Well, thank you all very much for joining us today, and I appreciate the opportunity to speak to you about some of the work that we've done related to transcriptional and epigenetic mechanisms of depression. What I'd like to do today is to illustrate how we've used genome-wide data sets to gain unique insight into the biology of depression and other stress-related illnesses. I'll begin with a very brief overview of what I mean by transcriptional and epigenetic regulation. I'll then have several themes on which to focus. One is to highlight new information about the dramatic sex differences that occur at the transcriptional level with respect to depression. It's been known for many years that depression is twice as common in women than men, but little has been known about the underlying mechanisms. Secondly, I wanted to highlight the distinction between what we refer to as stress susceptibility versus stress resilience. What are the mechanisms that govern an individual's own unique inherent vulnerability to different forms of chronic stress? And I'll talk about that by illustrating two examples, one a pro-resilience factor and another a transcription factor that mediates the effects of early life stress. First, let's review transcriptional and epigenetic regulation. What I'm depicting here is the DNA double helix. We know that a mammalian organism has about 3 billion nucleotides stretched out linearly would be about two meters of DNA. Yet somehow this DNA is compacted to an extraordinary degree to fit within a microscopic cell nucleus. And we've learned a great deal about that compaction, which describes the field of epigenetics. So we know that the DNA double helix is wrapped around octamers of histone proteins to form the unit of chromatin called a nucleosome. Chromatin is simply the word given to the material within a cell nucleus. The nucleosomes are then further organized in very rote ways in the most compacted version to form what we would recognize as a chromosome. And this organization or the degree of compaction of DNA is critically important for function because spans of DNA in regions of chromatin where the DNA is tightly packed cannot be functional. However, in spans of DNA, spans of chromatin where the DNA is more open, it's accessible to the transcriptional machinery and can be functional. And for that reason, we can use information about the chromatin architecture at specific parts of the genome to help us in many ways. This work can help us identify the specific genes that are affected both in human depression as well as in animal models of stress. Secondly, this information can provide the first ever glimpse into transcriptional mechanisms that occur in the brain in vivo. Prior research has focused, had to focus mechanistic studies on cell culture. Studies of chromatin now enable these mechanistic studies to be done on the intact brain. And by analogy with the cancer fields and developmental biology fields where certain types of epigenetic modifications that control the opening and closing of chromatin, as depicted in this cartoon, once they occur are permanent. The idea is that perhaps the same thing occurs in the brain in response to behavioral experience. That certain types of epigenetic modifications would convert open to closed chromatin or closed to open chromatin, thereby producing very long lasting changes in expression changes in genes. I'm going to dive one level more deeply into this information focused on this level of analysis where now these gold stripes are the DNA double helix wrapped around the octamers of histone proteins indicated by these spheres, showing the two ends of the spectrum of chromatin open, active, closed, inactive. And we've learned a great deal at the detailed molecular level at what governs the transition from one to the other. A type of protein called a transcription factor will bind to spans of DNA that are relatively open, binding to specific nucleotide sequences in that available DNA. Once a transcription factor, and I'm just listing examples of transcription factors that we've been interested in, like Delta FosB, CREB, beta-catenin. Once a transcription factor binds to its specific sequence in open DNA, it then recruits many coactivators to induce activation of that gene's transcription. I'm just illustrating one example, the recruitment of an enzyme called a histone acetyltransferase, which would add acetyl groups to histones, which would serve to move nucleosomes even further apart, recruit many other activators, and eventually the basal transcription complex containing RNA polymerase II to mediate transcription. Now, all told in simpler systems like yeast cells or stem cells, one or 200 different types of proteins might be recruited to a single gene in concert with its activation. So, these are very complex processes. Similar complexity mediates the condensation or inhibition of gene expression, where recruitment of many types of repressor proteins serve to condense chromatin and suppress gene transcription. Just