How the Social Becomes the Biological: The Interaction Between the Genome and the Environment

Dr. Legato: The discussion is meant to address the interaction between the genome and the environment and is essentially a discussion of the role of epigenetics in modifying the phenotype. We think it is one of the most important issues in molecular biology, and resolves the artificial division between what is biological sex and what the role of the environment is in producing the final phenotype.

So, with that introduction, I would like to introduce the moderator of this discussion, Gillian Einstein, the Wilfred and Joyce Posluns Chair in Women's Brain Health and Aging. Gillian, I am going to also introduce and describe the areas of interest of Robert-Paul Juster, Bruce McEwen, and Dr. Barros.

Dr. Juster's research program focuses on chronic stress among diverse social groups. Over the last decade, he has developed expertise in measuring chronic stress, known as the allostatic load, which describes the physiological dysregulations related to chronic stress and unhealthy behaviors.

Dr. Bruce McEwen is the Alfred Mirsky Professor and the Head of the Laboratory of Neuroendocrinology at the Rockefeller University. He has studied extensively the cellular and molecular mechanisms that underlie the effects of stress and sex hormones on the adult and developing brain. His particular contribution for the purposes of this roundtable is his creation of a new understanding of how the brain changes in adult life and during development, with particular emphasis on understanding the impact of stress on the brain and sex differences in human brain function.

Dr. Barros is an expert on the link between inflammation and cancer insights into the role of epigenetics in cell differentiation. She is in the Department of Periodontology at the University of North Carolina at Chapel Hill.

Gillian, would you like to begin the discussion?

Dr. Einstein: Yes. You might all say what is at the top of your list with respect to this topic. Bruce, do you want to go first?

Dr. McEwen: Sure. Many years ago, I was a student of Alfred E. Mirsky, and was introduced to what we now call epigenetics. Mirsky and his younger colleague, Vincent Allfrey, were among the first to show that histones are modified to alter gene expression. So I have lived with this in my whole career, which is one of the reasons we started to look at steroid hormones and their ability to affect gene expression.

It is through that that we then discovered the plasticity of the brain. The whole idea of epigenetics allows one to understand how the social environment, as well as the physical environment, can affect the development of the brain, modify brain structure, and, through brain–body interactions that we will talk about with Rob and the idea of the allostatic load, influence the entire body in a reciprocal way.

Dr. Einstein: That is great. Rob, what about you?

Dr. Juster: I would say that my work is very downstream of epigenetics, and my focus has been primarily on peripheral biology and combining multiple different perspectives to understand how social environments can lead to allostatic load. Throughout my PhD and now in my post doc research, I have been focusing on nuancing the different effects of biological sex, sex hormones, sociocultural gender, and sexual orientation, and how each of those different factors relate to stress physiology and allostatic load and mental health.

Dr. Einstein: I wanted first to ask something that has never been quite clear in the literature to me, especially if you consider animal models. It seems to me that epigenetics is primarily studied in males and not in females. Is that something that you all would agree with, and if there are some strong examples in females and if there are sex differences, could we talk about that?

Dr. McEwen: Maybe I will start first. Of course, sex differences have been a major part of what we have done over the years, and just this week, a paper that we submitted to Nature journal is in the process of being accepted in which we are looking at the ability of stress to regulate gene expression in both males and females in—one group of cells in the brain, in the hippocampus, that we know are vulnerable to stress. If the stress is too strong, as in seizures or a stroke, these cells self-destruct, but over the course of a normal period of stress, the dendrites shrink in males but not in females.

So we have been fascinated by that kind of sex difference for years, and now what we see looking at the level of gene expression, which is a reflection of epigenetics, is that the genes that are turned on and off in males, in this one group of cells, are quite different in males and in females. There is overlap. There are always certain immediate early genes that are turned on by the acute stress, but for the most part, there are large sex-specific differences in which genes are turned on and off.

In fact, one of the paradoxes is that among the overlap, the genes that are upregulated in the male are often downregulated in the female and vice versa. So it gives us a picture of how different even a group of cells can be that is not normally thought to be connected in any way to sex or sex differences. We presume this is happening to other cells throughout the brain, so it is quite a dramatically different view of the brain and sex differences.

Dr. Einstein: That is fantastic. What a great contribution. I go back to the early work looking at licking of the anus in mouse pups and behavioral changes with respect to that. In those studies, there was not a consideration of sex differences. I think most of the animal studies were done in males.

Dr. McEwen: Oh, of course, in the Michael Meaney adaptation of the pioneering Seymour Levine-Victor Denenberg handling studies, the observation was that for females, if you cross-foster between a good mother and a not-so-good mother, the female pups will acquire the characteristics of the mother. That is, if they are nurtured with a good mother, they become themselves good mothers and vice versa. In those studies, not too much was observed about the males, except that they are perhaps more laid-back if they are nurtured by a good mother.

Dr. Einstein: Interesting.

Dr. Legato: Bruce, what are the implications of your elegant studies on the impact of stress on the hippocampus and other parts of the brain on behavior or the response to stress?

