Science Talk

The Mind's Hidden Switches

Eric J. Nestler, director of the Friedman Brain Institute at the Mount Sinai Medical Center in New York City, talks about his article in the December issue of Scientific American magazine on epigenetics and human behavior, called "Hidden Switches in the Mind"

Podcast Transcription

Meet Dr. Bechard Nor, pioneer transplant surgeon and one of the many achievers helping to unlock human potential at Cutter Foundation.

Steve:         Okay, how do you do this again? You press this button, no wait. Okay, I'm going to try this one. No. And what about this one? Yes. Welcome to the Scientific American podcast, Science Talk, posted on November 22nd, 2011. I'm Steve Mirsky. This week on the podcast:

Nestler:         The ability of this chronic social stress to produce maladaptive changes in brain and behavior are mediated through epigenetic modifications of gene expression in particular emotional centers of the brain.

Steve:         That's Eric Nestler. He's a neuroscientist and the director of the Friedman Brain Institute at Mount Sinai Medical Center here in New York City. He's also the author of an article in the December issue of Scientific American magazine called, "Hidden Switches in the Mind". On November 21st, I visited Nestler in his office at Mount Sinai.

Steve:         The old nature versus nurture thing is very simplistic. I think most people understand that it's simplistic at this point. But this article really explains how they're dependent upon each other to an extent that most people probably still don't appreciate.

Nestler:         Yes, I think that's true. The idea that nurture modifies one's own nature has been around forever. Epigenetics provides an understanding of how that actually occurs. And you're absolutely right, in that it begins from the moment of conception and continues throughout life, and it's an iterative process. So that an individual's genetic constitution, his or her nature, defines how that individual will respond from the initial environmental exposures in a developing fetus, in utero, post birth and then throughout life.

Steve:         You come at this from a particular vantage point, because you're an addiction researcher.

Nestler:         Yes.

Steve:         But the findings that your lab generated would apply to great many biological phenomena.

Nestler:         That's right. Yeah, in fact, it would apply to all organ systems. And in fact, our work on addiction has very much relied on earlier advances, mainly in the cancer and developmental biology field, where a lot of epigenetic principles were first established. But then our work in addiction involving the brain, is directly relevant to all aspects of brain function, both in terms of normal activity, normal behavior, learning and memory, and across a wide spectrum of neurological and psychiatric disorders.

Steve:         So, let's talk about the nuts and bolts of the research. We have the genome, which most people are probably familiar with at this point, but it's a very dynamic system depending on what's getting activated and what is getting turned off.

Nestler:         Right. So to a first approximation, a person's genome is fixed from the moment of conception throughout life. A human genome contains roughly 3 billion nucleotides in linear sequence, and that is the same, again to a large approximation, in every cell type throughout life. During development, as an embryo forms differentiated tissues, liver cells, brain cells, muscle cells, the cells in those tissues begin to allow for the selective expression of genes contained in those same 3 billion nucleotides. So those three billion nucleotides are parsed typically into 20 to 25,000 genes. The same 20 to 25,000 genes are present in every tissue in the body, but during development, there's this selective expression of those genes that allow a stem cell to become a liver cell, brain cell, muscle cell and so on. And once those differentiated decisions are made, they persist through life. And we now know in detail that exactly how, through epigenetic mechanisms, that differentiation occurs.

Steve:         Right and development is the clearest example. So that you only have...

Nestler:         It's the best established by far.

Steve:         Right. You don't want a liver cell growing in our brain.

Nestler:         Exactly.

Steve:         But what's really been interesting in the last couple of decades is this appreciation of the fine tuning, of how, let's say, in the liver, certain genes will be activated for specific periods of time, so that the organ can fulfill its function.

Nestler:         Exactly. So the capacity of an organism to adapt to an environment is mediated through epigenetic changes in gene expression. Liver genes will only be expressed in liver. The liver will never express brain genes. But as you said, as an individual eats a high-fat diet or is exposed to other types of environmental exposures, the ability of the liver to adapt involves turning on and turning off certain liver genes to control the liver's activity, and that's mediated through these epigenetic mechanisms that modify the activity of established liver genes.

Steve:         So enough with the liver. Let's get back to the brain.

Nestler:         Yeah, yes.

Steve:         Which is what you study.

Nestler:         Yeah, I mean, by analogy, a person's behavioral experience could be seen for the brain just the way a high-fat diet is seen for the liver; that as an individual learns things, it is subjected to stress, it is subjected to other challenges in his or her environment. The brain attempts to adapt and respond to those challenges and that occurs in large part through changes in expression of brain genes through epigenetic mechanisms. And we now know that it is the same epigenetic mechanisms that occur in brain as occur in liver, just involving, only the specifics differ—brain genes in brain, liver genes in liver.

