The Psychopath Inside (9 page)

Normally, amygdala-activated stress will, in turn, end up activating the serotonin-producing neurons in the brain stem. So acute anxiety and stress may quickly lead to release of serotonin, ultimately offsetting the stress. When a normal person is brought to anger, they will soon relax in response to serotonin, which turns off the stress loop. In someone who is an impulsive hothead, the stressor that produces an angry response may not be turned off by serotonin if the brain areas normally responding to
serotonin have low function, or if not enough serotonin is released. One's reactions depend on the interaction of all of the serotonin-related chemicals (the transporter, MAO-A, serotonin receptor subtypes, enzymes) with limbic areas. For some, the anger response may go on for hours, not just a couple of minutes. Given all these variables, plus differences in the early development of the limbic structures, all affected by genetics and maternal stress, it is easy to see why there are so many types of stress and anger responses in different people. In psychopaths, subtypes will also exist, but with their underfunctioning amygdala and orbital/ventromedial and cingulate cortices, there is often little stress and anxiety to begin with.

The genes that regulate empathy, and are therefore of great interest in the understanding of the cause of psychopathy, include a range of alleles that affect the function of the hormones oxytocin and vasopressin in the brain. Oxytocin reduces the amygdala's fear response in social situations and allows for trust. It's released in high concentrations during childbirth, nursing, and sex, particularly in women. Vasopressin allows for pair bonding, particularly in men. Voles that have vasopressin receptors in the reward centers of their brains become monogamous. Studies in the laboratories of Elizabeth Hammock and Larry Young at Emory University, Bhismadev Chakrabarti and Simon Baron-Cohen (cousin of comedian Sacha Baron Cohen) at the Universities of Reading and Cambridge, Thomas Insel at the National Institute of Mental Health, Dacher Keltner's and Sarina Rodrigues's group at the University of California, Berkeley, and Paul Zak at the Claremont
Graduate University throughout the period of 2005 to 2010 began to show the role of these alleles in empathy. Paul Zak has recently shown that testosterone receptor genes also affect generosity and empathy.

The empathy and aggression traits and their associated gene alleles have shown some promise for understanding psychopathy, but genes have not been identified that strongly predict other important traits in psychopaths, such as grandiosity, glibness, pathological lying, and lack of morals and ethics. The route to understanding these traits will likely come first through analysis of brain anatomy and alterations in connections in the brain (what features of the brain are involved in these functions?), with the genetic information (which genes affect these features?) coming later.

By 2007 to 2009, researchers began to understand that most complex adaptive behaviors in humans were probably influenced not just by one gene, but by dozens. One lab might find one gene associated with schizophrenia and publish their result. Another lab might find another gene. Other labs would try to replicate these findings but not achieve statistical significance. It was frustrating, and people wondered if a lot of bad data was being published. It wasn't until researchers started pulling together large groups of subjects that they realized there weren't one or two genes responsible but maybe fifteen or twenty, each contributing a few percent of the variance in symptoms.

Most of the genes couldn't be characterized as dominant or recessive, unlike those controlling eye and hair color, for example.
Complex behaviors, which are controlled by numerous interacting genes, are also under the influence of a myriad of regulators of these genes. Aggression, for instance, is a complex behavior, influenced by the interaction of genes regulating serotonin, norepinephrine, dopamine, androgen, and a host of other cellular functions. No single gene can be flipped on or off to decide someone's fate as violent or passive. So
dominant
and
recessive
are not in the everyday lexicon when discussing these behavior-modifying genes and their regulators.

Given the complexity of these genetic systems, any conclusion drawn from determining whether I had the warrior gene (which a thorough genetic analysis would show) would be inconclusive. It would take a whole set of bad genes to faze me.

