Behave: The Biology of Humans at Our Best and Worst

by Robert M. Sapolsky

  So what does the extent of exposure (as assessed by the ratio) predict about adult behavior? Men with more “masculine” 2D:4D ratios tend toward higher levels of aggression and math scores; more assertive personalities; higher rates of ADHD and autism (diseases with strong male biases); and decreased risk of depression and anxiety (disorders with a female skew). The faces and handwriting of such men are judged to be more “masculine.” Furthermore, some reports show a decreased likelihood of being gay.

  Women having a more “feminine” ratio have less chance of autism and more of anorexia (a female-biased disease). They’re less likely to be left-handed (a male-skewed trait). Moreover, they exhibit less athletic ability and more attraction to highly masculine faces. And they’re more likely to be straight or, if lesbian, more likely to take stereotypical female sexual roles.72

  This constitutes some of the strongest evidence that (a) fetal androgen exposure has organizational effects on adult behavior in humans as in other species, and (b) that individual differences in the extent of such exposure predict individual differences in adult behavior.*73 Prenatal endocrine environment is destiny.

  Well, not exactly. These effects are small and variable, producing a meaningful relationship only when considering large numbers of individuals. Do testosterone’s organizational effects determine the quality and/or quantity of aggression? No. How about the organizational plus the activational effects? Not those either.

  Expanding the Scope of “Environment”

  Thus the fetal brain can be influenced by hormones secreted by the fetus. But in addition, the outside world alters a pregnant woman’s physiology, which in turn affects the fetal brain.

  The most obvious version of this is how food ingested by a pregnant female influences what nutrients are delivered to the fetal circulation.* At an extreme, maternal malnutrition broadly impairs fetal brain development.*74 Moreover, pathogens acquired by the mother can be passed to the fetus—for example, the protozoan parasite Toxoplasma gondii can infect someone pregnant (typically after exposure to infected cat feces) and eventually reach the fetal nervous system, potentially wreaking serious havoc. And this is also the world of maternal substance abuse producing heroin and crack babies or fetal alcohol syndrome.

  Importantly, maternal stress impacts fetal development. There are indirect routes—for example, stressed people consume less healthy diets and consume more substances of abuse. More directly, stress alters maternal blood pressure and immune defenses, which impact a fetus. Most important, stressed mothers secrete glucocorticoids, which enter fetal circulation and basically have the same bad consequences as in stressed infants and children.

  Glucocorticoids accomplish this through organizational effects on fetal brain construction and decreasing levels of growth factors, numbers of neurons and synapses, and so on. Just as prenatal testosterone exposure generates an adult brain that is more sensitive to environmental triggers of aggression, excessive prenatal glucocorticoid exposure produces an adult brain more sensitive to environmental triggers of depression and anxiety.

  In addition, prenatal glucocorticoid exposure has effects that blend classical developmental biology with molecular biology. To appreciate this, here’s a highly simplified version of the next chapter’s focus on genes: (a) each gene specifies the production of a specific type of protein; (b) a gene has to be “activated” for the protein to be produced and “deactivated” to stop producing it—thus genes come with on/off switches; (c) every cell in our bodies contains the same library of genes; (d) during development, the pattern of which genes are activated determines which cells turn into nose, which into toes, and so on; (e) forever after, nose, toes, and other cells retain distinctive patterns of gene activation.

  Chapter 4 discussed how some hormones have activational effects by altering on/off switches on particular genes (e.g., testosterone-activating genes related to increased growth in muscle cells). The field of “epigenetics” concerns how some hormonal organizational effects arise from permanently turning particular genes on or off in particular cells.75 Plenty more on this in the next chapter.

  This helps explain why your toes and nose work differently. More important, epigenetic changes also occur in the brain.