Dr. McEwen: Well, that is a good question. In this particular study, we were looking at only one cell group, and we have not looked at others. But there is a study by a German group headed by a woman named Uta Habel, who does functional imaging on men and women. And in this particular study, when men and women do three different tests of emotional sensitivity, they end up scoring the same. ¹ Now, while they are doing this test, they are in a functional imaging machine, and the areas of the brain that are activated by these tests are recorded. What you see is that the typical average pattern of activation in the woman's brain is quite different from in the man's brain. Different circuits are being activated.

Dr. Legato: That is fascinating.

Dr. McEwen: Somehow the brain knows what to do with the stimulus, reinforcing the notion that men and women do many of the same things equally well but they use different strategies. And that is kind of reminiscent of the Mars and Venus stuff that we hear.

Dr. Legato: Exactly. There are also data, Gillian, to bring in another environmental impact, that nutrition has sex-specific effects on embryos and their development. The Dutch Winter famine studies and more recently studies in Gambia show that the resulting epigenetic pattern is quite different in male and female embryos subjected to peri-conceptual and early developmental nutritional deprivation.

Gillian, could you talk about or lead the discussion about the heritability of these epigenetic phenomena?

Dr. Einstein: Yes, well, one of the things that is so remarkable about the findings that are now coming out is that we are learning that Lamarck might have been correct, that is, that there is heritability of acquired characteristics. And I wondered if any of you would like to comment on that.

Dr. McEwen: Well, there is a cover of Nature Neuroscience that shows Lamarck's picture, and it features an article by Kerry Ressler on modifying the olfactory system in a mouse in such a way that the modification was transferred across several generations as an epigenetic trait, presumably in the germline DNA. In a follow-up study, they actually showed that they could reverse this or treat these animals in such a way that would actually stop this progression. So, that is exactly along the lines that you were saying.

There are two studies on bariatric surgery: one on women and one on men. The studies on women show that if you do bariatric surgery and then the woman conceives a child, the child has a much lower probability of becoming obese up into the teenage years. Now, this may be a matter of life-style change, but it may also be something produced by an epigenetic impact on the germline.

There is a study on morbidly obese males who underwent bariatric surgery that showed that the pattern of cytosine methylation in the sperm was markedly altered by the bariatric surgery. Among the genes that were altered were genes associated with metabolism. Now, what is missing there is a long-term study of whether the progeny of such a man would be less likely to become obese. But there are suggestions there, and in some mouse studies as well, that this may be the case.

Dr. Einstein: Very interesting. I mean, it begins to make us think that these changes are permanent because I do not think we really know how to reverse these epigenetic phenomena if they are inherited. I do not know, Rob, if you have any thoughts about that, since you are studying populations that are under a lot of stress.

Dr. Juster: You know, epigenetics has not really been assessed in the allostatic literature because so much of the literature is based on animal models. But there was one study, Bruce, that you were part of with Goodman in 2005 that was published in Psychosomatic Medicine ² that looked at intergenerational associations of lower socioeconomic status.

What they found in this study was that non-Hispanic, Caucasian, and African American children with less educated parents had much higher allostatic load that was driven by insulin dysregulation. I think this was one of the first studies to really open the idea in the allostatic load literature that there might be epigenetic modifications going on.

Dr. Legato: Dr. Juster, can you discuss or are there any data about changes in the epigenetic characterization or the epigenetic picture in transgender patients as they transition from one gender to another?

Dr. Juster: That is really interesting. I do not know of any literature. I do know that there was some recent controversy about an epigenetic study looking at sexual orientation, but I am not familiar with whether that got published. It got a lot of media attention when it was presented at a conference, but I do not know of anything parallel in transgender individuals.

Dr. Legato: Bruce, do you?

Dr. McEwen: No, I don't. I mean, there are, of course, in a totally different domain, the studies of Rachel Yehuda on the children of Holocaust survivors and also children of survivors of 9/11, which are suggestive that there might be some epigenetic transmission. It may be behavioral or it might also reflect something more fundamental.

Dr. Legato: Has the persistence of epigenetic change been traced beyond the third generation, or do we still have work to do to see how permanent those changes are?

Dr. McEwen: I am not absolutely sure. I think in some of the studies, one other criterion is that they should go on to the third generation, but I cannot be exact in whether they have done it.

Dr. Einstein: I was wondering, Dr. Barros, if you know of continuing generational changes with respect to cancer effects.