one example might be a histone deacetylase, which removes these acetyl groups. The approach we've taken, then, is to let the biology guide us into what genes would be most important to study in the most open-ended, unbiased way possible. We would begin, for example, by using a technique called RNA sequencing or RNA-seq to identify sets of RNAs that are upregulated or another set that are downregulated, either in human depression or in stress models in mice. We would then overlay on top of those RNA-seq data results from a related approach called chip sequencing, which would analyze different types of histone modifications. I mentioned histone acetylation, there's histone methylation, DNA methylation, and so on. If a gene is transcriptionally activated to induce an RNA molecule, that change would be associated with either an increase in an activation or a decrease in a repressive mechanism of chromatin regulation. And then we would use chip sequencing to overlay on top of that genome-wide measures of the binding of these or other types of transcription factors. This is an enormous undertaking. All told, the data shown on this one slide, one time point, one type of stress, one brain region would involve many terabytes of sequencing information. Yet we believe that this is the right approach to take because when one digs into areas of overlap on these Venn diagrams, one can validate the genes identified with far greater accuracy than a reliance on one platform alone, say RNA sequencing or chip sequencing alone, each platform associated with relatively high false negative and false positive discoveries. Now, in talking about a syndrome like depression, what I've just explained to you is an extremely reductionistic approach focusing on the cell nucleus, the organization of DNA within chromatin, within that nucleus. But what I want to emphasize is that this reductionistic approach would provide an unbiased comprehensive analysis of all of the genes present in different cell types in the brain that are being altered in depression or in stress models in animals, giving us a complete collection, complete template of all the neurotransmitter receptors, ion channels, structural proteins, and all other classes of proteins that are being altered within specific regions of brain, which we could then use to study the functional consequences of this very basic transcriptional and epigenetic regulation. I'll come back to this at the very end of my talk. In order to carry out this work, one needs to know where in the brain to look for transcriptional and epigenetic regulation. Here we've relied on decades of research that I've identified broad regions of forebrain as being important, including regions of prefrontal cortex, the nucleus accumbens, which is a major reward center in brain, in the temporal lobe, the amygdala and hippocampus, the hypothalamus, and the innervation of all of these regions by the brain's monoamine systems, dopamine coming from the VTA, and norepinephrine and serotonin coming from the locus coeruleus and dorsal raphe. We also need to rely on chronic stress models in rodents. Now, almost all advances in biology, in medicine have relied on the use of animal models, yet this slide illustrates the unique challenges in animal models of a psychiatric disturbance. To this day, 2021, we still can diagnose a mental disturbance, depression, or any other. Solely based on behavioral abnormalities, we do not have any genetic test, blood test, or brain scan to assist with diagnosis. Yet the very behaviors that we ask our patients about are inaccessible in rodents as depicted as a joke on this slide. So this is, I think, one of the main reasons why progress in psychiatry has lagged behind that in most other medical specialties. Yet despite the challenges, we believe that it is possible to study stress susceptibility in resilience in rodent models in ways that are relevant to human depression, and that's what I will tell you about now. So one of the models that we've looked at is called the chronic social defeat model of stress susceptibility and resilience. In this model, we take a test mouse, a C57 mouse, place it in the cage of a bigger, meaner mouse and a retired breeder ICR mouse. Fighting ensues immediately. We limit the duration of that fighting to just a few minutes to limit physical injury and then separate the mice with a screen so that the test mouse is subjected to all the aggressive cues of the larger mouse for the rest of the day. And we repeat this period every day for 10 days, that test mouse being exposed to a different aggressor. And what we have found over the years is that we can induce, in a genetically normal C57 mouse, a behavioral syndrome characterized here. This is essentially my DSM criteria for the behavioral syndrome of chronic social defeat stress. And I want to credit Olivier Bertand, who's now at NIDA, and Nadia Tsankova, who's with us at Mount Sinai for introducing this paradigm in my lab. We say that the mice have anhedonia, that's a