  This domain of epigenetics was uncovered in a landmark 2004 study by Meaney and colleagues, one of the most cited papers published in the prestigious journal Nature Neuroscience. They had shown previously that offspring of more “attentive” rat mothers (those that frequently nurse, groom, and lick their pups) become adults with lower glucocorticoid levels, less anxiety, better learning, and delayed brain aging. The paper showed that these changes were epigenetic—that mothering style altered the on/off switch in a gene relevant to the brain’s stress response.* Whoa—mothering style alters gene regulation in pups’ brains. Remarkably, Meaney, along with Darlene Francis of the University of California, Berkeley, then showed that such rat pups, as adults, are more attentive mothers—passing this trait epigenetically to the next generation.* Thus, adult behavior produces persistent molecular brain changes in offspring, “programming” them to be likely to replicate that distinctive behavior in adulthood.76

  More findings flooded in, many provided by Meaney, his collaborator Moshe Szyf, also of McGill, and Frances Champagne of Columbia University.77 Hormonal responses to various fetal and childhood experiences have epigenetic effects on genes related to the growth factor BDNF, to the vasopressin and oxytocin system, and to estrogen sensitivity. These effects are pertinent to adult cognition, personality, emotionality, and psychiatric health. Childhood abuse, for example, causes epigenetic changes in hundreds of genes in the human hippocampus. Moreover, Stephen Suomi of the National Institutes of Health and Szyf found that mothering style in monkeys has epigenetic effects on more than a thousand frontocortical genes.*

  This is totally revolutionary. Sort of. Which segues to a chapter summary.

  CONCLUSIONS

  Epigenetic environmental effects on the developing brain are hugely exciting. Nonetheless, curbing of enthusiasm is needed. Findings have been overinterpreted, and as more researchers flock to the subject, the quality of studies has declined. Moreover, there is the temptation to conclude that epigenetics explains “everything,” whatever that might be; most effects of childhood experience on adult outcomes probably don’t involve epigenetics and (stay tuned) most epigenetic changes are transient. Particularly strong criticisms come from molecular geneticists rather than behavioral scientists (who generally embrace the topic); some of the negativity from the former, I suspect, is fueled by the indignity of having to incorporate the likes of rat mothers licking their pups into their beautiful world of gene regulation.

  But the excitement should be restrained on a deeper level, one relevant to the entire chapter. Stimulating environments, harsh parents, good neighborhoods, uninspiring teachers, optimal diets—all alter genes in the brain. Wow. And not that long ago the revolution was about how environment and experience change the excitability of synapses, their number, neuronal circuits, even the number of neurons. Whoa. And earlier the revolution was about how environment and experience can change the sizes of different parts of the brain. Amazing.

  But none of this is truly amazing. Because things must work these ways. While little in childhood determines an adult behavior, virtually everything in childhood changes propensities toward some adult behavior. Freud, Bowlby, Harlow, Meaney, from their differing perspectives, all make the same fundamental and once-revolutionary point: childhood matters. All that the likes of growth factors, on/off switches, and rates of myelination do is provide insights into the innards of that fact.

  Such insight is plenty useful. It shows the steps linking childhood point A to adult point Z. It shows how parents can produce offspring whose behaviors resemble their own. It identifies Achilles’ heels that explain how childhood adversity can make for damaged and damaging adults. And it hints at how bad outcomes might be revers
ed and good outcomes reinforced.

  There is another use. In chapter 2 I recounted how it required the demonstration of hippocampal volume loss in combat vets with PTSD to finally convince many in power that the disorder is “real.” Similarly, it shouldn’t require molecular genetics or neuroendocrinology factoids to prove that childhood matters and thus that it profoundly matters to provide childhoods filled with good health and safety, love and nurturance and opportunity. But insofar as it seems to require precisely that sort of scientific validation at times, more power to those factoids.

  Eight

  Back to When You Were Just a Fertilized Egg

  I’m reminded of a cartoon where one lab-coated scientist is telling the other, “You know how you’re on the phone, and the other person wants to get off but won’t say it, so they say, ‘Well, you probably need to get going,’ like you’re the one who wants to get off, when it’s really them? I think I found the gene for that.”

