Welcome to the session on mechanisms of comorbidity. I'm Tristan McClure-Begley. I am the Chief of the Integrative Neuroscience Branch at the National Institute on Drug Abuse. And I'd like to just give a few opening remarks to sort of set the narrative for our session and also to remind everybody that when we get to the Q&A portion of the presentations to please use the microphones in the aisles so that the recording captures your question. So when we think of comorbid psychiatric conditions, I think it's really important that we consider the power of perspective and how these things are grouped and how they appear to us based on how we're looking at them. In this example here, are we seeing three people in one boat, or are we seeing a different distribution based on our position and our perspective how these things are separated in time and space? Or is it really just a massive regatta and we have no hope of ever figuring out exactly what's going on? But really, what we're considering here is the degree to which different diagnosable entities are occupying the same individual, whether they be separated in time or co-occurring. And what do those coexisting conditions and the overlap of the features that are used both in their diagnosis as well as the discrete quantitative features that can be measured from these individuals along the course of their conditions? What does that tell us about the physiology? So it is undeniable that these things do, in fact, exist. What is an interesting research directive is looking at the nature of co-occurring conditions and their relative position in time. Does one thing precede another? What are the features of one diagnosable condition that precede the onset of another? Are they, in fact, distinct entities? Or are they just revealing themselves differently as part of another physiological maturation process? All of these things point towards particular types or physiological mechanisms that maybe have yet to be explored for therapeutic and diagnostic benefit. So with the bottom line up front here, we're going to take the position that psychiatric comorbidities are prevalent. I'm not going to bother bombing everybody with statistics except to point out that comorbid disorders are more often the rule rather than the exception. This is particularly true when it comes to substance use disorders, which are one of the more prevalent modifying co-occurring conditions. In many cases, when you have co-occurring disorders, they are associated with more severe symptoms and a more depressing clinical trajectory. Also that there is no either or. It is certainly possible that you're able to identify comorbid conditions within an individual based on diagnostic criteria overlap, but also that more than one diagnosable condition exists in the same person. They both can happen. And that recognizing the temporal patterns of comorbid features is key to research into underlying physiological mechanisms, and maybe for effective clinical management. So just to make some points here before we get into the talks, when we refer to mechanisms of comorbidity, what we're talking about here are the shared physiological systems and or the features of diagnosable conditions that impact the existence, the trajectory, and the management of those conditions. When we say things like causal, we don't mean exclusive. There can be more than one discrete physiological or molecular mechanism that converges on a common phenotype or behavior. And with that, I think I'll hand it over to Dr. Raj. Please take the stage. Hello, everyone. I'm Raj Gaurishankar. I'm a postdoctoral research fellow at the University of Washington in the laboratory of Michael Brukus. I'd like to thank Tristan for the wonderful introduction and for inviting me to take part in this panel and for you all for being here. What we're interested in in the lab and what I hope to do in my own research program in the future is try to understand how neuropeptides in the brain, specifically endogenous opioids, opioids our own brains make, how they, and that's pictured on your right here, imaged through a microprism using a two-photon microscope in a mouse, and how time scales at the level of seconds of this endogenous opioid can then influence circuit activity of the brain region that these opioids are being released into, as shown again in the same mouse using the same microprism using two-photon microscopy. So ultimately, we'd like to understand how endogenous opioids control flexibly, if you will, provide gain control for circuit activity and how that then refines behavior. Now, why am I talking about endogenous opioids? Why am I talking about specifically dynorphin as an endogenous opioid in a mechanisms of comorbidity panel? In the late 2000s, George Kub and others posited that the escalation in drug seeking and the genesis, if you will, of substance use disorders is due to a reduction in the affect, behavioral affect, as shown in the black trace right there, where as you get negative affect, drug seeking escalates. And that's what results in pathological drug seeking and substance use disorders, as opposed to the hedonic or the pleasurable components of what an animal or a person would obtain by seeking out a particular drug. Now, why dynorphin came into play is there has been a lot of preclinical research showing that if you antagonize the cognate receptor for dynorphin, the kappa opioid receptor, in mouse, rat, and monkey models, you can actually get a reduction in the escalation of drug seeking and a reduction in the relapse to drug seeking induced by stress. Now, that led people naturally and rationally to posit that the dynorphin kappa opioid system is actually involved in having an inverse relationship with behavioral affect, i.e., when negative affect goes up, you get an increase in dynorphin. And by blocking that, you can then reverse, if you will, these negative affect-related symptoms. This has led to the development of several short-acting, reversible kappa opioid receptor antagonists that are at multiple stages of clinical trials right now, yielding promising results. But for us as basic scientists, we began to take a step back and wonder, wait, we haven't actually tested the hypothesis of when dynorphin is exactly involved. Does dynorphin control drug seeking in some way? If you remove it from affect, and if you look at what dynorphin does during seeking of a reward, is it even involved? Where exactly in the brain does this process happen? Is it even involved, if you take a step back, in natural reward seeking? Forget drug seeking. If an animal seeks a natural reward, is there a mechanism set in place for dynorphin to control that process that may then be co-opted during the drug seeking process? And we think all of these questions would provide a useful window into therapeutic intervention. So now, what do we know about natural reward seeking in the brain? What many studies in the past have found is that the basal ganglia, and nestled in it the dorsomedial striatum, is actually essential for reward seeking. I.e., when I say reward seeking, I mean the dynamic process of us learning to perform a particular action to obtain a desirable outcome. And we can train animals to do this in a very simple task, which I'll show you in a bit. What we found is, within the dorsal striatum, there's multiple populations of cells, but a specific population of cells, called the D1 medium spiny neurons, D1 medium spiny neurons are essential and integral to this process, where as an animal learns this action outcome strategy, or this reward seeking behavior, you get an increase in the activity of these D1 medium spiny neurons. But what we don't know, and what we already know, is that over 95% of these D1 medium spiny neurons also express dynorphin. But what we don't know is what this dynorphin actually does. This dynorphin has largely been used as a marker to identify these cells, but its process in basic behavior is unknown. And what's been posited is that this dynorphin can either act in a local circuit feedback inhibition mechanism, where dynorphin binds to the kappa opioid receptor, thereby inhibiting vesicular release of whatever neurotransmitter that cell population releases, or it can act in a feed forward mechanism, as shown with the blue diagram, where it can then act on the axons of a circuit coming into that brain region, thereby inhibiting activity of those axons, and thereby refining activity of the circuit in that brain region. So what we set out to ask is, is dynorphin playing a role in reward seeking by refining circuit activity of the dorsal medial striatum? And to do this, we trained animals in a simple operant behavior task. We collaborated with Lin Tian, who is a pioneer in biosensor development, to develop an endogenous dynorphin biosensor, which we can then use to look at the time scales and the dynamics of dynorphin release. We then deleted selectively this dynorphin from the dorsal medial striatum and asked what that did to behavior. Conversely, we activated dynorphin release at the time scale specifically where we saw dynorphin release and asked what that does to behavior. And finally, we asked, what does this dynorphin release, how does it act on circuits within or to the dorsal striatum to then refine reward seeking behavior? And so when I mean reward seeking behavior, what I mean is an animal learns to associate a nose poke, which is simple foraging behavior. It's very easy to train mice to do this, to then nose poke followed by a cue that tells the animal it's about to get a reward, following which the animal gets a reward. And here's a video of a trained mouse doing exactly that behavior, where the animal makes a nose poke, it gets a cue, followed by the delivery of a reward that it can then go ahead and consume. So to ask what dynorphin actually does, we looked at chelate 1.3, which is this dynorphin biosensor that we developed in collaboration with Lin Tian. And we first asked how dynorphin may be involved in behavioral affect, i.e. an animal's affect is clearly changing as it's learning in this session to associate an action with an outcome. And so as this animal learns operant behavior, as shown by this operant index, which is an index of learning, what we find is that in the early sessions, where the animal hasn't quite learned operant conditioning, i.e. this action-outcome strategy, there's not that much of an increase of dynorphin throughout the session. This is a 30-minute session. But what you find is once the animal has learned, you get a much bigger rise in this dynorphin across this 30-minute session, suggesting that dynorphin is acting as some sort of occasion setter, if you will, giving the animal important information about the context. Now, at shorter timescales, as this animal is dynamically performing this behavior, what happens? And here, the same behavior, you see the trace where the white trace is the dynorphin, and the blue trace is a control trace. What you actually see is that after the animal makes a nose poke, you get this increase in dynorphin in anticipation of obtaining the reward. You'll see that again in a subsequent trial. You'll see that again in a subsequent trial. The animal makes a nose poke, and you get a huge rise in dynorphin that then stops once the animal retrieves and then consumes the pellet. And now, this was fascinating for us as basic neuroscientists, because we've always thought dynorphin to act at much slower timescales. But the timescales you're seeing here are at the level of seconds, which we've normally only associated fast neurotransmitters like glutamate or GABA to do. But dynorphin here, an endogenous opioid, a neuropeptide, which is much larger and much stickier, is acting at much faster timescales as well. Now, if you graph this up and you look at when exactly this is happening, what you see is, again, in the early sessions, you don't see much of an increase in dynorphin at all. But once the animal is trained in operant conditioning, you actually, as you saw in the video, you get this increase in anticipation of reward of dynorphin. OK, is this purely