Dr. Barros: Well, just from Yehuda's Holocaust-related studies. These are those that show transgenerational changes ³ and identification of intergenerational effects on methylation of FKBP5, demonstrating an association of preconception parental trauma with epigenetic alterations that are evident in both exposed parent and offspring. This study provided potential insight into how severe psychophysiological trauma can have intergenerational effects. FKBP5 genetic variation and altered expression levels can be associated with many different tumors.^4,5

In regards to generation changes, what we could observe initially in animal studies that were planned to address the association between periodontal disease, gum disease, low birth weight, prematurity, and pregnancy complications was that mice born from mothers infected with oral pathogens were born too early and too small. ⁶ The ones that survived presented with behavioral issues represented in behavioral assessments as altered rearing movements and acoustic startle tests. Both male and female Campylobacter rectus–exposed F1 mice had significant reductions in rearing movements, providing evidence that the maternal infection led to alterations in exploratory activity in a novel open field chamber. Bacteria-exposed F1 mice had a trend toward enhanced acoustic startle responses that closely approached significance. ⁷ The behavioral changes were consistent and significant. These assessments were performed in mice that were 4–6 months old for >2 months. We also have data from another study in humans. This was a study where we were trying to understand the effect of treating patients with periodontal disease. The rationale is that periodontal/gum disease leads to systemic infection. This can take effect by two pathways: (1) through elevated levels of inflammatory markers or cytokines, such as interleukin (IL)-6, IL-1, tumor necrosis factor alpha, C-reactive protein, or (2) a direct infectious effect.^8,9 In addition, there are publications showing that live oral bacteria can be found, for example, in atheroma plaques.^10,11

So, the goal of our research was to evaluate how to interfere with these pathways by treating pregnant women to potentially reduce prematurity and complications and also related morbidity in infants. The “Maternal Oral Therapy to Reduce Obstetric Risk” (MOTOR) ¹² study showed that treatment had no impact on prematurity, ¹³ opening a discussion about the precise timing of periodontal treatment in pregnant women. ¹⁴ MOTOR children, that is, 263 two-year olds born to women in the MOTOR study, were evaluated for neurodevelopment assessments using the Modified Checklist for Autism in Toddlers (MCHAT). We observed differences between the children of mothers who presented with periodontal disease. Toddlers from untreated mothers had significantly lower mean MCHAT composite risk scores compared to controls. Thus, treating maternal periodontal disease during pregnancy significantly lowered autism risk scores for toddlers, and this effect, evaluated at 24 months after birth, was statistically significant.^15,16

Dr. Einstein: One of the things I find fascinating about your work, just as I find fascinating about neuroendocrinology, is that it demonstrates that all body systems are interconnected: what happens in the mouth affects birth, affects the reproductive system, and influences genetics all over the body.

Dr. McEwen: I think that is really the essence of this concept of allostasis and allostatic load: the idea that the brain and body are continually communicating in a reciprocal way—that the systems that are engaged to help us adapt. Cortisol, the inflammatory, the immune system, the metabolic system, and the autonomic nervous system are all there to help us adapt. Without them, we would not survive. And yet, when they are overused and dysregulated, you find that they can also cause pathophysiology. Hence, it is our life-style, the experiences that we have, when we sleep, what we eat—so many things in our lives actually can affect how those systems treat us.

Now, I know Rob has done a marvelous job of applying this and looking at many different aspects besides the one he is working on now. Maybe he could comment a little bit further about the implications of that.

Dr. Juster: I think one of the first observations that Bruce, myself, and Sonia Lupien made when reviewing the allostatic load literature was just how inconsistent sex differences were in allostatic loading. Of course, that makes sense. Every single sample is going to be of a different age and of a different socioeconomic background. But among the studies that actually look at differences in individual biomarkers that comprise the allostatic load index, there are marked sex differences that are sometimes utterly ignored if sex is just thrown in as a covariate. In making this observation, we also recognized that a lot of the causes of allostatic load as well as some of the moderators/mediators, such as social support, religiosity, and occupational characteristics, were variables that were showing marked sex differences. So, we started thinking about how this was really more of a sociocultural gender effect.

And that is when I decided that my PhD work would focus on nuancing all these different characteristics. One thing that we have been doing recently (and it is not very common in the allostatic literature) is using sex-specific cutoffs in our analyses. In doing so, we are able to detect gender-based changes in allostatic load. So, we have been looking at how gender roles are different between the sexes. By taking a sex-specific approach, we are able to tease apart these gender-based associations.

We are finding the same thing as well when looking at sexual orientation. I am working on a paper with Teresa Seeman, Vickie Mays, and Susan Cochran from UCLA that is looking at the National Health and Nutrition Examination Survey. It is a sizable sample of 13,000 individuals. And when we do sex-specific analyses and look at sexual orientation, we find that there are no differences among women, but there are among men: bisexual men show the highest allostatic load, while gay men show the lowest.

We are starting to think about how we all are familiar with Shelley Taylor's model of tend-and-befriend, which argues for sex differences in bio-behavioral responses to stress. But we could think even further about differences as a function of sexual orientation, that is, where each of these different socially marginalized groups cope with stress in different ways. And that can ultimately lead to a different allostatic load profile.

Dr. Einstein: Very interesting. We are actually looking at women who are in what we call emotional kinds of jobs. We are looking at the 9/11 responders, most of whom are women, and they carry a huge emotional load. One of my graduate students, Arija Birze, is looking at allostatic load in these women. The data are not all in yet. She has just finished collecting blood, but it already looks as though cardiovascular parameters are different in these women. But in a sense, she is looking at gendered work, since this kind of work calls on what is thought to be a primarily female sort of response, that is, an emotionally caring response, and how it actually affects health down the line.

Dr. Juster: That is fascinating.

