Yale Psychiatry Grand Rounds: "Serotonin and Sociability"

March 22, 2024

Information

March 22, 2024

Aghajanian Lecture: "Serotonin and Sociability" Speaker: Robert Malenka, MD, PhD, Pritzker Professor of Psychiatry and Behavioral Sciences; Director, Nancy Pritzker Laboratory; Deputy Director, Wu Tsai Neurosciences Institute, Stanford University

ID11504

To CiteDCA Citation Guide

00:10OK. Well, thank you for that introduction.
00:13Thank you for the history of the department,
00:16John. I will note it's 1030,
00:20so I'm often accused of talking too long
00:23and going over my allotted time, 10:30.
00:26I will speak for 45 minutes and end at 11:15.
00:30And it's not my fault
00:31that we're going to 11:15.
00:33I also want to thank all of you for
00:35coming and those of you on Zoom,
00:37I particularly want to acknowledge the
00:39presence of Steve Waxman in the audience,
00:41who's many of you know was the
00:43chair of Neurology here for many
00:44years and some of you may not know.
00:46I did my PhD with Steve and Jeff Kosis in
00:49the late 70s and early 80s at Stanford,
00:52and I really owe my scientific career
00:54to them because I entered their lab
00:57as a first year medical student with
01:00essentially no scientific experience.
01:02I have no idea why they took me,
01:04but I'm very happy that they did OK
01:08and here are my acknowledgements.
01:11I'm an advisor to a few companies.
01:14I'm also currently on leave from Stanford
01:16to be the Chief Scientific Officer at a
01:18place called Bayshore Global Management,
01:20which is the family office in Palo Alto
01:23of a very wealthy Bay Area individual,
01:26which I'm happy to talk about in private,
01:28if any of you are interested, OK.
01:31We actually heard about George
01:33Agajanian and his contributions
01:36to psychiatric neuroscience.
01:38We heard that from John that
01:40he was a pioneer in the 60s,
01:43actually studying psychedelics
01:45in the 60s and 70s,
01:47studying serotonin neuromodulation.
01:49So he was prescient in that degree
01:54in the sense that we all know.
01:56There's a renaissance now of trying
01:58to understand both the mechanisms of
02:00psychedelics and their therapeutic potential.
02:03He also was a pioneer with Steve
02:05Bunny in studying dopamine neurons.
02:08And I want to shout out to Steve
02:11because I vividly remember in the
02:1370s when I was trying to figure out
02:15what I wanted to do with my life
02:17and I was beginning to think
02:19about neuroscience as a career.
02:22This is back in the days when
02:24there was no Internet.
02:25You had to actually go to the library.
02:26That's a place for the younger generation
02:29where there are books and they used to be
02:31journals because you couldn't go online.
02:33And I would,
02:34just because for a variety of reasons,
02:36I would skip class,
02:37go to the library and I started
02:39reading and there was a book.
02:42I think it was called Biological Psychiatry.
02:45Or it's the series of review papers.
02:47And I vividly remember reading a
02:49review paper from Bunny and Agajanian
02:51about recording single unit activity.
02:53I believe it was dopamine neurons
02:55and just thinking.
02:56And then I answer for easing different
02:58types of substances and drugs.
03:00And I was thinking, man, this is really cool,
03:02maybe I should do that someday.
03:04So again, thank you for your,
03:06your pioneering work.
03:08OK.
03:09And of course it's an honor
03:11to give an Agajanian
03:13lecture given his his contributions
03:15to the field. As Al alluded to,
03:17some of you who are probably older
03:20than 40 may know that I started my
03:23career studying the mechanisms of NMDA,
03:26dependent LTP and Ltd in the hippocampus.
03:29I say that because it's remarkable to me.
03:31I get students and postdocs now who have no
03:34idea that I actually worked in that field.
