gathering here in Missouri Auditorium and to those watching by videocast for our Wednesday afternoon lecture, our major lecture of the week. And we're fortunate today to have as our speaker Dr. Brigitte Kiefer. Her training and includes a PhD in organic chemistry and biochemistry from Universite Louis Pasteur in Strasbourg, France, and then postdoctoral training in Basel before returning to Strasbourg where she became professor and research director. director at Inserm. But in 2014, Dr. Kiefer moved to McGill University in Montreal, where she is currently a professor in the Department of Psychiatry, while continuing to hold a professorship at the University of Strasbourg in France.
I think it's not an overstatement to say that she is a giant in the study of the opioid system in the brain, a topic of considerable current interest and importance given the crisis we face in this country. of opioid overuse and addiction and opioid overdose deaths. She's the first person to isolate a gene encoding an opioid receptor back in the early 1990s and that opened up an entire research field that leads to much of our current understanding of the molecular basis of how opioids work, what the endogenous purpose of that receptor is and how medications affect it in ways that currently we're all trying to. to understand and in many instances trying to manage it.
Her genetic dissection of the opioid system has led to major advances in understanding pain, addiction, and mood disorders, and has also contributed a lot to molecular pharmacology and to G protein coupled receptor research since the opioid receptor is in that family. She has been widely collaborative, shared lots of her information and her genetic tools, has created innovative research. lines that have had an important impact in neuroscience.
We are proud to say that she is a grantee of the National Institutes of Health, despite her current location above the border. And she has also received numerous awards, the Lounsbury Award from the French and U.S. Academies of Science, the French Academy of Science Election in 2013, and in 2014, she received the International L'Oreal UNESCO Award for Women in Science. So her work has been truly transformative for our understanding of the world.
of the opioid system, but we have lots of work to do here. We at NIH are fond of saying we have all hands on deck trying to bring the best science to the problem that we currently face of opioid addiction, and the kind of work that she's going to tell you about today is going to be a critical part of how we're addressing that. Her title is Opioid Receptors and Brain Function, so please join me in welcoming Dr. Brigitte Kiefer. DR. Thank you very much.
Very nice introduction. Good afternoon everyone. Good afternoon to you. Thanks for coming and good afternoon for all those that are on the web.
I am very happy and honored to give this lecture, the world's lecture, and happy also because I meet many colleagues here at NIH, since long-standing colleagues. So I'm going to talk about the opioid system. I think this...
I have slides, you don't have them, so I don't know what I should do with that. Ah, here it comes. Could you turn on the lights, please? Thank you. So, this is a kind of timely topic, as you said.
Opium is a kind of magical substance that seems to relieve any kind of pain, physical pain, mental pain, social pain, and produces euphoria. And since the usage of opium, which goes back to millions of, no, sorry, thousands of years, I'm sorry. The active ingredient of opium was isolated 200 years ago by a German chemist. It was called morphine.
You see the morphine structure here. And from the isolation of morphine, morphine was used in the clinic to treat severe pain. And it was obvious that morphine produces both very efficient, highly efficient... analgesia was a very strong painkiller and is still the strongest pain killer, but also had strong addictive liability. So, since that time, the holy grail in the field is to look for a compound that would be as would be as good as morphine but would be devoid of addictive liability.
So the problem seemed to be solved in 1898 where Bayer, a German company, sorry, a chemist in Bayer decided to deacetylate morphine. So a very simple chemical reaction that produced a molecule that was commercialized. as the first non-addictive opiate compound, and it was called heroin.
So, of course, you know today that heroin is much more addictive than morphine because it crosses the blood-brain barrier much faster, and since that time, heroin became a problem. And in fact... Opioid, well opium and opioid drugs have been a challenge, a societal challenge, since many, many years. From the opium wars that you may know about, where British trading policies... devastated Chinese economy, leading to millions of opioid addicted individuals, to the heroin prohibition in the 1920s, now we have entered the era of the opioid crisis, where the opioid crisis is still a problem.
overprescription of opioids to treat pain has actually led to get got out of control leading many many people to become addicted to opioids and then to transition to heroin or fentanyl use and this alarming increase of overdose and I'm sure many of you have seen this cat type of curves so there's an acute problem here now during all these years neuroscience has dramatically evolved and And we have discovered a fascinating neurodegeneratory system that actually I will talk about today. The opioid system. So morphine acts by binding on the brain, on receptor sites.
And the first receptor sites were discovered in 1973, and they were called mu, delta, and kappa opioid receptor sites. These were simply binding sites. Then, because there were binding sites in the brain and this was the beginning of pharmacology, it became obvious that we were not born to live into poppy fields and that maybe these receptor sites had a role in our physiology and the first endogenous peptide that actually activate these or bind to these receptors were isolated two years later and these were metankephaline and luankephaline and they belong to a group of to a family of opioid peptides that derived from three genes, which are called POMC, preproencephaline, and preprodinorphine, that you may have heard about.
