Yale Psychiatry Grand Rounds: "Lustman Awards"

May 26, 2023

Lustman Awards with speakers AZA Stephen Allsop, MD, PhD; Sarah Jefferson, MD, PhD; and Yang Jae Lee, MD

Information

ID: 9969
To Cite: DCA Citation Guide

Download Transcript

00:00So with that, I'm going
00:03to okay. Welcome
00:05to those who are seeing
00:06us on the recording.
00:11I'm going to turn now
00:11to our first first winner and
00:141st Presenter AZ. I'll stop
00:18that anymore.
00:31Here we go.
00:34So Izzy Alsop is a graduating resident
00:36this year in the Neuroscience
00:38Research Training program and has
00:41really been a remarkable colleague,
00:43trainee and citizen of our community
00:45during his four years here.
00:46Now that doesn't surprise us.
00:49Because when we when we
00:50first saw his application,
00:53Aza did his his his bachelor's at North
00:56Carolina Central University before going
00:58to Harvard for his MD&amp;MIT for his PhD,
01:01where he worked with K Ty,
01:03who's a truly brilliant neuroscientist
01:05who's done groundbreaking
01:06work over the last 20 years,
01:07work with some of the leading
01:09people in our field.
01:10And I read a lot of letters.
01:12I read a lot of strong letters.
01:14A Ty's letter kind of floated up above
01:16the table because because it was
01:19so strong and so we expected great
01:21things of a CA when he came here
01:23and we have not been disappointed.
01:28Aza You'll hear about his basic
01:30science grounded in the work he
01:32did his PhD student with Kay,
01:33and then going on through some innovative,
01:35exciting new work that
01:36he's done while he's here.
01:37So I want to emphasize a couple
01:38of the other things. Actually,
01:39I haven't gone through all your slides.
01:40I don't know exactly what you're covering,
01:42but that you might hear a
01:43little bit less of today.
01:46One is that in addition
01:47to continuing basic science work
01:49growing out of his PhD studies,
01:51Aza has grown and entirely
01:54new clinical research.
01:57Line of work in in collaboration with
01:59Joy Hirsch and that involves the the
02:02use of near infrared spectroscopy to
02:05image the brain and try to understand
02:07the interactions of the people in dyads
02:10and also how that interacts with music.
02:12And so this is the second thing
02:13I want to tell you about A ZA,
02:15which is a spectacular musician.
02:16He's released several albums he performs
02:19regularly around around New Haven.
02:22The third thing I want to,
02:23I want to say that you may not see in the
02:25slides today is what a wonderful clinician,
02:27clinical leader and clinical citizen he is.
02:30A ZA was one of the chief residents
02:31on the Clinical Neuroscience Research
02:33Unit this year which I directed
02:35on which Doctor Cho is one of the
02:37attendings and it ran incredibly
02:38smoothly and introduced new innovations
02:40and how the unit ran to try to meet
02:43patients needs while we were there
02:45which was true leadership.
02:48And finally, that that
02:49dedication to citizenship
02:51goes goes far beyond AZA's academic
02:54work to his work in in advocacy,
02:57in leadership, in the community,
02:59in antiracist work and so forth.
03:03Aza is also extraordinarily
03:04dedicated to his family.
03:06I just got a chance to meet his mom today.
03:08Thank you for being here with us.
03:10And. Really I I mean people
03:12talk about triple threats.
03:13I kind of lost track of how
03:14many threats we got out here.
03:16So, so anyway it's so it's been
03:17a great pleasure to get to know
03:19Aza over the last four years.
03:22It is a great pleasure to introduce
03:23him to give this presentation and
03:25and the award today and it is a great
03:27pleasure that he will be staying
03:29with us as an assistant professor
03:31transition formally on July 1st,
03:33but that's in process now.
03:34So we look forward to many more years
03:36of of of comradeship and collaboration.
03:39Congratulations on this award.
03:41And I'm going to keep
03:51you this down.
03:52Thank you so much for that introduction.
03:54I feel really, really humbled
03:57and and grateful to be here.
03:59Thank you to the most high.
04:00Thank you to my family for being here,
04:02my mom, sister, my nephew who you've
04:05already heard from my brother,
04:06my wife watching online and
04:08other family support and.
04:10Thank you to the Lesman family
04:12and the committee and this
04:14community which has been an amazing
04:17place to to be for residency.
04:19We're going to talk a little bit
04:21about some of my work that takes
04:23a computational framework and
04:25looking to see how can we better
04:27understand that the code the brain
04:28uses to represent social information
04:30across different model systems?
04:32And then how might we then use
04:33that to be able to start treating
04:35mental health symptoms?
04:39These are my conflict of interest,
04:40which are not relevant to the work
04:41that I'm talking about today,
04:42but is to other research.
04:44And I'm going to start by just setting
04:46a basic frame ground in some of the
04:48work that I did during grad school and
04:50post that because it kind of sets up
04:52the problem that I then was looking to
04:54solve with this project during residency.
04:57And the solution that we're working with is
04:59called Function Functional Encoding Units.
05:01I'll talk about why we think this
05:03might be a really useful way to
05:05help represent neural information.
05:07But before we get into the data,
05:08I just want to just kind of share
05:10yields land acknowledgement,
05:11which just says that,
05:12you know the cool research and cool
05:14community that we get to have and
05:15build comes that at a real price to to
05:17nations and people that we can't forget.
05:20That even as we celebrate the
05:22things that we're able to accomplish
05:24while here on this land,
05:25it's amazing to be in this environment.
05:27Because it's Chris said like I get
05:29to kind of bring the spiritual,
05:30the musical aspects of myself
05:32into the work that I do.
05:33And even though I won't
05:34talk about that research.
05:35It informs how I think about science
05:37and what kinds of questions to ask
05:39and how we might go about asking those
05:41questions because I'm going to talk about,
05:43you know,
05:43neurons in the brain while animals are
05:45doing things and we're recording from them.
05:47But none of that really gets at
05:48the hard problem of consciousness,
05:49which is how do we have the specific
05:51quality of our conscious experience.
05:53And I think asking those kinds of
05:56questions really kind of push us
05:57towards these ideas of like quantum
05:59psychology and quantum psychiatry,
06:01which I think.
06:02Might be very relevant to how we
06:03think about social information
06:05and the way that our experiences
06:07become entangled with each other.
06:08And so with that frame,
06:11I wanna jump into thinking about
06:12why we wanna study the social brain.
06:14It's really at the heart of culture
06:17and how we evolve our social norms.
06:20And many of the big problems
06:22we have to solve today,
06:23like the coordination challenge around
06:25climate crisis really will require
06:27us to be able to to evolve the ways
06:30in which we socially relate to each
06:32other to to guide group behavior.
06:34And obviously as a clinician across
06:36a number of different diagnosis
06:38in psychiatry and neurology,
06:39we see altered social cognition,
06:41some that even are somewhat
06:44defined by by alterations in in
06:47social cognition and behavior.
06:49But.
06:49When we try to look at the brain
06:51and look at the kind of circuits
06:52and networks that might be involved,
06:54we're looking at the rodent brain on the
06:56left and the human brain on the right.
06:57It's it's very complex and many
06:59of these regions do things that
07:01are non social as well.
07:02And so one of the things
07:03that appealed to me
07:04during Graduate School was this idea
07:06that we can use systems near science
07:07tools like optogenetics to look at a
07:09specific circuit and ask what that
07:11circuit does in some behavior and
07:13then from there build build up a
07:15circuit understanding of behavior.
07:16And so I wanted to apply this to
07:19social behavior for my thesis work.
07:21And so to kind of reduce our framework
07:22to be able to make these kinds of
07:25potential hypothesis and conclusions,
07:26we focus on a specific element of
07:28social behavior, social learning.
07:29It's so critical for animals to be
07:31able to learn about things in the
07:33environment that are happening,
07:35specifically the kinds of things
07:36that they should avoid, right.
07:38There's a high cost to certain lessons
07:39and you want to be able to use the
07:41experiences of other animals to learn.
07:43So while this baby element has
07:45definitely been traumatized,
07:46it hasn't been physically injured and
07:48that gives it some a much harder.
07:50Higher chance of survival.
07:51And we already knew from experimental
07:53paradigms in the past that rodents
07:55will do this, that kind of behavior,
07:57and you can measure it in a lab setting.
07:59And we knew that areas like the amygdala
08:01and the anterior single cortex were involved,
08:04because if you do something like
08:05put lidocaine in those regions so
08:06that they're no longer active,
08:07animals aren't able to learn.
