Activity Transcript

Jonathan Meyer, MD: Hello, this is Dr Jonathan Meyer, Voluntary Clinical Professor of Psychiatry at the University of California here in San Diego with you to talk about TAAR1, a new target with potential for psychiatric indications. What are trace amines? If you've never heard of TAAR1, you have to understand that TAAR1 stands for trace amine associated receptor type one.

What the heck is trace amine? Well, these are compounds that are structurally related to classical neurotransmitters, such as norepinephrine, dopamine, et cetera, but the difference is they're expressed at much lower levels, and that's been an issue for studying them simply because dopamine, for example, is only 1% of our CNS neurotransmitter, and these are at even lower levels than something like dopamine. The other issue in studying TAAR and TAAR receptors is that the receptor location is intracellular, and that made it very difficult to first find them and then characterize their properties.

Finally, in 2001, there was a consensus about what these are and how we're going to name them. We now recognize that these are predominantly intracellular receptors. These are all G-protein coupled receptors, they're not ion channels, and there's 26 variations of these, of which only about 7 are present in human brains, but with relevance to what we're talking about in schizophrenia, where TAAR1 specifically is expressed is very relevant to what we believe is the pathophysiology of the positive symptoms of schizophrenia, but with implications with other neuro circuits that we'll talk about.

Early on, about 20 years ago, human genetic studies noticed some associations between polymorphisms of TAAR receptors and risk for schizophrenia. People then translated this into animal models, and they created what were called TAAR1 knockouts. These are animals which have nonfunctional TAAR1 receptors with the idea of studying the physiology and behavior when we have essentially 100% TAAR1 antagonism. Well, for one thing, they confirm that TAAR1 receptors do form what we call heterodimers with D2 receptors. They glum onto them in an intracellular location and moderate the signal from dopamine binding intracellularly. TAAR1 agonist and antagonists don't directly bind to the D2 receptor active site. They don't modulate the actual binding profile. They alter the downstream signal, and we noticed that already in the TAAR1 knockouts by changes in their dopamine physiology, and we'll go over that in a second.

Lastly, starting in 2011, chemists were able to synthesize TAAR1 agonists and also antagonists, but agonists were the area of interest because we felt that this was a model of an antipsychotic, and that was based on their ability to block cocaine-induced hyper-locomotion. They could also do properties similar to what olanzapine did on these behaviors and some other interesting aspects, some antidepressant activities, some pro-cognitive activities, some improvements in metabolic parameters. We'll talk about the latter maybe a little bit less, but these antipsychotic activities were very exciting and really stimulated research in this area.

Well, as I already alluded to, TAAR1 is expressed in brain regions involved with schizophrenia, but TAAR1 is really implicated in a number of other circuits as well. As most of you know, schizophrenia is not just a disorder of dopamine per se, although that dysfunction as we'll talk about may be a final common pathway for the positive symptoms of psychosis, but TAAR1 and schizophrenia involve a number of circuits. Some of these, of course, implicate glutamate. Many of you may have heard of the glutamate hypo-function hypothesis of schizophrenia, and the basis of that is the finding that at these GABAergic interneurons up in the prefrontal cortex, there is inadequate NMDA related glutamate activity. From that, people have actually synthesized compounds to try to increase NMDA activity with the goal of making schizophrenia better.

Well, it hasn't worked out so well, but one thing we know is that when you give animals and also humans NMDA antagonists, it's a great model for schizophrenia, especially the cognitive disruption associated with schizophrenia. Well, guess what? TAAR1 is involved in those pathways. TAAR1 is also involved in serotonergic pathways. This, of course, has implications for mood. There may be other aspects of schizophrenia pathology, including depressive symptoms, and then of course our favorite is dopamine, but in this case, what we're talking about is the mesocortical circuit. This is going from the midbrain up to the frontal cortex, and I think most of you now appreciate that schizophrenia is not just a disorder of positive symptoms. It's also a cognitive disorder, and dopamine hypofunction in this mesocortical circuit we think is foundational to the cognitive deficits of schizophrenia.