Dr. Legato: I was just thinking of Takotsubo cardiovascular syndrome, in which the initial case reports were of 20 people presented with a tremendous emotional shock (either positive or negative) experienced acute dilatation of the left ventricle that could sometimes result in myocardial infarction, but other times did not. Nineteen of the subjects who experienced the syndrome were women and only one was a man. I think that is a gendered and quite fascinating response to stress.

Dr. Einstein: Yes. Well, on that note, I would be really interested in hearing about how you think epigenetic changes might influence successful aging. Perhaps Dr. Barros, you would start the discussion on this.

Dr. Barros: Well, in addressing aging, the main associations we are investigating relate to the role of oxidative stress, inflammation, and infection. Oxidative stress plays a leading role in this process, with significant influence on a wide range of biological and molecular processes. We can find relevant information on the strong association of oxidative stress and aging and how epigenetics can influence this process. Global hypomethylation of the aged genome has been associated with decreased activity of DNA methylation enzymes. ¹⁷

We are currently investigating how much we can reverse epigenetic changes that occur in response to inflammation and infection. We have data on epigallocatechin-3-gallate (EGCG), the major catechin found in green tea, and other DNA methylation inhibitors (RG108) to try to reverse the effects of infection and inflammation in epithelial cells. By inducing inflammatory reaction in epithelial cells and then treating them with EGCG and RG108, we observed that exposure of cells to bacterial infection affected epithelial barrier function by increasing cell permeability. Permeability alteration was associated with increased methylation levels of the PKP2 gene, which plays an important role in desmosome assembly, accompanied by reduced gene expression. The DNMT inhibitors activity prevented alteration in cell adhesion in response to infection, minimizing the disturbance to the barrier function. ¹⁸

With regard to aging, dietary supplements have been investigated as substances that have the potential to reverse epigenetic changes accumulated throughout life, and in this way, prevent age-related diseases through epigenetic modifications.

Curcumin (diferuloylmethane), a polyphenolic compound, is a known anti-inflammatory and antioxidant, which has been shown to modulate several diseases via epigenetic regulations. ¹⁹

Other papers evaluating oxidative stress report that dietary curcumin supplementation ameliorates two clinically important markers of arterial dysfunction associated with aging: large elastic artery stiffening and endothelial dysfunction. ²⁰ With accumulating evidence showing that oxidative stress, which is caused by enhanced reactive oxygen and nitrogen species and reduced antioxidants machinery, is another underlying cause of neurodegeneration in Parkinson's disease, studies using drosophila and rat models of the disease presented data testing various doses of exposure to curcumin. The studies indicated a dose-dependent significant delay in the loss of activity pattern, reduction in the oxidative stress, and increase in the life-span in the drosophila model. ²¹ Treatment with curcumin and DFO attenuated the loss of dopamine and increased antioxidant enzymes, resulting in preservation of dopaminergic neurons in a rat model combining curcumin with desferroxamine, a chelating agent. ²² Has anyone studied that?

Dr. McEwen: You means in terms of pharmacologic interventions that impact the aging process?

Dr. Barros: Yes.

Dr. McEwen: No. I was thinking of the more psychosocial factors and the work that that suggests that having what is called a eudaimonic life-style (a life-style that has meaning and purpose). It was a study that that showed that having such a life-style actually results in lower level of inflammatory parameters in the blood and, by implication, perhaps, a longer health span. ²³ I think some of the work from The Rush Hospital Center ²⁴ in Chicago has supported the idea that meaning and purpose in life is a factor associated with a healthier aging pattern.

Dr. Juster: There has also been some literature looking at allostatic load which has similar findings: a sense of coherence is a variable related to lower allostatic load. There is one study published in PNAS by Zalli, ²⁵ with Andrew Steptoe as the senior author, which looked at a subset of individuals from the Whitehall II sample. They found that individuals with shorter telomeres and higher telomerase activity had higher allostatic load. The effect was independent of age, socioeconomic status, and body mass index. This is one of the first studies to link telomeres with allostatic load.

Dr. McEwen: And certainly telomeres are supposed to get shorter because of an increase in oxidative stress and inflammation.

Dr. Einstein: Has anybody done a genome-wide association study (GWAS) on these groups and made a link between modification of the genome and stress?

Dr. Juster: I do not know if that has been done with the Whitehall II sample, but I would be very surprised if it hasn't. I just have not read the article. I mean, the Whitehall II has just been phenomenal in measuring every cutting-edge biomarker they can get their hands on.

Dr. Legato: Would any of you like to comment on the emerging investigation of RNA epigenetics and its relationship to DNA epigenetics? I realize that the former is a little bit more in its infancy than are studies of epigenetic modifications of DNA, but are any of you looking at that?

Dr. McEwen: We use RNA epigenetics as a marker of gene regulation, whereas the studies of DNA epigenetics often involve studies of methylation or hydroxymethylation. And then when one looks at underlying mechanisms for epigenetics, you have, of course, not only the methylation of DNA, but also you have the modifications of histones that can either increase or decrease the repression of chromatin or the activation of chromatin.

