns will ever be used to murder anyone. The conversation about the origins of murder should also be directed toward understanding the behavior of that small percentage of society that commits those acts, and where and under what circumstances to understand that behavior it will undoubtedly be necessary to piece together multiple pieces of a causal puzzle that differs from one murder or other violent act to the next. Therefore oversimplifications such as “guns cause murder” are as toxic as overexaggeration of the role of some genetic variant. However, it also illustrates why the genotype is important even though it will only rarely lead to the event. The effects of genes are intertwined with male sex, wealth disparity, trauma, sexual competition, drug use, psychiatric illness, and weapons on hand that may include vehicles and guns. None of the predictors will act that reliably and yet we should want to identify all the predictors and their interactions.
42
5. Murder is multifactorial
Decades of careful and often enlightened scholarship by criminologists and historians on the environmental origins of murder have not led to consensus. The theories are diverse. Pieter Spierenburg, a professor of historical criminology, in A History of Murder: Personal Violence in Europe from the Middle Ages to the Present, hypothesizes that murder rates are inversely proportional to the progression of the “civilizing process,” the whole class of behaviors requiring self-control and the increasing ability of modern states to monopolize the use of force, disarm citizens, and enforce order (Spierenburg, 2008). The “uncivilized” or backward nature of the USA would thus explain its twofold higher murder rate compared to Western Europe. Indeed, in autocratic China, we probably do not know the true murder rate because in the main only solved cases are reported; however, the claimed rate is about one-tenth that of the United States. Ignoring that fact, and not counting crimes committed by the state, it would nevertheless seem that Spierenburg’s hypothesis has strong support. It is not just the guns or the monopolization of force by the state, it’s the culture and here again factors appear at work to facilitate violence in the U.S. Randolph Roth, in American Homicide, takes up a thread developed by Eric Monkkonen and concludes that the high rate of murder in the USA is due to our political institutions and public faith in them, trust in the integrity of public officials, social solidarity based on race, religion and political affiliation, and faith in the social hierarchy as an avenue to achieve respect (Roth, 2009). Democracy requires dissent, and in Roth’s view when people dissent from a president whose politics they don’t approve of, this leads to an increase in the murder rate. The message seems to be to elect “good” presidents, but we can always be sure that about half will disapprove and many will deny their legitimacy. Historian–criminologists reach some convenient destinations (Americans should be more civilized, Americans should elect good presidents, Americans should stop clinging to their guns). However, to make sense of this rare behavior it seems better to identify the factors and rank them by hierarchical importance. As mentioned, some factors that will ultimately be integrated with predisposing genotypes include economic disparity and male sexual competition, youth, guns, poor medical care, poverty, drugs, and mental illness. Taken together, these factors are pieces in the puzzle of behavioral causation and leading to the question of personal responsibility should we eventually understand how they fit together to predict the behavior of an individual.
Guns or people? Each one of the individual environmental determinants just listed is itself a complex story, as illustrated by guns. A gun can be found in almost one of two American homes, and on an annual basis approximately 10,000 of the 15,000 murders are committed with guns. Does this mean that guns
A fierce people
43
cause murder? Not really. Murder rates were never higher after guns were invented than they were before guns existed. In medieval Europe the murder rate was about 35 per 100,000 and as late as the sixteenth century it was still about 20 per 100,000. At present, and after the invention of the gun, and worse—guns with high rates of fire and large capacities for ammunition—the murder rate in Western Europe is less than 2 per 100,000 and has been for the past century. Of course this ignores state-sanctioned mass killings (for example, some 40 million civilians died during World War II, more than 8 million per year, millions having been executed or having died during bombing attacks targeting noncombatants). Several countries with low murder rates are among those with the highest gun ownerships. In Switzerland, young men between the ages of 21 and 32 train for a few weeks a year and receive M-57 assault rifles and ammunition, which they are required to keep in their homes; once discharged the M-57 is replaced with a bolt action rifle. In that country of seven million people gun crime is rare—so rare that politicians usually don’t have police protection. Whenever the Swiss example is mentioned, a series of differences between Switzerland and other countries is then raised by advocates of gun control: wealth, tradition, culture, lack of problems with drugs, lack of other social problems; however, that’s exactly the point. Switzerland is not unique in having high gun ownership and low murder rates—other examples are Israel and Finland, the US being an exception that, for gun control advocates, “proves the rule.”
A fierce people Kinship and sexual competition play a role. Controversial anthropologist Napoleon Chagnon, who died in 2019, studied the Yanomami, a remarkable and threatened people who live in parts of what is now Northern Brazil and Southern Venezuela. Distilling Chagnon’s life work on the Yanomami into a few sentences, these so-called Fierce People had extraordinarily high rates of murder and frequently there were violent struggles between villages. In his book and in a study published in Science titled “Life Histories, Blood Revenge, and Warfare in a Tribal Population,” Chagnon reported that one in four Yanomami men were killed by other Yanomami men and 44% of Yanomami men over the age of 25 had participated in a killing. Chagnon’s book Yanomamö: The Fierce People (1968), a best-selling anthropology text. However, in Darkness in El Dorado: How Scientists and Journalists Devastated the Amazon, journalist Patrick Tierney accused Chagnon of giving the Yanomami weapons and other serious misdeeds. In response, in the early 2000s, Chagnon was investigated by the American Anthropological association, but eventually was largely vindicated (except for perhaps continuing to misspell the name of the tribe), having been elected to the National Academy of Sciences in 2012.
44
5. Murder is multifactorial
From the perspective of genetic fitness one of the most important aspects of the intravillage and intervillage violence that Chagnon described was that it was strongly associated with sexual competition between males— those with the most kills had the most wives and children and struggles between villages often involved raids to capture women. Also, the likelihood of conflict between individuals and between villages decreased with higher kinship (genetic relationship). As villages increased in population these factors encouraged fission, even though a larger village might be better able to fend off attacks of neighboring villages. Within villages there were several levels of conflict resolution. The first involving one belligerent hitting the other up the side of the head and then the other reciprocating. In the second stage the flat side of an ax was used. Next was the “pole stage” in which a pole was brought down upon the opponent’s head (the obvious question—as in gentlemanly duels fought with pistols at 40 paces being, “Who gets to go first?”). The next stage, which was more an example of conflict than conflict resolution, involved poisoned arrows. However, at least they did not have guns. Despite its high rate of gun ownership and “gun culture,” the US does not lead the world in murder. Because many murders such as wife burnings are not recorded as murders, the murder rate in India (and other countries) is probably underestimated although still less than that of the US. However, the nongun murder rates of Taiwan, the Philippines, South Africa, Mexico, Brazil, Estonia, and Colombia exceed the US total murder rate, and the total murder rates of at least six countries exceed that of the US by greater than threefold. Personally, and I hope to hopefully to avoid the “uncivilized” label, I favor some gun control measures. No guns on airplanes or laboratories! However, this is not because of evidence that murder rates and crime rates are decreased by gun control. In some jurisdictions that instituted gun bans, homicide rates rose dramatically; for example, they doubled in Washington, DC, from 1976 to 1991 following a virtual ban on handguns. The ban was ineffective, but clearly some factors other than guns were responsible for the large increase. In India, guns are not widely available but according to Indian police authorities at least 2500 brides are burned alive every year, and it was reported by the Time magazine that in 1995 the number of deaths by wife burning was 5800. The usual procedure is that in-laws, angered because of lack of fulfillment of a dowry or demands for additional dowry, set the victim afire using kerosene intended for cooking stoves, or gasoline or whatever other flammable liquid was at hand. Perhaps vindictive relatives would be less likely to burn these defenseless women if a few of them unexpectedly turned out to have a gun. In a 1982 survey of imprisoned US criminals, one-third admitted to being “scared off, shot at, wounded, or captured by an armed victim.”
Civilizing people
45
Other than immolation and shooting, people find many alternative methods for killing. In the datasets I have collected and in the trials in which I testified, a variety of methods were reported to have been used to commit murder. If a gun was not available, and sometimes when it was, another weapon at hand was often used. Physicians know that guns cause horrific wounds, and some ammunition is designed to do so, for example by piercing protective armor or by expanding in the body, as hollow-point bullets do. I’ve treated patients with gun wounds. Annually, more than 100,000 Americans survive being shot and more than 3 million survive cut/stab wounds—an extraordinary toll of violence. However, as the burned brides could attest, if only they were still alive, the human body is fragile and easily damaged by bludgeoning, burning, stabbing, poisoning, being crushed by cars, pierced by poisoned arrows, or being pushed from high places. Also, an interacting factor with means of injury is medical care, which has had a major effect on the murder rate. Today, many assault victims survive who only a few generations ago would have died and been counted as murder victims. Frequently, those survivors who do not appear in the murder statistics have suffered devastating losses in quality of life, productivity, and longevity.
Civilizing people Closing the loop on Spierenburg’s civilizing influence and social control argument, there is a flip side to the sense of individual empowerment provided by gun ownership. An individual with the mindset to defend themselves may be more likely to act out on their impulses, perhaps with a gun or but also using whatever comes to hand, rather than seek the help of the system. One can judge for oneself whether that is a good thing. Some people would like a society where there is more social control and less potential for an individual to disrupt things. Others, of the “Ayn Rand school,” “Live Free or Die license plate slogan school,” “Eat my dust school of driving school,” or “How’s my driving? Call 1-800-E******,” believe that the occasional death by shooting—and even accepting that these are usually tragic accidents and suicides—is a worthwhile price to pay for individual empowerment. Reasonable people, including Supreme Court justices, disagree over whether the Second Amendment to the US constitution guarantees the right of individual citizens to own guns. Context matters. The Second Amendment is a plank in a Bill of Rights for the individual. In archaic American usage militia was a word for a group of citizens, not an army. The sea change was the extension of the Bill of Rights to women and former slaves, and to some extent, to corporations (groups of people), and in the future any of these individual rights could be nullified, as is the case in several countries.
46
5. Murder is multifactorial
Violent youth A strong factor that drove murder rates upwards was the baby boom after World War II and their second-generation offspring. With the aging of these cohorts, and the incarceration or death of some of the most violent, murder rates declined. This phenomenon was strongly observed in Washington, DC, nearby where I live. This dramatically reduced the number of hotheaded male perpetrators and potential victims, so it is not surprising that the murder rate has declined (Pinker, 2011). Police chiefs and politicians claim credit for that decrease, but they do not deserve praise any more than previous ones deserved the blame for the original epidemic. The same error is usually made in comparing transnational rates. Murder rates in countries with low birth rates and an age distribution weighted to more mature individuals, for example Western Europe and Japan, will naturally be lower. Also, these age effects are not merely arithmetically proportional. Over longer periods the age structure and demography of a population affect personal expectations, leading to a culture in which violence becomes more normative, as in the Yanomami. It is this complex arena in which genes altering propensity act. For example, in Mexico over the past several decades, economics as well as desperation have led to violence associated with the abuse and trade in drugs, and as has been driven by drug use in a neighboring country, namely the US. Changes in male/female sex ratio and level of polygyny in turn change the dynamics of access to females, and the socioeconomics of sexual competition are powerful drivers of violence of young men, and—in the vein of Napoleon Chagnon’s sexual competition, the dynamics of mating can help make sense of some of the horrific mass murders committed by socially isolated adolescent males who have failed in the evolutionarily paramount imperative to reproductively affiliate with females. In polyandrous species such as some birds in which females have multiple mates it is the female who tends to be larger and more aggressive (Jenni, 1974). This is not the case in humans and other great ape, lesser ape, and monkey species in which sexual dimorphism varies in degree but it is the male who is larger in size, fang, and aggression. Females can also be aggressive but across primate species are less aggressive, and as tracks with physical sexual dimorphism. In humans the male/female bodyweight ratio is 1.15, substantially less than some of our closest relatives. In bonobos and chimpanzees, the ratios are 1.36 and 1.29, but male gorillas and orangutans are more than twice as large as females, 2.37 and 2.23, male orangutans continuing to grow throughout their lifespan (https://carta.anthropogeny. org/moca/topics/sexual-body-size-dimorphism). Behavioral causality is both macro and micro. For the origins of violence, evolutionary psychologists Margo Wilson and Martin Daly at McMaster University identified both. At the micro level, they noted that
Violent youth
47
police and criminologists were more often startled by how ludicrously small and (as will be seen) seemingly trivial are the triggers for murder, and yet the murders were far more likely given broader context. Wolfgang (1958) classified 560 homicides in Philadelphia into 12 motive categories. Eighty-seven % of the murderers were men. “Altercation of relatively trivial origin; insult, curse, jostling, etc.” was the leading cause by a wide margin, accounting for 37% (Wolfgang, 1958). There is broad agreement that murder is usually committed on impulse and the spur of the moment. In “2B or not 2B” I described how a strongly functional genetic variant of HTR2B can predispose to murder, and the murders committed were of this seemingly senseless type. As far back as 1969, the staff report for the National Commission on the Causes and Prevention of Violence analyzed criminal homicide in 17 U.S. cities, finding that “altercations appeared to be the primary motivating forces both here and in previous studies. Ostensible reasons for disagreements are usually trivial, indicating that many homicides are spontaneous acts of passion, not products of a single determination to kill” (Mulvihill et al., 1969, p. 230). As conveyed by Wilson and Daly, the report quotes a Dallas detective: “Murders result from little ol’ arguments over nothing at all. Tempers flare. A fight starts, and somebody gets stabbed or shot. I’ve worked on cases where the principals had been arguing over a 10-cent record on a juke box, or over a one-dollar gambling debt from a dice game” (Mulvihill et al., 1969, p. 230). In identifying the “Young Male Syndrome” Wilson and Daly made sense out of many of these seemingly senseless murders (https://www.martindaly.ca/uploads/2/3/7/0/23707972/wilson___daly_1985_young_male_ syndrome.pdf). Although I never met Margot Wilson, throughout my career I tried to follow the advice of Martin Daly, her spouse and partner in research, and who was an influence although I only met him once at a small conference. Studying homicides in Detroit, Daly and Wilson drew from the work of sociologist Marie Wilt, observing the same patterns of senseless killing of young males by young males. But—digging more deeply—they found that “Face” and dominance status within a highly sexually competitive social milieu were usually at stake. Males compete for females as if they are a resource, because in evolutionary terms they are, the “fitness” of the male depending on reproductive access to females. In these competitions, the poor and disenfranchised have the least to lose and will thus—in modern neuroeconomic terms—risk the most. Analyzing 512 homicides, Wilson and Daly found that most were male on male, and usually victim and murderer were young (peaking in the early 20s), more than likely unemployed, and usually unmarried. More than two-thirds of the cases were caused by social conflict, and in many the motive directly identified was jealousy or the provocative situation was male-to-male escalation of one-upmanship.
48
5. Murder is multifactorial
The thesis of this book is that free will emerges from our neurogenetic individuality and self-guided neural plasticity. We can shape our lives and brains to become what we want to become, and resignation of oneself to the environmental milieu and our inborn limitations is itself a top-down decision. However, people are born unequal and with unequal opportunities. As we have learned from Wilson and Daly’s studies of homicides in Detroit, Chagnon’s research on the Yanomami and as will be discussed later, Wilson and Daly’s studies of parents and their adopted children (the “Cinderella effect”), the social milieu can profoundly influence choices, even to commit murder. Worse, this propensity is ingrained by evolutionary forces that have biased our brains toward violence against sexual competitors and those who we identify as not of our blood kin (Trivers, 1972).
References Bevilacqua, L., Doly, S., Kaprio, J., et al., 2010. Population-specific HTR2B stop codon predisposes to severe impulsivity. Nature 468, 1061–1066. Chagnon, N., 1968. The Fierce People. Jenni, D.A., 1974. Evolution of polyandry in birds. Am. Zool. 14, 129–144. Kleiman, M., 2009. When Brute Force Fails: How to Have Less Crime and Less Punishment. Mulvihill, D.J., Tumin, M.M., Curtis, L.A., 1969. A Staff Report Submitted to the National Commission on the Causes & Prevention of Violence. vol. 11 U.S. Government Printing Office, Washington, D.C. Pinker, S., 2003. Blank Slate: The Modern Denial of Human Nature. Pinker, S., 2011. The Better Angels of Our Nature: Why Violence Has Declined. Roth, R., 2009. American Homicide. Spierenburg, P., 2008. A History of Murder: Personal Violence in Europe from the Middle Ages to the Present. Trivers, R.L., 1972. Parental investment and sexual selection. In: Campbell, B. (Ed.), Sexual Selection and the Descent of Man 187/−197/. Aldine, Chicago. Wolfgang, M.E., 1958. Patterns in Criminal Homicide. University of Pennsylvania Press, Philadelphia.
6 Distorted capacity: The measure of the impaired will If someone talks of subconsciousness, I cannot tell whether he means the term topographically—to indicate something lying in the mind beneath consciousness—or qualitatively—to indicate another consciousness, a subterranean one, as it were. He is probably not clear about any of it. The only trustworthy antithesis is between conscious and unconscious. Sigmund Freud—The Question of Lay Analysis
Conscious and unconscious behavior All people make decisions, but some are inherently more patient and reflective, while others, as we learned in the “2B or not 2B” story, make decisions that are relatively impulsive. In each case a person has made a choice to engage in a conscious or explicit behavior. People also perform many actions implicitly, which is to say without conscious thought. Such actions can also be termed unconscious, but not subconscious, an ill-defined term. Breathing is an example of the way people ordinarily transition between unconscious and conscious behavior. We ordinarily do not think about when to take a breath, and certainly not while sleeping. Breathing is often unconscious. However, we can at will initiate or inhibit breathing, and often do when we are instructed to do so by a doctor, practice yoga or breathing exercises, prepare to dive under water, or hold our breath while beneath the surface. Everyone has experienced that the urge to breathe becomes progressively stronger, and suppressing a breath requires increasingly strong conscious control, until loss of consciousness and reflexive breathing. Similarly, a person may unconsciously withdraw a hand from hot metal or from icy cold water. However, people have different degrees of capacity to consciously hold their hand against the pain and can weigh the potential consequences. Impulsivity, which at its extreme is action without forethought, an “implicitization” of behavior, is an important, normal behavioral trait.
Our Genes, Our Choices https://doi.org/10.1016/B978-0-443-22161-3.00015-6
49
Copyright © 2024 David Goldman. Published by Elsevier Inc. All rights reserved.
50
6. Measurements of capacity
Hans Eysenck, the legendary psychologist who factor analyzed human personality, pointed out that humans need impulsivity to enable them to initiate behavior. Following in the paths blazed by Eysenck, Cattell, and others, psychologists can measure personality and temperament relatively easily using questionnaires. Although there are many such questionnaires and personality scales it is a remarkable fact that common personality factors emerge when the results are analyzed using a statistical methodology called factor analysis. Factor analysis takes the responses and asks, in a hypothesis-free fashion, whether there are response patterns that cluster together. Of course, whether a factor can be discovered depends to some extent on what questions asked. So, for example, the person who says that they are “afraid of spiders” is also more likely to report that they are uncomfortable speaking before large groups, leading to the identification of a factor associated with anxiety and introversion. One factor that emerges time and again even when the questions are varied is a trait associated with extroversion, novelty seeking, hyperarousal, and diminished behavioral restraint, or in other words impulsivity. We often worry about impulsivity and arousal in the context of aggression and violent behavior, and indeed some studies of impulsivity are motivated by interest in aggression, including aggression directed against self. However, impulsivity and arousal are core to normal existence, and to diverse psychopathologies including substance use disorders, borderline personality disorder, and antisocial personality disorder, as will be explored in this chapter. Interpersonal aggression because of the complexity of its origins, and because victim, accused, lawyer, jury, judge, lawmaker, parent, and teacher all join the conversation, is an arena where the origins of choice are disputed most intensely, and with lives and futures at stake. We often only become concerned about the behavior of children and adults at the precise moment they become aggressive and disruptive, and leading to evaluation, and often diagnosis and medication. Otherwise, we may ignore them. This observation cuts across diagnoses, age groups, and circumstances. We are not as perturbed by the homeless person with schizophrenia who is quietly suffering but are moved to action when they aggressively approach commuters on the subway. Such socially proscribed aggressive behavior seems to exemplify impairment of will and directly raises the issue of agency. We do not want to blame, or praise, people for behavior made merely on impulse. However, at some other times and people the “same” behavior can be thoughtfully, and even diabolically, motivated, and usually the origins of behavior are somewhere between these opposite poles. The neurogenetic contributions to aggression and dissocial behavior are thus politically sensitive, generating concerns for stigmatization and misuse.
Context appropriate and inappropriate behavior
51
The antipsychiatry movement believes that ordinary disruptive behavior of children is reified as attention deficit hyperactivity disorder, and thereby medicalized and medicated. Repeatedly, neurogenetics has been used to stigmatize the socially and economically disadvantaged who have higher rates of criminality and incarceration and to justify racism, and even genocide. All of this has been true at times, and in places, and not true at others. Many disruptive children are overmedicated or given medications they do not need at all, and on the other hand a larger number of children with attention deficits are helped through critical phases of their life when their academic and social development is at stake via medication and other accommodations (for example, more time to complete tests, seating them in proximity to the teacher), and if along the way they do not disrupt the entire class so much the better. As is coming later in this book, the main argument of Skinnerian neurogenetic determinism is that behavior is neurogenetically determined, and thereby people are “beyond freedom and dignity.” The decision to medicate away a dissocial behavior is easily taken if the person was not in the first place free. If an adherent of hard neurogenetic determinism treats other persons “as if” they are free, this is done out of some combination of expedience, condescension, or artifice. By exactingly defining the neurogenetic origins of dissocial behaviors it can be shown that such behaviors are usually constrained, constraint being a reality of existence, but free, in people who choose to exercise their free will.
Context appropriate and inappropriate behavior One reason aggression itself, and as contrasted with impulsivity and arousal, is not a good genetic phenotype is that there are many types of aggressive behavior, and the meaning of aggression is context dependent. Aggressive behaviors are diverse, are often best understood as adaptive or maladaptive social behaviors, and overlap strongly with other social behaviors. The protocol and physics of physical interaction must be learned, just as other social behaviors must be learned. Rough and tumble play is normal and healthy among children at a playground and by adults and children “playing” by the rules in sporting events. Play ineptitude often predicts social ineptitude, and to some extent this a cause-and-effect relationship. Some people are innately more facile in physical interactions and physical/ social interactions. At the other end of the spectrum are people with autism or autistic features who face great barriers. Innate levels of arousal and irritability modulate the ease with which children learn to interact physically and socially. Children with social phobia face a higher barrier. Physical interaction is just one part of a multidimensional continuum of social interaction that includes many facets in which behaviors
52
6. Measurements of capacity
at opposite ends of the scale make others uncomfortable or are directly perceived as threatening, rude or a sign of dishonesty, weakness, sickness, or unreliability. For example, volume, pitch, tone, modulation of voice, pace of speech, eye contact, listening versus speaking, touching, olfaction, proximity, facial expression, body habitus, and limb movement. As many an older politician has learned, what is fitting is context dependent. Books list rules of etiquette just as books list rules of sport but the nuances of social interaction are primarily learned not from books but via peer interaction. In every social interaction the child learns through observation both by conscious study and via the unconscious patterning of mirror neurons. They also model their behavior based on what they observe on media. They put what they observe into physical practice. This is one reason why it is beneficial for a family to have a pet. The child learns kindness and measured treatment by observing the behavior of others toward the pet, seeing—for example—that the fierce play of a kitten should not be misinterpreted. They learn from the behaviors of the pets themselves, and for example to relent once an opponent submits. They learn to refrain from force that would maim or kill. On the other hand, if their examples behave badly, the child is likely to emulate them. If all goes well, the end product of the meld of innate predisposition and environment is an adult whose behaviors are appropriate to a particular time and social context, and the most successful will be those able to take cues and adapt, while projecting a semblance of their own personality. Too often, the child is abused—themselves learning to abuse. Or the child is neglected such that they do not learn while the brain is most plastic. Or as is increasingly true, the child is coddled. Thus the neurogenetic determinism of choices impacted by arousal and impulsivity is as contingent on upbringing and experience as genes. Also, a person who wants to learn to handle social situations more adroitly can deliberately place themselves in circumstances where they can—or must—learn to do so better, and—previewing discussion later in this chapter—can practice cognitive strategies they may initially apply consciously and later as second nature unconscious behaviors. The adult brain is neuroplastic and capable of such adaptation, but the child’s brain is more so, and—as noted—the child who is dyssocial is at risk of a cascade of other problems. This leads to every parent’s dilemma as to whether they should let their child experience or protect them from the risks accompanying youth sports and the further risks of unregulated play, or later—to experiment with drugs or take part in other risky activities. Most children do not quite do fine left to their own devices (see Lord of the Flies) such that some guidance by adults and older peers is essential in the developmental expression of normal instincts for play, and as youth sports supervised by adults can encourage. However, whereas most
Personality types and choices
53
children benefit from latitude to freely interact, a few are too rough and tumble or too timid or meek to engage in play without more guidance, and a smaller proportion still (e.g., children with autism) need more intensive attention. Interestingly, that proportion is surging, either because of overdiagnosis, better recognition, or some change exposures causing autism; the frequency of autism in the U.S. is estimated by the CDC to be 1 in 189 for girls and 1 in 42 for boys. The CDC estimates that the prevalence of ADHD in children is 9.4%. Part of the attention needed can be to help the timid or indecisive child initiate behaviors—almost one in six children being under treatment for an anxiety disorder, and on the other hand, to help the overly aggressive or impulsive child to modulate theirs. In children, both timidity and aggression can be more intense because of the higher level of arousal, and because frontal cortical-mediated executive cognition will not be fully developed until late adolescence or young adulthood. During development, and throughout adulthood, impulsivity and arousal are normal, evolutionarily essential, behavioral traits, as Eysenck, who first factor analyzed human personality, pointed out. To survive, we must initiate behavior, even though almost any behavior entails some burden of risk. Failure to decide often translates into a coin flip guess that would better be made by deliberation. Sometimes, indecision leads to identification of a third or fourth way of doing things, the either/or decision representing a false binary. However, deadlines compel decisions, whether between the either/or in favor of the alternative that necessity has mothered. A person who cannot decide is further from free, and in some cases even if they think they decided not to decide, because that may not really be what happened. Impulsivity will always to some extent be socially defined, as even more so are aggressive behaviors that impulsivity can unleash. Although impulsivity (and arousal) driving behavior and executive cognition modulating impulse are in some ways the more fundamental behaviors they concern us less in our everyday lives, because it is the aggression that touches us and not the neural states influencing it.
Personality types and choices Two contradictory memes have long uneasily coexisted within psychology and are often espoused by the same expert. The first is the politically relevant belief in uniformity of developmental trajectory and primacy of the environment in shaping differences in life arcs and outcomes, such that equity and justice can be accurately graded by measuring disparities in outcome. If outcome differs, the social environment is unjust. The second demiurge is classificatory—the project of identifying personality “types.” First came the paper and pencil personality inventories in the 1920s. From answers to yes/no questions such as “Are you afraid of spiders” sufficient
54
6. Measurements of capacity
data could be collected to perform factor analysis, inductively identifying 3–5 factors, and similar results were seen when scales were constructed deductively, perhaps because observation of personality differences has always been a vital social skill. On physiology and personality, the views of Aelius Galenus, a Greek physician living in the Roman Empire, largely came down from Hippocrates and were unchallenged for more than 1300 years. Galen identified four humors influencing personality: blood (Sanguine), phlegm (Phlegmatic), yellow bile (choleric), and black bile (melancholia) that in unbalanced proportions lead to personalities characteristic of each of the humors (Nutton, 2005). Everyone knows representatives of Galen’s personality types. But, remarkably, a vast amount of data measuring personality scales not very different than the four humors find that roughly 4 (3–5) personality factors can account for about two-thirds of variance in responses. These scales are embodied in many tests, the Eysenck, the Minnesota Multiphasic Personality Inventory, the Myers–Briggs, the Tridimensional Character Inventory, the NEO PI-R, and many more. Importantly for many employers (including the army) the personality scales include useful gems such as conscientiousness. Importantly for biologists, heritability of a personality subscale score tends to range between 40% and 60%. This is very surprising given that many events in a life might be thought to be life—and personality—changing strongly indicating that personality is to a substantial extent innate. Furthermore, it might be supposed that paper and pencil tests could not measure personality very accurately. However, the high heritability of personality easily disproves that. Whatever is being measured is being measured with some high degree of accuracy such that resemblance for personality between pairs of individuals can correlate highly with coefficient of relationship, ranging from 0 from unrelated individuals, to 0.5 for siblings, to 1 for identical twins. We think we know impulsivity when we see it, but when we set out to measure it, either via questionnaires or performance, we are immediately confronted by the many routes to the same behavioral destination. Ultimately, the complexity of even “impulsive” choices says much about the neurological origins of choice, and whether “choices” are determined. The child left alone in a room with a plate of cookies (or marshmallows) may snatch one for many reasons. Or the same child may choose to take only one, wait many long minutes before giving in to hunger or curiosity, or—inexplicably—not take any. The child who ate the whole plate of them may be more impulsive but also may be hungrier, angrier, bored, rebellious, or curious as to the nature of the next cookie or what will happen if they eat all of them. The child may decide that there is at least a remote possibility that the treats are being counted as part of some sort of diabolical experiment, which might lead them to a variety of choices.
Personality types and choices
55
Perhaps having taken one, it occurs to the child that if all of them should disappear their answer can be “What cookies?” Obviously, a child is more likely to dream up such scenarios than is your average mouse. Thus if one is interested in measuring the magnitude and flavor of a biologic characteristic in humans it is wise to root the measure as closely as possible to brainstem and midbrain mechanisms that have been passed down to humans from their distant mammalian and reptilian ancestors rather than as modified by frontal cortical cognitive executive circuits. Those circuits, in “top-down” fashion, enable the brain to choose to modulate practically any behavior, or not to do so, letting impulse take its course. At the extreme ends of the behavioral spectrum are conditions labeled as personality disorders. As discussed in this chapter, hyperarousal is integrally related to borderline personality disorder, antisocial personality disorder, and to some extent, ADHD. It is also confused with psychopathy. Psychopathy has been recognized since ancient times—probably our brains are evolved to recognize it. For example, Allied Supreme Commander Eisenhower considered at least one of his generals, Montgomery, to be a psychopath, and during the Trump administration hardly a day went by without someone diagnosing him. Psychopath is an effective pejorative precisely because it is so ill defined, and a main virtue of calling someone a psychopath is that the more they deny it the more obvious it is that they are. Psychopathy is a construct refined and championed by Robert Hare, a Canadian psychologist, who noticed that some prisoners did not improve with punishment or efforts at rehabilitation. In Without Conscience: The Disturbing World of the Psychopaths Among Us (1993), Hare described psychopaths as social predators (Hare, 1993), and in Snakes in Suits: When Psychopaths Go to Work (2006) he and a coauthor observed that psychopaths are often successful in the business world, while harming those around them (Babiak and Hare, 2006). The impulsive child will probably take one of the cookies or marshmallows and genuinely regret it. However, the psychopathic child after eating the cookies may claim there were no cookies, fake regret, hint another child took them, or plead misunderstanding. The psychopathic child, suspecting a trick, may eat none, offer the cookies to others, or induce another child to take the rap. Such a good child. Unlike psychopathy, antisocial personality disorder (ASPD) is an actual DSM diagnosis and as such it is a label most likely to be applied to recidivistic felons, people who may never have been arrested but who are incorrigibly resistant to social control, and presidents we do not like. However, there are many reasons for irredeemable disruptive behavior, including hyperarousal, poverty, politics, cognitive deficiency, obsession, addiction, and impulsivity. Both psychopathy and ASPD have been attacked as a general theory of crime, but there is no general theory of crime. ASPD, a common psychiatric disorder, is on a genetic basis most strongly cross-transmitted with borderline personality disorder, the
56
6. Measurements of capacity
common link between these social behavioral disorders being hyperarousal—a comparatively distinct but still physiologically complex cause of impulsive behavior. Neither ASPD nor borderline personality disorder are mechanistically defined—they are behaviorally defined, but in both disorders behavior is unstable, and often marked by impulsive decision making, especially in response to minor irritations, and which is known as irritability. Patients with ASPD and BPD often have many things simultaneously going wrong in their lives—problems with others, the law, impulsive harm, and self-harm, and many of these debacles happen because their arousal or thymos, overcharged like the young of many species, has led them to actions without thinking. In this way, they behave less like a mature adult making top-down strategic choices, and more like a child acting on instinct, and on that basis not exercising free will. Furthermore, and as is part of the basis of Marsha Linehan’s dialectic behavior therapy (DBT) for BPD, the bad decision making is likely to have become engrained, and such DBT has helped many patients with BPD become better decision-makers. Psychopathy is not a behavior of this sort; it is a way of thinking at all deliberate speed. The Hare Psychopathy Checklist, the instrument most often used to measure psychopathy, captures the lack of empathy enjoyed by a certain small fraction of the population. In Finland, and as was described in Chapter 3, I and my colleagues Matti Virkkunen and Markku Linnoila studied impulsive criminals, almost all of whom had ASPD and alcohol use disorders (and many also had borderline personality disorder). They were not psychopaths. As was in part due to a stop codon they carried in their HTR2B serotonin receptor, they had committed impulsive crimes (usually murder) for no personal gain and almost invariably while drunk. Afterward, they were remorseful, although at another date they might commit another impulsive crime. They also tended to have a biomarker of impulsivity, namely, low levels of 5HIAA (5-hydroxyindoleacetic acid), a metabolite of serotonin. Later, it would be shown by Dee Higley and Steve Suomi at the NIH, in studies on which I was part of the team, that some of the same important gene × environment interactions that are important in human impulsivity occur in monkeys if they are exposed to early life stress via maternal deprivation. Higley observed that monkeys with many scars on their faces and that tended to take death-defying leaps between trees had lower 5HIAA in their spinal fluid, and in addition to being heritable, these low 5HIAA concentrations were also moderately heritable, from one generation of impulsive monkeys to the next (Higley et al., 1993). Such gene × environment interactions, and the measurement of their effects, will also be discussed later, but a narrative had been established that impulsivity, arousal, and aggression could be measured, but also that extrapolations from impulsivity to criminal behavior were overly reductive, erring by
Personality types and choices
57
directly comparing complex social behaviors in humans to those of other species and going wrong by implying simplicity of causality and behavioral prediction that remain out of reach. For understanding the neurogenetic origins of choice, it is important that impulsivity is pharmacologically manipulable. As noted, serotonin is a neurotransmitter that helps animals inhibit behavior, and when depleted, which can be done pharmacologically via the drug para- chlorophenylanine or by dietary deprivation of the serotonin precursor tryptophan, they become impulsive, irritable, and aggressive. In such states rats treated this way are more likely to bite. Psychopaths are not like that. Preying upon others as if without conscience they will bite with all due deliberation or sadistic enjoyment, as cat with mouse, not as cat fighting for its life with dog or other cat. Psychopathy—usually while conflated with ASPD—has for always been a concern of behavioral theorists at the frontiers of social behavior and the law, having recognized that a substantial portion of the population does not play “fair” in society’s tit-for-tat games. Instead of doing unto others as they would have them, psychopaths do unto others first. Modeling interactions of computer bots, it has been shown that in the social trust “titfor-tat” game it usually pays to be generous during an initial encounter with a stranger, only taking more than one’s share in retaliation with specific untrustworthy transactional partners. However, if too many untrustworthy entities are introduced into the common social milieu, rewards are reduced overall—the social fabric breaking down—but the players most damaged are ones that are the fairest. Over time, those players, if they are human, are likely to learn to behave in more untrustworthy fashion, making choices against their nature. Reciprocally, untrustworthy partners are more likely to behave in a trustworthy fashion under conditions of tighter surveillance and harsher penalties, even though at heart they are antisocial. Psychopathy is thus a different thing from impulsivity and hyperarousal altogether, even if the result is that it can predispose people to behave disagreeably, or malignantly, and as if it is their nature to do so. On an individual level, competitive aggression causes us to excel and accomplish, and for many people is an essential part of existence. Without aggression we would not have survived our millions of years of evolution, and without which we might not survive now, and to succeed in society some level of aggression is required, and the requirement varies depending on the society, and all the particulars of one’s situation. It is one thing to be a pacificist living in a gated community and quite another to be eking out a hardscrabble existence between a rock and a hard place where the hand of the law is only occasionally seen, or where the law and its enforcers are the threat. Jesus, perhaps, or certainly his disciples urged soft words to counter harshness, perhaps to deflect conflict but not
58
6. Measurements of capacity
to escalate it. However, it is well that pacificists and neutrals sometimes have someone to protect them, because otherwise too many of their voices might have been silenced, or—if countries—their borders violated. For example, by the Treaty of London (1839) other European powers guaranteed the perpetual neutrality of Belgium, but in two world wars Belgium was a battlefield and occupied. Sweden, remote from the European theater and supplying the Nazi’s with iron, was able to maintain its neutrality, as was Switzerland, which was not needed as a military thoroughfare, and its banking activities continued. However, during World War II, as in the war to end all wars that preceded it, other neutrals were sooner or later drawn in: Denmark, the Netherlands, Belgium, and Norway quickly, and, eventually, the United States. Clearly, nations—like children—also learn by the physical experience of having fought or having been attacked. Formerly neutral countries such as Belgium, the Netherlands, Luxembourg, and the United States became founding members of NATO, which to this day does not include Switzerland or Sweden, although—as I write—Sweden appears on the verge of joining NATO because of Russia’s recent invasions of Ukraine, supposedly intended to prevent the expansion of NATO. Until that happy time when everyone behaves nicely, it is for communities, nations, and coalitions of nations to stop some behaviors, at times using force but never in disproportion and always attempting to prevent and defuse rather than inflame. The idea is that force properly applied can reduce the need for force. Never has this distinction been more graphic than at the time (2020) I was revising this book, when via videos made with handheld devices much of the public was awakened to the use of disproportionate force by people—namely police—who should have been specifically trained and motivated to do better, and too often directed against people with black or brown skin. Thus when we speak of someone being “aggressive,” we are usually making a judgment that their behavior was disproportionate to the context.
The inheritance of impulsivity, and what it means In twin studies, including studies of twins raised apart, the inheritance of personality traits has consistently been observed to be 40%–60%. As discussed elsewhere, the basis of the twin method is to compare the resemblance of identical twins to nonidentical twins, and for personality traits identical twins tend to resemble each other, or correlate, much stronger. This is quite remarkable because of the crudity of the measurement tools and the multitude of nongenetic factors that could alter personality, or the ways individuals respond to questionnaires of this nature. There are at least two important lessons to be drawn from the heritability of personality. The first is that these little questionnaires do a remarkably good job
The inheritance of impulsivity, and what it means
59
at measuring something. We know this because heritability will always be limited by precision of measurement. For example, if the precision of measurement is 70%, the maximum heritability could be 50% (0.72), which is about what the heritability of personality factors has been observed to be. If environment also plays a role in personality, the precision of measurement of personality factors is probably higher than 70%. The lesson to be drawn from the heritability of personality is that there are genetic variants that influence personality that are inherited from parent to child, and whose sharing between siblings makes them more similar. The third point is more subtle: since even siblings and fraternal twins who share only half of their genes resemble each other in personality far more strongly than random pairs of individuals. This is important in understanding how the genes work. As will be discussed more later, one possibility for gene action on behavior is that only particular (epistatic) combinations of genetic variants, acting in concert, produce a given phenotype. The other possibility is that gene variants tend to be additive in action, in which case the effects of a particular genetic variant can be understood without necessarily understanding everything else. Additivity versus epistasis used to be constantly disputed among behavioral geneticists, but the argument subsided when it was seen that polygenic scores constructed by simply adding the effects of many individual genes appear to be predictive. In a rat model of arousal and impulsivity—itself predictive of liking for drugs of abuse such as cocaine—created by University of Michigan neuroscientist Huda Akil, we found (Zhou et al., 2019) that a half dozen genomically significant loci predicting two-thirds of the genetic variance in arousal in novel environments acted additively. The larger point is that data from human twins and other relative pairs—those data representing effects of all the genes—not just the ones mapped—had supported additivity. Indeed, for most behaviors, and whether in humans or rats, resemblance between pairs of individuals increases in arithmetic proportion to degree of relationship. Illustrating the power of genomics and the role of genes in choices, but still only a foretaste of discoveries to come, in Karlsen Linner et al. (2019) detected 106 genes increasing risk-taking, and 295 genes influencing risk tolerance, the latter in a sample of more than a million people. Surprisingly, given that its origins are so strongly influenced by social context, criminal violent behavior is itself moderately heritable. Harkening back to Kallikak family, criminality does tend to run in families, and the propensity is partly genetic. Some decades ago, BBC host Jeremy Paxman sprang another guest on me (or maybe I was sprung upon him). This man, convicted of murder, undermined his claim that such behavior was not inherited by revealing that he was not the only murderer in his family. Familiality arises because of shared family environment as well as genetics. However, adoption studies, for example by Michael Bohman and Robert
60
6. Measurements of capacity
Cloninger in Sweden, have long shown that genes are involved, not just family environment (Cloninger et al., 1981). Adoptive children tend to share the criminal behavior of biological parents, especially fathers, they have never met. Sharing of genetically influenced risks, and benefits, tends to correlate with degree of relationship, implying that the genes act additively rather than in only particular combinations (i.e., epistatically). The additivity of gene effects in behavior is fortunate for gene discovery because statistical analyses allowing for gene–gene interactions are inherently less powerful and more difficult to interpret. When many tests are performed, one must correct for the possibility of a lucky guess. Otherwise, science becomes like a horse race, in which the bettor who buys tickets for 100 combinations is more likely to hit the Trifecta for Win, Place, and Show, independent of skill at judging horses. As we try to identify effects of gene combinations we must take care to avoid “winner’s curse,” which is the phenomenon that if many people play a guessing game or millions of lottery tickets are purchased, someone has to win, and lucky winners can easily delude themselves that they “have a system.” With a million genetic markers, about 5 × 1011 two-way interactions can be tested, or with 25,000 genes, >300 million gene–gene interactions. Any two-gene interaction discovered by such blanket testing would require very strong statistical support to confidently state that the finding was real and did not happen merely because the geneticist bought millions of lottery tickets. The penalty to be paid for the blunt-force approach of testing all interactions is larger, deeper, and more expensive genetic studies. However, once having found genes “for” behavior such as HTR2B and MAOA, we can then begin the more difficult process of determining how they interact, but at least the number of two-way tests is limited, because so far only a few genes are known to alter behavior. Even with a small number of genes, some will wish to test for higher order interactions (three-way, four-way), again expanding the number of tests, each representing a lottery ticket.