loss of interest in pleasurable activities. I think we can measure this well in a rodent. The rodents show less interest in drinking sugar water, eating high fat food, having sex, running on an exercise wheel, and so on. We say that the animals exhibit anxiety-like symptoms. I think that's a huge inference. Really, what we're measuring is exploratory behavior. After social defeat, the mice explore less. There's hyperactivity of the hypothalamic pituitary adrenal axis, a hyperglucocorticoid state. There's a disruption of circadian rhythms, increased liability to self-administer drugs of abuse to themselves, an interesting metabolic syndrome where the mice become obese, they gain weight even though they seem to enjoy eating less, they develop a profound social avoidance, and we can treat these symptoms with antidepressants. We believe that the social defeat paradigm is uniquely useful to study human stress-related disorders for a variety of reasons. First, only a subset, maybe two-thirds, of C57 mice develop this full range of symptoms. We call those mice susceptible. The remaining third still show the anxiety-like symptoms or the decrease in exploratory behavior, but none of the other abnormalities that I talked about. They are relatively resilient or resistant to the bad effects of stress. Also, whereas most other forms of stress only produce behavioral changes for a few days once the stress is alleviated, chronic social defeat stress produces changes that are essentially lifelong. Many of these behavioral abnormalities last for at least six months, which is a long period of a mouse's life, the longest that we've looked. Unless these behavioral abnormalities last a long time, we can study whether antidepressants can treat the symptoms. This is different from most other chronic stress paradigms in rodents, where the antidepressants and the stress are given at the same time, testing really whether the antidepressants prevent the effects of stress, who we can really show whether the antidepressant is reversing the effects of stress, much more akin to the human situation. As I already mentioned, we can reverse the symptoms of lifelong susceptibility with the same range of antidepressant medications that work in human, chronic administration of tricyclic antidepressants, SSRIs, and so on. Anxiolytic drugs like benzodiazepines do not work, and single doses of ketamine do work about as effectively as the standard antidepressants. We think that these features of the chronic social defeat paradigm make it quite relevant to human stress-related disorders. Let me illustrate the ways in which now we've begun to use the open-ended datasets that I talked about at the outset within the context of this chronic social defeat model. This is work done by Rose Baggett in the lab, who subjected mice to 10 days of chronic social defeat stress, our standard protocol, took four brain regions, nucleus accumbens, medial prefrontal cortex, basolateral amygdala, ventral hippocampus, and two days after the last stress carried out RNA sequencing and identified genes in animals that are either susceptible or resilient to that stress. The heat maps show genes that are up-regulated in yellow or down-regulated in blue. For those of you who are not familiar looking at these heat maps, each of these vertical lines represents a single RNA. The genes, I'm sorry, the Venn diagrams at the right show the same data, only as Venn diagrams, where we are now illustrating the total number of genes that are being regulated in resilient animals, up or down, versus the number of genes that are regulated in susceptible animals. So, as you can see from the Venn diagrams and also from these heat maps, then in most of the brain regions studied, more genes are actually affected in resilient animals than in susceptible animals. Actually, the unique exception is the basolateral amygdala. This kind of high-level analysis is interesting to us because it shows that susceptibility in most brain areas is not mainly, predominantly, I'm sorry, that resilience is not predominantly the failure of an animal to show the bad effects of stress, but rather represents a more plastic state. But susceptibility, by contrast, may represent the failure of adaptive plasticity, at least in most of these brain regions studied. Now, I'm very mindful of the limitations in applying an animal model to depression, as that earlier picture of the rats showed. And for that reason, we have paid increasing attention over the last few years to always validate anything that we do in an animal model in the human syndrome. So, I wanted to draw your attention next to a large RNA sequencing study that was performed in my lab by Benoit Labonté, Benoit is now at Laval in Quebec City, where he took six brain regions of 100 humans, half-male, half-female, half-depressed, half-not-depressed, so a relatively small data set for heterogeneous syndrome-like depression, but the largest RNA sequencing data set still available. And what I'm showing you here are the genes