Guess what? TAAR1 is up there as well, but for many of us, the most important thing early on in treating a patient with schizophrenia is getting on top of the positive symptoms of psychosis. I think everyone listening to this is probably familiar with the term meso- limbic. That meso-limbic pathway goes from the ventral tegmental area or VTA up to the ventral striatum. Well, this is based on an animal model, and guess what? In humans, we have a slightly different nomenclature for that, which I'll talk about in a second, but for now I'll use the term meso-limbic, and we know that too much dopamine neurotransmission in this meso-limbic pathway, certainly in an animal model, is associated with the behaviors and phenotype of the positive symptoms and psychosis, and TAAR1 is heavily expressed in this area.

As I just mentioned, the meso-limbic term or the area of the striatum and the pathway where we have excess dopamine activity associated with the positive symptoms of psychosis really came from animal research and some early human imaging studies. When I say early, it's because our resolution wasn't so great. We couldn't distinguish all the parts of the striatum. In the animal model, especially the rodent model, the ventral striatum is much larger than it is in humans, and so the assumption is, well, these circuits going to the ventral striatum are analogous to what we're seeing in human beings, but we now understand that the dopamine overactivity in human striatum circuits is nothing to do with that meso-limbic pathway from the VTA essentially to the reward center in the ventral striatum. It actually has to do with dopamine overactivity in this dorsal media substantia nigra to the associative striatum, that middle structure, and the adjacent portion of the sensory motor striatum.

I'm going to call that circuit the SNASM, which goes from the substantia nigra to the associative and adjacent sensory motor striatum. If I use the term meso-limbic, I'm really referring to the animal data, which is where a lot of our information comes from, but you should know that in humans, the meso-limbic circuit is not the problem with the positive symptoms of psychosis. It is in the SNASM where we see dopamine overactivity.

Well, how does TAAR1 modulate this on the synaptic level? As I already mentioned, TAAR1 receptors glom onto D2 and they do this both presynaptically and postsynaptically. If we give a TAAR1 agonist, we see a couple of things. It decreases the rate of firing of these VTA neurons in the animal model of the positive symptoms of psychosis. That's good. That's what we want. We want less dopamine neurotransmission in that circuit, which is analogous to the human SNASM, and not surprisingly, TAAR1 antagonists make it worse. TAAR1 agonists presynaptically also increased the sensitivity of the auto receptor. I think many of you may have heard that we have D2 receptors not just postsynaptically, but also presynaptically. Presynaptically, they function to create feedback inhibition. When dopamine binds, it kind of moderates the dopamine release presynaptically. While TAAR1 agonists increase the sensitivity to feedback inhibition, and it's another way they turn off presynaptic dopamine release or at least decrease it.

What's also interesting postsynaptically is that TAAR1 agonist help moderate that overactive signal we saw in the TAAR1 knockout models. The way it happens postsynaptically is that, again, TAAR1 gloms onto the D2 receptor, forms a heterodimer, but what it does is it influences a downstream signal. When you have a G-protein coupled receptor, you of course have a G-protein related signal, which influences cyclic AMP and other pathways, and you also have a non-G-protein pathway, which goes down that right side, which implicates β-arrestin-2 and this thing called GSK3 beta. The idea is too much GSK3-β activity is associated with hyper-locomotion, and TAAR1 agonists really helped shut down that pathway, and we feel like this is another way of moderating the dopamine neurotransmission signal for any dopamine which reaches these post synaptic D2 receptors.