03:37Even though I'm going to brag,
03:39I think some of our findings are in
03:42textbooks now and my approach to LTP
03:44and Ltd was the assumption behind trying
03:46to understand the mechanisms of LTP and
03:48Ltd was a very simplifying assumption.
03:51It was that the BI directional
03:53control of synaptic strength,
03:54that is synaptic plasticity,
03:56must be fundamentally important for
03:59all forms of adaptive and pathological
04:02experience dependent plasticity.
04:04And therefore,
04:05it's probably important to delve into
04:07those underlying molecular mechanisms,
04:10because what we find hopefully
04:12will be important, as I said,
04:14for all sorts of important adaptive and
04:17pathological functions of the brain.
04:19In the probably in the late 80s,
04:23early 90s, I started thinking about
04:25other topics I wanted to approach,
04:27I wanted to pursue,
04:29especially as I established my
04:31independent lab and I needed to
04:33individuate from my work with Roger Nickel.
04:36And I started thinking about this topic,
04:40which is Thorndike's Law of effect,
04:41which is a very simple explanation,
04:45I would argue, of all of behavior.
04:48And the explanation you
04:49can just read it here,
04:50is very simple.
04:51We repeat behaviors that lead to a
04:54reward or positive reinforcement,
04:57and we avoid behaviors that lead to some
05:00form of punishment or aversive experience.
05:03And I would argue when all is said and done,
05:06a lot of animals behaviors and a lot
05:08of our behaviors can be explained
05:10by this very simple principle.
05:12And so if you accept that
05:15simplifying assumption,
05:15a very important question is what circuits
05:17in the brain mediate reward and aversion.
05:20And I was particularly interested
05:22in plasticity in these circuits for
05:25the hopefully the obvious reason
05:27that external stimuli,
05:28even internal stimuli that your your
05:31your effective response to those,
05:33whether you experience them as positive
05:36or negative, rewarding or aversive,
05:38is highly plastic.
05:41And I I'll you know,
05:42the joke I say is any of any of you who
05:45have been in an intimate relationship
05:47know exactly what I'm talking about.
05:49The the love of your life can be
05:50the love of your life and 30 seconds
05:53later something will happen and
05:55they're the bane of your existence.
05:56And as a neuroscientist,
05:58I believe those changes in your
06:00effective response to these external
06:03stimuli doesn't happen magically.
06:05It has to be because neural
06:07activity in your brain has changed,
06:09and it very likely is because the
06:12circuits that tell you what is rewarding
06:14and aversive are modifying their
06:16ensemble activity in interesting ways.
06:18So where do you look if you want
06:21to start studying the circuits that
06:23are contribute to the experience
06:25of rewarding or reversive stimuli.
06:27And obvious place to start is the
06:30classic mesolimbic reward circuitry that
06:32I'm sure all of you are familiar with,
06:35consisting of dopamine neurons in the VTA,
06:37projecting to the ventral striatum or
06:40nucleus accumbens, etcetera, etcetera.
06:41So many years ago I'm actually oh God,
06:46when was this? In the 90s,
06:50people in my lab started studying
06:52dopamine modulation of nucleus
06:53succumbing synaptic transmission.
06:55We started studying drugs of abuse
06:58as a model for very powerful
07:01reinforcing external stimuli that
07:03caused synaptic and circuit changes
07:05in this mesolympic reward circuitry.
07:07We thought about looking at feeding behavior.
07:10We actually did a lot of work that we
07:13never published on peptides and how
07:15they might influence reward circuitry.
07:18And most importantly,
07:19for the purposes of this talk,
07:20beginning literally 14 or 15 years ago,
07:24I started thinking about what are other
07:27interesting experiences that might
07:30impact this classic reward circuitry.
07:33And an obvious one is social interactions.
07:36The simplifying idea being that pro
07:39social non aggressive interactions
07:41for most of us are highly reinforcing.
07:45And if that is the case,
07:47then it must be the case that this
07:49circuit that has been evolutionarily
07:51conserved to tell us what is rewarding
07:54and aversive, must be modulated,
07:55must play a role in the positive or
07:58negative experiences we have when
08:01we socially interact with other
08:03members of our species.