that are upregulated in yellow, downregulated in blue. For example, in a region of medial prefrontal cortex, this is Broadman area 25, and how those same genes are affected in all the other brain regions that we analyzed. And the same thing for females. By far, the predominant high-altitude pattern that we noticed in this data set was a dramatic difference between the genes that are abnormal in male depression versus female depression, where there was only 5 to 10 percent overlap in the transcriptional abnormalities in men and women associated with depression. We did a lot of bioinformatic work to validate that this is, in fact, a bona fide result. We do believe it. And to gain even greater insight, we went to compare sex differences in stress responses in mouse models. And for this, I wanted to introduce still a different chronic stress model in mice, where we can expose mice to 21 days of chronic variable stress, exposing mice to a different stress each day. And at the end of the 21-day period, show that the male and female mice develop the same behavioral abnormalities. And what we were able to find are, again, a series of transcriptional changes associated in several brain areas as a consequence of the stress in males and females, but, again, very little overlap between the sexes, highlighting the fact that the stress responses we believe are fundamentally different at the molecular level between males and females, suggesting that stress-related disorders are fundamentally different in men and women, really driving the need for separate drug discovery efforts. I'll get back to that as we move along. Having these broad datasets also help us validate these animal models of stress that we've been relaying to the human syndrome. So we can show here, we can now, rather than argue whether a behavior exhibited by a mouse is similar to a behavioral abnormality exhibited by a depressed human, we can ask, to what extent are the transcriptional changes induced in each of these brain regions by different forms of chronic stress in a mouse overlapping with what we see in the human syndrome? You can see that each of these models can replicate a significant portion of the molecular pathology in the human syndrome, although a different portion largely for each chronic stress syndrome, perhaps consistent with the idea that each type of chronic stress is eliciting a different subtype, a different flavor, a different domain of abnormalities that the broad human syndrome captures. And this overlap with human depression is not seen between our stress datasets in mice with these other disorders relaying specifically to a stress-related disorder like depression. So now let me give you an idea of how we've utilized these datasets driving causal information. And I wanted to do this by telling you about one particular class of transcriptional regulation in what we call long non-coding RNAs. These are a relatively recently discovered class of RNAs. They were absent on all previous chip-based methods. We only detect them now because of the availability of sequencing. They're defined as being 200 nucleotides or more in length, and they have the same structural properties of RNAs or transcripts that encode proteins. Long non-coding RNAs play different types of roles across biology. They're defined based on their relationship to protein-coding genes. For example, long non-coding RNAs that are called links, long intergenic long non-coding RNAs are those that show no overlap with protein-coding genes as just one example. We think that long non-coding RNAs might be particularly important for psychiatric syndromes because they are more prevalent than protein-coding genes in the human genome. They're highly enriched in brain, and many arose uniquely in the primate lineage. So when we take now this human data set that I mentioned earlier and looked only at long non-coding RNAs, and this is work done by Orna Isler, who's with us at Mount Sinai, it turns out that about the same number of long non-coding RNAs are regulated in human depression, male and female, compared with protein-coding genes. These are the heat maps of these data showing only now the long non-coding RNAs. Once again, just as for protein-coding genes, there is virtually no overlap between the long non-coding RNAs that are altered in females versus males. We wanted to understand how it is that these long non-coding RNAs are affecting depression, and we needed to identify where to start. Again, we used the biology to guide us. So we started out by developing an algorithm to gear us toward the first long non-coding RNA to study. We'll eventually study the rest. We asked, of all the long non-coding RNAs that are expressed, which ones are more correlated with protein-coding genes might be more functional or more easily studied as being functional. We then asked which ones are expressed at higher levels. We asked which ones are differentially expressed as a result of depression. You can see a preponderance of changes in female brain versus