08:08And you can see that in this
08:10white line here where freezing is,
08:12is on the Y axis.
08:14So we also knew that in humans
08:15these same two regions,
08:16anterior singlet cortex and the mega,
08:17are involved in social learning and there's
08:19just showing you one example of that.
08:21And so let's zoom in a little
08:22bit on those two regions.
08:23The cingulate is an area that seems to
08:27be kind of in between the subcortical
08:29and cortical regions in certain ways
08:31and it kind of can be thought of as
08:33a self other hub that helps to look
08:35at contingencies in the environment
08:37and then make appropriate decisions.
08:39The amygdala again also has some degree of
08:41homology across different model systems.
08:43And now we think of it as sort of
08:46a behavioral hub where appropriate
08:48actions can be initiated depending on
08:51the right stimuli and associations.
08:54And so they seem to be reasons that
08:55would be important for this kind of behavior.
08:57OK.
08:57So we know that they're important for
08:59social learning in rodents, primates.
09:01I didn't show you that data and also humans.
09:04And we knew that they have anatomically
09:06reciprocal connections to each other.
09:08And so a large part of my thesis work
09:10was looking in a specific version
09:12of social learning in rodents and
09:14asking how do neurons actually encode
09:17information during this paradigm.
09:18So I'll show you what this paradigm
09:20is because it sets A-frame for the
09:21some of the analytical tools that
09:23I'll show you in a little bit.
09:25And so here an observer mouse
09:26receives 1 foot shot.
09:28It's then transferred to a
09:29plastic floor where it will
09:31no longer receive any foot shocks.
09:33And then a demonstrator mouse who's a cage,
09:35made this place and it goes undergoes
09:37basically Pavlovian cue fear conditioning,
09:39where a cue predicts the delivery
09:41of for shock to that animal.
09:42We can then test that animal on
09:44its own where we just play the cue
09:46and ask does it freeze to that cue.
09:48If it does, we're saying it learned
09:50that association through the
09:51experience of the other animals.
09:52We did a number of controls.
09:53I won't show you to really convince
09:55ourselves that that is indeed what was
09:57happening at the behavioral level. OK.
09:59So now we wanted to record from these
10:01neurons using in vivo electrophysiology
10:02to ask what are they doing,
10:03doing this kind of task.
10:05And we'll add this period called
10:07habituation here in the Gray where
10:08we're giving cues to the demonstrator,
10:10but they don't predict anything happening.
10:12And so those cues have no meaning.
10:14Then we'll start actually paying
10:15them to shock of the demonstrator
10:17and asking how do neurons change
10:19as animals actually learning that
10:21this cue has a predictive value.
10:22And so first,
10:24I just want to give you the sense
10:25these are rasters basically showing
10:26what these neurons are doing,
10:27doing this behavior.
10:28And I just want to show you that
10:29these neurons are responding to the
10:31things that we need for animals to
10:32make these associations, right?
10:33The cue, the shock of the demonstrator.
10:36And some neurons actually even
10:37respond to both.
10:38But I want you to already notice that
10:40the actual sort of even aesthetics
10:42of these waveforms and rasters look
10:43different across the different neurons,
10:45even if they're representing quote,
10:47UN quote, the same thing.
10:49We were really interested in these
10:51neurons that during habituation
10:52seemed to not really have a strong
10:55response to the cue.
10:56But then when animals are actually
10:57learning about the meaning of the cue,
10:58these neurons start to respond.
11:00So these are the kinds of neurons that
11:02might actually represent learning
11:03because they change their response
11:05to the cue as animals are learning.
11:07And so we honed in on these neurons and
11:10we wanted to take a tool from engineering.
11:12This is the same kind of tool that
11:15we use to look at GPS and ask,
11:17can we use this sort of state space
11:19approach to model how neurons are
11:21not just firing at a given trial,
11:23but how they actually change their
11:25firing during this sort of task?
11:27And it also allows us to get some
11:30confidence about what these estimates are.
11:32And so we can take the actual neuronal data,
11:34run it through our model,
11:35and it provides us an estimate at each trial.
11:38And then over time,
11:39we can track how these neurons change.
11:41Just like a G PS:,
11:42can track sort of your trajectory
11:43along a certain a certain path.
11:47So this is for one neuron.
11:48It allows us to identify when do
11:50these neurons actually learn.
11:52But I want you to appreciate here
11:53that when we
11:54record from multiple different neurons,
11:55we kind of run into a problem.
11:57You can look again just looking at the
11:59waveforms and the rasters that neurons
12:01that are excited or inhibited don't
12:03all do that in exactly the same way.
12:05And so we can take a gross approach and say,
12:07well, all of these neurons are inhibited,
12:08all these neurons are excited.
12:09We can say these neurons have basic
12:11responses or they have sustained responses.
12:14But that seems like a very gross
12:16level of analysis and unlikely to be
12:18the code that the brain is actually
12:20using to represent social information.
12:22If we dig down into each individual neuron,
12:24then it's also difficult to understand
12:26how each individual neuron will
12:28contribute to some sort of code.
12:30And so the idea that we had was,
12:31can we take the same kind of
12:33state space modeling approach,
12:34but combine it with unsupervised clustering
12:36approaches so that we can figure out
12:38in any given population of neurons,
12:40which ones are actually belonging together
12:42in terms of how they functional respond?
12:44Because these might be the putative
12:46actually units that are encoding
12:48the kind of information that the
12:50brain needs to encode.
12:51And so I teamed up with Alexander Lynn and
12:55and then Baba to create this pipeline for
12:58creating these functional encoding units.
13:00And so we can model the rate of each
13:03neuron like I showed you before.
13:04Then we're going to use an unsupervised
13:06clustering approach to figure out
13:07within this population of neurons,
13:09which neurons go together in terms
13:11of how they respond.
13:12And then this allows us to derive
13:14what the functional ensembles
13:16are within a given region.
13:18So the first thing we wanted
13:19to do was to test this and say,
13:20can it actually pull out ensembles
13:22if we know what the ensembles are?
13:24So we simulated 50 noisy neurons that we
13:26know belong to five ground truth clusters.
13:28Either they don't respond at all,
13:30they're excited either in
13:31a sustained or phasic way,
13:33or they're inhibited either in a sustained,
13:36yeah, a sustained or phasic way.
13:37And we want to ask, can this pick it out?
13:40So before I show you that,
13:41I just want to give you an example
13:42of what one of these functional
13:44encoding units actually look like.
13:45So this is an Fe from the ACC.
13:48That's 36 neurons in it and each color
13:49is an individual neuron and these
13:51rasters are just overlaid on each other.
13:53And for those of you familiar with
13:55looking at this sort of in vivo EFIS data,
13:57you'll see that it kind of resembles
13:59what a native neuron excitatory
14:01response would look like.
14:02It doesn't look like noise
14:03or anything like that.
14:03So this gives us some clue that we we
14:06might be moving in the right direction.
14:08OK,
14:08So what I'm going to show you here on
14:10the left is in green the parameters for
14:12each of these ensembles in the ground truth.
14:15And so these two parameters,
14:16which would be important because
14:18it allows us to actually make
14:19intuitions and hypothesis about
14:21what these ensembles are doing,
14:22are jump and phasicity.
14:24So jump says when a neuron actually responds,
14:28how strong of a response is it,
14:31and phasicity says how sustained.
14:34Or Phasic is responsive,
14:35it's above 5, we'll say that's phasic.
14:37If it's lower than five,
14:39we'll say that that's sustained.
14:41And So what you can see here is
14:42that we're able to actually pull
14:44out each of these five clusters
14:46and estimate the parameters.
14:47The parameter estimates aren't perfect,
14:49but they're close and relatively they
14:51retain the the relationship of the
14:54ensembles and we have a 96% accuracy.
14:55If you actually go into the data
14:57where the algorithm makes a mistake,
14:59it actually.
15:00Has low confidence in that estimate.
15:02And so you could actually intuit
15:04from the covariance matrix that
15:05it might be making a mistake here.
15:07And so this then just gives us
15:09a graphical representation of
15:11that entire population.
15:12And you can see when we pull out the rasters,
15:13we see those five responses inhibited
15:16in aphasic and sustained way,
15:18excited in a sustained and phasic way,
15:20and nonresponsive.
15:21OK, so this is our simulation.
15:23Now we wanted to say,
15:24what can we learn by applying this
15:26sort of method to actual data?
15:28And so we went back to this
15:29social learning experiment that
15:30I told you about and said, OK,
15:31if we take the neurons from the ACC and
15:34the BLA and we cluster them together,
15:36are neurons distributed across
15:37ensembles or do they segregate?
