Yale Psychiatry Grand Rounds: June 11, 2021

June 11, 2021

Information

John P. Flynn, PhD Lecture: "Neural Circuits of Motivational Valance Processing"

Kay M. Tye, PhD, Wylie Vale Professor, Systems Neuroscience Laboratory, Salk Institute for Biological Studies

ID6707

To CiteDCA Citation Guide

00:00For the annual Flynn Lecture an it's my
00:03pleasure to be able to introduce both
00:06the lecture and also our speaker today,
00:10so the Flint Lecture is in honor of
00:13a faculty member in our department.
00:16Doctor John P. Flynn,
00:17who was a pioneer in the understanding
00:20how the brain contributes to behavior,
00:23and it's hard to think now of
00:26how revolutionary his work was,
00:28but in particular.
00:30I look back to this paper from 1964,
00:34which Full disclosure I was already
00:37born but just barely and what he and
00:40his colleagues did was to stimulate
00:43part of the cat brain and to find that
00:47the cat would attack a rat or another.
00:51Other rat sorry a cat would attacker at
00:54yes and that this would be completely
00:58independent of any other stimulus.
01:01Any other thing happening in the environment?
01:04Just this electrical stimulation
01:06could result in attack,
01:08and it began to form our concept
01:11that aggression had a brain basis
01:14and this work is still inspiring.
01:16Today in 2011,
01:18a paper in which the hypothalamus
01:20was stimulated with light.
01:23In using novel molecular genetic
01:25tools was really based very much
01:28on this 1964 paper and Alan Lewis,
01:31who was a resident in our department
01:34and who worked in our lab,
01:37also followed up on this work
01:40using molecular,
01:41genetic and pharmacological tools.
01:43Looking at Astro choline and its
01:46receptors so seriously influential
01:48from the time that this work
01:51was done to the present day.
01:53I also want to note that we have
01:56Doctor Flynn's daughter Sarah Flynn.
01:58Joining us today and it's always
02:01been one of the real pleasures of
02:03this of this lecture for me to
02:06meet Doctor Flynn's late life,
02:08Holda and his daughter Sarah.
02:10I wish that we could be meeting in person,
02:14but certainly in the coming years.
02:16As as we move beyond this
02:19pandemic will move back,
02:20hopefully to to having
02:22this lecture in person.
02:24The winners of or the winners.
02:27The presenters of this lecture have been
02:30luminaries and our current presenter,
02:33Doctor Campti is is no less of a
02:37luminary than any of our previous
02:40presenters Ann Dr Ty did her work.
02:44Her undergraduate training at the at MIT,
02:47where she graduated with a major in
02:51brain and cognitive sciences in 2003.
02:55After taking a year off to travel,
02:58she joined Patricia Genics lab at UCSF
03:01and began her work in electrophysiological
03:04recording from brain areas.
03:06Important for emotional behavior
03:08and connecting those circuits too.
03:10Important behavioral outputs.
03:12She then moved to Carl Daesrath
03:15Laboratory at Stanford University
03:17as a postdoctoral fellow,
03:19where she began to use the
03:22molecular genetic approaches.
03:23The labeling approaches to identify
03:26specific neuronal subtypes.
03:27These stimulate stimula Tori approaches that
03:29allowed her to isolate specific sub circuits,
03:32and these tools allowed her
03:34to unleash your creativity.
03:35Ann became the basis of the work
03:37that her lab members she and
03:40her lab members have been doing.
03:42To this day,
03:43she became an assistant professor
03:45at MIT in 2012,
03:46was rapidly promoted to
03:48associate professor with tenure,
03:49and then moved her laboratory
03:51to the Salk Institute,
03:53where she was promoted to professor.
03:56Doctor Ty is currently professor
03:58and the Wiley Veiled Chair of
04:00Systems Neuroscience at the Salk
04:03Institute for Biological Sciences.
04:04She's also an adjunct faculty member at the
04:08University of California at San Diego and,
04:11as I mentioned,
04:12her research broadly is focused on
04:14understanding the neurobiological mechanisms
04:16underlying social and emotional processes
04:18at the circuit level at the cellular level,
04:22at the synaptic level,
04:24and particularly those relevant
04:26to psychiatric illness.
04:27Professor ties been recognized with a
04:30number of prestigious research awards.
04:32In fact,
04:32it's hard to find out an award that has
04:36not really pulled out her work to to honor.
04:40She received the NIH Director's
04:42New Innovator Award,
04:43the presidential early Career award.
04:45For scientists,
04:46scientists and engineering
04:47from the White House,
04:49the Society for Neuroscience
04:50is Young Investigator award.
04:52She was listed as technology
04:54reviews Top 35 in among.
04:57The reviews top 35 innovators under 35.
05:00She received the NIH
05:01Director's Pioneer Award,
05:03and she's also been recognized
05:05with a number of awards.
05:07From mentoring at the undergraduate,
05:09graduate and postdoctoral level,
05:11and this is a place where I know
05:15that she really has made incredible
05:18efforts and I want to highlight her
05:21trainees who have been at the forefront
05:24of the wiki project to recognize.
05:27Scientists who are generally
05:28under represented,
05:29both women and members of
05:31underrepresented groups in Wikipedia
05:33for the work that they've done,
05:35and it's no short.
05:37Nothing short of a revolution.
05:39What she and her trainees have done
05:42to further the efforts of outreach,
05:44diversity and inclusion in science.
05:47So without any further ado,
05:50I'm going to now.
05:53Turn the podium over to doctor
05:55ties presentation.
05:56She will be joining us
05:58for questions at the end,
06:00so please do put them in the chat.
06:04I will moderate the chat as after
06:07the presentation is finished and
06:09I hope that you will all join
06:12me in welcoming doctor Thai.
06:14And I will now. Share her presentation.
06:26Hi, it's such a pleasure to be here
06:28for the Flynn lecture at Yale,
06:31where I'm going to tell you about
06:33the neural circuits of motivational
06:35processing focusing on violence.
06:40Some stimuli carry innate emotional valence.
06:45No training or learning is required
06:47to have emotional responses to them.
06:50Other stimuli are initially neutral,
06:52unless paired with an emotionally
06:54significant stimulus.
06:55So to have an emotional response to this,
06:58we need to learn the associations between the
07:01Twin Towers and the horrific events of 911.
07:05There's been quite a bit of controversy in my
07:08field about whether emotion is experienced
07:11by humans and by animals in the same way.
07:15There is no way to ever really know
07:17how an animal experiences emotion.
07:19There is no way to say whether it
07:22is the same as a humans experience.
07:24There's also no way for me to ever
07:27truly know whether my subjectively
07:29experienced emotions are the same
07:32or different as another humans.
07:34So you know what I'm interested
07:36in is emotion.
07:37What I actually study is motivated behavior.
07:42I've always been interested in how
07:43we assign positive negative balance
07:45to environmental cues. For example,
07:47think of a track star and a war veteran.
07:50They both had the same auditory stimulus,
07:53bang a gunshot.
07:54The track star might have a rush of
07:57excitement while the war veteran
07:59might have sensations of panic or fear,
08:02someone presented with these stimuli many
08:04times without consequences may habituate.
08:06An experienced neutral
08:07valence and low arousal.
08:09So why would the same sound produce
08:11opposing reactions in different individuals?
08:14Sure, they've had different
08:16experiences in the past.
08:18That have changed the way they experience
08:20their environment now, but how?
08:21How do our brains tell us if
08:23something is good or bad?
08:25What actually makes our brains different?
08:27Solving this mystery would shed light on
08:30the fundamental principles of survival,
08:32as well as really help us understanding.
08:34The key hallmark features of many
08:37different psychiatric disease states.
08:39So by plotting things here in
08:41a simplified version of the two
08:43dimensional theory of emotion,
08:45where on one axis we have the
08:47intensity or the arousal,
08:49and on the other axis we have
08:51balance orthodontic value.
08:52This is the implication,
08:54and this is we can also compare this.
08:57Framework with another framework called
08:59the two factor theory of emotion that
09:02suggests an order of operations.
09:04And then she got back to theory of motion.
09:07We think about stimuli that coming
09:08in and the first question we have to
09:11ask ourselves is, is this stimulus important?