08:04And then again,
08:05I think it's obvious for those of us
08:08interested in psychiatry that there
08:10are a lot of pathological conditions,
08:13neuropsychiatric disorders,
08:15where there are maladaptive or
08:19abnormal social interactions,
08:20most obviously some of the major
08:23symptoms of autism spectrum disorder.
08:26So what I'm going to do today is
08:28take you on a whirlwind tour of
08:30a series of published papers.
08:32Because these are all published,
08:34I'm going to just give you some of the
08:36highlights from each of these papers,
08:37beginning with this work that really
08:39began 14 years ago in my lab,
08:41done by Gul Dolan when she was a postdoc.
08:44And I'll just give you one slide of that.
08:46Then I'll take you on a whirlwind tour.
08:48I'm already behind schedule,
08:49ending up in some unpublished work
08:52that actually just got accepted.
08:53So it should be coming out in the
08:55next month talking about the effects
08:57of MDMA on empathy like behaviors.
08:59OK,
09:00so the story begins with this work
09:02that Google Dolan did in my lab
09:05again beginning in 2010 or so,
09:08where the question was does reward circuitry,
09:11in this case in particular
09:13the nucleus accumbens.
09:14And I'm assuming this audience
09:16all knows these this terminology.
09:18If you don't,
09:19you're going to be a little lost,
09:20and Google wanted to ask was
09:23actually asking the question of
09:25whether oxytocin A neuropeptide.
09:27Many of you are aware of that
09:29is well established to play an
09:31important role in a variety of social
09:33behaviors and maternal behaviors.
09:34Whether it's actions in the nucleus
09:37accumbens might be important
09:39for what we back then defined as
09:43the reinforcing component of a
09:44same sex non aggressive social interaction,
09:48and if there's a lot of work behind this.
09:50But all of her data were consistent with
09:53a somewhat complicated hypothesis that
09:56during a same sex non aggressive what
09:59I'm going to call pro social interaction,
10:02that oxytocin is released in the incumbents
10:04and the way it influences ensemble activity
10:07in the incumbents is by causing the
10:10release of serotonin that magically is
10:13required in ways we still don't understand.
10:16And we're trying to figure out that that
10:19release of serotonin is very important for
10:22mediating or contributing to the rewarding
10:24effect of a positive pro social interaction.
10:27So back then,
10:28when this work was published in 2013,
10:31we reasoned that if this
10:33hypothesis is correct,
10:35that is oxytocin is released,
10:36but it's actually working to 'cause
10:40the release of serotonin from serotonin
10:42inputs into the nucleus secumbens,
10:45Then it must be the case that serotonin
10:47released in the nucleus secumbens
10:50should influence and perhaps enhance
10:52what I'm going to define as pro
10:55social interactions or enhance what
10:57I'm going to define as sociability.
11:00And again,
11:01I was trained by Roger Nicole mostly.
11:03And you know what I was trained
11:05is if there's an experiment that
11:07tests your hypothesis,
11:08that can prove you wrong,
11:10you are duty bound to do that experiment,
11:12that what good science is,
11:13is constantly trying to prove yourself
11:16wrong and that eventually you gain it.
11:18You know you can't.
11:19If you can't do that after 5
11:21or 10 or 15 years of work,
11:23then maybe your original
11:24hypothesis was right.
11:26OK, so hopefully the next few slides
11:28are all going to be very obvious.
11:30This was published several years ago.
11:32The major input into the nucleus accumbens
11:35comes from the dorsal Rath, a nucleus.
11:37Because the results that
11:39I'm going to present to you,
11:41when they first came into my lab,
11:42I was kind of skeptical because
11:44they are so robust.
11:45And, you know,
11:46when you're doing behavior in mice,
11:48it's hard to get robust behavioral responses.
11:51I was pretty nervous about them.