male brain. Which of these are enriched in neurons, again, for ease of study, and which ones of these neurally enriched genes are also enriched in neural progenitor cells, for the reasons that I'll illustrate in just a couple moments. That left us with one long non-coding RNA called LINC473. You can see in our data set that LINC473 is downregulated in all four brain regions, all cortical regions, not downregulated in nucleus accumbens, a change only seen in female brain, not in male brain. We validated this in a second cohort, downregulation of LINC473 in the ventral medial prefrontal cortex, Broadman area 25. In a second cohort of depressed females, no significant effect in males in that cohort either, although there was a clear trend. And in a third cohort, we worked with Carol Tamenga and Chun-Feng Tang at UT Southwestern to use fluorescent in situ hybridization to quantify levels of LINC473 RNA in cortex. Found, again, a reduction in LINC473 in the depressed female cortex, an effect not significant in the male cortex. This also enabled us to identify the cell types that show LINC473 staining indicated by these double puncta, which mainly are pyramidal neurons. I'll come back to that in a moment. So remember I mentioned that LINC473 is only expressed in primates, it's not expressed in a mouse. How do we study its function causally? Well, we still hypothesized that whatever function LINC473 plays in a human, it is through its linear RNA sequence, and that there would be the targets for that sequence in mouse brain, even if mice do not normally express that LINC473 sequence. So what Orna did was to develop a viral vector to overexpress LINC473 in ventral medial prefrontal cortical neurons of a mouse and asked what effect that had on the mouse's inherent vulnerability to stress. What she found is that when she overexpressed LINC473, she found that a period of chronic social defeat stress that induced social avoidance in a normal female mouse was no longer susceptible to that stress when it overexpressed LINC473. That LINC473 was pro-resilient, consistent with the fact that a loss of LINC473 is seen uniquely in depressed women. This effect of LINC473 overexpression was not seen in male brain, and this carried out in the other form of stress that I mentioned, chronic variable stress, where with several behavioral paradigms, the effect of chronic variable stress in a normal female, shown in these white bars, is not seen upon overexpression of LINC473, effects not apparent in males. So LINC473 is uniquely pro-resilient in females. We then asked, how is it that LINC473 is producing unique effects on behavior in females and not in males when overexpressed in mouse prefrontal cortex? We turned to physiology, collaborating with Yan Dong and his colleagues at the University of Pittsburgh, we overexpressed LINC473 with viral vectors in ventromedial prefrontal cortical neurons, cut slices recorded from those pyramidal neurons and measured their excitability. What I'm showing you here are recordings from brain slices from female mice, showing that there's no difference in the excitability of pyramidal neurons that are not affected with virus, that are infected with a control virus, but those neurons that are overexpressing LINC473 show reduced excitability, reduced frequency, and reduced amplitude of excitatory postsynaptic currents. Notice that this effect of reducing the excitability of prefrontal cortical pyramidal neurons in females is not seen in males, even though the degree of overexpression is the same. And what's so interesting to us is what's highlighted here in red, and that is that the reduced ventromedial prefrontal cortex pyramidal neuron excitability seen with LINC473 expression is consistent with several other manipulations where we show that when we go in and decrease this neuron excitability, we induce a pro-resilient behavioral effect. Now, to gain further insight into how it is that LINC473 is producing these electrophysiological and behavioral effects in female mice, we turn to RNA sequencing to examine all the genes whose transcription in ventromedial prefrontal cortex is being altered by LINC473 expression. So what we did in this case is subject mice to one type of stress paradigm, in this case chronic variable stress. I'm showing you data for female mice only in this case. We can identify a set of genes that are upregulated in yellow, downregulated in blue after a period of chronic variable stress in normal animals, animals injected with a control vector. And notice how expression of LINC473 dramatically blunts the genes that are upregulated and dramatically blunts the genes that are downregulated by chronic variable stress. You can see that in the Venn diagram where several thousand genes are normally altered up or down in this brain area by chronic variable stress, reduced to almost one-tenth that amount in the presence of LINC473. Again, consistent with behavioral effects that what LINC473 is doing in a mouse brain is preventing the majority of