On top of that, you have the microRNAs that regulate how RNA is processed. There is also something called RNA splicing and RNA editing, which are further ways of modifying the RNA when it is produced. Going back for the moment to the microRNAs, there is about 50% of the genome that used to be called “junk DNA,” which we know comes from transposons and retrotransposons: viral DNA that has been incorporated into the genome and really provides the regulatory signals that make us different from worms, since we do not have that many more structural genes than a Caenorhabditis elegans. Yet, we have this completely different body type.

So, I think that this whole area of epigenetics is burgeoning. One of the problems is that many people think of it only in terms of CPG methylation, but it involves so many other components.

Dr. Legato: Exactly. What about assisted reproductive techniques and the errors that happen in epigenetic labeling in those products of that technique? Any comments about that?

Dr. McEwen: There are, of course, studies such as the whole cloning of DNA from somatic cells. Even though the genome produced a whole organism, there may have been modifications, since it came from a somatic cell origin that affected things such as life-span and so forth. That is a bit different from what you were asking about.

Dr. Legato: Yes, but it is the same principle though.

Dr. McEwen: Same principle, yes.

Dr. Einstein: I am sure you have thought about this Bruce, what is the molecular pathway from exposure to trauma or stress and, now I am thinking of this, Christine Heims' study on changes in somatosensory cortex with early exposure to sexual abuse, and epigenetic genome modification? Biologically, how do we get from an experience such as sexual abuse to a reduction in cortical thickness, for example?

Dr. McEwen: I think we are now realizing that many parts of the brain are continually being shaped by experiences. For example, a chronic stressor such as simply restraining an animal every day for a few hours a day, even while it is in its resting state, actually causes neurons in the hippocampus and prefrontal cortex to shrink. ²⁶

At least in males (females are different), neurons in the basolateral amygdala grow, and the animal becomes more anxious. In neurons in the medial amygdala, the dendrites actually shrink, and the animals become socially avoidant. ²⁷

In a resilient brain, when the stress is over, there is some recovery of some resilience, even though the genome is, you know, continually changing. ²⁸ And yet in a non-resilient individual, the medial amygdala gets stuck in that state, and the animal then shows a depressive-like, socially avoidance behavior that has to be treated in some way as we think about human anxiety and depressive disorders. So, it is a loss of resilience while the brain is continually being shaped.

One of the best examples of that was the recent study of a woman's brain in relation to her delivering a child, showing the sculpting of the brain that involved the loss of volume in some areas of the brain. Actually the greater those changes were, the better the attachment was to the child. ²⁹

And then, of course, musicians show shaping of their motor and sensory control areas, which goes on all the time. ³⁰ It is a continuing process that is inexorably and inevitably connected to who we are and what we do.

Dr. Legato: This is a very relevant subject, given the concentration recently on the consequences of holding prisoners in solitary confinement for long periods of time.

Dr. McEwen: Absolutely.

Dr. Legato: Is anyone studying that in a systematic way? It might help to reverse that common practice in punishing offenses by prisoners.

Dr. McEwen: If we talk about the idea of therapeutic interventions and resilience, the poster child of changing the brain in a positive way involves regular exercise or physical activity on prefrontal, cortical, and hippocampal volume and function.^31,32 And, of course, regular physical activity has antidepressant effects. Mindfulness-based interventions have been shown to change amygdala activity and amygdala volume, as long as the person is also showing a reduction in a chronic anxiety disorder. ³³ So, there are things that have been documented to actually have positive, maybe ameliorative, effects for an existing condition.

Dr. Einstein: Just to go back to the brain reshaping and resilience, are you suggesting, Bruce, that there would be epigenetic modifications through stress and potentially stress hormones that would then change the ability of the brain to respond with successful reshaping?

Dr. McEwen: Yes. I mean, one example we have is looking at the methylation of lysine residue on one of the histones. When you acutely stress a young rat, you find that there is a huge increase in the methylation of this lysine residue in the hippocampus and particularly in the dentate gyrus, and not so much in any other brain region. When you do repeated stress, this methylation effect disappears. ³⁴ It also is not so present in an older animal. Now, one of the things that is associated with that methylation, if you trap the DNA that is repressed, turns out to be some of these retrotransposon-like elements, which are known to promote what are sometimes called “jumping genes” because the DNA can rearrange. And when it rearranges itself, it may expose actual structural genes, coding genes that actually are not normally active or cause other genes that normally would be active to disappear.

So, as this goes on, perhaps in the aging process, it may lead to loss or gain of function. The best example of DNA rearrangement is the generation of diverse antibodies through a DNA rearrangement process. This is something that is a future direction for more research on how those rearrangements actually alter genes and how this may change as a function of experience, or as a function of aging, and so on.

Dr. Juster: It would have huge implications as well for the concept of cognitive reserve: how individuals with neurodegenerative diseases will degenerate at a different rate impacted by their experiences. Do you know, Bruce, if any research has been done looking at cognitive decline in humans and epigenetic modifications?

Dr. McEwen: Not yet. I think it is a very good question.