And I wasn't in a rush to learn about them. Diane's parents had both died of Alzheimer's, so I'd initially been worried about Alzheimer's genes in her and in our kids. But their brains looked healthy and they showed no signs of cognitive decline, so getting our family's genetic results no longer seemed urgent. And learning of my violent family lineage was not enough to supplant the sense of urgency. Two months earlier, I'd been intrigued by my odd brain scan, but the novelty had quickly worn off. Plus I was busy with other projects. Not only had I been giving talks about serial killers, but I was also in the middle of getting two biotech companies off the ground and analyzing brain scans and genetic patterns as part of separate studies of Alzheimer's disease and schizophrenia.

This same time period of 2006 through 2009 was also
extraordinarily exciting for our imaging genetics studies. We were hot on the trail of discovering two new genes for schizophrenia and a new one for Alzheimer's, and in developing a whole new method for gene discovery itself. This method shortened, by two orders of magnitude, the time and expense of discovering genetic alleles associated with diseases, especially for the complex diseases and disorders of the mind we encounter in psychiatry. Typically, to locate a gene associated with a disease such as schizophrenia, you'd need something like three thousand subjects. Some of them would have symptoms of schizophrenia and some wouldn't, and a particular gene might be more common among those who do have symptoms. We created a statistical method that uses a set of equations to compare a subject's genetic information with brain-imaging data and psychological testing. This way, you'd need only three hundred, or even thirty, subjects to identify a candidate gene. The technique could also determine which patients would best respond to a drug or other treatment for schizophrenia, Parkinson's disease, or depression, and also quickly determine those patients most likely to have debilitating side effects of these treatments. Instead of a person being a guinea pig receiving different drugs one after another for six months, you'd get to the right one immediately, reducing suffering as well as medical costs.

The imaging genetics lab group of Steven Potkin, Fabio Macciardi, David Keator, Jessica Turner, and several high-end technicians and collaborators across the entire UC Irvine campus was firing on all cylinders at this time, and the drive and need to get papers and grants and patents and talks completed weighed
heavily on everyone's minds. Amid this maelstrom of scientific delights, it was all too easy to forget, at least temporarily, about my scan.

And testing for the warrior genes and other aggression-related genes is not a simple process. For instance, there's the GWAS (genome-wide association study), which samples from a few million single nucleotide polymorphisms (SNPs), whose proximate location to genes imply them in the etiology of traits and diseases. This technique is well standardized and relatively inexpensive. But there are significant limitations to GWAS in several important respects; there are more than three billion base pairs total per chromosome (six billion per chromosome pair), so there is less than 1 percent coverage of the genome, even considering SNPs as “proxies” for other DNA variants. Much can be missed with even the best GWAS sampling.

The only way to cover all of a person's genetic code is to do deep, whole genome sequencing of all three billion base pairs, plus sequencing of an alphabet soup of other elements. The cost of such analyses has dropped from hundreds of millions of dollars to several thousand per genome, and there are advertised rates of analysis less than a thousand dollars per person. However, this cost is highly misleading, as that low price comes with no real further analyses, just a listing of the sequences. This is like someone with an understanding of only English being handed a phone book a thousand pages in length written in Mongolian graphemes and Navajo syntax. For several thousand dollars there are commercially available analyses that translate the codes but not the syntax.
This interpretation of the meaning of the codes must still be done by a highly experienced team, including a geneticist, statistician, epidemiologist, cell biologist, and domain clinician (for example, a psychiatrist, cardiologist, or immunologist). This is where the real costs are hidden in full genetic analyses.

The time is ripe for application of sequencing projects integrated with cognitive, metabolic, and brain imaging methods to investigate complex human traits in disorders such as psychopathology in a way that was never possible before. Successful examples exist for complex diseases never previously understood through “omics” medicine, including genomics (genes and related nucleic acids in the nucleus), transcriptomics (various mRNA levels in tissues), proteomics (different levels of proteins and their interactions in relevant tissues), and metabolomics (blood and urine levels of several thousand hormones, metabolites, sugar, etc., and their dynamic interactions over time). Personal genomics and personalized medicine emerge as new feasible applications and not only as future possibilities, thanks to the developments in analyzing genomes and complex traits and visualizing these results into a unified framework.