just a consequence of the animal performing the behavior and the dynorphin's just there? Or is this actually causal to this process of operant learning? To ask that question, what we did is we used conditional dynorphin knockout mice, where we selectively knocked out dynorphin, or the precursor for dynorphin, from the dorsal striatum. And then we asked what that did to operant learning. And what we find is, when we do that, compared to the control animals, these animals have a much harder time learning operant conditioning, learning this action-outcome association. They eventually do. But even when they do, they don't perform actions or consume rewards at the level of the control animals. And we know, and this is important, that this isn't due to their innate preference for sucrose. When given sucrose in their home cage, they eat the same amount of sucrose as their control animals. And also, importantly, we find that we can rescue this behavior by just one systemic IP injection of U5488, which is a capital opioid receptor agonist. So again, suggesting that the dynorphin being there is important for providing information about the context that then helps the animal learn operant conditioning. Now conversely, and this is a bit of a blunt approach that I just showed you, what if we're able to fine-tune when exactly dynorphin is released to mimic what we saw in using the dynorphin biosensor, and ask if that is sufficient to now increase operant conditioning behavior? So we used optogenetics for this, where we can now stimulate these dynorphin neurons to produce dynorphin release specifically when we saw dynorphin release using the biosensor. And what we find is normally, when these animals are sucrose-naive, when they haven't learned to operant condition at all, we asked if stimulating these neurons is reinforcing by itself. And it's not, which is an important consideration to take. But once these animals have learned operant conditioning, you actually find a nice big increase in their operant index when you stimulate specifically the dynorphin neurons when we see dynorphin release using the biosensor. And importantly, what we find is we can reverse this increase by using a short-acting reversible kappa opioid receptor antagonist, atycoprine. So now I've told you that dynorphin is necessary and sufficient for operant learning and for operant behavior in mice. But how is dynorphin actually doing this? And to follow this up, we went with the feed-forward inhibition route, because there's a lot of evidence suggesting that local release of dynorphin onto circuits coming into the dorsal striatum, or any region, refines activity of those axons, thereby refining activity of the dorsal striatum thereby refining activity of that circuit entirely, resulting in behavior. And so one of our candidates was the basolateral amygdala, because there's an established pathway from the basolateral amygdala to the dorsomedial striatum that has already been shown to be involved in stimulus outcome associations, in reward-seeking behavior, and in determining outcome value. And what we also found is that about 60% of these BLA neurons that project to the dorsomedial striatum express kappa opioid receptors, and that's just doing that in reverse. So what do these BLA terminals do as an animal is performing operant conditioning? So to do that, we made use of GCaMP, which is a calcium-sensitive biosensor that fluoresces when there's an influx of calcium into the cell, which is a proxy for neuronal activity. And using fiber photometry, what we did was, now again, this is a trained mouse. And again, the white trace is GCaMP now, and the blue trace is a control trace. What you find is an animal is exploring. As it's approaching the no-spoke part, you get this nice big increase. And it makes the no-spoke. You get another increase. And it gets the Q-light telling it that it's going to get a reward. And as it goes and retrieves the reward, you get this nice big inhibition of activity at these BLA terminals. Again, this is another clue for us, suggesting that that dynorphin presumably is being released onto these terminals right around when we saw dynorphin being released, now binding to the capital opioid receptor, resulting in this inhibition in activity. And so if we graph that up, what we see is, in the early stages of operant conditioning, we see a nice big reduction in activity. That's when the animal actually consumes the pellet, suggesting that there might be some other mechanism at play here, presumably beyond dynorphin. But now when the animal is trained in operant conditioning, you get this nice big increase as an animal performs the operant task or the action, followed by a reduction in activity when the animal approaches, retrieves, and consumes the pellet. Now, so far, that cartoon that I've diagrammed in the top left is just a hypothesis, right? To be able to actually test this, what we did was we knocked out dynorphin selectively in the dorsomedial striatum, and then looked at activity of these BLA terminals in the dorsomedial striatum. And when we did that, again, we replicated the behavioral deficit, where we found that these mice did not learn operant conditioning as well. But what we now found is that that dip in activity that you see is severely and significantly blunted. And that peak is blunted as well, suggesting that that outcome-dependent inhibition is informing future action-dependent increases in this circuit. And that's something we're following up on. So to summarize, what I've showed you is as an animal learns an action outcome behavior to seek natural rewards, what we find is that you get an increase in dynorphin right around when the animal anticipates the delivery of the reward. And that increase then results in a reduction in the activity of the BLA terminals, presumably via dynorphin release binding to the kappa opioid receptors on these terminals. And then that reduction during the outcome then informs increases in activity when the animal learns to perform these actions to operant conditioning. With that, I'd like to thank the Brukus Lab, Michael, my PI, a whole host of undergrads without whom this would not have been possible, and our collaborators. Thank you so much. Thank you. Hey, good afternoon. So my name is Yann Miner. I am a research scientist in Marina Picciotto's lab at Yale University School of Medicine. And I'd like to thank Tristan and Michael Bigley for inviting me to give this talk. So as a preclinical scientist, I am interested in using animal models to investigate how cognitive dysregulation can lead to psychiatric disorders. But with a particular focus on stress-related phenotypes and mood disorders. But of course, cardiologic dysregulation is not unique to mood disorder and can be found in a lot of different conditions, including cognitive disorders and addiction, to cite a few. So understanding the cardiologic mechanisms that lead to mood disorders can also help us understand how the same mechanisms could lead to some of these other psychiatric conditions that therefore would be comorbid to mood disorder. So one of the first challenges when we are using animal models is, of course, that mood disorders are very complex, with many different traits that are different from patients to patients. And some of these traits are extremely unique to humans. So in order to try to model these disorders, we are following the so-called RDoC matrix put forth by the NIMH, where instead of trying to recapitulate everything, which is anyway not possible, we are trying to focus on elemental constructs, on relatively simple elements that we can easily measure, that are fairly naturalistic, and that we can compare between different species. And also an important aspect of this is that we are trying to look at these constructs not as a binary thing, where it's either healthy or pathological. But we are trying to look at it as a spectrum. And in order to evaluate our models, we are using different paradigms. And I will cite, for example, the tail suspension, where a mouse is hanging by the tail, and the forced swim test, where a mouse is put in a beaker of water that it cannot escape. And so this is extremely stressful. The animals will fight, will engage in active coping in order to try to escape. But eventually, they will reach a breaking point. And so these tests have, in fact, excellent predictive validity, meaning that if you give to these animals antidepressants that work in humans, you will increase the coping behavior. But the point I want to make here is that these are not tests of, quote unquote, depression. They are tests of antidepressant efficacy. And they are definitely not models of depression, because it's a five-minute test of a mouse that cannot escape a very bad situation. On the other hand, it is actually an interesting and very relevant paradigm if you want to provoke stress, if you want to provoke a coping strategy in an animal and see what are the mechanisms that are involved, and also if you want to really push these strategies up to a point where the animals will eventually give up. And so going back to the cholinergic dysregulation and depression, it's been known for several decades now that hypercholinergic tone in the brain can lead to depressive symptoms. So back in the 70s, Dave Janowski and collaborators found that when they injected patients with phytostigmine, phytostigmine is an acetylcholinesterase inhibitor, thereby preventing the degradation of acetylcholine. They found that these patients had an increase in depression symptoms. And in fact, it was something that had been known before that, back to almost the 50s. Because in populations that are living next to crops, to fields where there was a heavy use of pesticides, and many pesticides are in fact acetylcholinesterase inhibitors, there was an increase in, there was a higher prevalence of depression. And so on the right side, I just put one example in rural Mexico showing that in teenagers that are next to these fields with a lot of pesticides, there was an inverted, sorry, correlation between acetylcholinesterase activity and depression symptoms. But you could argue that, okay, this is iatogenic, it's induced by drugs. Is it physiologically relevant? What about in real life? And this is what this study tried to answer. So here, what we did was to use a PET tracer, a radioactive ligand that doesn't bind to acetylcholine, but it's binding to one of the most commonly expressed nicotinic, most commonly expressed cholinergic receptor, which is the alpha-4 beta-2 nicotinic acetylcholine receptor. And what we found here, and it's at the bottom, on the bottom line, is that depressed patients had less binding. So it was a bit paradoxical because we did not really expect that, and so what we did eventually was to look at post-mortem tissue. And what we found, in fact, is that, no, the number of receptor was, in fact, the same. And what you're observing here is, in fact, the competition of acetylcholine with the PET ligand, meaning that in depressed populations, the higher level of acetylcholine competed with the PET ligand, thereby leading to less receptor occupancy by this ligand in the brain. So this is confirming that in a, I would say, regular or normal, quote-unquote, depressed population, we can observe a hypercholinergic tone in the brain. So if we want to use animal model to tease out the fine molecular mechanisms involved in the connection between cholinergic dysregulation and mood disorders, can we recapitulate some of these effects in a mouse model? And the short answer is yes, we can. So here's one example of a tail suspension test where the animals are, again, hanging by the tail. They are coping, they are moving, they are trying to escape. And what happened when we injected the acetylcholinesterase inhibitor, Fisostigmine, was that the animals were not coping as much. And this effect was, in fact, dose-dependent, meaning that the more Fisostigmine, the less they were coping. And we could, in fact, completely reverse these effects by administering either nicotinic antagonist or muscarinic antagonist, so