11:53But I would say that at least
11:55my postdocs tell me all of the
11:57experiments I'm presenting you today
11:58are performed and analyzed blindly,
12:00not knowing whether you were doing an
12:03active manipulation or placebo manipulation.
12:06So again,
12:07the hypothesis is that release of
12:09serotonin in the nucleus accumbens
12:12promotes pro social behaviors is
12:14necessary for the reinforcing
12:17component of a social interaction.
12:20So what Jess Walsh did,
12:21who was a postdoc in my lab,
12:23she now has her own lab at the
12:24University of North Carolina,
12:26is to the obvious experiment.
12:28We got a cert.
12:29Cree mouse,
12:29that's a mouse where Cree recombinase
12:32is expressed in dorsal Rath,
12:33a serotonergic neuron.
12:35So we can use Cree dependent expression
12:38of trans genes such as optogenetic tools.
12:41In this case,
12:42she expressed channel rhodopsin
12:44and dorsal Rath A.
12:45She implanted the light pipe in
12:47the nucleus accumbens so she could
12:50activate or inhibit serotonergic
12:51inputs in the nucleus
12:53accumbens. And then she did
12:55a series of very simple,
12:57rudimentary social behavior, assays.
12:59One is called the juvenile interaction assay.
13:03It's very simple.
13:04You take your subject mouse and you bring
13:07into its home cage a little juvenile.
13:10You use a juvenile so there's
13:12less likelihood of an aggressive
13:13interaction and you just ask your
13:15subject mouse how much time do you
13:17want to hang out with your little
13:18buddy in a non aggressive fashion.
13:20She also did this, the classic three
13:23chamber social preference assay,
13:25which is what it is.
13:27That's where you have you put your
13:29mouse in a three chamber house
13:31you have two cups under 1 cup.
13:33There's another little mouse,
13:35a juvenile, and the other cup,
13:37it's either empty or there's
13:38an inanimate object.
13:39We like to use a toy mouse and
13:41you just ask your subject mouse,
13:43do you want to hang out with your
13:45little buddy or do you want to hang
13:47out with your inanimate object more?
13:49And then you're going to see
13:50a lot of data like this.
13:51You're going to see the individual
13:53animals in the left grass.
13:54You're going to see a summary and
13:57hopefully and again we always did the
13:59appropriate controls and hopefully
14:00the result is pretty apparent.
14:02If you activate serotonin inputs
14:04optogenetically in the nucleus
14:06accumbens during these assays,
14:08you get a very robust and
14:11very reliable enhancement of,
14:13I'm going to call sociability.
14:16And remarkably,
14:17if you inhibit the serotonergic inputs,
14:20in this case using Halorhodopsin,
14:22you get the opposite effect.
14:27We did a bunch of controls.
14:29These manipulations,
14:30activation or inhibit inhibition of
14:32serotonergic inputs in the incumbents did
14:35not affect novel objects, interactions,
14:37did not affect locomotor activity.
14:40And most importantly,
14:41I want to really emphasize this because
14:44I for me I think there's an enormous clue
14:46in this finding that unlike dopamine,
14:49So when you release dopamine and then the
14:51ventral striatum or nucleus succumbing,
14:52whether you do that with a drug of abuse
14:55like cocaine or amphetamine or morphine,
14:57whether you do that optogenetically,
14:59I think one of the most reliable
15:01behavioral findings in this field is
15:03that the release of dopamine at least
15:06in certain sub regions of the accumbens,
15:08I have to be careful here.
15:09It's highly reinforcing.
15:11The animals rules display condition,
15:13place preference.
15:14They will nose poke or press a
15:16bar in order to get that dopamine
15:19releasing the incumbents.
15:20What this shows,
15:21and we've done this over and over again,
15:23is that when you do the same
15:26manipulation with serotonin,
15:27it is not acutely reinforcing and
15:30just think about that dopamine release
15:32the incumbents, the animals love it.