15:40In other words,
15:41do neurons across different regions
15:44actually share firing rate properties?
15:49And so when we look at the
15:50neurons in the ACC and BLA,
15:52indeed what we find is a distributed
15:54representation where you have these various
15:56ensembles with different properties,
15:58but you have neurons from both
15:59the ACC and the BLA within them.
16:01Now this makes sense intuitively,
16:03although no one had shown this before.
16:05If we think about the fact that these new,
16:07these regions have reciprocal.
16:08Anatomical connections with each other,
16:10and they're both involved in learning.
16:12One might hypothesize that
16:13they would share ensembles,
16:15but we were able to actually
16:16demonstrate that that is the case here.
16:18Interestingly, though,
16:19there's a different sort of distribution of
16:22that representation between the regions,
16:24and so you can still see that there's
16:27some separation in sort of the Fe
16:29code between these two regions.
16:30So we wanted to go back into this idea
16:32of that there's certain neurons in the
16:34ACC that are actually tracking learning,
16:36and I showed you sort of
16:38individual neuron example of that.
16:40We hypothesize that if this is actually
16:43picking up real ensemble dynamics,
16:45it should be able to find some.
16:48Excuse me,
16:49It should be able to find some
16:51differences in learning that occur
16:52between habituation and conditioning.
16:54And as a control,
16:55we basically take all of the same data,
16:57we divide the sessions the same,
16:59but we give a false cue to the algorithm.
17:01So a cue that actually wasn't
17:02there and asked,
17:03is it going to still say that something
17:05happened or not as our control?
17:06And So what I'm showing you on the left here,
17:09Gray is the ensemble
17:11representation doing habituation,
17:12and in red is conditioning,
17:14and you can see that there's a shift.
17:16From this low phasicity space into a
17:18high phasicity space that occurs during
17:20learning this idea that neurons are
17:22kind of tightening their response to
17:24to the queue and you don't see that
17:27when you look at the control condition.
17:30And so if you just break this down
17:31into a uni dimensional analysis,
17:33you can see that there is a significant
17:35difference in the phasicity parameter
17:37for learning when you look at the actual
17:40queue versus the the control condition.
17:41And so we we can use this to
17:43start looking at the kinds of
17:45parameters that an ensemble might.
17:46Be using to represent learning
17:49as as animals are actually
17:51behaviorally demonstrating it.
17:52What's powerful is this is that we
17:54can then take these parameters and
17:56begin to form hypothesis about the
17:58kind of biological or bi physical
18:00properties that might be necessary
18:01to undergo this kind of change.
18:03And so we wanted to actually go into
18:05the data and see if that might be true.
18:07And so to look at that we're going
18:09to use a molecular biology approach
18:11here which is a a Krylox V system.
18:14To be able to express channel
18:17adoption within a specific circuit.
18:19And here we're looking at the
18:21anterior singlet neurons that project
18:24monosynaptically to the amygdala.
18:26And so we can express channel adoption here.
18:28And when we shine light,
18:29we see different populations.
18:30Neurons that are in this monosynaptic
18:32projection show very shortly
18:34and see responses to light,
18:36whereas neurons that are in
18:37this excited network and so they
18:39receive either reciprocal feedback.
18:40Or collaterals from these excitatory neurons,
18:43they show light responses that
18:44are a little bit longer latency,
18:46and then these inhibited network neurons are
18:48within the network but not directly involved,
18:51and they show inhibited
18:51responses to the light.
18:52OK, so we're able to record from neurons,
18:54identify where they are in the circuit,
18:56and the question is,
18:57will we see a difference in our
19:00ensemble representation when we are
19:02informing it about the anatomical
19:04properties of the neurons involved?
19:07And so sort of give you the
19:08graphical sense of what this is.
19:10This is the same population that
19:11we were looking at before where
19:13there's this low phasicity space
19:14that goes away during conditioning.
19:16But now we're also looking
19:17at the individual neurons,
19:19What network are they in?
19:20And So what is the distance that
19:22they travel as neurons are learning.
19:25This is another way of looking at it
19:26to just illustrate the point that when
19:28we know the network identity of the neurons,
19:30we're able to pick out significant
19:33differences in the parameters by
19:35which they're encoding learning.
19:37And here we provide evidence at the
19:40ensemble level to the hypothesis
19:42that people have had that I actually
19:44gabourgic signaling in the cortex
19:46and how that is shifting parameter
19:47responses that are enabling animals
19:49to learn during this sort of process.
19:52This ensemble data provides evidence
19:54for that,
19:54that sort of hypothesis.
19:56Could we see the greatest changes in
19:58phasicity actually happening within the
20:01inhibited network versus the others?
20:03And so a major motivation for developing
20:06this sort of approach was that it
20:08might be a very useful way to begin
20:11looking at the translational problem.
20:13And that problem is that,
20:15you know,
20:16we are able to look at behavioral
20:18systems across mics,
20:19humans and primates,
20:21but there's really differences in
20:22the ways in which they can behave
20:24and we're looking and searching.
20:25I'm collaborating with Steve China to
20:27really think about ways that one might
20:29develop social behavioral experiments
20:30that can go across a model systems.
20:32However,
20:33for humans and primates,
20:34we can actually design one to one
20:36human behavioral paradigms and
20:37I'll show you an example of that.
20:39We can get different kinds of
20:41neurophysiology data across these
20:42different model systems and that has led
20:45to different sorts of ways of analyzing.
20:47And we're hoping that this
20:48sort of states based approach,
20:49because it allows you to take
20:51neural observations and then derive
20:53actual state formulations for how
20:56population activity is changing,
20:57might provide a window to be able
20:59to ask across different model
21:01systems and different kinds of
21:03neuronal representations how animals
21:05representing social information and so.
21:07Towards this,
21:08I won't talk about the human
21:09side of this today,
21:10but I wanted to show you some of the work
21:12we're doing with primate
21:13data to apply this this tool.
21:15And so like I showed you,
21:16Steve Chang has developed this
21:18model where primates can look at
21:20each other or an object while he's
21:21recording from these different brain
21:23regions and importantly 2 of the same
21:25regions that we're interested in,
21:26the anti single cortex and the amygdala.
21:31He published this paper in Neuron where
21:33they're able to look at individual
21:34neuro responses and show a variety of
21:36ways in which neurons are responding to face.
21:38Eyes are object, but again,
21:40they have to really start to define a neuron
21:43as responding to face or object or eyes and.
21:47They're missing a lot of the nuances
21:48that are happening at the ensemble level,
21:50even though they're able to
21:51describe at the single unit level
21:53what's happening really well.
21:54And so we wanted to apply it to this data.
21:56And here's an example from the
21:58anterior singlet cortex where
21:59we see some things that again,
22:00intuitively make sense.
22:01There's some overlap between the
22:03face and eyes representation,
22:05although there's some places where
22:07we have segregated ensembles,
22:08And when we look at the object,
22:09we see that it's really sort of spatially
22:12distinct from the other representations.
22:16And again we can look at this at at
22:17the uni in the uni dimensional way
22:19and see that there's these trends to
22:21for these parameters to be different
22:23when animals are looking at social
22:25stimuli versus on non social stimuli.
22:27And we've now applied this to various
22:29regions including the dorsomedia,
22:31prefrontal cortex and the OFC.
22:33And again you begin to see that
22:35there might be different strategies
22:36where there's much more segregation
22:38in the representation between the
22:40ACC and the dorsomedio prefrontal
22:41cortex where there might be more
22:43overlap in the area like the amygdala
22:45or the orbital frontal cortex.
22:46And so we can kind of derive hypothesis
22:49about what this might mean and how
22:51these sorts of parameters across model
22:53systems are changing as animals are
22:55engaged in these sort of social behaviors.
22:57And so I've shown you some behavioral
22:59data in rodents and some behavioral
23:01data in non human primates where we're
23:04able to record from neurons as animals
23:06are engaged in social behavior and
23:08have now derived a new analytical
23:10approach that lets us know from these
23:13recordings what are the groups of ensembles.
23:15We can assign parameters to ensembles
23:17and then use that to begin better
23:19understanding how brains across model
23:21systems are are representing information.
23:24We think that this will be
23:26important for for translation.
23:27In the future.
23:29So I want to thank all of my mentors
23:32here in the department.
23:33This again has been really an
23:35amazing place to train.
23:36I feel really grateful to be able to
23:40to be here and to continue to think
23:43critically with everyone about how we
23:45can progress and evolve our community.
23:47And then my collaborator Steve Chang Demba,
23:51my PhD advisor,
23:53K Tai and.