Impulsivity differs from person to person and from species to species The heritability of impulsivity has a very important limitation. What is on average true for the group is often false as applied to the individual. Impulsivity and its related personality characteristics such as novelty seeking and arousal are descriptors that are in some obvious ways and to greater and lesser extents socially bounded. Whatever the overall heritability of impulsivity and other personality traits, there are certain individuals in whom habit and environmental exposure play the predominant or deciding role. Heritability is just the average g enetic contribution to the
Impulsivity differs from person to person and from species to species
61
trait. Where does the gorilla or the monarch sit? Wherever they want, and in one case because of genetic endowment (the gorilla) and in the other because of environmental endowment and experience. Some species are more impulsive and aroused, whereas others are more reflective or inhibited. As a species humans are patient. This is because most of us, and our ancestors who bequeathed their genes, are all our lives walking on social edges that require a high degree of impulse control. There are many ways and places for a person to fit in, but all the landscapes of socially acceptable human behavior are complex, and many are knife-edged. One step too far to the left or the right can result in social failure, ostracism, and worse. Humans are adapted for social life and to perform tasks that require long-term focus, planning, and patience. Thereby, human arousal and impulse are powerfully counterbalanced and channeled by impulse control. As compared to other primates, and even more dramatically as compared to rats, humans have overdeveloped frontal lobes. The overly impulsive and aroused individual is more likely to be become involved in social conflict and aggressive interactions. However, people with such deficits can compensate, going beyond instinct to plan for their futures and to avoid pathways that may in the short term be continually rewarding, but in the long term disastrous. Often a person has chosen a pathway for which they are not well suited. In the movie Full Frontal, one of the characters disparages another: “She drives a car that does not suit her,” by which she may have meant that there was a discordance between that person’s personality and life choices. A notable example of the need for humans to maintain impulse control over the course of a lifetime, and the consequences of rare missteps, is addiction. All modern humans are potentially exposed to highly addictive drugs and other addictive agents, for example gambling and cyberworlds. In the short term these agents reward, but in the long run can lead to misery. Impulsive individuals find it considerably more difficult to “just say no” to drugs, and once addicted are further impaired in their volitional capacity. As discussed in an earlier chapter, one human, or mouse, may be more impulsive or more cautious than another, even owing to the action of a single gene but even more to the combined actions of many. Considering the effects of such variation in impulsivity in the simpler context of the life of a rodent, caution and impulsivity have differential value in life situations. The incautious mouse is more likely to wander into an open field to be borne off in the talons of an owl or the jaws of a fox. On the other hand, the too-cautious mouse may not feed and thrive as vitally as the “moxy mouse” with just the right quotient of chutzpah. There should be no doubt that whereas the primal behavioral characteristic—impulsivity— may be the same in man and mouse, the range of c onsequence is vastly different in the two species, as is the decision making, as we will see next. Impulsive humans, unlike mice, are exposed to a vaster range of risks, and
62
6. Measurements of capacity .
many of the risk exposures are new ones to which there was no possibility of evolving instinctive protective responses.
Zero-trial learning Rather than relying on instinct, people choose life paths, but as they proceed along those paths they are constantly, even if partially, aware of the need for self-control, forethought, and sublimation of immediate wants. In this regard their internal mental state and the processes leading to the activation or inhibition, or delay of action, are to some extent and in many situations thoughtful and explicit, rather than unconscious, reflexive, and automatic. Placed in an uncomfortable position—like a cat on a “hot tin roof”—they may choose to endure the suffering because of an understanding that the alternative is worse. It is important to acknowledge that much of the behavior of other species is also noninstinctive. Inherent to the adaptive success of most species is the ability to learn, and to inhibit behaviors until the right moment comes to release them. This frequently includes the ability to emulate the behavior of another animal of its species, or related species—as in mixed flocks of birds. However, compared to humans, the learning of most other species is far more likely to be associative. These species are hampered by both their more limited conceptual range and cognitive language. Lacking these tools, a mouse may be trained to stay on the experimental apparatus which is the equivalent of the “hot tin roof,” which is a hotplate, rather than to jump. However, it takes repeated trials of training to create the associative learning linkage in the animal’s brain. It is comparatively difficult for one mouse to simply explain the situation to the next mouse. Associative learning is not the most important way that humans learn and regulate their own behavior. We often learn by example. Furthermore, we often do not need to see an example or be told. We can build a conceptual model of the situation enabling us to anticipate and avoid the consequences of things that have never been directly experienced. Such zero-trial learning, enabled by the much greater cognitive predictive capacity of the human brain, sets humans apart, or can, from other animals. Humans do not have to destroy the world’s ecology via global warming, because we can anticipate the consequences of our actions and modify them.
Impulsivity and aggression in context Impulsivity is strongly tied to aggression for the simple reason that much of a human’s environment does not consist of things but is a social network composed of people. If a person is prone to “lose it,” soon others
Measuring impulsivity and aggression
63
will experience the effects. The aggression may be physical, but it can also be verbal or emotional. Thus aggression is socially defined and significant in the context in which it occurs: the living room, the boardroom, the road, the playing field, the hunting ground, the battlefield. In different contexts the same behavior has a completely different meaning, and indeed impulsivity plays a very different role. The soldier who bayonets an enemy in battle is following his training and may have made himself a hero. The inebriated man who has put a knife into their best friend has just made the biggest mistake of their life. Without aggression humans would not have survived to discover the tale of their origins. To succeed in human societies, some level of aggression and its underlying mechanisms—arousal and decisiveness in decision making—are required, and the requirement varies depending on the society and all the particulars of one’s situation. However, the role of aggression and dominance has been vastly overestimated by some who believe they are the modern advocates of social Darwinism. As we shall see later, when discussing the genetics of sex, the meek not only may inherit the Earth but quite often do. Aggression, impulse, and decisiveness are essential behaviors among many, and all intermix into choices. The prospective mate may be more interested in the builder than the hunter. As will be discussed in Chapter 13, behavioral selection is both frequency dependent and context, that is, niche dependent. Regarding frequency dependency, it is often good to be a warrior if all around one are worriers.
Measuring impulsivity and aggression As a medical student, I had a one-on-one tutorial with Ernest Barratt, who developed the Barratt Impulsiveness Scale, a 30-item questionnaire that became the one most widely used measure of the so-called construct of impulsivity, one of the best, but hardly only, predictors of aggression. Confronted with a stranger whose parking space one has just stolen, what one wants to know, in addition to whether they are armed, dangerous, or politically connected, is whether they have a short fuse. One might also benefit by knowing if they had a difficult day or were desperate to make an appointment. Mike Ziegler was one of my early mentors in science. When I was a medical student in Ziegler’s lab in Galveston, he taught me basics of the catecholamine and serotonin neurotransmitters and their biosynthetic enzymes essential to monoamine neurotransmission. Ziegler was an expert on hypertension, but these phylogenetically ancient neurotransmitters are also foundational to arousal, impulsivity, aggression, and indeed most processes of emotion and cognition. After I left Texas to join the NIH as a postdoctoral fellow, I was then startled to spot a small Ziegler-relevant item in the Washington Post which I read thoroughly, in
64
6. Measurements of capacity
search of mischief. The item was a one-paragraph report concerning one of his former associates at NIMH, an outstanding neuroscientist. One random day this colleague was trying to park at Bethesda Naval Hospital (now Walter Reed) across the street from NIH when someone snuck in and stole “their” space. All this happened long ago, in that better time before NIH and Walter Reed were enclosed by high fences to keep out the unwanted, and the full facts will never be known. However, the news headline was “Psychiatrist Kicks Lady” (or woman—memory fails). The circumstances that brought the psychiatrist and the other driver together were unpredictable but predictively unpredictable. Sometimes an insect will land on our nose. In an instant one mistake can undo us, and impulsivity makes errors more likely but failure to venture out onto life’s high wires makes for a dull existence indeed. Ernie Barratt’s scale, the Barratt Impulsiveness Scale (BIS), makes some sense out of impulsivity by breaking it down into its components, or more accurately, different flavors, each with a potentially different neurobiological meaning, evolutionary origin, and meaning—when we come to consider the meaning of impulsivity for choice. In this way, the BIS presaged later constructs of impulsivity, notably by Trevor Robbins, who in rodents discerned six varieties of impulsivity that could be delineated via tests in the laboratory setting (Dalley and Robbins, 2017). One cannot ask a mouse how it feels expecting an intelligible response in any known language, but in humans with language the BIS questionnaire measures attentional impulsiveness, motor impulsiveness (the tendency to act “without thinking”), and lack of forethought and future planning. Surprisingly, given the questionable validity of impulsivity as a unitary construct, measures of impulsivity using questionnaires tend to align, and genes drive them. This alignment was seen by observers of human nature long before the advent of psychological assessment in the 20th century. Humans are a social species and success in the social milieu depends on the ability to make behavioral predictions. To do this we perform a state assessment of another’s behavior of the moment, and the nature of the circumstances or provocation. However, our experience also teaches us to perform a trait assessment of the person. Acute observers recognized that some people were obviously more thymic, choleric, sanguineous, and short-fused, and others less. We may recognize personality type from long observation, by secondhand reports or even by first impressions of a person’s body habitus, dress, tone, or the lines on their face. The ability to better predict the behaviors of a person does not necessarily mean that their behavior is neurogenetically determined, but directly establishes that behavior is constrained by the nature of the brain generating it and the immediate circumstances. Measurement by life history: When the person is old enough to come to the attention of the criminal justice system, assessing whether they have been
Measuring impulsivity and aggression
65
convicted of violating a law is procedurally much easier than measuring whether they have an aggressive or impulsive temperament. Therefore criminality is a crude measurement of behavior, but produces data more easily collected than some of the procedural assessments of impulsivity and aggression that we will discuss. Also, criminality represents the expression of predisposition over years of a person’s life, albeit an expressed behavior contaminated by many other factors. Some genetic studies on aggression have studied criminal violent behavior, and there is evidence of moderate heritability. In other words, it is true that criminality tends to run in families and that genetic factors are involved. A similar approach to measuring the “genetics of aggression” is to measure lifetime aggressive behavior. The Brown–Goodwin scale is one way of doing that. Like recording criminal convictions for aggressive or impulsive behavior, the Brown–Goodwin scale measures outcome rather than propensity. In the 1980s, the Brown–Goodwin scale, used in the MAOA by testosterone interaction study in Chapter 4, was created by my colleague and collaborator, LaVonne Brown. Brown, a gifted child psychiatrist, conceived the idea that instead of asking people questions about their propensity for violence he could learn more by measuring their history of it—because impulsive people accumulate many such incidents over a lifespan and generally manifest such tendencies early. Frederick Goodwin, an outstanding psychiatric researcher, was former director of the National Institute of Mental Health (NIMH) where he assembled an amazing and never to be matched concentration of the world’s foremost psychiatrists and neuroscientists. Using the Brown–Goodwin scale, Brown, Goodwin, and others conducted pioneering studies on the relationship of biological factors, including serotonin, to aggression. Measurement by experiment: More recently, the assessment of impulsivity and aggression has been greatly advanced by experimental measures that allow direct measurement under controlled circumstances. Such “human laboratory measures” parallel how the impulsivity, arousal, and novelty seeking of the Htr2b gene knockout mouse were measured. The person, like the mouse, is placed in a well-defined experimental situation, where more extraneous variables can be controlled. The brain can be imaged during such testing, revealing predictive differences in regional metabolic activity. These differences are telling us that some people’s brains work differently and localize brain regions that are important in inhibiting impulse. Other brain imaging studies involving probes for response to emotion and reward have revealed that people also differ in the activity of neural systems that mediate emotion, fear, and craving. Thus there is a balance between stimuli and brain systems that create behavioral urges and systems that enable us to moderate, delay, sublimate, or suppress these urges. Point-subtraction aggression paradigm: How aggressively will a person behave toward an anonymous rival? At the University of Houston, Don
66
6. Measurements of capacity
Cherek set out to test this behavior in people, and discovered that in an experimental setting many people with aggression and personality disorders would react very aggressively and even initiate aggression. In Cherek’s point-subtraction aggression paradigm (PSAP), the experimental subject is put in front of a computer with an option of pressing three buttons—one to earn money, one to subtract money from another research subject, and one to protect themselves from attacks. As is common in experimental psychology, and as may eventually destroy experimental psychology, the subject is deceived because there is no second person (Gerard Moeller et al., 1997). The PSAP, like the famous Milgram experiment in which participants were induced by the scientist in the white coat who was physically present to deliver shocks of increasing intensity, is deceptive because no one is in the adjacent room. In the Milgram experiment, 50% of people could be induced to deliver high intensity shock. That was a disturbing result, with broad implications for people’s willingness to follow orders and abdicate responsibility. In line with tit-for-tat principles, the most rational PSAP strategy is to press the button that generates rewards and only press the button protecting against attacks or retaliate if attacked. During the experiment, the experimental subject will be attacked leading to protective and retaliatory responses. Game theory teaches that it is rational to retaliate. However, some people become so irritated that they retaliate massively and disproportionately—“tits-for-tat,” one might say—and ultimately to their own detriment. Such behavior is highly reminiscent of revenge and vengeance behaviors that sustain needless ethnic and political conflicts in most parts of the world. There is something within many that makes them want to hit back harder or even disproportionately. As Gandhi memorably said, “An eye for an eye and pretty soon all the world is blind.” However, many people act, and even come to believe, that if you take one of ours, we should take 10 of yours. Even worse, and as Larry Siever, a world expert on impulsive and aggressive behavior at Mount Sinai, observed, when placed in this situation an exceptional few ASPD, psychopathic, bored, wealthy, or foolish individuals will immediately and without provocation begin pressing the button to subtract points from their imaginary opponent. Asked why, one said, to show they meant business, and intimidate their “opponent.” Thus even in a seemingly simple situation, divorced of most of life’s complexities, a person makes choices reflective of their innate propensities and the attitudes and strategies they apply on a given day. The Milgram experiment opened many eyes to the possibility of ordinary people doing evil, but people’s behavior in the PSAP is in its own way disturbing. Only money is involved, but on the other hand the experimental subject is following no one’s orders and there is no authority figure urging them on. In the social calculus, it is an unfortunate fact that we must account for avengers and intimidators. Indeed, if all the rest of us insisted
Measuring impulsivity and aggression
67
on retribution even just equivalent to the offense, the only p ossible outcome would be that all the world would become blind. Despite the human desire for revenge and retribution, justice must be tempered and perhaps the key achievement of civilization has been that retribution is usually in the hands of cooler heads motivated more by reason than by passion. As is worth mentioning a second time, both the PSAP and the Milgram experiment involve deception. The participant is only briefed afterward about what really happened. In human research, deception should be avoided whenever possible, even if it appears that the information to be realized is important. When a researcher lies to a person or is “economical with the truth,” they breach obligations to allow the human participant maximal autonomy as to whether to participate. Deception ultimately erodes the bond of trust between people and researchers. Delay discounting: To what extent does one discount the value of a larger reward with a longer waiting time versus one when a smaller reward might be more immediately accessed? As seen in our study of the Htr2b knockout mouse, delay discounting is a powerful measure of impulsivity in that animal model. The more impulsive mouse is unwilling to delay gratification for the same interval of time, even though it has been trained to understand that a longer wait will result in a larger reward. However, it is important to understand that a person’s willingness to delay gratification can be affected by more than genotype. It can also be affected by their experience and their perception of the environment. This could happen in the laboratory, but is especially true in real-life delay discounting situations, where the delays are frequently longer and the reward at stake is larger. If one lives in a chaotic environment where it is possible that the holder of the reward may skip town or conveniently forget their obligation, or where the reward is acutely needed in the short term, it may make more sense to put these factors into the mental calculus and take the smaller but more immediate reward. However, impulsive individuals, whether their impulsivity has a genetic or an environmental origin, have a pervasive tendency to take the smaller but more immediate reward, even in a laboratory setting where the longest delay is only a matter of minutes. Go–NoGo: What happens if a person is given an easy task requiring them to respond on certain trials and withhold from responding on others? Do impulsive people have problems withholding responses? It turns out that the answer is yes. Impulsivity measured in even such a simple context correlates with impulsive personality disorders and predisposition to addictions, which are obviously much more complicated behaviors to evaluate. The Go–NoGo test has many variants, including a new “Emotional Go–NoGo” that is being used to explore aspects of emotionality and emotional differences between people. In its form, the Go–NoGo task might include 100 trials, 60 calling for a “Go” button press and 40 calling for a “NoGo,” inhibition of response. The scientist
68
6. Measurements of capacity
can obtain several measures from this simple test, including the reaction time and the number of false alarms in response to NoGo signals. It is these false alarms that indicate disinhibition and impulsivity. Males and people at risk for addictions are, among others, more likely to show the disinhibited false alarm responses. Iowa gambling task: The Iowa Gambling Task was, not surprisingly, developed at the University of Iowa and has penetrated somewhat into public consciousness via its discussion in Antonio Damasio’s book Descartes’ Error (Damasio, 2005). The task involves four virtual card decks represented on a computer screen. The object of the game is to win money, and when a card is drawn the result may be to add or subtract. The four decks are not created equal. After a time, most people figure that out and begin exclusively sticking to “good decks,” but even before they are consciously aware of this, the “bad decks” begin generating higher stress responses, for example in terms of galvanic skin conductance when the computer cursor hovers over a “bad deck.” In contrast, people with frontal lobe dysfunction will continue to play the “bad decks” even though they are aware they are losing money by playing those decks and fail to show the stress responses associated with “bad decks.”
Integrating measures and genes Genes and neural systems play roles in diverse behaviors. Furthermore, impulsivity and aggression themselves have multiple origins at the level of attention, motivation, and emotionality, which are the underlying neurobiological domains. The mediating processes are themselves modified by multiple genes. Therefore if neuroscientists, who are even more impossible to herd than cats, were somehow all organized to desist from research on impulsivity and aggression or if a “ban” on such research were imposed, advances in the area would continue to proceed at about the same pace, albeit with different labeling. The neural nexus of integration of much work on impulsivity is the frontal lobe of the brain and its functions. In 2005 the US Supreme Court ruled that the execution of convicted teenagers was unconstitutional, citing their “underdeveloped sense of responsibility.” They were right. The frontal lobe, which mediates that cognitive control, is one of the last regions of the brain to complete its development, and genes and interacting environmental factors affect the rate and character of development of this critical brain region. The relative volatility and impulsivity of adolescents and children will need to be discussed further in the context of childhood psychiatric diagnoses involving dyscontrol, irritability, diminished attention span, and impulsivity.
Integrating measures and genes
69
The frontal lobe can be damaged by closed head injuries from rapid decelerations in collisions and falls, and blast injuries that many soldiers suffer. The frontal lobe is particularly vulnerable to coup/countercoup injury resulting from rapid acceleration/deceleration of the soft brain against hard bone. A safety principle behind auto seatbelts, airbags and crumple zones, and the helmets worn by cyclists is to absorb the energy of deceleration and to modify the effects of deceleration. Increasing attention has been turned to the protection of the brains of boxers, football players, and other athletes because the concussions and knockouts (KOs) they experience are due to decelerating injury, and autopsy studies have confirmed that the damage to their brains is cumulative and lasting. Frontal lobe damage is also part of the global damage caused by Alzheimer’s disease, a genetically influenced neurodegenerative disease that afflicts millions, and Pick’s disease, a much rarer neurodegenerative disease that, for reasons unknown, specifically affects this region of the brain. Two famous examples of damage to the frontal lobe are Phineas Gage and James Brady. Gage was a 25-year-old railroad construction foreman responsible for tamping the gunpowder, sand, and fuse into blasting boreholes. Something went wrong and the powder exploded, catapulting the iron tamping bar through the side of Gage’s face, out the back of his left eye, through the top of his head, and then through the air about 80 ft further. Later in his life Gage usually carried around the tamping bar, which was about three and a half feet long, and an inch and a quarter in diameter. The bar, which was smooth and had a javelin-like tapered point, destroyed Gage’s frontal lobe, although we do not know the exact extent of the injuries. After the accident, Gage spoke within minutes and lived another dozen years. Gage’s poorly documented behavioral changes, a story further distorted by the passage of a century and a half, made him a teaching example of the behavioral consequences of frontal lobe injury. The most well-known modern example of frontal lobe injury is Presidential Press secretary James Brady. Brady “took a bullet” for President Ronald Reagan. Brady’s suffering and behavioral change also had consequences, helping to launch a movement to restrict access to firearms. In 1981 one bullet fired by John Hinckley, Jr., shattered Brady’s left frontal bone, damaged his left frontal cortex, and lodged in his right frontal cortex, which ended up being destroyed. Brady was partially paralyzed for the rest of his life, but also had behavioral manifestations of frontal lobe injury including personality changes. Although neuroscience, and in particular the public perception of neuroscience, is strongly influenced by memorable anecdotes (anecdata) such as the stories of Gage and Brady, it progresses through systematic studies. The information base on effects of injury to the frontal lobe, and regions within it, is deep and wide. In the USA, millions of individuals experience
70
6. Measurements of capacity
some frontal lobe injury, and thousands have participated in systematic and detailed studies. Such studies use imaging techniques and postmortem autopsy studies on the brain itself to evaluate the extent of injury, and neuropsychological testing—including many new computer-based methods—to assess premorbid capacities, cognitive and emotional performance deficits, and recovery. New capabilities in functional brain imaging have enabled neuroscientists to tie frontal cortical inefficiency to cognitive deficits. Behavioral pharmacology, and to some extent genetics, have revealed roles of neurotransmitters in frontal function. Frontal lobe deficits include reduced attention span (attention deficit), poor working memory (short-term memory), and difficulty in executive cognitive functions, the key to which is the ability to switch between cognitive strategies. This leads directly to problems with planning, reasoning, and ability to inhibit emotions. Frontal lobe damage often leads to perseverative behavior and to inappropriate aggression and sexual behavior. It can also lead to Witzelsucht—“jokiness”—the trait of constantly, and often inappropriately, injecting humor into situations and the telling of pointless stories. Thinking about the prevalence of some of these behaviors, we can see the danger of neuroscience by anecdote. How easy it is to identify someone who is “uninhibited” sexually, perseverative (for example, completing a book despite indicators of futility), witty, or being prone to telling pointless stories. Weak genetic predictors of functional consequences of effects on the frontal lobes are now known in development and in head injury. In head injury, apolipoprotein E (APOE), a gene that predicts part of the vulnerability to Alzheimer’s disease, was shown to predict the cognitive impairments that occur in boxers. In professional boxing the head is unprotected; the head is a legitimate target. A KO is achieved by damaging the opponent’s brain. Recently, it has come to the attention of neurologists that football players have a very high frequency of head injury. The reason consciousness is disturbed (the boxer was hit and became “woozy” or a wide receiver was knocked unconscious) or lost is because of deceleration injury to innumerable neural networks, and this more frequently occurs without loss of consciousness or concussion syndrome. Later, cells die. Although frontal cognitive performance tends to be worse in head-injured patients, the combination of head injury and the cognitively unfavorable APOE genotype, and the catechol-O-methyltransferase (COMT) genotype, leads to worse outcome.
Measuring the brain Hare and Barratt were both pioneers in the science of motivation, looking beyond and from action and inward to cause, and one measuring psychopathy and the other impulsivity. In looking to motivation, both were
Measuring the brain
71
prescient, foreseeing a time when with brain imaging and genetic predictors it might be possible to understand internal states and behavioral origins, and as Freud had attempted to do in an earlier era prior to decades where experimental psychology dominated behavioral science. In the late 19th century, the psychologist William James had argued for free will based on internal state, “My first act of free will shall be to believe in free will” (Perry, 1935). This type of thinking was antithetical to the experimental psychologists who would dominate the first half of the 20th century. Inspired by Pavlov’s experiments conditioning dogs, John B. Watson published “The Behaviorist Manifesto” in 1913, writing. Psychology as the behaviorist views it is a purely objective experimental branch of natural science. Its theoretical goal is the prediction and control of behavior. Introspection forms no essential part of its methods, nor is the scientific value of its data dependent upon the readiness with which they lend themselves to interpretation in terms of consciousness. The behaviorist, in his efforts to get a unitary scheme of animal response, recognizes no dividing line between man and brute. The behavior of man, with its refinement and complexity, forms only a part of the behaviorist’s total scheme of investigation.
Watson’s objectivist approach set the stage for the radical behavioral determinism of Skinner that will be discussed further in this book, and for many was the death knell of free will. If one only measures external behavior, it is far easier to believe that behavior is determined rather than merely influenced by determinants. Behavior devolves into movements in a maze or open field, or lever presses. However, the neural basis of even the simplest human behavior is irretrievably complex and multilayered, with many pathways of top-down modulation. Thereby, the exact same behavior, for example a lever press, or an approach to a novel object, can emerge in a myriad way, so that two decisions to press a lever may be less equivalent than a decision not to press. By studying impulsivity and psychopathy, Barratt and Hare were uncovering differences in motivation. Motivation matters. We may observe that from childhood a person has a history of stealing, lying, torturing animals, drug use, and sexual promiscuity. Someone who has examined the child or who knows the child may report a variety of explanations for the behaviors, ranging from psychopathy and lack of conscience to attention seeking, cry for help, or whatever. The child may themselves believe one of these explanations, but that does not make it true. Prior to directly accessing activity of brain circuits, as Barratt was beginning to do via measurements of the electroencephalogram, or genes that influence them, classifications of antisocial and impulsive behaviors that depend on externally observed behavior have worked better. However, experimental psychology is now coming full circle and— in back to the 19th century fashion—with ever more detailed measures of
72
6. Measurements of capacity
neural circuit function during behavior. Neuroimaging connectivity, electroencephalography, magnetoencephalography, and even direct measurement of local neural field potentials via electrode arrays emplaced in the brain are accessing the activity and functional connectivity of the brain. Psychophysiological tests are widely employed to detect lies, wherein it is seen that although two people are answering a question identically one is truthful, but the other is lying, and furthermore people responding identically are widely divergent in how they decided to do so. The polygraph, invented almost a century ago by an American medical student and a police officer, typically measures respiratory rate, pulse/blood pressure, and skin conductivity. Nowadays, many other tools are available, including functional magnetic resonance imaging and the electroencephalogram. Polygraph testing is controversial, and some prominent neuroscientists, for example Mark Gazzaniga, have cast doubt on the future of the technology. The National Research Council reviewed polygraph testing in 2003 (https://www.nap.edu/read/10420/chapter/1). As reviewed by the National Academy committee, accuracy rates in mainly substandard studies ranged from 80 to 98%, overall, about 90%. The NRC’s report issued shortly after 9/11 and in which Wonder Woman even makes a cameo appearance, soberly concluded that polygraph testing distinguishes lies from truth “at rates well above chance though well below perfection,” and further that the accuracy of polygraph tests may be degraded by countermeasures, thus limiting their value in security screening. However, the science of psychophysiological assessment constantly advances. Multiple fMRI studies after the NRC report, and for example Ofen, et al. (2017), have found that lying and preparing to lie require activation of frontal and parietal cortical regions. fMRI and brain electrophysiology are not yet in wide use for lie detection, except as seen in science fiction. However, higher resolution scanners, scans performed during cognitive and emotional challenges to more incisively probe brain responses, and cheap, portable MRI machines and other devices such as near infrared scanners are becoming available. Wonder Woman’s rope remains elusive, but fortunately, or more likely, unfortunately, it is likely that we will become more and more accurate in the ability to detect deception. Another avenue to the inner workings of the brain is that certain so-called truth sera (e.g., sodium pentothal) that disinhibit behavior also can bring down a person’s guard, and information is more readily extracted, whether the person is psychopathically on guard or withholding for other reasons. It is generally felt that administration of such drugs during interrogations represents a form of torture and a violation of Fifth Amendment rights. Furthermore, the main effect of such drugs is to disinhibit responses of all kinds, both truth and lies. External behavior is, like an island, only the smallest and least interesting part of a large underwater mountain range that may be responsible for many small islands, and seamounts that do not quite broach
The arousal (thymos) of youth
73
the surface. When in the late 20th century experimental psychologists began using neuropsychology tools to look below the surface, they found that not only was there much more to be seen, but that often the reality was the opposite of what could be externally observed. For example, when studies of executive decision making were performed it could be seen that the brains of many people who made bad decisions were working overtime to make good decisions. This paradox will be discussed in Chapter 13, in which the action of a common polymorphism of the COMT gene to alter concentrations of dopamine in frontal cortex. If dopamine levels are suboptimal in frontal cortex, the seat of executive cognition, the brain must work harder, as was observed by magnetic resonance imaging (MRI) to generate equivalent performance on tasks such as the N-back (in which digits are remembered), Wisconsin Card Sort Task (assessing perseverative errors in a card classification task), or digits forward/reversed memorization, and as we showed was influenced by genotype. On the emotion side, and as again will be discussed in Chapter 13, two people may seem to be reacting similarly to images of faces and things, or to a painful stimulus, but MRI of the brain can reveal that one of them is responding far more strongly. Neural circuits, and the genes that enable their formation and influence their activities, predict behavior, but in the end the person may choose their behavior, even reshaping those circuits, by marshaling other resources, or they may choose not to do so. Following in Skinner’s footsteps, their first act of unfree may be to disbelieve in free will, but even then, and as we will later show via Skinner’s biography, they will be hard-pressed to unshackle themselves from the responsibility of having free will.
The arousal (thymos) of youth The young of most mammalian species are impulsive and disinhibited. They tend to play and explore, and—as they get a little older—to fight and engage in other dangerous activities that an older individual might consider reckless. As shown by developmental studies, inhibitory connections between frontal cortex and emotional and motor centers of the brain are not fully developed in adolescents. Furthermore, the “motor” or energy driving behavior is far stronger in youth, as is easily observed whether in children, pets, livestock, or animals in their natural environment. The Greeks named this innate energy thymos θυμός and it was recognized by Plato as one of the three constituents of the human psyche, the others being desires and reason. Innately, some species, and people, are more spirited, and one might even say “wild.” In my work on impulsivity, I have focused more on this side of the equation, rather than the top-down control mediated by the frontal lobes. Genes acting on decision
74
6. Measurements of capacity
making—determining choices if one will—are working within this developmental context. In effect, and as recognized both in law and social customs, decision making of the young is impaired by the greater thymos and still not fully developed top-down cognitive control of youth. Youth are more likely to volunteer for combat—and we take advantage of that—but they are also more likely to make unwise choices from which we attempt to protect them. In general (when it does not suit us) we try to help delay decisions that are irreversible. Thymos—of thalamic and midbrain origin, and deficits of executive cognition—of frontal lobe origin, do not destine people to make bad choices or only impulsive choices. Marsha Linehan invented dialectic behavior therapy, a cognitive therapy, for people with borderline personality disorder. BPD is a common, but underdiagnosed, psychiatric disease marked by emotional volatility leading to shifting, unstable personal relationships, veering from strong and unrealistic attachment to disillusionment and despair. Para-suicidal behavior is common, including cutting. BPD is strongly cross inherited with another personality disorder that is often marked by hyperarousal, namely antisocial personality disorder. On a closed psychiatric unit, such patients may be seen walking from one end of the space to another, much as a caged tiger may endlessly circle the edge of the enclosure. Any of these behaviors may be “controlled” by medications, and some drug users are self-medicating hyperarousal—however the insight of Marsha Linehan was to realize that emotionally volatile BPD patients could be helped— and help themselves—via a cognitive therapy approach that uses their capacity for reason.
Animal models of arousal, impulsivity, and aggression About the inevitable integration of genes into the equation, it is also important to understand the importance of studies conducted in animal models, and to understand the limitations of these models that obviously are not candidates for DBT. Animal models, especially those using rodents, have enabled invasive and genetically and environmentally well-controlled studies free of many of the confounding factors that impede human studies. Anyone acquainted with the differences in temperament between various breeds of dogs would understand that behavior is strongly determined by genetic factors, as illustrated by the following three jokes, two involving dogs and one a fox: Joke 1. Customer (timidly) to innkeeper: “Does your dog bite?” … “No?” (relieved) … “Ow! I thought you said your dog didn’t bite!” Innkeeper: “That’s not my dog.” Joke 2. Mailman (angrily): “I thought you said your dog’s bark was worse than its bite!”
Animal models of arousal, impulsivity, and aggression
75
Dog owner: “Wait ‘til you hear its bark.” Joke 3. Fox (puzzled): “Scorpion, you promised not to sting if I carried you across the river. Now we both drown. Why did you?” Scorpion: “It is my nature. I am a scorpion.” Unlike differences in behavior between humans, which can be environmentally determined, the behavioral differences between species and breeds are primarily genetic in origin. Of course, animal behavior is not always as clear-cut and predictable as that of the scorpion. Often, the results are muddled. Many studies conducted on aggression in mice and rats are inconsistent and difficult to interpret. It is hard to understand what is going on in a mouse’s mind (probably not much) when it is confronted with the unusual challenge of a second mouse dangled by its tail into its cage or when a second free moving “intruder mouse” is placed in its plastic box. The mouse is a poor model for primate aggression, and although the aggression scenarios are controlled, they are highly artificial. The main problem is that these sorts of “make the mice fight” experiments yield different results depending on the test. With one test one strain is more aggressive and with a different test another is more aggressive. Despite limitations of animal models, most of what we know about behavioral pharmacology and neurocircuits involved in reward, learning, emotionality, and behavioral control are from studies conducted in the mouse and rat, which it should also be said are from an evolutionary perspective no lowlier than the human, just differently adapted. Furthermore, there are now new insights into the origins of impulsive behavior in rats, from a group based at Oxford University and led by Trevor Robbins and Barry Everitt. The impulsivity has more than one origin, involving both an urgency component attributable to a region of the brain known as the striatum and a control component attributable to the frontal cortex. Importantly, the underpinnings of the impulsive behavior can be traced back to the dopamine neurotransmitter system and its receptors, and as discussed elsewhere in this book, other neurotransmitters such as serotonin also play important roles. We have come to understand that impulsivity emerges not from individual neurotransmitters but from the function of neurocircuits, as shown by dysfunction caused by frontal lobe damage. Measurement of frontal function enables prediction of impulsive behavior and cognitive control. With the methods that are available we can say that many people are impulsive not just because they are not trying hard enough or want to be that way, but because that is how their brain works. As will be discussed later, brain imaging studies are showing us that the brains of many people with frontal lobe impairment associated with a genetic variant in COMT are working harder, but not keeping up. At the genetic level, the animal models have also pointed directly to certain neurochemical factors that are important in neural pathways that modulate behavior, and that in turn has led to studies of variation in genes such as MAOA, COMT, and HTR2B.
76
6. Measurements of capacity
References Babiak, P., Hare, R., 2006. Snakes in Suits: When Psychopaths Go to Work. Harper Business. Cloninger, C., et al., 1981. Inheritance of alcohol abuse: cross-fostering analysis of adopted men. Arch. Gen. Psychiatry 38, 861–868. Dalley, J., Robbins, T., 2017. Fractionating impulsivity: neuropsychiatric implications. Nat. Rev. Neurosci. 18, 158–171. Damasio, R., 2005. Descartes’ Error. Penguin Books. Gerard Moeller, F., Dougherty, D.M., Rustin, T., Swann, A.C., Allen, T.J., Shah, N., Cherek, D.R., 1997. Antisocial Personality disorder and aggression in recently abstinent cocaine dependent subjects. Drug Alcohol Depend. 44 (2–3), 175–182. ISSN 0376-8716. Hare, R., 1993. Without Conscience: The Disturbing World of the Psychopaths Among Us. Taylor and Francis. Higley, J., et al., 1993. Paternal and maternal genetic and environmental contributions to cerebrospinal fluid monoamine metabolites in rhesus monkeys (Macaca mulatta). Arch. Gen. Psychiatry 50, 615–623. Karlsson Linnér, R., Biroli, P., Kong, E., SFW, M., Wedow, R., Fontana, M.A., Lebreton, M., Tino, S.P., Abdellaoui, A., Hammerschlag, A.R., Nivard, M.G., Okbay, A., Rietveld, C.A., Timshel, P.N., Trzaskowski, M., Vlaming, R., Zünd, C.L., Bao, Y., Buzdugan, L., Caplin, A.H., Chen, C.Y., Eibich, P., Fontanillas, P., Gonzalez, J.R., Joshi, P.K., Karhunen, V., Kleinman, A., Levin, R.Z., Lill, C.M., Meddens, G.A., Muntané, G., Sanchez-Roige, S., FJV, R., Taskesen, E., Wu, Y., Zhang, F., 23and Me Research Team; eQTLgen Consortium; International Cannabis Consortium, Social Science Genetic Association Consortium, Auton, A., Boardman, J.D., Clark, D.W., Conlin, A., Dolan, C.C., Fischbacher, U., PJF, G., Harris, K.M., Hasler, G., Hofman, A., Ikram, M.A., Jain, S., Karlsson, R., Kessler, R.C., Kooyman, M., MacKillop, J., Männikkö, M., Morcillo-Suarez, C., MB, M.Q., Schmidt, K.M., Smart, M.C., Sutter, M., Thurik, A.R., Uitterlinden, A.G., White, J., Wit, H., Yang, J., Bertram, L., Boomsma, D.I., Esko, T., Fehr, E., Hinds, D.A., Johannesson, M., Kumari, M., Laibson, D., PKE, M., Meyer, M.N., Navarro, A., Palmer, A.A., Pers, T.H., Posthuma, D., Schunk, D., Stein, M.B., Svento, R., Tiemeier, H., PRHJ, T., Turley, P., Ursano, R.J., Wagner, G.G., Wilson, J.F., Gratten, J., Lee, J.J., Cesarini, D., Benjamin, D.J., Koellinger, P.D., Beauchamp, J.P., 2019. Genome-wide association analyses of risk tolerance and risky behaviors in over 1 million individuals identify hundreds of loci and shared genetic influences. Nat. Genet. 51 (2), 245–257. https://doi.org/10.1038/s41588-018-0309-3. Epub 2019 Jan 14. PMID: 30643258; PMCID: PMC6713272. Nutton, V., 2005. The fatal embrace: Galen and the history of ancient medicine. Sci. Context. 18 (1), 111–121. Ofen, N., Whitfield-Gabrieli, S., Chai, X.J., Schwarzlose, R.F., Gabrieli, J.D., 2017. Neural correlates of deception: lying about past events and personal beliefs. Soc. Cogn. Affect. Neurosci. 12 (1), 116–127. https://doi.org/10.1093/scan/nsw151. PMID: 27798254; PMCID: PMC5390719. Perry, R.B., 1935. The Thought and Character of William James, p. 323.—Letters of William James 1, p. 147. Watson, J.B., 1913. Psychology as the behaviorist views it. Psychol. Rev. 20 (2), 158–177. https://doi.org/10.1037/h0074428. hdl:21.11116/0000-0001-9182-7. S2CID 145372026. Zhou, Z., Blandino, P., Yuan, Q., Shen, P.H., Hodgkinson, C.A., Virkkunen, M., Watson, S.J., Akil, H., Goldman, D., 2019. Exploratory locomotion, a predictor of addiction vulnerability, is oligogenic in rats selected for this phenotype. Proc. Natl. Acad. Sci. U. S. A. 116 (26), 13107–13115. https://doi.org/10.1073/pnas.1820410116 (Epub 2019 Jun 10).
7 Distorted capacity: Neuropsychiatric diseases and the impaired will I’m only happy when it rains. Garbage—“I’m Only Happy When it Rains” There are two colors in my head. Radiohead—“Everything in its Right Place” One reason psychiatric diseases are feared and stigmatized is that they often impair the capacity for choice. Distortion of capacity—craziness if you will—is a disturbing prospect. We try to help people who suffer in such ways. The alternative, which has been called antipsychiatry, is that we can eliminate the stigma by ridding ourselves of the idea. From the perspective of this intellectual hillock, psychiatric diseases represent the reclassification—or reification—of constellations of socially unacceptable behaviors as entities subjectable to medicalized control. Are psychiatric disorders valid disease entities? To what extent do psychiatric diseases predict behavior, and why? Which psychiatric diseases distort or impair volitional control? In this chapter we will see that having defined diseases on the basis that they alter behavior rather than by process, it is unsurprising that these diseases, and their alleged vulnerability genes, have become a focus of controversy. Thus substance use disorders are demarcated first by use and diseases such as schizophrenia, depression, and anxiety disorders by symptoms experienced and their effect on the lives of patients, and others, not by a brain scan, physiological measure or genotype.
Impulsivity, diminished capacity, and neuropsychiatric disease Neuropsychiatric illnesses associated with diminished capacity are of two main types: those that increase arousal—the impulse for a behavior— and those that diminish capacity to inhibit and evaluate. Sometimes, as
Our Genes, Our Choices https://doi.org/10.1016/B978-0-443-22161-3.00004-1
77
Copyright © 2024 David Goldman. Published by Elsevier Inc. All rights reserved.
78
7. Distorted capacity
with addictions and schizophrenia, both simultaneously come into play. Also, either may play a role at one time but not another. One reason to study the neurogenetics of impulsivity is to come to a more unified and accurate understanding of what is happening in brains of people with psychiatric disorders, but another is to better understand behavioral causality in people without. The behaviors that signal psychiatric diagnoses are telling us that there is an underlying problem, and psychiatric diagnoses identify groups of people who in common tend to benefit from the same medical interventions—both pharmacological and nonpharmacological. In thinking about the relationships between impulsivity, behavior, and free will, each of which is transdiagnostically important, we can be informed by disease classifications, but not bounded by them. Indeed, the diagnoses are “fuzzier” than we would like and over time have proven to be mutable or even invalid. A major task of biological psychiatry is to use new tools and information to transform and sharpen the process and categorization of psychiatric pathology for example in DSM-5 (The American Psychiatric Association, 2022). Psychiatric disorders that can affect competence to choose will be briefly sketched here toward the goal of understanding that there is a broad relationship between diagnosis and capacity, with important nuances. Probably the most important nuances are that diverse psychiatric disorders can impair capacity, there is wide variation in the effect of the same diagnosis on decision making, and even the same person may at some point in their life or in some situations be competent to decide, and in other cases be incompetent.
Affective disorders Depression and anxiety disorders are a broad group of diagnoses primarily defined by disordered mood (emotion) and affect (expression of emotion) that afflict a third or more of the population. Several are related in terms of genetic transmission and shared genes for genetic risk and by neurobiology, for example at the level of brain circuitry, neurochemistry, and neuropharmacology. Along with the addictions, mood and anxiety disorders make up the bulk of psychiatric practice. However, despite the abundance and causal complexity of these disorders, the clinical pharmacology of these disorders, and psychiatric disorders in general, is not very complicated, for example as compared to disorders treated by a doctor of internal medicine. When I was a resident, it seemed that psychiatry was like dermatology: “If it is wet, dry it. If it is dry, wet it. If that doesn’t work, try steroids.” Of course, that was never true of dermatology, where one must be constantly on the lookout for many specific problems, including skin cancers and signs of other serious diseases; nor in psychiatry, where the problem might well be an endocrine disturbance, drug induced, or reflective of an unseen cancer or infection.