15:34Serotonin release at least using
15:36those kinds of behavioral assays.
15:38It's a nerd.
15:39So this is just showing doing a
15:41condition place preference assay that
15:44optogenetic activation of serotonin
15:46inputs in the incumbents does not cause CPP.
15:49But in the same animals,
15:51it caused a very robust increase
15:54in juvenile interaction,
15:56showing that in the same animals it wasn't
15:58because there was some technical problem.
16:00OK,
16:01what does this have to do with autism?
16:04So hopefully it's apparent that
16:06if you're looking at effects of
16:09neuromodulators on social interactions,
16:12one of the most obvious neuropsychiatric
16:15disorders or syndromes that you
16:18might want to see if your findings
16:20are relevant to is autism.
16:22And, you know, I, I, I won't make my normal,
16:26whatever sarcastic comments
16:28about biological psychiatry.
16:29But in autism,
16:32there's a lot of evidence that
16:35serotonin systems might be involved,
16:37for a variety of reasons.
16:39We started by studying a mouse model of
16:43chromosomal copy number variation known
16:46as the 16 P 11.2 deletion syndrome.
16:49My genetic colleagues like Matt State,
16:52who used to be here at Yale and
16:53others at the Simons Foundation,
16:55tell me it makes up about 1% of the
16:59cases of autism spectrum disorder.
17:02The truth is,
17:03we chose this mouse model for two reasons.
17:06And I'm saying this for the
17:08trainees in the office audience.
17:10One is, it was available at Stanford,
17:13so it's easy to get two.
17:15As you're going to see.
17:16We had a flocked version of it.
17:18We had a conditional knockout
17:20of this syndrome 3A major
17:23Funding Agency in the field,
17:24the Simons Foundation, that is Safari.
17:27When I asked them which mouse
17:28model they wanted me to study,
17:30they suggested this one.
17:31So the combination of it being available,
17:33there being a conditional knockout,
17:35and the Funding Agency I was hoping to
17:37get money from told me to study this one.
17:39OK,
17:40we'll study this one and the
17:43details aren't important.
17:44As I said, we had a conditional knockout
17:47of this chromosome the the syntenic
17:49chromosomal region in the mouse that
17:51was made by Ricardo Dulmich's lab,
17:53who at the time was at Stanford.
17:55So what that allowed us to do is to ask
17:58the question of if we genetically delete
18:02this chromosomal segment only from dorsal
18:04rafta serotonin neurons do we see a
18:07sociability deficit in these two assays.
18:10And what these graphs show you is that yes,
18:12indeed we do.
18:13So we we we deleted this chromosomal
18:16segment in two different ways.
18:18One is we injected a virus expressing
18:22Cree into the Dorserathea nucleus.
18:26The other way is we cross this
18:29conditional knockout of the 16 P
18:3111.2 with a cert Cree driver line.
18:34So in one, we're just deleting it
18:36from cells in the dorsal Raphae.
18:38We can't specifically say
18:40they're serotonin neurons.
18:41The other manipulation,
18:42we're deleting it from all
18:45serotonergic neurons in the brain.
18:46The bottom line is if you look at these
18:49two graphs or these two bar graphs,
18:51that manipulation caused a
18:53pretty pronounced impairment of
18:55these two assays of sociability,
18:58the juvenile interaction and
19:00three chamber place preference.
19:02We then just did some simple slice
19:05electrophysiology where we deleted
19:07this chromosomal segment from dorsal
19:10Rafa serotonin neurons while also
19:13expressing GFP in the serotonin neuron.
19:17So then we could do patching
19:19from the serotonin neurons in the
19:22dorsal Rafa knowing that we had
19:24genetically deleted this segment.
19:25And again,
19:26all this shows is that the deletion
19:28of this chromosomal segment had
19:32electrophysiological effects.
19:33It decreased the excitability of the neurons.
19:36That's what's shown here.
19:37And then it decreased the amplitude of
19:40spontaneous miniature synaptic currents.