09:13You can think of that as absolute
09:15value if it's not important,
09:16we ignore it.
09:17If it is important,
09:18the next question we need to ask
09:20ourselves is how is it important?
09:22Is it something that is good or bad?
09:24Do we want to avoid this Daniels or
09:27approach it?
09:28And so this is what I want to focus on today.
09:31Where is the fork in the road and
09:33how do we determine what gets routed
09:35down one path or another?
09:39So if we think why am I so obsessed
09:41with understanding the perturbations
09:43that could occur with you know
09:45of motivational valence, really?
09:47Because I think this could be a
09:49common thread between many different
09:52psychiatric disease states.
09:53For example, in the case of anxiety,
09:56perhaps there is a shift to too
09:58much motivation to avoid potential
10:00negative consequences relative
10:01to motivation to seek rewards.
10:04The converse can also be true.
10:06Too much motivation to seek potential
10:08rewards and not enough motivation to
10:10avoid potential negative consequences.
10:12And you know, admittedly,
10:13this is a very simplified perspective.
10:16And then in the case of depression,
10:18perhaps we just have reduced
10:20motivation over all.
10:22So how do we implement this process?
10:24How does the brain carry out this very,
10:27very simple process of asking
10:29if something is good or bad?
10:31So there are a few different neural
10:33circuit motifs that we've seen,
10:35you know,
10:36recur and appear over and over
10:38again as the field of circuit
10:41neuroscience has developed and matured.
10:43And I'm not going to really talk
10:45about the labeled lines model just
10:47because it's sort of a straw man.
10:49When we think about balance processing
10:51because it doesn't really allow for
10:52learning to happen reversals to occur.
10:54You know,
10:55aside from just strengthening or
10:56weakening the lines of information
10:58flow as they already exist,
10:59you know it doesn't.
11:00It doesn't explain how we can learn to
11:03like a bitter taste like coffee or beer,
11:05or how we can adapt to changing
11:07conditions in the environment.
11:10So what we want other machines that
11:13we have explored include divergent has
11:16where a common scene is is coming on.
11:19To posit that send information to
11:21different downstream targets were
11:23different processes may occur,
11:24and pretty Danbury and Anabella
11:26have worked on this in my lab.
11:29Will talk about that momentarily.
11:31Another possible motif is that opposing
11:34components where the anatomical origin and
11:37and target the point A&B are the same,
11:39but the lines of.
11:40But the circuits that are are
11:42sending messages from point A to
11:44point B are different functionally.
11:46For example,
11:47if there glutamate and GABA
11:49going from point A to point B,
11:51what really matters is is the
11:53neurons integrate that information
11:55of those opponent components to
11:57sort of do a winner take all?
11:59And for that, my graduate student,
12:02former graduate student,
12:03Edward now who's now at Princeton,
12:05explored this in his thesis work,
12:07and then the 4th would be
12:09that of Neuromodulatory gain.
12:11And so we've looked at this
12:13in the prefrontal cortex,
12:15Caitlin Vanderweel and Cody
12:16Siciliano have explored this in and
12:18today I'm going to tell you about
12:20a new venture of understanding
12:22how neuropeptides can influence
12:24the divergent circuit motifs that
12:26we already know about within the
12:29basil lateral and make the left.
12:31For today I'm going to be focusing
12:33on these two motifs and how they're
12:36relevant for balance processing.
12:37So the road map for today are just three.
12:41Sort of big general questions.
12:43One is a little bit of review
12:45where to circuits encoding positive
12:47and negative bills diverge.
12:49There's many places where they could diverge,
12:52but will will go through one case study.
12:55What are the local interactions between
12:57these functionally distinct circuits?
12:59Just form give rise to function
13:01and then third how can narrow
13:04modulation participate in the
13:06process of valence assignment?
13:08And then well.
13:09Rabbit up,
13:10so the first hints of the Amygdala's
13:14involvement in emotional processing
13:16came from temporal lobe activities
13:18in primates over a century ago.
13:21Following Brown and cheaper work,
13:23Kluver Bucy coined the term
13:25psychic blindness to refer to what
13:27happens to these animals when they
13:29lose their innate fear snakes,
13:31or,
13:31you know,
13:32sort of lose interest in food and treat
13:34sort of inanimate objects like light bulbs,
13:37similarly to what would have been
13:40previously fearful rewarding stimuli or
13:42Larry Weiss grants then identified these.
13:44These phenotypes were actually
13:46specifically attributable to
13:47selective amygdala lesions later on.
13:50There's also the important
13:51case study patient ** who,
13:53following bilateral amygdala damage,
13:55lost fear to snakes and spiders.
13:58The ability to recognize
13:59emotion in other peoples faces,
14:01and even lost the fear.
14:03Fear responses to being mugged,
14:05which actually happened several
14:07times in her life, however.
14:09When she was at suffocation when
14:13she's experiencing suffocation,
14:14she still could have the autonomic
14:17responses increase in autonomic arousal,
14:20heart rate,
14:20sweating, etc.
14:21Suggesting the amygdala is important
14:23for the assignment of environmental
14:26significance to sensory stimuli,
14:28but not necessarily the production
14:31of the autonomic or arousal
14:33and physiological responses.
14:36So in my in my lab, we work in mice and
14:39if you take a look at the mouse brain,
14:41take a little section and look here,
14:44there are 13 sub nuclei of the amygdala,
14:46actually, but I'll just be focusing on
14:48two of them for simplicity of today.
14:54So where does circuits encoding
14:56positive and negative balance diverge?
14:58Well, I've already sort of hinted
14:59to you that the bill is important.
15:02Why you know what?
15:03What else about the amygdala?
15:05Do you make alot is kind of this
15:08primitive analog of of the cortical
15:10striatal circuit and some have
15:12referred to it as the alligator brain.
15:15It's it's primitive brain within the
15:18brain because the amygdala days auto
15:20mgla is 90% glutamatergic neurons
15:22and described the cortical cortical
15:24like in that it forms associations
15:26is not a laminated structure like
15:28the cortical like the cortex,
15:30but other than that there
15:32are some similarities.
15:33The central amygdala is
15:35striatal like because.
15:36It's 95% Gabaergic medium spiny
15:38neurons in a similar manner.
15:40So there is a very rich body of
15:43literature suggesting that the
15:45PLA is well positioned to be where
15:47valence encoding neurons exist.
15:49So we know that neurons in the
15:52PLA encode positive and negative
15:54balance that all modalities of
15:56sensory information converges in
15:58the basal lateral make the love that
16:01learning induces synaptic plasticity.
16:03And just a brief didactic direction here,
16:06so so.
16:08We believe plasticity occurs in
16:10synapses that were previously weak
16:12synapses that might carry information
16:14about the condition stimulus,
16:16such as an auditory tone,
16:18but once is paired with either
16:20a reward or punishment,
16:22will be strengthened,
16:23and so just to I'll come back to this later,
16:27but just very quickly.
16:29We are using ampata NMJ ratio as a
16:32proxy for glutamatergic synaptic strength,
16:34because after your garden variety LTP occurs,
16:37you will see an.
16:38Increase in Emperor Scepter
16:40mediated currents.
16:47We have already seen that fear
16:50conditioning increases Amp'd MD ratio
16:54in putative flammable DLA synapses.
16:57In this particular putative synapse
16:59I've also found when I was a graduate
17:03student that reward conditioning
17:04also increased ampata MDA ratio in
17:07putative clam Omega Listen apps is
17:09in the same computer the senses,
17:11and so when I was a grad student.
17:15I presented these data.
17:16This is my first time meeting and
17:19one of my colleagues came up to
17:21me afterwards and just, you know,
17:22gave me gave me a lot of Flack about
17:25you know how does this make sense?
17:27How can the same mechanism underlying
17:29fear and reward conditioning?
17:33Well, there's a few different possibilities.
17:36Maybe the amygdala just encode salients and
17:39it it you know anything that's important,
17:42good or bad, will induce plasticity.
17:46Or maybe the amygdala is actually
17:47the site of valence assignment,
17:49but it occurs by a distinct projections
17:51that we just didn't discriminate between
17:52when we were patching blindly. So dumb.
17:57We know that the essential medial nucleus
18:00is critical for the expression of fear.
18:03We know that up genetically stimulating,
18:05central medial nucleus neurons
18:07evokes freezing responses.