transcriptional changes that chronic variable stress induces in this brain region consistent with a pro-resilient effect. Now we're moving one step further by analyzing neural progenitor cells. Instead of expressing LINC473 really to gain more robust information, we would want to knock it out. Yet we can't knock it out in a mouse because it's not normally expressed in a mouse. So in order to have a knockdown system, we need to use human tissue. So what we did, again, work from ORNA, collaborating with Kristin Brennan, is to obtain neural progenitor cells from control human subjects, both females shown in purple and males shown in green. These induced pluripotent stem cells are converted to neural progenitor cells, and then we induce a knockdown of endogenous LINC473 using an antisense oligonucleotide. And what I'm showing you in the upper right is the ability of this antisense oligonucleotide to knock down normal levels of LINC473 to about the same extent in female-derived and male-derived cells. Yet when we do RNA sequencing on these cells, we show regulation of close to 800 RNAs in female-derived cells, both up and down, with very few genes being affected in male cells. So for whatever reason, when we knock down LINC473 from male cells, it has very limited transcriptional consequences, even though basal levels of expression of LINC473 are roughly the same between male and female cells. So what we're doing now to follow up is a technique called CHIRP, Chromatin Isolation by RNA Purification, to identify the direct binding partners of LINC473. Orna has preliminary evidence that LINC473 is binding to both genomic and protein targets. We're interested in identifying those targets to figure out how LINC473 is exerting its pro-resilient effect. Secondly, we are very interested in understanding the sex-specific nature of this regulation. Two questions. Why is it that LINC473 is selectively downregulated in depressed females, not in depressed men? And then secondly, in both human and mouse tissue, why is it that LINC473 produces so many more dramatic effects in females than in males? Is this due to chromosomal differences between males and females, to hormonal differences? We need to understand what is driving that fundamental sex difference. Nevertheless, the sex difference that I've highlighted illustrates one very critical point, and that is that if the transcriptional features of depression in humans and stress responses in mice are so fundamentally different between males and females, as I've shown, why is it that both the behavioral presentations and treatment of depression are, to the most part, similar in humans? And I think there are several answers to that question. One is the fact that even though there's only 5% or 10% overlap, that 5% or 10% overlap is still several hundred genes, and the therapies that we've developed historically have looked at both sexes, so not surprisingly, we have found treatments that work generally as well in men and women. Also, even where different genes may be affected, some of the same gene pathways or biochemical pathways may be affected. And finally, even in some cases where a gene is affected only in females or only in males, its manipulation may produce a therapeutic effect in both sexes, although that's not the case for LINC473. It is the case for many other sex factors, and that's something that would further explain this question. Nevertheless, we believe that depression, human depression, is fundamentally different at the molecular level between males and females, and really drives now a drug discovery effort that pays attention to these differences. In the remaining few minutes, I would like to tell you about one other area where we've seen dramatic sex differences, and that is in the area of early life stress. We know that stress early in life, adverse life events, is one of the greatest risk factors for human depression, but the underlying mechanisms remain unknown. One of the unique features that we wanted to recapitulate in a rodent model is to subject animals to early life stress, but in a milder way, where the effects of stress are latent and only become apparent when the animals are subjected to a second hit of stress in adulthood, which seems to replicate what is seen in many humans. So this is work done by Kate Pena, who's now at Princeton, where she identified a certain period of pre-weaning development, which I'm referring as early life stress II, second half of pre-weaning phase, where when we subjected these mice, pups, to maternal deprivation, basically separated them from their mothers, just look at this middle composite behavioral score, we were able to make these animals more susceptible to a second hit of stress later in life. These are in male pups. The same effect was seen in female pups, that the combination of this early life stress II and a second hit of stress revealed susceptibility in female mice. We performed RNA sequencing on multiple brain regions, showing a very interesting pattern in all three