basically, antagonists that are blocking the two main subclasses of cholinergic receptors. And by the way, I just want to mention that these compounds, these antagonists, have been used in short and small clinical trials in patients unable to respond to classical therapy. And in these patients, there was an improvement of some of the depression symptoms. So now, what about the connection between stress and acetylcholine? So stress is a major risk factor for the development of mood disorders. And we wanted to understand whether there could be a connection with acetylcholine. So to do this, we used a biosensor, so basically, a little molecule that is fluorescent. And when it binds to acetylcholine, it's becoming more fluorescent. And we can put this sensor in the brain. So on the left, we actually infused it in the ventral and dorsal hippocampus, but we've done it in other areas. And we also implanted an optic fiber on the vicinity of that biosensor so that we could record the fluorescence as a proxy for the fluctuation of acetylcholine in the brain. And so the animals are alive and well, freely moving, and we just administered a very short foot shock just to stress them out. And what we observed, and you can see it in the middle, is that following that foot shock, there was a transient increase of acetylcholine. And in fact, these observations were not fully new because back in the 70s, researchers had used restrained stress where they really blocked an animal in a tube for about two hours, and they measured the level of acetylcholine in the limbic system, so the amygdala, the prefrontal cortex, and the hippocampus. And they found, similarly to what we saw, that there was an increase of acetylcholine during restrained stress. But here's the interesting piece of data. If you remove the adrenals, something you wouldn't do in a clinical trial, you can get the exact same release of acetylcholine. So immediately, it's telling you a few things here, is that first of all, stress can lead to an increase in acetylcholine completely independently of the HPA axis and the quote-unquote stress system. But what it's also suggesting is that given these data, we can potentially get some kind of vicious cycle or maybe something that I would call a self-perpetuating loop where acetylcholine can trigger the activation of the HPA axis. This is something that has been done, that's been observed, including in humans. But then, the activation of the HPA axis can lead to a physiological stress that in turn can also lead to more release of acetylcholine. And also, what it can suggest is that stress, acetylcholine probably mediates stress and it can be done via the activation of the HPA axis, but it's most likely also doing it via other networks, completely independent of the HPA axis. And here, it's opening also this idea that because we are targeting other networks, we are most likely not targeting networks that would be unique in the control of mood disorder, but most likely other psychiatric conditions that would be comorbid to mood disorders. So the next step for us was to figure out where in the brain does that matter. So we looked at different brain areas, but I will spare you the suspense, and we focused on the hippocampus by using a genetic strategy to knock down acetylcholinesterase activity. So we tried to replicate genetically what we did with the pharmacology. And here, what we observed is that when the mice had less activity of the acetylcholinesterase, so in theory, they had more acetylcholine only in the hippocampus, they were showing, for example, more avoidance in a light, dark box test. So this test is very simple. You have a black box that is safe on the animal and you have a bright box that the mice explore, but not too much. When the animals had the knockdown of acetylcholinesterase, they were spending more time avoiding the bright compartment. And in the TEL suspension, the Foreswim test, the knockdown animals showed less coping behavior in these two tests, suggesting that increasing the cholinergic tone only in the hippocampus was sufficient to decrease coping strategy. And I am a system neuroscientist, so the next step for us was to figure out what are the circuits in the brain that could be involved in this phenomenon. So here in green, you have the representation simplified of the cholinergic system, and the cholinergic neurons are grouped in these little bundles. And about 70% of the acetylcholine in the hippocampus, at least in a mouse, comes from the medial septum and the ventral diagonal band at the bottom. And so what we did here was to use a chemogenetic strategy so that we could, on demand, activate these cholinergic neurons projecting to the hippocampus. And in the end, this is what we did. These are the pathways in red that we activated on demand. And we didn't find anything. So that was a bit surprising here, because as I said, most of the acetylcholine comes from there. But if you look at this diagram, you can observe that these projections do not go just to the hippocampus, but to the cortical areas, and also the medial ebonyla, and other places. So what's possible is that we have, in fact, some kind of positive effect that is induced by the cholinergic stimulation of these non-hippocampal pathway. And so to investigate this possibility, we used about the same strategy, except this time we modified and we activated only the cholinergic neurons that are projecting to the hippocampus and nowhere else. And this time, that worked. We could completely recapitulate the same phenotypes we had observed with visostigmine or the acetylcholine esterase knockdown, where, once again, the animals showed more avoidance for a bright and scary environment, and less active coping in the foreswim and the tail suspension. And we also observed in another test, because we do a battery of tests, they were showing less social interaction, knowing that mice are normally very social. And so that suggests that hyperactivity of the cholinergic system, at least in the medial septum, may not be just a risk factor as is, because, as I mentioned before, the cholinergic stimulation of these other areas that are not the hippocampus can, in fact, have some kind of counteracting effect to the negative effect induced by the hypercholinergic tone in the hippocampus. And finally, we did also one small experiment where we targeted, this time, the very, very few cholinergic interneurons of the hippocampus, and we activated them using the same strategy. And we found basically exactly the same results. So mechanistically, what it suggests is that this is not the circuit that matters as much as how much acetylcholine you have in the hippocampus. And by extension, I think this suggests that mechanistically, we are more talking about something that is related to a slow volume transmission, maybe a global neuromodulation, as opposed to a very specific point-to-point connection between specific cholinergic neurons and receiving postsynaptic hippocampal neurons. And finally, I just want to finish on a few, a bit of data on the amygdala, just to make sure you don't think it's too simple. So we've done a lot of work on the amygdala, and actually much more than the hippocampus. And hyperactivity of the amygdala is known to be a strong risk factor for mood disorders. And we have done a lot of experiments where we, for example, blocked genetically, with pharmacology, the cholinergic receptors in the amygdala. And we could induce some very robust antidepressant-like effect. But what we wanted to know was, what are the cholinergic patterns in the amygdala in response to the scoping strategy during a tail suspension test? So once again, we used our little biosensor, detecting in real time acetylcholine, but this time in the amygdala. And then we looked at the data, and what we plotted, the phases when the mice were active, and the phases when the mice were just not coping. And what we found is that during the active coping bout, when the mice were mobile, was that acetylcholine was actually higher. And when the mice were immobile, there was less acetylcholine. So it seemed a bit paradoxical, especially because, of course, the amygdala is not the hippocampus, so it may be different. But I just told you that if we block the cholinergic receptors, we can induce an antidepressant-like effect. So it was a bit puzzling at first, but then we looked at the data in more detail. And this time what we did was to look at the fluctuation of acetylcholine a few seconds before and after a change of activity. So basically what you see on the right is when the mice were immobile, and then suddenly they start to engage into an active coping strategy. And not too surprising, we found that there was an increase of acetylcholine. But if you look closely, the T0 is when they start moving. And you can see that the deflection of acetylcholine stops before the behavior, suggesting that acetylcholine may not be the direct trigger, but it's at least part of a network that then lead to the active coping strategy. So if I try to recapitulate everything and put it into a broader context, what it suggests is that a quick and transient increase in acetylcholine may not be a bad thing. It's probably part of an adaptive mechanism that allows to engage proper coping strategy. But it's most likely the sustained elevation of acetylcholine, as can be observed during chronic stress, for example, that eventually can lead to the development of mood disorders. And most likely to the perturbation of multiple networks that not only will sustain these mood disorders, but most likely engage other pathways that will lead to other conditions that would then become comorbid to mood disorders. So in conclusion, acetylcholine dysregulation in stress-related phonocytes is fairly complex, multifaceted, and involves multiple neuronal pathways. But it's a sustained acetylcholine dysregulation, as observed in mood disorders, that most likely can lead to secondary alterations of other systems that may not even be cholinergic that will allow to sustain these mood disorders and trigger multiple comorbid disorders secondary to the mood disorder. And also for this crowd in particular, I hope I convinced you that the crosstalk between clinical and preclinical researchers can allow to suggest new causal mechanisms, neuronal substrates, network perturbation, and potentially even temporal pattern underlying the comorbid conditions observed in psychiatric disorders. And to finish, I want to thank again Dr. Michael Begley for inviting me for this talk, and all the people involved in the work, and especially my lab at Yale University, and our PI, Marina Picciotto. Thank you. Do I have it? There we go. Hi, I'm Diana Martinez. I'm actually a psychiatrist. And my talk will be far more clinical, start with clinical and get to a little bit of basic neuroscience a little bit. I've been doing clinical research in substance use disorders for about 20 years. And so I'm going to be talking to you today about conduct disorder and antisocial personality disorder. So I just want to start by pointing out that antisocial personality disorder is remarkably common comorbidity when it comes to substance use disorders. And this was a recent meta-analysis that was done looking at opioid use disorder. And it showed that 43% of men have antisocial personality disorder. This compares to about 4% to 6% of the general population. And in fact, in this meta-analysis, antisocial personality disorder was the most common comorbidity in opioid use disorder. When it comes to women, 29% of women had antisocial personality disorder, which is generally about 1% of the general population. And it was as common as PTSD and some other comorbidities in women. When we look at other disorders, such as alcohol use disorder, cocaine, and cannabis use disorder, it's about 15% to 25% of patients have antisocial personality disorder. So even though ASPD or antisocial personality disorder is very common in substance use disorders, it's remarkably understudied. And I won't go over all the criteria for an antisocial personality disorder. But