18:09We also know that disconnecting the basal
18:12lateral amygdala from the central medial
18:15nucleus will abolished fear expression.
18:17But of course,
18:18the central amygdala does many other things,
18:20and there are a number of
18:23fantastic studies that show this.
18:25So it's not this simple.
18:26Same thing for the nucleus incumbents.
18:28Of course we know that the nucleus income
18:31is important for diversity of functions,
18:33but it's best known for its importance
18:36in reward related processes.
18:39We know that Papa genetically
18:40stimulating VLA terminals and
18:42nuclear comments supports self
18:43stimulation in place preference.
18:48When Praneeth Embery Annabelle or
18:49came to my lab, they came ready to
18:52tackle this question about what the
18:54circuit mechanism is for signing
18:56positive and negative balance.
18:58So the hypothesis is simple,
19:00just that the specific downstream
19:02projection neuron matters.
19:03So if you pair a tone with
19:05foot shock for example,
19:07you'll see a strengthening of synapses.
19:11Coming on to see M projectors.
19:13Whereas if we are.
19:16I'm going to pair the tone
19:18with a reward like sucrose.
19:20We would see strengthening of
19:21synapses coming on to feeling neurons,
19:23projecting the nucleus of
19:25Commons so very simply,
19:26we just used retrogradely traveling for us.
19:28It feeds and check into them into
19:30either the nucleus of comments
19:32or the central medial nucleus,
19:34then train animals and either
19:35a fear conditioning task or
19:37reward conditioning task.
19:38Simulated putative built Alanic
19:40inputs to the amygdala and
19:42then recorded just as before.
19:44Except now we just know where
19:47the targets are very simple.
19:50And what we found was that after fear
19:53conditioning we see a button increase
19:55in synaptic strength on to see M neurons,
19:58CN projectors excuse me and decrease
20:01in amber ratio on to PLA to see M
20:04neurons after reward conditioning.
20:08Conversely, for being
20:09late to a Commons neurons,
20:11we saw synaptic strengthening on to
20:13them or some Excuse me Ltd a reduction
20:16and Anthony ratio after fear conditioning,
20:18but an increase in synaptic strength
20:20after reward conditioning.
20:21So just the opposite.
20:24And importantly,
20:25what I want to point out is that even
20:29conditions like food restriction can
20:31change the basil amput empty ratio.
20:34So if the visa lottery legal is truly
20:36the site at which we translate sensory
20:39information into motivated behavior,
20:40then be light inputs should be able to
20:43drive motor behaviors in the naive animal.
20:48OK, but is there a causal relationship
20:51and we show that there is.
20:53If we fail to stimulate
20:55village Commons animals,
20:56we get intracranial self stimulation.
20:58Just replicating work that gets tuber
21:01and John Britton have already shown.
21:05And we also show that if we simulate dealing
21:07arms protecting the central medial nucleus,
21:10we actually get punishment
21:11avoidance place avoidance.
21:13So OK, if it's true that if this whole
21:16hypothesis is true and this is relying
21:19on an MDA receptor dependent mechanism,
21:21what happens with with an MD or soldier LDP?
21:25I'm sure you know,
21:27but if you don't know glutamate is released,
21:30it will bind to gloomy receptors,
21:32including an perceptors.
21:34An MDA receptors,
21:35however NMD a receptors are have a
21:38magnesium blockade that sits in the poor
21:40of energy receptors and the depolarization.
21:43Caused by glutamate binding
21:44and amp receptors.
21:45Allows the influx of sodium.
21:47Then it becomes depolarized.
21:49This allows the magnesium blockade to
21:51come out of the NBA receptor poor,
21:53allowing calcium influx to occur.
21:56So if this is,
21:58you know,
21:59this is all true that this is
22:01how energy receptors.
22:02This is how NBA receptor dependent LTP works.
22:06Then hyperpolarizing the postsynaptic
22:07neuron at a critical time point here would
22:10then prevent learning because we couldn't,
22:13you wouldn't be able to remove
22:16that that magnesium blockade.
22:18So we did this experiment.
22:20This is bilateral Weezer dual virus
22:23approach to express halorhodopsin
22:25in either Bealach na Si Productions
22:27or belay to see M Productions
22:29and then train them again on a
22:32fear or reward conditioning task.
22:34Photo inhibiting only when the
22:36US the unconditioned stimulus was
22:38presented either for shock or sucrose.
22:41And what we found.
22:44I've sort of what I said for the
22:46fear conditioning and that when we
22:48photo inhibited be late to a comment,
22:50something happened.
22:50But if we silenced Beladice
22:52central medial nucleus projections,
22:53we saw an impairment in fear conditioning.
22:57When we looked at reward conditioning,
22:59however, something sort
23:01of unexpected happened.
23:02It was surprising because we
23:03actually didn't see an impairment
23:05to belay and astie inhibition,
23:07as I would have expected,
23:09we actually saw in enhancement
23:11when feeling to see an innovation.
23:14OK, so that was a little confusing,
23:16but it'll make sense later.
23:18So so just an interim summary is
23:21that I've told you that opposite
23:23synaptic changes map onto
23:25production following either fear,
23:27conditioning or reward conditioning.
23:29If we just, you know,
23:31circumvent all of this upstream
23:33plasticity and just cut right
23:35to the chase and manipulate
23:37these projector populations,
23:38activation of projections will
23:40either cause avoidance or approach.
23:43If we then bilaterally inhibit
23:45either of these productions,
23:47will see that inhibition of central
23:50medial projectors impairs fear
23:53but enhances reward learning OK?
23:56But is it really that simple?
23:58And of course the answer is no.
24:00There's a lot of heterogeneity.
24:02It's it's, you know,
24:03all these neurons that are
24:05anatomically defined,
24:05have a lot of functional heterogeneity.
24:08We do address this that even though
24:10it's really exciting that we can target
24:12things more specifically than ever before,
24:14we're still possibly only
24:16observing a majority vote.
24:17There you know,
24:18just because an offer genetic
24:20manipulation can produce only one
24:22measurable behavior in a given task,
24:24this does not in any way
24:26suggest functional homogeneity.
24:27Of neurons anymore than having you know,
24:30one president of the country suggest
24:33that everyone agrees who it should be.
24:36So Eve Marder,
24:37once said very wisely,
24:39you know very artistic I
24:40love I love this club,
24:42that optic tools tell us what neurons can do,
24:45not what neurons do do.
24:47So it's very important we need to
24:48be able to explore the endogenous
24:50functional properties of neurons.
24:52Really understand them.
24:53So the minimal the absolute minimal
24:56criteria for violence in coding
24:58a single cells is that number 10,
25:00and I should say before I get
25:02into this set on one axis,
25:04there's the response to rewarding CS is.
25:07On the other access,
25:08there's a response to aversive
25:10CS is in either case,
25:11and in both cases we can either have
25:13innovation, no response or an excitation.
25:16Uhm?
25:19So the so,
25:20given these combinations of responses.
25:22Of course, if.
25:25Just.
25:28They conclude that it fails in
25:29coding a little bit harder is to say
25:32that it's differential responding.
25:33You want to say if it's
25:35responding in a similar,
25:37indistinguishable manner to to
25:38both positive and negative stimuli
25:40that it is not encoding balance.
25:41This is pretty important.
25:43Oftentimes studies will only show or
25:45response to one or the other within
25:47a given cell an you really need
25:49responses to both without seeing
25:51responses to both in the same cell
25:53you you can't differentiate between
25:55salience responses and balance responses.
25:58And then finally it needs to be
26:00independent of sensory features.
26:01It's not just tracking the modality
26:04or the specific frequency of
26:07the tone or whatever.
26:08OK, so this just underscores the
26:11importance of seller resolution recordings.
26:13So on a baler performed this
26:16and it had fixed task an.
26:18In this task,
26:20one tone predicts sucrose delivery.
26:23As you can see here and then in another,
26:25another town predicts quiet time delivery,
26:27which is a no go trial.
26:32So we did a match modality,
26:35build task where both the CSS and
26:38both US is were of the same modality.
26:42Auditory stimuli and gusta Tori outcomes.