These genes encode large precursor protein that are later cleaved in smaller peptides, and I show you here three major opioid peptides. You see beta-endorphin on the top, you see metanilencephaline, and you see dinorphins. And these all share a common N-terminal sequence that interact into the receptor site. Now, identifying receptor genes was much more difficult, so it took 20 years from the discovery of binding sites to the isolation of genes, and the first gene was isolated by expression cloning. It was done independently and at the same time by the lab of Chris Evans in UCLA in my own lab.
We published actually the same week, which was fortunate, so we got equal recognition, and since then we collaborate in the context actually of NIH CenterGround, which is great. and soon after the cloning of the delta opioid receptor gene homology cloning provided the other genes and now we have three genes encoding mu delta and kappa receptors and it took another 20 years until we could produce enough of the recombinant protein to be able to crystallize these receptors and obtain and solve their structure by x-ray sorry x-ray crystallography and this was achieved in 2012 both by the lab of Brian Kobylka and the lab of Ray Steven that solved the structure of mu, delta, and kappa and the ORL receptor. I'm not going to talk about this fourth receptor.
It's an homologous receptor. This structure was solved, the receptor bound to antagonists. So we have the atomic structure of inactive receptors and... Three years later, we got the first structure of the active mu-opiate receptor.
So now we know a lot about this protein. We know them atom by atom. And these receptors belong to the big G-protein-coupled receptor family. We have thousands of gene encoding for G-protein-coupled receptors throughout the body.
They are extremely important as biomedical targets. Half of the drugs that are used to treat people are agonist or antagonist of G-protein-coupled receptors. Thank you. and the opioid receptors are three members of this big, big family. So if you want to think about the opioid system as in a unitary fashion, we can think about this system as a system that exists in our brain to teach us beneficial behaviors and help us cope with stress.
So this system is important to regulate reward and aversion processes, and I will talk a lot about that in the upcoming... talk. It's also a system that is extremely important to reduce pain and cope with stress.
I will talk a little bit less about that. And it also regulates a number of autonomic or peripheral functions, like the immune system, respiration, and other things on the periphery. And you see from this slide that the three receptors are distributed very differently in the mouse brain on the left and in the human brain on the right, so clearly indicating that they fulfill different functions.
So this leads me to the outline of this talk. The first part will be focused on the mu-opiate receptor because this is our problem in the opiate crisis. The second part will be more general about the role of the free-opiate receptors in physiology. And I will focus on mood state and pathology drug abuse. And the third part will be, now we'll enter more specifically into brain circuit to see how this receptor operates.
within brain circuits. And the fourth part will be about a new approach we developed in our lab, which I call translational neuroimaging, where we try to understand in mice, how does this receptor shape brain functional connectivity? So let's start with the first part.
So I told you we have three receptors, highly homologous, differently distributed. And they fulfill many functions. So which receptor is doing that? So this is a kind of our founder work, creating mice lacking a gene.
So we created mice lacking the mu opioid receptor gene, and we found that any activity of morphine was purely and simply ablated in these animals. Morphine was not producing analgesia, was not pleasurable. There was no morphine dependence. We've repeated exposure.
There was no constipation. There was no respiratory depression. There was nothing. We could even inject 25 the little dose of morphine, these animals would not die from overdose.
So it was clear from this very simple experiment that this mu receptor is actually the single and the essential molecular target that mediates both the therapeutic effect and the adverse effect of the opioids that we use in the clinic or that are abused in the street. Actually, I extend this. This was the original experiment, but there's many more experiments later confirmed this.
So what you know about opioid effects in the clinic, or in terms of opioid use disorders, are effects that are mediated by this particular receptor. And by the way, if you use mice that lack either the delta or the kappa receptor, they respond perfectly normally to morphine. So this was really disappointing at that time because people thought, well, it would have been nice to have two receptors, one receptor for pain or nociceptive control and one receptor for euphoria.
And then we would design drugs that would actually be specific for one of the receptors. But what if now it's a single receptor protein that mediates all these effects? Happily, science goes on, and there were a few revolutions in the field of G-protein-copper receptors.
And one revolution is in signaling. So let me very simply go through. What happens when an opioid receptor signals?
The agonist, the drug, will bind on to the receptor, which is a 7-transmobrine receptor. This will recruit a number of effectors. Some are G-proteins, and in this case, these are inhibitory G-proteins, which recruit other effectors.
Or there may be a recruitment of non-G-proteins, like beta-restin. So there's two main pathways. But in fact, okay.
So clearly, there's a whole complex that is responsible... to inhibit the neuronal activity. So I don't go into details.
These effectors can be ion channel, kybincyclic AMP, every. event of activating an opioid receptor will inhibit the neuron and there will be long-term effect because there will be a number of cascades that are engaged, go to the nucleus and modify transcription. So typically we have an inhibitory G protein GI coupled receptor.
So what is important is not only the receptor, it's the agonist receptor effector complex. And this gives you an example of a very recent list we did for a search in the literature. of all the signaling effectors that have been identified as effectors of the mu opioid receptor.