23:56Many people have been able
23:57to work with me while
23:59I was here and really enable
24:01a lot of this work. So thank you all
24:04for listening and I appreciate it.
24:26I was wondering since you see most of
24:28the changes at least in these brain
24:30areas and phasicity and you have that
24:32baseline change between the excited
24:34and the inhibited in terms of jump.
24:37What what would you predict
24:39across other brain regions?
24:41Is phasicity the thing that
24:42would change most with learning?
24:44Is jump more stable?
24:45What would it take to change jump?
24:48Those are exactly the kinds of
24:51questions that we're asking.
24:52Phasicity seems like it might be
24:54a really interesting parameter
24:56when we think about how.
24:57Neurons might use things like
24:59oscillations to communicate information.
25:01And so in shifting phasicity,
25:03you might better enable neurons
25:04within a certain ensemble to be able
25:06to oscillate at a certain frequency
25:08or get inputs at a certain timing.
25:11Whereas jump might be more defined by
25:14sort of what kinds of ion channels you have,
25:17how you know great of
25:19and how many spikes can you
25:21generate. And so some of what we're
25:23going to do actually is to use
25:25optogenetics to actually entrain.
25:26Different circuits at specific
25:28frequencies and ask how does that change
25:30and shift the phasicity parameter?
25:32If we over expressed time adoption,
25:34how does that shift the jump parameter
25:35to really start getting at this question
25:38of how are these parameters directly
25:39related to biophysical properties
25:41of the ensembles, which I think
25:43is probably one of the most exciting
25:44ways this could be helpful.
25:52I was really struck in the earlier
25:54slide by the effect where it seemed
25:56that the intrasingulate response
25:58to the observed cue was much more
26:01sustained as compared with the BLA
26:03and I think the similar thing has
26:05been observed in predator threat.
26:07So I I just was curious your
26:09speculation as to why there is a sustained
26:11response in prefrontal cortical
26:12circuits to these aversive
26:14or socially aversive stimuli?
26:17Actually this question is part of
26:19what got me interested in this
26:20work like looking at in vivo
26:21data and seeing that there's some
26:22new other seem very sustained,
26:24some that seem very basic.
26:26It's like what is the different
26:28strategy for representation that
26:30And so one idea I think is that the
26:32more sustained firing is how the
26:34brain might represent state shifts.
26:36So now I'm in an aversive state or
26:38now I need to attend to the state
26:40of this other animal that might
26:42be better represented by something
26:44that has like a very sustained.
26:46Property, while a stimuli that comes
26:49on or off or that comes in and out of
26:51attention might better be represented
26:52by sort of more phasic property
26:55and that's something that we can
26:57like actually directly test. Thank
27:04you.
27:06Any other questions? No.
27:08OK, we will move on then.
27:11Thank you a CA for a great presentation.
27:19So our next tree is there.
27:22Jefferson is a wonderful rising
27:23third year resident Alex Pond is
27:25going to introduce her over Zoom.
27:29So I think Alex, if you're able to,
27:38yeah. Are we ready? It works.
27:45Yeah, we can see. I hear you.
27:47That's. I'm on my. I
27:52just have to end this show.
27:58We'll do it like this for now.
28:01OK. You're good. Alex. Go ahead.
28:05OK. Yeah. It is a pleasure
28:07to introduce Sarah Jefferson.
28:10Sarah received her MD,
28:12PhD from Penn State.
28:14She came to Yale.
28:15About three years ago to
28:16start her residency with NRTP.
28:19Yeah, I remember when we recruited
28:21her, everyone was very thrilled
28:22because of her PhD work,
28:24which involved a series of very
28:26elegant study on how GABA ergic
28:28in the neurons is involved with
28:30depression and antidepressant actions.
28:33And there's obviously a long history
28:34on this research topic at Yale,
28:36starting from work with John
28:38Crystal and Peter Muharram And then.
28:40Later on on Ron Duman and then
28:42more recently myself as well.
28:43So it's very exciting to see that
28:45Sarah can carry this torch and then
28:48also move it to new directions.
28:50Sarah's current research focused
28:51on the potential therapeutic
28:53effects of psychedelics.
28:54So today, as you see in the title,
28:57she'll talk about her work on one
28:59of the more classic but also lesser
29:02study compound called 5 Methoxy DMT.
29:05And I think there's a lot of
29:06opportunity with this area, right.
29:07And Sarah will tell you more about it.
29:09But this compound has several
29:11very fascinating properties.
29:12Its effect is very short lasting in humans.
29:15It only lasts for on the order
29:17of 10s of minutes.
29:18And then the subjective experience
29:20also very intense and nonvisual,
29:22very unlike other psychedelics
29:24like psilocybin and LSC.
29:26Currently there are a number of phase two
29:28clinical trials to study this compound,
29:30but there's really not a whole lot
29:32known about it in neurobiology.
29:34So what Sarah will tell you,
29:36I believe is actually one of the
29:38first more rigorous published study
29:39that actually looks very closely
29:41now at what this compound does in
29:43terms of its neural and behavioral
29:44effects in an animal model.
29:48So beyond, you know, just the work itself,
29:49I want to mention that, you know,
29:52Sarah started this project at
29:53a difficult time.
29:54I mean,
29:55my lab was just about to move and she
29:56had to complete all of the things we
29:58should tell you actually within a year.
30:00So this really shows you.
30:03How a sense in terms of his her
30:06ability to just lean and execute.
30:08She had to learn new methods in terms
30:10of using two photon microscopy and then
30:13also do all the experiment analysis.
30:15She's also extremely innovative
30:17now working with Al K,
30:20Chris Pinger and Marina Paciotto.
30:22She's now trying to combine these
30:24microscopy methods with molecular methods.
30:26To continue studying 5 MU DMT
30:28as well as other psychedelics.
30:30So I really look forward to,
30:32you know, seeing what should achieve
30:34in the coming years.
30:35So with that,
30:36yeah,
30:37Please join me to congratulate
30:38and welcome Sarah Jefferson.
30:40All
30:49right. Thank you so much, Alex.
30:51When I get this up, here we go. OK.
30:58So you know, first of all, just I,
31:00I really want to thank the Lessman family,
31:01the Lessman Award Committee.
31:02It's such an honor to be here and be able
31:05to share this work with all of you today.
31:07And it's also an honor to follow
31:09such an incredible neuroscience
31:11neuroscientist and person as a ZA.
31:14So I'm really grateful for this opportunity.
31:17So as Alex mentioned,
31:18I'm going to be discussing a study that
31:21was completed in his lab and I'm now
31:24carrying for this work in Al K's group.
31:28And it's focused on the short acting
31:30psychedelic called 5 Methoxy DMT and
31:33its effects on any behaviors and on
31:35structural plasticity and mouse models.
31:41In terms of disclosures,
31:42I do have an SRA with Freedom
31:45Biosciences looking at this drug.
31:48So as many people in this room know,
31:51psychedelics have really been the the
31:53focus of this resurgence and interest
31:55in their potential uses as therapeutics
31:57for range of mental health disorders
32:00ranging from mood disorders to PTSD,
32:02substance use disorders and beyond.
32:05And I think you know thorough review of
32:06that topic is beyond our scope today,
32:08but I wanted to point out.
32:10That two of these drugs have now
32:12reached phase three clinical trials and
32:15those are MDMA for PTSD and psilocybin
32:17for treatment resistant depression,
32:19and these are both used in
32:22combination with psychotherapy.
32:24And my interests have,
32:26as Alex said,
32:27been in understanding the neurobiology
32:30of depression in particular and in
32:32the development of novel therapeutics
32:33for treating a resistant depression.
32:35So I'm going to really focus on on
32:38this category called the tryptamine
32:40structural class of psychedelics,
32:42of which psilocybin is probably
32:44the best known member.
32:46But I'm going to be talking more about
32:49this less studied but very interesting
32:52compound called fibrothoxy DMT.
32:54So you know what is the evidence from
32:56the clinical studies look like for the
32:58use of psychedelics in the treatment
33:00of mood disorders in particular.
33:01So there have now been a number
33:03of clinical trials through phase
33:05two and into phase three with
33:07psilocybin assisted psychotherapy
33:09for treatment resistant depression.
33:11And some of the more remarkable
33:13aspects of these these drugs are
33:15their ability to induce a very
33:17rapid antidepressant effect,
33:19so one day after a single dose or two doses.
33:23As well as their enduring benefits and
33:26these vary from study to study a bit,
33:29but the earlier studies showed
33:31antidepressant effects of psilocybin out
33:33to three months following a dosing session.