The knockout models really have been very instructive in understanding how all of this might work. Behaviorally, TAAR1 knockouts look very similar to wild type mice, but when you give them a dopamine releaser such as amphetamine, they have much higher sensitivity for this, more hyper-locomotion and much more presynaptic release. The idea is you see both the presynaptic and postsynaptic problems when you have too little TAAR1 activity. As I alluded to before, TAAR1 knockouts also have increased firing rates of dopamine neurons, but it's very interesting. Where this occurs is just in the ventral striatum. That's the part of the mouse, striatum, item associated with the positive symptoms of psychosis, but not in the dorsal striatum. What this means is that if I'm giving a TAAR1 agonist, I will preferentially moderate dopamine neurotransmission in the area of the striatum associated with the positive symptoms of psychosis, but not in the motor area as opposed to a D2 antagonist, which is sort of indiscriminate in where it binds and where it causes its effects.

Postsynaptically, again, we see with the TAAR1 knockout mice that they have more active GSK3-β, and if you think about this, they just have a mouse which is more agitated if you want to feel that way, more hyper-locomotion, and that's part of what we see when dopamine binds and that GSK3 beta pathway is highly active. We also see some other interesting aspects which relate to mood and maybe even to addiction. We certainly see in TAAR1 knockout mice increased serotonergic neuron firing. They perform much worse on anxiety tests. They also perform much worse on memory tests, and we think this might be a glutamate pathway. One thing we know from a lot of animal studies is that if I give a glutamate antagonist, specifically NMDA antagonists like PCP or experimental compounds like MK801, they cause cognitive dysfunction.

Guess what? We see that in the TAAR1 knockout mouse, and we feel like some of these properties also relate to the propensity for addictive behaviors, and they study this by giving them alcohol and showing how much more, they like it compared to the wild-type mouse. Just to reinforce, again, when I say meso-limbic, I'm referring to the mouse model, but now we understand for the human striatum that it's the middle section where there's too much dopamine neurotransmission, and it's the associative striatum and that adjacent portion of the sensory motor striatum. That tract I'm going to call this SNASM, but you should know that the science has advanced and we're not treating mice, but we think the problem is still the same, too much presynaptic dopamine release in a certain part of the striatum in the human brain.

Well, what's the evidence that a TAAR1 agonist can actually make an antipsychotic? Well, here's an example. SEP363856 was given in a model where they used ketamine. Again, ketamine, like amphetamines, will cause increased pre-synaptic dopamine release. Ketamine is also good at causing cognitive disruption. Here we show that use of this TAAR1 agonist reduced ketamine induced increases in pre-synaptic dopamine release. Exactly what we would want. What's wonderful for us really as clinicians is that when you give a TAAR1 agonist to animals, so to speak, who don't have schizophrenia, meaning we haven't pretreated them with ketamine, we don't see this effect on dopamine neurotransmission.

It shows you that TAAR1 agonism functions as a homeostatic mechanism. It moderates when things are out of whack, so to speak. In this case, they're out of whack because we gave the animal ketamine, but if we gave a TAAR1 agonist to somebody without schizophrenia or an animal which wasn't getting ketamine or amphetamine, we do not see significant effects. That also is exciting. As you know, if I give a D2 antagonist to somebody who doesn't have schizophrenia, not only do I not get a benefit, I get the downside of a D2 blockade. That's something we don't think we're going to get from the TAAR1 agonists.

To sum it all up, we certainly see evidence that TAAR1 agonists block the behavioral effects of stimulants or NMDA blockers like PCP. This implication is obviously for antipsychotic activity, as well as maybe as a role for substance use disorders. I didn't show this data, but TAAR agonists clearly potentiate the effects of other antipsychotics, but TAAR1 agonists don't induce catalepsy, and they actually reduce the catalepsy from a strong D2 antagonist like haloperidol. I already alluded to some serotonergic properties. We see in pre-clinical models evidence of TAAR1 agonists having antidepressant effects and also some pro-cognitive effects. An exciting thing that we see in animal models is that it actually improves the metabolic profile of TAAR1 knockouts and reduces some of the metabolic adverse effects of agents such as olanzapine.