So this receptor is able to act on all these pathways. So this leads us to what we call biased signaling of functional connectivity. This concept has been developed in the G-protein copper receptor field, so it's not unique to opioid receptors. It's general for these receptors. The concept says that depending on the drug, The active conformation of the receptor will differ and different signaling effectors will be engaged.
So this is just a kind of theoretical example where drug 1 would for example recruit A and B and efficacy would be low, drug 2 would recruit A only and we would have an optimal drug and drug 3 would recruit C and B and this would be really bad because it would mediate adverse effect. So this led people to think about trying to design a drug. by targeting specific signaling pathway and in this way separate good and bad effects. So regarding the mu opioid receptors, there is a very nice effort in this direction.
You see a mu receptor here, you see the two pathway, the GIGO and the beta-arrestin, and a very early work from Laura Bond showed that if you delete beta-arrestin in a mouse, morphine has better effect, nice analgesia, less tolerance. less respiratory depression. So it seems from this early study that if you avoid the beta-arrestin pathway, you may be able to have a better drug.
So this is how the company Trivena developed a drug through classical screening efforts that would be biased for GIGO. So basically it binds, it does activate specifically GIGO, it doesn't engage beta-arrestin, and this drug has actually better properties and is now entering clinical trial phase three. There's another drug that has now been developed at Scripps by Laura Bone, which is also highly GI biased. It's again a kind of screening and medicinal chemistry effort. And I'd like to mention this third approach that was used.
It's very important because it's the kind of first case where virtual docking was performed on the 3D structure of the receptor. So this is extremely efficient because now you can dock virtually millions of molecules. on the atomic structure of the receptor.
You can select thousands of compounds. You can rank their ability to bind to the receptor. And then you can specifically pick chemical families, chemical structures that are totally new.
So this way, this group, it was mainly Kobylka and Brian Roth and Brian Stoichet, were able to isolate about 2,500 new chemical structures that potentially were good. And now you go to the wet lab. you synthesize the first 20, and you start working with them.
And they found, they developed this molecule, PZDEM21, which actually is totally GI-biased, and which, interestingly also, has different properties, so that respiratory depression seems to be lower. So why is this important? Because I think if we think in the context of G-protein-copy receptor, this type of approach allows now to accelerate the pace of discovery.
So this is why the structure of these receptors are important. We now have drugs that have very different biology, in fact. I don't know if you can call them opioids, but they bind to the mu receptor for sure.
And there's a last question I'd like to mention here, which is a really hotly debated issue in the field. Is the fact that this compound in cell system show biased activity. which means they don't activate all the effectors but only a part of them, is this responsible for their differential effects in vivo? There's a big gap between the two. We're not sure.
So at the moment, correlations are made. And we hope in the future to address this by looking at the brain connectivity. I'll go back to this in my very last slide. And this last slide is to tell you that there's different ways to modulate the mu opioid receptor.
mu opioid receptor signaling and try to develop novel drugs beyond biased agonist. One way is, for example, to develop allosteric modulator that you see on the left. So they basically... facilitate endogenous signaling.
There are agonist allosteric modulators that have been built for the myoreceptor, but they don't work in vivo as yet, so to be developed. There's another way, which is to increase the level of endogenous peptide. You see RB101. This is a blocker of enkephalin degradation.
So that's another way to increase myoreceptor activity in a natural manner. But for now, there's nothing really on the market. Or there's effort to develop peripheral opioids, of course, that do not enter the brain, or also to develop kind of bivalent ligand and use over-opioid receptor to modulate the activity. So overall, the take-home message, too, is that there's many strategies underway to try and reduce the adverse effect mediated by activated mu receptors.
So I'm going to stop with this part, and I'm going to move to the second part. So opioid receptors in physiology. So when you look at, when you use mu receptor knockout mice, so they have no mu receptors, I told you that they don't respond to morphine, so morphine is not pleasurable. Fine.
If now you expose these animals to overdrive... of abuse that normally act at their own receptors, you will find, you will see that these mice are not interesting, or these drugs are not reinforcing when the mu receptor is gone. So this is interesting because it tells.
us that mu receptors are not only mediating rewarding properties of the opiate that binds to it but also rewarding properties of other drugs of abuse likely through endogenous on kefalon release this has not really been proved so we have a general mediator of drug reward and if we push a little bit further we could look at natural reward what are natural rewards natural rewarding stimuli or situation that that actually are important for us to learn that we need to do this for our species or our self-survivor. Like eating is a natural reward, sexual activity is a natural reward, social interaction is a natural reward. So social interaction actually are extremely rewarding and social rejection is extremely painful. That's a really developing research area.
And I'd just like to mention here that we tested the possibility that social interaction interactions are mediated, social rewards, sorry, are mediated partly at least for the myo-opioid receptor. And this was actually demonstrated by a paper we published a few years ago, which showed that myo-opioid receptor pups that are four or eight days old show reduced maternal attachment. So to explain simply, maternal behavior is rewarding for the pup.