In any case, it would be a while before someone took the time to give my genes a good looking-over. In the years between sending my blood sample and receiving my genetic results, I did consider now and again what the genetics results might tell me about what my brain scans meant. But I was not concerned about following in the footsteps of Thomas Cornell.

S
o I had the brain of a psychopath. And I had the family history and possibly the genes of a psychopath. And yet I had turned out very different from the serial killers I'd been studying. Something wasn't lining up, and that makes a scientist look for answers.

Although the loss of specific brain function in my limbic areas was in agreement with the neural profile of psychopaths from my lab and others, I noticed over the following year that there were individual case reports of people with such brain damage who were not murderers or psychopaths. This suggested to me that although the specific type of brain damage or functional loss might be
necessary
to cause psychopathy, it may not be
sufficient
to cause it. Other factors must be present.

Even if my DNA looked dangerous, that wouldn't be enough to turn me to the dark side, either. There was still no real link between genes and psychopathy, only between the MAOA allele and the potential for violence. I looked at all the case studies I could find in the literature and in my work, and saw that for all the psychopaths, including dictators, who had psychiatric reports from their youth, all had been abused and often had lost one or
more of their biological parents. While there may be cases where this is not true, I could not find any proven ones. There were cases where the murderer denied early abuse, but many people will deny such abuse, only for it to be discovered later that either they were too embarrassed to admit it, or they were protecting the abusing adult, typically a family member.

It was also becoming known from many studies that there was a high incidence of early childhood physical, emotional, or sexual abuse in the prison population of psychopaths. A small survey of thirty-five psychopathic offenders in youth detention facilities found that 70 percent reported serious mistreatment throughout childhood. Given that the onset of reliable memory for childhood events in adults may reach back to three to four years of age, this implied that a higher percentage of adult criminal psychopaths actually experienced significant abuse earlier than that. As such, it was possible that more than 90 percent of them were abused at some point in their early life. Add to this those psychopaths who protect their abusers, and the percentage could approach 99 percent, or so I reasoned. This was when I first started to consider
why
I might not be a full-blown psychopath. The killers had been abused, and I had not. Despite my hard-line conviction that we are shaped by nature and not nurture, I began to think that upbringing might play a significant role in creating a criminal after all.

•   •   •

The environment can interact with genes during development in a number of ways. One of those is through what's called a genotype-environment correlation. A child with genes
predisposing him to aggression may frequently misbehave, drawing hostility and abuse from his caretaker. Or an aggressive parent may pass along genes for hostility and also behave in a belligerent way toward his kids, and then both the genes and the antisocial attitude continue down the line. Such a pattern could explain my murderous lines of ancestors. Even if genes for aggression washed out over the generations, an expectation that families always act like this could have remained.

Another form of gene-environment interaction is what's known as epigenetic marking. Seemingly out of nowhere, your teenage daughter, who doesn't have the svelte shape of you or even your mother, starts to put on weight and looks very much like your grandmother, her great-grandmother. To figure out why, you all decide to take standard DNA tests to determine your respective genetic codes. But it turns out that the DNA code controlling the appetite and obesity of your plumping daughter is more similar to the DNA code of yourself and your lean mother than to that of your fleshy grandmother. So the genetics don't seem to explain your daughter's teenage-onset obesity. And she doesn't eat much more than an average person. Something else unexpected must be going on. Perhaps her metabolism is malfunctioning. But how and why? Then your niece, who is studying genetics in her doctoral work, suggests something may have been passed down from great-grandmother to grandmother to you, and then to your daughter. That something is not the genetic code itself, but a small extra bit, or tag, of chemical information stuck on to several genes controlling obesity and metabolism.