26:46And generally speaking,
26:47for the non specific PLA recordings,
26:50we just reproduced publications that already
26:52existed from Dan Saltzman and Trajanic slabs,
26:55but we didn't, you know,
26:56go through all this trouble.
26:58We did record 1600 neurons,
27:00just we producers,
27:01although it's always nice,
27:03we wanted to know what neurons
27:05of Goan production target did.
27:08And so we used an approach
27:10developed in Tony Zeiders lab,
27:12which he termed Photostimulation assisted
27:14identification of neuronal populations
27:16which many people would refer to as
27:18photo identification or photo tagging.
27:20And basically the idea is
27:22you've got an electrode.
27:23You've got an optical fiber,
27:25and you've got some specific
27:27cell population that you have are
27:30expressing opposition and so first.
27:32You know it's near on a spikes.
27:35We can record that if neuron B spikes,
27:37we can record that too and we just
27:39want to record all the endogenous
27:42naturally occurring activity during
27:44the task after the task is over,
27:46we can shine light and get
27:48spikes out of the neuron.
27:50That is expressing option.
27:53Then we can look at that and determine
27:55whether there was a photoresponse or not.
27:57Importantly, very,
27:58very importantly,
27:58we need to control for recurrent excitation,
28:01particularly if you're doing photo tagging
28:04in any place known to have it like cortex,
28:08hippocampus, amygdala.
28:10What does this mean?
28:11Well,
28:11it's just referring to this compound
28:13that can occur if you have secret to
28:15expressing ones that are protected from,
28:17for example, BL A2 incumbents.
28:18Let's say that's great.
28:19Where were expressing in those cells?
28:21We can verify that anatomically post hoc,
28:23and it's easy in vivo to tell the
28:25difference between a teacher to expressing
28:27cell and a non expressing neighbor
28:29that is not responsive to light at all.
28:32However,
28:32there may also be non expressing neighbors
28:36that receive recurrent excitation
28:38from the opsin expressing neuron.
28:41And so when we pass,
28:42we can validate this.
28:44That doesn't help us in vivo necessarily,
28:46you'd think.
28:47But let's look at this and we can.
28:50We can patch.
28:50We can look at different light
28:52powers and determine a distribution
28:54of photo response latency's,
28:56because if it is through
28:57recurrent excitation,
28:58there's this added synapse.
29:01So sometimes the response to non
29:03expressing neighbors that are receiving
29:05input which you can determine and
29:07know about unequivocally ex vivo,
29:09you get the distribution.
29:11This distribution is completely
29:12non overlapping with those cells
29:14are actually trying to record from
29:16this feature to expressing neurons.
29:18That's great. That's often the case.
29:21Sometimes, though,
29:22the case is that there's some partial
29:24overlap where you for wrestle with
29:25once a pot with a false positive,
29:27once a false negative,
29:28and then some cases there's
29:29so much overlap issues.
29:30Different approach,
29:31like maybe you know genetically incredible
29:33calcium indicators or something.
29:35So we did this.
29:36I'm going to skip.
29:37I'm going to skip to the
29:38punchline of this story
29:39to tell you about new,
29:41unpublished data, but.
29:42We use multiple different types
29:44of analysis and found that in
29:46general it's true that village,
29:47an AC predominantly encodes positive balance.
29:49Be latest central amygdala predominantly
29:51encodes negative balance, and we've
29:53shown this in a bunch of different ways.
29:57So it is heterogeneous genius.
29:58There is a lot of heterogeneity,
30:00but the general trend is history.
30:02So what about the local interactions between
30:04these functionally distinct circuits?
30:05Why do we have them all like salt
30:07and pepper mingled together in
30:09the basal lateral make alone?
30:13What is the microcircuit architecture,
30:16so to speak well?
30:20If I have here in green reward encoding you
30:22know principle drones or production runs,
30:25fear encoding principles and then of course.
30:29Some interneurons that are that
30:32are Gabaergic interneurons,
30:33you might have one motif where there's
30:36mutual inhibition, where reward
30:38encoding neurons can silence fear,
30:40encoding neurons, and vice versa.
30:43You might actually also have a uni
30:46directional or asymmetric inhibition,
30:47where fear encoding neurons would
30:49silence reward encoding runs,
30:50but not vice versa.
30:51And honestly, when I first drew this picture,
30:54I just drew it 'cause I thought like
30:57technically it's another possibility.
30:58I didn't actually think it was
31:01going to be the answer.
31:03And then of course,
31:04like the third possibility,
31:05being that there's a neutral
31:07excitatory response.
31:08Felt thanks to Klong Hoon Chung.
31:10We were able to look at this and you
31:12can see here some bling NEC ability
31:14to see an cells within the lateral
31:17nebula intermingled so much together,
31:19but also you know there's a gradient.
31:22So indeed there are gradients.
31:25We can quantify them,
31:27and they are intermingled.
31:29So this sort of underscores the need
31:32to think about the relationship between
31:34these different populations of neurons,
31:36so I'll just show you one little tidbit here.
31:40But when we looked at the latest
31:42central amygdala neurons that
31:44were photo identified here,
31:46we've got 33 units.
31:47We can also see neurons that
31:49are photo excited,
31:51meaning they have slower latency,
31:52excitation, and response to light.
31:55Order on that were photo inhibited,
31:58meaning they were silenced by the activation
32:00of the late Essential Omega neurons.
32:02So if you look at these numbers,
32:05you can see that the the the amount
32:08of neurons that it's silent.
32:10These bileta CN neurons or silencing
32:12is greater than them themselves,
32:14and so there's this huge ability for them
32:17to suppress the responses of other neurons.
32:20So.
32:21You know,
32:22perhaps it's true,
32:23like if we compare these numbers that
32:25I just showed you for the normalizing
32:28to the neurons that actually expressed
32:30channelrhodopsin and then looking at
32:32the proportion of neurons are excited
32:34or inhibited by those projectors.
32:36You can see that the latest Central
32:39McDonald's inhibit far more neurons
32:40than do either of the other productions.
32:43So maybe when it comes down to
32:45a majority vote, so so to speak,
32:48at the Bellator CN population is
32:50more influential than local network.
32:52And maybe their votes count more similar,
32:55you know,
32:55so that might be like something
32:57that we want to consider.
32:59It's not necessarily that all
33:01votes are equal,
33:02not necessarily just a majority vote.
33:04Some neurons are positioned to be
33:06able to exert greater influence
33:07over their neighbors, like be late,
33:10essential in Glen Rose.
33:13Now, why won't the brain work this way?
33:16If we record from this is published
33:18in in Gwendolyn Calhouns preprint.
33:21But when we record from from NEC
33:24Projectors and green Eyes,
33:25just abbreviated it end and patch
33:28a record from from any project and
33:31stimulate CN protectors or vice versa.
33:33You know, recording stimulate
33:35from the opposite pairing you.
33:37We actually see that they have
33:39different asymmetric impact on
33:41each other when you stimulate.
33:43Uhm? A common projectors recording
33:46from central appraisal of
33:49projectors you actually facilitate.
33:51Excitation, whereas in the other
33:53direction you suppress it, so why would?
33:55Why would the brain work this way?
33:58I did not expect to see an asymmetric
34:00unidirectional relationship,
34:01but maybe that makes sense because you
34:03know you can mate or eat or drink later.
34:06You need to escape from that predator
34:08right now that is the most immediate,
34:11most urgent demand or threat
34:12on your survival.
34:13So you have to address that.
34:18Further, maybe it makes sense 'cause
34:19reward seeking is just inherently risky.
34:21If I'm a mouse that lives in in my
34:24borough and I need to, I need to.
34:26I want to forage for food or water.
34:28Then you know, providing a skip
34:30could be a good insurance policy.
34:33OK, so that's some speculation
34:34and just contacts, but how does
34:36neuromodulation play a role in here?
34:38So this is the third final questions.
34:41Main big question or focus on today.
34:45And for that meeting that I want to
34:47actually show you the the questions
34:49that that make me lose sleep at night.
34:51So it's great that we found that
34:53you know after fear conditioning
34:54we see strengthening of synapses
34:56going down one path afterward,
34:58conditioning with the strengthening of steps,
35:00going down another path, but.
35:02When this is happening in real life,
35:05you know animals are capable of 1
35:08trial learning. How is this possible?
35:11You know?