And this is important for the pup survival. The pup will react to this to have the mother doing this again and again. And this is how bonding takes place.
These knockout pups do not experience the pleasurable feeling of mother maternal behavior as well. As a consequence, bonding doesn't take place as well. And this actually has consequences and we recently demonstrated that adult muon knockout mice show an autistic-like behavior.
So there is an important role of the mu opioid receptor beyond drug... drug abuse, which has to do with natural reward. It's very important.
Now, very briefly, if we do all these experiments using delta-opioid receptor knockout mice, we will find that drug reward is not very much changed. But on the other hand, we will find that these mice show high level of anxiety and a despair behavior, which we do not see in mice like in the mu or the kappa-opioid receptors. So we found at that time this was in 2000, that in fact, this receptor has potentially anxiolytic and antidepressant activity, which led drug companies or medicinal chemists to move their drugs, their delta agonists that were developed to treat chronic pain, into psychiatry. And there were clinical trials, unsuccessful up to now at least, to develop delta agonists as antidepressants.
So now I basically summarize studies. made by many, many labs throughout the world. We know that activating the mu receptor is pleasurable, and this clearly is important in the initiation of addiction.
I'll show this to you in the next slide. We know, on the other hand, I didn't talk about kappa. I summarized here. that activating the kappa receptor is highly dysphoric and that the two receptors oppose each other lightly through modulation, opposing modulation of dopaminergic transmission. We also know, if we now switch to mood state, that repeated stress or repeated exposure to drug increases kappa opioid receptor activity, kappa opioid receptors and dinorphin activity, and this contributes to the negative effect of, let's say, stress and addicted individuals.
On the other hand, we know that activating delta receptors, anxiolytic and antidepressants. So in drug discovery effort, it is recognized today that activating delta or blocking kappa or feasible strategy to help to improve mood. Okay, so this is physiology. Let's now go to pathology and to the addiction cycle. So I apologize for my colleague Warren, the addiction field, because they know this by heart, but I'd like to remind you that addiction is a complex brain disorder.
Addiction develops from recreational or social drug use. that gradually switch to binge intoxication episode. These binge intoxication episodes then are followed by a withdrawal state, withdrawal when the drug leaves out of the body, and this is a very aversive state, which is then followed by what we call a preoccupation anticipation period, that is commonly known as craving, and then craving leads you to consume again and go in a new binge intoxication episode. As you go through this cycle, the brain adapts.
Tolerance develops to reward. The drug is less rewarding, but aversive effects of withdrawal are getting stronger. Craving is getting stronger.
Motivation gets biased towards the drug instead of natural rewards. Self-control is reduced, and all this leads to reinforce the addiction cycle. Within this pathological process, it is clear now we know, so now I position the opioid receptors. The mu receptor is involved in the early step of drug consumption.
It is well established that the kappa receptor and dynorphine, which is his favorite peptide, are involved in the low mood of withdrawal. And it is likely that delta receptor are involved. There's less evidence for that in that part. Maybe delta activity is deficient, but this is not really known.
And I did put on the other side of the cycle mu, delta, and kappa because there is some literature about these receptors also acting in higher order processing. But there's much more to do in this aspect. I think this is a field of research that will develop in the future. So the take-home four is that these receptors contribute very differently to addiction.
Now, let's go back to the mu opioid receptor. So where do these receptors operate? And in fact, now, well, up to that time, we've been doing very simple things. We took the receptor away, when they look at behavior, in the middle we have a big black box, and we say, oh, this receptor is important for this behavior, but we cannot understand the mechanism underlying the change in behavior if you don't understand the brain circuits involved.
So there's two ways you can do this. You can either move in and then look at all... every little tiny neuron and neuronal pathway.
So these are approaches that you can do using conditional gene knockout approach, for example, and then optogenetic to manipulate the neurons. So we did that, and I'm going to show you a little bit of that. And there's another way is to zoom out and look at the whole brain and see what happens and how does the mu receptor activity shape the brain connectome. So we are doing both.
So I'm going to summarize the micro scale in just a few slides. Before you do a conditional knockout of a receptor, you know that you need to use CRE-MICE, that express query combinase, in specific cell population, and this is where you're going to target the knockout. In the case of opioid receptor, it's a problem because we don't really know in which cells this receptor are expressed.
This is a general problem in the G-protein copper receptor field. Antibodies. in vivo are not good and it is extremely difficult to get cellular resolution. So we developed a tool and this was Gregory Scherer, a student in my lab who is doing extremely well now.
He has become truly American and at that time we developed this in collaboration with Alan Basbaum and the idea was well why don't we look at the cell Look at the receptor. It would be nice to see them. So the idea was to develop a knock-in mouse model where we would introduce a fluorescent gene in frame, within the sequence, just before the stop codon so that these antibodies... would naturally produce a receptor which is visible.
And of course, well, we had to be sure that the receptor is functional, so we created this C-terminal fusion, and this is now shown for the delta receptor. this was our first attempt, and this worked extremely well. So we created these mice that express a functional delta receptor, which is expressed at physiological levels throughout the nervous system. And now you can see the receptor.