Impulsivity, diminished capacity, and neuropsychiatric disease
79
Increasingly, psychiatry looks more than “skin deep” and accesses a range of therapeutic options. Still, it is amazing how often the words “Prozac” and “talk to the patient” summarize the therapy of mood and anxiety disorders. Psychiatry has further to go. The mood and anxiety disorders include major depression, dysthymia (milder depression), panic disorder, generalized anxiety, and phobic disorders. They are genetically influenced and environmentally programmed. These disorders of mood can shape the thinking of the person at an unconscious level. Therefore it does not much help a depressed person to tell them to “cheer up.” As Helen Mayberg and Sid Kennedy have shown, there is a neurocircuitry of depression. In groundbreaking studies in which they stimulated a specific region of the frontal cortex with a deep electrode, they and their neurosurgical colleagues brought into remission several patients whose severe depressions had been resistant to all other therapy. During the actual surgery and at the appropriate depth and frequency of stimulation at a specific location in the frontal lobe, the depression of some patients lifted, they remained free of depression for extended periods, and depression returned if the battery failed. Similarly, there is a well-validated neurochemistry of depression. People whose catecholamine neurotransmitters were depleted by reserpine, a drug used to treat high blood pressure, began to exhibit signs of depression: decreased mood, appetite, interest in sex, and energy. Most tellingly, they started having depressed thoughts. In other words, while it is true that depressing events and thinking can make us depressed, it is also true that biochemical depression can drive depressed cognition and other symptoms such as lack of energy, appetite, and interest in sex. It is a vicious cycle. Also, through the work of Rene Hen at Columbia we know that the therapeutic effectiveness in depression of selective serotonin reuptake inhibitors (SSRIs) such as Prozac (fluoxetine) depends on the ability of these drugs to release brain-derived neurotrophic factor, a protein that stimulates neuroplasticity and neurogenesis (the production of new neurons) in the adult brain. One of the virtues of antidepressant treatments including medications and electroconvulsive therapy, which works in about 95% of depressed patients and has saved thousands of lives, is that these treatments also address the cognitive aspects of depression. Unfortunately, and as has been a special concern in adolescents, who, as discussed, have less well developed executive cognitive control and are on the other hand more aroused, depressed patients sometimes commit suicide after treatment, and during therapy as they are beginning to recover. Bipolar disorder, affecting approximately one in 200 individuals, is the most mysterious of the mood disorders. As indicated by its name, bipolar disorder leads to alternating cycles of depression and mania, when the person becomes energetic, hypersexual, grandiose, and delusional. In the manic phase there is pressure of speech and flight of ideas, but no disorder of thought except for the coloring of thought by happy optimism
80
7. Distorted capacity
and whatever delusions may be present. Manic episodes are more likely to occur in early adulthood, and as the person becomes older episodes of major depression predominate. One prominent psychiatrist, who of course studied bipolar disorder, illustrated its effects. “Geraldo” was usually manic, which added to his brilliance, creativity, and energy. On the other hand, they were irritable and grandiose, and not a particularly good sport (except when winning!). Like the Russian grandmaster Bogolyubov, Geraldo might have leapt atop the table, hurled their opponent’s queen against the wall and shouted, “Why must I lose to this idiot?” Except that Geraldo didn’t have the patience for chess. Entering a restaurant, they often flitted from table to table shaking everyone’s hand, imparting happy or irritable energy. “Geraldo” was quick to charm the audience but in the next moment to create a thunderstorm. Many esteemed psychiatric geneticists believe that there is a genetic link between bipolar disorder and schizophrenia. Clearly there is a shadowland between, and part of this borderland is an entity called schizoaffective disorder. Also, acutely psychotic patients with either disorder respond to the same antipsychotic dopamine receptor blockers. Finally, some genes play a role in both, as has been increasingly evident from studies using polygenic scores. However, the virtues of this type of diagnostic conflation, or syncretism, are questionable and it would seem to lead to contradictions in both research and clinical practice. By analogy, the BRCA1/2 genes play a role in both breast cancer and ovarian cancer, but no one claims these are the same disease even though the genetic connection has been proven. Breast cancer and ovarian cancer have many other distinct causes and consequences, as do bipolar disorder and schizophrenia. There are several clear distinctions between classic bipolar disorder and schizophrenia. One is a disorder of mood and affect and the other a disorder of thought. Clinically, they “feel” completely different. Bipolar disorder frequently responds to lithium salts. This was an amazing and wholly fortuitous discovery, which points to a different cellular molecular pathology in at least the bipolar patients who do respond. That molecular pathology is still unknown, but ultimately it may drive the brains of bipolar patients to alternating episodes of high and low mood. As a geneticist I am particularly moved by patterns of genetic transmission of bipolar disorder and schizophrenia in families because cross- transmission would point directly to common causality, and the lack of it to the opposite. In general, and despite the sharing of effects of polygenic scores that on an overall basis are not yet capturing a large part of the heritability of these diseases, there is a lack of cross-transmission of the two disorders. Schizophrenia most strongly predicts familial risk of schizophrenia, and bipolar disorder most strongly predicts risk of bipolar disorder. That would argue that many of the undiscovered
Impulsivity, diminished capacity, and neuropsychiatric disease
81
genes will work relatively independently, even though there will be some elements that contribute risk to both diseases, in the fashion that molecules such as BRCA1/2 and estrogen create some modicum of overlap between cancer of the breast and ovaries but the risk profiles and cellular biology of these two cancers remain distinct. The present state of knowledge of bipolar disorder is that for unknown reasons a sizeable proportion of all human populations has a severe disorder of mood. Frequently, the disorder is so severe as to make them a danger to themselves and others. Then they can be hospitalized, if necessary, against their will. However, what about lesser degrees of mania or depression? Perhaps the manic patient is throwing away the family’s life savings. Situations such as this present an ethical dilemma because the capacities for thought and decision are intact. and yet every thought and decision are colored as if seen through glasses that are tinted blue (depression) or rose colored (mania). Like mood disorders, anxiety disorders also are color thinking in ways that can be highly specific and to some extent, determinative. As someone with acrophobia, I can reliably report that someone with this phobia is not going to respond nonchalantly as did the person who had jumped off the top of a skyscraper and was briefly interviewed as they passed the 59th floor: “How are you doing?” “All right so far.” In High Anxiety (forget Vertigo), Mel Brooks is an acrophobic Harvard psychiatrist newly appointed director of The Psycho-Neurotic Institute for the Very, Very Nervous. Dr. Thorndike was booked in a top floor room at the vertigo- inducing San Francisco Hyatt. Probably because of the movie, I never had a problem at that hotel. However, checking around the room at an Arizona airport hotel, I noticed there was a sliding glass door, but no balcony or guardrail. This reminded me of a Jack Handey joke, “Hey, instead of a trap door, how about a trap window? You know, one that if you lean on it, you fall out. Oh, wait a minute, that’s how windows already work.” Being “very scientific” I tested the door, and it slid smoothly open leaving me teetering at the edge of a multistory drop. The soles of my feet tingled but that happens even if I see another person at a precipice. I backed away. To close the door, I approached on hands and knees. As illustrated, an interesting feature of acrophobia, and other phobias, is that behavior is not necessarily prevented, but the choices that are made are distorted and colored by predisposition.
Substance use disorders and other addictions The genesis and advancing course of substance use disorders can be related to urge and impulse control as well as disordered mood. These three neuroscience-derived dimensions—negative emotionality, reward, and executive control—are accessed in the “Addictions Neuroclinical
82
7. Distorted capacity
Assessment” that our group at the NIH assembled (Kwako et al., 2016) and validated—these three dimensions having emerged from factor analysis of our patients with substance use disorders and controls without (Kwako et al., 2019). However, in the process of addiction there is a transition to habit formation and further to compulsive behavior. Addictions are common worldwide, affecting more than 1 in 10 individuals. Addictions occur to many different agents, only some of which are drugs, but it is important to understand that many addictive agents are more addictive than others. As compared to other agents that are pleasurable—food, chocolate, or perhaps horseback riding (cleverly called “Equasy” by David Nutt)— psychoactive addictive drugs release dopamine far more powerfully in the main reward region of the brain, the nucleus accumbens, also known as the ventral striatum. Having learned that an agent is rewarding, both stress and cues specifically associated with the addictive agent create an impulse which can be consciously experienced as craving. Whatever their level of impulse control, it is much more difficult for the nicotine addict to resist the temptation to smoke a cigarette than it was for me to for example give up chocolate, which I had to do for health reasons. The chocolate was pleasurable, and I still enjoy cues associated with it, but there was never a strong primary pharmacological effect of the food. In smokers who have been studied using brain imaging methods we know that the nicotine cue will release dopamine in the nucleus accumbens, the brain’s reward center, stimulating craving for the “real thing” as well as conditioned cues. The chocoholic’s brain will also show some of the same responses to chocolate, but these responses are not initiated or pharmacologically maintained in the same powerful way as the addictive drug. Therefore addictions can represent an example of both disordered impulse and disordered impulse control. Also, these impairments can be circumscribed to a particular addictive agent, and they differ for agents of different addictive potential, and depending on the person’s experience with the addictive agent the craving may be severe or mild. To some extent the strength of these individual responses to addictive agents can now be measured by brain imaging, both using specific dopamine receptor ligands that enable the measurement of dopamine release in the nucleus accumbens, as Nora Volkow of the National Institute on Drug Abuse (NIDA, where Volkow is director) and NIAAA (where she is a lab chief) has done, and via the study of metabolic activations of the brain with functional magnetic resonance imaging (fMRI). Already, it is being learned that genetic variants can alter these reward activations, as Vijay Ramchandani, Dan Hommer, Reza Momenan, and Markus Heilig, present and former colleagues at the National Institutes of Health, have shown. Furthermore, a genetically influenced circuit involving the nucleus accumbens and a regulatory region of the frontal
Impulsivity, diminished capacity, and neuropsychiatric disease
83
lobe was discovered by Elliot Stein (at NIDA) and Eliot Hong (at the University of Maryland), and together we found that weakness of this circuit predicted greater craving for nicotine. Some people are born with greater propensity to addictions and craving, and we are beginning to pinpoint the genes that are involved and the brain circuits they affect (Hong et al., 2010). An important aspect of addictive agents is that many, or even most, are not consequence free. Some addictive agents, for example alcohol, lead to impaired cognitive capacity, and thereby to a more general disinhibition of behavior. When people choose to consume alcohol, they have made a choice that can lead to problems that follow from behavioral disinhibition and impaired judgment. Furthermore, when an alcohol use disorder patient “chooses” to drink, his choice has been strongly colored by their addiction to the drug.
Obsessive-compulsive disorder Obsessive-compulsive disorder (OCD) is a common, severe, genetically influenced and usually lifelong psychiatric disease. It is often resistant to treatment. Perhaps one in 200 people meet the clinical criteria for OCD, and many more have milder obsessions and compulsions. In contrast with the milder forms of perfectionism and compulsion that many experience, and that are useful in many situations in life, the obsessions of OCD are disruptive of normal life and happiness. The compulsions vary from one person to the next. However, whatever the compulsion—cleaning, checking, counting, or a myriad of others—the obsession and the activity required to fulfill it can come to dominate existence. As anyone who has watched Tony Shalhoub in the role of detective monk, there is nothing wrong with the impulse control or judgment of most people with OCD. Also, there is not a generalized disorder of impulse. Like a person with an addiction, an OCD patient experiences a very strong urge to perform a specific act and feels miserable if they cannot.
Tourette syndrome Tourette syndrome is a relatively common neurological disease leading to disordered motor movements. These can include facial tics, movements of the extremities, and vocal tics. The vocal tics often involve coprolalia, which is the exclamation of curse words. Like behaviors in OCD and addictions, the Tourette’s tics can also be voluntarily suppressed, as well as diminished or even eliminated by drugs that block dopamine receptors. By focusing their concentration, the person with Tourette syndrome can eliminate or reduce the tics, but they reemerge as soon as they relax. This is a general problem with compulsions, whether in addictions, OCD, or
84
7. Distorted capacity
Tourette syndrome. With constant monitoring and conscious effort, the behavior can, with a struggle, be suppressed, but ultimately the tics tend to emerge.
Common origins of disorders of impulse? It is important to note that Tourette syndrome, OCD, and stuttering all tend to cooccur in certain families. Many geneticists feel that this indicates that these seemingly very different behaviors are influenced by some of the same genes. A key to effective treatment of these disorders seems to lie in the inhibition or removal of the impulse. For example, the impulse to use the addictive agent can eventually fade in the addict. SSRIs help some OCD patients and in those who do respond, both the obsessions and the compulsive acts subside, so that the person experiences some cognitive relief. Also, Tourette patients successfully treated with dopamine receptor blockers experience a diminished pressure to make the motor movement.
Schizophrenia Schizophrenia is a psychotic disorder affecting approximately 1% of people. Clearly, the hallucinations and delusions of a schizophrenic patient can lead to bad decisions, in a way that is beyond the control of the person with schizophrenia. It is precisely these delusions and hallucinations that are most responsive to treatment with antipsychotic drugs, the original prototype being chlorpromazine. Even while psychotic, the schizophrenic patient may have adequate (which is to say normal) judgment and control in many areas. When treated with the antipsychotic drug, these other areas are not much affected, and people without schizophrenia do not experience the same sorts of benefits—they only experience side effects such as sedation and motor impairment. This focal effect of antipsychotic drugs on delusions and hallucinations is how one can be sure that antipsychiatrists are incorrect in arguing that schizophrenia is merely the reification of thoughts and behaviors that are socially unacceptable or inconvenient. Also, the ability of certain drugs and sensory deprivation to reproduce delusions and hallucinations, including the same ones experienced during psychotic episodes, strongly argues that there is something happening in the brain that is beyond the control of the person with schizophrenia, and for which help is needed. The objective fact is that there are no space aliens inserting thoughts into the brain of the paranoid schizophrenic, and it makes life much more difficult if one goes around thinking that there are. However, schizophrenia is also a thought disorder, affecting executive cognition and the ability to make judgments. Successful relief of psychotic symptoms of schizophrenia often leaves a significant cognitive component that is usually considered to be a deficit. Louis Sass, in a brilliant and
Impulsivity, diminished capacity, and neuropsychiatric disease
85
entertaining book Madness and Modernism—sent me after a day touring Pompeii’s ruins—proposed that this difference in the residual thought patterns of schizophrenia patients was not a deficit, but a gain of function, with its own peculiar consequences (Sass, 1992). Both modernism and the cognitive madness of schizophrenics are reflexive and recursive. Encapsulating Sass’s conception, reflexivity is the type of thinking involved in saying, “I think that I think that I think that I think that I am.” Or to put it more efficiently, “(I think)n that I am,” where n is any number of iterations of “I think.” Anyone who ever watched the original Star Trek knows the trap of infinite cognitive loops. By this stratagem, Captain Kirk defeated several supercomputers, both alien and human made. I’m sympathetic to Sass’s idea that reflexivity is a strength because the idea that certain cognitive styles fit niches has appeal, and humans face demands for reflexive thinking that our primate ancestors did not. Reflexive thinking is on the rise culturally and may have been favored genetically in our human lineage. The modern word processor has worsened our reflexivity—unrestrained we tend to infinitely correct and “perfect.” Because there is a little bit of schizophrenia in my own family, I am even more sympathetic, or perhaps biased. It is pleasant to view one’s cognitive style as plus rather than minus. However, one consequence of reflexivity is difficulty in decision making. A cognitive hallmark of schizophrenia is ambivalence. Alas, the cognitive deficits of schizophrenia go deeper than reflexivity. There are problems with gating of stimuli. Most tellingly, deep studies of frontal lobe function, for example by Daniel Weinberger and his group, formerly at NIMH and now comprising the Lieber Institute, reveal deficits not only in schizophrenics but also in their siblings free of the disease but having some of the same genetic loading. As will be discussed more, the frontal lobe specifically mediates executive cognitive processes that include not only behavioral inhibition but also switching between different cognitive strategies. Often when talking to patients with schizophrenia it is apparent that they have difficulty in switching to an appropriate cognitive strategy and as a result make perseverative errors. As they say, but as is refuted somewhere else in this book, the definition of insanity is the expectation that one can do something that has failed over and over and get a different outcome. As will be discussed elsewhere and in the context of a gene, namely catechol-O-methyltransferase (COMT), which moderates frontal cognitive function, task switching can be measured experimentally. When this was done on schizophrenic patients and their well siblings by Weinberger and Michael Egan, we made a very surprising discovery, which was that even when the schizophrenics and their siblings were performing an easy frontal task that they could complete as well as the next person, their frontal cortex was having to work much harder to do it (Egan et al., 2001). About a decade later, my son Ariel and I traveled with, or as I told one border
86
7. Distorted capacity
guard, “against,” the Weinbergers—Danny, his wife Leslie, and his son Collin in the Middle East. This trip taught me much about culture and decision making, but especially valuable were Danny’s insights on neurocognitive deficits in schizophrenia and the way that altered frontal lobe function which might be crucial in the onset of schizophrenia in adolescence, a time of rapid neurodevelopmental change and adaptation. It also taught me a little about risk management. Noticing a pistol on the front seat of the taxicab I pointed out, to Weinberger’s chagrin, that such guns were not very accurate. Picking up the gun the driver replied, “At this range, it doesn’t have to be.” As Danny pointed out, this is also how dopamine works in the frontal lobe—it often diffuses and acts extrasynaptically.
Delusional disorder Delusional disorder is not the condition of holding a political opinion different than one’s own. With a frequency of about 1%, it certainly is too uncommon for that description to apply. Unlike schizophrenia, delusional disorder is not marked by cognitive deficits. The problem is the delusion itself, which can create the impulse for the person to harm themselves or others. Delusional disorder usually afflicts high functioning people and has onset later in life, although there may be subtle precursors. Typically, a person with delusional disorder believes there is a vast conspiracy directed against them, and like schizophrenia patients, they have ideas of reference. Paranoid fears torment them and may lead to all kinds of behaviors harmful to themselves, their families, and others as they struggle to prevent poisonings, defend themselves and their children, keep the enemy agents and their devices out of their homes and bodies, and so on. Many people with delusional disorder do not come to the attention of physicians. Delusional disorder can express itself in sexual behavior, both in morbid jealousy and in sexual attachment. One of the first psychiatric patients I saw as a medical student was a man with delusional disorder. “Mr. Hilltop” was cared for by Harry Davis, a wise psychiatrist who practiced and taught at the University of Texas Medical Branch in Galveston. Davis was a flexible thinker who dealt with situations for which there was no simple formula or “standard of care.” For this reason, he had some of the “best cases,” and he assigned a few of the most fascinating to me. “Mr. Hilltop” was a self-made millionaire in the oil industry. In his late thirties he began to suspect that the government was out to get him. It started with a tax audit. As the saying goes, just because you are paranoid doesn’t mean someone isn’t after you. Then he began noticing that people around him were “in on it.” Being a Texan, he had guns, but bought more and began sleeping with them. He constructed a pillbox to defend his property. He was distracted at work—it wasn’t easy to keep up a façade
Disorders of impulse control
87
when many of his so-called associates were actors and spies. He deduced that his own family was “in on it.” If not, why did they ignore the obvious truth? For unclear reasons, Mr. Hilltop placed a revolver against his cheek and shot himself; the bullet passing clean through and leaving him otherwise unscathed. Treated with a dopamine receptor blocker, Mr. Hilltop’s delusions subsided, and although they were always, as he put it, “in the back of my mind,” he was able to get back into his normal life. Three decades later, as I reflect on his unusual behavior, Mr. Hilltop’s persecutory delusions were a necessary ingredient in the harm he did himself and could easily have led him to harm others. If he shot someone, I believe that a judge should have taken his illness into account as a mitigating factor or modified his sentence so that the primary cause of his behavior should be addressed. However, it would be wrong to say that he had lost all ability to control his behavior. Throughout his ordeal, he had a strong ability to control his impulses, and outside the sphere of his delusions he was as sane as the next person. Also, the choices he made were constantly determined by his individual character and perspective, of which his paranoia was a part, and in the end he did not, like a Charles Whitman, choose to shoot people from some Texas Tower, nor did he quite take his own life. Instead, Mr. Hilltop made a completely individual, bizarre, and I think unpredictable, choice.
Disorders of impulse control The opponent process to urge is control. There is probably no other area of psychiatric diagnosis where there has been so much inconsistency, change, and confusion as in the disorders of impulse control in which either urge or control, or both, may be disturbed. Yet, there are millions of children and adults who have these disorders, there are effective, life changing and even lifesaving interventions for individuals who are accurately diagnosed, and genetic factors and life experiences (trauma) play a strong role in their genesis. Disorders of impulse control start in childhood with attention deficits, hyperactivity, and irritability, and begin later in life with neurological syndromes associated with brain injury. We have already discussed the impact of brain injury (e.g., Phineas Gage, James Brady). In neurodegenerative disease, particularly frontotemporal dementias, impulse control can be greatly weakened. As we learned, damage to the frontal lobe plays an especially important role in releasing behavior, and in leading to impulsive choice. However, the development of the frontal lobe is not complete until early adulthood. Furthermore, children are naturally more aroused and fuller of vital energy. This is a powerful reason why behavioral choice is different in children, and
88
7. Distorted capacity
why they are treated differently when they are convicted. It is a reason why adult disorders of impulse control are likely to have their onset in childhood. Finally, it is a reason why the childhood diagnoses of impulse control, hyperirritability, and “bipolar,” all of which are likely to lead to the long-term treatment of the child with psychoactive medications, are controversial.
Antisocial personality disorder Antisocial personality disorder (ASPD) is the psychiatric diagnosis that has been classically associated with aggression and impulsivity, but its definition has been the most mutable and confusing. A hallmark of ASPD is preexisting childhood conduct disorder—some children with childhood conduct disorder grow out of it, but adults with ASPD get their start by lying, stealing, torturing animals, skipping school, attacking other children, and in general behaving as little hellions. The criteria for ASPD can be met based on childhood conduct disorder plus past behaviors including failure to conform to social norms, irresponsibility, and deceitfulness. With behaviors such as this, it is unsurprising that more than four-fifths of incarcerated criminals meet criteria for ASPD, and that this common diagnosis is not particularly useful for predicting recidivism. On the other hand, Robert Hare’s conception of ASPD was primarily psychological, not behavioral. Hare’s scale attempted to measure the inner state of remorselessness, alienation from normal social control, and lack of conscience. In that regard it was ahead of its day, pointing to a time when with brain imaging and genetic predictors it might be possible to understand behavioral origins. However, at the time it was put forward Hare’s scheme ran against the tradition established by the great psychologist John Watson, who had emphasized the value of measurable external manifestations of the brain, as compared to explanations based on unmeasured internal states. This is a very important point, and therefore it is also touched upon earlier in this book (see Index). Scientists first and foremost study what they can measure and are attracted to classification schemes based on measurements. We may observe that a defendant has a history of childhood stealing, lying, torturing animals, early drug use, and sexual promiscuity. Someone who has examined the child may then report a variety of explanations for the behaviors, ranging from lack of conscience to attention seeking, or a cry for help. From Westside Story: Dear kindly Sergeant Krupke, You gotta understand, It’s just our bringin’ up-ke That gets us out of hand. Our mothers all are junkies, Our fathers all are drunks. Golly Moses, natcherly we’re punks.
Disorders of impulse control
89
The child may convince Sergeant Krupke, but that does not necessarily make it so. Prior to directly accessing activity of brain circuits or genes that influence them, classifications of antisocial and impulsive behaviors that depend on externally observed behavior have worked better. The ability to identify genes and identify neurochemical factors that influence ASPD, and the ability to image the activity of the brain, for example while the brain is performing tasks requiring behavioral inhibition and executive cognitive control, is arguably bridging the gap between internal states and external ASPD behavior. In that regard, Adrian Raine, my colleague James Blair at the National Institute of Mental Health (NIMH), and others have shown that many with Hare-type ASPD do not have deficits of frontal function and have not lost the ability to control their impulses. Instead, some of these ASPD individuals appear to be “Hannibal Lecters.” These are the conscience-free, remorseless, and alienated people that met the Hare definition of ASPD. They may become serial killers. Obviously, the deficit in such individuals is not in their ability to plan or defer immediate action in favor of long-term reward. However, there are three other psychiatric disorders that primarily involve a deficit of impulse control.
Intermittent explosive disorder Intermittent explosive disorder (IED) was discussed in Chapter 3 in the context of the “2B or not 2B” story, in which the 5-hydroxytryptamine (serotonin) receptor 2B (HTR2B) stop codon was found to contribute to severe impulsive behavior, and even violent, senseless, murders. IED of whatever cause is common relative to diseases such as schizophrenia and bipolar disorder, and it can cooccur with bipolar disorder and other diagnoses. It is marked by extreme expressions of anger, often to the point of uncontrollable rage, and the behavior is disproportionate to the situation at hand. IED outbursts are brief and are often accompanied by signs of heightened autonomic activation such as sweating, chest tightness, twitching, and palpitations. Typically, the person is remorseful afterward.
Borderline personality disorder Borderline personality disorder (BPD) probably affects one in 50 people and as many as one in 25 people, although it frequently cooccurs with other disorders. Like ASPD, with which it can cooccur, and is genetically cross-transmitted, BPD also is marked by powerful, poorly regulated emotionality, in addition to impulsive behavior. BPD is a disorder of arousal. BPD patients are often in severe emotional distress. They form strong, but shifting and unstable emotional bonds. Often the emotional attachment is unreasonable and unreciprocated, leading to tragic disappointments. Our understanding of BPD and the advances in its treatment, limited
90
7. Distorted capacity
though they may be, are due to work of dedicated and brilliant psychiatrists. Don Klein, at Columbia University, was a pioneer in showing that pharmacotherapy was one way of helping the BPD patient gain at least some measure and sense of control. Larry Siever, working with Antonia New at Mt. Sinai Hospital, used brain imaging to measure the differences in the brains of BPD patients. There are extraordinary obstacles in the way of such studies. BPD patients are not convenient to study and in the end, and despite the moderate heritability of BPD, it will probably be discovered that many of these patients suffer from emotional dysregulation because of early life trauma. Here, I am reminded of multiple personality disorder, for which we really have no understanding of mechanism, are sympathetic to the suffering of people who have it, and are thoughtful of the likelihood that many people who have it probably had some terrible early life trauma. Psychiatrists indeed appear to be making headway, finding that BPD patients have differences in regional brain metabolic activity correlating with their deficits in cognitive and emotional control. Again, the frontal lobe is implicated, but with BPD the ability of the frontal lobe to modulate emotion may also be coming into play. Most importantly, Marsha Linehan pioneered, and many other psychiatrists followed, in showing that BPD patients could be helped by a cognitive therapy, dialectical behavior therapy. This observation carries with it the powerful implication that even the most emotionally volatile people can choose and can be helped to develop strategies to choose better.
Childhood conduct disorder Childhood disorders of impulse control, irritability, and misconduct are the most controversial. Childhood conduct disorder (CD) is a bit of a catchall but is the precursor for ASPD in adults. Children with CD lie, cheat, and steal. They are likely to be truant, commit vandalism, and engage in early sexual behavior and drug use. Childhood CD also may frequently include individuals destined to be Hare-type psychopaths because these children are cruel and remorseless to both pets and people. However, a major problem with the CD label is that most children engage in such behaviors, especially when they are very young. As we have discussed, children have not developed executive cognitive control to enable them to regulate their behaviors well. This is what parents are trying to help their children learn. Also, while awake, they tend to be more aroused and energetic (thymic). Approximately 5% of children for whatever reasons— genetic, family environment, environmental toxins, socioeconomic— establish a pattern of repeated and continuing CD behavior through late childhood and adolescence. Unfortunately, labeling them as CD has so far yielded little benefit because there is no consensus on how to treat it. Clearly, there is a strong role of culture because rates of delinquency vary
Disorders of impulse control
91
widely between different societies. Those who progress to criminal violence are overwhelmingly male. Other than the Y chromosome, the role of genetic factors is not well understood.
Attention deficit hyperactivity disorder Attention deficit hyperactivity disorder (ADHD) is a neurobiologically based disorder. It is moderately to highly heritable. Further out on the frontiers of childhood disorders are other diagnoses such as childhood bipolar disorder. The behavior of children is notoriously difficult to assess, and usually entangled in the web of family interactions (Rutter et al., 2010). Most children with the diagnosis of childhood bipolar would perhaps be helped more by receiving another diagnosis, or none. However, many children with ADHD have deficits in brain function and are performing as well as they can under the circumstances. However, the ADHD diagnosis is currently not made based on process but is made by behavioral assessment by the parent or teacher on the basis that the symptoms appearing before the age of seven are persistent and disruptive. The behaviors are in three domains: inattention, for example, forgetfulness, distractibility, and losing things; hyperactivity, for example, fidgeting, inability to stay seated, restlessness, and excessive talking; and impulsiveness, for example, not waiting one’s turn. Studies have shown that ADHD diagnoses made in this fashion are reliable, but such a behaviorally based diagnosis might be made in a way that is reliable but nonspecific to a frontal lobe deficit or an excess of arousal. It is also easy to see that such diagnostic criteria have led to overdiagnosis. There are many other causes of attention deficit and hyperactivity in children: child abuse, sensory deficits, sleep disorders, discipline and rearing practices, and family stress and disruption among them. That has very important consequences because the attention deficit aspect of ADHD is treatable, with methylphenidate, a medication that in amphetamine-like fashion augments levels of dopamine in the frontal lobe. When child abuse or other family dysfunction are at issue those problems will not be solved with a drug. In ADHD, frontal lobe deficit makes it very difficult to attend to tasks or play activities. One way of thinking about ADHD is that it is a developmental lag disorder, with children with ADHD being 3–5 years behind their classmates in executive cognitive control. In fact, the younger children within a grade are more likely to receive a diagnosis of ADHD. However, although the frontal lobe continues to develop in children, the attention deficit can continue into adulthood in perhaps 60% of cases, with varying degrees of adaptation. This would imply that there is something different about their brains, and the issue is not just developmental timing. The future of ADHD assessment will hopefully embrace neuropsychology, brain imaging, and genetics.
92
7. Distorted capacity
One alley for ADHD assessment that appears to be a blind one is pharmacological response, representing therapeutic diagnosis, which is anyway a bit backwards because it is obviously better to avoid the administration of stimulant drugs to children who do not require them. It has been claimed that the effect of stimulants to calm children with ADHD is “paradoxical” and that a favorable response to treatment therefore confirms the diagnosis, but this is not the case. In line with the idea that many children and adults diagnosed with ADHD do not have a specific frontal deficit is the fact that the effect of methylphenidate to improve frontal function is not specific to ADHD. That is one of the problems with stimulant drugs. It is tempting to use them, and the price to be paid may be one only to be forfeited in the long term. Today, 1 in 12 Major League Baseball players receive physician-prescribed stimulant medication for “adult ADHD.” This epidemic of adult ADHD in baseball players followed the banning of stimulants in that sport. Illegal amphetamine (“greenies”) helped hitters maintain the intense focus of concentration necessary to pick up the rotation of a baseball as it leaves a pitcher’s hand, and in milliseconds, process whether it is a 98 mile per hour fastball or an 85 mile per hour curveball. ADHD children treated with stimulants do show improved cognitive performance; however, as Judy Rapoport at the NIMH showed more than three decades ago, so do many normal or even above-average children. Rapoport is a world expert on childhood psychiatric disorders, including ADHD and childhood schizophrenia. When she administered amphetamine to children with ADHD, they substantially improved cognitively, as expected. Because of the controversial nature of studies involving children who are normal controls, the children without ADHD in her study were already decidedly above average cognitively, being offspring of well-informed parents who were themselves cognitively above average. Nevertheless, the cognitive performance of the control children, who were already doing well, also improved to a similar extent as the ADHD children. The cognitive response of the ADHD children to stimulants is not “paradoxical,” although the behavioral response (they may sit still better) can be viewed that way. While cognitive function of many people will improve following stimulant drugs, we will see later that whether they do can depend on the stressfulness of the situation and the difficulty of the cognitive task, and at least one gene, COMT, plays a role. Gene findings such as this, and the role of genes such as monoamine oxidase A (MAOA) and HTR2B in impulse control and genes that modulate stress response and emotionality, will help us revisit the role for genetic influence on both impulse and impulse control, but in a somewhat different light. The effects of genes on these other mechanisms will provide further explanation for the heritability of the disorders that involve impulse control.
References 93
Overall, the impulse control disorders, and other psychiatric diseases we have discussed, represent difficult challenges for individuals and families who must live with them and for physicians and all the others— teachers, nurses, social workers, and coworkers—who help them cope. In deciding whether people have free will, it is crucial to understand that there are disorders that impinge in specific ways on the choices they make. These disorders are endemic among us and our families. However, I will argue that in the final analysis these disorders are impinging factors, not deciding ones, and are also to be embraced as part of our neurogenetic individuality.
References Egan, M.F., Goldberg, T.E., Kolachana, B.S., et al., 2001. The effect of COMT Val108/158Met genotype on frontal lobe function and risk for schizophrenia. Proc. Natl Acad. Sci. USA 98, 6917–6922. Hong, L.E., Hodgkinson, C.A., Yang, Y., et al., 2010. A genetically modulated, intrinsic cingulate circuit supports human nicotine addiction. Proc. Natl. Acad. Sci. U. S. A. 107, 13509–13514. Kwako, L.E., Momenan, R., Litten, R.Z., Koob, G.F., Goldman, D., 2016. Addictions neuroclinical assessment: a neuroscience-based framework for addictive disorders. Biol. Psychiatry 80 (3), 179–189. https://doi.org/10.1016/j.biopsych.2015.10.024. Epub 2015 Nov 17 26772405. PMC4870153. Kwako, L.E., Schwandt, M.L., Ramchandani, V.A., Diazgranados, N., Koob, G.F., Volkow, N.D., Blanco, C., Goldman, D., 2019. Neurofunctional domains derived from deep behavioral phenotyping in alcohol use disorder. Am. J. Psychiatry 176 (9), 744–753. https:// doi.org/10.1176/appi.ajp.2018.18030357. Epub 2019 Jan 4. Erratum in: Am J Psychiatry. 2019 Jun 1;176(6):489 30606047. PMCID: PMC6609498. Rutter, M., Bishop, D., Pine, D., Scott, S., 2010. Rutter’s Child and Adolescent Psychiatry. Sass, L.A., 1992. Madness and Modernism: Insanity in the Light of Modern Art, Literature and Thought. The American Psychiatric Association, 2022. Diagnostic and Statistical Manual of Mental Disorders, fifth ed. The American Psychiatric Association.
This page intentionally left blank
8 Inheritance of behavior and genes “for” behavior: Gene wars Life is perpetual instruction in cause and effect. Emerson Progress in behavioral genetics asks for the wisdom of many disciplines: psychology, ethics, psychiatry, history, molecular biology and statistics. However, sometimes the result of bringing together these disparate viewpoints is an intellectual Tower of Babel. When discussions devolve into ritual combat, the best we can hope is that we are “Dee” rather than “Dum” and try to keep in mind that one accrues scant credit for winning a wrestling match with someone half your size. Decades into the genetic revolution, and even despite genomic studies since the first edition of this book identifying hundreds of gene locations of genes influencing behavior, we struggle to master a cross disciplinary language of genetics and to reach consensus on fundamental elements: heritability of behavior, environmentality, reaction range, and the significance of individual and population variation. Can we agree on some of the cultural, religious, philosophical, and ideological influences on behavioral genetics and the uses and misuses of genetics? Whatever our backgrounds, our opinions on psychiatric genetics, including more politically charged areas such as the genetics of cognitive ability (Murray and Herrnstein, 1994) or antisocial behaviors, should be informed by data. To what extent are psychiatric geneticists “following the data” and to what extent are the motivations of their inquiries, and the conclusions, colored by ideological and social agendas? What are the motives of its critics, and are they relevant? Psychiatric genetics is sometimes a toxic environment in which some geneticists and psychiatrists have been accused of being cogs in a government plot to tranquilize people after reifying “undesirable” behaviors as psychiatric diagnoses. Adding genetics to the mix inflames the concern. Because both psychiatry and genetics have contributed to legacies of racism and genocide, disparities between ethnicities, social groups, and economic strata feed the thesis that modern psychiatric genetics serves the stigmatization and suppression of the socially and economically
Our Genes, Our Choices https://doi.org/10.1016/B978-0-443-22161-3.00018-1
95
Copyright © 2024 David Goldman. Published by Elsevier Inc. All rights reserved.
96
8. Inheritance of behavior and genes “for” behavior: Gene wars
isadvantaged. Can the science be disentangled from the politics, or do d we want it to be? Shouldn’t science inform policy and philosophy? The ideological war on behavioral genetics reached a dubious peak in 1995 at the Wye Conference on the Genetics of Aggression, an event organized by David Wasserman, now a bioethicist at NIH and working right around the corner from me. The Wye Conference was a “happening.” Even before demonstrators arrived, Irv Gottesman, a legendary figure in the field, had been characterized as a Nazi. Gottesman performed important twin studies on aggression and introduced to behavioral genetics the concept of endophenotype (an inherited, disease-associated intermediate phenotype). He was also Jewish. The demonstrators were better organized than us. They brought proceedings to a halt (with my slides in the projector), chanting, “Maryland conference, you can’t hide; we know you’re pushing genocide.” A demonstrator made the most apt comment of the meeting. A name-badged attendee had become agitated even before the demonstrators arrived. Nothing to write home about, but now the scientist had gone off-kilter, pushing toward a demonstrator, who had the wit to yell, “Get that scientist’s DNA!”. Later, after things had cooled down, the angry person was persuaded to leave. The inducement: a bottle of wine, as Wasserman also remembers.
The debate on the heritability of behavior The picture’s pretty bleak, gentlemen… the world’s climate is changing, the mammals are taking over and we all have brains the size of walnuts. Gary Larson
A main issue dividing geneticists at the Wye Conference from historians and philosophers was the heritability of behavior. The argument was complicated by the fact that the geneticists tended to emphasize the caveats and unknowns, in addition to the knowns about inheritance and genes. Also, the discussion leapt forward to the threatening consequences instead of proceeding stepwise: behavior → inheritance → genes → genetic predictor → identify consequences → deal with consequences. Having worked our way halfway toward the end of the chain, someone would derail the discussion, sending the whole group back to one of the earlier points. For example, it was claimed that behavior could not be defined, it was not heritable, or if genes were involved they could not be shown to be causal. Controversy over the heritability of behavior continues to the present day, with exaggerated claims on both sides despite clarifying popular expositions such as Michael Rutter’s Genes and Behavior: Nature–Nurture Interplay Explained, published in 2006.
The genome encodes reaction range
97
There was a standard discussion of the heritability of behavior. Worldwide, identical twins are rather common, and the frequency of identical twinning is remarkably consistent: about one in 250 births. As an aside, it has occasionally been observed that despite concerns about the dangers of human cloning the existence of so many clones has thus far not resulted in calamities. In the 1990s, but also later there was a tendency to frame the debate on genetic causality as if it hinged on twin studies, and to the neglect of other evidence of causality such as gene manipulations that enhance or eliminate behaviors, some gene manipulations pleiotropically altering more than one behavior, cross-species behavioral variation, and much else besides. In the 1990s the mapping of that predict behaviors was nascent. At the Wye Conference we could have argued that many of these genetic variations alter the profile of vulnerability—probabilistic determinism if you will. However, we did not get that far at the Wye Conference—not because of the demonstration but because this was a time when psychiatric genetics was much less advanced in identifying genes and quantitating their often-modest effects on behavior. Not everyone with an identical twin is aware they have one, although the existence of unknown relatives, even distant cousins, is increasingly being identified by direct-to-consumer genomics and genealogical databases. However, even in the days before such technological innovations behavioral similarities sometimes led twins separated at birth back to each other. For example, the identical twins James Alan Lewis and James Allan Springer famously discovered each other at the age of 39. Eerily, both had first born sons by the same name, married and divorced women named Linda, remarried women named Betty, had pets named Toy, chewed their nails to the nub, worked part time in law enforcement, and vacationed at the same beach. Another set of twins, Oskar and Jacob, were separated at birth. One was raised as a Catholic in Germany where he joined the Hitler Youth and the other as a Jew in Trinidad and eventually spent time on a kibbutz. Like the legendary Mallifert twins in the Charles Addams cartoon, Oskar and Jacob found remarkable similarities when they finally met at the airport. They both read magazines from back to front, stored rubber bands around their wrists and shared a variety of other idiosyncrasies and mannerisms.
The genome encodes reaction range When people observe dramatic effects of extreme environments on behavior it is natural they should doubt the preponderance of evidence that diverse behaviors, ranging from personality factors to cognitive
98
8. Inheritance of behavior and genes “for” behavior: Gene wars
a bility to psychiatric diagnoses, are moderately (30%–40%) to highly (60%–70%) heritable. A useful conception enunciated at the Wye Conference by geneticist Marcus Feldman is that genes lead to heritability by determining a person’s “reaction range.” Depending on the environment, the behavior can thus be very different even if the person has the same genotype. The reaction range can indeed be very wide. However, the fact that a range of environmental exposures can be imagined does not mean that they happen. Heritability measures what the role of gene and environment is, rather than what it could be. Although the range of potential reactions can be very wide, reaction range also must be understood probabilistically. Given a certain exposure, some reactions are much less likely than others. Both factors lead to the observed moderate to strong inheritance of behavior: people do not experience many of the more extreme environmental exposures we could imagine, and for a given exposure people with the same genotype tend to respond in certain ways, even if their potential range of reaction is much wider. Because the environmental factors to which populations of people are exposed are subject to change, reaction range is the main reason why heritability is not a fixed attribute of behavior, as unfortunately has sometimes been assumed (Murray and Herrnstein, 1994). If the environment is changed such that no one or everyone exhibits the behavior, then heritability of the behavior would drop to zero. This is true because many heritable behaviors are contingent on exposure, and conversely some heritable behaviors found in only a fraction of the population can theoretically be induced in nearly all via the appropriate environmental manipulation. An additional consequence of reaction range is that if environment is varied dramatically and randomly, the environmental variance will increase, and the role of inheritance (heritability) will drop close to zero. This is because heritability is a ratio of variance attributable to genetic factors divided by overall variance. This is shown in Fig. 8.1. Heritability is a Rao Addive genec variance Total variance FIG. 8.1 Heritability as a ratio.
Reaction range and free will
99
Choice and reaction range Depending on the environment, and as Feldman did not say, epending on individual choice, the behavior can be very different d even if the person has the same genotype. Also, reaction range can be very wide, but is itself changeable by manipulating the environment. The ability to choose one’s responses and shape one’s environment within a genetically determined reaction range is not an abandonment of causality or resort to magical thinking. It is recognition that topdown choices of environments and behaviors, as well as genes that set reaction range, are integral components of the chains of causality that lead to behaviors. By increasing environmental variance, the effect of most genes is diluted, lowering heritability, since in the short term increasing the environmental variance does not lead to an increased variability in genotypes. The exception is that some genetic effects only occur following an environmental exposure, so that if the environment is not so varied as to expose some individuals the effect of the gene will never be visible.
Reaction range and free will Reaction range and environmental interaction can also become a clever person’s explanation for why a gene cannot “cause” a behavior. However, this theoretical argument is trumped by the moderate to high heritability of temperament and psychiatric disease. Heritability does not have magical origins but is directly attributable to genes, most of which are not yet identified. In proportion to their degree of genetic relationship, twins and other blood relatives share a complement of genetic variants inherited in common (identical by descent) from their parents. We do not understand exactly how inherited genetic differences lead to behavioral similarities within the reaction range framework or other theoretical frameworks. However, we should have the humility to admit that this is because we are yet unable to measure many of the specific genetic and environmental factors and because these factors have not been adequately integrated into the framework of gene × environment interaction (Caspi et al., 2002). Similarly, the proportion of variance in behavior that is explained by genes that have been securely identified remains small—those genetic markers are not highly predictive of behaviors such as anxiety or arousal, nor of psychiatric diagnoses. However, we also must understand that the effect sizes of many of the individual genes are very small, requiring very large samples to detect them. Furthermore, the genomic methods that have been mainly applied so far are mainly useful for detecting common variants.