35:11So I think the timescales how do
35:13the synapses know which postsynaptic
35:15neurons are which,
35:16and what you're supposed to do,
35:18so you know if you think about
35:21spike timing dependent plasticity,
35:22you know shifting things from
35:24just 50 or 100 milliseconds.
35:26Can dramatically impact whether
35:28it's LTP or Ltd.
35:29If we record in vivo,
35:31you can see that after the onset of
35:33the Q this the the auditory tone
35:36just 100 or 200 milliseconds later,
35:38we're back down to baseline.
35:40We don't have.
35:41You know the neurons are no longer
35:43active even as the Q continues to play.
35:46And then let's think about the
35:48most common paradigms used in the
35:50field where condition stimulus
35:52will play for 10 or 20 seconds,
35:54the beeline neurons that are
35:56critical for the plasticity.
35:58Please tell a very transient activation
36:00at the onset of the condition stimulus,
36:03then then go back down to baseline and
36:05you know could be 10 or 20 seconds.
36:08The unconditioned stimulus is presented.
36:10So how does the brain solve
36:12the valence assignment problem?
36:13How does the brain attach these things
36:16that are so desperate in time together?
36:18So first there's the question of how do they?
36:22How does the brain know which
36:24synapses route information to wear?
36:25So there's a few different options here so.
36:28In terms of knowing,
36:29knowing who's who in terms
36:31of the postsynaptic neuron,
36:33we wanted to explore this and so we
36:36sequenced violate any simulated CM neurons.
36:38And look for surface receptor specifically.
36:41Indeed,
36:41one service provider that we identified as
36:44interesting is the Neurotensin receptor,
36:46one which we see to be enriched have
36:49enriched expression and Sian projectors.
36:51Now why am I going to focus in on
36:54the nearest ends in one receptor?
36:58Well,
36:58we know that there is a one receptor
37:01controls a G protein coupled receptor GP CR.
37:04The function of this receptor has
37:06been linked to both LDP in the PLA and
37:09also contextual fear conditioning.
37:11And the thing that really got me was
37:13done by Kimberly Kempadoo and she found that.
37:17Different concentrations of narrow tense,
37:19and this is a different cell population,
37:22but different concentrations are
37:24intense and could actually either
37:26facilitate glutamatergic transmission
37:28or suppress glutamatergic transmission.
37:31So if this is possible, you know then?
37:35We hypothesize that but the same
37:38concentration of their attention could then,
37:41you know, drive effects differently on
37:43these projector populations that have
37:45different receptor expression profiles.
37:47OK, so if different concentrations
37:49are neurons at the same profile,
37:51can can make glutamatergic transmission
37:54either be facilitated or suppressed,
37:56then maybe one concentration of their
37:58attention can shift the balance from
38:00one population to another if they have
38:03different receptor expression profiles.
38:08So first we want to explore
38:10this an we performed.
38:12We did some pharmacological manipulations
38:14where we administered the audience
38:17or want agonist in today's letter
38:19will make the lab and we actually saw
38:22an enhancement of reward learning.
38:25We saw a nonsignificant trend towards
38:27an impairment of fear conditioning.
38:29OK, so that's potentially, you know,
38:32suggested that narrative could
38:33play a role in balance processing,
38:35and we also see that indeed there
38:38are different effects you know
38:40match for credit concentration of
38:42Billy and Acnb licien neurons.
38:44We do indeed see different
38:47dose response curves.
38:49These differences at a given concentration
38:51of tendon, animal or near attention,
38:54we can see an increase of billet and AC.
38:59Amplitudes EPS evoked from Billy the
39:02NACS neurons, and a suppression of EPS,
39:04sees a bug from the legacy M neurons,
39:07and this is blocked by the NPS
39:10are one antagonist.
39:11OK, great, but where do we get near?
39:14It ends in front.
39:16Where does it come from?
39:17So we we trace upstream and looked for
39:20colocalization of Tracer suggesting
39:21that you were protecting the leg as
39:24well as expression of neural tension
39:26in the New York Times and cream mask.
39:30And so we found that the PDT was a
39:33prime target and photo stimulating
39:35PDT to belay neurons produces valence
39:38specific effects that we actually.
39:41If we stimulate this PT input to be alive,
39:45which includes neurotensin neurons,
39:47we see a facilitation of reward learning
39:51and an impairment of fear looming.
39:54Next to this huge body of work
39:56that I'm about to tell you about to
39:58get into is a by Hailee a postdoc
40:01in my lab here at salt.
40:03And what,
40:03how found was that the PDT to bill a
40:06nuro 10 synergic projection contributes
40:09heavily to Billy computations,
40:11and so he said to look at this and
40:14interrogate the contribution of Neuro
40:1610 synergic input from this specific pathway,
40:20we first Chris Byrd.
40:22We used CRISPR to knock down
40:24that near attend and Gene.
40:26And so the PDT input would either have
40:29knocked down or control of your attention.
40:33Then we wanted to be able to perform
40:36recordings in the basil lateral amygdala,
40:39in both control an crisper animals
40:41and be able to photo identify
40:43the different populations.
40:45So first I'm just showing you that the
40:48crisper works we can inactivate the
40:50near 10s and gene and preserve glutamate.
40:55Narrative the neurons coberly's glutamate.
41:00And we found that the crisper animals
41:02will be knocked down the narrow
41:04Tenzin gene within the PD gbla we
41:07see an impairment of reward learning,
41:09and this also promotes fear.
41:11Learning this is again suggesting this role
41:14for near Tencent in Valence processing.
41:17OK, but how is this role actually like?
41:20What does this look like?
41:22What? How does?
41:23How do the computations that
41:25are being performed in the PLA?
41:27How are the computations being
41:29modified when you know there's
41:31more attention or not?
41:35And so dumb. How recorded from hundreds of
41:38neurons from the basal lateral amygdala,
41:41in a task where animals had to just
41:44differentiate between a reward and
41:46shock predictive cues, as well as a
41:49neutral Q that didn't predict anything,
41:51and we perform functional agglomerative
41:53clustering, and we can see that
41:55across different types of clusters.
41:57Here, Gray is controlled and blue is
42:00crisper and across the board. But you know,
42:03there's a diversity of responses.
42:06But what you can see across the
42:08board in each of these clusters
42:11is that the response is blunted.
42:13When we Chris Brown your
42:15attention on both directions,
42:16both for excitations an ambitions,
42:18the absolute value of the
42:20amplitude of the response,
42:22the change in physical activity
42:23in either direction to reward
42:25or Punish Inc uses reduced,
42:27so every all the valence
42:29processing just looks blunted by
42:31the knockdown of their attention.
42:34OK, So what about specifically
42:37to BLAN AC projectors?
42:40We see that indeed.
42:44Control animals we replicate this general
42:46effect of having a reward you know,
42:49a positive balance biased encoding
42:52property which drops sort of to neutral.
42:55When we Chris Brown are tense and
42:57same thing central projectors.
42:59We reproduced our initial response
43:01of having a negative balance biased
43:04response of this whole population,
43:06which again is sort of neutralized
43:09or attenuated with the knockdown
43:11of crisper injustice.
43:12Pacific PDT to be late input.
43:15Neurotensin everywhere else
43:17in the brain is intact.
43:19OK,
43:19so this is just showing the opposite effects
43:22of the different projector populations.
43:25And then what I want to show you here is,
43:28this is what normally
43:30happens for each trial type.
43:31We've got sucrose neutral,
43:33an shock condition stimuli,
43:34and each trajectory begins here in time.
43:36Then the Q will be on set and
43:39then the response will occur.
43:41And what am I showing you here?
43:43This is looking at the neural ensemble
43:45as visualized as a neural trajectory.
43:47So what does that mean?
43:50How do you?
43:51How do we think about an ensemble of neurons?
43:52So if you if you had 100 neurons,
43:54I mean here we have.
43:56Sam called 700 rounds,
43:57but if you have 100 rounds and you want
44:00to see what the whole population is doing,
44:02you could plot that in 100 dimensional
44:04space and then just look at each neuron
44:07as representing a dimension and then
44:09just look at the trajectory of that
44:11ensemble in activity space across time.
44:13It's an aerial trajectory.