Dr. Einstein: I do not think such research has been done. I was just at the Alzheimer's meeting in London, and I did not hear anything about that, which is really an interesting, important point.

Dr. McEwen: Yes, I think this has not really reached that community yet, and it should.

Dr. Juster: Yes, I mean Yaakov Stern, who is a really big name in cognitive preserve, is here at Columbia, and our group is hoping to collaborate with him. We will suggest it.

Dr. Einstein: I went to a talk by Yaakov Stern and there was no mention of this. I think it is a really great idea.

Dr. McEwen: This goes back to Barbara McClintock, who studied the different colors on corn maize, which are really the products of these jumping genes, which expose or hide genes that express the different pigments.

Dr. Legato: I was going to ask all of you whether the action of a hormone is to directly engender methylation or some change in the modification of genes? Is it a direct effect of the hormone, or a trigger for methylation, for example?

Dr. McEwen: There is a whole host of enzymes that put these different modification groups on cytosine residues and DNA and on lysine residues and histones. And their activity can be regulated through hormones but also by other cellular mechanisms. The most common modification of histones that leads to gene activation is putting an acetyl group on a lysine residue, and acetylation is produced by the same acetyl coenzyme A that is involved in oxidative metabolism in the mitochondria.

There is a reserve molecule called acetyl-L-carnitine (LAC), which we are studying because it is deficient in some people with depression, and supplementing it in animal models with depression can lead to very rapid relief of depressive-like symptoms. One of the mechanisms appears to be the acetylation of a lysine residue, which leads to the upregulation of an important gene that regulates glutamate release.^35,36 The consequence of having too much glutamate is to shrink dendrites and to suppress neurogenesis, which are some of the things that happen, we think, in a depressed brain. Usually, in the brain at least, hormones work synergistically with a whole host of other mediators to produce their effects. So, it is not the hormone working by itself.

Dr. McEwen: For example, estrogens will cause a rise and fall of synapses in the hippocampus and prefrontal cortex during the estrous cycle, and they do so by a synergy with excitatory amino acids and NMDA receptors. ³⁷ And if you block the NMDA receptors, you see no estrogen effect. So, it is always in concert with other cellular processes.

The same thing is true of the stress-induced loss of spines or dendrites in the hippocampus, which we can reproduce by giving excess glucocorticoids. ³⁸ But even those actions are blocked if you block the excitatory amino acid receptors while you are giving the glucocorticoid.

Dr. Einstein: I am wondering if anybody has looked at the role of endocrine disrupters in affecting gene expression and how exposure to toxic waters might affect methylation or demethylation of genes and epigenetics?

Dr. McEwen: Well, Csaba Leranth at Yale and Neil MacLuskey, now at the University of at Guelph in Ontario, have shown that bisphenol disrupts the estrogen effect I spoke of on synapse formation. ³⁹ Now, I am not sure that we know mechanistically how it does so, but I believe that Marija Kundakovic and Frances Champagne at Columbia ⁴⁰ have studied some of these agents for their effect at the epigenetic level.

Dr. Einstein: That takes me to the next question, which is a little bit far afield but always present, which is the moral, the bioethical, and the ethical repercussions of understanding the epigenetic changes due to environment due to gender roles, or other kinds of roles in life. I am wondering if you all have thought about that. I am sure, Rob, you have.

Dr. Juster: I have certainly thought about it in the context of sexual and gender minorities. There is always a big divide of nature versus nurture, and conversations get very heated one way or the other. But I think epigenetics represents a really interesting way of thinking about how stigma and prejudice can affect particular genotypes and lead to different health outcomes. I think one study that I would love to do at some point in my life would be to focus on epigenetics among sexual and gender minorities based on candidate genes involved in the stress response. I think such an approach would get away from trying to understand what gene explains why someone is the way they are, but more how the interaction between particular types of genes (e.g., those related to glucocorticoids or brain-derived neurotrophic factor or serotonin transporter), and how that would be related to experiences of stigma, that then would influence mental and physical health.

We know that the lesbian, gay, bisexual, and transgender (LGBT) community experiences much higher rates of psychiatric disorders, but at the same time, it is a community like many other marginalized communities that can show remarkable resilience. And so, I think taking an epigenetic approach would really allow people to understand how the environment modifies the gene and makes them hardy.

Dr. Einstein: I was wondering, Dr. Barros, if you had thought about this with respect to cancers.

Dr. Barros: Not to cancers. We have not been directly focusing on cancers, but we do have recent data on a large group of patients. We are in the process of analyzing samples from 900 subjects through an epigenome-wide association study (EWAS) with the objective of identifying methylation signals associated with periodontal disease, with smoking as the possible mediator of this association. Data are showing DNA methylation association with genes involved with inflammation and oxidative stress pathways, which have been also associated with cigarette smoking.

We will complete the EWAS analysis in African Americans and Caucasians to confirm this effect of tobacco smoking as environmental exposure in association with periodontal disease.

Dr. Legato: This whole topic of how gene expression and regulation is the link between the environment and the phenotype really puts a whole different light on the consequences of mishandling of individuals by society and the environment. It not only makes it much more understandable, but it gives us some hope of changing it and, with it, how to change the effects of a malignant environment on behavior.