I just want to point out that when we look at DSM-5, the first criteria is a failure to conform to social norms, deceitfulness, impulsivity, aggressiveness, and reckless disregard for others. An antisocial personality disorder does require the diagnosis of conduct disorder. So children have to be diagnosed with ASPD. You have to have had conduct disorder as a teenager. And then go on to have ASPD. However, we do know that there's a lot of patients out there who meet criteria for AABS, which is Adult Antisocial Behavioral Syndrome. So these are folks who have antisocial personality disorder or meet all their criteria, except for that they did not have conduct disorder or documented conduct disorder as teenagers. So if we take into consideration AABS, the percent of patients with substance use disorder who have either ASPD or AABS goes up tremendously, between 20% and 25% to 30% and 35% in cocaine, alcohol, and cannabis use disorder. So this is one of the most interesting studies that I think I've read in the past few years. It was a remarkable study. It was published in 2022. This was a study that interviewed 3,000 patients with antisocial personality disorder or AABS. And it talked about all these different symptoms. So here we have playing hooky. We have stealing. We have getting into fights. We have staying out, losing your temper, getting into fights, lying significantly, lying all the time. Goodness gracious. And they divided it by both ASPD and AABS. And the thing that struck me most about this graph was that the symptoms that we generally consider to be pathognomonic or part of the diagnosis of ASPD were the least common. So things like cruelty to animals, spending a lot of time in prison, disregard for others, and being a bully were the least common. I'm not going to mess with the mouse anymore because it's so, you know, the markers. So you can see that those were the least common symptoms. And the most common symptoms were things which are like getting into fights, being impulsive, not holding a job, not meeting obligations. So I decided to ask the question, what if we rethink antisocial personality disorder? In fact, just the name antisocial personality disorders implies that it's unfixable and it's unchangeable. But what if we don't consider it to be psychopathy or sociopathy, which is generally the way it's described? And I don't think that description fits because, as I mentioned, only 25% of patients with this disorder actually have so-called psychopathy. But what if we relooked at antisocial personality disorder? What if we reframe it in terms of impulsivity and impaired social cognition? What would happen then? So I went back and I decided to start with conduct disorder. And as I read more and more about conduct disorder, it was very clear to me that it was a disorder And as I read more and more about conduct disorder, it certainly became clear that there were a lot of similarities between ADHD, garden variety ADHD, and conduct disorder. And in fact, if you look at the literature from back at the 80s, conduct disorder was sort of viewed as being part of a spectrum with ADHD, which actually I think fits. There's a lot of similarities. Kids with ADHD are likely to at least about 30% to 40% have conduct disorder to comorbid with ADHD. When we talk about kids with conduct disorder, about 50% have ADHD. Both ADHD and conduct disorder are a risk factor for developing antisocial personality disorder down the road. And both of them are a tremendous risk factor for developing a substance use disorder. So a child with either one of these disorders at a very high risk of developing a substance use disorder as an adult. And then I found this study, which I think is one of the more overlooked studies that I've seen in the literature. This was published in 1997. It was a study that took about 40 kids, I'm sorry, 83 kids ages six to 15. They all had conduct disorder. Two thirds of them had ADHD plus conduct disorder, but they all had conduct disorder. And they divided them into two groups, methylphenidate treated 60 milligrams a day versus placebo. And the outcomes were really remarkable. What they found is that not only did their school performance improve, but their conduct disorder symptoms improved tremendously. And I have over here, this is the teacher ratings of how children improved. And you can see their aggression improved, their conduct disorder improved, their academic problems improved, not as much actually as their behavioral problems. And in fact, this improvement was seen not only in the kids who had comorbid ADHD, but even the kids who had conduct disorder alone improved significantly with the treatment with methylphenidate. As I mentioned, this was published in 1997. I've never seen a replication of this. And I actually haven't even seen it move into clinical practice, which in some ways as an addiction psychiatrist is very depressing because the number of patients I see or number of participants in my study who had conduct disorder as children is profound. And usually almost none had any sort of treatment whatsoever. And now I wanna switch and talk a little bit more about social cognition. I'm gonna start with the Netflix show Chimp Empire. I'm not sure how many of you have seen it. It's one of the most remarkably amazing shows that I've seen lately. And this is Gus from Chimp Empire. And Gus is a teenager who really has a hard time with social cognition. He can't figure out the hierarchy in the clan of chimpanzees. He's always on the outside. He can't figure out the rules for grooming and not grooming. It's pretty remarkable to watch him struggle with this in this documentary series. And frankly, it reminded me to an extent of some of my patients or my participants in my study. And I'll use an example of a participant I had a few years ago with alcohol use disorder. So in our studies, participants come into a research unit for about a month. And I have one person who came in. He was at our study. He was remarkable. He did a great time on the unit. It's a psychiatric unit. So there's the nurses, there's social works. And we have a lot of social work interns because it's a teaching hospital. And he did all the procedures. Everything was going well. And it was his last day to be in this study. And all of a sudden, there's this blow up on the unit. I'm called, the nurses, they're calling security. My patient has to be escorted off the unit. I'm like, what happened? So I run over there to see what happened. And it turned out that my patient had an interaction or an altercation with a social work intern. He was very upset because he had asked her out on a date. He was leaving the hospital. He asked her out on a date. She took it badly and it escalated from there. So I bring my patient to my office and try to get a sense of the situation. And he explains to me, well, a social work intern, she came and saw me every day. She laughed at my jokes. She asked me about my family members. I assumed that she wanted to see me. And of course, that kind of interaction at a party might imply that somebody wants to see you. But when it's a social work intern at a hospital, it doesn't have that same implication. And this really got me thinking more about sort of the way that our patients with substance use disorders struggle with social cognition. And to be frank with you, social cognition is really hard. There's a lot of steps. There's emotion processing, where you have to recognize facial expressions, body language, tone of voice. There's social perception, where you have to know what the social norms and rules are and how those social cues fit into those norms. There's theory of mind, where you need to see a situation from another person's point of view. And then there's attribution, where you make sense of all of this and how it fits within your own particular culture. Despite the complexity, I decided to look more into looking at social cognition. Social cognition is measured many different ways. There's the reading of the mind in the eyes test. There's looking at facial recognition. There's a number of tests where you read a social vignette and you ask the person what they think of it. There's some videos where you can have these social interactions and detect whether or not people can figure out what's happening with these social interactions. And there's also effective picture systems. To put a long story short, there's been a lot of research in conduct disorder showing that there are social cognition deficits. There's actually very few in antisocial personality disorder. The few studies there are tend to use incarcerated individuals, which isn't really indicative of the full spectrum of the disorder. But I wanted to start, I wanted to focus on this one study. I found this study remarkably compelling. This looked at social cognition in children with conduct disorder and autism. So it was adolescents. They had three groups. They had children who were normally developing, autism children, children with autism disorder, and children with conduct disorder. And it was a simple task. They showed them the pictures. The pictures here, they have neutral, positive and negative valence, and they asked them to respond quickly when they saw a positive versus negative valence picture. Ultimately, the results showed that the control group of kids performed much better than the kids with autism and the kids with conduct disorder. In fact, the children with autism and conduct disorder performed about the same. But what's compelling to me about this is that when it comes to autism, we refer to the lack of social cognition as being a form of mind blindness, as though it's not anybody's fault. But when it comes to conduct disorder, we generally characterize these children as being callous and unemotional, which implies that they read the social situation and react to it with harm, rather than understanding that they really don't read the social situation at all. I think we have a huge amount of data to show that they don't read the social situation at all, so therefore can't really be considered callous and unemotional. So now I'm gonna go back to ADHD, social cognition, and stimulant treatment. And I'm gonna go back to this study that was published in 1937. It's a remarkable study. It was a man, a psychiatrist named Charles Bradley, who was working at a hospital in Rhode Island, and the hospital in Rhode Island was for disturbed children. They had to be cognitively normal, they were intellectually normal, but they were so disruptive in their behavior that they were sent to live in this hospital. And he started testing dexedrine, which is amphetamine. He was giving it for headaches, which was kind of the big thing in the 30s. Everybody used amphetamine for everything. He tested it for headaches and discovered that the children responded remarkably to this, to amphetamine. Certainly there was a major improvement in their schoolwork, they attended school, but there was a remarkable improvement in their social behavior, and the way that they became less aggressive, more compliant, less irritable, and more able to get along with their peers. And actually, if you look closely at the data with ADHD and stimulants, the data with ADHD and stimulants, it's very plus minus when it comes to academic performance. Some kids do better, some kids don't. But when it comes to social function, the data is very strong. So treating kids with ADHD and stimulants makes their social function improve significantly, and there's actually some studies showing that it can improve theory of mind. And this down here at the last corner, there's a meta-analysis. This is children with ADHD who have oppositional aggressive behavior, and this meta-analysis shows that treating with stimulants has a pretty profound effect on their ability to function socially. So now I'm just gonna go back to, this is my one sort of neuroscience slide. What does dopamine do anyway? And it's interesting, because for much of my life, I did imaging of dopamine receptors and substance use disorders. Lately, I've been doing more other things, but I went back