The added tag, called an epigenetic tag, might have been added on to several of her great-grandmother's genes while she was a young child, enduring starvation during a decade-long famine in Ireland, Poland, Bosnia, or the Bronx, nearly a century ago. Her great-grandmother's cellular response to the great stresses of that starvation may have been to change her metabolic machinery to more efficiently use energy and store fat, and to increase appetite once food was plentiful again. So your daughter, her great-granddaughter, under other teenage stressors, and with a plentiful food supply, responded by putting on weight to the point where she now resembles the plump but hearty teenager her great-grandmother became when the famine ended in her homeland eighty years ago. Some of these effects are dependent on whether the ancestor was male or female, since certain genes are “imprinted” on either the paternal or maternal side of the family.

The epigenetic tag is one of many alterations to the genetic code that can be induced by environmental stressors. This is one of the core mechanisms underlying the interaction of nature and nurture.

While there have been numerous recent studies on the role of epigenetic interactions on metabolism, cancer, and susceptibility to infectious and immune diseases, it is also a key to understanding some psychiatric disorders, from schizophrenia to psychopathy. One of my favorite scenes from the 1968 film
Charly
that so affected my career choice is the one in which the cognitively awakened title character goes to his teacher/therapist's
chalkboard and writes, “that that is is that that is not is not is that it it is,” and asks her what it says. She is unable to decipher the quip, and then he goes to the board and punctuates it into: “That that is, is. That that is not, is not. Is that it? It is.”

This puzzle offers a good analogy for what the epigenome does. The raw DNA base pair code in this analogy is “thatthatisisthatthatisnotisnotisthatititis,” and the way this raw sequence is laid out directs the code to be transcribed into the sequence of words but not quite a sentence. Normally the transcribed message from the DNA to the RNA would be translated into the protein, here the mature and sensical sentences “That that is, is. That that is not, is not. Is that it? It is.” But environmental stressors can induce epigenetic tags to be added on to some of the original genetic DNA, so that the punctuation, the spacing of the words, the text formatting in general, can be altered to produce a slightly different meaning: “That that is, is. That that is not, is not. Is that it? It is?” Same words, same sequence, but a final question mark added changes the thrust of the message. This slight “epigenetic” change to the sentence's intended “genetic” meaning is different from an actual mutation. In a mutation, the actual spelling of the sentence is changed, either by inserting a letter (or more) or deleting an existing letter. Such a change can, of course, radically alter the function of the sentence, which may now become, “That that is, is. That that is not, is snot. Is that it? It is.” In a similar sense, the genome is the book you inherited at birth, the epigenome is the way you read that book.

Another way of looking at the epigenome function is to
consider the new car you buy from the dealer. All that original hardware is like your genome, while alterations you might make to soup it up, give it some more pep, or, for your daughter, slow it down, are like the epigenetic modifications.

Epigenetic alterations are one of several reasons why identical twins are not identical. Even with identical raw genetic codes, differences in early environment, whether overly stressful or more positively enriching, can change their behaviors down the line as teenagers and adults. Identical twins can also have different numbers of the same genes inherited from one parent or another, and this can also alter how the identical twins look and behave. A third mechanism can involve a seemingly otherworldly phenomenon caused by “retrotransposons.”

Retrotransposons are short bits of DNA or RNA present in the nucleus of the cell surrounding the genes themselves. Once thought to be junk DNA with no apparent purpose, these odd snippets of information are not fixed in place but can move around, like grains of rice in soup. They are capable of connecting widely separate genes, even on different chromosomes, and they can alter cellular function. They can rearrange the “sentences” our DNA types out, and in doing so can ultimately change, usually subtly, human behavior, and account for not only differences in how identical twins act, but also what makes schizophrenics psychotic and perhaps why certain depressives become suicidal.