33:36So that's pretty remarkable.
33:37And even in the phase two COMPASS
33:39trial that I'm showing here,
33:41we do see that the effects continue
33:43to be quite enduring even with
33:45these larger studies.
33:46The drawback with the use of psilocybin
33:49in a clinical setting is that these
33:51dosing sessions take at least six hours.
33:53You know,
33:54they're very labor intensive.
33:55A trained therapist needs to be
33:57with the person the entire time.
33:59So if we're thinking about ways
34:00that we could potentially translate
34:02the use of these to the clinic,
34:03the way in terms of scaling up and
34:07improving access for more patients,
34:09I think looking at a shorter acting
34:12compound does potentially very interesting.
34:15So with that in mind,
34:16I began this work on this short
34:19acting seritinergic psychedelic
34:21called 5 Methoxy DMD.
34:22If you've heard of this at all,
34:23you've probably seen the association
34:25with the Colorado River toad
34:28that I'm showing here.
34:29This produces this compound
34:31and its parotid glands and now
34:34it's mostly synthesized
34:38synthetically so. So you know,
34:40we we don't get it from the toads anymore.
34:43And what's? Some of the pharmacokinetics
34:45are really what make this remarkable.
34:48So this drug is typically used either through
34:52inhalation or intranasal insufflation.
34:54So the onset of action is very rapid and
34:57the duration of the effects are very short,
35:00usually resolving in about 20 minutes.
35:02So this makes this really appealing
35:04if we're thinking about potentially
35:05getting this drug to more patients
35:07and being able to really effectively
35:10implement this in the clinical setting.
35:12And I just wanted to point out that in
35:15terms of the pharmacology of this drug,
35:18it targets serotonin 2A receptors
35:20just like all of these classical
35:23serotonergic psychedelics do.
35:24The 2A receptors are thought to be
35:26responsible for the psychedelic effects.
35:28There's still, I think,
35:29an ongoing debate about their
35:32necessity for therapeutic effects.
35:34And this drug also targets serotonin 1A
35:37receptors with a pretty high affinity,
35:38which makes it a little unique
35:41compared to psilocybin.
35:43So looking at clinical studies of
35:45five methoxy DMT as Alex said,
35:48sorry these are are pretty early.
35:51We have a few observational studies
35:53that do suggest that a single dose
35:56of this drug can produce relief of
35:58depression and anxiety symptoms and
36:01potentially a long lasting way at
36:02least from the data that we have.
36:04And the day that I'm showing here
36:06is from an observational study in
36:08a naturalistic setting where people
36:10use an inhaled form of five methoxy
36:12DMT and then they self reported
36:14symptoms of depression,
36:15anxiety and stress at various time points.
36:18And they did see improvement in all of
36:21these measures at 30 days following the drug.
36:24As Alex also mentioned,
36:25there have been some phase two
36:27clinical trials.
36:28There's one that's been completed.
36:29So far we don't have the full data set,
36:32but based on the press release.
36:34They had some promising results and
36:36that was published from GH Research.
36:39It's a study of eight patients with
36:42treatment resistant depression.
36:43And the data that we have so far showed
36:46that seven out of eight patients
36:48achieved remission in their depressive
36:50symptoms at seven days following an
36:52escalating dose regimen of this drug.
36:55So they basically gave them enough to
36:57achieve a strong psychedelic effect and
37:00then looked at their depression symptoms.
37:02So I'm really interested in
37:04understanding on a mechanistic level,
37:06how a single dose of a psychedelic
37:08drug can produce these very long
37:10lasting improvements and symptoms.
37:12And one potential explanation for this is
37:15through enhancement of neuroplasticity.
37:17So neuroplasticity is generally an
37:19increase in the number or the strength
37:22of synaptic connections between neurons.
37:24Deficits in neuroplasticity have been
37:26noted in patients with depression.
37:28And furthermore,
37:29we know that antidepressants can
37:31enhance neural plasticity on a
37:33time scale that seems to correlate
37:35with their therapeutic effects.
37:37So if we're looking at measures of sort
37:39of antidepressant efficacy that can
37:41translate across from humans to rodents,
37:43this is a useful one to look at.
37:46We know that chronic use of
37:48fluoxetine that can increase
37:49the density of these dendrotic spines,
37:51which is are these protrusions where most of
37:55synaptic connections form on the dendrite.
37:58And antidepressant doses of ketamine acutely
38:01increase the density of these spines.
38:03So again, this correlates with the time
38:05scale of their therapeutic effects.
38:07So what do we know about the effects of
38:10psychedelics on neuroplasticity so far?
38:12So this is taken from a really
38:14nice paper that was published in
38:17Neuron from Alex's group.
38:18It was led by Link Shaw Show.
38:21And they looked at changes in the density
38:24of dendritic spines over a prolonged time
38:26period after a single dose of psilocybin.
38:29They're focused on the mouse
38:31medial frontal cortex,
38:32which is an area that's been shown to
38:34be modulated by antidepressants before.
38:37And if you focus on the the red graph here,
38:40you can see that after a single dose,
38:42an injection of psilocybin,
38:44they had noticed an increase in
38:46dendritic spine density at one day.
38:48And this persisted very long term,
38:50over a month after that fall,
38:52that single injection.
38:53So that was pretty remarkable.
38:56So the question that we wanted
38:57to answer in my study was whether
38:59this psychedelic was very short
39:01acting psychedelic effects could
39:03similarly alter neural plasticity.
39:05If so,
39:06what are the time scales
39:07of those effects and.
39:11The first thing that we needed to
39:13do in this study was to evaluate,
39:15you know, what is a psychedelic
39:17dose of this drug in a mouse.
39:19You know, there's sort of limited
39:20literature on this drug prior to
39:22this and obviously we can't ask
39:25the mouse if it's experiencing
39:27mystical type experiences or oceanic
39:30boundlessness as we do in people.
39:31So we focus on this particular
39:33behavior that we can measure
39:35called the head twitch response.
39:37And as you can see here in this video,
39:39it's this rapid side to side
39:41motion of the head.
39:42That has typically been used as a marker
39:45of psychedelic effects in rodent studies.
39:48Obviously, this doesn't really have peace
39:50validity for psychedelic effects in humans.
39:52This isn't what psychedelic
39:53effects look like in humans,
39:55but has some predictability.
39:56We know that drugs that target
39:59the serotonin to a receptor that
40:01have psychedelic effects in humans
40:03induced this behavioral response,
40:05whereas drugs that target those same
40:07receptors that lack psychedelic effects
40:09in humans do not produce this behavioral.
40:11Change.
40:14So the first part of the study was
40:17measuring the head twitch response
40:18with a range of doses of five
40:20methoxy DMT and we compared to
40:22psilocybin as a positive control.
40:24Psilocybin's been better characterized
40:27in this assay before and we're actually
40:30able to do this using an automated
40:32magnetic ear tag based technique.
40:36This was initially developed
40:38in Javier Gonzalez Myzo's lab,
40:39and it was actually kind of built
40:42from scratch by Mark Dibbs,
40:43one of the students who was rotating
40:45in the lab at the time.
40:47This takes advantage of the fact that
40:49the movement of a magnet within a
40:51copper coil can generate A voltage that
40:53can then be automatically detected,
40:55and we essentially just place the
40:58magnetic ear tag on the mouse.
41:00We put it into a container that's
41:02lined with a copper coil and we get.
41:04An output on the number of head
41:05twitches which is allows us to
41:07really ramp these experiments up and
41:09look at an extended period of time.
41:11So first we just validated this
41:13approach with a range of doses of
41:15five methoxy DMT and with psilocybin
41:17and compared to hands board videos
41:19and showed that this automated
41:21system is working very well.
41:23Those two measures are highly correlated and
41:26then if we get into the actual data here,
41:28so this is our time course over 60
41:31minutes of the head twitch response.
41:33And you can see the doses of five methoxy
41:36DMT in blue with psilocybin in red.
41:39And the two things that I think are
41:41important to take away from this
41:42are number 15 Methoxy DMT induces
41:44a head twitch response that is
41:46consistently brief regardless
41:47of the dose that we're testing,
41:50usually results within about 10 minutes.
41:53And that's compared to psilocybin that has
41:56this more long kind of protracted response.
41:59Interestingly,
41:59increasing doses of the drug produce
42:02increasing amounts of head twitch.
42:03This seems to indicate that this behavior
42:06could correlate also with the intensity
42:08of the psychedelic effects as well.
42:11And for our further studies,
42:13we wanted to choose a dose of
42:14five Methoxy DMT that was roughly
42:16equivalent to psilocybin in terms of
42:18the number of head twitches induced.