online, started reading about dopamine, mostly because a friend of mine called and said that her daughter was doing a dopamine detox, and did I have an opinion on this? And my opinion was, what the heck is a dopamine detox? And it turns out that there's quite a bit of stuff online saying that a dopamine hit gives you pleasure, and if you get too many dopamine hits, you need to do a dopamine detox. But of course, dopamine doesn't really, doesn't code reward or pleasure per se. Dopamine is about 1% of neurons in the brain. That's a big task for this small population of neurons. But what it does do is it sort of predicts reward. And I found this as a very interesting study from 2020. This was a study done in healthy controls, giving them a complex cognitive task called the MBAC, which is a very hard task. And what they discovered that amphetamine or stimulants did was it didn't make people perform any better on the cognitive task, but it made them more willing to try, more willing to take on a challenge, more motivated. Their expectation or their view of the effort required to obtain this task went down, so they were more likely to engage in this task. And this is consistent with what we see from PET studies that I used to do all the time. But PET studies in substance use disorders, what we were able to show is that patients or those with substance use disorders have greater dopamine signaling in the brain were more likely to choose between competing reinforcers. And those who had low dopamine were more likely to choose the drug. So we could show that low dopamine meant that participants were more likely to choose drug over money. It also meant that they were less likely to succeed in a treatment that used rewards to get them to where they needed to be. And this has been replicated in other groups. So when I talk about dopamine signaling, this is now what I use when I talk to my medical students. Low dopamine signaling, you see the big shiny red reward that's right in front of you, and it's really hard to see the rest. And when we increase dopamine signaling, you're able to see, consider and see alternative reinforcers that might be on the periphery. So in summary, these are my clinical and research questions. I would say the first thing that I do truly think needs to be addressed is does treating conduct disorder have any impact when it comes to developing substance use disorders? In general, children with conduct disorder are viewed as being lacking empathy, which I don't think the data supports. And in fact, we have some data showing that neurobiology plays a role, and that this disorder might be able to be fixed or at least ameliorated to an extent with a very simple treatment. We know that if we treat children with ADHD with stimulants, especially from the time that they're young and consistently, you can decrease the risk of developing a substance use disorder down the road, which is a pretty remarkable finding, and we don't know if we can do the same for conduct disorder. When it comes to antisocial personality disorder, there's a huge field to be figured out. We really don't know whether or not this comorbidity predicts treatment response or not. We don't know even what the overlap is, frankly, between antisocial personality disorder and adult-onset ADHD. And if you look at some of the data that I pointed out, it seems to be a lot of overlap. And we know that treating ADHD in patients who have substance use disorder can improve their outcomes when it comes to the substance use disorder. Lastly, when it comes to social cognition, we have very little data showing that it can be changed. There are hints of things. I showed the data with stimulants. There's also some data with MLOR5 antagonists in Fragile X syndrome. There's some data with nicotine, like nicotine patches and schizophrenia can improve social cognition. And I think we really owe it to our patients to do more research on this to sort of see, can we modify this factor? And then lastly, when it comes to animal models, I would just point out that impaired social cognition is common across psychiatric disorders, whether we're talking schizophrenia, ADHD, depression. The impaired ability to read a social situation is common. And I do wonder if we can model more of this in our animal models. We could be able to get a better sense of how it could be addressed in the human populations. And this is just by lab. So I'd like to thank everybody who is in my lab and thank you for inviting me to be here. Thank you. All right, let's see if I can, whoops. See if I can get this, deal with the technology. I'm Ned Nunes, I'm also at Columbia with Deanna, and I'm an addiction psychiatrist who's spent my career doing treatment research, and I've really enjoyed the presentation so far. I always enjoy the effort to pull together what's going on in the basic world with the clinical world. It caused me to reflect that when I was an undergraduate in studying psychology many years ago, it was a very behavioral department, and the brain was a black box. It was too complicated, you were never going to be able to see it. There was one course in physiological psychology where you were going to dissect a sheep brain, I think it was, and I signed up for it, but then I chickened out, so it was the road not taken. And then after residency, I fell in with a group that was doing clinical research in depression, headed by Don Klein, who was one of the fathers of clinical psychopharmacology, but he used to give a very similar message, which was, no, the brain's too complicated, just go to the clinic and look for serendipitous clinical observations, and you'll figure out what works. And of course, a lot of what happens is you have an observation like Deanna talked about, that early study with benzadrine, and then you go backwards and look at the brain and try to figure out what benzadrine's doing. So anyway, I'm just going to make some comments about how I look at substance use disorders clinically, and then we can get into some questions and answers. So the main points I want to make is that drugs and substances are addictive because they function as reinforcers, either positive reinforcers or negative reinforcers or punishers. And there's also an element of habit formation, where even after something's not very reinforcing anymore, it's become a habit. There's a lot of Pavlovian conditioning going on, where cues in the environment trigger craving. Other mental disorders may make a drug more reinforcing, which is why you may be part of why you have comorbidity between mental disorders like depression or even antisocial personality or attention deficit disorder, for example. Drugs have toxic effects. They function as physiological stressors, and they tend to make psychiatric symptoms, depression, anxiety, executive functioning, cognitive functioning, tend to make that worse with chronic exposure in extreme doses anyway. And stress and trauma are common underlying risk factors for most psychiatric disorders. Stress and trauma bring them out. And for the substance use disorders as well, stress and trauma bring them out. And actually, I was reflecting, listening to Deanna's talk, that some of the first substance use disorder patients I met were in a methadone maintenance clinic. They were opiate-dependent patients. And we were there to study the treatment of depression in opiate-dependent patients. And they all had the same history. They got in a lot of trouble in school. They had trouble learning. They got in trouble with the teachers. They had behavioral problems. They ended up dropping out of school in about somewhere around seventh or eighth grade. And they got mixed up with a bad peer group, and they got into drugs. And they ended up with opiate use disorder and then methadone maintenance. But for all the world, they had antisocial personality, or they had disorders in that spectrum, if you will. So drugs as reinforcers. Drugs are positive reinforcers. And one of the things that always amazes me about those opiate-addicted patients that I saw at the methadone clinic, I'd always ask them, well, how did you get into this? How did heroin make you feel when you first took it? And I was expecting to hear, what? Oh, it was so relaxing. No. Most of them would say, I had never felt better. I had so much energy. I was effective. I could get things done. You would think they were describing a stimulant. And they clearly had a different response than the average person to the drug. And a very positively reinforcing, where they would say, where can I get more of this stuff? I've got to get more of this. I've never felt better. Drugs act as punishers. A lot of patients who are addicted to one drug will talk about another drug and say, I never want to see it again. It was horrible. They experience drugs as having really bad effects. And the point being that there's a tremendous amount of individual variation in the human population in how people respond to drugs. And then there's negative reinforcement, which is withdrawal syndromes, which drive drug taking because the person wants to take drugs to try to avoid or get out of the withdrawal syndrome. So what are the risk factors for substance use disorders? There's complex genetics, and twin studies show that it's at least 50% heritable. What's inherited? Probably how reinforcing the person finds the drug and the extent that they don't find the drug to be punishing and so forth. Stress is a big driver, and sensitivity to stress of people in the human population varying how sensitive they are to stress. Trauma, there's a lot of trauma behind both substance use disorders and other psychiatric disorders. And finally, executive function deficits, or cognitive function, and I think that's again, Dion, a part of what you're talking about in terms of the social cognition, that there's impulsivity, inattention. There were a number of longitudinal studies of children who were picked up in elementary school and followed into adolescence and adulthood. One of my favorites is the CEDARS study, which was Tartar and Moss, and they found this syndrome that they called neurobehavioral inhibition, which was impulsivity, inattention, and aggressiveness. They got in fights and that sort of thing. And that was a risk factor for going on to develop substance use disorders in adolescence and adulthood. And then there are other things like delayed discounting, tending to value an immediate reward, the apple right in front of you, rather than the spectrum of other rewards that might be out there that might be harder to get, you have to delay and work harder and longer to get. So I like to think about the bell curve in all of these things, because all of these traits that I've been talking about in the human population, they're complex traits, and they manifest in the population like a bell curve. So this is just one example, good drug effect. To what extent does the individual find a particular drug, whether it's alcohol or opiates, pleasurable? The majority of the population finds the drug neutral. Most people who take opiates are very happy to have pain relief when they have a broken rib or whatever, but they don't feel much else one way or the other. Many people on the left-hand side of the bell curve will say, it's really unpleasant, my head's foggy, I can't think straight, it's horrible, I was happy for the pain relief, but I couldn't wait to get off of it. Now on the other end of the bell curve, you have those people, like those methadone patients that I met early in my career, where it just feels really good, they've never felt better, it's just extraordinary. So there's this bell curve, and I would maintain that there's this sort of bell curve for all of these features of a population that become risk factors, sensitivity to stress, how unpleasant people find a drug, and so forth. So who's at risk to become addicted are the people who fall on the wrong side of the bell curve, if you will, or the risky side of the bell curve, and one probably more of these different traits. And then what are the risk factors for depression, anxiety, and so