100
8. Inheritance of behavior and genes “for” behavior: Gene wars
Twin studies and controversies they provoked People who care about racism and stigmatization were alarmed and saddened by Murray and Hernstein’s book emphasizing genetic differences in cognition between races—differences that are more likely attributable to other factors but which in any case do not reveal much about the cognitive ability of any individual. However, at the root of the discomfort is the point that inherited variation in cognitive ability is antithetical to the idea that given equal opportunities everyone will do equally well. Maybe they can, if they will, but they do not. Ironically people perform equally only if treated unequally. If the goal is equality of outcome, or even just for everyone to do well, some people need more help, and different help—perhaps guided by genotype or measures of cognition and personality. What do twin studies tell us about the inheritance of behavior, and what don’t they tell us? In the decades after the Wye Conference, several genes influencing behavior—all of larger effect—were securely identified, some by me, and in the past several years, a series of very large genome-wide association studies have yielded hundreds of genes (almost all of very small effect) influencing behaviors and hundreds of genes discovered by this method have been combined into polygenic scores predicting behavior. In line with Feldman’s reaction range they do so probabilistically. However, at the time of the Wye Conference people who did not like the idea that behavior and cognition are substantially inherited could still argue the point if they could undermine the twin studies. People who oppose the idea that behavior and cognition are substantially inherited do not like twin studies. Many are justifiably troubled by Murray and Herrnstein’s book The Bell Curve, emphasizing racial differences in cognitive performance (or to say it more specifically, IQ test scores) (Murray and Herrnstein, 1994). These differences were tracked against vaguely defined “race” rather than population. They are more likely attributable to nongenetic factors. Also, the group differences do not tell us very much about the cognitive ability of any individual or the inheritance of cognition. Given equal opportunities, would everyone perform equally well on cognitive tests? The evidence is that many can perform equally only if they are treated unequally. In a fashion like what we have learned of the inheritance of personality and cognition, if we want everyone to do well, some will require more help; and different help, perhaps guided by genotype. Let us explore in more detail the twin method that has generated this conclusion. The twin method is based on the idea that we can learn about inheritance by comparing the similarity of identical (monozygotic, MZ) twins to the similarity of fraternal (dizygotic, DZ) twins. As mentioned earlier, identical twins are often remarkably similar in behavior, even when raised apart. Fraternal twins tend to be less similar, although resembling each
The debate on genes “for” behavior
101
other more closely than two people selected randomly. Heritability can be computed from the ratio of similarity (concordance) of monozygotic to dizygotic twins: the MZ/DZ concordance ratio. For many personality traits and cognitive abilities, this ratio is approximately 2:1 and leads directly to an estimated heritability of 40%–60%. However, there are flaws in the twin design. The major defect is that identical twins tend to experience the same environment. However, the role of shared environment in behavior is not large. The effect of increased sharing of environment by identical twins has been measured in a surprising way. Twins have been studied who thought they were identical but who were fraternal twins. The degree of resemblance for personality and cognition of such inauthentic identical twins approximated that of fraternal twins. From the standpoint of the heritability of behavior and cognition, thinking you are an identical twin is not the same as being an identical twin. If behavior is inherited, then this means that certain genes that twins coinherit from their parents are responsible. This suggests that we should eventually be able to identify how those genes “for” behavior can build a brain and how variations in these genes can influence behavioral differences. Even if the effect of the gene is small and probabilistic, it is the reliability of the gene effects on a population basis that enables evolution to work to act on the frequency of genetic variants by selecting genes “for” behavior. To understand how a functional DNA variant can influence behavior, we will later navigate the route from DNA to molecule to cell to brain to behavior. At that point, we will start with DNA and its natural variations, and the process by which DNA is transcribed into RNA and then decoded by cells to enable the synthesis of proteins and eventually the complex structure of brain circuits that enable behavior.
The debate on genes “for” behavior Apart from the problem of heritability, there is the problem of the relationship of the heritable elements (the genes) to outcome. During the Wye proceedings a philosopher argued that it was always wrong to refer to a gene as a gene “for” behavior (a statement that is theoretically seductive, but false). In a way, I agreed, because the more correct formulation is “influence,” but “influence” is cousin to “cause.” The argument didn’t stop there: supposedly no gene could be said to be a “cause” for anything. Getting rid of “cause” was an idea that did not gain traction: I said that if we were not careful, we could next find ourselves doing away with “effect,” at which point we might as well retreat to caves. Paraphrasing Leon Trotsky, everyone has a right to be stupid, but not to overuse the privilege. A decade later, and as I learned in discussions with a coauthor on a
102
8. Inheritance of behavior and genes “for” behavior: Gene wars
ight-have-been textbook, there was still vigorous debate as to whether m there are genes “for” behavior. There are two sides to the argument, but the crucial issue is whether evolution has selected certain genetic variations for the express purpose of altering behavior, thus making these genes “for.” I discuss examples of such genes, including the COMT “warrior/worrier” gene, elsewhere in this book. Genes are selfish things, and we are their vessels (Dawkins, 1976, 1996). However, in one sense the philosopher and the coauthor were both correct. There are many genetic variations that did not evolve for the purpose of causing a psychiatric disease, but that contribute to disease if combined with an unfavorable mix of other genes and environmental exposures. For example, it is perfectly obvious that there is no gene “for” cocaine addiction, in the sense that humans were only recently exposed to cocaine, and no advantage of being a cocaine addict has been identified (Goldman et al., 2005).
People are not monkeys A controversy that almost derailed the Wye Conference before it was convened, but that was not directly dealt with at the meeting, was the use of animal genetic models of aggression, and extrapolation (charitably) or conflation (uncharitably) of findings in the animal models to behavior in people. As already discussed in the context of the “2B or not 2B” stop codon (Bevilacqua et al., 2010), given the limitations of studies in humans, it is very important that we understand the importance of animal models. They can provide a key to the validation of genetic behavioral findings, including the overall role of inheritance and the roles of specific genes. I have primarily used animal models to follow up genetic findings made in people. Man-made models such as the Htr2b gene knockout mouse can provide an opportunity to test the predictive validity of observations made in people. Also, naturally occurring genetic variants in monkeys and other species such as mice and monkeys can be extremely valuable. At times these sequence variants found in other species are orthologous, that is of the same evolutionary origin and function. The rhesus macaque monkey has naturally occurring gene variants at monoamine oxidase A, the mu opioid receptor, and the serotonin transporter that are orthologous to or closely mimic the functional variant found in humans, and the list of natural and man-made genetic variants available in other species that are informative for human behavior is constantly growing. In such animal genetic models, it has been possible to confirm gene-by-environment interactions, particularly gene-by-stress interactions, in a way that would have been impossible in human studies. There is also an overarching and foundational role of animal studies in neuroscience that is sometimes not appreciated, and this is that most of what we know about the relationships
People are not monkeys
103
of brain neurochemistry and structure to behavior has been learned from studies in animals. It is true that brain imaging and other psychophysiological measures are increasingly enabling access to the human brain; however, this access is, if anything, lagging further behind what is feasible in animal models, because of a variety of powerful new methods that can be used in other species, which frequently depend on the use of genetic tricks. For example, the “connectome” images of Jeffrey Lichtman shown elsewhere in this book and which are enabling neuroscientists visualize and explore the connections between neurons in an unprecedented way can only have been generated in animal models where the genetic manipulations are feasible and ethical. Similarly, the optogenetics revolution pioneered by Karl Deisseroth, in which specific neurons and circuits are switched on and off using light, depends on the ability to perform a genetic manipulation and to place a probe into the brain. The reasons for the widespread exploitation of animal models in neuroscience therefore do not include either a desire to experiment on animals or overoptimism about the ability of animal models to replicate the human experience. For many of the higher order behaviors of humans, animal models fall short. People are not mice or monkeys. Our culture matters but so do our genes (Pinker, 2003) and brains (Churchland, 2011; Damasio, 2005). Emphasizing the difficulty of establishing animal models of behavior, there are no good animal models for several psychiatric diseases, including schizophrenia and bipolar disorder. Mice and monkeys can be induced to drink to inebriation and become hooked on drugs of abuse, and much of what we know about the neurobiology of addiction has been learned from these models. However, the alcohol-related behaviors of a rodent or monkey are not equivalent to alcohol use disorder, which is a human phenomenon that is social in its contexts, expression, diagnosis, and course, and human genetic in the sense of the functional sequence variants that humans have and that influence onset and course. The problem of using other species as models for human behavior is emphasized by the fact that even one person’s behavior has a different origin than the next and thus there is no one model for human psychiatric diseases. My schizophrenia or normalcy has a different constellation of causes than does yours, even if we superficially resemble each other. We are not like cars of the same model manufactured in the same plant. Under the hood, our engines are different. Therefore if an animal model should somehow exactly capture the features of your situation it will probably fall short for mine. While we can learn much about the basic neuroscience from animal models, we constantly return to the human for genotype, context, and an understanding of outcome. Knowing the sophistication of Frederick Goodwin in these issues, he clearly was aware of these nuances and limitations. However, bad things can happen when a scientist speaks in shorthand and in public forums. Goodwin’s downfall as director of the NIMH came unexpectedly, and as
104
8. Inheritance of behavior and genes “for” behavior: Gene wars
so often happens from a combination of good intentions and getting too far ahead of history. Arguably, hubris played a role. Goodwin had f ostered the Violence Initiative, a loosely associated group of studies aimed at understanding the origins of violence, which as discussed elsewhere is a behavior that is strongly context dependent in its meaning. In 1992 he was quoted as comparing youth living in the inner city to “hyperaggressive” and “hypersexual” monkeys living in a jungle. Without a doubt his aim was to state that something could be learned about aggression by studying nonhuman primates—but the result was disastrous. Monkeys are in fact an important animal model of aggression: their behaviors are closer to those of humans than the behavior of rats, with aggression exhibited in some of the same social contexts. Also, the aggression and impulsivity of monkeys is increased by some of the same factors that increase human aggression and impulsivity: low social status, deprivation, maleness, and low levels of a particular neurotransmitter, namely serotonin. Marie Asberg at the Karolinska Institute and Markku Linnoila and Matti Virkkunen at the National Institutes of Health and University of Helsinki had found that low levels of a serotonin metabolite, 5-hydroxyindoleacetic acid (5HIAA), in cerebrospinal fluid were associated with higher levels of aggression and impulsivity, and especially as evidenced by suicide attempts. Similar correlations between 5HIAA and impulsive and aggressive behaviors were seen in monkeys. Years later, but too late to play a role in this controversy, it would be shown that gene-by-environment interactions important in human impulsivity also occur in monkeys exposed to early life social stress under controlled conditions. Perhaps it would have made a difference to the outcome of the controversy if some benefit of the research, namely the importance of preventing early life stress exposures, had been shown. These gene-by-environment interactions, and the measurement of their effects, will be discussed, but the end of the story at the time of Goodwin’s resignation was that the field had scales for measuring aggression and impulsivity, a solid biochemical predictor (5HIAA) and models for neurobehavioral research on aggression and impulsivity. However, a blow had been dealt to the credibility and ideological basis of the research.
The politics of behavioral genetics When demonstrators burst into the Wye Conference center, the leader carried a red flag and yelled, “Workers of the world, unite. Throw off the yoke of your oppressors!” Most of the scientists in the audience were puzzled. Were we the workers or the oppressors? One may well ask what communism had to do with genetics. However, the answer is quite a bit. Behavioral genetics played a crucial role in a clash of ideas between extremes represented by communism and fascism. Both misused genetics.
The politics of behavioral genetics
105
The Nazis believed in the superman who through innate talent and will forged on to victory and ascendancy, overcoming oppressors. The “oppressor” was devious and vile but in the end conveniently weak and often Jewish. Siegfried easily overcame Mime once Mime’s evil was exposed to the light of day. Whereas Mime was garrulous and deceitful, Alberich, Mime’s brother dwarf, embodied the tropes that Jews were cold, renouncing love for gold—so many Jewish names having gelt, gold, silver, or even diamond in them. It was no accident that Nazis turned to social Darwinism and eugenics, and it was a foregone conclusion that their genetic, and—it should also be noted—historical and cultural investigations, would confirm their innate and supposedly Aryan superiority. From this false platform it was a short and illogical step to persecute and murder the “inferior,” who had no right to whatever they had and who despite their weakness and inferiority were somehow a threat, if only because they were weakening the genetic stock of the volk. As a consultant to Berlin’s Charite Hospital Department of Psychiatry I advised on its shameful legacy of Nazi psychiatry. What to do about the portrait of Maximinus de Crinis, its former director, who was the leading academic instrument of the sterilization and even killing of the psychiatrically ill. The Director of the Psychiatry Department Andreas Heinz opted for detailed and impossible to ignore contextualization such that no who passed by could ignore or forget that history. Indeed, people who want to confront that history can easily do so in Berlin—Berlin’s Holocaust Museum with its forest of inhuman monoliths into which a walk is a profound experience. On the other hand, Communists, and their so-called fellow travelers, socialists, in the extreme but often in the main believe that equality of outcome is possible and desirable. In this view, inequalities of outcome reflect discrimination, stratification, and exploitation. Interestingly, they also tend to have a different approach to history than the Germans have adopted—if one doesn’t like it, erase it. To judge equality of treatment, it is sufficient to measure outcome, be it income or a test score. The solution is to manipulate the input until the output is uniform. Equally rich, equally interesting, equally poor, equally mediocre, equally boring. As depicted by George Orwell in 1984 and by Arthur Koestler in Darkness at Noon, an intermediate goal of the collectivist state is the destruction of individuality: in the machinery of the state, people are generic and replaceable parts. The logical next step, implemented by Stalin, was the slaughter of the intelligentsia who, after all, should have been easily reestablished from loyal party stock. The gremlin of individuality was never exorcized from the socialist machine. There was always the problem of outstanding ability and accomplishment. Individuality, as well as history, could not quite be erased. Worker heroes were honored but in honoring them a seed of doubt was planted. The leaders were themselves not the equal of the “masses” in innate intelligence and capability, and in their hearts never believed
106
8. Inheritance of behavior and genes “for” behavior: Gene wars
they were. By their exceptional natures the leaders of these movements posed a problem for a central tenet of their dogma, which, it may be said, had already been run over by their karma. Consciously or unconsciously, innate differences between people had registered in them in the same way that even a schoolchild is aware that some of his classmates are fast, some smart, some slow, and some stupid. The inconsistency between innate variation and the goal of building the perfect socialist state festered. How, or why, can one assure (or enforce) equality of outcome if capability and predisposition vary? What is good for the goose may be good for the gander, but not necessarily for the peacock. In the Soviet Union the concern about individual variation, and the threat of the Darwinist view that something was inherited, led to a peculiar mutilation of genetics known as Lysenkoism. Lysenkoism reprised the Lamarckian idea that experiences of one generation adaptively transmit to following generations, perfecting the adaptation of the species to its environment. Elsewhere in this book, we will see that the modern science of epigenetics has taught us that gene expression and phenotype can be altered by epigenetic change outside the DNA code, and yet we have also learned that the “slate” of such epigenetic change is largely, and perhaps completely, wiped clean in each generation. Although the politburo is still out on that question, Lysenkoism remains as wrong today as it was in 1950. Also, Lysenko wasn’t engaged in genetic analysis at any molecular level, he was simply promulgating science by ukase. Lysenko like his predecessor Lamarck taught that if a giraffe repeatedly stretches its neck to grasp leaves, its progeny will have a longer neck and tongue. This provided a scientific basis to hope that after a few generations of everyone just trying hard all children could have the “right stuff” to be worker heroes or cosmonauts. The main problem for Lysenkoism was that Darwinism had already happened. A central precept of Darwinism was that inferiority is eliminated not by being rewarded but by being weeded out. Giraffes always carried genetic variation that made some taller. Generations of taller giraffes were less likely to go hungry and thereby more likely to pass on their genes (Dawkins, 1996; Dennett, 1996). Here in America, the strongest predictor of the yen to redistribute wealth is wealth. A laudable impulse compels the most successful, who sometimes are like the “lucky winners” described earlier but who often are talented and striving, to lecture the less wealthy. John Updike wrote of a certain type of wealthy person who was always striving for the success of redistributionist policies and candidates, but who never succeeded. More remarkably, and positively, many decide to engage in philanthropy— while retaining a few hundred million dollars, just in case. However, there seems to be something within almost all people, from the poorest to the wealthiest, that makes them strive to accumulate money, cars, houses, airplanes, boats, “collectibles” of all sorts, and even other people. To change
Antipsychiatry: Are psychiatric diagnoses valid?
107
the impulse to accumulate (the selfish impulse), the social system would have to change, but for better or worse, we are stuck with the human impulses we have, rather than those that might best fit any political system.
Antipsychiatry: Are psychiatric diagnoses valid? A foundational critique of psychiatric genetics came from the antipsychiatrists, who argued that psychiatric diagnoses are not real, so why study their inheritance or identify genes that contribute to them? If this is one’s position, it might be disconcerting to discover that whatever is measured as a psychiatric diagnosis shows moderate to high heritability. However, this is not true. There are many heritable traits (for example, height, eye color) that are not diseases. Again, it is somewhat inconsistent that geneticists and neuroscientists have begun to identify highly specific genetic and neurobiological causes of psychiatric diseases, but it is not fatal to the argument. Therefore it is important to take on this foundational critique, which in the end does help impel improved categorization of behavioral pathology. Critics following the trail of Peter Breggin and Thomas Szasz condemn both the pharmacotherapy and diagnosis of psychiatric disease. Breggin, a Harvard educated, published psychiatrist who some call “the conscience of psychiatry,” advocated behavioral treatments are almost always better than drugs and electroconvulsive therapy (ECT) used to treat depression, and that in particular it is wrong to medicate children with attention deficit hyperactivity disorder (Breggin, 1994). Breggin was partly on target. As discussed in Chapter 7, psychiatric diseases are frequently misdiagnosed and overdiagnosed. Pharmacotherapy often is ineffective or makes the situation worse, especially in the context of misdiagnosis, and other treatment modalities are not applied often enough or well enough. Psychiatrists are overly focused on identifying the right pill instead of the many other factors ranging from social dislocation, drug and alcohol use, past and ongoing trauma, lack of exercise and healthy pursuits, grieving loss, physical disability, and lack of purpose that weigh on people. In my time as a clinical director with oversight of a unit treating patients with alcohol use disorder, I have seen time and again that for most patients there is no one key to recovery. Substance use disorders such as AUD, and other psychiatric disorders, are multidimensional in nature. However, while having some merit, antipsychiatry is tendentious and harmful. Many children are indeed misdiagnosed with ADHD or other psychiatric disorders but many others—perhaps the majority—are diagnosed correctly, and for such children medications that increase dopamine levels in frontal cortex can be critical to their ability to learn. By the failure to recognize depression or the withholding of pharmacotherapy and
108
8. Inheritance of behavior and genes “for” behavior: Gene wars
ECT, thousands of depressed patients would die each year—many having taken their own lives—and many more would needlessly suffer. Instead, hundreds of thousands of people with depression are restored to happier and more productive existences and restored also to their families. Similarly, failure to medicate children with attention deficit hyperactivity disorder (ADHD) would set back many in education and socialization, and while it is true that there is a social context of ADHD it is important to recognize that in many cases the uncontrolled hyperactivity would lead to social exclusion. Thomas Szasz (1920–2010, 2011–2012), a Hungarian American psychiatrist, and like Breggin also a lifetime member of the American Psychiatric Association, had a more fundamental critique of psychiatry. It is an analysis of the weaknesses of the discipline that in certain ways cuts right to the quick, and we would all do well to take the criticism to heart even while ultimately rejecting its unhelpful nihilism. In Szasz’s view, psychiatric diagnoses are the reification of socially undesirable behavior and “problems with living” into categories so that those individuals and their behaviors can be “managed.” Thus homosexuality once appeared as a psychiatric disorder in the Diagnostic and Statistical Manual (DSM) of the American Psychiatric Association. When homosexuality became a more socially accepted behavior, it was duly removed. As will be discussed in the next paragraph, this is progress. On the other hand, schizophrenia is a disabling condition that prevents the victim from functioning in most social contexts, regardless of whether we would wish to become accepting of it. Recognition of a disability is an important step in an individual’s ability to integrate and become more functional, even if no specific treatment is available, and for the patient with schizophrenia treatment is available. It is one thing to be uncomfortable around people with impaired capacity. That is wrong. However, it is also wrong to fail to recognize their problem, make allowances, and offer help. The purpose of diagnosis is not to stigmatize but to enable greater specificity in recognition and intervention. Cogently, Szasz believed that to be termed a disease, a condition had to demonstrate pathology at the molecular or cellular level. Szasz would have recoiled at the reconfiguration of addictions as “use disorders.” He died before the advent of large-scale genomic and other molecular and neuroimaging studies that have found genes influencing most of the conditions, including schizophrenia, ADHD, and depression, whose validity as diseases he had questioned. Indeed, and despite recent trends in the DSM classification, it is a main goal of biological psychiatry not just to classify people based on external behaviors and drug them to make the behavior stop but to uncover mechanisms of psychopathology, leading to new treatments, targeting of treatments, and refinement of diagnosis. Szasz was not wrong to insist that psychiatric diseases should be defined with the rigor other diseases are defined. His error was in
Antipsychiatry: Are psychiatric diagnoses valid?
109
maintaining that a heritable, lifelong condition such as schizophrenia was not a disease simply because most of the important aspects of the illness had yet to be identified. During Szasz’s lifetime, and afterward, psychiatry had far to go to catch up with other fields of medicine, being perhaps a hundred years behind. However, the relative backwardness of psychiatry did not seriously challenge the validity of the disease diagnosis of schizophrenia, any more than the inability to see or culture a virus or bacterium challenged the existence of viral and bacterial diseases during the first half of the 19th century. A surgeon could still learn to sterilize their knives. Today, many genes predisposing to schizophrenia have been identified, and furthermore there have been great advances in understanding the neurobiology of this disease. The same is true of other psychiatric diseases, flawed though their nosology and diagnosis remain. In the DSM, many or all psychiatric diagnoses have indeed undergone substantial modification, and new disorders such as borderline personality disorder, premenstrual dysphoric disorder, and seasonal affective disorder have been added. This process continues with each new version. However, experts in psychiatric diagnosis (nosology) are painfully aware of the weaknesses and inconsistencies but think that the continuing evolution of the DSM is a sign of the health of the field, and not a defect. Change is inherent to any discipline of medicine, especially one as immature as psychiatry. A goal of “biological psychiatrists” such as me is to remake psychiatric diagnosis to reflect etiologies, rather than the surface appearance of the same clusters of symptoms. Diagnoses matter, and with both benefits and pitfalls. The U.S. recently endured an epidemic of Childhood Bipolar Disorder (CBD) except that most of the children diagnosed with this disorder probably do not have it, with a 40-fold increase in CHD in the decade before 2003, but only in the US (Moreno et al. 2007) Bipolar Disorder (BD) is moderately to highly heritable, enabling study of children at high risk to develop CBD or BD. Via longitudinal studies of children in high-risk families in which BD is genetically transmitted it has been observed that prior to onset their social, cognitive, and academic performance are typical of children in the general population. In the children in these families who develop BD the first episode typically occurs in mid-adolescence with an episode of depression, and the first episode of mania in late adolescence or young adulthood. As Duffy et al. pointed out, CBD is largely a U.S.-driven phenomenon, children with chronic emotional lability and irritability being the ones most likely to be given this diagnosis by US psychiatrists but not psychiatrists from other countries shown the same interview tapes (Duffy et al., 2020). Many of the children diagnosed with CBD probably have ADHD, and there is little continuity with adult BD, a disease marked by recurrent, acute episodes. Something is wrong, but in
110
8. Inheritance of behavior and genes “for” behavior: Gene wars
most cases the diagnosis of CBD mainly makes lives of children and their families worse, and more medicalized. Already mentioned in Chapter 7 was a major change in criteria for antisocial personality disorder (ASPD). Hare’s psychopathy of only a few decades ago required that the individual be cold and remorseless. Such people are still with us, although there is now no category that specifically names them. The modern definition of ASPD is functional, ignoring internal state, and focusing on problems with behavior. In addition, the diagnosis of intermittent explosive disorder (IED) has disappeared. Too bad, because some people have violent outbursts, often self-injurious or suicidal, initiated by minimal provocation and for no gain. However, there was always a problem with the diagnosis because if the person exhibiting the explosive behavior was inebriated the IED was ruled out. In real life, much explosive behavior occurs while people are inebriated, and most inebriated people do not become violent. Today there is no category that covers IED, and ASPD is a diagnosis that can be made reliably, but not necessarily with as much meaning as previously. Progress in psychiatric nosology has been slow, or retrograde, and awaits application of new types of information from brain imaging, neurocognitive testing, and genetics. In this important regard, the next version of the DSM is unlikely to advance far beyond the present. Diagnoses matter, and misdiagnosis and overdiagnosis have negative consequences. Children diagnosed with ADHD frequently do not have it and may have other origins for their disruptive behaviors that are critical to address. These include abuse and neglect, drug abuse, deficits in vision and hearing, and specific learning problems. In children, there has been a recent epidemic of childhood bipolar disorder, except that most children diagnosed with this disorder probably do not have it. Their parents have read about the new disorder and found a psychiatrist to prescribe medication. The drugs often make things worse, usually do not make things better and almost always making things more complicated. Something is wrong with these children, but treatment of childhood bipolar disorder seldom contributes to the solution. Also, psychiatrists, neurologists, and other physicians can help; however, very little of it can be easily resolved either with medication or with talking therapies. Therefore it is never a good thing for people to establish one of these behavioral patterns each of which compensates, in some very indirect and tortured way, for the developmental insult that they received. The longer the tree grows in its bonsaied shape, the more irreconcilably distorted it becomes. When there is a popular movie about multiple personality disorder (e.g., Three faces of Eve) it becomes more likely that a person will adopt the disorder as the expression of her angst. If people hear accounts of alien abductions and bodily violations, then expect an epidemic of such cases. In my admittedly limited experience with patients with multiple
Antipsychiatry: Are psychiatric diagnoses valid?
111
personality disorder, none was ever in a personality that was enthusiastic about having her or his blood drawn. Each of these stress-induced disorders, including MPD, represents an instance in which an individual, with whatever genetic vulnerability or resilience, was channeled into a pathological, unchosen, pattern of behavior. However, the peculiar challenge for physicians in evaluating so-called factitious behaviors is to avoid making matters worse by failing to recognize the origin and nature of the all too real problems patients with these disorders are experiencing. The physician can either expand or restore a patient’s ability to choose or delimit or destroy. The behavioral problem, be it swallowing, seizures, multiple personalities, or consumption of a drug, may be reinforced by misguided treatments or the mislabeling. Hans Castorp, Thomas Mann’s protagonist, who was perhaps only ever sick in soul, might have been self-confined for the rest of his life in the tuberculosis sanitorium atop Magic Mountain. For many, primarily women, the popularization of multiple personality disorder resulted in it becoming the expression of internal torment. Let us say that a person experiences an early life trauma such as sexual abuse. Surveys, for example by my colleague University of Arizona professor Mary Koss, have revealed that between one-third and one-half of women are sexually abused, and perhaps half as many boys. Some are less resilient or experience worse or more prolonged abuse or are abused during a more critical time in development. Acutely, the sexually abused child may have expressed their distress in a variety of ways including depression, anxiety, suicidality, and other types of behavioral acting out. The Three Faces of Eve, a popular movie about multiple personality disorder, probably led some people to adopt this mechanism as the expression of their angst. Perhaps many had better alternatives. The multiple personality disorder itself can lead to further difficulties and complications. If people hear accounts of abductions and nasty sexual experiments by aliens, then expect an epidemic of alien abductions, with attendant details. Multiple personality disorder is an extreme manifestation of internal distress, but victims of sexual abuse can manifest their internal turmoil in a variety of other remarkable ways including hysterical blindness, seizures (pseudoseizures), and movement disorders (psychogenic movement disorders). Each of these secondary problems is severe, has its own disabling consequences, and can lead to inappropriately targeted medical intervention that only complicates matters without addressing cause. Simple treatments, for example with placebos (as a resident I temporarily cured hysterical blindness with a tuning fork), can play a useful role if the root cause, for example continuing sexual abuse, is prevented. Later in life, and as we found in research in Native American communities in studies both with Koss (Koss et al., 2003) and also with Rob Robin (Robin et al., 1997), the effect of childhood sexual abuse is to increase the risk of several psychiatric disorders, all of which were already common in women.
112
8. Inheritance of behavior and genes “for” behavior: Gene wars
A woman who was sexually abused as a child is two to three times more likely to have depression, alcoholism, other addictions, antisocial personality disorder, or posttraumatic stress disorder. She is also more likely to have somatic symptoms such as headache and fibromyalgia, and to have psychogenic movement disorders, pseudoseizures, mysterious weaknesses, or hysterical blindness. Sometimes the person becomes nearly unable to swallow, with either no or minimal organic cause. Therefore childhood trauma, and especially sexual abuse, can be regarded as a rising tide that lifts all psychiatric disease ships. The consequences of these secondary psychiatric disorders can be devastating, leading to the transgenerational and lateral transmission of problems with family networks, for example if the result is alcohol use disorder. A girl with pseudoseizures or a psychogenic movement disorder can be incapacitated, with her ability to form good social relationships and perform effectively at school seriously damaged. Psychiatrists, neurologists, and other physicians can help; however, very little can be easily resolved either with medication or with talking therapies. Therefore it is never a good thing for people to establish one of these behavioral patterns, each of which compensates, in some very indirect and tortured way, for a developmental insult. The longer the tree grows in its bonsaied shape, the more irreconcilably distorted it becomes, and it distorts the environment around it. Medical diagnosis powerfully interacts with choice, as anyone who has signed an advance directive or seen someone committed is aware. Diagnoses, especially psychiatric ones, but the same is true for all, represent conditions that delimit the ability to choose, expand the range of options people have in their lives. However, diagnosis itself can expand or delimit choice. Firstly, though the exceptions to this principle are critical—psychiatry has repeatedly been used to repress, incarcerate, marginalize, and demonize. The damage done far exceeds that done by people to themselves who impose false limits on themselves, and for example by deciding via whatever convoluted thought process, that they do not have free will (Kane, 2005). Psychiatry has often been used by others to impose those limits, including by convincing people who may be a little sick that they are defined by a diagnosis. It is often essential for a person with an illness, and for example alcohol use disorder or schizophrenia, to acknowledge their illness and thus to begin living a better life, and as defined by opening more options for choice. However, the patient with alcohol use disorder is a patient with AUD, not an “alcoholic.” The patient with schizophrenia is not a “schizophrenic.” Thomas Mann’s Hans Castorp was a man with mild fever, perhaps from fevered glances at a woman with Kyrgyz eyes. Even if Castorp had tuberculosis, a consumptive disease, he was not a “consumptive.” Salient examples in which the labels of psychiatry have been used to define, repress, and even exterminate people demand acknowledgment.
Antipsychiatry: Are psychiatric diagnoses valid?
113
They include the eugenics of Nazi Germany, in which de Crinis was not an isolated example. Vast numbers of psychiatric patients were sterilized or murdered. In the Soviet Union, dissidents were psychiatrically imprisoned. The sterilization, brutalization, and warehousing of the p sychiatrically ill lasted well into the 20th century in the United States and most other nations. However, no branch of medicine is without its historical absurdities and flagrant cruelties. Trephination and bloodletting helped no one. For syphilis, mercury was the remedy of choice for 400 years and into the 20th century, finally being replaced only in 1910 by Nobel Laureate Paul Ehrlich’s partly effective Salvarsan whose active ingredient was an arsenic derivative, and only in the 1940s by penicillin. If psychiatry has lagged a few decades behind other fields of medicine (only in 1905 did Schaudinn and Hoffmann discover the cause of syphilis), it is because psychiatric illnesses were more mysterious, intractable, and frightening. Psychiatric disease is more frightening, and stigmatizing, in part because within any insightful person another person’s psychiatric illness resonates with an inner feeling that the seeds of such illness are also within each of us, perhaps to germinate, and leading to the loss of control and inability to self- actualize that we can observe in those who are psychiatrically ill. However, 21st century psychiatry is not 19th century psychiatry or mid-20th century psychiatry. Among the most important lessons of history is that things change. Nowhere is this truer than in medicine. Step by step, and with treatments discovered serendipitously such as lithium for bipolar disorder, and more largely by design such as clozapine for schizophrenia and SSRI antidepressants and ECT for depression, and drugs such as naltrexone and varenicline to ease addiction to—yes—other drugs (i.e. ethanol and nicotine) psychiatry has moved forward, and it has become possible to empty asylums largely because of that, much as tuberculosis sanitoriums and leprosy colonies (e.g. the island of Molokai) could be shuttered not because doctors treating TB patients became more beneficent but because they had new treatments and diagnostics. Yearly in the United States and worldwide, millions of lives are saved, and lives expanded by the diagnosis and appropriate treatment of depression. More than two million Americans have schizophrenia, and the majority are treated, freeing these patients from the coercion of false voices and delusions, and in turn freeing their families and caregivers, who otherwise are at many moments of the day in thrall to pathology, and would by necessity be confined to long-term psychiatric care facilities (asylums). The substance use disorders, including alcohol use disorder, gambling, and various illicit and prescribed drugs, affect >20 million Americans. Addicted people have a relapsing–remitting disease that constantly channels their behaviors, and their very thoughts, down repetitive and wellworn pathways. Many are helped by treatment. Each of these diseases is substantially heritable and the neural basis of several came to be (partly)
114
8. Inheritance of behavior and genes “for” behavior: Gene wars
understood in the late 20th century and early 21st century, much as the basis of syphilis came to be understood only in the early 20th century. In the bootstrap operation to refine and improve psychiatric diagnosis and treatment the views of people such as Szasz and Breggin, while authentically resonating with parts of what is wrong in psychiatry, are unbalanced by the good psychiatry does. Psychiatry is thus a medical discipline that is flawed and incomplete, probably always a work in progress(Kendler, 2008, 2012), and always inconsistent across and within generations of its practitioners, but that nevertheless helps people. Psychiatric diagnosis is a way station on the path to something better, but for all its imperfection it is better than the nihilistic alternative. The answer is not to roll back two centuries to a darker era when there was little recognition and no help for problems such as anxiety, depression, addictions, eating disorders, and psychoses. No serious person can think that it was better when the mentally ill were labeled the village idiot, packed off to asylums, burned at the stake as witches, or neglected. Each year in the US, thousands of lives are saved by diagnosis and appropriate treatment of depression. More than two million Americans have schizophrenia, and the majority are treated. Family life with an untreated person with schizophrenia is difficult, and that person is suffering. The addictions, including alcohol use disorder, other substance use disorders, and gambling, affect over 20 million Americans with relapsing–remitting disease. It is incorrect to say that they are being diagnosed for social convenience. They are hurting and many are helped by treatment. Most psychiatric diseases are substantially heritable and the neural basis of several is at least partly understood. In the bootstrap operation to refine and improve psychiatric diagnosis—where gains have been slow and have had to be measured across generations—Szasz and Breggin are far off the mark. Yet their critiques have value in representing a challenge to dogma and to remind us of the failures of psychiatry. Putting it most simply, they are saying we need less, but what we need is more. Having moved past the denial stage, we can approach the challenge of achieving a more complete understanding of the genetic and environmental forces, and neurobiological processes, that underlie psychiatric disease and thereby better prevent and treat them. The focus on process and etiology will not provide all the answers to problems that represent the end stage and late stage of pathological processes, but already is yielding dividends.
References Bevilacqua, L., Doly, S., Kaprio, J., et al., 2010. Population-specific HTR2B stop codon predisposes to severe impulsivity. Nature 468, 1061–1066. Breggin, P., 1994. Toxic Psychiatry: Why Therapy, Empathy and Love Must Replace the Drugs, Electroshock and Biochemical Theories of the “New Psychiatry”.
References 115
Caspi, A., McClay, J., Moffitt, T.E., et al., 2002. Role of genotype in the cycle of violence in maltreated children. Science 297, 851–854. Churchland, P., 2011. Braintrust: What Neuroscience Tells Us about Morality. Princeton Press. Damasio, A., 2005. Descartes’ Error: Emotion, Reason and the Human Brain. Dawkins, R., 1976. The Selfish Gene. Dawkins, R., 1996. The Blind Watchmaker. Dennett, D., 1996. Darwin’s Dangerous Idea: Evolution and the Meaning of Life. Duffy, A., et al., 2020. Pre-pubertal bipolar disorder: origins and current status of the controversy. Int. J. Bipolar Disord. 6, 18. https://doi.org/10.1186/s40345-020-00185. Goldman, D., Oroszi, G., Ducci, F., 2005. The genetics of addictions: uncovering the genes. Nat. Rev. Genet. 6, 521–532. Kane, R., 2005. A Contemporary Introduction to Free Will. Kendler, K.S., 2008. Explanatory models for psychiatric illness. Am. J. Psychiatry 165, 695–702. Kendler, K.S., 2012. Levels of explanation in psychiatric and substance use disorders: implications for the development of an etiologically based nosology. Mol. Psychiatry 17 (1), 11–21. https://doi.org/10.1038/mp.2011.70. Koss, M.P., Yuan, N.P., Dightman, D., Prince, R.J., Polacca, M., Sanderson, B., Goldman, D., 2003. Adverse childhood exposures and alcohol dependence among seven Native American tribes. Am. J. Prev. Med. 25 (3), 238–244. https://doi.org/10.1016/s07493797(03)00195-8. 14507531. Moreno, C., Laje, G., Blanco, C., Jiang, H., Schmidt, A.B., Olfson, M., 2007. National trends in the outpatient diagnosis and treatment of bipolar disorder in youth. Arch. Gen. Psychiatry 64 (9), 1032–1039. Murray, R.J., Herrnstein, C., 1994. The Bell Curve: Intelligence and Class Structure in American Life. Pinker, S., 2003. The Blank Slate: The Modern Denial of Human Nature. Robin, R.W., Chester, B., Rasmussen, J.K., Jaranson, J.M., Goldman, D., 1997. Prevalence, characteristics, and impact of childhood sexual abuse in a southwestern American Indian tribe. Child Abuse Negl. 21 (8), 769–787. https://doi.org/10.1016/s0145-2134(97)000380. 9280382. Rutter, M., 2006. Genes and Behavior: Nature–Nurture Interplay Explained. Szasz, T., 2010. The Myth of Mental Illness: Foundations of a Theory of Personal Conduct.
This page intentionally left blank
9 The scientific and historic basis of genethics Ring them bells for the chosen few Who will judge the many when the game is through—Bob Dylan Who Watches the Watchmen?—Alan Moore Quis custodiet ipsos custodes?—Juvenal Some scientists—whether from neglect, ignorance, ideology, greed, ambition, or sheer Tralfamadorian curiosity—cannot police themselves or the constituencies they represent. Then, on whom can we rely? The most durable and generally applicable model we have is a system of rigorous human research review and protection that was originally created in response to the exploitation of captive populations during the 1940s, and that ever since we have been trying to perfect. Remarkably, the principles, which are few, have been found adaptable to diverse ethical challenges, including the protection of people taking part in genetic and psychiatric studies. In addition to laying out the historic framework of genetic research, and the events that necessitated human protection, the meta-purpose of this chapter is to explore how questions of free will are foundational to any system of ethical research. This is done by recursion to specific problems in the process and ethics of genetic research and application of genetic findings. The guiding principle of clinical research is to treat all individuals as autonomous and worthy of respect as consenting agents, directly implying that we regard people as having capacity for free choice. Otherwise, informed consent is an impossibility. The broader societal genethic context involves respect for privacy and freedom from genetic discrimination. If people are slaves to causation, why should we treat them with dignity, as if they are autonomous agents? If they are not free, why not eugenically replace or redesign them, as we do machines (Witkowski and Inglis, 2008)? We do not ask the consent of our car before we drive it or redesign it, because a car is a machine. It may not start when we turn the ignition key, but not because it has chosen. For human research, as in other human-to-human interactions, are
Our Genes, Our Choices https://doi.org/10.1016/B978-0-443-22161-3.00023-5
117
Copyright © 2024 David Goldman. Published by Elsevier Inc. All rights reserved.
118
9. The scientific and historic basis of genethics
our “ethical” practices only an expedient contrivance to perhaps be disregarded when it becomes inconvenient, and as happened before, or do these rest on deeper bedrock? Although the ethical conduct of human research is founded on three simple principles, its structure and procedures continually evolve in response to new capabilities, complexities, dangers, and potential benefits inherent to modern biomedical research. Arguably, the few principles that are at the core of human research ethics, like Isaac Asimov’s famous “Three Laws of Robotics” (which, as I demonstrate in “Immortal,” do not really work!) (Goldman, 2021), are sufficient to guide oversight of human research into the far future. The Belmont Report (The Belmont Report, 1978) put forward three “laws”: respect for persons, beneficence, and justice. Everything that follows, including informed consent, minimization of risk and maximization of benefit, and equality of treatment and access to research, is built on these cornerstones. Although the ethical conduct of human research is not essential for ethical medicine, these are also intertwined: people expect the right of consent as to what is done with their bodies, and with their lives at stake. We will briefly sketch historical evidence that the conduct of human research cannot depend merely on the individual goodwill and judgment of the scientist. We will see that the need for independent gatekeeping also applies to the conduct of science and its funding and publication, with which genetic and all human research is closely intertwined.
Standards of science and evidence The quality and validity of science is the first issue to consider in deciding whether any proposed human research is ethical or any finding is robust enough to be used in society. To evaluate science, the best system that has evolved is peer review. If the results that will be produced are trivial or invalid, then the risk/benefit ratio of the project is unacceptable. This is precisely the claim of many critics of psychiatric genetics: the science is questionable, so any human research performed to investigate genes and behavior is also to be questioned. However, they would probably be less happy with how this is accomplished because the question of what can be done, what can be published, and how valid and applicable is mainly governed by peer review. Peer review is a process whereby on a constant basis scientists struggle to overcome and learn from usually anonymous critique. It works surprisingly well despite glaring examples of scientific fraud and misconduct that leak through the barriers. The main problem is work that is not up to standard but that for whatever reason survives the winnowing. Typically, new human research protocols are reviewed by scientific review c ommittees (SRC) prior to review by the IRB—the Human Research Protocol Committee.
Standards of science and evidence
119
For a decade apiece, I chaired the IRB of an NIH institute (NIAAA) and an SRC representing two institutes (NIAAA and NIDA). I saw that, surprisingly, a review process that is critical, and even harsh, is well accepted by scientists—the main issue being consistency and a perception that people on both sides of the review are genuinely trying their best. Most scientists bemoan the fact that too much error survives, and their concern is healthy and vital: we must look at the landscape of science, and ourselves in the mirror, and see at least some defects as they are. This means valuing gadflies of science, even while shooing them away when they swarm too closely. One reason peer review works is that scientists are much better at detecting flaws in what someone else has done and given anonymity, and largely even if named, are not shy in identifying defects. Anyone who has built a house will know that the carpenter, plumber, electrician, painter, and drywaller will with little prompting point out defects in the others’ work. Peer review in action: Ann Elk (Free University of Python) develops a theory of brontosaurus, the name of which was inaptly changed to apatosaurus: “All brontosauruses are thin at one end; much, much thicker in the middle and then thin again at the far end. That is the theory that I have, and which is mine and what it is, too.” This theory becomes known to the public but is never (to my knowledge) reported in a scientific journal. Also, it does not lead to a coherent, fundable project. Sadly, Dr. Elk had comparatively little impact on the field of paleontology. Commitment to the integrity of the scientific process and a feeling that everyone should play by the same rules explains why scientists are peeved when mediocre work or a bad idea survives peer review or is reported through the press. Of course, it is precisely those studies that make outlandish and improbable claims that are most likely to be highly publicized. Humorous examples might include the “Theory of the brontosaurus,” “Neanderthals invented Football,” “Effect of negative reinforcement on cognitive ESP ability,” “Oprah’s roots: Thirtieth generation granddaughter of the Queen of Sheba,” “God gene discovered, and it is a vengeful gene.” But, although Ann Elk was merely a funny story, in my lifetime I have lived through the unfunny “discoveries” of cold fusion and life on Mars. These examples illustrate that science is self-correcting, but it takes time and effort. Scientific review does not eliminate, but reduces, the level of background noise, be it “cold fusion” or proof of “ESP cognitive ability.” This is worth defending, and it is logical that people are passionate in their defense of it. As compared to most types of science, clinical research is far more expensive, usually requiring both institutional support and peer-reviewed funding. An Ann Elk-type clinical study purporting to demonstrate ESP cognitive ability recently survived peer review and was published in a psychology journal. However, defective clinical research studies often do not survive peer review for publication, or only some aspects survive. Critically, all publicly and privately funded clinical trials are now registered on Clinicaltrials.gov, such that major endpoints are
120
9. The scientific and historic basis of genethics
defined at the beginning and summary results reported at the end of the study. Compliance is imperfect but is improving yearly. In a field such as psychiatric genetics, researchers worldwide are engaged in a highly competitive, and ultimately self-correcting, race for knowledge. No criticisms of nonexpert gadflies approach the difficulties raised by the insightful probing by editors and reviewers of high-level scientific journals. Suspicion that science, ideology, and other motivations are inextricably mixed is historically based. In past times the line between data and synthesis was more often blurred. It was not uncommon for experiment, analysis, and speculation to appear in the form of a book. It is a great advantage to demarcate synthesis from results, and while it has undoubtedly cost us some great books and led to scientific articles being rigidly organized (Introduction, Methods, Results, Discussion and, nowadays, 100 pages of Supplementary Material), we have gained in other ways. In writing about psychiatric genetics, I have been able to rely on findings “battle-tested” by peer review, but interpretation of these discoveries is a different matter. One may reasonably question individual results but trying to undermine the overall genetic and neurobiological basis of these conditions soon forces one to part company with a massive and multidimensional body of evidence. On the other hand, there are other legitimate perspectives on the meaning of that evidence than the theory of emergent free will that I am elaborating in this book. One of my purposes in writing Our Genes, Our Choices was to present a synthesis beyond what would be appropriate for any scientific journal, in terms of scope, tone, and speculative content. This book certainly would not survive the demands of traditional peer review, and yet is built on foundations created by peer-reviewed science.