19:43We didn't pursue this anymore.
19:45There's a lot of obvious questions.
19:46I used to make a living doing this,
19:49but we just wanted to be sure that
19:51something was happening in these neurons
19:54when we did this genetic manipulation.
19:56Perhaps more importantly,
19:58this was our first attempt.
20:00So I mean,
20:01I think later I'll show you better
20:03fiber photometry experiments,
20:05but in this experiment what we did
20:07is we again genetically deleted
20:09this chromosomal segment from
20:11dorsal Raffy serotonin neurons.
20:13We expressed G camp in the
20:16dorsal Rafa serotonin neurons.
20:18We then did what's known as fibre photometry,
20:20which is where you're measuring
20:22the bulk fluorescence signal from
20:25the cells expressing G camp as a
20:27surrogate for neural activity.
20:29And all this shows you is that
20:32during a social interaction when the
20:35subject mouse approaches a little
20:37juvenile in the control animals,
20:40you get a pretty robust increase
20:42that begins a few seconds before
20:44the the social interaction happens.
20:47And you know the details I can
20:49talk about if you're interested in.
20:51But in the knockout animals,
20:53in the animals that show a sociability
20:56deficit,
20:56the magnitude of that increase
20:59is clearly reduced.
21:00So all of this is consistent with the
21:03idea that this genetic manipulation of
21:06an autism of a segment of a chromosome
21:09that's implicated in autism that
21:12causes sociability deficits in the mice
21:14does cause changes in the function of
21:17these dorsal raphae serotonin neurons.
21:19So that obviously raises the question of
21:22can we rescue the sociability deficits by
21:25activating the serotonin inputs in these
21:28mice that have a sociability deficit?
21:30And what this shows you is that we get
21:33a very robust rescue of the sociability
21:36deficits when we activate serotonin
21:38inputs in the nucleus accumbens in
21:40the mice in which we've genetically
21:43deleted this chromosomal segment
21:45from the serotonin neurons.
21:47And I know I'm going fast,
21:48but this has been published five or six year,
21:50whatever it is 6 years ago.
21:53So the obvious next question is if,
21:56if serotonin release in the incumbents
21:59is rescuing the sociability deficits.
22:01You know there's 16 different
22:03serotonin receptors,
22:04a bunch of them are expressed
22:05in the incumbents,
22:06which ones might be necessary?
22:08I know I'm going fast because I have a
22:10lot to cover in the next 25 minutes.
22:13What this shows is that if we in these
22:16mice that have a sociability deficit,
22:19if we activate now the somas of the
22:23serotonin neurons and we infuse
22:26either saline or serotonin 1B receptor
22:29antagonist into the nucleus accumbens.
22:31So the black bars are the control animals
22:34where we're injecting saline and we
22:36get this is a different set of animals.
22:39And again you can see how robust
22:41the effect is.
22:42I mean it's actually crazy robust.
22:44And this effect was completely
22:46blocked by infusion of A1B antagonist
22:49into the nucleus accumbens,
22:52suggesting that obviously that one
22:54serotonin 1B receptors in the accumbens
22:56are at least necessary for this
22:59rescue of the sociability deficits.
23:01So that led to Jess doing an experiment
23:03I told her would never work,
23:05but luckily she didn't listen to me
23:07where where she asked the question
23:09of whether,
23:10if she infused A1 serotonin 1B receptor
23:13agonist directly into the incumbents,
23:16could she rescue the sociability deficits.
23:19And what these green bars show is that,
23:21yes, indeed that was the case.
23:25So then subsequently we pursued this
23:27in a paper that was published 2 year
23:29three years ago in your cycle farm,
23:32where does this effect of serotonin
23:351B receptor agonist?
23:37Does it generalize to other mouse
23:40models of autism? And in fact it does.
23:43What this shows is three additional
23:45mouse models of autism.
23:47Again, the details aren't important.
23:49We actually studied 5 different
23:51genetic mouse models of autism too.