forth? Well, stress and trauma that we've talked about. Substances serve as physiological stressors, and the substance-induced lifestyle itself is likely very stressful, hustling on the street to try to find heroin, and dealing with all kinds of dangerous characters. And then personality and neuropsychological functioning, which is getting back to what Deanna was talking about, I think, a great deal, including these so-called personality disorders, like so-called antisocial personality, and again, this impulsive, inattentive, dysregulated cluster. So what are the relationships between substance and other mental disorders from a clinical perspective? I like the DSM-5 approach, where DSM-5 divides psychiatric syndromes that you see in someone using substances into substance-induced disorders and independent disorders. And I would encourage you to actually review DSM-5 and see what DSM-5 says about that. Among other things, Tristan was saying what comes first, that DSM-5 leans heavily on that. Does the so-called mental disorder come first, which, of course, in the case of attention deficit disorder or conduct disorder, it does. They have onset in childhood. Often a depression will have its onset before the onset of heavy substance use, or the other way around. You can have heavy substance use, and then the depression happens afterwards. So what happens if you remove the substances? So this is one of the studies I consider sort of a classic study, Mark Shuckett and Sondra Brown on a VA inpatient unit. I think this was in the late 80s, early 90s, and they had veterans with alcohol use disorder who were admitted to this inpatient unit for three or four weeks, and they were washed out of alcohol. So what this shows you is what happens to the Hamilton depression scale on the vertical axis with weeks of abstinence, admission week 1, 2, 3, 4. And a score of 5 to 10 on the Hamilton scale is mild to not depressed. Most of us probably walk around with at least a Hamilton score, somewhere in the 5 or 5-ish range anyway. 15 is a moderate depression, and when you get into 20, 25, 30, you're getting into severe depression. And what you see is that there's three groups here. The two groups on the bottom with the brighter lines, those are people who were diagnosed at baseline either with what they called at the time secondary depression or no depression. And secondary depression means that in the person's history over the lifetime, the substance use came first, then the depression came after. And this is the majority of the sample, actually. And what you see is the depression gets much better over three weeks' time. You've gone from a moderate to severe depression down to a mild to not depressed. And this leads to one of the major clinical recommendations, which is the first thing you do when you see someone who's using substances and has any kind of psychiatric syndrome is try to get them off the substance. See if you can get them abstinent and see what happens to the psychiatric syndrome. It's likely to get better. The one exception to that is the red curve up top. Those were the people in the sample with a primary depression. So by history, their depression came first in their lifetime. And the alcohol problems came after. And for them, the depression's more severe to begin with. You can see it's up around 25. It doesn't get much better over the course of three weeks of abstinence. And that's a group that really needs specific antidepressant treatment. So you can have, actually, if you think about it, a lot of different relationships between substance and other mental disorders. You could have a substance-independent disorder, like the red line in that graph we just saw. A patient can have two independent disorders. They're all common disorders. So somebody can have two disorders just by chance. You've got common underlying risk factors like stress and trauma. And there's even some evidence for common genetic factors that drive substance and mental disorders together. And then you've got, you know, I was interested in hearing from Jan and Raj about some of these animal models. And what is it, the tail-hanging test and the forced swim test? You know, those are tests for depression. And like you said, antidepressant medications increase the amount of time that a rat in forced swim will swim before they give up, right? That kind of stress also, stress like that in experimental animals, as I understand it, also increases the propensity of those animals to take substances. And that's another similar kind of animal model. So you see stress in an animal, that bell curve shifts. And they're more likely to actually find a drug positively reinforcing it. And similarly, when you stress human populations or when you stress individuals, they're more likely to take substances. So this is a sort of graph of all this. You've got common factors, stress, trauma, and genetics. You've got depression and post-traumatic stress disorder and ADHD and so forth on the one side and substance use disorders on the other. Those are interacting with each other and they have common risk factors driving both of them. So implications for diagnosis and treatment. Treat the substance use disorder. If you can get by history that it's an independent mood or other mental disorder, by all means treat it. It's really important to take a developmental history in all of this because if you dig into the developmental history, you may find ADHD in the school history. You may find this kind of neurobehavioral disinhibition or antisocial type stuff. Or you may find early anxiety in elementary school or junior high school, long before the age of risk for substance use disorders. Then you know you're dealing with an independent disorder. And so that's really, in a nutshell, the implications for diagnosis and treatment. Treat the substance use disorder with medications or with psychosocial treatments or both. And if you've got an independent mood disorder, treat it. Obviously one of the ways you can find out if you've got an independent disorder is if you can get the person off the substances and they still have the disorder, depression or whatever it is, like the red graph in that thing. And then the last slide I want to show is this one. This is a meta-analysis that Francis Levin and I did a bunch of years ago looking at studies of antidepressant medication on the outcome of depression in patients with substance use disorders. A mixture of alcohol use disorder, opiate use disorder. One or two of the studies were cocaine use disorder. But what they all had in common was they were placebo-controlled trials of antidepressants. And this is a forest plot. So each dot is a study. Each row and each dot is a study. Zero is no effect. No difference between medication and placebo. And you can see there's this bottom cluster of studies where there appears to be no effect. They cluster around zero. Those were studies that had high placebo response rates. Patients on placebo got really better. And then you've got this upper cluster of studies where the effect size, and this is the effect size, by the way, which colons D is the difference between the medication placebo and standard deviation units. An effect size of one is really big. One standard deviation unit is a big effect in a clinical study. You can see that that upper group is somewhere between .5 and one most of those studies. And that's a highly significant effect. And those were studies with low placebo response. So what low placebo response probably means is that these were independent disorders. These patients were all getting treatment for their substance use disorders. So you treat the substance use disorder. You've got low placebo response, the depression doesn't get better. Then the antidepressants look like they work well. And what were most of those effective antidepressant trials testing? They were testing tricyclic antidepressants, which of course are very anticholinergic. And so it's really fun to think about all these possible intertwining relationships. So I'm going to stop there and hopefully we can have a little bit of discussion and back and forth, whether it's on the clinical issues or the theory. Do we want to bring the panel up to the, yeah, yeah, yeah. Why don't you guys come on up to the stage? And any questions? Yes. Hi. Ramatzee Saunders, UCSF. Great talks. My question relates to, okay, so we talked about comorbidity of personality disorder or personality factors and substance use disorders. But something else that one observes as a clinician, and it seems intuitive, is that for someone who's been using substances for a really long time, their behaviors change. And after a long period of substance use disorder, it almost looks like there's a substance-induced personality disorder. And I think it's really hard to parse what came before, what came after, what was learned. So I wondered what your thoughts were about that. Actually, my first thought is habit formation, which I don't know if you guys, because that's one, you know, a long-term pattern of behavior becomes habitual, right, and then you get a pattern of behavior that grows over time by force of habit, and habit formation is something that grows out of learning. I always think about, you train a dog to be house trained, you have to reinforce him at first, and then after a while, it's just a habit, you don't have to. Yeah. And in the basic neuroscience, behavioral neuroscience world with animal models, we think about that a lot in the interplay between goal-directed behavior, if you will, and habit formation. And, you know, there's definitely a school of theories, and there's competing theories that, you know, I would say largely there's three competing theories right now about how substance use develops, substance use disorders develop, is, you know, your classical, it was goal-directed, and then it transitioned to being habitual, and now they can't get out of it. There's another theory, which, you know, needs more work, and, you know, but I like, is it's actually over-weighting of goals, and it's just excessive goal-directed behavior brought upon by, you know, the negative behavioral affect that the person is undergoing when they're not on the substance that they have now tended to value the most compared to everything else around them, and that's certainly something people are looking at as well. And then removed from habitual formation and sustaining habits is this idea of just compulsive behavior, perseverative behavior, which is different from habit formation, where now this person or this animal or, you know, this being that we're testing is going after this substance in a compulsive, perseverative fashion that's more automated, if you will, as opposed to habitual. And at least in behavioral neuroscience, with all the tools we have now, we can, we have multipose estimation with machine learning that can then, you know, pick apart different animals all the way to monkeys that can look at the genesis of, you know, these behaviors that we've just posited, right? Because the way you test if an animal has a habit is not based on key features. One feature that people use is whether it's overtrained. So now you devalue the outcome that the animal is supposed to get, but if they still keep performing the action, that means it's habitual. But, you know, is it habitual or is it automated? And do animals inherently, like people, develop automated behaviors on their own that they have trouble getting out of, right? And so we're at this sort of crossroads, I think, in basic behavioral neuroscience. But I hope that answered your question. No, it was certainly helpful, and thank you. Yeah, and the other thing I'd say is that from a sort of clinical perspective, the person starts to use substances, it disrupts their life. And then it can kind of become a vicious cycle, right? Their life is chaotic, and so they're in a stressful situation, which then drives more substance use, and their life becomes unraveled in many ways to the point where when somebody seeks treatment, there's the old idea, right, that somebody seeks treatment because