  This chapter is about progress in finding “the gene for that.”

  —

  Our prototypical behavior has occurred. How was it influenced by events when the egg and sperm that formed that person joined, creating their genome—the chromosomes, the sequences of DNA—destined to be duplicated in every cell in that future person’s body? What role did those genes play in causing that behavior?

  Genes are relevant to, say, aggression, which is why we’re less alarmed if a toddler pulls at the ears of a basset hound rather than a pit bull. Genes are relevant to everything in this book. Many neurotransmitters and hormones are coded for by genes. As are molecules that construct or degrade those messengers, as are their receptors. Ditto for growth factors guiding brain plasticity. Genes typically come in different versions; we each consist of an individuated orchestration of the different versions of our approximately twenty thousand genes.

  This topic carries two burdens. The first reflects many people being troubled by linking genes with behavior—in one incident from my academic youth, a federally funded conference was canceled for suggesting that genes were pertinent to violence. This suspicion of gene/behavior links exists because of the pseudoscientific genetics used to justify various “isms,” prejudice, and discrimination. Such pseudoscience has fostered racism and sexism, birthed eugenics and forced sterilizations, allowed scientifically meaningless versions of words like “innate” to justify the neglect of have-nots. And monstrous distortions of genetics have fueled those who lynch, ethnically cleanse, or march children into gas chambers.*1

  But studying the genetics of behavior also carries the opposite burden of people who are overly enthusiastic about the subject. After all, this is the genomics era, with personalized genomic medicine, people getting their genomes sequenced, and popular writing about genomics giddy with terms like “the holy grail” and “the code of codes.” In a reductionist view, understanding something complex requires breaking it down into its components; understand those parts, add them together, and you’ll understand the big picture. And in this reductionist world, to understand cells, organs, bodies, and behavior, the best constituent part to study is genes.

  Overenthusiasm for genes can reflect a sense that people possess an immutable, distinctive essence (although essentialism predates genomics). Consider a study concerning “moral spillover” based on kinship.2 Suppose a person harmed people two generations ago; are this person’s grandchildren obliged to help his victims’ grandchildren? Subjects viewed a biological grandchild as more obligated than one adopted into the family at birth; the biological relationship carried a taint. Moreover, subjects were more willing to jail two long-lost identical twins for a crime committed by one of them than to jail two unrelated but perfect look-alikes—the former, raised in different environments, share a moral taint because of their identical genes. People see essentialism embedded in bloodlines—i.e., genes.*

  This chapter threads between these two extremes, concluding that while genes are important to this book’s concerns, they’re far less so than often thought. The chapter first introduces gene function and regulation, showing the limits of genes’ power. Next it examines genetic influences on behavior in general. Finally we’ll examine genetic influences on our best and worst behaviors.

  PART I: GENES FROM THE BOTTOM UP

  We start by considering the limited power of genes. If you are shaky about topics such as the central dogma (DNA codes for RNA, which codes for protein sequence), protein structure determining function, the three-nucleotide codon code, or the basics of point, insertion, and deletion mutations, first read the primer in appendix 3.

  Do Genes Know What They Are Doing? The Triumph of the Environment

  So genes specify protein structure, shape, and function. And since proteins do virtually everything, this makes DNA the holy grail of life. But no—genes don’t “decide” when a new protein is made.

  Dogma was that there’d be a stretch of DNA in a chromosome, constituting a single gene, followed by a stop codon, followed immediately by the next gene, and then the next. . . . But genes don’t actually come one after another—not all DNA constitutes genes. Instead there are stretches of DNA between genes that are noncoding, that are not “transcribed.”* And now a flabbergasting number—95 percent of DNA is noncoding. Ninety-five percent.