Ethics of research: Trust, but verify Clinical investigators are constantly reminded and educated about principles and practices essential to the protection of human subjects, and most investigators are meticulous and ethical. Nevertheless, we rely on thirdparty review and oversight for compliance. The requirements faced in clinical research, which is also called translational research, are formidable, and it can be reasonably asserted that many of the techniques, and underlying knowledge, of modern medicine (transfusions, most surgical techniques, vaccinations) would be unknown or vastly delayed except that these were tried unsystematically and without formal institutional approvals. Today, standards of scientific evidence and ethical approvals define a completely different landscape. At the minimum, the researcher seeks the sanction of several committees, including a human research committee and a scientific review committee. These committees review human research protocols that define hypotheses, procedures, analytic design, hazards, and precautions.
Ethics of research: Trust, but verify
121
As will be seen, this infrastructure of research oversight—now constantly being perfected—was constructed beginning after World War II (Emanuel et al., 2011; The Belmont Report, 1978). The informed consent is the most important part of a research protocol because it provides the prospective participant with the information needed to make a meaningful decision as to whether to participate, and to realistically anticipate benefits and hazards. It is the informed consent that most dramatically expresses recognition that the research participant is a free agent, and pains are taken to evaluate and protect individuals who for whatever reasons—impaired cognition, youth, economic or physical vulnerability—are regarded as unable to give consent (even if capable of assent, as children are) or who are vulnerable to coercion (for example, prisoners) (Emanuel et al., 2011). The written consent is a component of a process that also includes discussions with the researcher or other staff. Yet, much hinges on it, and therefore it is surprising that we do not know the best length, level of detail, or format. It can run to many pages and contain several complex procedures, advanced concepts, and esoteric terms. Faced with this complexity, many people revert to their preliminary decision to participate, and that decision is based largely on a general understanding of the study and faith in the researcher’s integrity and beneficence. Investigators and those responsible for oversight are aware of the problem, but there is a constant battle between simplification and the need for complexification such that the consent contains everything needed to make an informed decision. Human research committees include people of different backgrounds, including medical, scientific, and from outside the medical/ scientific establishment who can understand the protocol and consent in all its aspects, and who endeavor to make the consent complete, understandable, and meaningful. Additional oversight is usual. Use of an investigational drug or device requires Food and Drug Administration (FDA) approval. Use of ionizing radiation triggers special review. An overarching concern is risk/benefit ratio. Scientific review is required on the importance of the science and validity of the design: Does the study pose an important question, and can it answer the question it poses? Expedience, as well as misguided sense of beneficence, exerts a constant pressure against ethical oversight. The great majority of clinical investigators, even those who are top in their field, passionately hate this process which occupies so much of their time, and which not only delays science and adds to the exorbitant cost but also stops many studies from ever happening. As a group, clinical investigators are exceptional individuals, with the temperament, scientific expertise, training in research (often from a mentor experienced in human studies), ability to write and communicate, and resources. Nevertheless, much interesting, and potentially beneficial
122
9. The scientific and historic basis of genethics
research never leaves the starting gate. Many studies that would benefit humankind and some of the research subjects never happen either because of the resources that are diverted or because of barriers that cannot be overcome. Perhaps it is not surprising that as a psychiatric geneticist I find that studies involving genetics and psychiatry are among the ones that attract disproportionate concern. Psychiatric disorders, particularly addictions, are feared and stigmatized, so there is often more care in reviewing a study of schizophrenia than a study of AIDS. Greater attention is paid to so-called genetic studies even though nowadays almost every study is in some sense a genetic study, even if no DNA is collected. However, genetic studies involving DNA deserve very careful scrutiny. In no study involving DNA or the collection of a biological sample containing DNA can we ever provide absolute assurance of confidentiality, because the DNA itself can identify the individual. Modern genetic research is rarely marred by violations of patient confidentiality or interest. An exception was the Jesse Gelsinger case, which occurred in the context of a clinical trial (Stohlberg, 1999). More recently, in 2018, He Jiankui performed CCER5 gene editing of twin girls in an unnecessary attempt to protect them from HIV, which may actually lower their life expectancy (https://www.nature.com/articles/d41586-019-01739-w). The Gelsinger gene therapy study survived human protocol review, and He’s study was done without proper review, but as will be discussed next, the history of science proves that intensive review is a very necessary evil, to prevent greater evils, including the erosion of the conception of humans as individually free, autonomous, and worthy of respect—foundations articulated and systematized in the Nuremberg Code and later in the Declaration of Helsinki and Belmont Report (Emanuel et al., 2011; The Belmont Report, 1978). It is a natural inclination to hold the “good” of scientific advancement and the public welfare above the welfare of the individual. It is natural myopia to think that one’s own research is justified based on the “general good.” Nazi scientists, who constituted a major portion of the German scientific research establishment, conducted a variety of involuntary studies designed to test the limits of human endurance. They justified the involuntary suffering of their victims with the thoughts that their horrific experiments were for the greater good of those who were like themselves (Lifton, 2000). However, the only good thing that ever came out of human research by Nazi scientists was the Nuremberg Code. At the Nuremberg trials, the defendants were able to point out examples of highly reprehensible studies carried out in other countries, and frequently on unsuspecting individuals and vulnerable individuals: children, orphans, the poor, the mentally incompetent, and those socially marginalized by race. Furthermore, such experiments went on for decades afterward. There were many of these experiments and if the goals might have been laudable—to prove the origin of scurvy or the role of an
Genes, jobs, and groups
123
infectious agent in disease—their ends did not justify the means. After the introduction of the Nuremberg Code, at Willowbrook State School in the 1950s and 1960s, Saul Krugman investigated hepatitis by injecting newly arrived, severely retarded children with different strains of hepatitis. The study was an advance in research ethics in that the parents “consented,” but was the research ethical? As detailed in 1995 by the Advisory Commission on Human Radiation Experiments, appointed by President Clinton, in the dozens of studies conducted in the Cold War era, thousands of unknowing individuals including children were injected with radioactive isotopes or even exposed by deliberate radioactive contamination of the air to study the effects of radioisotopes and their elimination from the human body.
Genes, jobs, and groups For most genetic studies conducted so far it is difficult to show harm, although one can imagine how harm would occur. In 1972 the National Sickle Cell Anemia Control Act was passed by Congress, withholding funding from states unless testing for this disease was voluntary, and discouraging mandated testing which might stigmatize Black Americans. However, almost genetic studies involve only the collection of information, and the information is not predictive enough to have much of an impact, should confidentiality be breached. Therefore the risk is mainly in the domain of information rather than effect, and the genetic markers are not yet so informative compared to other indicators such as measurements of levels of virus in blood, drug screening results, and computed tomography (CT) scans. Simply put, the information that someone has a condition is far more powerful than a genetic marker that only weakly predicts the risk of that condition. Having discovered a genetic marker for disease or vulnerability, what is to stop man from using it in any way that may be expedient? Only rules, based on a premise that individuals are deserving of respect and privacy. No case of employment discrimination on a genetic basis has ever been brought before a court, although the Equal Employment Opportunity Commission (EEOC) successfully settled one such lawsuit. In that lawsuit, the EEOC maintained that the Burlington Northern Santa Fe Railroad had secretly tested employees for a rare genetic condition that leads to repetitive stress injuries, including carpal tunnel syndrome, and the Railroad was also screening for common medical conditions such as alcoholism and diabetes. In 1998 a lawsuit on preemployment genetic testing at the Lawrence Berkeley National Laboratory was decided in favor of the employees. The Council for Responsible Genetics claims to have documented hundreds of instances of genetic discrimination that did not go to court. Although, as pointed out in a paper published in a
124
9. The scientific and historic basis of genethics
genetics journal in 2003, these claims are almost all unverified and might be seen in a different way if all the facts were known, the examples are representative of what people are concerned about and are instructive. For example, a genetic test revealed that a boy had fragile X syndrome, leading an insurance company to drop him from cover based on a preexisting condition, and in another case, it is stated that a social worker lost her job because she mentioned that her mother had Huntington’s disease, giving her a 50% chance of developing it. Note the nuances. In the Railroad case, the most powerful information collected by the company was that on diabetes, Alcohol Use Disorder, and other medical conditions. The yield on the test for the rare genetic condition was probably zero. In the case of the boy with fragile X syndrome, it is incorrect to say that the genetic test was “predictive” of fragile X syndrome. The boy already had the clinical symptoms of this syndrome, which has a genetic basis and for which there is a genetic test. He also might have been more crudely diagnosed with fragile X syndrome by other means. In the case of the social worker at 50% risk for Huntington’s disease due to family history, no genetic test was performed, and in fact a genetic test might have revealed that she was not at risk. However, she was discriminated against because of her family history. In the same way, she could have suffered discrimination because she had relatives with breast cancer, schizophrenia, obesity, or alcoholism: all are genetically influenced even though there is no reliable genetic test for any of them. However, this should not make us feel very secure in the long term, looking toward those more distant horizons when such tests will be available. For example, polygenic scores are today highly predictive of risk of morbid obesity (BMI >40), a condition common in the U.S. and associated with higher health care costs, loss of workdays, and reduced longevity. As shown in the UK Biobank sample, and by applying summary statistics from GIANT, a very large GWAS study of physical characteristics, UK residents in the highest decile of BMI polygenic score had a 7% risk of morbid obesity, and those in the lowest decile were at virtually no risk—at least for obesity. Such information has profound implications for use and misuse, and in the same ways that information on smoking, age, and sex are actuarially applied.
The Genetic Information Nondiscrimination Act The Genetic Information Nondiscrimination Act (GINA), sponsored by Representative Louise Slaughter, and signed into law by President Bush in 2008, remains the key development so far in the protection of individuals from genetic discrimination (https://www.eeoc.gov/statutes/genetic- information-nondiscrimination-act-2008). GINA, rather than the genome or his role in helping to lead the response to the COVID-19 pandemic
The Genetic Information Nondiscrimination Act
125
(which despite upwards of a million deaths in the US alone, arguably saved lives), perhaps will be seen as Francis Collins’ greatest legacy as director of the National Human Genome Research Institute (NHGRI). The informative full title of GINA is “An act to prohibit discrimination on the basis of genetic information with respect to health insurance and employment,” which includes denying insurance coverage or charging higher premiums or making job hiring, firing, promotion, or placement decisions. Some gaps remain. GINA does not cover life, disability, or longterm care insurance policies and it does limit an employee’s ability to sue. Worryingly, employers and insurers cannot discriminate based on the genetic finding, but they can do so based on the disease itself, and as mentioned, most diseases are genetically influenced. In many cases the boundaries between genotype and disease will be hard to define, if not impossible. For example, if a person has a genetic marker for an enzyme deficiency or an unfavorable polygenic score for obesity, GINA will prevent an employer from selectively refusing to hire them on that basis. Why would an employer do this? Because healthy individuals are more productive. However, what if they had the elevated cholesterol levels and evidence of early atherosclerosis that correlated with that marker? What if they are overweight, a condition that is strongly influenced by genes, as the polygenic scores and twin studies show? Or what if they had very early signs of Parkinson’s disease that correlated with a different genetic marker? Even though no genetic testing had been done, GINA would prevent the social worker from being fired based on her family history of Huntington’s disease, because the law prohibits information on one member of a group from being used to provide information about others. GINA protects people against misuse of their family histories. However, GINA does not protect them from signs of risk already manifest in some other way, even molecular. A person can lose insurance or employment based on a blood chemistry result, a brain scan, a psychometric test, or body mass index directly related to genetic risk. A person’s body mass index can be more informative than a polygenic score predicting body mass index. This is a very fine, and somewhat illogical line created in good part by lack of universal health care. If a child is found to have sickle cell anemia by a DNA test showing that they are a hemoglobin S homozygote (both copies of the beta globin gene carry the valine amino acid substitution as determined by DNA, RNA, or protein analysis), then GINA will shield them from insurance discrimination. However, if a technician puts a drop of blood under a microscope and determines that the child’s red blood cells distort into a sickle shape, making the diagnosis of sickle cell anemia, then the child has a preexisting condition. Also, perhaps the child’s race or parentage tipped off the physician. Does the physician also order a DNA test? Probably. Why not? It is inexpensive and precise. Does the child benefit from having the diagnosis made? Definitely. It could save them much s uffering
126
9. The scientific and historic basis of genethics
and could save their life. Does the child benefit from being labeled, for insurance purposes, as having a preexisting condition? No. It is illogical that the legal status of a genetically influenced disease depends on whether the information is defined as “genetic.” However, the culprit is not GINA, which is a stopgap remedy to reduce discrimination. The long-term solution is a system of health care in which people with preexisting conditions have equal access. When that goal is fully realized, the method—“genetic” or “nongenetic”—by which the diagnosis was made will be rendered irrelevant, doctors can better get on with using the most efficient and accurate tool available, and individuals will not, because of the expedient need of a third party, be subject to genetic discrimination.
Gene therapy The likelihood that a typical research genetic study will have a significant impact on a patient’s health is low, and largely confined to the sphere of genetic information and discrimination, which we have just discussed. However, there is one domain that stands out as an exception, and this is gene therapy studies. In these interventional studies, the consequences can be profound, for good or bad, and the temptation to use individuals can lead to abuses. In 1999, and as part of a gene therapy trial, Jesse Gelsinger was administered an adenovirus that contained the ornithine transcarbamylase (OTC) gene (Stohlberg, 1999). Four days later Gelsinger died of massive liver failure. It was already known that two patients had serious side effects, and the informed consent did not mention that some monkeys who received the treatment had died. Gelsinger suffered from OTC deficiency, an X-linked enzyme deficiency impairing the metabolism of ammonia, which is made when the body digests proteins. However, unlike infants with severe deficiency, Gelsinger was a young man in reasonably good health on medications and dietary protein restriction. This was because some of his cells were normal, his being a genetic “mosaic” (here is one of the exceptions; not every cell in one’s body necessarily has the same DNA). His participation in the study was therefore altruistic. He did not stand to gain as much from treatment, and furthermore he had high ammonia levels that according to the protocol should have excluded him. According to the father, in U.S. Senate testimony, investigators had said “the worst that could happen in the trial would be that he would have flu-like symptoms for a week.” The lead investigator and the University of Pennsylvania both had financial stakes that were not disclosed to the research participants and probably not to the institutional review board (IRB), under rules in place at that time. The university later paid the family an undisclosed amount.
Group consent and individual consent
127
Group consent and individual consent In studies I have done on alcohol use disorder and related psychiatric diseases in Native Americans, my lab has always carried out the research in the context of the tribal community, with group consent as well as the consent of individuals. We made this decision because of the strong group identification of many of the participants and nonparticipants who might also be affected by the research. This is a dramatically different way of conducting a human research study, as it is not strictly necessary and is certainly not expedient. For example, many Native Americans, including many who we studied, do not live on reservations. They could be studied under auspices of protocols that do not necessarily target Native American tribes, but which include and identify some of their participants as Native Americans or even as descendants of a particular tribe. The concept of group consent is potentially applicable to other groups, but it is debatable whether it should be. In similar fashion, researchers study Ashkenazi Jews, Japanese Americans, Chinese Americans, and individuals of ancestries and cultures. Even if they have not revealed their ancestry, it can be determined via ancestry informative markers, and indeed this is important for the validity of most genetic studies and for many studies that are not on the surface genetic. In this regard, both ethic identity and measured ancestry have become essential variables in human research, ethnicity being to ancestry as gender is to sex. For example, the polygenic scores tend to perform (predict) more poorly in people of different ancestry than in which the underlying statistics for the polygenic score was originally derived, A second type of “class” focus human studies can have, and often do, is selection of participants with a certain disease. Thus all people with Tourette syndrome (a movement disorder in which there are also vocal tics) can be considered a class, and they might be affected by any discovery relevant to their disease or if someone should write about their disease in such a way that they could be stigmatized. For example, David Comings, an expert in Tourette syndrome and diseases involving behavioral dyscontrol, reported that the gene for Tourette’s makes people hypersexual (Comings, 1994; Comings and Comings, 1985). Comings’ gene, and the assertion about sexuality of Tourette patients, remains unconfirmed. A third kind of class focus is the family. Whenever someone in a family is studied for a genetically influenced trait, and especially if the study is a genetic study, other members of the family are affected. This is also true if the trait is culturally transmitted. In most f amily studies a family history is obtained. The person interviewed provides information, including sensitive information on other family members. Those relatives are thus studied without consent.
128
9. The scientific and historic basis of genethics
Finally, there is the dilemma that studies done on any individual affect all of us. This is true because in the genomic era any study performed will be informative for some of what is going on in each of us. This is seen most profoundly in genetic studies, although it is also true in other studies. Each of us carries about three million genetic variations in our three billion DNA nucleotides, and almost all these genetic variants are found in other people. By studying others, we also learn about ourselves and about others in the wider “human family.” Often, but unpredictably, we will be specially affected by the new information. It is this dilemma that may eventually put an end to the genetic aspect of the discussion of group consent. The BRCA1/2 mutations leading to breast and ovarian cancer are far more common in Ashkenazi Jews but it is also found elsewhere. No one ethnicity or population “owns” BRCA1/2. The weakening of the potential of genetics to stigmatize individual populations is, however, futuristic. We are not there yet, and even when we get there the first principles of human research and the application of the knowledge are to acknowledge autonomy, preserve dignity, and exercise beneficence. By enabling people to give informed consent, they decide issues for themselves even if in our opinion their decision is not the best one. By respecting privacy and autonomy we forgo some uses of genetic information that would represent efficiencies. When it is possible to obtain consent of the affected group, this would seem to enhance that. In a few instances that has happened, but multiple issues are raised by a decision to seek group consent. What should group consent consist of? Should it require most of the group? Who should be empowered to give group consent? How does one balance an individual’s autonomy of choice with the need to protect groups? In this regard I believe that Native American tribes offer sound alternatives. To represent the “will of the people,” the better organized tribes have tribal health committees, human research committees (IRBs), and lawyers. It is an imperfect system but one that can improve itself. In my own studies on Native Americans, we augmented this with the use of community focus groups in combination with tribal health committees and the rest, working with tribes on a community basis for up to a year to study and improve the content and methods of studies.
Beyond a pretense of autonomy In human research we treat consenting people and groups “as if” they are autonomous and deserving of respect. Is this a convenient pretense, a temporary expedient until conditions change or a way of buying such treatment for ourselves? Alternatively, are people free, requiring us to treat them as if they are? There are many countries in which freedom is not acknowledged. Punishment for prohibited behavior is not acknowledgment
Beyond a pretense of autonomy
129
of freedom of behavior; it is behavior modification, social statement, and revenge. The reality of free will and individual autonomy will be the major focus of the latter half of this book. Was it frustrating when we had to start over with a new tribal administration or if council was not in session because bearskins had been taken and a ceremony was to take place? Was it frustrating in the first place to have had to vet the study through a human research committee? Of course, but as they say, the bears come with the territory, and we are better off having survived the process. However, what if we felt differently? Is there any fundamental reason to accept the autonomy of others, other than the Kantian imperative to do as one would want done generally, or the situational ethical imperative—a modified utilitarianism, that Derek Parfit endeavored to expand into a general imperative, or is it a matter of hearing and obeying “Do unto others as you would have them do unto you” (Parfit, 1986)? Each of these is an essentially faith based, and if we retrace the circuit of Parfit’s argument to its convergence with situational ethics, each is also in the end an inconstant and ad hoc basis on which to construct a moral system for how people should behave toward one another. By and large we are doing the right thing, but we need to do it for the right reasons. We should treat people as free and autonomous, not because we are following diktat or are persuaded to so by the Nuremberg Code, Declaration of Helsinki, Belmont Report, Bible, or the type of practical ethics argued by Parfit, but because we are persuaded to do so because of what people are. If we do not believe that persons are autonomous, our commitment to their autonomy is as unreliable as it is expedient. The moment that the wrong dictator, war, social crisis, or “important” scientific question arrives, and it becomes inconvenient, our “respect” for the autonomy of persons or that group of persons will inevitably erode, because in the first place our moral framework was constructed on a foundation of words. Also, it is illogical to believe that thinking persons who are treating others “as if” they are free will not constantly be behaving “as if” rather than “as is.” That is exactly what I would do. I would behave “as if” you are free, but as will be seen, we all indeed are. Millions of words have been written and more will be about individual autonomy and free will and the foundations of moral behavior, but this book’s approach to the problem will take a science-based path by establishing the neurogenetic roots of individuals’ ability to make free choices. Will that be the end of the discussion? Hardly, and hopefully it will excite at least these following attacks: (1) it has all been said before (and it probably has in one way or another), (2) he is ignorant of moral philosophy and that which he is criticizing (not quite true), (3) many questions in moral ethics are not automatically resolved by this observation (true), and (4) people are not free. All behavior is understood as the product of causally constrained automata (and therein lies the heart of the debate). But first we must build a brain from the instructions in a molecule.
130
9. The scientific and historic basis of genethics
References Comings, D.E., 1994. Role of genetic factors in human sexual behavior based on studies of Tourette syndrome and ADHD probands and their relatives. Am. J. Med. Genet. 54, 227–241. Comings, D.E., Comings, B.G., 1985. Tourette syndrome: clinical and psychological aspects of 250 cases. Am. J. Hum. Genet. 37 (3), 435–450. Erratum in: 1985 Jul; 37(4): 718 3859204. PMCID: PMC1684588. Emanuel, E., Grady, C., Crouch, R., Lie, R., 2011. The Oxford Textbook of Clinical Research Ethics. Goldman, D., 2021. Immortal: Our Cells, DNA, and Bodies. Elsevier. Lifton, R.J., 2000. The Nazi Doctors: Medical Killing and the Psychology of Genocide. Parfit, D., 1986. Reasons and Persons. Oxford University Press. Stohlberg, S., November 28, 1999. The Biotech Death of Jesse Gelsinger. New York Times. The Belmont Report, 1978. Ethical Principles and Guidelines for the Protection of Human Subjects of Research, by the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. Witkowski, J., Inglis, J.R. (Eds.), 2008. Davenport’s Dream: 21st Century Reflections on Heredity and Eugenics. Cold Spring Harbor Laboratory Press.
10 The world is double helical: DNA, RNA, and proteins, in a few easy pieces Stars in my pocket like grains of sand. Samuel R. Delany In the molecules composing our cells is ample genetic variation to distinguish any of us from others, except if we have an identical twin. Most molecules in this universe consist of only a few atoms. For example, a water molecule (H2O) consists of only three: two hydrogen atoms and one oxygen atom. These small molecules are inadequate for the task of carrying the information that makes us individual. However, surrounded by the water of human cells are some huge biomolecules: DNA, RNA, and proteins. Each is a large polymer consisting of hundreds, or even billions, of smaller molecules that have been enzymatically linked together into a chain. In this regard the biopolymers are more akin to synthetic nylon or rayon polymers or natural silk molecules in the fabric of clothing or in the strands of rope than they are to small molecules such as water. However, DNA, RNA, and protein biopolymers also differ from nylon and rayon polymers in two crucial ways. Each is a heteropolymer in which the subunits are chained together in defined sequences. With four DNA and four RNA bases (and many epigenetic chemical modifications leading to other bases that are functionally different) and 20 protein-building amino acids (plus two more in some species, and many posttranslational amino acid modifications such as phosphorylation and glycosylation) even a short sequence of DNA, RNA, or protein can be highly varied, and for example a sequence of 10 DNA or RNA bases generates more than a million different combinations (410). Typically, the protein-coding regions of genes are several thousand nucleotides in length, and the proteins themselves are usually hundreds of amino acids in length. DNA, RNA, and protein can thus take many forms, enabling them to carry information and perform diverse functions, even serving as e nzymes
Our Genes, Our Choices https://doi.org/10.1016/B978-0-443-22161-3.00007-7
131
Copyright © 2024 David Goldman. Published by Elsevier Inc. All rights reserved.
132
10. The world is double helical: DNA, RNA, and proteins
to themselves catalyze chemical reactions. They represent a highly ordered type of information storage, and that information can be decoded by proteins that form molecular machines in our cells, or by a someone armed with a DNA sequencer and knowledge of the code. Also, although much of the emphasis of genetics is on the decoding of the linear sequence of these biopolymers, which is their primary structure, there is also much more information inherent in their higher order structures and modifications. This is due to the ways, often variable from person to person, cell to cell, or even molecule to molecule, in which they can be altered. They can be secondarily modified, for example by the attachment of methyl groups to DNA or sugar groups to proteins. They are folded in complex ways, which is their secondary structure. There are also other higher order interactions between molecules leading to very large, complex assemblies of different biomolecules. There are tertiary and quaternary structures, and beyond. To understand how this information in DNA and proteins is read and interpreted we need to “follow the molecules” and can do this in a way that is in part historical. This is because scientists first studied the external manifestations of genes, then the proteins that determined those effects, and finally, after Oswald Avery, in 1944, discovered that DNA was the carrier of genetic information, the genes that encode the proteins. With the sequencing of the human genome, the challenge in human genetics has paradoxically reversed from the level of DNA sequence back to the more complex levels of gene expression and the external manifestations (such as human behavior) determined by gene expression. Later, and to close the loop, we will discuss the interactions of these biomolecules with each other, and with the environment, to encode the complex phenotypes that are externally visible and that were the original starting point for human genetics.
DNA recipes DNA is a complex molecule, but in the end a chemical. This means that if a set of procedures is carefully applied ordinarily there will be a dependable outcome. In other words, and contrary to what some frustrated scientists in my lab claim, genetics is often akin to following a cookbook. For example, here is one a molecular geneticist’s recipe for good Italian food. Step 1: Take one pack frozen lasagna. Step 2: Open it. Perhaps the simplicity of molecular biology is misleadingly overdramatized by this example, but there is no voodoo involved, if simply you do what you do so well. As shown in Figs. 10.1 and 10.2, a human genome consists of some three billion base or nucleotide pairs. The DNA chains are comprised largely of four nucleotides: the purines adenosine (A) and guanine (G)
133
DNA recipes
Transcripon
Translaon
Base Pairs
DNA
RNA Transcripts
Proteins
FIG. 10.1 Introducing the genome. Double-helical DNA has four principal DNA bases cytosine, adenine, guanine, and thymine, paired as A:T and G:C via hydrogen bonding. Genes—at least 25,000—are transcribed to distinct single-stranded RNAs. Many RNAs are structural and regulatory in function, but an even larger number are translated—at ribosomes—into proteins. Created with Biorender.com.
Within-person variaon at 1/1250 DNA bases Human variaon 23 chromosome pairs 3 billion DNA base pairs FIG. 10.2 Introducing genomic variation. Many of the 3 billion DNA base pairs, organized into 23 chromosome pairs, are genetically variable. The variations are of many types, ranging from single base substitutions to variable numbers of tandem repeated sequences to very large duplications, insertions, and deletions, but most involve single-nucleotide substitutions of one DNA base for another. In most individuals, one in 1250 base pairs is heterozygous, with a different DNA base inherited on the chromosome derived from each parent.
and the smaller pyrimidines thymine (T) and cytosine (C), and these nucleotides are linked together into long polymeric strands. More recently it has also become recognized that many of the cytosines have become methylated and hydroxy methylated, and subject to other epigenetic secondary modifications, leading to important differences in gene regulation, but although it can be stated that there are at least six relatively common
134
10. The world is double helical: DNA, RNA, and proteins
DNA bases, for purposes at this level of discussion, we can say that there are four DNA bases. Human DNA does not exist as a single polynucleotide strand but is organized into 23 chromosome pairs, one member of each pair having been inherited from the mother and the other from the father. Each chromosome is a long DNA duplex strand wound around protein cores composed of histones and supercoiled, compacting it and forming a complex structure critical to gene expression and to DNA stability, replication, and repair. During the formation of egg and sperm, new combinations of genetic variations that are already present in the DNA are generated by crossing over between maternally and paternally derived chromosomes in chromosome pairs, most human chromosomes being found in pairs. Also, new mutations constantly occur, although most unplanned DNA changes are corrected by repair processes. In eggs and sperm, one copy of each chromosome is present, and in other cells there are usually two. Thus, in humans, the haploid chromosome number (found in eggs and sperm) is 23 and the diploid number (found in the rest of our cells) is 46. As also in the figure, DNA is a double-stranded, twisted ladder with two types of nucleotide pairings between opposite strands: A–T and G–C (Fig. 10.1). The moderately sticky hydrogen bonding between the nucleotides on paired DNA strands allows the DNA to be melted apart into single strands and then reannealed into double-stranded DNA. The melting can be accomplished by raising the temperature, by altering concentrations of ions that are in the water surrounding the DNA or by proteins that can facilitate, as well as inhibit, the opening of the DNA. The ability of the d ouble-stranded helix to split into single strands enables the DNA sequence to be faithfully replicated from a single-stranded DNA template into either RNA, a messenger molecule within the cell, or a new DNA strand, so that the cell can divide and transmit a DNA duplex to each daughter. Note that each daughter cell receives a DNA duplex that is half old and half new. This important consequence of the double- stranded nature of DNA was immediately pointed out by James Watson and Francis Crick when they discovered the structure of DNA, an event that happened so very recently, in 1953 (Watson and Crick, 1953). Their insight was to realize that the double-helical nature of DNA would allow it to unzip and replicate copies of itself. As an aside, be careful around such genius. I have heard Watson exclaim (while someone else was speaking), “I don’t believe this!” in a way characteristic of him (Watson, 2001). It would also turn out that the double-stranded structure of DNA structure that Watson and Crick discovered was critical to the ability to probe specific regions of DNA for genetic variation and to artificially amplify and thereby purify selected DNA regions, thus vastly expanding access to human genetic variation. The study of DNA leads to functional genomics. Our DNA contains most of the instructions for building and operating a human and is contained in the cell nucleus. The nucleus also contains much of the
Polymorphism
135
molecular apparatus for the maintenance, repair, and replication of DNA and its transcription (copying) into RNA, a single-stranded polynucleotide consisting of ribonucleotides (R) instead of deoxyribonucleotides (D) because a different sugar molecule is attached to the nucleotide and substituting uracil for thymine. An increasingly larger class of these RNA molecules has been recognized as independently functional, regulating DNA and each other and even serving structural and enzymatic (catalytic) functions. However, much of the RNA is messenger RNA, which will be in turn processed, translocated from nucleus to the cytoplasm of the cell. Once in the cytoplasm, the RNA can be translated into protein by a large, multisubunit structure known as the ribosome (Fig. 10.1). Every cell spends a substantial part of its energy manufacturing new ribosomes, and every genome contains multiple copies of ribosomal genes, including genes that encode structural ribosomal RNAs. For cells and tissues of the body, the proteins made in the ribosome and further processed in other cellular structures such as the Golgi apparatus serve as building blocks, signaling molecules and machinery. As mentioned, the same DNA is found in all cells of the body, except most human erythrocytes and platelets, two constituents of blood. Otherwise, hair follicles, white blood cells from blood, brain, fat, and muscle all contain the same DNA. The RNA and protein compositions of cells are, however, very different. Cells have different mixtures and quantities of RNA and protein types, and it is the differences in these molecules that build and operate a cell that make a muscle cell different from a neuron.
Polymorphism Both DNA and the RNA and protein molecules it encodes are subject to genetic variation, and any of these genetic variations can be measured by sequencing the DNA or by genotyping (a limited form of sequencing) specific sites. The human genome contains at least 22 million common genetic variations (more are being discovered), which are called polymorphisms (Fig. 10.2). A polymorphism is a genetic variation with a frequency of greater than 1%. However, there are also very large numbers of rarer genetic variants. For 22 of our 23 chromosomes (the autosomes) almost all humans have two copies, thus giving us two chances to have the polymorphic variant (allele). The exceptions are the X chromosome, of which most males only have one copy (inherited from the mother), and the Y chromosome, which males inherit from their father. Also, many of us have large extra and missing pieces of some chromosomes, an example being Down’s syndrome, which is caused by inheritance of an extra copy of chromosome 21 and sometimes by an extra piece of chromosome 21. Speaking of extra or missing pieces of chromosomes, it has recently been
136
10. The world is double helical: DNA, RNA, and proteins
recognized that all humans also have many large DNA deletions and insertions, called copy-number variants (CNVs). Most of these are too small to be seen using a light microscope but can involve the deletion or insertion of thousands of DNA nucleotides. Such CNVs can delete copies of genes that are encoded by that DNA or lead to gene duplications as in the chromosomal region implicated in the type of Down’s syndrome associated with an extra, translocated, piece of chromosome 21. There is no canonical genome, the human genomes first sequenced and sequenced later being a few of the distinct genomes out of billions (Lander et al., 2001; Venter et al., 2001). Except for identical twins, triplets, and such, no two people inherit the same constellation of some 22 million common genetic variants, and many rarer ones. At the chromosomal level, 50% of the population—females—have a different constellation of chromosomes than the other 50%—males, and chromosome anomalies such as extra X and Y chromosomes and missing X chromosomes (the latter which is Turner syndrome) are common in live births and even more common in utero. Naturally, such large chromosomal variations can alter the nature of the brain, and later in this book sex-influenced behavioral variation will be discussed. Smaller genetic variations—any of which are less capable of altering function—are legion. For the autosomal, and in females—X chromosomal DNA sequences present in two copies, and except in instances of inbreeding between close biological relatives, there are many alleles that distinguish the copy of the DNA inherited from the father from that inherited from the mother. On average and across the three billion DNA nucleotides that constitute the genome there is an allelic difference between the maternal and paternal DNA copy once every 1250 DNA bases, the exception being people born to parents who were themselves related and thereby sharing alleles identical by descent, a phenomenon known more nimbly as consanguinity or less as inbreeding. Each position in the genome that may be genetically variable is said to be a locus. If two identical copies are present the locus is homozygous but if the maternally derived copy differs from the copy transmitted from father the locus is heterozygous, meaning that the sperm and egg (the zygotes) that formed the human were different (heterogeneous) at that locus. Thus there is not only genetic diversity between genomes, but also diversity within (Fig. 10.2). Variants of the single-nucleotide polymorphism (SNP) and less common single-nucleotide variant (SNV) type are locally linked in constellations. Fig. 10.3 illustrates how different copies of the same chromosome can have different SNP alleles at the same site. Furthermore, it shows a new level of DNA characterization that is increasingly recognized as powerful. This is the combination of several of these SNP alleles in a local region into a haplotype, which as the figure suggests is like a DNA barcode that can be used to identify the contents of a DNA sequence in the same way that barcodes in supermarkets can be scanned to reveal what is in a package. For most small regions (packages) of the genome, say 5000 DNA bases,
Polymorphism
137
A T T A G A C C A A T G G A T G C A G C T A T G G C C T T A A A T T G G A C C A A C G G A T G C A G T T A T G G C C T C A A A T T A G A C C A A T G G A T G C A G T T A T G G C C T C A A A T T G G A C C A A T G G A T G C A G C T A T G G C C T T A A
A T C T G C T C A T T C G T C T
FIG. 10.3 DNA barcodes: DNA variants in a local region form combinations (haplotypes), condensed versions of the four-locus haplotypes in this region are depicted at bottom.
there are relatively few (2–5) common haplotypes, and relatively few SNPs, called tag SNPs, need to be genotyped to capture these haplotypes, even though the haplotype may consist of many markers. The barcode nature of haplotypes is shown in Fig. 10.3. The reason for the restricted number of haplotypes in a region, and by inference the correlation between genotypes of SNPs in a local region, comes down to the rate that new haplotype combinations are generated versus their elimination by genetic drift. Via neutral drift the genome has been shaped by the locality of humans in earlier times, and such that Africans, the population with the largest effective size (Ne) had Ne of only 11,000. In a local DNA region one might easily be concerned by the number of haplotypes formed from 20 SNPs, that number yielding about one million potential haplotypes (220 permutations). However, a population with an effective size of 11,000 could at most provide space for 22,000 haplotypes—two per person. Worse, with each generation many of the haplotypes existing in only a few individuals are by chance not transmitted. After many generations, genetic drift will eliminate all but one haplotype in a local DNA region, except that two other forces—point mutations and occasional recombinant reshufflings—work to occasionally generate new haplotypes, and in addition, a few haplotypes can be maintained in a population by natural selection, and thus the COMT Warrior/Worrier Val158Met alleles are on ancient haplotypes of opposite configuration. The balance of forces between generation of new haplotypes and their elimination by drift and pruning by natural selection has led to the genome having a haplotype “blocky” structure. Thereby, local regions of 5–40 kilobases tend to have only a few common haplotypes. Larger regions naturally have more haplotype diversity because of the likely recombinant DNA shuffling during the meiotic generation of the gametes, and of course because a bigger region represents a larger mutational target.
138
10. The world is double helical: DNA, RNA, and proteins
These events have likelihoods, the best estimate of likelihood of a DNA nucleotide mutation being 2.5 × 10(−8) or 175 mutations per diploid genome per generation, and the likelihood of a recombination event being roughly 1% per one million DNA bases or about 66 per generation (more in females, the homogametic sex than in males, the heterogametic sex, and as varies substantially across the genome). It should also be noted that new haplotypes can be introduced into a population by admixture with other populations, and in this way, some Neanderthal and Denisovan haplotypes are still within many of us, those haplotypes serving as barcodes for ancient functional genetic variants in the same local regions of the genome. In the end the haplotypes can be tagged by a few SNPs that serve as proxies for other SNPs in the local DNA region, and which is why GWAS works—at least for detection of effects of common variants—with “only” a million SNPs genotyped, rather than 22 million. The relationship between nucleotide heterozygosity and effective population size under genetic drift is given by Motoo Kimura’s equation H = 4Neμ|(1+4Neμ) where Ne is the effective population size, μ is the nucleotide mutation rate, and H= nucleotide heterozygosity, and “4” derives from the fact that each breeding pair has four copies of an autosomal locus, and for the corresponding X chromosome equation “3” is therefore substituted for “4.” As might be expected, the type and level of variation varies widely across species according to the size of their populations, mating patterns, and—to a minor extent—mutation rates. Humans could easily be less neurogenetically individual, or more. Animals that have been inbred by the deliberate mating of closely related individuals for many generations (for example, inbred strains of mice produced by 20 generations of brother–sister mating) have lost almost all their genetic diversity. At each locus these inbred animals are homozygous, and furthermore all “individuals” from the same inbred strain are genetically identical, like twins. Also, mammalian species on a natural basis differ widely in their overall levels of genetic diversity. Species with very large breeding populations tend to be far more variable (O’Brien, 2003). For example, the Red-backed Salamander (Plethodon cinereus) that used to be found in prodigious numbers in the Maryland woods at the time I wrote the first edition of this book. Way back then—10 years ago—the Red-backed Salamander was both plentiful and genetically highly variable (as any survivors probably still are), and furthermore they were accompanied by other salamander species, now also hard to find, if not gone. Species with large population sizes—think periodic cicadas that emerge every 17 years in epic swarms—tend to be more genetically diverse, whereas species with smaller populations or that have suffered population bottlenecks have less diversity. In general, genetic variation is good: it makes species more adaptable, and in that regard, humans are much
Protein polymorphism
139
better off than certain species, such as the endangered cheetah and Florida panther whose dangerously low genetic diversity was discovered by Steve O’Brien (O’Brien, 2003). I helped with a paternity problem Steve encountered. It involved giant pandas (Ailuropoda melanoleuca) at the Washington National Zoo. Chia-Chia was the mother. However, was Ling-Ling, the London male whose sperm was used to artificially inseminate her, the father? Or was it Hsing-Hsing, Washington’s male panda whose ability to successfully mate had been impugned? By studying the pandas’ proteins, we found several that were useful as genetic markers and heterozygous in the baby panda; this was not easy because the genetic diversity of the pandas’ proteins was indeed greatly reduced. Two of these markers could only have been transmitted from Hsing-Hsing, the Washington male. One panda’s reputation was restored (O’Brien et al., 1984). Unlike humans, pandas, cheetahs, and Florida panthers have undergone severe population bottlenecks such that these species have lost almost all their diversity and are particularly vulnerable to infectious diseases that at any time could kill the last of their kind. However, because humans haven’t always been so numerous, we are much less diverse than some of our ape relatives. If this is not quite clear, then as Carl Merril, my mentor, would say, “Not to worry.” This was only the first of three essentials I learned under his benevolent tutelage, the others being “Sounds good to me” and “What do I know?” The take-home message or moral of the story of human sequence variation is that our genomes are a trove of sequence variation, and all the cells in our body, save some in the ovaries and testes, have a nearly identical pattern of variation that can distinguish us from all other individuals, unless we have an identical twin.
Protein polymorphism As just alluded to in the story of paternity in the giant panda, the first molecule whose genetic variation was studied by geneticists was not DNA but protein. Proteins are encoded by the DNA, but a particular protein is found only in some cells and tissues of the body, whereas the DNA is distributed nearly universally. Proteins are encoded by regions of DNA called genes. In some cells, these regions are active owing to differences in regulatory proteins bound to DNA, differences in chromatin (DNA– protein) structure, and differences in DNA modifications including DNA methylation. Active gene DNA in the nucleus is copied or transcribed into RNA, the single-stranded messenger molecule. The RNA is processed and then translated into a protein sequence in the cytoplasm. Protein synthesis takes place at the ribosome, where the message on the RNA is decoded by small, intricately folded molecules called transfer RNAs, each capable of recognizing an amino acid, and thus bridging between the worlds of RNA
140
10. The world is double helical: DNA, RNA, and proteins
and protein. The protein is synthesized one amino acid at a time, building a large heteropolymeric (multiple different subunit) protein molecule whose sequence faithfully reflects that of the original DNA molecule. The whole process can be likened to the conversion of sheet music to the music we hear from a player piano. The sheet music (DNA) is transcribed to player piano tape (RNA) and the tape is translated by the player piano into individual notes (amino acids) linked together as music (protein). Marshall Nirenberg, a lab chief at the National Institutes of Health who encouraged me to write, won the Nobel Prize for being one of several scientists to first decode one of these protein notes, namely the DNA sequence AAA. That sequence, when transcribed to UUU in RNA, is recognized as the triplet nucleotide code for the amino acid phenylalanine. Recently it has become increasingly recognized that both DNA and RNA can be directly modified, and for example by the methylation and hydroxymethylation of cytosines that are part of CpG dinucleotide combinations. This epigenetic change, unlike genetic changes (e.g., mutations), is readily reversible but enables the genome to vastly expand its potential for regulatory control, and even to modify RNAs to alter their stability, structure, or the amino acids they encode. If an amino acid in a protein or peptide (a short protein) is changed because of a nucleotide change, there is often a detectable alteration in the protein’s structure. Variants that change structure often alter function. For example, an acidically charged amino acid can substitute for a neutrally or basically charged amino acid, altering the mobility of the protein in an electric field, and enabling separation of the two genetic forms of the protein by electrophoresis. Protein structure can also be altered by posttranslational modification. For example, the ABO blood type is determined by whether sugar molecules are added to the protein at certain positions. This occurs after the protein has been synthesized, or posttranslationally. Protein polymorphisms are genotyped in several ways. Protein alleles (variants) can be detected by electrophoresis, which is separation by charge or size in an electrical field. However, several protein polymorphisms are genotyped by highly specific and sensitive immunological methods. The immunological methods depend on the use of an antibody molecule that recognizes the small difference in a protein (the antigen). The antibody can be generated by immunizing another individual or animal, by injecting it with the foreign protein. The immune serum or purified antibody can then be used for the serological test. In 1900 the Austrian immunologist, pathologist, and Nobel laureate Karl Landsteiner discovered the first of the serological polymorphisms, the ABO blood polymorphism, and 25 years later Felix Bernstein worked out the Mendelian inheritance. Many people know their ABO genotype. Their blood type might, for example, be printed on their military ID or driver’s
DNA polymorphism
141
license. Red blood cells carry the H protein, which if not glycosylated produces blood type O and otherwise can exist in the A or B form: three alleles (variants) at one protein locus. With three alleles, people can have up to four different ABO serological genotypes: O, A, B (all of which are homozygous genotypes) and AB (a heterozygous genotype). The OA and OB genotypes are also serologically genotyped as A and B. The ABO polymorphism is a functionally very important one because common bacteria in our gut (Escherichia coli) make the A and B antigens. For this reason, an individual who does not carry these antigens themselves will already have become naturally immunized against the antigens. If a type O individual is transfused with type A, type B, or type AB blood, they will immediately reject it. In contrast, type O (universal donor) blood can be transfused into individuals with the other genotypes, and a type AB person is a universal recipient. To perform the test, a small amount of blood containing red cells is needed. Another important blood group antigen is the Rh antigen. People who have it are Rh positive. Many Rh-negative people have become immunized against the Rh antigen and will reject Rh-positive blood.