42:21So what's kind of a?
42:22A similar intensity psychedelic dose
42:25of this drug and we ended up choosing
42:27the 20 mig per kick dose based on
42:30the total number of hedgewatches.
42:35So there's kind of limited behavioral
42:39data that characterizes psychedelics
42:42in behaviors outside of the hedgewitch.
42:45And I think it's kind of important to
42:47study a wider range of behaviors because,
42:49you know, these can give us
42:51insights into the mechanisms
42:52underlying these psychedelic drugs.
42:54And I think it's also important as people
42:56are thinking about screening novel compounds,
42:58novel psychedelics potentially.
43:00Now people are interested in drugs
43:02that have sort of psychoplastogen
43:05effects without psychedelic effects.
43:07So I think having a battery of behavioral
43:09assays that help us understand how these
43:11drugs work is it's important and so.
43:14Along those lines,
43:16a post back in the lab who's
43:18now gone on to grad school,
43:19Ian Gregg conducted this study of
43:23social ultrasonic vocalizations.
43:25And this is a social behavior that's
43:28produced during mating by males
43:30when they're exposed to a female.
43:32The females produce these
43:33ultrasonic vocalizations,
43:34but to a much lesser extent.
43:37And he wanted to look at the changes in
43:40social USV's with classical psychedelics,
43:42so psilocybin and five methoxy DMT,
43:45as well as with ketamine,
43:47which has kind of similar effects
43:48on neuroplasticity.
43:49But I would expect the sort of acute
43:52psychoactive effects to be quite
43:53different from those other two drugs.
43:55And he essentially measures this over
43:583 prerecording sessions and then a
44:01session that's immediately after the drug.
44:04And interestingly,
44:05what what was found was that all of
44:07these drugs suppressed social US fees,
44:09but to very different degrees and in
44:13particular 5 methoxy DMT almost entirely
44:16suppressed ultrasonic vocalizations
44:18produced during meeting behavior.
44:20Now thinking about the different
44:21interpretations of this,
44:22I think there are there are multiple options,
44:26but one possibility that we're toying with
44:28is that this is really representative of
44:31the intensity of the psychedelic effect.
44:34Which would track from what
44:35we know from humans,
44:37which is that even though this is
44:38a short acting psychedelic drug,
44:40the quality of the the psychedelic
44:42experience is very intense and
44:44indifferent from psilocybin.
44:47And I'm not going to go into this in detail,
44:49but essentially you can characterize
44:52different forms of ultrasonic vocalizations,
44:55different patterns and we found
44:58that these different.
45:00Novel antidepressants can modulate the
45:03pattern of ultrasonic vocalizations
45:05as well in particular ways.
45:07OK,
45:08so the biggest question we wanted
45:09to answer in the study was related
45:12to changes in neural plasticity.
45:13So the way we're able to do this is
45:16really this elegant approach that
45:19utilizes in vivo longitudinal 2 photon
45:22imaging of dendritic spines over time.
45:25And the way we do this is we
45:27insert a cranial window over an
45:30area of the medial frontal cortex,
45:32CG1M2,
45:32which again is that same region that
45:35they investigated with psilocybin
45:36that's been shown to be modulated
45:39with antidepressants.
45:40And we do this in a thy 1G FP
45:43mouse where there's a fluorescent
45:45reporter in layer 5 framinal neurons.
45:48These neurons are important in receiving
45:50inputs across different cortical layers.
45:52They're major output neuron from the cortex.
45:55That go to deeper brain
45:56structures and they they project
45:57to the other parts of the cortex as well.
45:59So these are very important neurons and they
46:02also highly express serotonin 2A receptors.
46:05So we think that it makes sense that
46:07these could be like a direct or
46:09immediate target of psychedelic effects.
46:11So we're actually able to image their
46:14apical dendrites in layer one as
46:16shown here and this is our timing of.
46:20Time scale of our experiments.
46:22So we take two baseline imaging
46:24sessions prior to drug delivery,
46:26which allows us to look at the
46:28spine dynamics prior to drug.
46:30These spines are dynamically being
46:31formed and eliminated all of the time.
46:34So it's important to sort of get
46:35a sense of the baseline there.
46:37And then we are able to image
46:39those exact same dendrites,
46:40the exact same spines every
46:42two days for seven days.
46:44And then we take a longer time point
46:46at over a month and I think we could
46:49actually carry these experiments out.
46:50At longer time points as well,
46:52something I'm curious about doing in
46:55the future and this is just an example
46:57of the beautiful images that we got.
47:01So we did this in drugs treated with
47:04five methoxy DMT versus vehicle and
47:06we found surprisingly that treatment
47:08with five methoxy DMT is again a
47:11single dose increases dendritic spine
47:14density at one day following injection.
47:17And similar to psilocybin,
47:18this response really persists for over a
47:21month following that single injection.
47:22So even though there was acute effects of
47:24the drug are wearing off within 2030 minutes,
47:27we're getting this really sustained
47:30enhancement of neuroplasticity which
47:32is pretty interesting and remarkable.
47:34The other measure of synaptic strength
47:36that we can look at is the size of
47:39spine heads or the spine head with.
47:41This is a measure that is enhanced
47:43by psilocybin.
47:44We found that this wasn't altered
47:45with five methoxy DMT,
47:46which is also kind of interesting
47:48that there's potentially these
47:50divergent effects on the way,
47:52or at least different effects on the way
47:54that these drugs affect neuroplasticity.
47:58Finally, spine density is a factor of the
48:00rate of formation of new spines and the
48:03rate of elimination of existing spines.
48:05So we wanted to understand what was
48:07underlying this change in spine density.
48:10And we found that with five methoxy DMT,
48:13the formation rate of new
48:15spines was increased at one and
48:17three days post drug injection,
48:19whereas the elimination rate was unchanged.
48:22So really what's happening with
48:23the drug is that we're getting
48:24a lot of new spines formed.
48:25We have an increase of about 15%,
48:28which is pretty significant
48:30and they seem to be persisting
48:33longterm and not being eliminated.
48:37So a summary of what I've shown so far
48:40is that this short acting psychedelic
48:435 Methoxy DMT elicits a very brief
48:45head twitch response which mirrors
48:47its psychedelic effects in humans.
48:49In terms of the duration,
48:515 Methoxy DMT substantially suppresses
48:54social USB's which potentially could
48:57could represent the intensity of
49:00this acute psychedelic effects.
49:02And a single dose of the short
49:04acting drug can produce long lasting
49:07enhancements in neuroplasticity.
49:08And just to touch on a couple of
49:10the things that we're actively
49:12working on right now.
49:13So as Alex said, we're,
49:15I'm pretty interested in combining sort
49:18of systems neuroscience tools with more
49:22molecular neuroscience tools as well.
49:25My background is more in molecular
49:27and behavioral neuroscience.
49:29And the questions I'm interested
49:31in addressing are related to #1.
49:33What's responsible for the acute
49:35effects of this drug on plasticity?
49:37And maybe more importantly,
49:39why are there such enduring effects
49:42of this drug on neuroplasticity?
49:44So trying to understand, you know,
49:45what is it about these centritic spines
49:48that's causing them to last so long term?
49:51And if people have questions,
49:52I can talk more about the the
49:54specifics of these, you know, finally.
49:56What are the behavioral consequences
49:58of these changes?
49:59I think it's overly simplistic to
50:01say that plasticity is all good and
50:03all therapeutic in every region.
50:04So really understanding, you know,
50:06#1,
50:07is this change specific to this brain region?
50:09And if so,
50:12what are the behavioral consequences?
50:15If it's happening brain wide
50:16that would also be interesting.
50:18I think there's there's a lot
50:20of open questions there.
50:21And then finally,
50:23we're also interested in looking at other
50:25classes of psychedelics and looking
50:27at their effects on neuroplasticity.
50:29I will mention that we've now
50:32completed a study with MDMA looking
50:34at neuroplasticity over time and
50:36we've actually found that MDMA looks.
50:41Pretty interesting as well.
50:43It induces very robust increases in
50:45neuroplasticity in the short term,
50:46but the dynamics of the changes
50:49over time are different from the
50:52serotonergic psychedelics Okay.
50:55And just a thank you again to
50:57the Westman Award Committee to my
51:00plethora of amazing mentors that
51:02I've had here at Yale.
51:04Alex, Al Marina,
51:06Chris.