Steve:         The on-off switch system is surprisingly—I don't want to say, simple—but straightforward, in that we have these methyl groups or acetyl groups, and depending on which one you have in a particular area of the protein surrounding the DNA, you're either going to activate the gene or turn it off.

Nestler:         It's, to a first level of approximation, it's simple and straightforward, that there are a set number of modifications that could be made to the DNA itself or to the proteins that bind to DNA and control its structure, and activity and that those small number of modifications either turn a gene on or off. In actual fact, one of the lessons over the last 10 years is just how complicated these epigenetic processes are. We now know that there are probably about a thousand proteins that are involved in epigenetic regulations. So if you think that a human only has only 20 or 25,000 genes, each gene gives rise to one or more proteins. The fact that a thousand proteins are involved in epigenetic regulation is pretty astounding, because that's just a genes wrapped up in controlling genes. And then in conjunction with that realization has been the observation that in response to an environmental stimulus, whether it's learning something or being exposed to high fat, that turning on a gene might involve hundreds of individual proteins binding to a given gene that's being controlled. So it's really quite complicated, and we're just beginning to scratch the surface and understand those mechanisms.

Steve:         So let's talk about, you have an example right at the top of the article about twins…

Nestler:         Yes.

Steve:         Who have the same exact genome…

Nestler:         Right.

Steve:         But it's the epigenetics that come into play when one of them becomes addicted to cocaine and one doesn't.

Nestler:         So let me say, first is a caveat that identical twins to a first approximation have identical DNA—those 3 billion nucleotides are the same in one twin as in the other. However, we know that there are other types of changes that might occur in a very, very minute number of situations, where the actual DNA sequence might differ due to mutations that occur even after fertilization…

Steve:         But during development.

Nestler:         Of the same egg. But during development. However, with the exception of those rare cases, we think that these epigenetic changes can explain why identical twins can differ so much in many features, and there are many cases that have been around for centuries of identical twins, where one grows to a tall, robust stature and another one is dwarfed. And that's due to growth hormone deficiencies that one twin might develop that the other one doesn't and so on, so that we know those principles can occur.

Steve:         Or the stereotypical evil twin. (laughs)

Nestler:         Or the stereotypical evil twin, that's right. So, now to the case that I mentioned in the article, we know that in most cases, if an identical twin has an addiction problem to a drug of abuse, his or her twin is probably not addicted, because environmental exposures are just as important in order for the genetic potential for addiction to become manifest. And unfortunately, there are some twins who are going to be more vulnerable and the question becomes what is the basis of that twin's enhanced vulnerability. One possibility is that it's just random events during development, that as a few neural stem cells in a fetus give rise to a hundred billion nerve cells in an adult human brain, a lot of stuff happens. And those kinds of random changes during development explain, for example, why the pattern of gyruses in the brain are very different even between identical twins. So clearly, the brains of identical twins are different, presumably just through random events.

Steve:         On a grosser level, their fingerprints are different.

Nestler:         Exactly. In fact, fingerprints are, I see as entirely analogous to the patterns of gyruses in the brain. And then there may be different exposures that the twins would have throughout life. One twin would have a different viral infection and another twin would be exposed to different peer groups in school, getting to very complex interactions, leading ultimately to one twin being very vulnerable and the other twin not being vulnerable. In fact, both twins might try cocaine experimentally, but only the more vulnerable one would begin to show changes in the brain and behavior that would enter that twin into a downward spiral, toward a severe addiction.

Steve:         And what is happening on the molecular level in these twin situations?

Nestler:         What we believe, and we don't really know for sure, yet, but what we believe is that in the vulnerable twin, there is altered patterns of gene expression mediated through these epigenetic mechanism that occur gradually and progressively in response to a wide variety of factors; as we said, random factors, viral exposures, things in the diet, peer groups and so on, teachers. And individual genes in individual regions of the brain would start to show epigenetic alternations that would change the degree to which those genes were expressed, leading to changes in behavior and ultimately to the vulnerability to a drug of abuse.

Steve:         And the use of the drug itself can spark some of these changes that can then become ingrained.

Nestler:         Absolutely. We see the drug of abuse as a key etiological factor in the process of addiction. Individuals who are a very high genetic, maybe even epigenetic, risk for addiction, but who are never exposed to a drug of abuse, that would simply remain a latent risk never expressed. And we are gaining more and more information as to how the drugs actually interact with that preexisting epigenetic state to lead to an addiction syndrome.

Steve:         We also have a situation you describe in the article, where you give mice the equivalent of depression.

Nestler:         Yes.

Steve:         And that's a really fascinating area as well, because of the epigenetic involvement with depression and resilience against depression. So let's talk about that for a while.