DNA polymorphism There are several types of DNA sequence variation, but two are most frequently important. The single-nucleotide polymorphism (SNP) is a single substitution in the DNA code of one nucleotide for another, for example the substitution of an A for a G at a one position in the DNA sequence. Almost all SNPs are biallelic, meaning that there are two possible allelic forms (say allele 1 and allele 2) and three possible genotypes: 1–1, 1–2, and 2–2. For this reason, they are not as individually informative in many genetic analyses as the short tandem repeat type of DNA diversity. However, SNPs are perhaps the most important type of variation for two reasons. First, they are by far the most abundant, constituting approximately 99% of sequence variation. As mentioned, for the three billion nucleotides in a woman’s genome, on average about one in 1250 is heterozygous, a different form having been inherited from her mother than from her father, so people contain about three million SNPs that are variable within their own genomes, that is from one chromosome to the second member of the chromosome pair. Second, SNPs are a type of variation that is very efficiently assayed or genotyped. Commercially available genotyping arrays already enable up to five million genotypes to be simultaneously determined. Array-based genotyping of this type has only been available since the beginning of this century, but more than 99.9% of genotypes that have ever been generated have been performed in this way. The cost is very low, less than a penny a genotype. Array-based genotyping of hundreds of thousands of markers is also available direct to consumer via
142
10. The world is double helical: DNA, RNA, and proteins
commercial services that process the DNA from an easily collected saliva sample and offer rapid turnaround, and as has led to revolutions in genealogy, genetic risk profiling, and forensic science. One may notice that genotyping is itself a method of DNA sequencing, where the sequencing effort is focused on places in the genome that are known to be genetically variable. Increasingly, the whole DNA sequence of a person is readable, especially by massively parallel DNA sequencing. When the human genome was first sequenced, a few hundred DNA fragments were read at a time. Nowadays, by sequencing about a billion small DNA fragments simultaneously, we can rapidly sequence an entire human genome, which is some three billion DNA bases in size, and with a sufficient level of redundancy to ensure accuracy. Once the DNA bases that comprise the DNA code have been accurately transcribed into letters, A, G, C, and T, they can be read like any text and in paper or electronic form. Whole genome sequencing is now also readily available to the affluent or for clinical research and searches for the causation of undiagnosed rare diseases (Turro et al., 2020). For many purposes a whole genome sequence is unnecessary. The exome—the 55 million base portion of the genome containing 25,000 or so genes—may be selectively sequenced and has led to thousands of genetically based diagnoses. For forensic identification, and for genealogy or ancestry investigations, the millions of genetic variants that are detected by whole genome sequencing represent a vast overkill. Polygenic scores for diverse traits are computed based on genotyping of a million SNPs or so, rather than requiring a whole genome sequence. Thus predictive genomics is well within the grasp of anyone who wants to access it. Furthermore, if a person’s genome is sequenced once, for example at birth, the need for later genotyping of that individual would be largely eliminated throughout the rest of their lifetime. Furthermore, genetic information obtained at birth could be used prospectively, and years before the infant tested had any voice in the choices made based on it. Testing of epigenetic markers will likely remain important throughout life, and there are several powerful and accessible ways to do it. Upwards of 850,000 CpG dinucleotides, out of some 28 million genome-wide, can be evaluated for cytosine methylation via arrays. Those methylation values can then be scores reflecting exposure to stress, alcohol, nicotine, and to measure accelerated cellular aging, with an accuracy of about + − 5 yrs (Bocklandt et al., 2011). Recently, nanopore sequencing has been used to simultaneously sequence DNA and detect its epigenetic nucleotide modifications (Wang et al., 2021). The technology is packaged in a case about the size of a cellphone, or Star Trek “Tricorder,” and about as portable. Each type of genetic marker, from protein to DNA, and each method for “genotyping” and sequencing, has its pitfalls and advantages. In research, there is constant innovation and diversification of methods, but
Measured ancestry
143
when genotypes are used in medicine and in the courtroom the methods must be stabilized and highly refined for quality, to minimize the possibility of error.
Measured ancestry More than nine-tenths of genetic variation is interindividual and only one-tenth is predictable based on population of origin. This is a compelling argument against racism: on an objective basis one knows very little about a person merely from their race, skin color, hair, or the shape of their eyes. However, because of the number of polymorphisms, the population differences that do exist translate into a very powerful tool for measuring the ancestral origins of an individual’s genome, and even the ancestry of chromosomal regions. In this regard ancestry is to race (or ethnicity) as sex is to gender. Both ancestry and ethnicity are biologically and medically meaningful, but although intertwined offer different information, much as the sex of an individual does not tell us their gender. To measure ancestry, geneticists use ancestry informative markers (AIMs). An AIM is a genetic locus specifically selected for its ability to distinguish human populations, a quantity known in quantitative genetics as FST. Even with a small number of AIMs, only a fraction of what is available—measured ancestry tracks with geography and population of origin in a remarkably powerful way (Fig. 10.4). Reading from left to right are African populations (blue). Next (red) are North Africans and Middle Eastern peoples. North Africans have a Middle Eastern ancestry component (red) but also some African ancestry (blue). Next are a series of Asian populations, with different components of Asian-specific factors (Hodgkinson et al., 2008). For example, both Chinese and Japanese are predominantly one of the Asian factors (dark blue). Continuing rightward (tan) are two populations from Oceania: Melanesians and Micronesians. European populations (red) and Native American populations (yellow) are at the right of the figure. This represents a seven-factor view of worldwide ancestry. Deeper views with more factors further subdivide some of the populations, for example Finns are closer to other Europeans than to other populations but nevertheless are distinct from other European populations. The ability to measure ancestry is a breakthrough in studies aimed at understanding the genetic origins of disease because many disease genes are common only in one population. Ancestry measurements, misinterpreted, thus can become a modern justification for alienation of individuals of different genetic origin, even being used to justify racism or facilitate genocide, as cruder identifiers such as skin, hair, and eye color, and other physiognomic have historically been misused, and are misused to this day. Benignly, one can cultivate
144
10. The world is double helical: DNA, RNA, and proteins
cultural choices in line with one’s genealogy or read much into the fact that one is a 30th generation descendent of some historical personage, although in biological terms that relationship amounts to nothing, or little. Less benignly, one’s DNA ancestry becomes another race-related identifier that a person carries around with them, regardless of the choices they would prefer to make. Genocide can be carried out based on an individual genetic variation, and for example some authoritarian government might decide to kill all the tall people, or all the bald people, or all the myopic people, and societies have done such things. More often, genocide is directed against the identifiable “other.” In modern times, the solution to racism is not elimination of race words (black, white, Asian, Hispanic, etc.). These are socially based constructs but for all of us who cannot escape them they have constant consequences in people’s lives. If we ever did get rid of them, measured ancestry is right there waiting in the wings, better capturing genetic effects and to a much worse extent capturing cultural determinants that partly track with ancestry (Fig. 10.4). Although people from different populations are readily distinguished using ancestry informative markers, most human genetic variation, more than 90%, exists between individuals rather than between populations. Also, the population genetic variation does not fit even nearly to racial classification schemes. In this sense, critics of the use of “race” are correct. Race is to ancestry as gender is to sex. They measure independent information and they correlate, but far from perfectly. Other descriptors are more precise, even in individuals who are admixed in ancestry or who were raised in more than one ethnic–cultural background. Furthermore, even though the population-specific component of human diversity is relatively small, this component is functionally important. Certain functional polymorphisms are strongly tied to populations, or even found in these populations exclusively. For example, the hemoglobin S allele, the cause
FIG. 10.4 Ancestry informative markers (in this case only 186) enable computation of ancestry factors (in this case, seven). Demonstrating that ancestry tends to track with geographical origin.Data from Hodgkinson, C.A., Yuan, Q., Xu, K., et al., 2008. Addictions biology: haplotype-based analysis for 130 candidate genes on a single array. Alcohol Alcohol. 43, 505–515. https://doi.org/10.1093/alcalc/agn032.
Measured ancestry
145
of sickle cell anemia, is common in individuals of West African origin and also elsewhere in places where malaria is endemic. The hemoglobin S allele is a single-nucleotide variant; substitution of one DNA nucleotide being sufficient to encode valine rather than isoleucine at the sixth amino acid of the hemoglobin beta protein. That one seemingly innocent change of one neutrally charged amino acid for another—which in 1949 Nobel Laureate Linus Pauling was able to detect by zone electrophoresis of hemoglobin—causes hemoglobin to form aggregates under conditions of low oxygen, leading to catastrophic damage to red blood cells (Pauling et al., 1949). This was classical hypothesis-driven discovery (and thus quite unlike most modern genomics). In a leap of insight Pauling realized that the abnormal sickle shape of red cells in sickle cell anemia was likely caused by a structural abnormality of hemoglobin, by far the most abundant protein of those cells. Later, studies showed a single evolutionary origin for almost all hemoglobin S alleles, and we even know why it spread and became highly abundant in West Africa: hemoglobin S confers resistance to malaria, and malaria is common there. Interestingly, West Africans and other malaria-afflicted populations have additional genetic variants that confer resistance against malaria, and these are found both in beta globin and in other genes such as G6PD. In each case the variants are primarily population restricted. Every population appears to carry variants that under some conditions confer an advantage and under other conditions lead to disease. Cystic fibrosis, a disease causing sinus infections and pneumonia, diarrhea, and infertility, is abundant in northern Europeans and rarer in other populations. Owing to the discovery of the cause of cystic fibrosis by Lap-Chee Tsui, Collins, and other geneticists (Kerem et al., 1989; Riordan et al., 1989), we know today that one in 25 European Americans carries a single copy disease allele of the most common mutation Delta F508, a three-nucleotide deletion that in turn deletes the amino acid phenylalanine at position 508 of the cystic fibrosis transmembrane conductance regulator (CFTR) protein. Also, it has been learned that CFTR protein controls membrane chloride transport in many cells of the body, accounting for the diverse manifestations of the disease. The carriers of a single copy of Delta F508 or one of the many rarer alleles that also cause cystic fibrosis are unaffected because the disease is recessive. However, some 30,000 Americans have inherited two copies, and therefore suffer from cystic fibrosis. Approximately one in 65 Hispanics and one in 90 Asians carry a cystic fibrosis mutation, but this translates to an even greater reduction in the frequency of the disease in those populations because of the recessive nature of cystic fibrosis. Also, different mutations are more common elsewhere in the world. Tracing the population associations of cystic fibrosis mutations may eventually clarify why the Delta F508 variant is so common in northern Europeans. The most likely explanation is that heterozygous carriers again have some
146
10. The world is double helical: DNA, RNA, and proteins
advantage against infectious disease, possibly tuberculosis, cholera, or typhoid fever, or even against lactose intolerance. Parsing individual genetic ancestry increases the power of genetic studies via better targeting of studies, increasing power to detect effects of ancestry-restricted variants, and preventing results being confounded by ancestry (for example, if patients were compared to controls of a different ancestry false conclusions might easily be reached). Furthermore, geneticists interested in the origins of human populations can reconstruct those relationships based on the measured ancestry of living descendants of founders of those populations. It can therefore be appreciated that the development of the capability to measure ancestry has been accelerated by scientists interested in disease gene mapping and in broad questions about population origins. Through these studies, geneticists have been able to confirm and greatly deepen our appreciation of the phenomenon of genetic individuality, as well as the genetic affinities of people who share common origin.
References Bocklandt, S., Lin, W., Sehl, M.E., Sánchez, F.J., Sinsheimer, J.S., Horvath, S., Vilain, E., 2011. Epigenetic predictor of age. PLoS One 6 (6), e14821. https://doi.org/10.1371/journal. pone.0014821. PMC: 3120753. PMID: 21731603. Hodgkinson, C.A., Yuan, Q., Xu, K., et al., 2008. Addictions biology: haplotype-based analysis for 130 candidate genes on a single array. Alcohol Alcohol. 43, 505–515. https://doi. org/10.1093/alcalc/agn032. Kerem, B., Rommens, J.M., Buchanan, J.A., et al., 1989. Identification of the cystic fibrosis gene: genetic analysis. Science 245, 1073–1080. https://doi.org/10.1126/science.2570460. Lander, E.S., Linton, L.M., Birren, B., et al., 2001. International human genome sequencing consortium. Initial sequencing and analysis of the human genome. Nature 409, 860–921. O’Brien, S., 2003. Tears of the Cheetah. O’Brien, S., Goldman, D., Knight, J., et al., 1984. Giant panda paternity. Science 223, 1127– 1128. https://doi.org/10.1126/science.6701515. Pauling, L., Itano, H.A., Singer, S.J., Wells, I.C., 1949. Sickle cell anemia, a molecular disease. Science 110, 543–548. Riordan, J.R., Rommens, J.M., Kerem, B., et al., 1989. Identification of the cystic fibrosis gene: cloning and characterization of complementary DNA. Science 245, 1066–1073. https:// doi.org/10.1126/science.2475911. Turro, E., Astle, W.J., Megy, K., et al., 2020. Whole-genome sequencing of patients with rare diseases in a national health system. Nature 583, 96–102. https://doi.org/10.1038/ s41586-020-2434-2. Venter, J.C., Venter, J.C., Adams, M.D., et al., 2001. The sequence of the human genome. Science 291, 1304–1351. Wang, Y., Zhao, Y., Bollas, A., et al., 2021. Nanopore sequencing technology, bioinformatics and applications. Nat. Biotechnol. 39, 1348–1365. https://doi.org/10.1038/ s41587-021-01108-x. Watson, J.D., 2001. The Double Helix: A Personal Account of the Discovery of the Structure of DNA. Watson, J.D., Crick, F., 1953. The molecular structure of nucleic acids: a structure for deoxyribose nucleic acid. Nature 171, 737–738.
11 The stochastic brain: From DNA blueprint to behavior Blind man breakin’ out of a trance Puts both his hands in the pockets of chance Hopin’ to find one circumstance Of dignity. Bob Dylan—Dignity Clouds are not spheres, mountains are not cones, coastlines are not circles, and bark is not smooth, nor does lightning travel in a straight line. Benoît B. Mandelbrot—The Fractal Geometry of Nature This chapter describes how the brain, with its astronomically complex interconnections and lifelong plasticity, builds itself using a restricted instruction set of some 25,000 genes, and perhaps an equally large number of regulatory RNA molecules. A crude statement of the problem of complexity is that the human brain contains about 100 billion cells that must be “properly wired” into some 1015 (one million billion) connections. However, that vastly understates the intricacies inherent in the multineural networks necessary for memory, cognition, motor control, cognitive, and sensory functions. For the time being, and with our present state of knowledge, it is most accurate to refrain from even putting a number on that number. Also, an effort such as this one to explain the process by which the brain develops can only sketch broad outlines and principles. Fortunately, other books have more completely developed the story of how the brain encodes emotion and cognitive functions, and for emotion I would point to Joseph LeDoux’s The Emotional Brain (Ledoux, 1998), as a general introduction Michael Gazzaniga’s Nature’s Mind (Gazzaniga, 1994), and for the development of the brain and how the brain, through fundamental processes including neuronal plasticity builds our “self,” LeDoux’s Synaptic Self (Ledoux, 2003). For those making comparisons, my emphasis is on the unfolding of the genetic program.
Our Genes, Our Choices https://doi.org/10.1016/B978-0-443-22161-3.00012-0
147
Copyright © 2024 David Goldman. Published by Elsevier Inc. All rights reserved.
148
11. The stochastic brain: From DNA blueprint to behavior
Self-assembly The brain is not built like a car. Cars do not self-assemble. The components of a car are assembled by the hands of robots and humans, who have programmed the robots. We can learn about complexity by studying mechanical constructions but cannot understand how biological entities develop and function merely by studying cars and similar inanimate constructions. The most amazing—but as we will see, explainable—aspect of the brain is that it is self-assembled. It does not suffice to have the “right stuff”: the stuff of life must be in the right place and at the right time. Unlike components of a car, for example the steering wheel, the brain’s molecules, cells, and networks of cells have autonomous properties that enable them to recognize, respond, and adapt to build, through a hierarchical series of interactions, the much more complex structure of the brain (Kuffler, 2001). These things are alive! Neither the steering wheel of your car nor the granule cell or Purkinje cell of your cerebellum has any understanding of the overall system in which it will play a part. Also, the steering wheel can be put aside until it is ready to be installed, but the Purkinje cell must be maintained and nurtured through interactions with other cells. However, because the Purkinje cell is itself a living entity it is also capable of responding in a myriad way to its environment and neighboring cells with which it synapses and crosstalks metabolically, and if not needed the Purkinje cell’s disappearance (death) can be readily programmed. The steering wheel has only a limited “conversation” with the column to which it is attached and if the car is constructed without a steering wheel the steering column is oblivious. It will just sit there as if it is still doing its job. On the other hand, it is a rare instance in which a car is delivered that has two steering wheels. In development this type of error can easily happen because the functions guiding self-assembly are simpler and more readily confused if, for example, a chemical gradient is temporarily disrupted at the critical moment. The conversations between cerebellar granule cells and Purkinje cells are both nuanced and needful. They guide the development and function of both cells and enable the formation of hierarchically more complex levels which further drive the assembly of the brain toward its ultimate complex structure and capabilities. However, there is no third-party standing by to say, “Ach! A car with two steering wheels—better fix that one.” Our complement of genes succeeds as a self-assembling instruction set only because it does not try to do everything at once. The DNA instruction set encodes molecules, which make cells, and then cells build networks of interacting cells. All along during this process there are feedback loops of a complexity that is probably infinite, such that molecular expression and the differentiation, positioning, and interaction of cells are in constant adjustment and refinement. Even in adult life there is no point at which a brain is static. The brain is constantly adapting at molecular, cellular, and network levels.
Cell assembly
149
As will be seen, the brain’s DNA blueprint succeeds because each genetic element in the instruction set is subject to subtle regulation, including alternative molecular forms. The genes talk to each other and guide each other. The gene products (proteins and RNA molecules) interact in what has come to be known as an “interactome,” the new and often useful fad being to place the suffix “ome” behind any field which attempts to encompass all features of a like kind, and for example, genes, proteins, lipids, connections, or methylations, and thereby forming a new field of scientific inquiry (a new piece of the “inquirome”?). Unlike the tools that humans have traditionally manufactured, brain development involves transformation from one developmental stage toward the next, and it is self-organizing. Like the molecules, the neurons also talk to each other and guide each other, but their interactions are far more complex and nuanced. Scientists have physical models for many molecule–molecule interactions but are far from being able to model cell–cell interactions with anything approaching that accuracy and precision, as will soon be seen when we discuss the neuronal “connectome.” The development and maintenance of neural organization is guided by function in accord with “fire together, link together” and “use it or lose it” principles and this depends on the ability of neurons to interact, and ultimately translate their interactions into molecular changes that may reinforce connections, weaken them, or even signal an individual neuron that it is time for that neuron to die. Such “programmed cell death,” or apoptosis, involves the activation of a particular molecular cascade, with many components. Governing the whole process of brain development and its ongoing plasticity is a mysterious and ultimately fruitful interplay between our innate and individual complement of genetic variation, and stochastic (random) processes that enable infinite divergence in the neural networks of our brains. This stochasticity is the raw stuff that neuronal developmental processes mold into unique circuits. In this way, the story of the development of a brain, or even a local neuronal network, never unfolds in the same way, no matter how many times the story is told (Kuffler, 2001; Ledoux, 2003). At an intrinsic genetic level, most humans, except for identical (clonal) twins, are unique, but during development we all become individual and, in humans, that individuality is over time amplified by the choices we make, which soon have the effect of shaping our brains in ways that we have chosen, whether wisely or not.
Cell assembly Neurons are complex entities—any individual neuron is too complex for us to give a full accounting (Alberts et al., 2002; Byrne and Roberts, 2009; Damasio, 2010; Kandel, 2000). Yet, to understand the self-assembly of a brain we must first understand the self-assembly of neurons. The
150
11. The stochastic brain: From DNA blueprint to behavior
building blocks of cells are proteins, structural and regulatory RNAs (ribonucleic acid), and small molecules, such as the lipids in cell membranes, that proteins build or properly position, compartmentalize, and modify. The “central dogma of molecular biology” (shown later) is DNA → RNA → protein. DNA is transcribed to messenger RNA by an enzyme known as RNA polymerase. RNA is processed and then translated into protein by cell machinery known as the ribosome. The proteins in turn fulfill myriad roles, forming the structure of the cell and scaffolding, barriers, and transport highways for other proteins and small molecules, catalyzing chemical reactions, transporting other molecules, regulating genes and each other, signaling distant cells, and serving as receptors for such signals. This “central dogma” has been an extremely fruitful and powerful explanatory model for the capacity of DNA to direct the molecular construction and function of cells. However, keep in mind that there is always a good chance that someone’s karma will run over your dogma, and over time important exceptions to the central dogma have been discovered. Exceptions to the central dogma include the ability of some RNAs to be reverse transcribed into DNA (RNA → DNA) and the discovery of RNA molecules with intrinsic structural, regulatory, and catalytic (enzymatic)activities. It has recently been appreciated that there are probably at least as many regulatory RNA molecules as protein-coding ones. These new functions for RNA extend the array of potential mechanisms by which DNA can ultimately direct cellular structure and function. In addition, alternative processed forms of RNAs can enable a “single gene” to encode a multitude of protein forms, even before posttranslational modifications of proteins, including additions of phosphate and carbohydrate groups, alter their structure and functional properties. Ultimately, the DNA directly or indirectly encodes 25,000 (the approximate number of genes), or perhaps an order of magnitude more other large molecules—proteins, structural RNAs, and regulatory RNAs. It does so in a fashion that is under strong and specific regulatory control. Thus there is a great range of molecular plasticity for the cell to exploit during development and differentiation, as it faces different functional challenges. The epigenetic plasticity of the genome will be taken up further later, and in the more global (rather than cellular) context of gene-by-environment interaction.
Interactomes The proteins and RNA molecules encoded by genes work together to build cell structures and functional molecular networks. Scientists of several related disciplines—molecular cell biology, biochemistry, neurobiology, endocrinology, and genetics—have now achieved a significant, but
Interactomes
151
partial, understanding of the molecular networks that govern the function of cells. Examples of some of the better understood pathways include regulation of the cell cycle and DNA synthesis, and cell energy generation including oxidative phosphorylation and glycolysis (the process by which sugars are metabolized to regenerate ATP, providing energy for many other functions). Increasingly, it is understood that many of these systems of interacting proteins are better understood as networks, rather than as linear pathways. A major task of genetics and molecular cell biology is to fill in the missing pieces of molecular networks and indeed there are still many thousands of genes whose functions are unknown and that must represent such missing pieces or new networks. Even proteins of unknown function are amenable to many types of analysis. The expression of the RNAs that encode the proteins and that subserve other functions has been accurately measured. This universe of RNA transcripts is the transcriptome, differing in every cell and within the cell at different points in the cell cycle and in response to changes in the cell’s environment, and cellular aging. The expression of RNA transcripts is intercorrelated, forming gene modules that often relate to broad and specific categories such as cell cycle, oxidative phosphorylation, synapse, and such functional pathways and genes associated with them have been systematically collated. A main goal of cell biology is to unravel those gene networks and to tie them to function, including mechanisms of disease, and targets for cures. The physical interactions of the proteins themselves can be directly measured and they can be colocalized within cellular compartments. Increasingly, these molecular explorations are under way in more specific cellular contexts and for example transcriptomics—the sequencing of the RNA—can be performed in specific subtypes of cells or by simultaneous single-cell sequencing of the RNA transcripts of thousands of cells. The unknown players are being found to connect to known molecular networks, where they play new, and sometimes more cell-specific roles. The complexity of the cell’s molecular networks is immediately obvious to any biochemistry student forced to memorize the glycolytic pathway (but why should anyone need to be forced?). However, the new RNA and protein technologies just mentioned, and systematic analyses of the published results on interactions, have enabled the concept of the protein interactome to be taken to the global cellular level. For those who don’t believe in reductionism, accommodate yourselves to the fact that after all our studies of the protein interactome, where any protein can be related to the function of any other through six degrees of separation, what we end up with is actor Kevin Bacon (Fig. 11.1). If the manufacturing process is perfected, an automobile can be built from perfectly machined parts such that it works each time, and one car of the same make and color is initially indistinguishable from the next. Having come this far, to the level of the protein interactome, let us step
152
11. The stochastic brain: From DNA blueprint to behavior
FIG. 11.1 A molecular interactome with six degrees of separation.
back and summarize why Kevin Bacon’s brain cannot be built that way. There are three answers to this question. The first is that the developing brain is a living functioning thing, not a piece of incomplete machinery on an assembly line. Therefore it must develop from one viable stage to the next, as the developing body partially recapitulates its phylogenetic origins. Second, nothing as complex as a brain could be specified from an instruction set of only 25,000 or so genes and perhaps an equal number of noncoding RNAs. The amazing fact is that brain development can be programmed from such an instruction set. Third, the plasticity inherent in a construction by a developmental process, rather than a piece-by-piece assembly, has given living organisms an enormous advantage in being able to adapt to different conditions present during development, and indeed our brains continue to exploit this plasticity throughout adult life.
Stochasticity How do coherent, adaptive behaviors emerge out of the complexity of random molecular chaos? Turning the question around, how does the chaotic, unpredictable setting and unfolding of our genetic program lead to individuality and choice? Is there something about our brains and genomes such that when we put “both hands in the pockets of chance,” free will appears as an emergent property of the human condition? Plasticity, which is the capacity for change, and random variation (also known as stochasticity, from the Greek stokhos—a guess) are hallmarks of the human brain. The fingers of identical twins closely resemble each other in length and shape. One twin’s fingers predict whether the others
Stochasticity
153
are long or short or webbed at the base. However, twins’ fingerprints are easily distinguished one from the other. If your evil twin wants to frame you for a crime, he should leave his DNA, not his fingerprints. The pattern of fingerprints is subject to general rules but is not programmed in details of the position and number of every ridge and whorl. The same goes for other external features of a person. Hairiness, hair color, and general distributions and whorled patterns are inherited but the positions of each hair on one’s head and body are random. Similarly, brain size and major characteristics of personality, cognition, and underlying neural mechanisms are strongly inherited; however, we are everyone—as individual as snowflakes. Within a snowstorm are billions of unique snowflakes and within our brains are billions of neurons, each with a unique pattern of synaptic arborization (tree-like branching). The general form and behavior of snowflakes and neurons obey certain general rules. Owing to the packing angles of water molecules, snowflakes have a general hexagonal symmetry and thus could never be mistaken for crystals formed from certain other molecules, for example the needle-like crystals of uric acid that plague gout patients. Pyramidal neurons of the cerebral cortex which send yard-long axons to the ventral root ganglion (neuronal cell mass) of the spinal cord all resemble each other much more strongly than they do other neurons or other cells of the body, and again their group coherence is due to the action of molecules that comprise them. There the similarity ends. Pyramidal neurons, and other neurons, are each comprised of a complex, constantly shifting, and reactive network of molecules adapted to work together in pathways and networks that function within the cell and across cells. Within the cell these build structures: nuclei, nucleoli, mitochondria, ribosomes, vesicles, Golgi apparatuses, which themselves vary in position of shape from one cell to the next. A snowflake might be compressed against its neighbors to be physically broken or fused, but it experiences only a shadow of the interactions of a neuron with its neighboring neurons and with the multitude of other cells that compose its environment. The snowflake does not have the “inner life” of a neuron, and thus by comparison its repertoire of possible responses is extremely limited. Snowflakes can melt and neurons can die, but neurons have a much wider range of reaction, and potential for random variation to alter that. Each neuron is not merely pressed against its neighbors but is dynamically interconnected, communicating electrochemically and metabolically with nearby cells and forming vastly complex networks that in the end enable the brain to produce all the outputs of the mind. A neuron is constantly being structurally and biochemically shaped by its neighbors but is also shaping them, and networks of neurons are shaping other networks, and the cells within them.
154
11. The stochastic brain: From DNA blueprint to behavior
Cascades, chaos, and great attractors Complex systems are not necessarily self-correcting. Complexity can instead make systems more vulnerable. One change, or defect, can lead to a cascade of effects. For example, with all its complex components, a space shuttle may fail because of a single glitch in a computer code. Or perhaps someone hits the wrong button (famous last words on a certain flight recorder: “Hey, what’s this button do?”). Recognizing that fact, backup systems are built into a space shuttle and our brain also has backup systems and mechanisms of compensation. Nevertheless, despite the backup systems, some errors lead to disaster. To make matters better, and understand how the brain can work, it is necessary to descend into chaos theory. Chaos theory is closely associated with the French American mathematician Benoit Mandelbrot, the father of fractal geometry (Mandelbrot, 1982). This theory defines some key features of the behavior of complex systems. An important feature of natural systems is their irregularity both at the macrolevel and at progressively finer degrees of magnification, as seen in the shapes of leaves and blood vessels, and the neurons’ complex branchings, internal structures, and biochemistry. The complexity visible at the macrolevel is reproduced at the microlevel—a phenomenon known as scale invariance—and leads to essentially infinite complexity. Thus the irregularity of a shoreline observable from outer space is also observable at the microscopic level. In his paper “How long is the coast of Britain? Statistical self-similarity and fractional dimension,” Mandelbrot showed that Britain’s coastline is infinite in length, if one uses an infinitely small measuring ruler. This infinite complexity of natural systems such as the brain is closely related to the occurrence of chaotic behavior in such systems. Chaotic behavior, which is often called the “butterfly effect,” is a property of certain types of change within certain types of dynamic system. As Frank Herbert observed (Herbert, 1965), “A beginning is a very delicate time.” A very small change in the “starting” state of the system can lead to a very large difference in the outcome. One thing leads to another. The result can be that even though all the component interactions within the system are deterministic the system appears to behave randomly, because the very small changes in initial starting state were essentially unmeasurable. The seriousness of the effects of chaos for understanding how the brain (which has billions of neurons) functions is emphasized by the fact that the mathematician Poincaré first discovered chaos in the 1880s when studying the three-body problem!
Brain assembly The locations, arborization structures, and biochemistries of billions of neurons cannot be specified, but their development can be programmed through a genome-guided developmental process. The brain programs itself, and it does so by a process of neural development by cellular
Brain assembly
155
interaction, beginning from the earliest moments in development when a cell that happens to be on the outside of a cell mass experiences a different environment than a cell on the inside of the same cell mass. These small starting differences lead to a cascade of changes. Also, and because of the unprogrammed molecular and structural complexity of cells, even two cells that are positioned so that they receive nearly identical influences will begin to diverge in subtle ways. Immature neurons derived from an embryonic structure known as the neural crest migrate to their final positions along a scaffolding of radial glial cells—crawling along these glial cells for long distances. Neurons born later migrate past those that have come before, occupying more remote territories, and the cells that are born later are already intrinsically different, leading to differences in the connections they “want” to form. After cells reach their final positions, they begin to grow, guided to the appropriate targets by contact with other cells and by combinations of adhesion molecules that are fixed in place and chemotactic molecular gradients of small molecules diffusing in the space between cells, all of which the amoeba-like neuronal growth cone encounters as it explores the environment. The growth cone of an axon often takes a route pioneered by the axon of another neuron, leading to the formation of neuronal fiber tracts. However, after neurons migrate into place and complete the process of forming axonal and dendritic connections with other cells, as many as half of them die. This massive cell death is thought to be due to competition between neurons for the attention of target neurons, including neurotrophic factors that are necessary for survival and that are secreted by target neurons. In many parts of the brain a dense network of synaptic connections is then formed which will later be refined by the elimination of synapses, in a process guided by use. Developmental sequence and basins of attraction thus tame a chaotic landscape of unpredictability. The brain is not created all at once or in a random way, but emerges in a developmental sequence, one moment in development setting the stage for the next. In adults, neuronal plasticity, and even neurogenesis—the generation of new neurons—is not finished, and adult neuronal plasticity is important in many ways. As elegantly detailed by Jay Giedd and colleagues at the National Institute of Mental Health (NIMH), and by others, myelination increases dramatically after birth, and there is progressive synaptic pruning through young adulthood (Gogtay et al., 2004). This is seen in Fig. 11.2, where through brain imaging the gray matter density can be seen to “cool down” from red and yellow colors to green and blue in these heatmapped images. Adult neurogenesis was first well appreciated in songbirds. Birds capable of learning new songs require new neurons. However, it has recently been appreciated that adult neurogenesis is also important in humans and other mammals, and in at least in two areas of the brain, the olfactory bulb and hippocampus. This is illustrated in Fig. 11.3, where in both cases the new neurons must migrate considerable distances to
156
11. The stochastic brain: From DNA blueprint to behavior
FIG. 11.2 Gray matter change from childhood to young adulthood. Source: Gogtay, N., Giedd, J.N., Lusk, L., et al., 2004. Dynamic mapping of human cortical development during childhood through early adulthood. Proc. Natl. Acad. Sci. U. S. A. 101, 8174–8179.
Human Brain Cerebral cortex Corpus callosum
Cerebellum
Lateral ventricle
Lateral Neural ventricle stem cells
Hippocampus Hippocampus
Olfactory bulb
New neurons
Rat Brain FIG. 11.3 Neuronal stem cells and adult neurogenesis in the hippocampus (left) and olfactory bulb (right). Source: Crews, F.T., Nixon, K., 2003. Alcohol, neural stem cells, and adult neurogenesis. Alcohol Health Res. World 27, 197–204.
Fractal neurons
157
find the proper locations, and then somehow integrate with neuronal networks. As shown by Rene Hen and his colleagues at Columbia University (Santarelli et al., 2003), adult neurogenesis in the hippocampus is crucial for recovery from depression, at least after treatment with the widely used selective serotonin reuptake inhibitor (SSRI) antidepressant drugs. This was proven in a very specific way: by irradiating the small region of the brain that has the neural stem cells that repopulate the hippocampus, Hen was able to block recovery in a rat model of depression. As scientists have artificially staged it, the developmental sequence of the brain resembles an expedition to climb a great mountain. The climbers do not scale the peak all at once, but move from base camp to progressively higher way stations, which they huddle in tents. Some are helped more by other climbers. Some turn back on the way or even die. In the development of the human brain, more than half the cells do not survive.
Fire together, wire together As noted by neuroscientist Carla Shatz, a major self-organizing principle of the brain is cells that “fire together, wire together,” in a way first conceived by DO Hebb (1949) the preservation of synapses depends on their excitation. Synapses and circuits that are used are strengthened and those unused are lost. Early in life, we pass through stages of great plasticity when language, music, and motor skills are most readily acquired. This is dramatically seen in the visual cortex, where the Nobel Laureate neurophysiologists Hubel and Wiesel (1962) famously showed that there was a narrow window of time—only about a week—when the occipital cortex must receive input to properly develop. If one side of the visual cortex is deprived, it is difficult to reverse the imbalance (amblyopia) that develops. Although amblyopia is called “lazy eye” it is actually “lazy brain.” Throughout adult life, thousands of new neurons are generated each day by neurogenesis, but most are lost within a few weeks of their birth, and the best way to retain unless there is effortful use of memory.
Fractal neurons The larger structures of the brain are programmed, but not the locations and interconnections of individual neurons or the fine structure and function of networks in the developed brain. In this regard, the human brain is very different than the brain of simpler organisms such as the nematode worm Caenorhabditis elegans. The identity and position of each of the 302 neurons of C. elegans can be forecast in advance. Much less can we predict the shapes of the extensions of the neurons in our
158
11. The stochastic brain: From DNA blueprint to behavior
brain, the synapses with other cells to form neural circuits, the locations of neurotransmitter-activated channels within a synapse, or the positioning and exact functional efficiency of individual secondary signaling proteins within the cytoplasm. Indeed, each of these is constantly changing in dynamic response to functional demands (Kandel, 2007). However, each of these phenomena is regulated by chemical and electrical gradients, and by interactions with other cells, and these interactions unfold in a temporal sequence. Rules shape the randomness. The shape of the maple leaf is fractal, but to the glory of Canada each maple leaf looks the same because of gradients of chemical and cellular interaction during development. If a gene programming the pattern of the maple leaf should be altered, the jagged form of the maple leaf might be smoothed to resemble the leaf of some other tree more closely. Similarly, there are many different types of neurons, several which have highly distinctive shapes to enable them to carry out their functions. The variety and complexity of neuronal shapes were first appreciated by the legendary Spanish neuroscientist Santiago Ramon y Cajal when he exploited the silver stain (the “black reaction”) invented by Golgi, who shared the Nobel Prize with him. Curiously, only one or a few neurons will intensely stain, allowing the forms of these neurons to be picked out from the surrounding cell mass (Sotelo, 2003). As Floyd Bloom said, “The gain in brain is mainly in the stain.” When I was a postdoctoral fellow, I helped my mentor Carl Merril adapt the Golgi stain for the staining of proteins and nucleic acids in electrophoretic gels, and it very quickly became a widely used, ultrasensitive and quantitative method (Merril et al., 1981)—probably many more protein gels than brains have been silver stained. But in the brain, where the silver stain detects cells, the structure that is revealed is far more interesting. The variety of neuronal cell types and configurations detected with the Golgi stain—each of which is an individual three-dimensional fractal structure—is illustrated in the two-dimensional line drawings in Fig. 11.4 (Sotelo, 2003).
Stochasticity in higher order brain structure As compared to a leaf’s individual patterns, or those of neurons, the higher order patterns of the brain, dubbed the connectome, are much more complicated. In brain structure, identical twins are more like each other than they are to people to whom they are unrelated. However, fingerprints of identical twins are not identical, and neither is the folding of their cerebral cortices or details of their wiring. The Technicolor brain slice images in Figs. 11.5 and 11.6, which are called “brainbow images,” illustrate the complexity of neuronal layering and interaction at different levels of the brain’s architecture.
Stochasticity in higher order brain structure
159
FIG. 11.4 Some neuronal cell types.
FIG. 11.5 Brainbow images of a neuron (left) and neurons (right). Courtesy of Lichtman, J.W., Sanes, J.R., Liver, J., 2008. A technicolor approach to the connectome. Nat. Rev. Neurosci. 9, 417–422.
This revolution in simultaneously imaging all the neurons in a brain region is the product of the imagination and technical skill of Jeffrey Lichtman and his colleagues and may well lead to another Nobel Prize, particularly if sufficient scanning and computational resources can ever be brought to enable the deciphering of the almost impossibly complex connectomes of even very small regions of the brain. These are not simply painted images of cells. Through a genetic trick, the cells paint themselves. A cassette of fluorescent protein genes is inserted and in different cells these genes are rearranged in different ways. The fluorescent proteins are of several colors, and as they are simultaneously
160
11. The stochastic brain: From DNA blueprint to behavior
FIG. 11.6 Brainbow images of neuronal fields. Courtesy of Lichtman, J.W., Sanes, J.R., Liver, J., 2008. A technicolor approach to the connectome. Nat. Rev. Neurosci. 9, 417–422.
expressed, they generate a palette of different colors so that nearby cells and their processes can be individually distinguished (Lichtman et al., 2008). At the first level (Fig. 11.5) is an individual neuron fluorescing red and embedded in a field of processes from neurons emitting other colors. These images enable reconstruction of not only the complex forms of individual neurons but also their interconnections. Thus far, the technology has been most successfully applied to supposedly simple systems, in particular the neuromuscular junction, where enormous and functionally meaningful variation has nevertheless been observed in the way neuronal connections are formed—variations that are beyond the ability of genes to specifically control. The images in Fig. 11.6 hint at higher levels, and different types, of architectures formed by neurons. Multiple neurons in a volume of brain are on the right, where there is a layer of neuronal cell bodies, and the axons have the same predominant (downward) orientation. Images such as this are inadequate to the task of defining the connectome in more than general terms but are already sufficient to confirm that the brain has infinite randomness, as well as an architectural framework to guide, or constrain, the randomness.
The stochastic basis of individual and group intelligence By now it should have become obvious that it is fundamentally wrong to think of the stochastic brain as a random system in the sense of the movement and collisions of dust motes. The brain is a system with an evolved “intelligence” that guides the behavior of its individual particles. A murmuration of starlings wheeling in the sky or a school of fish moving in unison
The stochastic basis of individual and group intelligence
161
has randomness but also cohesion. Murmurations wheel and turn, transform constantly in shape and extent, and make the most fantastical patterns beyond the comprehension, and certainly the planning of the individuals who comprise them. However, group cohesion and patterning are driven by the dim awareness of each individual element—bird or fish or neuron. Like a neuron, each element of the flock has extensive homeostatic and repair capabilities. The individual bird instinctively feeds and reproduces so that after it has died there is another to replace it. During breeding season individual birds pair and nest, “replenishing the ranks.” Yet each fall and as days become short the flock reassembles and behaves as it has for many years past. Each winter evening in my part of Maryland a living river of hundreds of thousands of black birds fly from harvested fields to roosting trees. These European Starlings and American Crows behave in this complex way because of genetically encoded general programs of behavior. Other winter-resident species, such as Carolina Chickadees, Blue Jays, Carolina Wrens, Barred Owls, Eastern Bluebirds, Pileated Woodpeckers, and Yellow-bellied Sapsuckers, have their own distinct behavioral patterns, and—remarkably—some of these diverse species have evolved to form mixed flocks of Titmice, Chickadees, and Nuthatches— different species acting in league much as different neurons cooperate as well as compete in the brain. Each evening the people who live beneath the trees favored by the Starlings await the arrival of their fine feathered friends. Such flocks avoid woods such as mine, which are too wild and jungle-like. Arriving at their park-like destination, the flying horde coalesces into a cloud that wheels, turns, and morphs into fantastical shapes until it finally settles into the trees where it makes a great cacophony and a considerable mess. The birds pack as close together as Hitchcock’s on the jungle gym of the little school where Annie Hayworth taught. Bicycling home each autumn I hear taped bird distress calls from one troubled house. Assuming that the homeowner is attempting to drive off the birds, I somewhat sympathize. However, the group mind of the flock is stubborn. While the behavior of any individual element of the flock would never be predictable, the landowner can predict with a very high degree of certainty that although he has not seen them for six months the birds will be back. The brain is also complex and unpredictable at the microlevel, and yet more predictable than a murmuration of birds at the macrolevel. Major brain structures and types of organization, for example the cortical layers, are found in almost all human brains. Brain circuits of reward, emotion, implicit memory, explicit memory, vestibular, autonomic, motor, sensory, auditory, and olfactory were structured by millions of years of selective pressure and display remarkable resilience in finding their way back to a conserved pattern of interaction and function. The effect of the blueprint is strong and part of the story of this book is the effect of gene variations
162
11. The stochastic brain: From DNA blueprint to behavior
and the ability of a person to make choices within that larger, genetically defined, pattern. However, the exact connections and outputs consequent to that blueprint are not predictable in detail. More importantly, neither are they consequent. Regarding predictability, at the DNA and biochemical level there is submicroscopic randomness (stochasticity) that individualizes all neurons one from the other, but really this is only the molecular signature of the functional individuality of each cell. In this regard, humans are unlike digital computing devices that rely on very large numbers of structurally and functionally identical components that can be linked together in different ways. The mind’s computational complexity is not merely a matter of its being composed of analog units capable of graded responses—computers can also be constructed from such components (Minsky, 2006). We are also unique because of the variability in our parts. No two people can be alike because no two of our neurons are even alike. However, our free will does not derive from randomness. Randomness is only the raw material from which a new ordered and individual neuronal complexity is extracted, and the rules set down in our genomes guide that process to destinations that are unique, but human, and self-guided.