brain regions in both sexes. Notice how early life stress produces changes in adulthood, so these are lifelong changes now that are induced in an adult that seem to replicate the effects of stress in adulthood. So these are animals that are raised normally exposed to stress in adulthood. These are animals that are exposed to early life stress, then allowed to age until normal adulthood. Notice the very similar transcriptional responses seen, almost as if early life stress primes a stress-like state in adulthood, consistent with the fact that the behavioral state of these animals is similar, both are primed for greater stress susceptibility. However, when we look molecularly, we still see this traumatic sex differences. So again, the general theme is the behavioral responses are similar, the behavioral patterns are similar, the overall transcriptional patterns are similar, but the specific genes that comprise those transcriptional patterns are mostly different between the two sexes. This is my last data slide. I just want to quickly tell you about one gene that we have identified as driving this priming effect of early life stress. This only holds for one brain region, the ventral tegmental area, and only in male mice. This is OTX2. It plays an important role in the development of VTA dopamine neurons, but never before studied in stress or depression. And we showed that it is the main factor that underscores the commonality of regulation mediating the effects of early life stress for a lifelong change in stress susceptibility, leading OTX2 to be part of what we call a chromatin scar, tagging sets of genes early in life to remain different for adulthood so that in response to a second stress, these genes now respond differently, underlying behavioral susceptibility. I'm not going to tell you more about OTX2 in the interest of time. Suffice it to say that we're now performing chip sequencing on OTX2 and on multiple types of histone modifications to understand the identity of these chromatin scars and have a better understanding of the biological basis of how early life experience reprograms an individual for a lifetime through epigenetic mechanisms. So to conclude, I want to make a few points. I focused today on using genome-wide datasets to provide unbiased information about mechanisms of stress susceptibility and resilience. We're very interested in mechanisms of natural resilience. While the field has focused on ways to undo the bad effects of stress in susceptible individuals, perhaps replicating mechanisms of natural resilience represents an additional path for antidepressant drug discovery. We focused on dramatic sex differences, effective early life stress, and we're also interested in antidepressant action. We believe that chromatin changes are one key feature of how behavioral experience produces long-lasting regulation, and of course, we're very interested in utilizing this very reductionistic approach to drive an improved understanding and treatment of the human syndrome. So let me just give you one last point. I talked about how this reductionistic approach can provide insight into receptors, channels, and other proteins affected in depression. Well, it turned out that in some earlier work, and this was done over a decade ago by my lab, work by Chris Krishnan and Ming-Hu Han, looking at the ventral tegmental area, we found unique induction of a series of potassium channels in resilient mice, no effect in susceptible mice. We validated this finding in several ways, importantly showing that small molecule potentiators of KCNQ channels produced a pro-resilient effect in mice. Well, with our colleagues at Mount Sinai, and it took us 10 years to get this study done, we now have been able to show similar pro-resilient or antidepressant effects of one particular KCNQ potentiator called azogabine, which is in the physician's desk reference as an anti-seizure medication, producing an antidepressant effect in an open-label study, and consistent with that, producing changes in functional brain imaging in the reward circuitry as a measure of anhedonia, correlating the behavioral change in antidepressant effect with this change in brain activity. And I'm very happy to say that James and his colleagues now have a study in press in the American Journal of Psychiatry that has replicated this antidepressant effect of azogabine in double-blind clinical study. So this illustrates how one can indeed take a reductionistic work that I've talked about so far and use it to derive information that is useful for clinical application. And with that, I want to once again thank the organizers for having me speak here today, and I'm very sorry that I can't be doing this in person, but now we'll end the recording and go live, and I'd be very happy to answer questions that you have for me. Thank you very much.

transcriptional regulation

epigenetic mechanisms

depression research

gene expression

sex differences in depression

stress susceptibility

RNA sequencing

chronic stress models