Nestler:         Sure. A lot of the work that we do relies on the use of rodents, animal models, rats and mice in particular. And it's always hard to develop a rodent model that is close to the human condition, because the human conditions are far more complicated. Nevertheless, in the case of drug abuse, it's more straightforward because you can simply allow a rodent to self-administer a drug to itself if it chooses to do so, and a subset of rodents actually will do so. They'll press a lever to get cocaine, some of them lose control over it, and even overdose as a result. Animal models of depression are more difficult because some of the symptoms that we use to define depression—guilt, suicidality, sadness—are inaccessible in an animal model. Nevertheless, there are other measures of depression in humans that can be assessed in rodents—like the ability to enjoy pleasurable activities, which is a cardinal feature of depression in humans—you can measure that in a mouse to see how much a mouse likes to eat a sweet treat, have sex, how well it sleeps and so on. And it turns out that when you take a normal mice or rats, and expose them to high levels, particularly of social stress, that the animals do show these signs and symptoms that are quite reminiscent of human depression. And interestingly medications that are used to treat depression in humans work in these mouse and rat models. So we feel that the models, while imperfect, do offer some insight into the human conditions. What we found is that the ability of this chronic social stress to produce maladaptive changes in brain and behavior—loss of pleasure, inability to sleep normally and so on—are mediated through epigenetic modifications of gene expression, in particular, emotional centers of the brain. And we've been able to demonstrate that the altered expression of those particular genes actually mediate the symptoms seen. One of the interesting finding is that a subset of mice or rats that are subjected to these high levels of social stress don't develop these behavioral abnormalities. We call them resilient because they're able to maintain normal functioning despite being subjected to high levels of stress. That's consistent with observations in humans, that a large majority of humans are actually amazingly resilient, they're able to do pretty well despite being subjected to horrendous degrees of trauma and stress. Identifying mechanisms of resilience can be quite important, because as we learn what make some individuals resilient, we can actually take that information and perhaps develop new treatments that would induce that resilience in more susceptible individuals. It turns out in our works to date that we've been able to find epigenetic modifications at particular genes in emotional centers of the brain that mediate that resilience. So one of the interests that we have now is to use that information to potentially develop these pro-resilience treatments.

Steve:         Almost a vaccine against depression.

Nestler:         Exactly.

Steve:         One of the fascinating things that depression work found was that the treatment of the—we'll call them depressed mice even though that's a loose term—the treatment of the depressed mice, restored their epigenetic profile so that it was similar to the native state of the resilient mice.

Nestler:         That's right. That was a surprising finding. We didn't know whether, how the two would compare, how the mechanisms of antidepressant drug action would compare to mechanisms of resilience. There's roughly a 30 to 50 percent overlap in the two mechanisms, which is very interesting because it suggests that one of the ways in which antidepressant drugs work is by inducing in depressed individuals some of the same modifications that occur naturally in those individuals who are inherently more robust or resilient.

Steve:         And I know you haven't done this work, nor could you, but you would think that talk therapy when it is successful, might have the same epigenetic effect.

Nestler:         That would certainly be the expectation, and there is every reason to believe that talk therapy, when it's effective, works by producing the same range of changes in the brain, as seen with antidepressant medications or might even occur naturally in people who are inherently resilient.

Steve:         We should say, we are in a hospital setting, so if you hear sirens, it's not surprising. (laughs)

Nestler:         Yes.

Steve:         One of the really fascinating things in the article is this discussion of, it's not truly hereditary in the Mendelian sense, but you can perpetuate a dysfunction from generation to generation, sort of, indirectly through the epigenetic effect. And why don't you explain that a little bit, because that's really fascinating.

Nestler:         Yes it is. And it's still very early and very provocative, but it raises the use of the term epigenetics, specifically to refer to the ability to pass traits on to offspring, but not through changes in DNA sequence. The best example of this occurs in what's called gene imprinting, where for example, females have two X chromosomes and one of those two X chromosomes must be inactivated in a cell in order for the cell's normal function. That means a girl has two X chromosomes, one from the mother one from the father. In every cell in her body, one of those X chromosomes is inactivated. Those selections of inactivation, which seem to be somewhat random, there may be other guiding principles, can be passed on from the girl's egg cells to her offspring, which means that two girls with identical DNA may have passed on very different traits to their offspring because the paternal X chromosome may be inactivated in one girl, but the mother's X chromosome may be inactivated in the other and leading to different organ functions. That's been well established. What's come out in recent years are findings that experiences in an adult organism can, through epigenetic modifications in the sperm and egg cells, be passed onto subsequent generations, leading to some provocative notions. For example, if a female rat or mouse is fed a high-fat diet and becomes obese there are reports that her offspring are more vulnerable to becoming obese with a high-fat diet, and in turn that could then be passed on to several more generations. In a similar way, we and other groups have early data that exposure of rats and mice to high levels of social stress can make the offspring of those animals more susceptible to social stress, again something that persists for several generations.