I mean, this was a miracle at that time. It took so much time to clone this receptor gene a long time ago, and now we could see it. It was a miracle. So you see this cell is coming from the brain of these mice, and you can see the receptor, which is concentrated on the...
surface and now I throw in an opioid and you can see that the receptors in real time cluster and then these endocytic vesicles go in so this is happening in 20 minutes I just show you this contracting in eight minutes so after that you can try and come with a Delta agonist there will be no response the animal is totally tolerant the risk receptor has disappeared it's inside the cell it's not accessible anymore That's a really interesting process. So just to summarize many studies we did, this tool allows us first to revise receptor anatomy, and there were many, many things that were revisited once we could see the receptor. Second thing, we could see real-time trafficking, which allowed us to differentiate between agonists that were all able to reduce anxiety, but some were internalizing the receptors, some were not, and in fact tolerance...
develop very differently depending on whether the drug is able to internalize the receptor or not. And finally, it allowed us to see that physiological internalization, which is internalization induced by endogenous opioid, is very different from pharmacological internalization. Here I show you pharmacology.
The drug hits the system, the receptor all hide inside the cell. In physiology, it's different. When opioid, we try to do that and we show showed that in fact under physiological situation when endogenous peptide are released, the receptor disappears only or internalized only in a very small proportion on neuron and in a very transient way. And you never wipe out all the receptor from the surface.
So there is no real tolerance using endogenous peptide in a physiological setup. So that's the three messages. So we made this with other receptors. So we also created new receptors that were Red and we now have new receptors that are yellow. So we have mice with colored receptors They are all functional and this allows us now to know in which cells our receptors are expressed So to make a long story short with the kind of studies we've been doing I Show you a very simplistic view of the brain and I show you brain regions Which are involved in the different stages of the addiction cycle.
So brain region engaged in the binge intoxication stage in orange brain region involved in the aversive state of withdrawal in blue, and those higher order circuits that are involved in the craving anticipation step. And all these regions express my opioid receptors. So when the brain is exposed to opioids, you modify actually neurons for all these circuits. So which are the receptors that are responsible for reward processing? Where are they?
Well, the traditional hypothesis and this... This was done since, I mean, this is research since 50 years that clearly showed, using many approaches, that mu receptors here in the VTA, which hosts dopaminergic neurons, are able to disinhibit dopamine and then release dopamine, help the release, and this contributes to reward processing. Now, in terms of genetic manipulation, we've been able, and with colleagues, to either delete the receptor in the... the dopamine-minoceptive field and see what those receptors do there, and to delete the receptor or rescue the receptors in D1 medium spiny neurons. So this is a specific population in the nucleus accumbens where receptors are mostly expressed.
And we were able to show that these receptors here are extremely important in motivational state. Now, I'd like to say at this stage that there's certainly other sites for opioid reward that have not been studied there. there yet and that are dopamine maybe independent, not necessarily dopamine dependent. And some sites could be, and this is the beauty now because this comes from the mu cherry mice, mice with fluorescent mu receptor.
They have beautiful mu receptor expressing neurons in the lateral hypothalamus and in the septum which are known to be hedonic hotspot. So there's many places where mu receptors could produce or facilitate reward processing. Now.
So the TACOM-5 is that mu facilitate drug seeking and taking within reward circuit. But these mu receptors are also expressed in what we call aversion centers, the brain centers in blue. So the question is, what do these mu receptors do there? What happens if they are excited by a drug, an opioid drug, or overstimulated by chronic exposure to opioids?
So I will just tell you two little stories. One is the story of Habenula. You may have heard about these tiny brain structures which connect four brain regions with midbrain regions.
And habenula is considered a brain aversion center, which actually is activated under conditions of, well, in different situations. Sorry, I just lost it. What you can see here are mu receptors.
in the medulla benulla, and in fact, this is the brain area that expresses the highest density of mu receptors. Most receptors are there. Actually, no one studied them here.
So we have an aversion center, and what we found, so we were able to delete mu receptors specifically in the medulla benulla, while these receptors are perfectly there, intact, everywhere else in the brain. And what we found is that naloxone, Mu receptor blockade, which is normally aversive, is much less aversive. So, to make a long story short, the way we interpret these results is that we believe that mu receptors in this aversion center normally inhibit aversive encoding, and this is very important, and this is a new role of the mu receptor, which could act as an anti-aversion receptor. So it can be pro-reward in reward circuitry, it can be anti-aversion in aversion circuitry. Another last thing I'd like to talk about is that we have been developing a model, what we call a mouse model of protracted abstinence to opioids.
You can expose mice to repeated opioids that make them very dependent, highly physically dependent for a week. Could be morphine, could be heroin. You can then leave them go through spontaneous withdrawal. for a week, two weeks, three weeks, four weeks. And then what you're going to see is the development of a despair-like behavior and social withdrawal, which has good face validity, in fact.
So we were able to show that this is very long-lasting. It will last four weeks, ten weeks. For a mouse, it's very long. We actually don't know how long.