One of the most common ways the epigenome functions is when environmental stressors, especially early in life, wrap DNA
filaments around spools of protein called histones. Stressors can also add or remove minuscule chemical side groups, called methyl and acetyl, to or from genes. These are just small groups of atoms that latch on to DNA strands. Such alterations can stop, slow down, or speed up a gene's ability to be read and do its job. Changing a gene's action alters the amount of proteins that are made, and therefore changes the balance of neurotransmitters in brain circuits, leading to changes in thoughts, emotions, and behaviors. These modifications are a big deal and have become a major focus in the understanding of the interaction of genes and environment, and are the key to understanding the nature-nurture problem. One of the main environmental stimuli that add these methyl and acetyl groups is stress, and these stimuli can include abuse, prenatal maternal anxiety, drugs, and even some foods. Stress releases the hormone cortisol, which transfers methyl and acetyl groups from donor molecules on to DNA.

Such additions may be a key in understanding the etiology, or cause(s), of psychopathy. When these side groups are added or removed from the regulators of the genes, the genes' function is temporarily altered, sometimes for hours or weeks, sometimes for years. Thus, early in-utero stressors like maternal use of alcohol, illegal drugs, or psychoactive medications can alter the later behavior of that child. But stressors occurring close to the time of birth can have the greatest deleterious effects. Furthermore, the later the stressors, like emotional or physical abuse, occur, the less the effect will tend to be. So emotional abuse or abandonment at the age of a year or two is far more deleterious to the child's
development and later behavior as a teen and adult than abuse or abandonment at age six or ten.

In perusing the literature on environment and psychopathy, I remembered a classic 2002 paper by Avshalom Caspi then of King's College London and his colleagues, showing what I considered to be the best demonstration of the interaction of nature and nurture. Caspi looked at data from the Dunedin Multidisciplinary Health and Development Study, a long-running study of about a thousand people born over the course of a year (1972–1973) in Dunedin, New Zealand, who have been assessed on several health and psychological measures every few years since they were three. Caspi looked at three factors: whether the subjects had the warrior gene, whether they'd been maltreated as children, and whether they displayed antisocial behavior. (Antisocial behavior was measured by combining a diagnosis of adolescent conduct disorder, convictions for violent crimes, a psychological assessment of a violent personality at age twenty-six, and reports of antisocial behavior from people who knew the subject well.) Caspi found that maltreatment, as expected, increased antisocial behavior. But the increase was much greater in males with the warrior gene. Twelve percent of the guys had this combination of abuse and the warrior gene, but they were responsible for 44 percent of the men's violent convictions, doing four times their share of the damage. Overall, 85 percent of the males with the warrior gene who were severely maltreated became antisocial. A similar pattern was seen in females, though they were less violent. A later meta-analysis Caspi and his colleagues conducted of similar studies
showed that even without abuse, the warrior gene does increase aggression, but its effect on its own is much smaller.

Those several months following birth are sometimes called the “fourth trimester,” and this extended period of what should have been prenatal development means that early environment for a human infant is particularly important. In fact, the most vulnerable time for a human's brain development in terms of environmental impact is from the period of birth and for several months after, in this fourth trimester. It is in this time that a human needs to avoid serious stressors, and it is when nurturing is so critical. There is a continued need for protection throughout childhood, of course, but the closer to the day of birth, the more important affection is.

Damage to the brain also shapes psychopathology in different ways depending on when it occurs. If at the age of two a child suffers damage to the orbital cortex, which is involved in ethics and morality, he may never develop a sense of right and wrong and may become profoundly psychopathic. If the damage occurs at the age of eight, the person's orbital cortex may have helped other parts of the brain understand right and wrong, but he won't be able to stop himself from committing wrong, as the orbital cortex is also involved in inhibition. If the damage occurs as a teenager or adult, the person will know right from wrong, and other areas of the brain involved in inhibition will be mature enough to help control impulsivity when the orbital cortex fails, but stressful conditions could easily push him over the edge.

Even without specific brain damage, several psychiatric diseases can rear their heads later in life.