Rules guiding the chaos of brain evolution and development There are rules for the development of a brain, and genes define some of the rules, enabling the human brain to form completely different structures than a fruit fly’s brain, and even allowing humans to choose in ways that fruit flies cannot. There are rules by which the brain evolved. However, both the human brain and the fruit fly brain develop in a similar general context of physics and biochemistry, and by some of the same molecular tricks. Within the molecular and cellular chaos are “great attractors,” both genetic and physical, that lead to confluences of outcomes in the structures and functional capabilities of brains. The fruit fly brain has a mushroom body; ours does not. Some of the constraints on structure are physical and some are accidents of evolutionary precedent. It might be handy, as Dr. Evil might have wished, for a shark to have evolved a laser beam with which to kill its prey, but the physical requirements are high and the evolutionary precedents available for adaptation or exaptation were nonexistent. Similarly, it would be advantageous for a person to be able to read others’ minds, so why has that never evolved? Obviously because it was not so easy to get to that adaptive peak, even though with powerful technologies we may eventually be able to reach it. When the confluence of physics and genetics has enabled a complex structure such as a brain to be built, that structure can be easily disrupted if some piece is missing. It is easier to disrupt or destroy a thing than to
Sense of self
163
build one. Lacking any of a myriad necessary nutrients and aspects of physical environment, the brain cannot develop properly. If a key gene is missing, the result can be a gross failure in the programming and in the resulting structure of the brain. For example, lissencephaly (an inherited disease marked by a smooth cortical surface) can result from the action of single gene defects, so can microcephaly (small brain). Human brains have important similarities to the brains of other mammals, including nonhuman primates. The brains of all these mammalian species develop the same generally similar large structures such as hippocampus and cerebellum, and these structures serve the same functions. However, certain key genes have recently been identified that are responsible for some of the large-scale functional differences between humans and even our nearest relatives. One dream of deep sequencing the genome of man’s closest relatives is to identify gene differences that are critical to human language and creativity. By sequencing ancient DNA, Svante Paabo at the Max Planck Institute has discovered that some Neanderthals and Denisovans, extinct relatives of man, cross-bred with ancestral humans such that some of their DNA survives in the human genome (Slon et al., 2018). But what are the genetic differences that may have limited Neanderthals and led to their ultimate displacement? Profound differences in language and cognitive abilities are ultimately attributable to these gene-programmed differences. Neanderthal’s neurons could not simply do whatever they wanted to do, or what was needed when faced by competition with Homo sapiens. Natural selection took the raw material of random mutation and genetic variation derived from such cross-breeding and made of that variation the human brain.
Sense of self Most people have a sense of self that extends all the way back to the earliest years of childhood. We feel that we are the “same person” now that we were then. Perhaps we even are that person (Rutter, 2010). An old man (me!) looks in the mirror and instead of seeing himself as he is, sees the boy who became the man (Hofstadter, 1980, 1985). We are a disorderly archive of memories, both explicit (conscious) and implicit (unconscious) of all our earlier versions. We may well have the same temperament and abilities (good with numbers, a genius at music, irresistible to the opposite sex—if this is what one sees in the mirror could it even be narcissism?), but in certain profound ways we are not the person we were. Each day we live, neurons and neuronal networks change, many neurons die, and some are replaced by other cells in a phenomenon known as adult neurogenesis. While we sleep and dream our brains reshape themselves. We start each day anew. In the ultimate abdication of responsibility, how can the brain person who awoke this morning be held to account for actions (positive or
164
11. The stochastic brain: From DNA blueprint to behavior
egative) of its predecessor? Was not that “someone else”? No. There are n two answers. First, the brain adaptively and randomly changes, but also fights for homeostasis, which is a process that lasts if we live. Second, our “self” is not a thing but a system of numberless interacting things. Much is transformed, but self endures. The leopard cannot change its spots, but more to the point the leopard changes from day to day but never into a giraffe. The flock of blackbirds discussed earlier continually replaces its individual elements but preserves continuity of behavior through space and across seasons. However, it is a misconception to imagine that physical continuity is even a requirement for conservation of self. Many “things,” such as ocean waves, require no continuity of physical elements as they translate through space and time. The great wave that crashed against the headland was the rising swell that could have been observed a mile offshore, although it brought with it none of the water molecules across that space. Indeed, like ocean waves, humans are complex waveforms translating through space–time, and subtly or more dramatically altering in form and function as we move forward. One of the true tragedies of severe brain diseases, and death, is this loss of self as the complex waveform dissipates, even as some effects of our being do translate forward far into the future and affect other lives. Metaphorically, a human self is a hurricane born as a confluence of rainstorms south of the Canaries, organizing and gaining strength as it moves into the Eastern Caribbean, thundering across leagues of ocean, and making landfall in South Texas, where it loses all distinct being. Throughout its “life” the hurricane’s structure and behavior are defined by physical laws, which is why hurricanes resemble each other as a class and why weather forecasters can roughly predict their strength and course. Their exact places and times of birth and death are impossible to predict. This lack of predictability does not, however, make a hurricane free, and for example to humanely avoid drowning people, or distributing rain where it is most needed. It is unaware of its past and its possible futures and of consequences and does not select between them; it goes where the steering currents take it. Like the waveform of a hurricane, the human brain develops and reacts in accordance with physical laws, but in addition it is informed by genetic and cultural instruction sets and is to a small but vital extent self-controlled. Hurricanes exceed in size and power of all living entities on Earth but generation after generation they remain the same. The hurricane has no progeny and leaves no genetic or cultural legacy to the next generation. It sings a song of wind and wave, but the song remains the same. It never learns unconsciously or consciously to alter its behavior by observing itself or others of its kind, for example to learn to avoid the mountainous island of Hispaniola. Natural selection, Dawkin’s “Blind Watchmaker” of evolution, cannot make one generation of hurricanes more powerful, wiser, or kinder than the previous ones (Dawkins, 1986). If there were no human
Sense of self
165
alive to hear its wind song, the hurricane and its winds would have existed, but is due to conscious observation by people that there is a lore and science of the hurricane, and this lore grows from generation to generation within a human framework outside a hurricane’s ken and kind. It is a mystery how one could define the moment of beginning and ending for any complex entity, but it is a special puzzle to demarcate the temporal boundaries of the human self (Hofstadter, 1980, 1985). It is the thesis of this book that neuroscience and neurogenetics yield the conclusion that humans are free, self-determining entities, deserving of respect. If this is true, then it is also crucial to define our beginnings and endings, and there are endless arguments on this point. If we become humans worthy of respect as free entities, when precisely do we deserve this reward, and responsibility? Is it the moment when haploid eggs and sperm, already genetically unique due to recombination and mutation, are formed? Does it happen at conception when sperm fertilizes egg and we become a single-celled, diploid life form? Or is it the eight-cell stage, the blastula, the gastrula, the quickening, birth, or perhaps when a child becomes capable of perceiving his own folly? I have explored life, death, self, and the definition of life in a later book, Immortal (Goldman, 2021), having stated in the first edition of this book that puzzles of beginnings and endings require more careful exploration. For this book, it is perhaps sufficient to observe that categorizing complex systems as alive versus unalive, free versus unfree, or conscious versus unconscious are all exercises that are inherently fraught with error and worthy of argument. We can state the precise moment when a cluster of thunderstorms become a tropical storm and a tropical storm a hurricane only because people have chosen to arbitrarily define this on the basis of sustained wind velocity. At some point the victim of Alzheimer’s disease loses touch with enough parts of their own self that the body and brain that go forward are no longer the former self, and at some point, there is no mind to make free choices. Long before that point there may have been sufficient impairment to justify legal guardianship, but here I am addressing a deeper and more difficult point. However, even as we contemplate beginnings and endings, we are also conscious that brain diseases can impair continuity of self in limited and sometimes reversible ways. Like a troubled electrical system, power can fade in and out, and our consciousness and will exist only as physical manifestations—there is nothing ethereal about them. Turn off the power and the machine stops. In an amazing demonstration of the fact that the brain is the person, any of us may suffer temporary amnesia or loss of a motor or sensory or language ability and recover. As discussed earlier in much greater depth, psychiatric disorders can distort our sense of self and locus of control. Addictions can make people become unlike their “true selves” in certain ways. The addict may steal, lie, crave and suffer misery because of their addiction and to a limited extent we can now help people with addictions
166
11. The stochastic brain: From DNA blueprint to behavior
and restore the addicted brain, but much better tools need to be developed beyond just the strategy of prevention. The depths of depression, the whirlwind chaos of mania, and the bizarreness of psychosis all disrupt the mind and are reversible, often spontaneously and often with the help of pharmacological or other intervention—even electroconvulsive therapy or electrical stimulation of regions of brain (Weinberger, 1986). Ultimately, as in the deep stages of dementia or in the death of the body that is the destiny for all of us, the thread of continuity between what we were and the diminished thing that we become is frayed. At the end of life, and like the thread of fate woven by the Norns in Norse mythology, it will break (I argue otherwise in Immortal but that is a longer conversation and a different book). As observed in Citizen Kane, aging is the one disease for which one does not want the cure, but the end of life, and the aging process, inexorably approach: at least superficially, no one gets out alive. However, if we are part of a continuing society of people, or perhaps a component of a community of other intelligences, something lives on of what we did and the spirit that guided us. The blackbird dies, but the flock lives. Also, like the Neanderthals who are now all long dead, we may transmit a genetic legacy to new generations who carry within themselves part of our DNA code, our “selfish genes” that through their selective advantage may achieve a type of immortality. Also, and again as conceived by Richard Dawkins, our selfish memes may survive, and to be transmitted both horizontally in society and as a legacy to generations that follow.
References Alberts, B., Johnson, A., Lewis, J., Raff, M., Roberts, K., Walter, P., 2002. Molecular Biology of the Cell, fourth ed. Byrne, J., Roberts, J., 2009. From Molecules to Networks: An Introduction to Cellular and Molecular Neuroscience, second ed. Crews, F.T., Nixon, K., 2003. Alcohol, neural stem cells, and adult neurogenesis. Alcohol Health Res. World 27, 197–204. Damasio, A., 2010. Self Comes to Mind: Constructing the Conscious Brain. Dawkins, R., 1986. The Blind Watchmaker. W.W. Norton & Co. Gazzaniga, M., 1994. Nature’s Mind: The Biological Roots of Thinking, Emotions, Sexuality, Language, and Intelligence. Gogtay, N., Giedd, J.N., Lusk, L., et al., 2004. Dynamic mapping of human cortical development during childhood through early adulthood. Proc. Natl. Acad. Sci. U. S. A. 101, 8174–8179. Goldman, D., 2021. Immortal. Elsevier. Hebb, D.O., 1949. The Organization of Behavior: A Neuropsychological Theory. Wiley, New York. Herbert, F., 1965. Dune. Hofstadter, D.R., 1980. Godel, Escher, Bach: An Eternal Golden Braid. Hofstadter, D.R., 1985. Metamagical Themas: Questing for the Essence of Mind and Pattern. Hubel, D.H., Wiesel, T.N., 1962. Receptive fields, binocular interaction, and functional architecture in the cat’s visual cortex. J. Physiol. 160 (1), 106–154.
References 167
Kandel, E.R., 2000. Principles of Neural Science. Kandel, E.R., 2007. In Search of Memory: The Emergence of a New Science of Mind. Kuffler, S., 2001. From Neuron to Brain: A Cellular and Molecular Approach to the Function of the Nervous System, fourth ed. Ledoux, J., 1998. The Emotional Brain: The Mysterious Underpinnings of Emotional Life. Ledoux, J., 2003. Synaptic Self: How Our Brains Become Who We Are. Lichtman, J.W., Sanes, J.R., Liver, J., 2008. A technicolor approach to the connectome. Nat. Rev. Neurosci. 9, 417–422. Mandelbrot, B.B., 1982. The Fractal Geometry of Nature. Merril, C.R., Goldman, D., Sedman, S.A., Ebert, M.H., 1981. Ultrasensitive stain for proteins in polyacrylamide gels shows regional variation in cerebrospinal fluid proteins. Science 211 (4489), 1437–1438. Minsky, M., 2006. The Emotion Machine: Commonsense Thinking, Artificial Intelligence, and the Future of the Human Mind. Rutter, M., 2010. Dorothy Bishop, Daniel Pine and Steven Scott, Rutter’s Child and Adolescent Psychiatry. Santarelli, L., Saxe, M., Gross, C., et al., 2003. Requirement of hippocampal neurogenesis for the behavioral effects of antidepressants. Science 301, 805–809. Slon, V., Mafessoni, F., Vernot, B., et al., 2018. The genome of the offspring of a Neanderthal mother and a Denisovan father. Nature 561, 113–116. https://doi.org/10.1038/ s41586-018-0455-x. Sotelo, C., 2003. Viewing the brain through the master hand of Ramon y Cajal. Nat. Rev. Neurosci. 4, 71–77. Weinberger, D.R., 1986. The pathogenesis of schizophrenia: a neurodevelopmental theory. In: Nasrallah, R.A., Weinberger, D.R. (Eds.), The Neurology of Schizophrenia. Elsevier, pp. 387–405.
This page intentionally left blank
12 Reintroducing genes and behavior One for sorrow, two for joy, three for a girl, four for a boy, five for silver, six for gold, seven for a secret never to be told. Traditional rhyme The tide of genetic behavioral prediction is rising, raising with it the question of whether behavior is chosen, or that we “come into this world with sealed orders,” as Soren Kierkegaard said. What are the limits of genetic prediction? Are some secrets of the mind forever inaccessible to all our genetic and physiologic probing? In the not-so futuristic movie, GATTACA, the protagonist Vincent Freeman sees his ambitions threatened by his “inferior” genome. His struggle is as much against the presumption of inferiority as with his innate handicaps. In a GATTACA world, genomes of the well-to-do are engineered not only to eliminate genetic glitches that lead to diseases but to enhance intelligence, stamina, and special talents, as exemplified by a pianist with genetically engineered polydactyly. From birth, people are tracked by their DNA and the assessment of a potential romantic partner may consist of a surreptitious DNA analysis. One’s talents or suitability may be judged based on predisposition, not disposition. At a job interview to be an astronaut, DNA is analyzed but Freeman substitutes the urine sample of a man with a superior genome—a so-called borrowed ladder. Afterward, he asks when the job interview will begin and is told, “You just had it.”
Our Genes, Our Choices https://doi.org/10.1016/B978-0-443-22161-3.00024-7
169
Copyright © 2024 David Goldman. Published by Elsevier Inc. All rights reserved.
170
12. Reintroducing genes and behavior
*HQRW\SLQJ RQHPLOOLRQ'1$PDUNHUV 6HTXHQFLQJ ZKROHJHQRPH (SLJHQHWLFV ZKROHJHQRPH
Halfway through the journey of this book, this is what has happened so far, if one picked it up and randomly read a page located just past the midpoint. We started with a gene—HTR2B—“for” behavior. A severe and common genetic variation in this neurotransmitter receptor was partially predictive of severe impulsivity and posed a challenge to the conception of free will. Next, we brooded over the measurement of behavior and the reification of some patterns of behavior into psychiatric diseases. We thrashed out the politics of behavior genetics, including the gene–eugenics connection, and the genethics of psychiatric genetics, and showed the hollowness and inconstancy of conceptions of an ethics in which people are treated as free even though in our hearts and brains we never acknowledge their autonomy. We visited the bestiary of psychiatric diseases that distort an individual’s ability to choose. This was followed by a step-by-step analysis of how a genetic blueprint can build a brain, and how intrinsic to that process is stochasticity that amplifies the neurogenetic individuality with which we began, and yet has rules that, by and large, guide the self-assembly of the brain successfully, using the dim awareness of its individual elements. We now turn to genetic and epigenetic tools that extend the prospect of prediction. These technologies include the simultaneous genotyping of upwards of a million genetic markers as well as massively parallel DNA sequencing to sequence the entire genome. Large-scale genotyping, both voluntary, involuntary, and by genetic proxy of relationship to someone else in databases, began almost simultaneously, but the power and scope
Behavioral prediction, a science imperfect
171
of DNA identification has been vastly expanded by the precipitous decrease in cost of the genotyping, followed by involuntary collection of millions of genotypes into forensic databases and by voluntary contributions of millions more to genealogic databases. Going beyond identification by DNA fingerprint, the character and characteristics of a person increasingly can be predicted. Also, and beyond the dystopian future of GATTACA, there is a new frontier of epigenetic studies in which the imprint of environmental experience on the genome can be measured and integrated with genetic information, to predict a person’s unique characteristics based on how genes and experience have interacted.
Behavioral prediction, a science imperfect Human behavior is notoriously unpredictable, as it will always be a problem for entities that have free will. Yet, psychiatrists and judges are routinely given the task of behavioral prediction, sometimes with life and death at stake. Repressive governments wish to identify and thwart behavior that threatens “stability” (a watchword in China) of the social order, the dream being to prevent unwanted behavior before it happens. This policing of thought as well as predisposition is exemplified by China in the present moment. However, even in more enlightened societies, or societies that are “less advanced” in surveillance and control, and for example the United States, scientifically based behavioral prediction is often enlisted by the criminal justice system. In the penalty (sentencing) phase following criminal conviction, how should predisposition, compulsion, and likelihood of recidivism be weighed? Should a pedophile be paroled given their exemplary behavior while incarcerated with other adults? What should be the treatment/punishment for a person with alcohol use disorder who has been arrested for driving under the influence, and does it matter if they have a polygenic risk score predicting alcohol use disorder (as is presently possible) or (slightly futuristically) noncompliance? Should a child be returned to a potentially abusive parent? Is it appropriate to try a 17-year-old accused as a murderer as an adult? DNA technologies can identify who we are but the more subtle and potentially powerful side of genetic analysis, in combination with a plethora of other measures ranging from structural and functional neuroimaging to neuropsychological testing, is its ability to tell us what we are. Should whole genome epigenetic analysis, by which the impact of experience including stress and mutagenic exposures can be measured, be used to help adjudicate cases in which plaintiffs ask for leniency at sentencing because life experiences have traumatized them and predisposed them to whatever crime they have been convicted?
172
12. Reintroducing genes and behavior
Commercialization of behavioral prediction Genetic prediction of behavior is upon us, appearing as a rosy- fingered dawn or dark threat to some, and for others an opportunity to make money. Allowing words to speak for themselves, on the website “DocBlum—Nutraceuticals for the Millennium,” Steve Allen, a television personality who once hosted the TV show “The greatest minds of all time,” named Dr. Kenneth Blum the scientist who “would change the world as we see it today.” In a widely unreplicated AND replicated study, a genetic marker lying outside and downstream from the dopamine receptor gene (DRD2) was associated with alcohol use disorder (Blum et al., 1990) and later this DRD2 marker was associated to many other conditions, lumped together as “Reward Deficiency Syndrome—RDS” and patented together with “nutraceuticals” to treat RDS. The “patent-protected, clinically-proven, and all-natural anti-craving ingredient complex” was a “dopamine neuronal-release promoter” taken as pill, powder, sublingual tablet, fruit juice, intravenously, intramuscularly, or intrarectally—more options than available for insulin. Because many genes, and gene x environment interactions, underpin behavioral variation, a single gene such as DRD2 is unlikely to predict a Reward Deficiency Syndrome, if such a thing exists. Indeed, from large genome-wide association studies (GWAS) we now know that the DRD2 gene is linked to several behaviors, including depression, neuroticism, and addictions. However, the predictive effects are small, and the genetic markers are different than the one Blum identified—that discovery and other replications that followed during the era of the candidate gene/ candidate marker were almost certainly driven by not-so-well-hidden stratification of cases and controls for differences in ancestry, and large interpopulation variation in the frequency of the chosen genetic marker, which was the first easily genotyped marker near DRD2, as elegantly shown by a brilliant young scientist in my lab, Yonwoo Jung (Jung et al., 2019). Most importantly, we now know that although the DRD2 gene is linked to some behaviors, the effects are weak—for example, the linkage to unipolar depression is based on a odds ratio of 1.03, which, while large in the field of behavioral genetics, is not, in real life, much better than a reading by a mystic. Genetic prediction by a growing range of companies also started at the single gene level, and as such was not of much consequence except as a business model, but now it has moved to an approach that is at once more powerful and useful, and at the same time therefore potentially more misused. This is the polygenic score (PGS) based on constellations of genetic markers. In psychiatry, the use of PGS is still nascent because of the
Commercialization of behavioral prediction
173
r elatively early state of the genomic studies, and as I have written, aiming to curb premature enthusiasm (Goldman, 2017, 2020). However, polygenic scores are an increasingly powerful, and having already been used to discover cross-inheritance of clinically distinct psychiatric disorders by genetic variants acting pleiotropically or transdiagnostically (Lee et al., 2021). Initially, and as shown in the “Clinical Reports” and “Research Reports” lists from 23andMe, a direct-to-consumer genetic testing company, the emphasis was on single gene-trait relationships. Even at this level, and with the low cost and convenience of DNA testing from a saliva sample, it is easy to see that some people out of curiosity or because of legal complications might invest in testing. A defendant might turn up something on this list such as dyslexia or alcohol dependence that would be mitigating or at least confusing. “Earwax type”—probably not, but alcohol-induced flushing is protective against alcohol use disorder and carried by more than half a billion people, implying that those who do not carry it are thereby more liable. Also, on 23andMe’s list as “Research Reports” were heroin addiction, avoidance of errors, memory, pain sensitivity, and back pain. The GWAS revolution, followed closely by whole exome and whole genome sequencing, has vastly expanded the number and variety of phenotypes for which genetic predictors are available. For example, as of 2021 the UK Biobank contains thousands of phenotypes measured in hundreds of thousands of volunteers who have donated their time, DNA, access to their electronic health records, and many of whom have volunteered for other physiologic and biometric measures. For example, half a million participants having completed assessment of reaction time, blood pressure, hand grip strength, various cell counts, and blood chemistries. In the United States, collections of >one million individuals are presently ongoing (https://allofus.nih.gov/). Genetic variations altering how choices are made have been mapped. From UK Biobank (https://genetics.opentargets.org/) we have evidence for more than 60 genetic loci influencing risk-taking behavior. Individually, their effects are small (the largest effect gene predicted 3% of the variance), but they can be combined into PGS, potentially to predict risk-taking behavior more robustly. There are pitfalls, including the diminished predictive value of PGS computed from one population when used in a second, and failures to consider gene-by-environment interactions and effects of age. However, the individual genetic findings are freely available and potentially of personal relevance to any person who has engaged in direct-to-consumer genetic testing. Furthermore, combinations of genetic markers can be made simplistically, for example by adding the beta values (variances predicted by markers a person carries) or in more sophisticated ways for computing PGS.
174
12. Reintroducing genes and behavior
The future of genetic behavioral prediction Because decisions to praise, blame, or guide behavior are made thousands of times a day, we should want to augment instinct with data, while shunning false panaceas. Seldom is DNA destiny, but it is a misappreciation of the science to disregard the importance of inheritance. For example, an important role for inheritance is known for several of the behaviors that lead to criminal behavior. These behaviors include alcohol use disorder and other substance use disorders, disorders of impulsivity including antisocial personality disorder, schizophrenia, and diminished cognitive capacity. Individual predictive genetic variants, some of which are already known, will be integrated with clinical history and neuropsychological testing to better define behavioral disorders, and better predict treatment response and outcome. They will also provide a more complete picture of the causation of criminal behavior. The understanding that a criminal behavior was strongly influenced, or determined, might be used to mitigate punishment. Also, the goal of biological psychiatry, of which genetics is a part, is to improve treatment and prevention by identifying new treatments and targeting the right treatment to the right individual. In a courtroom, evidence of predisposition is not necessarily exculpatory. It could reveal that a person is unlikely to change. This is especially true now when treatments, mainly, are only partially successful. The treatment of psychiatric diseases and behavioral problems is a vast but relatively unsuccessful industry. For several psychiatric diseases (depression, addictions) many of the FDA-approved treatments are only marginally better than placebo, although the consensus is that some good is done, especially in conjunction with other interventions. The diseases are complex in etiology and are poorly defined. When a doctor treats a patient who is homicidal, suicidal, or addicted, the treatment is based on a blurry diagnosis. When pediatricians, neurologists, endocrinologists, dermatologists, oncologists, or hematologists diagnose a patient, they choose from a palette of hundreds of specific diagnoses. In many cases they can rely on specific (pathognomonic) indicators such that the diagnosis is not only specific, but accurate in terms of identifying an etiologically homogeneous disease group. When physicians treat patients with psychiatric diseases, they usually lack these tools. Overall, medicine is also not so far advanced beyond its beginnings only a few millennia ago. Physicians have thousands of specific cures to offer; however, there are whole areas of medicine that, relatively speaking and like psychiatry, are still in the Dark Ages. There are numerous cancers that are poorly responsive to treatment. Alzheimer’s disease and other neurodegenerative disorders eventually will destroy most of our brains, if we live long enough, but mainly the treatments available delay, palliate, and partially mitigate. Millions of people suffer from chronic, largely
The future of genetic behavioral prediction
175
unrelieved, pain originating in the back, bones, joints, or other tissues for a variety of reasons some of which are slightly understood but usually not treatable at the root cause. Most paralyzed people stay paralyzed. Ultimately there is the problem of aging, which, as we were reminded by Citizen Kane’s attorney, is one disease of which no one wants to be cured. To diagnose and treat psychiatric diseases, access to the brain is partly the problem, but the larger obstacle is complexity. The brain is relatively inaccessible and animal models of human psychiatric diseases are for the most part missing (there is no adequate animal model of schizophrenia or bipolar disorder) or flawed (no mouse ever went to a bar, got drunk, crashed the car, came home, and beat their spouse). However, the functional complexity of 1 ounce of brain tissue is the greater problem, the brain’s complexity being greater than the rest of the body. The brain is living proof of Gödel’s theorem that no system can explain itself; to do so requires the creation of something even more complicated. The brain’s complex development, plasticity, and structural complexity are all factors that make it a very different and far more mysterious structure than any other organ in the body. The brain, with its billions of neurons, billions of interacting glial cells, and trillions of synapses, must be largely self- organizing, starting from an instruction set of only 25,000 genes. On the other hand, this also helps enable the brain to grow adaptively and learn, as already discussed in Chapter 11. From the complexity of the brain, it follows that any individual gene that predicts behavior will not necessarily be explanatory, and this is especially true because behavioral variation is polygenic. Further, it does not necessarily follow that linkage between a DNA marker to a psychiatric disease leads to an understanding of how the genetic variation alters behavior. If the understanding of the intervening steps is ever achieved this often comes years after the discovery of the genetic marker and long after it has been used as a predictor. Examples are legion, from monogenic diseases such as Huntington’s disease, which causes movement disorder and psychosis and Lesch–Nyhan syndrome, which causes self-mutilation and kidney stones, to chromosomal anomalies such as Trisomy 21 or velocardiofacial syndrome (whose genetic origin is deletion of a small but multigenic region of chromosome 22 leading to variable musculoskeletal anomalies and a high incidence of schizophrenia), which are precisely defined at the genetic level but poorly understood at the level of cell molecular physiology. It is a little as if a policeman notices that the red cars are faster than blue ones. Perhaps they now begin to pay more attention to the red cars. When physicians use superficial characteristics, we are less concerned about stereotyping and more interested in whether they are diagnosing the condition. Increasingly, neurogenetics is turning to a variety of tools, deepening our grasp of the phenotype from superficial layers of appearance and externally manifested behavior to deep measures that
176
12. Reintroducing genes and behavior
represent measures of processes that mediate between genes and those externally visible phenotypes. These intermediate phenotypes, sometimes mistakenly called endophenotypes, include transcriptomics (measurement of the RNA transcripts of cells) to functional and structural neuroimaging. Increasingly, genetic variants are studied in model systems, ranging from cultured cells and organoids to genetically engineered animals, to identify the molecular and neuronal network mechanisms by which genes alter behavior. Although the brain’s complexity is beginning to be unraveled, it is still more likely that the mechanism of action of a genetic variant predicting brain function will remain mysterious long after we understand why some cars are faster than others. The mechanisms of many genetic diseases that affect other organs also remain mysterious, but some of the peculiar challenges of unraveling mechanisms of brain diseases can be illustrated by contrasting progress made in understanding a few diseases whose genes have been known for decades (Rimoin et al., 2006; Scriver et al., 2000; Speicher et al., 2010). Here, I contrast knowledge and the discovery pathways for a gene causing acute hemolytic anemia (G6PD deficiency) versus different genes causing aggression and self-mutilation (specifically, Lesch–Nyhan syndrome, arising from HPRT deficiency) and cognitive deficiency (phenylketonuria, usually caused by PAH deficiency). These gene discoveries were made about 50 years ago, so there has been ample time to work out an understanding of mechanisms or to establish that what we do not know will be difficult to uncover. Also, these discoveries have led to specific approaches to prevent or mitigate disease.
A gene causing anemia One cause of acute hemolytic anemia is inherited deficiency of lucose-6-phosphate dehydrogenase (G6PD). G6PD has several importg ant functions, one of which is to regenerate NADPH—a molecule that can undo the oxidation of sulfhydryl groups found on many proteins of the cell. That is important, because otherwise the damaged proteins do not work and can even become toxic if not removed. The red cell is particularly susceptible to oxidative damage after a person eats certain foods (e.g., the always delicious fava bean) or takes certain medicines (e.g., antimalarial drugs). During the Korean War, a large fraction of African American male soldiers suffered massive hemolysis (destruction) of their red cells after being given chloroquine for malaria prophylaxis. It was noted that if the medicine was maintained, in time the hemolysis subsided and red blood counts again increased toward preexposure levels because the older, more fragile red cells had disappeared and the production of erythrocytes by the bone marrow had ramped up, their blood now having a higher proportion of young red blood cells known as reticulocytes, because polyribosomes
A gene causing self-mutilation
177
still busily manufacturing hemoglobin and other red cell proteins gave them a characteristic appearance. The molecular signals and process by which red cell production was increased also came to be understood at the molecular level and provided the intellectual property basis for one of the world’s first successful biotech companies (Amgen), which was built on the hormone erythropoietin. Two common genetic types of G6PD deficiency, one endemic to West Africa and the other found in peoples of the Mediterranean basin, were identified, and the mechanism for both was quickly worked out. Further, hundreds of other rarer G6PD mutations were found. Finally, it was shown that G6PD variants triggering hemolytic anemia are common in West Africa and the Mediterranean Basin because they confer resistance against malaria, part of whose life cycle in the body involves malaria merozoites infecting red blood cells where they proliferate, and apparently even communicate cell to cell while in the human body (Regev-Rudzki et al., 2013). A circle, illustrating the “poison to poison” principle of ancient Chinese medicine, was closed: the red cell fragility mechanism by which chloroquine, a mainstay of malaria treatment, triggered hemolysis in G6PD—carriers were connected to the mechanism by which G6PD deficiency conferred resistance to malaria. Malaria does not do as well in red cells made fragile by a gene or a drug. Things have not worked out nearly so neatly for genetic mental disorders even when the specific gene is known and even when the molecular function of the gene is known. Two examples are Lesch–Nyhan syndrome and phenylketonuria. Both, like G6PD deficiency, are biochemical genetic disorders that disturb well-trodden enzyme pathways. The genetic variants responsible have been pinpointed and we know how they impair or block the functions of the enzymes. Also, alterations in levels of small molecules the enzymes act upon have been precisely measured. The mystery is how the altered biochemistry translates to altered behavior.
A gene causing self-mutilation Lesch–Nyhan syndrome is caused by deficiency of the HPRT enzyme, whose name, hypoxanthine-guanine phosphoribosyltransferase, tells the tale of the main reaction it catalyzes. HPRT enables the salvage of purines, which we earlier learned are important as building blocks for DNA and ATP (adenosine triphosphate, the famous molecule that plays a major role in cellular energy transfer). The disease was first recognized in 1964 by a medical student at Johns Hopkins, Michael Lesch, working with his mentor William Nyhan, a biochemical geneticist. When I was a clinical fellow training at the National Institutes of Mental Health, I was fortunate to see several of these patients, as part of a biochemical genetic study. Carl Merril and I were studying these boys precisely because they represented
178
12. Reintroducing genes and behavior
one of the very few instances at that time—some four decades ago—in which a specific genetic variation was known to cause a complex behavioral phenotype. It was the intermediate steps between the genetic variants and behavior that were not understood, and they are still unclear. When Carl and I arrived at Rosewood Hospital, where the affected boys were in chronic care, they were restrained in chairs to prevent them from mutilating themselves or behaving aggressively as they especially tended to do when made anxious, for example by encountering strangers in white coats. We had not touched them, and they were emotionally calm and not unhappy, but nevertheless they grimaced and twisted their arms. Sadly, the boys had mutilated the ends of their fingers. The mutilation of fingertips in Lesch–Nyhan patients is much worse than nail-biting, something I did as a young doctor. When I drew the blood sample one of the boys pushed his arm toward the needle. I had stuck needles in many arms, but that had never happened. Using cells from these boys, I found that there was a proteomic signature of Lesch–Nyhan syndrome—these boys could be identified not only by the primary enzyme deficiency but by other proteins that were differentially regulated in their cells (Merril et al., 1981). The behavioral syndrome of Lesch–Nyhan is just one part of a disease whose biochemistry involves the unleashing of the synthesis of purines from normal restraints. Although Lesch–Nyhan syndrome is rare it shares biochemistry and some of its nonbehavioral symptoms with the much more common disease of gout. Gout was known to the ancient Egyptians and has been recognized throughout medical antiquity as the arthritis of the rich and powerful. The mechanisms of nonbehavioral manifestations of purine diseases have been worked out. Purines that are key to gout and Lesch– Nyhan syndrome are synthesized in the body and directly derived from the diet, particularly foods rich in DNA—especially dark muscle, fish, and other meat tissues that are rich in mitochondria, which has its own DNA genome. An excess of these foods leads to an excess of uric acid. This metabolic product of purines is normally excreted by the kidneys, and to a lesser extent the gut, but high concentrations of uric acid cause a problem, whether those high concentrations are due to diet or genetics. If the urine is acidified, crystals of sodium urate precipitate from solution in the kidneys, causing stones. In the joints (especially the big toe), needle-like crystals trigger inflammation and pain. Crystalline deposits in other tissues are called tophi. In gout and Lesch–Nyhan syndrome the uric acid crystals are both macroscopic and microscopic. In gout, microscopic uric acid crystals were first seen by Antonie von Leeuwenhoek, who thereby—and already in the 17th century—was directly observing a disease molecule causing pain. You have to be pretty careful when you choose a new name for yourself. Von Leeuwenhoek, who was born Thonis Philipzoon in 1632 but renamed himself “Anthony from the Lion’s Corner” used Hooke’s recently invented microscope to make a series of amazing discoveries: protozoans, red blood cells, the
A gene causing cognitive deficiency
179
compound eyes of insects, microparasites, sperm, parthenogenesis, bacteria in dental tartar, blood capillaries. He justly became world renowned despite never having published a scientific paper or presenting his work at a scientific conference. With his microscope it was von Leeuwenhoek who in 1679 observed the distinctive needle-like crystals of sodium urate in tophi from a gout patient, and he correctly guessed that these were the cause of the pain. In contrast to gout or serum uric acid levels, the latter for which there are already 183 loci discovered by genome-wide association studies (Tin et al., 2019), Lesch–Nyhan syndrome is an X-linked recessive disease occurring in 100–200 of U.S. males and identified on a case-by-case basis (Hoffmann et al., 2010; Khattak et al., 1998). Lesch–Nyhan syndrome is even rarer in females because the HPRT gene is located on the X chromosome, normally found in two copies in females. The predicted frequency of the Lesch– Nyhan syndrome in females is therefore the square of the risk of the disease in males—the odds both X chromosomes carry a defective HPRT allele. However, females are twice as likely to carry at least one, and such unaffected carriers have a 50% chance of transmitting the disease variant to any of their offspring. Lesch–Nyhan syndrome is a severe, multisystem disorder. Due to the blockade of the purine salvage pathway, by which the body recycles purines, the de novo synthesis of purines is vastly accelerated as is the downstream wastage of uric acid when these purines are metabolized. Infants with Lesch–Nyhan syndrome can be spotted because of crystals of uric acid in the diapers. Later they endure gout and kidney stones, moderate mental retardation, poor muscle control, spasticity and involuntary grimacing, and the writhing movements known as choreoathetosis. Most never walk. Drugs such as allopurinol help prevent the gout and kidney stones but at 2–3 years of age self-mutilation and aggression emerge in about 85% of the patients such that most have their teeth extracted. Although so far not detected as a cause of behavioral differences in genome-wide association studies, possibly because the gene variations are still too rare to generate detectable linkage signals by this method or because the HPRT gene is on the usually neglected X chromosome, milder defective variants of HPRT can lead to elevations of uric acid, and gout without the neurologic manifestations of Lesch–Nyhan syndrome, and as is known as the Kelley–Seegmiller syndrome (Khattak et al., 1998).
A gene causing cognitive deficiency Children with Phenylketonuria (PKU) tend to have blue eyes and fair skin because without phenylalanine hydroxylase they cannot convert phenylalanine to tyrosine, an essential metabolic step in the synthesis of melanin. Untreated, they can suffer from a variety of problems including intellectual disability, emotional problems, eczema, and seizures and have
180
12. Reintroducing genes and behavior
a mousy, musty odor. Nevertheless, and still for unknown reasons, PKU is relatively common in Caucasians, occurring in one in 10,000 newborns. Furthermore, because PKU is an autosomal recessive, as many as 1 in 100 Caucasians are carriers who although unaffected by PKU may have a heterozygous advantage over noncarriers, as has long been speculated but still unproven. PKU is one of a group of inborn errors that disrupt the metabolism of various amino acids, leading to their accumulation in the urine, and as such these diseases, including isovaleric acidemia, maple syrup urine disease (branched chain ketoaciduria caused by mutations in at least four different genes), methylmalonic acidemia, and propionic acidemia, were among the first to be understood at the molecular level. A key to solving the mystery of the PKU disease itself was that the children’s urine had a musty odor. This led Norwegian physician Asbjørn Følling to discover the molecular cause in 1934 when the mother of two intellectually impaired children directly asked him whether their problems might be related to the musty smell of their urine. When Følling added ferric chloride to the children’s urine, it turned dark green. He extracted and purified the responsible molecule, finding that it was phenylpyruvic acid, an alternative metabolite of phenylalanine if its conversion to tyrosine is blocked. Later, levels of other phenylalanine metabolites, phenyllactate and phenylacetate, were also shown increased in PKU. About a decade after Følling’s discovery, Jervis showed that PKU is usually caused by deficiency of the enzyme, phenylalanine hydroxylase (PAH). PKU is also more rarely caused by deficiencies of enzymes that synthesize PAH’s main cofactor. If one has PAH deficiency, high levels of phenylalanine and the metabolite phenylpyruvic acid build up in the body, and most importantly, the brain. However, Bickel (1953) showed that this build up, and the consequent effects, can be prevented by limiting phenylalanine in the diet, a course made possible by detection of PKU by the early screening of infants, as will be described. PKU is thus a common (in some populations), diagnosable, treatable, genetic behavioral disorder. However, the story does not end with treatment of children, whose brains are developmentally vulnerable. Fetuses of mothers with PKU can suffer developmental impairment due to toxic metabolites passing from mother to fetus. Kaufman’s letter: When I was a postdoctoral fellow interested in understanding genetic metabolic diseases affecting the brain, a letter by Seymour Kaufman, an authority on neurotransmitter biochemistry, was published in the Washington Post. Kaufman was a principal investigator in the NIMH lab of Giulio Cantoni (the discoverer of the methyl donor S-adenosyl methionine), where I was being mentored by Carl Merril, and enjoying hearing Kaufman, Werner Klee, David Neville, Howard Nash, and other outstanding scientists who had been lumped into Cantoni’s lab hold forth at lunch and in journal club. Kaufman claimed that Phenylketonuria (PKU) was not an example of genetics unraveling a behavioral disease.
A gene causing cognitive deficiency
181
Kaufman’s argument was that the cause of PKU was unraveled not by study of DNA but via biochemical findings and enzymology—studies of small molecules and proteins. I felt this was a false distinction; although PKU was not a DNA-based discovery it was a prime example of biochemical genetics in action. Transmission of an enzyme deficit is genetic in the same way that transmission of a blood marker such as ABO detected via an antibody reaction is genetic even though neither is DNA based. To work out the (sometimes violated) laws of inheritance, Mendel measured visible characteristics of his peas, not DNA, or even molecules, such as phenylpyruvic acid, that are more directly regulated by DNA. Kaufman was not persuaded, but over many years I persuaded some. Genetic testing and tests with genetic implications are frequently unrecognized as such. However, direct measurement of DNA sequence immediately leads to all kinds of concerns about genetic testing even if it adds no new information. For example, if a person is diagnosed with PKU it is almost certain that both parents are heterozygous carriers of this autosomal recessive disease. Even more specifically, hemolytic anemia due to G6PD deficiency, an X-linked disease, is usually transmitted from unaffected carrier mothers to affected sons and unaffected carrier daughters. Thus it is almost certain that his mother was an unaffected carrier. The diagnosis of G6PD deficiency can be made clinically and via a measurement of his enzyme level but there is a powerful genetic implication for his mother, and any of his male siblings could immediately be known to be at least at 50% risk of G6PD deficiency, with no testing performed at all. G6PD deficiency, like PKU, is an example of a functional genetic test, with all the usual implications including ethnic genetics—G6PD deficiency is far more common in Africans as well as in Southern Europeans. There is no perfect boundary between the disciplines of genetics and cellular molecular biochemistry—genetics illuminates biochemistry and understanding the molecular pathways and networks of the cell is the endgame of genetics: the isolation of the functional locus and its mechanism of action (Alberts et al., 2002). However, for either G6PD deficiency or PKU it was a genetic disease that was being unraveled—by the means available. Although measurement of phenylalanine hydroxylase enzyme activity is itself a genetic measurement, the screening for PKU in infants has been accomplished even more cleverly and efficiently for the past half century by a bioassay using another life form in the original neonatal screening assay by Guthrie (Guthrie and Susi, 1963). Guthrie was motivated by having fathered a child initially thought to have PKU. Although his son did not, a niece did, and was cognitively disabled as a result. Growth of Bacillus subtilis in a Petri dish was inhibited by B2 thienylalanine and this inhibition was reversed by phenylalanine in the blood of the infant being screened, the circle of growth being proportional to the amount of phenylalanine in the infant’s blood. In the 1960s the Guthrie test was widely adopted
182
12. Reintroducing genes and behavior
worldwide (although not everywhere), paving the way to screening of newborns for other genetic disorders and, in the case of PKU, leading to the protection of thousands of children via dietary restriction of phenylalanine. Modifications of the Guthrie test followed, with enzymatic assays and assays of phenylalanine by mass spectrometry replacing the growth, which enables the simultaneous screening for other metabolic enzyme deficiencies. The overall public health impact of screening for PKU, and the follow-up dietary restrictions, is to prevent more than 27,000 cases of cognitive impairment annually, in the United States alone. Nowadays, PAH mutations causing PKU are readily detectable by direct-to-consumer DNA sequencing or array-based genotyping. Using such genotypes, another way of appreciating the high frequency and effect size of the common phenylalanine hydroxylase variant in Europeans is that in large genome-wide association studies PAH is readily detectable as a gene of large effect on plasma amino acid levels (http://europepmc.org/article/ MED/27005778). We have now seen that specific genes can predict behavior. Next, and perhaps as already predicted by the fact that the impact of deficiencies of G6PD and PAH can be limited by modifying the diet, we will see that other functional loci are more highly predictive in the context of an environmental exposure.