Now, whether those latter three properties, antidepressant, pro-cognitive, metabolic translate to human beings remains to be seen, but what I can say 100% is that clearly TAAR1 agonists work as antipsychotics because we have human data and let me show you what's being studied right now. We have a couple of agents which are in phases 2 or 3 of clinical development. There's one compound called Ralmitaront. It's also called RO6889450. This is a TAAR1 partial agonist, and this TAAR1 partial agonist was developed from an older compound which had some kinetic issues. It's currently in a phase 2 study for the treatment of an acute exacerbation of adults with schizophrenia or schizoaffective disorder. We also have another compound which is now in phase 3 called Ulotaront. Its earlier name was SEP363856, and that was the preclinical data I showed you with the ketamine model. This is a full TAAR1 agonist, and interestingly, it also has some serotonin 1 agonist activity as well. They already published their phase 2 data in New England Journal in 2020, and we'll go over that in a second.

This is the phase 2 program for Ralmitaront. You can see they're studying it vs placebo in adults with acute exacerbation of schizophrenia age 18 to 45. With most schizophrenia trials, the primary outcome measure is the positive and negative syndrome scale, or the PANSS. People who are on placebo can then be re-randomized to receive active drug after completion of the 4-week study. They're hoping to finish the study by April of next year and then we'll get some data. They also will have some arms of risperidone 4 as an active comparator.

We're very interested in the role of TAAR1 agonists or in this case a partial agonist in patients with negative symptoms, so they also have a negative symptom trial, in stable adults aged 18 to 55, they're using the Brief Negative Symptom Scale or BNSS and looking primarily at the avolition apathy sub-score at week 12. We feel like of all the domains of negative symptoms, this is the one most amenable to pharmacotherapy. They're looking at both a monotherapy as well as an adjunctive study, and we hope we'll have these data in May of next year.

Well, Ulotaront is the most advanced of all the TAAR1 compounds in clinical trials. As I mentioned, it is an agonist. It's not a partial agonist. It also has some serotonin 1A agonist activities, and what this shows you is that the antagonist activity really is primarily related to the TAAR1 properties, but serotonin 1A doesn't hurt. It may be helpful in that regard. It certainly is very helpful as an anxiolytic property, so it's a bonus that we get. We don't feel like you have to have serotonin 1A activity to help out the TAAR1 agonism for the antipsychotic property, but we get this bonus of anxiolytic activity and that's not so bad.

Their clinical trial was so exciting, was published in the New England Journal of Medicine in 2020. What we have published is so far the clinical trial phase 2B study, which was a double blind, 4-week placebo-controlled trial of dosage of 50 or 75 milligrams in adults with an acute exacerbation of schizophrenia. That was published in the New England Journal. The following year in 2021, we had the open label extension, and as most of you know, open label extensions are done for the purposes of gathering predominantly safety data, but we do get a signal for ongoing efficacy as well.

Well, Study 201 was the double blind, 4-week phase 2B study. You can see the patients that were enrolled were somewhat younger, mean age 30, predominantly male, also predominantly white. Mean DMI is 25 because they were a younger population, and this was a sick group. Mean PANSS score was almost exactly 100, very ill population of patients with schizophrenia. Here is the effect over 4 weeks of study. Now, this is an inpatient study. Whether this failure to separate by week 2 is a property of being an inpatient study or something that relates to maybe the different mechanism of action of TAAR1 agonist, we'll just have to wait and see from the phase 3 data, but there's no question that it worked, and we're continuing to see acceleration of separation even at the week 4 endpoint with an effect size which was moderate at 0.45.

Most people were on the highest dose of 75 milligrams, and this is the effect on the negative symptoms. Notice the effect size is the same as total symptoms. That's something we're not used to. We're used to being better generally at total symptoms or positive symptoms and negative symptoms. Also, something very exciting about this novel mechanism, and of course we'll wait to see the replication in the phase 3 study. Most importantly, the tolerability and safety profile are very different than we've had from prior antipsychotics.