44:15It's hard to plot 100 dimensional space,
44:17so of course what we can do is perform
44:20dimensionality reduction principle
44:22components analysis and then reduce
44:24reduce the dimensions down to those
44:27that can offer the most covariance.
44:29So that's what we're doing here.
44:31Principal component one principal
44:32component to it that you
44:34components that offer contributed
44:36most covariance here and then.
44:37You can see that the sucrose trajectory's
44:40once the the tone comes on you can
44:43see the animals are showing this.
44:45This divergent of the trajectory's.
44:48These are the control animals,
44:49though. What happens when you put Chris Rock,
44:52Chris Brown? You say?
44:53Well actually it's this little blob right
44:55here that you can't even see anything.
44:58It's so small, so I'm going to
45:00blow it up in this inset,
45:02but here the arbitrary units were on
45:04a different scale. It's so small.
45:06So basically all of these trajectories
45:08shrivel up into into nothing, so when when,
45:10when trajectory length shrink,
45:12that suggests that the ensemble
45:13is either less dynamic,
45:14changing less or changing less quickly.
45:17So. The dynamics of the of
45:20amygdala neurons are blunted
45:22when we knocked down or Tencent.
45:27OK, so Chris bring out the New York Times
45:30and Gene in specifically the PDT tibial.
45:33A population reduces decoding accuracy
45:35of behavior from feeling relativity,
45:37so control animals.
45:38Once the Q comes on, it's very clear
45:41like what what trial type it was.
45:45However, if we do this with crisper
45:47animals and this is, you know,
45:49training with controls testing controls,
45:51you know different different
45:52training and testing of course.
45:54But from the same group.
45:58If we train and test on crisper animals,
46:01they perform significantly less well,
46:03but still are able to code above chance.
46:06Importantly, if we train the decoder
46:08with data from the control and then
46:11test on crisper or train with CRISPER
46:14data and then test on the control data,
46:16we get chance essentially.
46:19Suggesting that crisper and control
46:21animals are using different coding rules
46:25to determine what's going on. OK Anna.
46:28Nice ball in my lab who has has done a
46:31lot of work on another project relating
46:33to a real time Alpha tracker which I
46:36don't have time to talk about right now.
46:39Explored this for analyzing,
46:41analyzing BLTS with more regularity
46:43here and so this is pretty simple.
46:45We're looking at reward trials
46:46and shock trials.
46:47Should be simple, right?
46:49Just punishment and reward,
46:50but yet, even with the responses
46:52that we get from punishment reward,
46:54there are different responses.
46:56For example in with shock conditioning.
46:58You could either get freezing a passive,
47:00you know coping state,
47:01or you can get darting,
47:03which might be thought of as
47:05a more active state.
47:07Similarly,
47:07this is true for reward trials
47:09could be hanging out,
47:11or you can come and actively
47:13approach the port.
47:16So what we found here is that the
47:19neural responses to the rewards,
47:22yes, or the shocks, yes.
47:24For each type each behavioral.
47:29Type of response is is blunted overall,
47:32but specific to each type,
47:33and if we quantify this,
47:35what you can see.
47:37Is that there's a greater impact of
47:41Christopher ING out near attention
47:44in these low motivation States and
47:48or active avoidance states here.
47:51OK, so back to the outline.
47:54Where does circuits you know encoding
47:56positive and negative balance diverge?
47:58The basil lateral make list
48:00one of these prime targets,
48:02but now you know what criteria to look for.
48:06What are the local interactions?
48:08There are absolutely lots of microcircuit
48:11interactions of inhibition and or
48:13facilitation locally at the level of the
48:15lay of these functionally distinct circuits.
48:18And now we're beginning to
48:20explore how narrow modulation.
48:22Can play a role in balance assignment
48:25specially in these sort of initial trials,
48:28and so I don't really think
48:30neuromodulatory systems or neuropeptides
48:32will contribute within trial per say,
48:34but between trials, certainly.
48:38OK, so where are we now?
48:40What?
48:41What can we take away?
48:43What's the takeaway message from a,
48:46you know therapeutic standpoint here.
48:48Well, just that identifying
48:50differentially expressed genes and
48:52functionally characterized functionally
48:53distinct circuits can be leveraged
48:55for selective control of neuronal population.
48:58So instead of a shooting in the dark,
49:01like trial and error
49:03strategy for drug discovery,
49:05why don't we use a circuit based?
49:08Just drug discovery.
49:10Find small molecule targets on circuit
49:14components with known function.
49:16So again, this is just bringing things back.
49:20We've in the amygdala,
49:21Howley, Praneeth, Ambrian,
49:23Jake Olson are contributing to understanding
49:26the role of Neuromodulatory gang.
49:29OK,
49:30so why should you even care about all this?
49:34How does this ultimately matter for people?
49:37Well,
49:37despite the prevalence of anxiety disorders
49:40and other mental health disorders.
49:43Current treatments are not effective
49:45for the entire patient population,
49:46or at least have a lot of
49:49undesirable side effects.
49:51So how do we get to where we are now?
49:54From you know where we are now
49:55with with these sort of no no
49:57treatment is perfect that works.
49:59There's no such thing as a team that works
50:01for everyone that is free of side effects.
50:03How do we get from that to something
50:05where that is possible with every
50:07single individual will find a treatment
50:09that is effective for them that
50:10lasts and doesn't have side effects?
50:12Sounds like.
50:13You know, far away,
50:15how do we get there?
50:17Well,
50:17we have to change our current approach,
50:20which currently involves testing
50:22drugs that act nonspecifically
50:23throughout the entire brain and body,
50:26and doing so in a sort of
50:28again trial and error way.
50:30I think another huge problem with
50:33the way that we're tackling this.
50:35This challenge is poor
50:37disease classification,
50:38poor understanding of what actually
50:40gives rise to the disease,
50:42comorbidity that we see.
50:44We currently rely on symptom
50:46based categorical definitions of
50:48mental illness and this is just
50:49a barrier to a true
50:50biological understanding of psychiatric
50:52diseases as researchers are sort of
50:55discouraged from exploring comorbid
50:56Lee Express symptoms when they may
50:59in fact be meaningful clues that
51:01guide us to distinct ideologies
51:03that call for different treatments.
51:05All these issues route back to the
51:08inadequate understanding that we
51:10currently have of how the brain
51:12even gives rise to behavior.
51:14Super BASIC, but fortunately we
51:16now have the ability to identify
51:18specific neuronal or synaptic targets
51:21and test their role in behavior.
51:23That's what modern neuroscience
51:25technologies will allow,
51:26and perhaps high rates of
51:28comorbidity can be explained by
51:31common neural circuit perturbations.
51:33And then to translate our basic
51:36insights into a better type of therapy,
51:38we will need to apply the knowledge
51:40we gained informed therapeutic
51:42development and take advantage
51:44of we know about plasticity,
51:46resilience and the new adaptations
51:47that occur with stress and
51:49basically focus on a neural circuit
51:51based therapeutic development.
51:53It may be pharmacological,
51:54it may not be.
51:58OK, so regardless of if you're
52:00listening to this talk with,
52:02you know translation here
52:04or basic science here.
52:05I guess I would say from a
52:08basic science perspective.
52:10If we can't understand simple
52:11circuits like those in the amygdala,
52:14and simple questions like,
52:15how do we tell if something is good or bad?
52:19How are we going to understand more complex
52:21circuits and more complex questions and
52:23then from a translational perspective,
52:25the amygdala has been so well
52:27conserved across evolution,
52:29and you know,
52:29a lot of these circuits are still
52:32consistent in species that have
52:34scarcely changed for 70 million years.
52:36This this gives this lens a lot of hope,
52:39that the.
52:40The changes that we see in in mice
52:42will also have relevance to humans,
52:45given the common homology.
52:47OK, so with that I will thank
52:49all the members of my laboratory
52:52and you for your attention.
52:54My collaborators,
52:55my funding sources,
52:56and thanks for listening and I believe
52:59I will take some questions right now.
53:03Thanks.
53:09Thank you very much everyone
53:12for joining us for the talk.
53:14Thank you Kate for a great talk.
53:17I believe that we are going to have to wait
53:22a few moments for Doctor Tide to join us.
53:26She is in California and is on a delay.
53:29I do see that there is one message in
53:33the chat from Doctor Clayton Barnes.
53:36But if you have any other questions,
53:38please add them in.