References Alberts, B., Johnson, A., Lewis, J., Raff, M., Roberts, K., Walter, P., 2002. Molecular Biology of the Cell, fourth ed. Bickel, H., 1953. Influence of phenylalanine intake on phenylketonuria. Lancet 5, 812–813. Blum, K., et al., 1990. Allelic association of human dopamine D2 receptor gene in alcoholism. JAMA 263, 2055–2060. Goldman, D., 2017. Polygenic risk scores in psychiatry. Biol. Psychiatry 82 (10), 698–699. Goldman, D., 2020. Predicting suicide. Am. J. Psychiatry 177 (10), 881–883. Guthrie, R., Susi, A., 1963. A simple phenylalanine method for detecting phenylketonuria in large populations of newborn infants. Pediatrics 32, 338–343. Hoffmann, G., Zschocke, J., Nyhan, W., 2010. Inherited Metabolic Diseases: A Clinical Approach. Jung, Y., Montel, R.A., Shen, P.H., Mash, D.C., Goldman, D., 2019. Assessment of the association of D2 dopamine receptor gene and reported allele frequencies with alcohol use disorders: a systematic review and meta-analysis. JAMA Netw. Open 2 (11), e1914940. https:// doi.org/10.1001/jamanetworkopen.2019.14940. 31702801. PMCID: PMC6902783. Khattak, F.H., Morris, I.M., Harris, K., 1998. Kelley-Seegmiller syndrome: a case report and review of the literature. Br. J. Rheumatol. 37 (5), 580–581. https://doi.org/10.1093/rheumatology/37.5.580c. 9651092. Lee, P., et al., 2021. Pleiotropy and cross-disorder genetics among psychiatric disorders. Biol. Psychiatry 89, 20–31. https://doi.org/10.1016/j.biopsych.2020.09.026. Merril, C.R., Goldman, D., Ebert, M., 1981. Protein variations associated with Lesch-Nyhan syndrome. Proc. Natl. Acad. Sci. U. S. A. 78 (10), 6471–6475. https://doi.org/10.1073/ pnas.78.10.6471. 6947238. PMCID: PMC349061.
References 183
Regev-Rudzki, N., Wilson, D.W., Carvalho, T.G., Sisquella, X., Coleman, B.M., Rug, M., Bursac, D., Angrisano, F., Gee, M., Hill, A.F., Baum, J., Cowman, A.F., 2013. Cell-cell communication between malaria-infected red blood cells via exosome-like vesicles. Cell 153 (5), 1120–1133. https://doi.org/10.1016/j.cell.2013.04.029. Epub 2013 May 15. PMID: 23683579. Rimoin, D., Connor, J.M., Pyeritz, R., Korf, B., 2006. Emery and Rimoin’s Principles and Practice of Medical Genetics and Genomics. Churchill Livingstone. Scriver, C., Sly, W., Childs, B., Beaudet, A., 2000. The Metabolic and Molecular Bases of Inherited Disease. Speicher, M., Antonarakis, S., Motulsky, A., 2010. Human Genetics, fourth ed. Tin, A., Marten, J., Halperin Kuhns, V.L., Li, Y., Wuttke, M., Kirsten, H., Sieber, K.B., Qiu, C., Gorski, M., Yu, Z., Giri, A., Sveinbjornsson, G., Li, M., Chu, A.Y., Hoppmann, A., O'Connor, L.J., Prins, B., Nutile, T., Noce, D., Akiyama, M., Cocca, M., Ghasemi, S., van der Most, P.J., Horn, K., Xu, Y., Fuchsberger, C., Sedaghat, S., Afaq, S., Amin, N., Ärnlöv, J., Bakker, S.J.L., Bansal, N., Baptista, D., Bergmann, S., Biggs, M.L., Biino, G., Boerwinkle, E., Bottinger, E.P., Boutin, T.S., Brumat, M., Burkhardt, R., Campana, E., Campbell, A., Campbell, H., Carroll, R.J., Catamo, E., Chambers, J.C., Ciullo, M., Concas, M.P., Coresh, J., Corre, T., Cusi, D., Felicita, S.C., de Borst, M.H., De Grandi, A., de Mutsert, R., de Vries, A.P.J., Delgado, G., Demirkan, A., Devuyst, O., Dittrich, K., Eckardt, K.U., Ehret, G., Endlich, K., Evans, M.K., Gansevoort, R.T., Gasparini, P., Giedraitis, V., Gieger, C., Girotto, G., Gögele, M., Gordon, S.D., Gudbjartsson, D.F., Gudnason, V., German Chronic Kidney Disease Study, Haller, T., Hamet, P., Harris, T.B., Hayward, C., Hicks, A.A., Hofer, E., Holm, H., Huang, W., Hutri-Kähönen, N., Hwang, S.J., Ikram, M.A., Lewis, R.M., Ingelsson, E., Jakobsdottir, J., Jonsdottir, I., Jonsson, H., Joshi, P.K., Josyula, N.S., Jung, B., Kähönen, M., Kamatani, Y., Kanai, M., Kerr, S.M., Kiess, W., Kleber, M.E., Koenig, W., Kooner, J.S., Körner, A., Kovacs, P., Krämer, B.K., Kronenberg, F., Kubo, M., Kühnel, B., La Bianca, M., Lange, L.A., Lehne, B., Lehtimäki, T., Study, L.C., Liu, J., Loeffler, M., Loos, R.J.F., Lyytikäinen, L.P., Magi, R., Mahajan, A., Martin, N.G., März, W., Mascalzoni, D., Matsuda, K., Meisinger, C., Meitinger, T., Metspalu, A., Milaneschi, Y., Program, V.A.M.V., O'Donnell, C.J., Wilson, O.D., Gaziano, J.M., Mishra, P.P., Mohlke, K.L., Mononen, N., Montgomery, G.W., Mook-Kanamori, D.O., Müller-Nurasyid, M., Nadkarni, G.N., Nalls, M.A., Nauck, M., Nikus, K., Ning, B., Nolte, I.M., Noordam, R., O'Connell, J.R., Olafsson, I., Padmanabhan, S., Penninx, B.W.J.H., Perls, T., Peters, A., Pirastu, M., Pirastu, N., Pistis, G., Polasek, O., Ponte, B., Porteous, D.J., Poulain, T., Preuss, M.H., Rabelink, T.J., Raffield, L.M., Raitakari, O.T., Rettig, R., Rheinberger, M., Rice, K.M., Rizzi, F., Robino, A., Rudan, I., Krajcoviechova, A., Cifkova, R., Rueedi, R., Ruggiero, D., Ryan, K.A., Saba, Y., Salvi, E., Schmidt, H., Schmidt, R., Shaffer, C.M., Smith, A.V., Smith, B.H., Spracklen, C.N., Strauch, K., Stumvoll, M., Sulem, P., Tajuddin, S.M., Teren, A., Thiery, J., Thio, C.H.L., Thorsteinsdottir, U., Toniolo, D., Tönjes, A., Tremblay, J., Uitterlinden, A.G., Vaccargiu, S., van der Harst, P., van Duijn, C.M., Verweij, N., Völker, U., Vollenweider, P., Waeber, G., Waldenberger, M., Whitfield, J.B., Wild, S.H., Wilson, J.F., Yang, Q., Zhang, W., Zonderman, A.B., Bochud, M., Wilson, J.G., Pendergrass, S.A., Ho, K., Parsa, A., Pramstaller, P.P., Psaty, B.M., Böger, C.A., Snieder, H., Butterworth, A.S., Okada, Y., Edwards, T.L., Stefansson, K., Susztak, K., Scholz, M., Heid, I.M., Hung, A.M., Teumer, A., Pattaro, C., Woodward, O.M., Vitart, V., Köttgen, A., 2019. Target genes, variants, tissues and transcriptional pathways influencing human serum urate levels. Nat. Genet. 51 (10), 1459–1474. https://doi.org/10.1038/s41588-019-0504-x. Epub 2019 Oct 2. PMID: 31578528; PMCID: PMC6858555.
This page intentionally left blank
13 Warriors and worriers An eye for an eye makes the whole world blind. Gandhi Why should there be common functional genetic variants that cause some people to behave differently? Why are there common functional variants at genes such as the serotonin transporter (SLC6A4) (Caspi et al., 2003), monoamine oxidase A (MAOA) (Caspi et al., 2002), FKBP5 (a chaperonin of the cortisol receptor that down-regulates stress responses) (Binder et al., 2008), Neuropeptide Y (NPY) (Zhou et al., 2008), and Brainderived Neurotrophic Factor (BDNF) (Egan et al., 2003)? Sometimes, the genetic variation exists for a different reason, for example resistance to an infectious disease, but pleiotropically also alters behavior. That may happen only when environmental conditions change, and for example when human learned to ferment alcohol, and suddenly the common ALDH2 variant causing flushing became a protective gene against alcohol use disorder. Before alcohol, it was not. The same gene, complement C4, that helps mediate intrinsic immunity (a type of immunity) against pathogens, has been implicated in schizophrenia, although it is doubtful that the gene evolved for millions of years to cause schizophrenia. Behavior is also influenced by a plethora of rare and uncommon genetic variants, many that may have originated recently by mutation and not been eliminated by natural selection. But sometimes the genetic variants are genes “For” behavior, having been selected for their ability to predispose to one behavioral strength or another, and not representing an accidental effect on behavior. Such variants are often both ancient and common worldwide. The reason is that communities composed of varieties of people were more successful, but at the level of the individual selfish gene the explanation is that the gene for a behavior may be selectively advantageous when the behavior is rare but disadvantageous when the behavior is common, and that variants influencing behavior are advantageous at certain times and in certain niches, but not in others. In this chapter we will discuss this mechanism: balanced selection. It has maintained two common forms of a gene that influences both cognition and emotion, and in these two behavioral domains the
Our Genes, Our Choices https://doi.org/10.1016/B978-0-443-22161-3.00006-5
185
Copyright © 2024 David Goldman. Published by Elsevier Inc. All rights reserved.
186
13. Warriors and worriers
a dvantages of the two forms of the gene counterbalance, leading to “warriors” and “worriers,” each of whom may find their niche within society. Evolutionarily, the advantage goes to the individual best suited to occupy an available opportunity in the community, but the result is that no human society is entirely composed of Attilas or Gandhis. We are both warriors and worriers, and fortunately the balance has sometimes worked, after seasons of violence and vengeance. The off-target (pleiotropic) effects and relatively rare new mutations do not account for many of the common genetic variations that influence behavior, and much of the intrinsic variation in behavior measured in the heritability studies, and now specifically identified in genomic (GWAS) and candidate gene studies. Much behavioral variation exists because human societies built of people with varied cognitive strengths and personalities are more successful. Evolution has validated the value of diversity. People are both warriors and worriers, and fortunately the balance has sometimes worked, after seasons of violence and vengeance.
A common genetic variant “for” warriors and worriers One reason that many common genetic variants can be surprisingly easily tied to behavior is that these gene variations exist for the purpose of causing behavioral variation, and not just variation in some obscure biochemical indicator or mediating brain function. In other words, we can observe the behavioral effect and not just the effect on molecules or a subtle psychophysiological measure because the genetic variant exists to cause behavioral variation of the type we observe in everyday life. Evolution acts through biochemistry and neurocircuitry to alter behavior but does not select variants whose only effect is to alter these intermediates, though they are “closer” to the gene. The variant may exert a stronger effect on the biochemical intermediate phenotype, but it exists to alter behavior. Individual genes are selfish (Dawkins, 1976), but evolution is inherently a “top down” process, natural selection judging the fitness of a gene based on external outcomes, even if the basis is alteration of internal processes. Evolution “chooses” based on outcome, and it has repeatedly been shown that a variety of mutations, at the same gene and at different genes, can suffice to achieve the same outcome. By one mutation or another, bears, foxes, owls, birds, and rabbits that live for enough generations in polar regions are selected to be white. By one mutation or another, humans who live for enough generations in a conflict zone are more likely to become warrior-like. However, in human societies, and even in ones facing conflict (which is to say, most) there is always room for worriers, and in frequency-dependent fashion they are selected. A common genetic variation in catechol-O-methyltransferase (COMT), an enzyme that pleiotropically influences both cognition and emotional
Executive cognitive function
187
resilience, is maintained by the selective advantage of behavioral variation. This “warrior/worrier” polymorphism involves two alternative amino acids at position 158 in the protein sequence—Valine158 or Methionine158. The effects of the alleles counterbalance each other and as a result both alleles are maintained at high frequencies (from 40% to 60%) in populations worldwide. COMT was discovered by Nobel laureate Julius Axelrod—a genius of neurochemistry who began as a technician but ended up training and inspiring a generation of neuroscientists, including myself. I did not work with Axelrod directly, nor other NIH Nobel laureates such as Marshall Nirenberg, Harvey Levin, and Carleton Gajdusek who were contemporaneous, but as a postdoctoral fellow I came close to running over Axelrod over as he jaywalked preoccupied. Probably, no one would have believed that Axelrod “darted out in front of my car.” Another inspirational pioneer in neuroscience, Patricia Goldman-Rakic—whose work on dopamine’s role in the frontal cortex is also key to the COMT story, did die after being hit by a car, so perhaps people carrying genes predisposing them to be absentminded professors should stick to the lab. In the brain, the main function of COMT is to metabolize catecholamine neurotransmitters. COMT catalyzes one reaction by which the action of several neurotransmitter molecules is ended. In the frontal cortex, the action of dopamine is particularly dependent on COMT activity because the dopamine transport, another protein that clears dopamine from the synapse, is not active there. The COMT Met allele is less effective than the Val allele, leading to higher levels of dopamine in the brain’s frontal cortex, especially in people like myself (about one in six of the population), who have two copies of the Met allele (Met158/Met158 homozygotes). About two decades ago, a team that included me, members of my lab, and Daniel Weinberger’s lab at the National Institute of Mental Health found that people with this Met/Met genotype (all other things being equal!) tend to have better—slightly better—frontal cortical function than people with other genotypes (Egan et al., 2001). This, all other things being equal as they so seldom are, translates into better performance on cognitive tasks that are executed by the frontal lobe.
Executive cognitive function In combination with genotype, performance on any of these tasks might be used to predict behavior influenced by frontal lobe function, and when it comes to frontal lobe tasks we are spoiled for choice, as is also a frontal lobe task. The frontal cortex drives working memory, which is the ability to remember items for short intervals of time and is much like a computer’s write/erase cache memory being used in the “Now”
188
13. Warriors and worriers
and as contrasted with hippocampal episodic memory, which is like that sometimes more slowly retrievable computer disc memory of the “Then.” Working memory can be measured in a variety of ways, for example by having a person repeat digits forward or backward. Good performance on the digits reversed task might be seven digits—not easy! Another measure of working memory is the N-back task, in which the person is asked to recall a stimulus presented one event ago (1-back), two events ago (2-back), etc. As the task becomes more difficult, the metabolic activity of the frontal cortex intensifies, stress increases, and performance eventually declines. When the task is easy (for example the 1-back test) and everyone is performing the task accurately, a person with the Met/Met genotype still differs in an important way—showing lower metabolic activity, reflecting their higher cortical efficiency—Pat Goldman-Rakic’s conception. The frontal cortex is also the master switchboard for task m anagement— enabling people to apply the cognitive strategy appropriate to the situation—and it is thus the executive cognitive center of the brain. This latter ability is crucial to the capacity of people to make adaptive responses to their environment, so that if one approach is not working, they can switch to a different strategy instead of making perseverative errors. One executive cognitive strategy is to wait, and the frontal cortex helps people wait in a different way than the amygdala, a deep brain structure that can cause people to freeze in response to fear, a response that is “neuro”—as is everything the brain does, but, unlike frontal cognition, not especially cognitive. There are many of measures of executive cognitive function. Two popular ones are the Stroop Test and the Wisconsin Card Sort Test. The Stroop Test is often illustrated by speakers at scientific meetings, and in pre-COVID-19 times when scientists gathered in rooms and some whispered annoyingly, I always try to prepare my friends. It could be anticipated that a talk on executive function would, as if there was no choice in the matter, feature a slide such as this, with a request to the audience to call out what they see (Fig. 13.1). Most say, “Red.” A few mavericks say, “Blue.” But we exclaim, “Stroop test!”. The image can be classified as either word or color, but in many other ways besides. The test can be refined by measuring interference of word and color classification via response time, the word more greatly interfering with ability to name the color under some circumstances and for example if the word red is written in blue, or even by measuring the metabolic activity of the brain during the test. As just mentioned, the Stroop Test doesn’t measure it, but someone’s brain might also be processing the
5(' FIG. 13.1 What do you see? A test of frontal cognitive function.
Cognitive flexibility and free will
189
stimulus as a cognitive test and also in a myriad of other ways (photons!, word!, slide!, Courier font! and so on). Regardless of how well one task switches, or how strangely—and in this regard there is an important term in psychiatry called “loose association,” it is the frontal lobe that enables us to task-switch from one classification strategy to another, and to home in (not “hone in”) on context-relevant associations and ignore perfectly true but practically useless associations such as “word!”
Cognitive flexibility and free will Central to free will is the unique cognitive flexibility of humans. One person chooses to classify an image by color, and another as to whether it is an aspect of a test. Scientists who measure cognitive flexibility are usually themselves cognitively flexible but to access the way people think they use measures that tap into how people usually think, recognizing those usual responses. People with loose associations are outliers of cognitive flexibility, and to their disadvantage in many situations. To produce objective and reliable data on cognition across many different types of populations, we require simple, standardized tests. Faced with the task of designing experiments producing objective and reliable data on cognition across diverse populations, cognitive psychologists devised simple, standardized tests that, naturally or by constraint, will elicit a limited repertoire of responses, best only two, from most people. The day of the Rorschach test in which the test subject says whatever may come into their mind, is largely past. However, just because we know a person’s A/B response, we don’t know how they got there. They might have been stung by a “bee.” They might have decided to alternate between A and B. Under ordinary circumstances, when we try to guess what algorithm a person is running to solve a problem we often fail, and some people and choices are more difficult to decipher than others. Socially, we attempt prediction of human responses, but as with machine learning programs of the self-driving that can accurately predict that a person is about to cross the road, our mode of prediction can nevertheless be alien to the agent being predicted. Furthermore, we do not necessarily understand how we are predicting behavior of others, even if we have more self-understanding than the self-driving car. However, although we probably have little understanding of what algorithms a person is running to solve problems, we can at least determine what parts of the brain are activated during cognitive tasks, thus isolating the locations of the circuits involved. The amount of time required for the decision making and the sequence of involvement of brain regions tell us more. We can discover genes, neurotransmitters, and drugs that modulate these processes. We can define that some are
190
13. Warriors and worriers
better at performing a task than others and that two people who perform equally may be working at very different levels of intensity. This is seen, for example, when brain frontal metabolic activity is measured by functional magnetic resonance imaging (fMRI) during tasks that depend on frontal lobe function. Some people’s frontal cortices are more efficient or better “tuned” such that they require less metabolic effort reflecting less neural activity to perform a given task. We are all familiar with this concept. Humans have invented a variety of devices—levers, pulleys, wheels—to perform tasks with less effort, and natural selection has ergonomically perfected the body in countless ways. However, for any given characteristic—sensory, motor, cognitive, metabolic, immune reproductive, and more—some bodies are more efficient than others and such that in certain circumstances a task can be performed by some and not by others, or not repetitively performed with the same accuracy.. The same is true of cognitive tasks dependent on neural networks that are wired differently and that may function efficiently or less efficiently because of an individual’s unique mix of innate genetic programming, environment, and training. As discussed elsewhere in this book, the function of the brain is surprisingly flexible, practice triggers “use it” neuroplastic changes reinforcing synapses and cells enabling task performance and disuse leads to “lose it” neuroplastic changes. However, innate differences matter. In schizophrenia, cognition has been studied by leading neuropsychiatrists, including a consortium studying cognition in schizophrenia led by David Braff, a psychiatrist at UCLA, and a group at NIMH led by Daniel Weinberger and including Michael Egan and cognitive psychologist Terry Goldberg; these scientists found that patients with schizophrenia and their well siblings tend to perform worse on executive cognitive tasks executed by frontal lobe neurons. Later, Weinberger’s group showed that the frontal cortex of many of the poorer performers was less efficient and could be manipulated pharmacologically, for example by augmenting dopamine with amphetamine like drugs or an inhibitor of the dopamine-metabolizing enzyme, COMT. In a collaboration with my lab, it was found that a common polymorphism of COMT alters that set point, frontal cortical efficiency, performance on frontal cognitive tasks, and response to drugs that augment dopamine (Egan et al., 2001). Increasingly, and regardless of their genotype or neurofunctional indicators that might guide intervention, people attempt to boost cognitive performance with drugs. The drugs include everyday caffeine and nicotine; prescription stimulants such as amphetamine, methylphenidate, and modafinil; and illicit amphetamine cooked up in rogue labs a la “Breaking Bad.” Repeated use of any one of these drugs probably has consequences, as the brain adapts. Any daily user of these drugs can appreciate their addictive potential. But we have little understanding of
Cognitive flexibility and free will
191
what it means to be frontal cortical inefficient—why one person benefits more and another less, and—despite the identification of many genes, why one person is more liable to addiction is largely unknown. What is the brain doing that it appears to be inefficient or more liable to addiction? I would like to believe (and this is an admission that what I say next is speculative) that sometimes cortical “inefficiency” is a manifestation of alternative p atterns of cognition—including patterns less rigid and patterns more flexible—and that such alternative patterns may be invaluable under different circumstances than are represented by the psychologist’s test, or perhaps any test devised. Similarly, if we could “cure” liability to addiction we might sacrifice important behavioral traits such as extraversion/risk-taking or even empathy. If we eliminated schizophrenia, we might subtract a valuable part of human cognitive flexibility. In drugging cognitive function, it is not quite true that we are “meddling with forces we cannot possibly comprehend” but this is still a close approximation of the truth. For Our Genes, Our Choices this means that presently people who make life choices and those who would assist them still have to be guided more by general experience (e.g., amphetamine is addictive but can temporarily improve cognitive performance) than by specific genetic or neurocognitive indicators (e.g., COMT genotype, a polygenic score for cognition or addiction liability, readout from an fMRI scan during a frontal cognitive task). However, the science of cognition and addiction is rapidly expanding, creating the real possibility that drugs and even other interventions, including new drugs and devices, can be targeted to those who will benefit, expanding their range of meaningful choices. The difficulty of assessing cognitive flexibility is well illustrated by the inflexibility of some of the tasks that have been designed to test it. For example, let’s reconsider the Red version of the Stroop test. This is a fine illustration of the greater range of potential responses that are possible. For example, is there anything wrong with the response: one-syllable word? It is a generic answer, but so is the acceptable answer “blue.” The answer is that these responses may be a little weird or unusual, but they are intrinsically accurate, and in fact could be called for in certain situations. Failure to understand the true range of potential accurate responses is one way that we underestimate human intelligence and underestimate the decision making that our brains constantly perform. We covet cognitive flexibility, but do not do well at encouraging it or assessing it. Starting early in school, teachers may not only look for the “correct” answer, but at times punish a child who arrives at the answer by a different method. This was a cardinal error of “new math.” Or we want to assess the cognitive flexibility of a job applicant. Can they “think outside the box”? A typical question—the answer to which is easily found on job interview prep websites—is:
192
13. Warriors and worriers
Why are manholes round? One applicant sits silently, even though they know “playing possum” is not a viable strategy in a job interview. Another compliments the interviewer on their question and indicates enthusiasm for the general concept of thinking out of the box. This could be a good idea when interviewing for an “assistant” position. Next is the person who gives an expected out of the box answer, “Because discs don’t fall into manholes when turned this way or that as compared to squares” (as if a square was the only alternative shape). So, all well and good. This applicant is “creative.” However, the next interviewee answers, “Because round covers don’t fall in when turned this way or that and land pointy end-down on someone’s head.” Too much information, but nevertheless everything may be okay. The next says “Round covers can still fall into holes but at least they don’t have pointy corners.” However, what is to be done with the applicant with too many explanations? They speculate that round covers are easily rolled into place, suggest that human bodies are roughly cylindrical (the round peg/round hole theory), discs conserve iron and being lighter are cheaper to ship, discs fit whichever way one orients them, tradition. And so on. These responses illustrate the vast difference between humans and other species that might be asked the manhole question. Mostly, they just stare at you uncomprehendingly, like a cow at a passing train. How would we decipher the pathways by which such diverse, very human, responses are generated, or whether a response is “good” or “bad,” or a way of thinking that tends to generate on-target (practical) or off-target (whimsical) responses is to be desired, and as is likely to be context dependent. Each of these responses listed before, and others not, represents a choice, and it is here as elsewhere that we see the vast difference between humans and other species that are sometimes advanced as models for human behavior. Humans’ range of choice, and means of getting there, is vastly greater, more diverse, and top down and led by our frontal lobes is how those choices are made.
Perseveration Einstein claimed that the definition of insanity is doing something that does not work over and again and expecting a different result (second edition of a book?). Wrong, as usual. Tell that to Edison. Few inventors would succeed, and few species would survive, without a strong streak of perseveration. The hunting success of tigers and polar bears is less than 10%. No person had a year like Einstein in 1905. In this so-called miracle year, he published papers on the photoelectric effect (showing that light was both a wave and a particle), Brownian motion (mathematically
Perseveration
193
confirming that matter was composed of atoms and molecules), Special Relativity (showing that the speed of light is constant but everything else: mass, distance, and time are relative), and mass–energy equivalence (E = mc2). Coupled with brilliance, Einstein’s perseverance paid off—even as a 26-year-old clerk in the Swiss patent office who had recently had his dissertation rejected. Psychologists in total have never had a year such as Einstein’s 1905 but have discovered good ways of measuring the tendency to perseverate, and perseveration errors. The Wisconsin Card Sort Test measures executive cognitive flexibility in a different way than the Stroop Test—by perseveration. The test subject matches one of several cards to a target in a situation where more than one matching strategy is available. For example, and only slightly facetiously, suppose the test card is a halibut (a type of fish) and the test targets are a cat, a trawler, and a tossed salad. The person might match animal cards, “fishing cards,” or components of a dinner (most people not eating cats). Whichever matching strategy is used the person is informed the answer is wrong. People with frontal lobe damage perseverate, continuing to use the strategy that has been labeled incorrect on the next test in the series. In a remarkable finding that has been replicated by many others and in many types of populations: “normal controls,” patients with schizophrenia, siblings of schizophrenic patients, and head-injured patients, we found that people with the less efficient Val158/ Val158 genotype tended to make more perseverative “errors.” It is particularly interesting that patients with schizophrenia, well siblings of schizophrenia patients, and head-injured patients show the genotype effect on cognition, because all already tend to have deficits in frontal lobe function, as manifested by their ability to perform the Wisconsin Card Sort Test and other tasks that are performed by that brain region (Egan et al., 2001). Around the time we were doing these studies, one of my sons, Ariel, and I traveled to Eilat and Petra with Danny Weinberger, Danny’s son Collin, and his wife Leslie. Danny was and is perhaps the world’s premier “biological psychiatrist” and an expert on schizophrenia, and I had joined with him to make the first discoveries using “imaging genetics.” One key to schizophrenia is the function of the frontal lobe of the brain, and a major determinant of frontal lobe function is the neurotransmitter dopamine. Frequently, the function of the frontal lobe can be augmented by boosting frontal dopamine levels, as happens if one takes the drugs amphetamine or methylphenidate. Thus it is that methylphenidate is used in the treatment of attention deficit hyperactivity disorder (ADHD), and it often helps. While in Petra, a Jordanian taxicab driver illustrated how dopamine works in the frontal lobe, helping to explain why a deficit in activity of the COMT enzyme would augment dopamine levels and frontal function. It turns out that unlike most neurotransmitters, dopamine diffuses far from the synapse where it is released. It doesn’t have to be precisely aimed at its
194
13. Warriors and worriers
target across a narrow synaptic space. This is especially true in the frontal cortex, which largely lacks a dopamine transporter that takes up released dopamine so that the neurotransmitter can no longer bind neuronal receptors on either side of the synapse. So, one may say that the aim of the cell releasing the dopamine does not have to be very good. It will hit a target even if released anywhere in the general vicinity. Germane to all that science, I noticed a handgun on the front seat of the taxi. I asked the driver about his “conversation piece.” It turned out that he was a “retired” intelligence officer. To Danny’s further chagrin I remarked that this type of handgun was not very accurate. The driver stopped the cab, waved the gun, and said, “At this range it’s accurate.” Later, and after run-ins with Sami the guide, who destroyed my camera, camel races, and a scuba patrol in Egypt led by a former commando in the IDF we found ourselves back at the border hoping to reenter Israel. A well-armed Israeli soldier suspiciously eyed my diving knife and with a meaningful look asked Danny, “Are you traveling with each other?” “We are traveling against each other,” I corrected, and from then on that’s how Danny described the trip. Following our travels in the Middle East, Weinberger and I did not seem to work together as often as before. He left the NIH to found the Lieber Institute but before that his NIMH lab went on to make a very important additional discovery about the effect of the COMT polymorphism on cognition, which was that the cognitive advantage of the Met158/ Met158 genotype disappeared under some circumstances that often occur in everyday life. The reason is shown in Fig. 13.2 which conveys the model developed by Pat Goldman-Rakic. At a particular dopamine concentration frontal cognitive performance is maximal. Under ordinary conditions, people with the Met158/Met158 genotype tend to have higher dopamine levels and are thus closer to this cognitive optimum. However,
ŽŐŶŝƚŝǀĞWĞƌĨŽƌŵĂŶĐĞ
^ƚƌĞƐƐŽƌ^ƚŝŵƵůĂŶƚ
KDd sĂůϭϱϴ
KDd DĞƚϭϱϴ
&ƌŽŶƚĂůĚŽƉĂŵŝŶĞ
FIG. 13.2 COMT Val158 (Warrior) and Met158 (Worrier) variants, stress, and the inverted U effect of dopamine on frontal cognitive performance.
Worriers and warriors
195
the relationship between dopamine concentrations and cognitive performance is an inverted U-shaped curve. Too much of a good thing is a bad thing. When stressed or if they take amphetamine-like drugs that boost dopamine levels and improve cognition in Val158 carriers, the dopamine levels of Met158/Met158 homozygotes become excessive, and their performance deteriorates (Mattay et al., 2003). In the frontal cortex, as Goldman-Rakic predicted, dopamine tunes cortical efficiency, there being an inverted U-shaped curve of effect of dopamine, and with drugs, genotypes, and environmental stresses having the ability to alter that tuning.
Worriers and warriors Gandhi worried that an eye for an eye leaves the whole world blind. But fortunately natural selection has made some people worriers and others warriors. As we have just seen, Met158/Met158 homozygotes have a cognitive advantage, but a disadvantage in cognitive resiliency under stress. At about the same time, I and colleagues in my lab found the downside to the Met158/Met158 genotype. The basis of emotions are the circuits (Ledoux, 1998) and genes (Bevilacqua and Goldman, 2011) that mediate them. People with the Met158//Met158 genotype tend to be less resilient to emotion and pain (Zubieta et al., 2003). The lower pain threshold of Met158/Met158 homozygotes and their stronger emotional response to equally perceived pain are also observable, and can partly be understood, via brain imaging. Jon-Kar Zubieta, now at the University of Utah but then at the University of Michigan, had shown that pain threshold is predicted by the brain’s ability to release endorphin (a natural opioid). As shown in the p seudo-colored image later, opioid receptors are found throughout the brain but especially in regions important in pain perception, such as the thalamus. But as is directly relevant to the addictive potential of opioids, opioid receptors are also abundant throughout the limbic system, an evolutionarily ancient set of interconnected regions that generate emotions and responses to emotional stimuli. These limbic regions, all highly expressing high opioid receptors, include the amygdala, hippocampus, and cingulate cortex. Emotional regulation is crucial in the interpretation of pain—any person in emotional distress having a lower pain threshold. Giving emotional comfort to people in pain can ease their pain. Reciprocally, chronic pain directly leads to emotional problems, especially depression and anxiety. Blocking pain or treating pain at its source can ease emotional distress. To measure pain threshold, Zubieta infused hypertonic saline (salty water) into the jaw muscle via a needle, calibrating the rate of infusion so that the pain threshold can be accurately measured. People with the Met158/ Met158 genotype had lower pain thresholds and stronger emotional
196
13. Warriors and worriers
FIG. 13.3 Opioid receptors are found in brain regions involved in pain and emotion.Imaged by [11C]carfentanil positron emission tomography. Image courtesy of Jon-Kar Zubieta.
r esponses to the equivalent amount of pain. Via the [11C]-Carfentanil PET Zubieta and I found that people with this more pain-sensitive COMT genotype could release less endorphin following the painful stimulus (Zubieta et al., 2003). Perhaps the endorphin release of these worriers was already maximal just from being placed in the experimental situation (Fig. 13.3). Both warriors and worriers always had niches in human populations, and in hominid apes from whom humans evolved. Interestingly, the Val allele is the evolutionarily more ancient form of COMT, being found in other great apes. Therefore, from an evolutionary perspective, the Met158 allele is the derived latecomer. Having appeared in an ancient ancestor, Met158 spread throughout the world, such that both COMT alleles are abundant everywhere. COMT Val158Met is a gene “for” behavior, but the gene effect is small. All other things being equal, Met158/Met158 “worriers” are cognitively advantaged but less resilient to pain and emotional distress, Val158/ Val158 “warriors” are more resilient, and Val158/Met158 heterozygotes are somewhere between (see Fig. 13.4). Probably, natural selection for these genotypes is frequency dependent and contingent on experiences lived by individuals, but also kindreds and communities. If a clan lacked either warriors or worriers it was in trouble, and because ancestral social groups were small it was necessary that the genetic variants producing the alternative behavioral patterns would be relatively common. One hunted
197
Worriers and warriors
KDdsĂůϭϱϴ >Žǁ^ƚƌĞƐƐ KDdsĂůϭϱϴ ,ŝŐŚ^ƚƌĞƐƐ
ŽŐŶŝƚŝǀĞƉĞƌĨŽƌŵĂŶĐĞ
KDdDĞƚϭϱϴ
KDdDĞƚϭϱϴ
FIG. 13.4 Top: low stress—advantage goes to Met158/Met158 homozygous “worriers.” Bottom: high stress—advantage goes to Val158/Val158 homozygous “warriors.”
and another tilled the soil and kept the fire, and both needed to be good at what they did. Many evolutionary forces, from hybrid dysgenesis to the loss of favorable polygenic combinations and especially the Hamiltonian imperative to share resources with one’s blood relatives, thus promoting the transmission of one’s own genes, favor endogamy, the “choice” to mate within one’s community and clan. However, the maintenance of functionally advantageous diversity at genes such as COMT leads to exogamy: the “choice” to mate with an outsider. COMT—a gene that alters behavior—is one of a legion of genes at which diversity has been essential to the resilience of human populations. Those genes include ones whose variants can protect some fraction of the community when we are exposed to a new pathogen, such as HIV (e.g., the CCR5 gene). They may not directly alter behavior themselves, but the evolutionary need to maintain diversity at them has indirectly encouraged exogamy. Genes and mating behavior are entwined, but this can lead to both assortative and disassortative choices. Contrary to some match-making algorithms and team building strategies, people affiliate and thrive both because their interests, strengths, and natures coincide, and because they complement each other. For most human activities, diversity pays off. Yet the imperative to choose diversity, although genetically encode, will always compete with the even more powerfully encoded tendency to identify and favor kin. Via personal genomics, and going beyond family histories and the crude and largely biologically meaningless “racial” identifiers such as skin pigmentation, people increasingly pinpoint those with whom they are most closely related and quantify those identities. A lesson of genes such as COMT is that anyone engaged in such personal genomics should also consider that the genes are telling us that diversity is a virtue. The rare COMT “worrier” or individual with HIVresistant CCR5 genotype would be more likely to have the advantage.
198
13. Warriors and worriers
References Bevilacqua, L., Goldman, D., 2011. Genetics of emotion. Trends Cogn. Sci. 15, 401–408. https://doi.org/10.1016/j.tics.2011.07.009. Binder, E.B., Bradley, R.G., Liu, W., et al., 2008. Association of FKBP5 polymorphisms and childhood abuse with risk of posttraumatic stress disorder symptoms in adults. J. Am. Med. Asoc. 299, 1291–1305. Caspi, A., McClay, J., Moffitt, T.E., et al., 2002. Role of genotype in the cycle of violence in maltreated children. Science 297, 851–854. Caspi, A., Sugden, K., Moffitt, T.E., et al., 2003. Influence of life stress on depression: moderation by a polymorphism in the 5-HTT gene. Science 301, 386–389. Dawkins, R., 1976. The Selfish Gene. Egan, M.F., Goldberg, T.E., Kolachana, B.S., et al., 2001. The effect of COMT Val108/158Met geno type on frontal lobe function and risk for schizophrenia. Proc. Natl. Acad. Sci. U. S. A. 98, 6917–6922. Egan, M.F., Kojima, M., Callicott, J.H., et al., 2003. The BDNF Val66Met polymorphism affects activitydependent secretion of BDNF and human memory and hippocampal function. Cell 112, 257–269. Ledoux, J., 1998. The Emotional Brain: The Mysterious Underpinnings of Emotional Life. Mattay, V., Goldberg, T.E., Fera, F., et al., 2003. Catechol O-methyltransferase val158-met genotype and individual variation in the brain response to amphetamine. Proc. Natl. Acad. Sci. U. S. A. 100, 6186–6191. Zhou, Z., Zhu, G., Hariri, A.R., et al., 2008. Genetic variation in human NPY expression affects stress response and emotion. Nature 452, 997–1001. Zubieta, J.K., Heitzeg, M.M., Smith, Y.R., et al., 2003. COMT Val158Met genotype affects μ-opioid neurotransmitter responses to a pain stressor. Science 299, 1240–1243.
14 How many genes does it take to make a behavior? Three Rings for the Elven-kings under the sky, Seven for the Dwarf-lords in their halls of stone, Nine for Mortal Men doomed to die … J.R.R. Tolkien Genes, to the extent that they influence behavior, are determinants of behavior. Scorpion and wasps sting because gene constellations make them born to do so, and ladybugs do not. Do genes that predispose to behavior or that determine behavior act alone or in combination? Can one gene variant, like Sargon’s ring that ruled over the rings of man, dwarf, and elf, overwhelm the actions of others? More specifically, is one genetic predictor sufficient, or will they have to be used in polygenic combinations or in particular—which is to say, epistatic—combinations? In this chapter we will see that, variously, genes that alter complex behavior act both alone and in combination, but not often in epistatic combinations. Alone, because as units of selection, genes and their allelic variants are selected “for” behavioral phenotypes as well for other traits, but then leading to secondary effects on behaviors. Together, because all genes act in concert with others, and usually the additive and nonadditive (epistatic) effects of many genes lead to any inherited difference in behavior. It is the combined action of many genes and environmental influences that leads to the observation that behavioral traits are almost invariably inherited in non-Mendelian fashion. One gene does not make a behavior, and yet natural selection is able to “see” many genetic variants, eliminating them, bringing them to fixation, or maintaining them at high abundance. Although humans for millennia had performed artificial selection, breeding ever-more fruitful, tasty, and hardy plants and larger, tamer, fiercer, woolier, more colorful, and swifter animals, little was known about the nature of inheritance except that offspring tended to resemble their parents. Astoundingly, from time to time a new variety or hopeful monster would unexpectedly appear, the latter often being taken as an evil portent.
Our Genes, Our Choices https://doi.org/10.1016/B978-0-443-22161-3.00016-8
199
Copyright © 2024 David Goldman. Published by Elsevier Inc. All rights reserved.
200
14. How many genes does it take to make a behavior?
Even Darwin when he published the On the Origin of Species (1859) as an explicit “Theory of Natural Selection” was unaware of the patterns and basis of genetic transmission. The roles of genes and DNA would not be learned until the mid-20th century, but unknown to Darwin, who had himself only carried out some rudimentary experiments on corn, foundations of modern genetics were already being laid. Gregor Mendel, a Czech abbot, who in other hours of the day was a mathematician, meteorologist, and biologist, presenting his work in 1865 and publishing it a year later, to little notice. Mendel’s experiments, rediscovered only at the turn of the century, were exhaustive and exacting, and by their nature revealed that he was perhaps already a Mendelian, having become aware of patterns of inheritance by less formal observations. Mendel crossed peas differing in dichotomous traits such as large/small, yellow/green, and round/wrinkled, establishing rules of heredity for monogenic traits. Mendel’s rules were (1) the principle of uniformity—like being inherited from like; (2) the principle of segregation— recessive traits absent in the first-generation offspring of a cross reappearing in the second generation when two copies of the recessive allele again segregated together in predictable proportions; and (3) the principle of independent assortment such that the proportions of offspring sharing two different traits could be computed by multiplying their independent probabilities. In nature, and probably never more than in behavior, each one of Mendel’s rules is violated in important ways. Violating the rule of uniformity, mutations are observed, regression to the mean and extreme outliers occur for polygenic traits such as cognitive ability, and parental imprinting effects are occasionally observed as observed for Prader–Willi/Angelman syndromes, where two perfectly normal genes lead to distinct behavioral diseases with distinct phenotypes because of the mere fact that both gene copies are inherited from the same parent. Violating the rule of segregation, most characteristics are not Mendelian in nature—instead mating black and white is more likely to yield gray or brown than black or white. Regarding the law of independent assortment, Mendel’s seven phenotypes all happened to be caused by genes that are not in physical proximity, and that therefore did assort independently. In 1911, Thomas Hunt Morgan would discover pairs of phenotypes in fruit flies that while otherwise Mendelian in nature transmitted to the next generation nonindependently, and as if coupled together or in repulsion. These violations were the basis for the construction of the first linkage maps of genomes, showing how genes must align, and that later led to the ability to map genes influencing phenotypes, including behavioral characteristics of many kinds. None of this is a widely held criticism, and deviations from Mendelian transmission are not mentioned here to detract from Mendel’s monumental contribution. Genetics has advanced via discovery of the basis for each of these exceptions. Rather, criticism of Mendel, and again by his admirers, has centered on the proportions of phenotypes he observed in his crosses. Across Mendel’s experiments, these proportions may be too close
Single genes
201
to perfect to be accounted for by chance. This critique was first made in 1936 by the peerless mathematical geneticist R.A. Fisher who while praising Mendel, called the data “abominable” and “cooked.” Fisher calculated that the likelihood that the data fit as well as it did across many independent experiments was WZ &