  What is that 95 percent? Some is junk—remnants of pseudogenes inactivated by evolution.*3 But buried in that are the keys to the kingdom, the instruction manual for when to transcribe particular genes, the on/off switches for gene transcription. A gene doesn’t “decide” when to be photocopied into RNA, to generate its protein. Instead, before the start of the stretch of DNA coding for that gene is a short stretch called a promoter*—the “on” switch. What turns the promoter switch on? Something called a transcription factor (TF) binds to the promoter. This causes the recruitment of enzymes that transcribe the gene into RNA. Meanwhile, other transcription factors deactivate genes.

  This is huge. Saying that a gene “decides” when it is transcribed* is like saying that a recipe decides when a cake is baked.

  Thus transcription factors regulate genes. What regulates transcription factors? The answer devastates the concept of genetic determinism: the environment.

  To start unexcitingly, “environment” can mean intracellular environment. Suppose a hardworking neuron is low on energy. This state activates a particular transcription factor, which binds to a specific promoter, which activates the next gene in line (the “downstream” gene). This gene codes for a glucose transporter; more glucose transporter proteins are made and inserted into the cell membrane, improving the neuron’s ability to access circulating glucose.

  Next consider “environment,” including the neuron next door, which releases serotonin onto the neuron in question. Suppose less serotonin has been released lately. Sentinel transcription factors in dendritic spines sense this, travel to the DNA, and bind to the promoter upstream of the serotonin receptor gene. More receptor is made and placed in the dendritic spines, and they become more sensitive to the faint serotonin signal.

  Sometimes “environment” can be far-flung within an organism. A male secretes testosterone, which travels through the bloodstream and binds to androgen receptors in muscle cells. This activates a transcription-factor cascade that results in more intracellular scaffolding proteins, enlarging the cell (i.e., muscle mass increases).

  Finally, and most important, there is “environment,” meaning the outside world. A female smells her newborn, meaning that odorant molecules that floated off the baby bind to receptors in her nose. The receptors activate and (many steps later in the hypothalamus) a transcription factor activates, leading to the production of more oxytocin. Once secreted, the oxytocin causes milk letdown. Genes are not the deterministic holy grail if they can be regulated by the smell of a baby’s tushy. Genes are regulated by all the incarnations of environment.

  In other words, genes don’t make sense outside the
context of environment. Promoters and transcription factor introduce if/then clauses: “If you smell your baby, then activate the oxytocin gene.”

  Now the plot thickens.

  There are multiple types of transcription factors in a cell, each binding to a particular DNA sequence constituting a particular promoter.

  Consider a genome containing one gene. In that imaginary organism there is only a single profile of transcription (i.e., the gene is transcribed), requiring only one transcription factor.

  Now consider a genome consisting of genes A and B, meaning three different transcription profiles—A is transcribed, B is transcribed, A and B are transcribed—requiring three different TFs (assuming you activate only one at a time).

  Three genes, seven transcription profiles: A, B, C, A + B, A + C, B + C, A + B + C. Seven different TFs.

  Four genes, fifteen profiles. Five genes, thirty-one profiles.*

  As the number of genes in a genome increases, the number of possible expression profiles increases exponentially. As does the number of TFs needed to produce those profiles.

  Now another wrinkle that, in the lingo of an ancient generation, will blow your mind.

  TFs are usually proteins, coded for by genes. Back to genes A and B. To fully exploit them, you need the TF that activates gene A, and the TF that activates gene B, and the TF that activates genes A and B. Thus there must exist three more genes, each coding for one of those TFs. Requiring TFs that activate those genes. And TFs for the genes coding for those TFs . . .

  Whoa. Genomes aren’t infinite; instead TFs regulate one another’s transcription, solving that pesky infinity problem. Importantly, across the species whose genomes have been sequenced, the longer the genome (i.e., roughly the more genes there are), the greater the percentage of genes coding for TFs. In other words, the more genomically complex the organism, the larger the percentage of the genome devoted to gene regulation by the environment.