51:07And I'm only mentioning people in
51:09the Quan lab who directly contributed
51:11to this study because their group
51:13is ever expanding and then as well
51:16to to the K lab in particular,
51:18Patrick is a post grad working with
51:21me who really did the the majority
51:24of the MDMA spine work and we're
51:26going to be sad to to lose him
51:29to grad school next year.
51:30So I'm happy to take any questions.
51:32Thank you.
51:48glad that nobody has to
51:49lick the toads to do this.
51:51this. is great, great, talk.
51:54I'm wondering if this finding that you
51:57have because it was so substantial,
51:59the increase in spine creation
52:01on that 24 hour time point.
52:04Whether there are PET imaging possibilities
52:06in humans that could be parallel.
52:08I know the SV2 PET leg and mainly looks at
52:11presynaptic markers of synapse structure.
52:14Is there something that you can
52:17look at that's more presynaptic just
52:19to parallel that postsynaptic that
52:21might actually translate to a human
52:23imaging study pretty quickly, Yeah,
52:27that that translational aspect is is
52:29very interesting to me as well, yeah.
52:32I think that you know certainly with
52:36kind of more classical immunohisto
52:38chemistry looking at pre synaptic
52:40markers that could be like a kind of
52:43quicker way of getting at that question.
52:46You know we could just take for example
52:49a shorter time point and just see if
52:51the changes in pre synaptic marker is
52:53parallel the the increases in the.
52:57Starting point, I know a lot
52:59of people are looking at SV2A
53:01with the psychedelics as well.
53:03So, so I think yeah,
53:06I think that that's definitely a
53:08nice pairing that we could do with
53:11wonderful studies and just to follow
53:13up on that are people who are taking.
53:175 Methoxy DMT that the drug
53:20do they get very quiet?
53:21I mean do they they have a very
53:24internal psychedelic experience
53:25that would parallel the the loss of
53:28of sub vocalizations of subsonic,
53:31ultrasonic Yeah that's that's
53:33an interesting question.
53:34I I think that from from what
53:37I've read about this,
53:39it seems that they do have a
53:41pretty internal experience.
53:42It's.
53:43You know,
53:44almost to the point of like ego
53:46dissolution for some it's it's a
53:48very intense psychedelic experience.
53:49So I would imagine that they
53:52they probably are not really
53:54interacting with their environment
53:55in a substantial way.
53:57Yeah,
54:00I had a brief question
54:06I wanted to ask about the plot
54:08with the different doses of
54:10five MEODMT and the psilocybin.
54:13So just looking at that pot,
54:14I I wouldn't necessarily pull
54:16out the differences that people
54:18talk about with DMT lasting much,
54:20much shorter than psilocybin
54:23because all the curves seem to
54:24come back to baseline in about
54:26an hour or a little less.
54:29What do you think are the differences there?
54:33Like if is it,
54:35is it a dosing difference between
54:36what people usually take and
54:37what's been giving given to the
54:39mice or is it something else?
54:43Yeah. So, well I think that the
54:47dose is definitely an important
54:48point to to touch on here.
54:50You know I I see,
54:52I think the other point you're
54:53making is about the the duration.
54:54Does the duration with this assay
54:57really fully reflects the duration of
54:59the human psychedelic effect And and I
55:02think that that that's a fair point.
55:04I think it's not a one to one.
55:06You know certainly in humans
55:07the the effects of psilocybin
55:09are going lasting for hours.
55:11You know in terms of the dosing,
55:13I think that's a really relevant
55:15question to address because the the
55:18clinical studies with five methoxy
55:20DMT that we have used doses of like
55:246/12/18 milligrams and typically
55:27whereas the doses that we're using
55:29here like the 20 mig per kig dose
55:32would roughly be equivalent to
55:33like 90 milligrams in a human.
55:35So this is a really quite a high
55:37dose if you look at it that way.
55:39So I think that that is also a
55:41pretty important question going
55:42forward is can we achieve similar
55:45effects with doses that are more
55:47relevant to the human studies?
55:48Great.
55:49Thank you.
55:49Thank
56:07you, Sarah.
56:12Oh, perfect.
56:12Oh, that would be what are we
56:16looking for here for a video?
56:19OK, it's not in the PowerPoint.
56:21It's a separate form.
56:23I think it is our next speaker.
56:26It's right here.
56:27Our next speaker right here, Yep,
56:28our next speaker is Jay Lee,
56:30who was the 2nd year resident and
56:33he has done some really innovative
56:35work in Uganda bringing some
56:37work there to help alleviate
56:39stigma and psychiatric treatment.
56:41We will be introduced by Dr.
56:43Robert Rosenpeck.
56:52So I should mention for the historical
56:55record that 50 years ago Jeff Lesman,
56:58the son of Seymour Lesman and I.
57:01Began our careers at Yale,
57:04walking down the Carters of the VA
57:06to do what every resident must do.
57:09Get fingerprinted.
57:13So, Jeff, if you're out there,
57:15here's the longevity.
57:19Jay Lee is a remarkable person.
57:22The ordinary.
57:22So we we shift now to global mental health.
57:26Jay Lee was born in Korea.
57:30Came to the United States,
57:31was in a military family and traveled,
57:33lived in many places around the country.
57:37And what he developed as a kind of a
57:41modus Vivendi was a way of fitting into
57:44places where he shouldn't have fit in.
57:47And it wasn't enough to do this in
57:50the US and in the American South.
57:53He had to go further.
57:54And so he worked in China.
57:56Now he didn't go to China
57:57as a tourist.
57:58This is in college.
58:00Medical school Over the
58:01years he went and lived
58:03there and was involved.
58:06China was mild,
58:07so he was looking for a place where
58:09he could have a new experience
58:11and he chose Africa,
58:13spent some time in Ghana,
58:15then ended up in Uganda.
58:17He's an incredible networker,
58:19connected with people and created Empower
58:23Health in a rural area in Uganda,
58:27now most researchers.
58:29Service researchers find a
58:31health system and study it.
58:33Jay, with his incredible talents,
58:35created a health system and
58:38now is beginning to study it.
58:41He started a program in which
58:44thirty students from both
58:47Uganda and the US are paying to
58:50go spend their time with him.
58:54They are doing research and
58:56learning clinical work.
58:57Many of these are undergraduates
58:59and graduates.
59:00So he's
59:01a social entrepreneur of incredible
59:03talent and this is the beginning of what
59:08will be a research career which will
59:11be layered on top of that. Thank you.
59:18Greetings from Uganda.
59:18I apologize that I cannot join in person,
59:21but it's such an honor to be
59:23recognized for this award.
59:24My name is Jay, I'm a second year
59:26resident in the program and today
59:28I'm going to be talking about a
59:30novel intervention that we conducted
59:31to destigmatize mental illness in a
59:34rural area of a low income country.
59:37So I wanted to give you a little
59:39bit of context of this work.
59:41So I've worked in rural Uganda on
59:44various health related interventions,
59:46specifically the Basuga region in the
59:48eastern region of Uganda since 2015.
59:52And in 2018, you know,
59:54I started A51C3 organization
59:56called Empowered Through Help.
59:58And currently, we provide healthcare,
60:00General Healthcare to a catchment
01:00:02area of 70,000 people,
01:00:03mental healthcare to a catchment
01:00:05area of 400,000 people and we
01:00:08conduct experiential fellowships
01:00:09for pre doctoral Ugandan and
01:00:11American students this summer.
01:00:13We have approximately 55 students in
01:00:16total and they work collaboratively
01:00:19on global mental health projects.
01:00:21And in 2021, we started providing
01:00:24mental healthcare you know,
01:00:26to people out of our Health Center
01:00:29and this is a very rural area.
01:00:30You know, there was no,
01:00:32you know,
01:00:33evidence based on mental healthcare
01:00:35system available in this area.
01:00:36So we realized that you know in order
01:00:39to increase mental healthcare seeking,
01:00:41we need to do some educational programs
01:00:45and some sensitization programs to to
01:00:48let people know that you know we have.
01:00:50Medications that that can work,
01:00:51that can help for things like
01:00:54psychosis and mania.
01:00:56So I looked at,
01:00:57you know what other interventions
01:00:59have been done to encourage
01:01:01healthcare seeking for mental
01:01:02illnesses in low income countries.
01:01:04And frankly I didn't find very much
01:01:07but the but the condition that has
01:01:09had better funding and global health
01:01:11that's got many parallels in my opinion
01:01:13to chronic mental illnesses like
01:01:15schizophrenia and bipolar disorder is HIV,
01:01:18AIDS.
01:01:19They're similar in that they're both chronic,
01:01:21they're both heavily stigmatized.