they hit bottom. What does hitting bottom mean? It means that their life has really come apart. They've lost their family, they've lost their job. They've damaged their health. It's chaos. So I think some of it is sort of the social and interpersonal toxicity, if that makes any sense. Thank you. And there's also some studies in the animal world, in the rat and mouse world, that aren't published yet, but they're about to, looking at social rank in mouse cages and how that influences drug-seeking behavior, like natural reward-seeking behavior and drug-seeking behavior. And there's interesting correlations between social rank and if you disrupt that social rank, what that does to drug-seeking. Thanks a lot. Mark Famador from Rowan, New Jersey. Thank you so much for stimulating our brain cells. But I was wondering, how optimistic are you that your findings, or maybe are we getting closer to using acetylcholine as a biomarker? The criticism with DSM-5 is it's all descriptive, and the research that we're doing here hopefully will not only convince ourselves, but be an ambassador to others, that our science, our practice, or our field is science. It's not bogus or voodoo or whatever. Do you think we're getting close to having these findings as biomarkers? The CAP opioid receptor antagonist that I spoke about, there was a NIMH fast-fail study that happened recently, out of which they developed a biomarker using the CAP opioid receptor antagonist. And what they showed there was that in patients with anhedonia compared to the placebo, that they had self-reported lower scores of anhedonia, and were also able to look at activation in different parts of the brain using this as a tracer, as a ligand, and show activity and changes in activity as these patients were prescribed this compound multiple times, where you saw increases in activity in the ventral stratum, which you normally associate with mood and reward processing. And so we're getting there. I think what we're hoping to do also, at least on the basic neuroscience side, is to really not put the cart before the horse, and to really tease apart exactly what's going on in the brain. I mean, I think the days are behind us, where we attribute one function to one molecule or one brain region. It's a lot more complex than that. And we're hoping to disentangle that more. And I guess, just to add a wrench, you know, I've heard Roger McIntyre with Bipolar describing it as a heterotypic condition, unlike diabetes, which is homotypic. So that also makes it more complicated, because in homotypic, you have the hemoglobin A1C. And the heterotypic, now what do you do? So just to add to that, biomarkers are often useful in a couple of different ways, right? Either they're useful for improving accuracy of diagnosis. They're useful for monitoring response to treatment. One of the things that's interesting is actually looking historically at prognostic value for certain biomarkers, and then making a determination on what kind of quantitative metric for biomarkers for positive predictive value does it take for something to actually be implemented, right? So if we look at, you know, standard medical practice in a lot of cases, and you look at the positive predictive value of what are considered, you know, standard biomarker tests, things like Pap smears, for example, well, you know, that's about 70%. You know, it's not 100. It's 70. Most things are, you know, are less than, you know, dead on certain, but they're good enough that people are, you know, clinicians are willing to say the investment in additional care or a second test is met based on the accuracy of this particular biomarker. And so when it comes to, you know, complex psychiatric conditions, I think it becomes important to think not only about what is the thing and what is its tether to a particular mechanism, but what is the threshold beyond which we accept that being indicative of response to treatment or investment in, you know, energy? Are we benefiting the patient population in terms of their risk, in terms of our investment in, you know, care, et cetera, in response to that particular biomarker? So I think, you know, to be optimistic about it, I think we're entering an era where the average cost of biomarker discovery is decreasing, right, as, you know, molecular measurements become cheaper, instrumentation for molecular measurement at that scale becomes more democratized as, you know, is not just occurring at select sites but rather can be done at multiple sites. You need to get to a critical mass of that kind of information, though, in order to see the sort of cross-validation and then, you know, get the provider community to actually agree on a particular level, so, yeah. Hello, Steven Gilman from NICHD. I wonder, I guess I have more of a conceptual, theoretical, philosophical question about comorbidity and the concept of comorbidity, so as I understand it, the definition of comorbidity is that there are multiple disorders present in the individual at the same point in time, and it's, I'm wondering what your thoughts are on, are there conceptual weaknesses there, especially when some of the examples here where there's symptom overlap, some of the discussion earlier is these things cause each other within the individual over time, conduct problems lead to substance use disorders, the first symptom of antisocial is violation of social norms, it was interesting to see that in your slide, and it just seems that there are situations where it's not as clear-cut as cancer and cardiovascular disease present together, so I wonder what your thoughts are on, really, this idea of multiple conditions, disorders, present in the individual. Well, actually, I thought that part of what was interesting about thinking of it as multiple disorders is you can sort of triangulate in a way, to try to understand what's going on, if that makes any sense. Oh, that diagram you had in the... Yeah, yeah, I mean, you could make the, I guess you could make the argument that it's sort of one disorder where the substance and the conduct problems are manifestations of the same thing, I guess. Well, certainly when it comes to children, we generally know that the disorder precedes a substance use disorder, especially if it takes, if it's onset is pretty early in life, which is generally the case with ADHD, anxiety and depression, and even conduct disorder, whether or not we include ODD in that or not. So in that respect, we can see a trajectory where you have a disorder, and then the substance use, because adolescents don't really have access to drugs until they get a little bit older, and then it's, then everything sort of takes off. So I think when it comes to children, at least, we can sort of make an assessment that you start with some sort of psychiatric disorder, the onset of which is often early in life, and then, you know, that leads to self-medication, a big believer in self-medication, and then you end up sort of in the cycle, one worsens the other. And so one of the, I think, the advantages of trying to tease apart symptoms without necessarily putting them into a diagnosis is treating symptoms. So even though we don't have, like, a biomarker, per se, of impulsivity, we know that it responds really well to stimulants. And so in some ways, it's kind of a cheap biomarker, right? If somebody does really well with stimulants, then you're like, okay, this is a way to move forward. And I understand the reluctance to give patients with substance use disorder stimulants, but on the other hand, you know, there are ways to sort of treat the impulsivity and see if you can get further with their treatment. Certainly not if they have bipolar disorder, but, you know, a quick litmus test. You know, with stimulants, you know pretty quickly if they work or if they don't. And I would just urge that, you know, we think of antisocial personality disorder to sort of think of it more in terms of impulsivity and see if we can address that issue and then work on some of the social factors. Because as I mentioned, a lot of what was in that graph of symptoms had more to do with just not following through than it had to do with being actually an evil person. I'm Justin Hoxhall. I'm in Denver, Colorado. I'm glad you mentioned about treating people with substance use disorders for ADHD because that was going to be one of my questions about, you know, if the addiction train has left the station, what do you do in that circumstance? Certainly there's evidence to support that if you intervene early enough in the course of ADHD, you lower the likelihood of developing a substance use disorder. But what do you how do you approach the patients who've already developed one, may still benefit from treating their ADHD, but are at a higher risk of especially with the reinforcement of the circuitry and the development of the whichever pathway you're going to invoke for that would put them at higher risk. There's a number of clinical trials in the literature looking at methylphenidate or amphetamines for treatment of patients, adults with substance use disorders who also have ADHD. So you're treating the two concurrently and you're worried about what are you doing, giving a potentially addictive drug to someone with a substance use disorder. But by and large, they all use extended release formulations so that, you know, the formulations of drugs that are addictive tend to be immediate release. And so the extended release formulations are preferred. The trials tend to show that you don't see worsening of addiction. So Frances Levin at our place in Columbia and her colleagues have done a couple of trials with extended release amphetamine for cocaine use disorder in adults. And it's safe. There's no evidence that the patients become amphetamine addicted. And it's a pretty good effect size that the cocaine that patients treated with active amphetamine are more likely to get abstinent than placebo. It's a significant effect. Whether you're measuring complete abstinence or reduction in substance use that, you know, the ADHD gets better. There's also a lot of negative trials in the literature. They tend to use lower doses. They sort of wimp out on the dose, you know. And of course, one of Dr. Levin's points she makes is, you know, don't push, you know, if you're not getting a response, push the dose. So I think, you know, the answer is there's evidence that extended release and, you know, be careful. But there seems to be evidence that it's helpful. Particularly those studies establish the childhood diagnosis. So they do a careful developmental history. So you get the diagnosis that there's an independent disorder. But I wouldn't exclude, I wouldn't say conduct disorder excludes ADHD. I mean, I think some of the strongest data is with stimulant use disorder. Yeah. Stimulant use disorder and nicotine use disorder. If you treat the ADHD, those disorders tend to get better with stimulants. And we know that stimulants are, you know, when it comes to ADHD, remarkably effective. But there are other medications as well. So you could try adamoxetine. You can try guanfacine in patients who you may not, you may be wary of giving a stimulant to. They're not as, the effect sizes aren't as large. But if it works, it's definitely a place to start. And those are norepinephrine uptake inhibitors. Right. So I do really think that if we sort of try to piece apart like the impulsivity, like can we address the impulsivity with drug use, which doesn't apply to all substance users, but applies to some. And then try and build in like, you know, some sort of social supports. Often patients have lost this. But piece by piece, if you sort of decrease the substance use, you might not get to abstinence. But as long as you decrease it bit by bit and work on the impulsivity and then encourage patients to try and piece back some of those social structures that they might have lost. Hi. Frank Summers. I represent kind of the older mode of treaters here. That I've been listening to the biomarker team and the experimenters with great interest. But I felt like you stopped short of connecting your data