All drugs have side effects, not a surprise, but the way to look at it is partly the numeric differences from placebo and numbers needed to harm. You want to see that number needed to harm to be as high as possible, and certainly when it's 500 definitely, you know it definitely is favoring your drug, or if it's negative, it means placebo was actually worse, and not surprisingly, placebo was worse in terms of worsening of schizophrenia or suicidality, but overall the numbers needed to harm and then merit differences strongly favor Ulotaront being a very well tolerated medication, and that translates to the discontinuation rates due to adverse events which were differed by 1.9% vs placebo.

Also, we don't see a big signal for clinically significant weight gain or impact on metabolic parameters either. These are the things we've come to expect from newer antipsychotics. Another benefit from a TAAR1 agonist is that we don't incur a lot of tolerability issues with regard to cardio-metabolic risk. I alluded to before we feel like the preferential action of TAAR1 agonism is in that area of the striatum associated with the positive symptoms of psychosis and not so much in the dorsal striatum. That's what's born out when we look at what we used to call EPS, but now we just call it what it is; Parkinsonism, akathisia, etc.… Again, numbers which look almost exactly like placebo. I think that's important. This is something we recognize as a limitation of our current direct D2 acting antipsychotics, and while we are modulating dopamine neurotransmission, we are preferentially doing that in the area of the brain where there's excessive dopamine activity. Again, the homeostatic effects of TAAR1 agonists are really borne out by these really low rates of neuromotor adverse effects related to effects in dopamine neurotransmission.

This is the long-term open label extension data, which was published last year. Looking at safety, we see weight remains flat. Prolactin went down numerically. We think this is probably removal of a prior effect. Cholesterol, not clinically significant. Glucose, not clinically significant. Again, everyone's going to get monitored because they have schizophrenia, but this is a compound which is not associated with clinically significant changes in metabolic parameters or prolactin over a long-term study.

Here is the long-term efficacy signal. Now, as you know, the sample size does drop off some over the 6 months of treatment, but the retention in this study was quite significant. About 75% of the people made it actually to week 26 who are already at week 4 showing ongoing efficacy. You don't want to over-read this data too much because it is open label and of course people who don't like the drug tend to drop out early on, but certainly we're seeing a sustained response throughout the 6 weeks of this open label extension trial. TAAR1 agonism really is an exciting new mechanism that treats schizophrenia, which involves reduction of presynaptic dopamine release, and also modulating of this oversensitivity to dopamine signaling postsynaptically. It seems to preferentially do it in those areas of the striatum associated with the positive symptoms of psychosis. In the mouse, we'll call that the meso-limbic pathway. In humans, we'll call it the SNASM, but the idea is it does it without directly binding D2 receptors or directly altering how dopamine binds to the receptor. What we're doing is modulating that downstream signal.

We have abundant data from both human genetic studies and the mouse TAAR1 knockout models, which talks about TAAR1 underactivity as a factor in schizophrenia pathophysiology with a lot of abnormalities in the mouse models and TAAR1 knockouts and dopamine neurotransmission. The preclinical data also showed that TAAR1 agonists were effective for our psychosis models, whether it's amphetamine or ketamine without catalepsy induction, and at least in the mouse, beneficial effects on mood, cognition, and metabolic parameters.

The most advanced compounds are in phase 2 and 3 studies. Ralmitaront, which is a partial agonist, is in phase 2 studies, whereas Ulotaront, which is a TAAR1 full agonist and a serotonin 1A agonist is now in phase 3. The phase 2 data for Ulotaront was published in the New England Journal of Medicine in 2020, showing an effect size for total symptoms, which was moderate, but also the effect size for negative symptoms was equivalent to total symptom reduction with a tolerability profile showing very low rates of neurological side effects in an absence of clinically significant metabolic effects. This is really what we've been looking for in an antipsychotic, is an absence of some of these neuromotor adverse effects, also without significant metabolic issues. We're hoping to have phase 3 data over the next few months, and I think this represents one of these exciting new mechanisms to treat schizophrenia without all of the baggage of D2 binding.

This is Dr Jonathan Meyer. Thank you very much for your participation. Please proceed to answer the post activity assessment questions and take a moment to complete the program evaluation.

This transcript has not been copyedited.