01:01:23So I looked at what interventions
01:01:25have been conducted before for HIV,
01:01:27AIDS and there was some really
01:01:29good evidence for creative based
01:01:31interventions and among these
01:01:32the other had the most evidence.
01:01:34So we subsequently developed and
01:01:36conducted the first evaluation
01:01:38of a creative based intervention
01:01:40to reduce mental illness stigma
01:01:42in a low income country aimed at
01:01:44the general population.
01:01:48So here are the methods that we utilize.
01:01:50So the first step was to find the
01:01:52current beliefs and attitudes for
01:01:54severe mental illness in the population
01:01:56and utilizing that information we
01:01:58developed the criteria for intervention
01:02:01which is the other intervention.
01:02:03And then we evaluated the effectiveness
01:02:05of the intervention using a
01:02:08single arm pre post study design.
01:02:12So to start off with,
01:02:13we conducted 4 focus groups with
01:02:15community members from from the area
01:02:17that we are that we are working and we
01:02:20utilize this information to develop
01:02:22a criteria for the intervention.
01:02:23So our aim was to not. Change beliefs,
01:02:26but to work within existing beliefs,
01:02:28to add the belief that medications
01:02:31can be helpful for conditions such
01:02:33as like such as psychosis, Amania,
01:02:37and to incorporate those new beliefs
01:02:40within the existing belief and
01:02:43then the theatrical intervention.
01:02:44So from the focus groups we developed
01:02:48the criteria of you know what we want
01:02:50to see in the seen that the others get.
01:02:52So outside of that was up
01:02:54to the participants,
01:02:55up to the audience or up to
01:02:58the community to come up with,
01:03:00you know,
01:03:01useful the other skits that they found
01:03:03was entertaining and that they found
01:03:06was that that they found was helpful.
01:03:08So we had a competition for
01:03:10the best the other skit.
01:03:12So four different teams competed
01:03:15and and the winner was a group that
01:03:18portrayed A relatable story of an
01:03:21individual with mental illness.
01:03:22So if he had what looks like psychosis,
01:03:25so at first he took him to
01:03:27traditional healer, He was still sick.
01:03:29After that they took him to
01:03:31religious leader for prayers.
01:03:32He was still sick after that.
01:03:34And then he got some medication
01:03:35and then he got better and he got
01:03:37functional enough so that he could hold
01:03:40down the job as a motorcycle taxi driver,
01:03:42which is a wellregarded profession
01:03:44in the village setting.
01:03:48And to evaluate the
01:03:50effectiveness of this program,
01:03:51we utilize the 61 item questionnaire as
01:03:55measuring demographics as well as stigma.
01:03:58So the Sigma questionnaire in
01:04:00particular was taken by from a
01:04:02study that Doctor Rosenhack and Dr.
01:04:03Iannacho conducted in Nigeria
01:04:05a few years ago.
01:04:07And we also divided that
01:04:09up into two categories,
01:04:11broad acceptance scale and the
01:04:13Personal Acceptance Scale.
01:04:14And so in general,
01:04:16broad acceptance scale reflected
01:04:18structural stigma and personal acceptance
01:04:21scale reflected public public stigma.
01:04:24So this is kind of like the
01:04:25flow chart of the evaluation.
01:04:26So initially we selected 101
01:04:29participants randomly to get the
01:04:31initial questionnaire and of the 100,
01:04:33one 77 watch the intervention.
01:04:36And 57 of the 77 were administered
01:04:39A questionnaire one week later and
01:04:4246 of the 57 were administered A
01:04:45questionnaire one year later to
01:04:47evaluate the long term effect.
01:04:49So the results, so as you can see here,
01:04:53the broad acceptance scale,
01:04:55so people begin accepting those
01:04:57with mental illness.
01:04:58We're having more accepting beliefs
01:05:01towards those with mental illness
01:05:03quite significantly as you can see.
01:05:05And also for the Personal
01:05:07Acceptance Scale too.
01:05:08And these are some effect sizes.
01:05:11As you can see,
01:05:12the coins D reflects almost like
01:05:13a difference in one standard
01:05:15deviation for both broad and
01:05:17Personal Acceptance Scale.
01:05:18And it also shows that that that change
01:05:21has persisted not over like one week,
01:05:23but over the course of one year as well.
01:05:30So some significant findings of this
01:05:31study is that this was the first
01:05:33study that evaluated effectiveness
01:05:35of a creative based intervention
01:05:36for reducing mental illness stigma
01:05:38in low income countries and it was
01:05:43very notable the persistent and
01:05:45significant and large effect sizes
01:05:48throughout the personal and broad
01:05:50acceptance scale at both time points.
01:05:52And you know, it's notable that this
01:05:54was much greater than most comfortable
01:05:57interventions that have been conducted
01:05:58in high income countries that showed
01:06:00more of a modest effect size.
01:06:02And a possible explanation could
01:06:04be that in high income countries
01:06:06there's just so much media to consume,
01:06:08whereas in the area that we work,
01:06:10there's one TV for the village that's
01:06:13a communal TV that people utilize to
01:06:15watch Uganda play soccer or you know,
01:06:18have seen major presidential addresses.
01:06:22But there's not a lot of media that's
01:06:25available which could help explain
01:06:27the a very large effect size as well.
01:06:30The strength of the study was
01:06:32this community based approach.
01:06:33So you know,
01:06:34we didn't design the intervention ourselves.
01:06:36We just gave a criteria of what the
01:06:39intervention should include and it was
01:06:40up to the community to develop something
01:06:42that was relatable to the audience.
01:06:44And so you know that was certainly a
01:06:47strength and could have like also contributed
01:06:49to the strong effect size as well.
01:06:51And a weakness of the study was
01:06:53that there is no control group
01:06:55just by the study design.
01:06:56This is a single arm pre post design,
01:06:59so social desirability bias could
01:07:02certainly factor in to factor in
01:07:05to biasing the results here.
01:07:08And another possibility is that
01:07:09attrition could be nonrandom,
01:07:11although there is no statistically
01:07:13significant differences in demographics
01:07:14as well as initial stigma level
01:07:17between the groups.
01:07:20And here are some references
01:07:21and acknowledgments.
01:07:25Appreciation goes out to the entire
01:07:26Empowered Through Health team,
01:07:28as well as from the
01:07:29community Health workers,
01:07:30my many mentors and friends.
01:07:33Thank you. So
01:07:42Jay is in Uganda at this moment.
01:07:44I think he might be on Zoom,
01:07:46though. Are you on Zoom?
01:07:47Jay? I'm here. You're here.
01:07:51Oh, wonderful. Okay wonderful.
01:07:59Any questions for Jay?
01:08:08I'm just curious of what a
01:08:10placebo would look like.
01:08:11Would it just be a a presentation
01:08:13that was not on the topic?
01:08:17How, how would that be designed?
01:08:20Yeah. So first of all,
01:08:22I think like the first part of
01:08:24my presentation where I think,
01:08:25you know, like the Lustman Committee
01:08:26and Lustman family as well as,
01:08:29as well as even acknowledging
01:08:31the honor I was cut out.
01:08:33So I'd like to do that.
01:08:35And I'd also like to thank Doctor
01:08:36Rose and that for his very
01:08:37generous introduction as well.
01:08:39So we're doing a follow up study.
01:08:42This is going to be like a cluster
01:08:45randomized control trial of a radio
01:08:47intervention to measure you know
01:08:49the difference between stickman.
01:08:50What we're doing there is a control
01:08:53group is listening to programs that
01:08:55are not related to mental illness,
01:08:57but they're they're they're you know,
01:08:59listening to some random programs and
01:09:02you know we're serving pre and post.
01:09:05And the idea is that the idea is
01:09:09that that could be a control group.
01:09:22So I see a question on the chat
01:09:24about major depress of this order.
01:09:26In the community that we work
01:09:29major depressive disorder is not
01:09:30really regarded as mental illness.
01:09:33You know generally when people think
01:09:35of mental illness or you know they
01:09:37call it disease disease of the skull.
01:09:40Do you think of four things once
01:09:42bipolar disorder, schizophrenia,
01:09:44very severe alcohol use disorder
01:09:47and epilepsy.
01:09:48So those four are considered
01:09:50to be like mental illnesses.
01:09:51So you know that that could be like the
01:09:54topic for a future studies but you know at.
01:09:561st we need to know more about like
01:09:58the local conceptions of depression
01:09:59and how to and how to address that.
01:10:02And we're doing,
01:10:03we're partly doing that this summer,
01:10:05You know,
01:10:05like we're not doing an intervention,
01:10:06but we're, you know,
01:10:07studying it more in terms of like
01:10:09the local attitudes and beliefs.