Dr. Barros: And we do have data on former smokers to access that reversal of the hypomethylation found in smoking-related genes.

Dr. Einstein: I was just going to ask about reversal.

Dr. Barros: Yes, we could access that. Still, they are preliminary data, but we could see this reversal effect. Our findings suggest the existence of dynamic, reversible site-specific methylation changes in response to cigarette smoking, indicating that after smoking cessation, methylation levels increase approaching those levels of never smokers. In this EWAS analysis of periodontal disease, tobacco is associated with decreased methylation levels of smoking-associated genes and with severity of the disease.

Dr. Legato: Does age have any impact on the degree of reversibility that is possible?

Dr. Barros: In this population, we could not identify age as a significant factor. We are working with a population that is relatively older. We do not have comparative data from a younger population.

Dr. Einstein: I have a futuristic question. In a family where there had been generations of mental illness or a number of generations that have been affected by progenitor stress, do you think that people will begin to try to do the reverse methylation of genes to reverse these effects for future generations?

Dr. Legato: I think that is the implication of what we are uncovering about how the environment works on the individual.

Dr. McEwen: Of course, the big thing in terms of this epigenetic view is the life course, as exemplified in a beautiful review article by Neal Halfon, ⁴¹ who looks at the view of medicine from the standpoint of three systems, using the analogy of computer operating systems. So, the first system was the view of things, such as pathogens, and finding magic bullets, such as penicillin.

The second system is the idea the life-style, the disease of modern life: what we eat, how well we sleep—all these things. And we are in the middle of that era right now. Beyond that is the operating system three, to realize it all begins even before conception, and it is a one-way epigenetic street in which you cannot roll back the clock, but you can shift the trajectories or directions that the brain and body will take by interventions at different stages of the life course. Early interventions are likely to be more effective, and while interventions later can shift the life course, fortunately, they are not going to be able to put you back to the state that you might have been if things had not happened the way they did early in life.

So, you get into the whole area of early life abuse and neglect, known as adverse childhood experiences (ACEs) pioneered by Anda et al. in 1998. ⁴² But then beyond that, growing up in poverty and in polluted, ugly environments, the lack of social support, loneliness, many factors that all influence how our brains and bodies function and the trajectory that they take.^43,44

All of this is the new view that is now altering how we think about interventions. It is not just about using drugs; we know that drugs can only have certain effects, and they often cause side effects. The interventions that are related to changes that we have already talked about—exercise, improving sleep pattern, mindfulness, dietary, you name it—are beginning to be incorporated into what is called integrated medicine. ⁴⁵

This really is the kind of model we should all think about. One of the things that I do is to participate in the National Scientific Council on the Developing Child, headed by Jack Shonkoff at Harvard, which, by the way, has a wonderful website: http://developingchild.harvard.edu/science/national-scientific-council-on-the-developing-child. And the whole focus there has been on prevention of adversity, interventions early in life such as the Nurse–Family Partnership, developed by David Olds, to try to improve the interaction between the parent, or the mother and child, to avoid the development of these ACEs that we know are so bad (www.nursefamilypartnership.org).

There are interventions that are related to reversing poverty, moving to opportunity—things like that—that actually seem to have long-term benefits to help people have a more fulfilling life. But then the problem comes when you develop some of the ways in which you try to change the trajectory there.

Dr. Einstein: Yes, exactly. Marianne, I do not have any more specific questions for our participants, but I would be interested in going around and asking everyone what they think are the key questions we should be asking at this point, if we have not covered them.

Dr. Legato: Wonderful question. Dr. Barros, do you want to start? What questions would you most like to be asked and answered?

Dr. Barros: Well, we would like to concentrate on pregnancy complications and try to understand how much we can prevent and/or reverse the effect of infection and inflammation in newborns, since in utero exposure can potentially lead to long-term life health effects, including cardiovascular disease and diabetes.

Measurement of the potential impact of periodontal disease on general health can be very complex, especially when focusing on pregnancy and all the intricate balances that go on during that phase with DNA methylation potentially mediating Th1- and Th2- cytokine production in T-cells. Epigenetics can change that balance and thus interfere with many adverse health effects of prenatal exposures that could persist throughout life.

Other questions include how we can measure and better understand the epigenetic effects associated with maternal periodontal infection and inflammation on the behavior of children who were exposed to maternal infection and smoking. Considering that current concepts of autism causality implicate genetic and intrauterine exposures, another research focus would be on the follow-up analysis assessing neurodevelopment on the same children assessed through MCHAT to try to identify the pregnancy exposures that could be implicated with autism or autism spectrum in these children.

Another project would be to evaluate methylation in genes such as AHRR and GPR15 as markers for potentially clinical relevant predictions of smoking-associated alterations and evaluate how much we can interfere with disease progression by reversing the effects of smoking on these entities.

Dr. Einstein: Thank you. Rob, what about you?