And it can be prevented by fluoxetine, which is a serotonin uptake inhibitor. so there's a serotonin mechanism there, and it can be blocked by neuroBNI, which is blockage of the kappa-opioid receptor. So the interesting experiment I want to tell you about is that we actually deleted the mu receptor in the dorsal raphe nucleus, which is the center, the home, for serotoninergic neurons. We then exposed the animals repeatedly to the opioid. We let them go for four weeks, and what happened is that the social withdrawal that we see in control animals does not develop in these animals, while other signs of abstinence develop.
So this means that the chronic activation of mu receptor in a serotoninergic center of the brain is responsible, in this case, for the social interaction deficit, or the emotional deficit we find in opioid abstinence. So this is to say that mu is not only a reward, but when you activate mu receptors... you activate and you modify many, many cells and neurons in the brain.
And those that are in the aversion center are probably as important and contribute to the problem, what we call the problem. Now I'm going to move to the last part. Okay, doing pretty well. So from microscale to mesoscale. Can we move now from...
Why wouldn't we look... at the whole brain and maybe be non-invasive and translational. This is a kind of frisky project I entered into about 10 years ago or maybe less, seven years ago. In human research, neuroimaging is instrumental and is maybe one of the few ways ways to look at brain functioning. And there's a host of literature about human imaging studies that are able to link genotype to phenotype, to show biomarker for disease, and even to show...
show drug effects. This is a little bit more ethically difficult in humans. Now, why wouldn't we transfer this type of science into rodents? And in addition, and cherry on the cake, we would be making... mechanistic because in mice for example you can delete any gene you want anywhere you want you can modify and Activate or inhibit any targeted neuronal population So there's a lot of manipulation you can do to understand what happens and to look at the consequences on the whole brain So we decided to go for gene and drug effects on so WBFC means whole brain functional connectivity, so this is how networks or neuronal population talk together.
So we have a little problem with mice because what you see here is a human brain and what you see at the bottom right here is a mouse brain. So it's very, very tiny. So it means you need a very big magnet. And I met Juergen Hennig about 10 years ago.
He was one of the king of MRI in Europe, in Germany. He had a big group. Most people worked as usually in humans, and there were a few brave people working in rodents, and Laura Arsene was working on the mouse brain. She was the only one, and at that time, she had developed incredible high-resolution ways to look at the mouse brain. One way was to look at resting states, so you look at slow blood, bold oxygen level fluctuations in the brain and the resting state, and the other type of approaches she used are fiber tracking.
and I just show you, this is structural MRI, just to give you an idea of the type of resolution. Her resolution is similar to what you can get in humans. So we decided to try and compare control animals with knockout animals.
So basically a normal animal with an animal that is lacking one gene. Is there any difference we could see? And we compared resting state activity of his brain.
It's a completely data-driven approach. so we had no hypothesis. This is what I really like.
So we look at the whole brain. We have no preconceived hypothesis. She was able, using independent component analysis, to find about 100 components, which means these are groups of voxels that fluctuate together. They have a synchronous activity.
And in fact, these voxels correspond to anatomical brain regions, which is good. So now when we have these components, we can look at long range... communication. So we look at synchrony, how these components fluctuate with time, and whether we have synchrony or not.
And we can now separate the group. We have wild type, we have knockout, and we can build the correlation metric. How does each of these components talk to the other component?
And then we can subtract the two matrices, and we can have this type of fingerprint. This barcode or fingerprint shows you the differences between a wild type and a knockout and a knockout. The fact that the myoperic receptor was deleted reshapes the brain connectome.
And it looks like it's happening everywhere. It's very widespread. But if you use statistical analysis, you will see that most changes occur in brain area that process negative valence information.
So negative effect. You have pain region, periodical gray. You have amygdala.
You have anterior cingulate, amygdala, habenula. So this fits extremely well. what we know from the function of this receptor.
We went on, so now we are at the stage of methodological development. We found, we made a volumetric analysis, and we found a number of volumetric changes in the knockout that correspond to the actually functional changes, so it made total sense. We also can work and develop hypothesis-driven analysis, and for example, we can sit on the prefrontal cortex and look at how the prefrontal cortex talks to the rest of the brain. And what we can see is that the major changes in these knockout mice is the decreased dialogue between prefrontal cortex and nucleus accumbens, which fits totally with the fact that these mice show lower motivation for rewards.
And we can do other studies. I don't want to go into details. The important last thing I want to say is that we can now use a knockout animal for another receptor, which has a totally different function. This is a GPR88 receptor.
and we find a very different... Different fingerprint in this case major changes are in the cortex which fits extremely well with these animals that have ADHD type phenotype. So the take-home message 7 is that gene to connect to mapping is feasible in the mouse. So perspective well first understanding of course we're going to apply this technology now to abstinent animals or animal that are exposed to opiates become abstinent and then try to treat.
these animals and see what is changing in the brain. We can try to think about translation and use this fingerprint as biomarkers, why not? So we have started studies with humans and compared humans that have a polymorphism in the opioid receptor genes. We were able to find using a brain bank at the Douglas that in post-mortem tissue the mu receptor is very strongly expressed in the human abenula. as it is in the mouse, we were able to compare different phenotypes of human individuals, healthy individuals, and we were able to find the Habenula component as a coherent entity, and found that if you compare the groups, Habenula talks differently to other parts of the brain in the GEG variant individuals.