with the patients I deal with. Although I must say, I have been thinking about certain patients I would like to suspend over a tank of water. And we can talk about that later. But so could you offer a takeaway in what... I hate to hear myself describing myself as an old-fashioned practitioner. What we should learn from your data that will help us in formulating a combination of talk therapy or whatever. But I kind of felt like you guys felt that we all got the connection instantly. And I must say, I didn't. So could you... What is it I should take away from your biomarker data studies? How about the enriched environment studies? Experimental animals and enriched environments. You mean for neurogenesis? No, the rats of NIH. I'm sorry. I'm talking about the speed, the change in the curve as to when the animals would respond, would show a biological response to the stimulus after they had been trained versus before. Right. And I kept waiting for that. And therefore, here's what you should... So take away. So I'll set the stage for you here and basically point out that what you're highlighting is what we in the research field consider a research gap, right? Is that we are examining a physiological phenomenon where we have now, thanks to this type of work, where we have very good evidence for the involvement of a very discrete biochemical signaling process. Now, that biochemical signaling process, in the case of both the dynorphin system and the acetylcholine system, has potential applicability to an array of behavioral phenotypes that are consistent across psychiatric conditions. What we're missing there are the tools and methods that allow us to engage those biochemical pathways to see what the impact on the behaviors are going to be so that we can make precisely that kind of recommendation to the clinician. Or even at this stage in the work, giving us suggestions where we can actually experiment, I should say, experiment on our own patients. But as far as it's gotten, Frank, here's what you might want to consider, even though it's not cast in concrete. I mean, I pitched it to Diana because you do work with the Cuban subject. So to what extent do you find the kinds of, you know, fundamental physiology informative to your design of experimentation? So certainly, you know, when we go from the rodent to the human, there's a huge gap. And then when we go from human to what we can actually prescribe, that's the even bigger gap. But one thing I did like about the dynorphin talk that resonated with me is that it just sort of shows to me why I think buprenorphine is probably a fantastic medication, at least if we're talking about, you know, opioid use disorder. We tend to think of the mu partial agonism, but the kappa antagonism may have a very promising role when it comes to getting people to shift their behavior from these, you know, sort of entrenched rewards. And then when it comes to, you know, acetylcholine, I'm a huge fan of nicotine patches. And I always argue that nicotine is not the same as tobacco. Nicotine patch, I don't even view as being addictive. If you go into your average CVS around the corner from a high school and look at the nicotine patches, there's dust on the top. Nobody steals them, right? So there's a big difference between a cigarette and a nicotine patch. And I think we often don't really look at the power of nicotine patches when it comes to not just tobacco use disorder, but also, you know, there's interesting data with mild cognitive impairment. There's interesting data with refractory depression. Tourette syndrome. Right. And so, you know, I do feel that we underutilize it, particularly because we have this idea of nicotine. But it is something that is available right here and now that could be looked at. I can give you an example, actually, from our lab. It started in animals and then was translated to a small trial in adolescents with, I think it was referred to as explosive behavior. So we had the hypothesis that targeting the nicotinic system with patches or with nicotine or any kind of agonist could change aggressive behavior in animals based on what we know about the circuits that are involved. We tested it in animals and it worked great. Managed to get an IRB, put patches on adolescents that really had this kind of explosive behavior and it worked as well. So that doesn't necessarily help the practice immediately, but I think also the value of what we are doing is that if we move forward with hypotheses about what are the circuits that are involved in this patho, I mean, I don't know, in this physiological, I don't know how to phrase it exactly, but basically what we believe underlies this disorder, we can really use the animal models to very painstakingly take away elements and see how we could modulate these circuits and then go back to the clinic and try to use, as you said, compounds that may not be available in the human, but try to use strategy to then modulate these circuits that, once again, we can hypothesize actually regulate this abnormal pathological behavior. So there is a, but yeah, I think the issue as well that you're pointing out is really the temporal aspect. And yeah, that is something where I'm not really sure to answer. And if I may add one thing to this is, I mean, ultimately all our goals as basic neuroscientists is to benefit and improve the quality of life of people suffering from these neuropsychiatric disorders, but I also want to really stress the importance of basic science for the sake of basic science. The fact that dopamine is not pleasure and that dopamine purely signals anticipation of a reward, and now the theories are that dopamine might just be salience. It might not have an associated valence to it even. That happened because of basic science for basic science. And that in the future is going to inform development of better treatment strategies to help people with all of these disorders as well. And I think it's really important for us to also really consider what a particular molecule or what a particular brain region is doing. Before there's an insult, before trying to understand whether that system is co-opted in some neuropsychiatric disorder, I think it's important for us to really understand that aspect as well. Wouldn't you say it's true too that most of the treatments we have in psychiatry now were discovered by serendipitous clinical observation, and then you take that back to neuroscience and try to reverse engineer it? So fluorazine was discovered for psychosis because it was said that it was being used in an operating room, and somebody gave it to an emergency room to a patient with a manic psychosis, and the psychosis cleared up. And they're like, oh, maybe this is actually an antipsychotic, right? And benzadrine was synthesized by some chemist in a red stimulant, and then somebody gives it to kids in Rhode Island in 1936? 37. 37. And then sometimes those clinical observations get lost, which is sort of what you're describing, right? They get lost. In fact, there's an argument that lithium was discovered four times in human history and forgotten. So they say that the Greeks and Romans, classical Greeks and Romans, knew there were certain mineral waters that were good for mood disorders. They didn't know why. And then apparently lithium was used in the 1800s, but they didn't understand the toxicity. So they made people really sick. You'd get the dose 200, and so it was abandoned. And then the third time was a modern story. It was an Australian kidney physiologist, and his buddy was a psychiatrist working in the mental hospital. And he was studying lithium in the kidneys because lithium's a salt, and the rats got sleepy. So he said to his buddy, why don't you try to give it to your agitated patients in the hospital? He gave it, and then people with bipolar disorder got better. So it's serendipitous clinical stuff. Then you go and you try to figure out how lithium works, which I don't think anybody still has a clue, right? It changes. Yeah, but some of these things, like the acetylcholine system and the dopamine system and the noradrenaline system, we study noradrenaline and dopamine a lot because that's what the drugs that treat depression and attention deficit disorder seem to be hitting initially. Those drugs were discovered by accident, but now you can do neuroscience and start to reverse engineer how they're working. And maybe understand better the mechanism. Talking about the classical antidepressants and rooting for my own work, we know that a lot of the either TCA or even SSRIs are excellent nicotinic receptor antagonists. And if you combine, actually, in animal models, but I think there are small clinical trials as well, and I know we worked on that with Pfizer, where you combine a nicotinic receptor antagonist to a classical antidepressant, you can get some kind of additive effect. So this is what the animal models really allows us to do. I like the idea that you put forth about the reverse engineering, because we've been doing that, for example, with guanfacin. I mean, we were part of a central grant where they use guanfacin in stress-related nicotine use and looking at sex as a biological variable. But then we were able to go back to the animal models and really try to tease out which circuits were affected by the treatments. And for example, at the time, so not only that allows you to get maybe a better definition and maybe find along the pathway better targets, as opposed to hitting something that may be really at the very top of multiple pathways, but also to understand, and it was the case in the guanfacin studies that, because we are looking at sex, that the pattern of neuronal activation between males and females who were receiving guanfacin was absolutely, completely different. But the output, phenotypically, were strictly the same. So that also opens a can of worms. But then there's the reality that a great many patients in treatment for mood or anxiety disorders are on multiple medications. Average person with bipolar disorder is on three medications. They're all doing different things. Like you're saying, you have a complicated pathway and maybe different parts of that pathway if you can with medications. But I feel like it also goes back, going back to even the biomarkers, to some extent about, and again, I'm not a clinician, but about definitions. I mean, as far as I know, depression, it's not, I think you should put an S. You have different endophenotypes, different traits, different traits to diagnose that are not the same from one person to another. And my maybe naive assumption is that these traits, of course, involve potentially different markers or different pathways. So trying to just target this big entity as a whole might be the first issue. As opposed to teasing the part. But Deanna, like you were saying, pick out something like impulsivity from a larger syndrome and see if you can target that, right? Impulsivity is not a singular therapy. No, it's not, but I do think that, I do think in a short period of time with treatment, we know if our medication is having an effect or not, right? So one advantage of impulsivity is it tends to change quickly if your medication is making it better. As opposed to, let's say, an antidepressant, we're like, okay, three, four weeks later, right, we're going to see what's happening. It's probably some risk that we're going to make a diagnosis based on treatment response. We can get some information, but we wouldn't assume that that medication had any response. Also, it did respond to the syndrome, but it doesn't have ADHD. Not at all. Not at all. Right. But if you suspect there's hints of ADHD in there, you can know pretty quickly whether or not your treatment, the treatments we have right now for ADHD are having an impact. It's just not a long wait, is my point.

comorbidity

psychiatric conditions

substance use disorders

integrative neuroscience

endogenous opioids

dynorphin

neuropeptides

brain circuits

cholinergic dysregulation

acetylcholine

mood disorders

clinical management

translational research

personality traits

cognitive factors