Steve:         Let's talk about the non-nurturing mother scenario.

Nestler:         Yeah. Okay. So that' the most straight forward situation, and in fact, we would have a good idea of how that occurs but it's not really through an epigenetic mechanism. So the way, this is work done by Michael Meaney and Frances Champagne and many others, who have done research in rats, where they've shown different styles of mothering, where some mothers tend to groom their pups to a lesser extent than others. You can call them low-licking and high-licking mothers, for example. It turns out that the female offspring raised from low-licking mothers tend to be low-licking themselves, and the offspring from high-licking mothers tend to be high-licking themselves. But that seems to be passed on through the behavior, behavioral experience of the pup because if you did cross-fostering experiments, you'd see very different effects. So a female pup that's licked a little, the process of having less-nurturing care would produce changes in that pup's brain that would persist a lifetime. Those changes are epigenetic based; and then that leads to essentially permanent changes in maternal behavior in that pup once it grows up that then is passed onto the next generation through that offspring's licking of her pups.

Steve:         Right. This can be repeated ad infinitum.

Nestler:         And it can be repeated ad infinitum. I say that's understandable because there's a very clear behavioral mechanism by which that can occur. Obviously, the mothering experience of a pup is going to have profound implications for that pup's own behavior when it becomes a mother and so on. The way epigenetics helps understand that process is that it provides the mechanism by which maternal experience of a pup can essentially produce lifelong changes in the brain because many epigenetic changes are that stable.

Steve:         And if you can somehow treat an individual in this generational cycle, you could theoretically break the cycle.

Nestler:         That's right, and that's been demonstrated experimentally in these rats, that if you take a rat, for example, that's been raised by a low-licking mother, and do something to it—it could be a medication, it could be behavioral change—then you can cause a behavioral switch in that animal so that it now behaves as if it had a high-licking mother.

Steve:         It's really fascinating. So, where do you go from here in your research?

Nestler:         There are several key challenges. One is to delineate this complex code of epigenetic change. As I had mentioned earlier, it's a very, very complicated process involving hundreds of proteins, a  thousand or so genes, and we really need to delineate exactly all of the changes that are occurring. As we identify those changes our hope is, expectation is, that that will help us identify new ways to intervene, to develop better treatments for addiction and depression and other neuropsychiatric disorders. There could be broader implications. So, we're very interested in seeing whether some of these epigenetic modifications can go from one generation to the next, but that remains much more controversial.

Steve:         I have to ask you, you have a Bill Parcells–signed NFL football there?

Nestler:         Yes I do.

Steve:         How did that wind up in your office?

Nestler:         Well after last night's football game, I'm sorry to say that…

Steve:         Shouldn't have brought it up.

Nestler:         I am a New York Giants fan and when I left Yale after many years, my colleagues bought me an autographed Bill Parcells football. It's actually from Ron Duman, who's a close friend and colleague of mine at Yale, who is a Pittsburgh Steelers fan, but I think, appreciated my being a fan of the Giants.

Steve:         He's got more Super Bowl rings…

Nestler:         Yes.

Steve:         So he's okay with it.

Nestler:         Exactly.

Steve:         Well that's interesting, I'm sure we could do a lot of epigenetic research on NFL football players.

Nestler:         Oh yes. Oh yes, indeed.

Steve:         We'll be right back.

Voice:          Station this is Houston, are you ready for the event?

DiChristina:         Hello I'm Marietta DiChristina, editor in chief of Scientific American. Win the space competition of a lifetime. See your experiment done on the International Space Station. Get a zero-g flight, specialized astronaut training and more. To find out how, go to

Voice:        Lift off.

Steve:         If you're regular listener, you've probably noticed that we've been conspicuous by our absence. We've been away for a while, but we plan to be back with a vengeance with plenty of new episodes. In addition to our usual long-form interviews, we'll be rolling out additional shorter episodes, and look for the imminent return of our quiz, TOTALL……. Y BOGUS. Now let's check in with Kerri Smith, to hear what's on the next Nature podcast.

Smith:         On the Nature podcast this week, Egyptian archeology after the revolution; how the immune system helps keep us warm, and what the Earth's fondant filling is made of.

Steve:         Which you can get to on iTunes and at

Well that's it for this episode. Get your science news at, where you can read Alex Wild's piece about why that particular image of the pepper spraying policeman is already iconic. It's called "Why One Pepper Spraying Cop Image Dominates." And follow us on Twitter, where you'll get a tweet every time a new article hits the Web site. Our Twitter name is @sciam. For Science Talk, the Scientific American podcast, I'm Steve Mirsky. Thanks for clicking on us.

Share this Article:


You must sign in or register as a member to submit a comment.

Email this Article