So this is kind of prospective, this is what we're developing. And the last part I'd like to talk about is the drug effects. We are trying to see now how mu opioid agonists modify the brain connectome.
And, for example, we started a big study where we compare how morphine, fentanyl, duprenorphine, and more drugs, which all activate mu receptor, change the brain. So what is interesting here, first, you cannot do this in humans. There's no way you can have ethical approval to test as many drugs.
It is very, very heavy. You can do this in mice. Second, so the commonality is that they all act from your receptors. They all show analgesia.
They all reduce pain. They have differences in their pharmacokinetics. They are not all as rewarding when you do behavioral assay. And some of them have biased signaling properties.
So they don't activate signaling effectors the same way. So. So the way we designed the experiment is to try and capture the effect of an acute drug injection in the scanner. So the animal is recorded under resting state, then receive the drug, we keep on recording, acquiring images, and we do this both in control groups or in mu knockout mice, because if we do this in mu knockout mice, we're gonna be able to track non-specific effect or off-target effects, and then Make sure we look at mu receptor mediated effect. And just to show you a few results, we are analyzing this data for now.
We can, for example, look at 14 different brain regions, which are involved, actually the color code here corresponds to the addiction cycle. So different regions involved in the different aspect of opioid use disorders, let's put it that way. And each time, and now we can compare, look at the drug effect. Each time you see a line here, it tells you that connectivity or synchrony between these two regions is decreased by the drug. So nothing happens in saline control.
We see that morphine actually changes connectivity dramatically the first time period, and then it gradually fades away. This is a different time period. And then buprenorphine comes up much later. So this tells us that we can capture the pharmacokinetic of the drugs. Now, we can go on, we can sit on abenula, for example, and see how the drug changes abenula talking to the rest of the brain.
And you will see here that morphine, fentanyl, and buprenorphine change things a lot. Okay, let's be vague at this stage. There's less effect for buprenorphine, but what is interesting is if we do the exact same thing in the mu knockout mice, so these mice lack the mu receptor, morphine does nothing, fentanyl does nothing, so they are really mu selective.
But buprenorphine has a lot of effects that are not mediated by the mu receptor, and this is interesting because we know that buprenorphine also acts at kappa receptors. We can then, for example, focus, compare, and superimpose drug effects on the same brain, and you can see images like this where you can see, for example, morphine, fentanyl, buprenorphine effects on the way habanula talks to the rest of the brain. and we see commonalities and differences. So we are at this stage for the moment.
These are very preliminary data. We know there's imitation. We know we're anesthetizing the animal. We know that there's many things that have explained differences across the drug. We know that they have commonalities.
So we're... Developing, I would say, a lot of new approaches to analyze this data. We are also better characterizing the brain state of these mice, etc.
In the long term, I think this is critically important because we'll be able to better understand the opioid effects on the brain. We'll be able maybe to predict, we hope to be able to do that, what is going to be the behavioral effect of the drug. by looking at fingerprints, direct fingerprints on functional connectivity.
And we hope that this will be translational in the future. For now, there's only one study in the world that has compared buprenorphine effect in rats and humans. This was done by the lab of David Borsuk. So, I will conclude.
The first part told you that... The mu receptor is the guilty one, the guilty protein. It mediates both analgesic and adverse effect of opioids.
But many strategies are underway to reduce this adverse effect. So below I noted a little recommendation. I think that's things we should do. There should be many more organism-level studies for this drug development.
The second point was that the free receptor played distinct roles in hedonic homeostasis and emotional... control. I didn't talk about pain.
This is still a whole field. I think what is underestimated and understudied are the opioid peptides. We know much less about the peptides compared to what we know about the receptors.
The third part was about myopoid receptor in circuits of addiction. Things are much more complex than we thought originally. They don't only facilitate reward.
They do other things in other places of the brain. It's going to be important to be able to tackle. the different neurons that respond to the opioids.
And I think that, so I insisted on that, that reversing the negative effect of protracted abstinence will be very important in the strategies. And the fourth part is more methodological. I think we can work in the mouse from drugs and gene to connectomes. I think we're there at the crossroads of mechanizing and biomarker research. And yeah, developing new methodologies is key to the progress.
I'm a big fan of methodology. And I'd like to thank my current team at McGill. I'd like to thank, I would like to insist on the fact that NIH has been funding me since when I was in France, when I moved to Canada.
allowed me to do very risky projects. I think this was very productive. I am deeply thankful to them and to also other funding sources. And I'm deeply thankful to my guys. I mean, they did all the work.