53:40And Trisha, if you can,
53:42let me know when Doctor Ty joins us.
53:45We will switch to the question
53:47and answer portion of the talk.
53:49And there are a couple of
53:52questions for you in the chat.
53:54If I can read them out to you, I will.
53:58I will start out by has the KAYTIE has
54:00that I lab thought about developing
54:03a dynamical systems model of the
54:05interactions between projection
54:07specific subpopulations in the VLA.
54:12And you're muted.
54:14Thank you, great question.
54:15I think we've approached looking at
54:17predicted dynamical models in other systems.
54:19We have not yet begun to build one
54:21in the DLA, but we absolutely should.
54:23I think you're absolutely right.
54:25Like you know there are these
54:26projections that exist.
54:27There are neurons that have hard wiring,
54:29but you can think of that as the
54:31roads that exist and we want
54:33to think about the traffic.
54:35So coming to that is something we want to do.
54:38We started to do that in the prefrontal
54:40cortex and you know we're able to
54:42predict a lot of different people.
54:44Cortex VLA data is a little bit different.
54:47Not to say that we couldn't do it, but.
54:50We haven't done it yet,
54:51but we we absolutely
54:52should. Thanks Kate,
54:54the next question is about the
54:56recruitment of the central amygdala
54:59projections during rewarding conditions.
55:01It looks like it might be a check
55:04on reward behaviors are there?
55:07Is there evidence of deficits
55:09in central amygdala projections
55:11in unchecked reward behaviors,
55:13like an addiction?
55:15Well, interesting. Interesting, you know.
55:18So first the projections to Central
55:21central itself is super complicated
55:23and I think of it as a ministrator.
55:25There's so many different mirror
55:27peptidergic neurons there,
55:28and we found a lot of different
55:31roles from central amygdala that are,
55:33you know, widely varying from, you know,
55:36aversive behavior to social behavior
55:37and other other components, but.
55:41Or the specific projection from
55:43the PLA to the central limit,
55:45which is predominantly coding for a version,
55:48but it's not exclusively doing so.
55:50We haven't seen that in.
55:52We haven't looked in addiction
55:54to see if it is,
55:56it is specifically modified in terms
55:58of firing rates we have looked.
56:05Look and actually this is something I
56:07didn't talk about in this particular talk.
56:09It's pretty new data,
56:11but with social isolation,
56:12a stress in this can potentiate the minerals
56:14and correlate's with increased drinking.
56:16So when we stimulate the light
56:18inputs to the prefrontal cortex,
56:20we now find that this this is sufficient
56:22to drive alcohol drinking as opposed
56:24to water in a 2 bottle choice.
56:26So that's something, but it's,
56:28you know, still super preliminary
56:29and thinking about this particular
56:31projection from DLA decentral,
56:32we haven't studied that one.
56:35In in the in the context of addiction,
56:38it is sort of this projection.
56:40These neurons have the capacity to sort of.
56:43Ubiquitously inhibit
56:49birthday very broadly, and had,
56:50of course also recruit another local network,
56:52but the local feedforward inhibition
56:54coming from the latest central England
56:56is is much more potent than other
56:58other productions that we've studied.
56:59So I guess I don't know the
57:01answer to that question either.
57:03That's a great question.
57:04We haven't done that specific experiment.
57:07Next question is sort of related to that.
57:10Do you see individual differences among.
57:13Subjects in the balance
57:15between the two circuits.
57:16Have you been able to match that in
57:19any way to susceptible ability versus
57:22resilience for stress disorders?
57:23That's a great question.
57:25That's great, but so we definitely
57:27do see individual variability.
57:29We have not systematically correlated
57:31that with pre-existing behavioral.
57:33Features, but we should do that.
57:35That's a very good expansion
57:37that's I love that idea. The next
57:40question is more clinical,
57:41so I will actually ask John Crystal to
57:44jump in and help or ask other clinicians
57:47in the audience to help with K.
57:49You may have some ideas.
57:51So the question is help me translate
57:53this into clinical practice of patient
57:55presents with generalized anxiety disorder.
57:58How do you know whether the treatment of
58:00choice is pharmacology psychoanalysis,
58:02cognitive therapy, behavioral therapy?
58:03Learning theory,
58:04or a combination of all of these approaches,
58:07with the theories that you put forward
58:10around looking at a circuit analysis help to
58:13make those kinds of decisions I know not now,
58:16but perhaps in the future,
58:18and I will recruit John to
58:20help back you up on this one.
58:23Yeah, it's John.
58:24Do you want to start her?
58:26I
58:26have go ahead, go ahead once you start.
58:29I mean it's so some of this work.
58:32I'm actually referring to colleagues work so.
58:35I'm thinking of Amy Atkin, for example.
58:37We've had a lot of interesting conversy
58:40thinking about different types of anxiety,
58:42different types of you know psychiatric
58:45conditions where you're you're put into
58:47buckets with some symptoms, but that's not.
58:53Basically, with the biological perturbations,
58:55so using resting state,
58:57fMRI is is something that a meeting has
59:00been able to do to sort of predict,
59:02you know what treatment strategy
59:04would be most successful with
59:05this particular individual.
59:07I'm a big proponent of individual
59:09individualized medicine.
59:10I understand us residency.
59:11fMRI might not be, you know.
59:13We need something that
59:15could be more affordable,
59:16more easily applicable to a broader
59:18population that can be rapidly used,
59:19so I understand that's sort of,
59:21you know, a proof of principle,
59:23maybe not the immediate thing that
59:24we would put right into practice,
59:26but that's sort of what I would
59:28love to see happen in the future
59:30of mental health treatment.
59:31Go ahead, John.
59:33Well first K, What an amazing awesome,
59:37fantastic, wonderful, stupendous,
59:38exceptional presentation.
59:39It was just so enlightening.
59:42And and and such a great example
59:45of the way you approach this work.
59:50And and because it was so fantastic,
59:52it's about 10 years of head of
59:55anything that we're able to to do
59:57in the clinic while you were away.
60:00The thought of we,
01:00:01we got into a little discussion about how,
01:00:04how could we even think about
01:00:06developing biomarkers of new retention
01:00:07signaling in the human brain,
01:00:09and which would be kind of the
01:00:11first step in guiding a precision
01:00:13medicine approach.
01:00:14But we're all in with you on
01:00:17on the idea that.
01:00:19That if we we treat illnesses,
01:00:22we treat pathology in other areas
01:00:25of medicine and we treat people in
01:00:28in psychiatry and we need to be able
01:00:32to incorporate an understanding of
01:00:34the pathology that people actually
01:00:36have if we want to get better
01:00:39and more specific treatments.
01:00:41So that's that's clearly a
01:00:43critical step in the pathway.
01:00:45So what a wonderful lecture
01:00:47really appreciate it.
01:00:49Thank you.
01:00:49Yeah,
01:00:50I mean I this also didn't make it
01:00:52into the seminar because it's still.
01:00:59Have collected so recently but we are.
01:01:01We've now started using the
01:01:02tensor from you lonely.
01:01:04You along these lab created.
01:01:05You know all of his tools work so well
01:01:08and he created a new retention sensor
01:01:10and so obviously you know I'm not
01:01:13necessarily a proponent of, you know.
01:01:15I think there's a lot of safety
01:01:17testing that needs to be put in place.
01:01:20You know long term effects of of using
01:01:22viral or genetic tools in humans,
01:01:24even though there's systemic administration
01:01:26potentially possible long term effects.
01:01:28I'm just having a transgene and creating
01:01:30proteins needs to be something that is
01:01:32dealt with a great amount of caution,
01:01:35so I will make that big disclaimer.
01:01:37Circle it underline highlight.
01:01:38However, you know,
01:01:39potentially you could see you know
01:01:42this isn't this is one way that you
01:01:45could be able to to sort of say what
01:01:47is going on in terms of the dynamics
01:01:50and so it offers some some window into
01:01:52potentially using neurotransmitters.
01:01:54Biomarker even this really needs
01:01:55to be stressed out and approved.
01:02:01Away from that, but that's sort of my long
01:02:04term forward looking view. Thanks John.
01:02:07I want to apologize that I haven't been
01:02:09reading out the names of the questioners,
01:02:11so I will try to start reading out the names
01:02:14of the questioners as I asked the questions.