Dr. Juster: I am trying to answer three specific questions throughout my postdoctoral work. The first is asking how geopolitical variation in stigmatizing state-level policies affects the health and allostatic load of LGBT individuals.

I am working with a gentleman in public health by the name of Mark Hatzenbuehler, who stratifies state-by-state different policies related to the laws for the LGBT community. What we are hoping to do is to understand how allostatic load is different based on these different state-level policies.

Another research question that I am trying to answer is whether social support buffers against stigma and bolsters the health of older lesbian and gay individuals. Living in New York, there is a great opportunity to understand successful aging of people who fought for LBGT rights here, and so we are trying to understand how cognitive reserve and successful aging are related to their health.

The last research question that I am trying to answer, which is really my little baby that I am most excited about, is understanding the stress-related neurological signatures of transgender individuals. So, we are going to recruit transgender individuals and expose them to a mild stressor. The point is to understand how stress regulatory regions are related to their cortisol and heart-rate variability changes.

With those questions, I am focusing heavily on the LGBT community, but I think more broadly a question that I would love to have answered at some point in my career is do specific experiences of “otherisms” leave similar or dissimilar biological scars. And I think studying the LGBT community offers a really interesting opportunity to understand stigma. I believe that thinking about other groups of marginalized individuals and trying to understand where and how their biological signatures are different, where are they similar, how are different coping strategies used by different groups, and how does that relate to their health are all questions that I am hoping to answer.

Dr. Einstein: Great. Thank you. And Bruce?

Dr. McEwen: Well, a current major focus for us has been driven both by studies in animal models and now studies in depressed people. The big problem is the increasing frequency of obesity, diabetes, depression and dementia—the three Ds—which at least in some people may be related in that the predisposition toward insulin resistance. ⁴⁶ Even prediabetes and frank diabetes in some people are associated with a major depression. And in those people, there is also an increased risk for dementia.

All of these are increasing in our society and in other societies because of what we eat, our life-style, and many other factors, and there are huge healthcare costs associated with all of these. Also, because early life adversity increases the frequency of these, particularly the first two (i.e., diabetes and depression), as well as substance abuse.

There is a huge problem that needs to be dealt with not only with interventions early in life, as the National Scientific Council recommends, but also the thing that we are interested in, which is this mechanistic linkage between diabetes, depression, and dementia. There is a collaborative group that has formed that has the acronym PALS: Pathophysiology Allostatic Load over the Life Course. The person who has really initiated this is Natalie Rasgon, who is a professor of psychiatry at Stanford who not only sees patients but does research, and she is one of the people who originally saw the connection of the three Ds.

Now, we have a human study, not only related to PALS but also related to something called the Hope for Depression Research Foundation, in which we are actually finding people who are probably treatment resistant with normal antidepressants who have severe major depression and who are deficient in acetyl-L-carnitine. And we are trying to see if we can see if they also have a propensity toward insulin resistance or metabolic disruption. ⁴⁷ If we do, then a new direction of therapy has opened up perhaps that might actually help to slow down this transition and, at least for some people, slow down the progression not only to depression but also to dementia.

Dr. Legato: What about you, Gillian? What are your questions?

Dr. Einstein: Okay, fair enough. I am a bit with Rob in that I am interested in gender. We are beginning to develop contacts within the trans community, and we are asking about differences in cognition and organizational and activational effects of hormone treatment. A study of the genome prior to and post transition would be very interesting to do to determine any epigenetic changes or differences between trans and cis. I would be delighted to talk further with any one of you about this because I think we could do such a study in Toronto.

But I am also interested in how the gendered life experiences of women, both traditional cultural experiences, as well as even surgical experiences, end up affecting women's biology. I think this is one of the big questions that we really need to understand in terms of, as Bruce pointed out, the trajectory of health over the life course, including developing dementia. Currently, we are studying women who have had their ovaries removed prior to natural menopause because they carry the BRCA mutation. We are finding significant changes in cognition, which worsen over time, especially in women who do not have hormone replacement.

I think of ovarian removal as a gendered practice because the ovaries are thought of as only important for reproduction. In fact, as we all know, they have effects all over the body throughout life. So, I think that the practice of ovarian removal really needs to stop, and we have to open up other courses of treatment for women. So, I am interested in how tradition and culture and gendered practices affect the biologies of women over the life course.

Dr. McEwen: One thing that I might add is that a study by John Morrison's lab on rhesus monkeys showed that surgical ovariectomy—surgical menopause, so to speak—led in the brain to mitochondria that have a circular shape, and that shape is known to be associated with increased free radical and inflammatory production.^48,49 And this is unlike what appears to happen in natural menopause, not such a dramatic change.

Of course, then it argues not only against surgical menopause but at least for treatments, including estrogen replacement that actually might reduce that inflammatory oxidative stress tone.

Dr. Einstein: I really appreciate you saying that, Bruce, because I am following John Morrison's work very carefully. And in fact, he is a scientific advisor for the chair that I have. We are very interested in understanding whether inflammation is causing some of the brain changes we are seeing.

Dr. Legato: This has been a very valuable and interesting conversation, and I would like to thank you all for participating.