I love them. Thank you very much. Note that one of them, I stuck there because he was not there that day. You see, one is different there. He's doing the post-processing MRI.
So, Dr. Collins had to leave. Unfortunately, he had a phone call, but we want to thank you. so much for the illuminating talk. We have about six minutes for questions.
And then there's going to be a reception in the library that everyone is invited to. So please come. You can talk individually to Dr. Kiefer at that time.
But now we have questions. Thanks. Thank you. Very thought-provoking lecture. From the overall effect of this opioid, what I picked up was it probably has a diverse and multi-effect, not only on some of the naturally existing analgistic, but it also has...
maybe effect on some other receptors. You mentioned depression of the respiratory effect and also constipation. The molecular side or immunological side of it probably is suggesting that inhibiting the natural analgistic side of it is a good idea. Probably induces the genes that are not just very specific for one receptor, but also affect others.
There are also some data that it does affect the diabetic patient probably also through some receptor maybe competition. And induction of, I don't know, enhancement of insulin receptor. This is just a thought. Yeah, I mean, we know that these receptors are also expressed.
Oops. Yeah. These receptors.
are also expressed peripherally, peripheral nerve endings, but also in the peripheral nervous system, and maybe peripheral organs. They are less well studied here because their expression is very low, but there may be, of course, consequences. peripherally that's for sure is there any time frame when the depression of respiratory the depression occurred probably through maybe mitochondrial dysfunction what no no the the networks in the brainstem that are responsible for the respiratory depression are well known the several nuclei in the brain stem that express mere receptors and this is very well studied so these research receptors or in different nuclei and they inhibit the cells and respiration drops. So it's these mechanisms are all membrane. They originate from receptors on the cell surface and then these receptors signal there and it has been shown that once they are endocytosis they keep on signaling inside the cell.
And we don't know about mitochondria, but their internal signaling also mechanism, yeah. Thank you. That could happen in any neuron or in any cells. Thank you.
Yes, at the gene level of the three receptors, is there any significant variability like single nucleotide polymorphisms that would put individuals into this subgroup or that subgroup? Yeah. So I showed you very, very rapidly one SNP that is very well known for the mu opioid receptor.
This SNP is called A1NTG. It is changing the N-terminal sequence of the protein. from an asparagine to an aspartic acid. It modifies the expression of the receptor on the surface and probably modifies function. This polymorphism is occurring, is pretty frequent.
It occurs in 10 to 15% of the population, and it is the most studied. There are other polymorphisms also that have been less well studied, but they are for the free receptor genes, yes. Okay, thank you.
So, there's a lot of medicinal chemistry efforts going on with mu and kappa and delta molecules, but many people study also heteromeric agonists like a mu kappa. antagonists. And I'm wondering, what do you think the prospects are for things like that in terms of successful drugs as well? And maybe you could interpret the answer in terms of the knockout animals.
Well, the bivalent drugs, you know, happen in... pharmacologically sane animals to have specific pharmacological properties that may be interesting. You never have the real proof that you actually bind onto receptors, you know, but why not? You know, whether the receptor heterodimerize, this is always, this question comes back all the time. It's very difficult to show in vivo.
So I'm not sure this happens really. Okay. So why not, I would say. Yeah, Rita.
Thanks for a great talk. So you show very nicely that there are mu receptors that can mediate both a rewarding effect as well as aversive effects. Why does the rewarding effect win in the beginning?
So they're initially rewarding, but both effects should be stimulated by the mu receptor. Yeah, but in fact, it depends on which cell the receptor is expressed. What I said is that in a naive animal, the receptor that is, for example, in VTA neurons that will, you know, these inhibit dopamine probably will be rewarding.
In the habenula, what the receptor probably do is they inhibit or shut down aversive neurons. So they do the same thing except, well, no, they don't do, it goes in the same direction, okay? So that no... Normally you have an anti-aversive mu and a pro-reward mu. Now what happens when the animal is exposed to chronic opioids?
These two receptors will adapt. No one has really tested, do they become tolerant? We know that rewarding effect of drugs is getting lower.
We don't know which receptors get tolerant where. What if the receptors in the medial albinula get tolerant? So they are not able to inhibit these aversive neurons anymore and then you have a negative effect that develops. So what has never been studied, something that we are able to do now because we have mice which expository recombinase specifically in the mu neurons.
What happens in those cells in different circuits when the receptor is repeatedly exposed? We don't know. So this is what we have to do now.
Thank you. Can I ask you one general question before we... Given the role of dopamine in reward as well, why aren't dopamine antagonists, I mean agonists, drugs of abuse? They're not at all, right? Well, only...
So how does that work? Well, psychostimulants are drugs of abuse. So they increase dopamine, right, because they inhibit the reuptake of dopamine within the neurons.
So they are drugs of abuse. Pure agonists are not. Yeah, so, you know, no, I can't answer that question. Okay, just curious. No, no, I'm sure there's an answer to that.
I have to think about it. Okay, well, I want to thank you again. This is really a brilliant talk. And come to the reception and talk to Brigitte one-on-one.