01:02:17I also want to sort of echo Johns
01:02:19phenomenal work. Phenomenal talk.
01:02:20Thank you so much for being with us.
01:02:23Kay Ann. I want to introduce you again
01:02:25to Doctor Flynn's daughter, Sarah Flynn,
01:02:27who's been able to join us today,
01:02:29and it's really been a pleasure to have
01:02:32a representative of the Flynn family.
01:02:34Hi Sarah. Icera it's an honor.
01:02:38So the next question is from Doctor Tech,
01:02:41both the function and structure
01:02:43of amygdala seem to be impaired
01:02:45in subjects with schizophrenia.
01:02:47Have you ever used,
01:02:48or do you plan to use any models
01:02:51of schizophrenia,
01:02:53and in this preclinical work such
01:02:55as phencyclidine administration
01:02:56or some other way of getting it,
01:02:59that that link?
01:03:02I have not. I guess the short answer is no.
01:03:05I have not done that,
01:03:07but I think when I think about schizophrenia
01:03:09and what what is really happening,
01:03:11I guess the the way that I've
01:03:14thought about the most is sort of a
01:03:16computational framework, you know,
01:03:18kind of drawing from like roles
01:03:20and and oppression.
01:03:22With Patricia Coleman,
01:03:23rookies excuse me and thinking about just
01:03:26how do you modulate signal to noise?
01:03:28How do neuromodulators are near peptides?
01:03:31Influence.
01:03:32You know where you draw the
01:03:34line for spike threshold,
01:03:35so you know there's always going
01:03:37to be some noise of inputs.
01:03:38There's going to be some signal,
01:03:40and one way that I think about this is.
01:03:42Pattern completion there's
01:03:43always going to be some noise,
01:03:45and when there's more noise and.
01:03:50No. You know membrane potentials are closer,
01:03:53despite thresholds.
01:03:54Then you might just get spiking
01:03:56and therefore pattern completion,
01:03:57and that's what the brain's
01:03:58really good at doing.
01:04:00This is not just limited Nicola.
01:04:01I'm thinking of amygdala, hippocampus,
01:04:03prefrontal cortex as a connected circuit,
01:04:05but this is something this is sort
01:04:06of the way I've been thinking about
01:04:08how schizophrenia can can come
01:04:10to be in the context of dopamine
01:04:12and other neuropeptides as well.
01:04:14But the signal to noise concept is really.
01:04:17Gotten into the infected,
01:04:18my brain and an I've been thinking
01:04:20about a lot in that terms.
01:04:22It makes a lot of intuitive sense
01:04:24to me how you could get there.
01:04:26So just just how hallucination is
01:04:28constructed is fascinating to me,
01:04:30but we haven't done those experiments.
01:04:33Yeah, that that's an active area of
01:04:36investigation in the in the department.
01:04:39Jane Taylor and Phil Corlette are working
01:04:42on models that really do sort of address
01:04:46some of those behavioral aspects of.
01:04:49Not necessarily pattern completion,
01:04:50but I I like that idea of
01:04:53signal to noise a lot.
01:04:55So John Crystal asks in humans
01:04:57emotions are quite differentiated.
01:04:59So for example fear is
01:05:01different from discussed,
01:05:02which of course we do have discussed
01:05:04circuits in the brain stem of mice that
01:05:08are differentiated from fear circuits.
01:05:10There different from frustration,
01:05:11which of course we can distinguish
01:05:14behaviourally in humans.
01:05:15This higher level differentiation of
01:05:17emotional states seems to involve insula.
01:05:19And have you thought about how
01:05:21the insula might fit in with
01:05:22the story that you told today?
01:05:25That's an excellent question.
01:05:26I am not presently working on the insulin,
01:05:28but my first postdoc from the lab on a
01:05:31Baylor who now has her own lab in Bordeaux,
01:05:34has focused a lot on the
01:05:36insula and amygdala circuitry,
01:05:37so I guess I would encourage you to
01:05:40check out her work, but absolutely,
01:05:42it's in its intricately involved in
01:05:44reciprocally connected with a lot
01:05:46of the different sort of, you know,
01:05:47the emotional triad circuitry, I think.
01:05:49Also Nadine Gogol's work in the insula is.
01:05:56But I haven't done that, but there is
01:05:59beautiful work being done by a number
01:06:02of rising stars in the
01:06:04field. And Mark Anderman also has
01:06:06some market as well. Yes, absolutely.
01:06:08So another question from the chat allcare
01:06:11asks or says behavioral activation
01:06:13is 1 therapeutic form of activation
01:06:16of positive valence neurons using
01:06:18psychotherapy which is used to treat
01:06:20negative balance disorders like depression.
01:06:22And he asks, you might,
01:06:24if you think this might rely on
01:06:27the competition between the BLA.
01:06:29Nucleus incumbents in the BL.
01:06:31A central amygdala circuits and I would add.
01:06:33If so, how might you measure that in
01:06:35a person who is undergoing therapy?
01:06:38Oh, oh,
01:06:38the second part of your question.
01:06:40I'm going to come back to that that in a
01:06:43moment I think this is a great question.
01:06:46It it actually brings us back to Harkins
01:06:48back to a few questions ago regarding
01:06:50clinical treatment, and I think.
01:06:53The reality is. That that.
01:06:58I've been told that pharmacological
01:07:00treatment, in my opinion not specific,
01:07:02and so this is. It's not.
01:07:03It's never going to be maximally
01:07:05effective in my opinion on its own.
01:07:07I think pharmacological treatments are
01:07:09always going to be better in combination with
01:07:11some form of cognitive behavioral therapy,
01:07:13and I'm using that a very
01:07:14very broad term and again,
01:07:16of course I'm not a medical professional
01:07:18and not qualified to give medical advice,
01:07:20but my perception and my understanding
01:07:22of the circuitry would suggest this dual
01:07:24approach would be what is necessary.
01:07:26I do think it's very likely that there's
01:07:28going to be competition between circuits.
01:07:30There's going to be interaction because both.
01:07:33BLANECNBLICE central amygdala
01:07:34neurons are not only rip,
01:07:36typically connected model synaptically
01:07:38with excitatory connections,
01:07:40but also through feedforward connections
01:07:42and that dual connectivity allows a
01:07:45lot of opportunity for regulation of
01:07:47the net strength of connections between
01:07:50those two populations of neurons,
01:07:52and so we do see a lot of plus.
01:08:00I did not object into that study today,
01:08:03but but we have seen that and simple
01:08:06manipulations, food deprivation,
01:08:07social isolation of these things
01:08:09will change the balance of, you know,
01:08:11the the plasticity between these two
01:08:13populations of neurons in their ability to
01:08:16affect each other in a very qualitative
01:08:18way where you know it can become from a
01:08:21suppression to a facilitation based on,
01:08:23for example, food deprivation.
01:08:25And so this is something this is.
01:08:27Gwendolyn calhouns.
01:08:29A study that's that's been
01:08:30bar by archive for awhile.
01:08:31But yeah, I think I think that's
01:08:33kind of how I would imagine
01:08:35this circuitry playing into it,
01:08:36but it's going to be broader than that.
01:08:38Of course, you know,
01:08:39I don't think this is the end all and be
01:08:42all of Dylan's processing by any means.
01:08:44I don't mean to imply that.
01:08:49And a representative of how the brain
01:08:51will apply biological implementational
01:08:52strategies to solve this problem.
01:08:54And I think we'll see that motif
01:08:56appear again and again in different in
01:08:59different circuits as well as the glow.
01:09:05Therefore.
01:09:07That's all the questions from the chat,
01:09:09except for some comments to
01:09:11say brilliant presentation.
01:09:12Does anyone else have any
01:09:13questions they'd like to ask live?
01:09:19Well, if not thank you so much Kay,
01:09:22for being with us with a 3 hour
01:09:25difference in time early early on
01:09:27the West Coast it was a pleasure to
01:09:30be able to hear about your work and
01:09:33particularly to have our audience that
01:09:36involves everything from a molecule
01:09:38to a patient to to be able to stab.
01:09:41Your synthesis has been really helpful,
01:09:43so thanks so much everyone in
01:09:45Marina for hosting and having me.
01:09:50Wonderful God care.