23:54I just showed you the 16 P 11.2 deletion.
23:58These are two other genetic models.
24:00We also did a model called the valproic
24:03acid model and again the the take
24:05home message is these red bars and
24:08all of these mouse models of autism
24:10that show sociability deficits.
24:13In this case giving A1B agonist
24:16parenterally IP rescued their
24:19sociability deficits.
24:20So this is the summary of this first
24:23part of the talk I'll publish that bi
24:26directional modulation of serotonin
24:27releasing the incumbents bi directionally
24:30influences pro social behaviors.
24:32We can delete this chromosomal
24:34segment only from dorserath A
24:36serotonin neurons and
24:38influence their sociability,
24:40behavior and influence on their excitability.
24:44And we can rescue these sociability
24:47deficits by in activating serotonin
24:50inputs or giving the animals a serotonin,
24:53one being agonist.
24:54So what does this have to do with MDMA?
24:57So around the time we were
24:59finishing these experiments,
25:00Boris Heifetz joined my lab.
25:02Boris is an MDPHD.
25:04Actually, he's the only anaesthesiologist
25:06in the world that I know who
25:08should have been a psychiatrist,
25:09but I think he just wanted to make
25:11more money than psychiatrists make.
25:13So he became.
25:14But man, he everything he does
25:16is related to psychiatry.
25:17So he joined my lab as a
25:20postdoc after finishing his
25:22anesthesia residency at Stanford.
25:25He actually came to my lab because his
25:27training was with Pablo Castillo at Einstein.
25:30Pablo was an ex postdoc of mine,
25:31so he came with from a slice
25:33of Physiology background.
25:34Long story short,
25:35we figured at somehow we started
25:38talking about all sorts of topics
25:40and it turned out we both had
25:43a real interest in using drugs
25:45as probes of brain function.
25:48We both had a very strong interest in MDMA,
25:51hopefully for obvious reasons.
25:53As most of you know, MDMA,
25:55also known as Ecstasy or Molly,
25:58has been used by human beings now for
26:00whatever it is, well for decades.
26:02And as many of you know,
26:04it has very powerful pro social effects.
26:08So there's a clue there.
26:10How can a drug that has molecular
26:13targets have this profound
26:14effect on social interactions?
26:17So we were real in in the advantage of
26:20using drugs as powerful probes of brain
26:22function is they have molecular targets.
26:25So if you're doing this in mice,
26:26you can use all the modern
26:29tools of genetic neurobiology,
26:30of optogenetics and chemogenetics
26:32to really figure out which molecular
26:34targets are most importantly,
26:36which cell,
26:37which circuits and cell types within
26:39those circuits are responsible for
26:41the behavioral effects of those drugs.
26:43And then depending on the
26:45drug you're studying,
26:46you can give it to human beings and
26:48study their effects on human beings.
26:49So I really think drugs are perhaps
26:53underutilized probes of important
26:55behavioral functions in in human beings.
26:59And then obviously hopefully
27:00as many of you know,
27:02we were also interested in
27:05MDMA because due to the due to
27:08the pioneering work of MAPS,
27:09it's been extensively studied in
27:11phase three trials as an adjunct
27:13to psychotherapy for PTSD.
27:15Most of us are anticipating the
27:17FDA will approve its use and
27:20then mechanistically I,
27:21you know it influences the
27:24three neuromodulators,
27:25I'm most interested in serotonin,
27:27dopamine and oxytocin.
27:29So it just,
27:31you know,
27:31it was obvious that we should start
27:34studying this and I should say I
27:36went public with my interest in 2016
27:38and it is unusual for a journal like Cell,
27:42which many of you know is kind of a
27:44******** cellular molecular journal.
27:46It's pretty unusual for them
27:48to publish a commentary about
27:50I don't like to call
27:52MDMAA psychedelic,
27:53although I will use that term.
27:55It's actually an intactogen and
27:56I'm going to give you evidence that
27:58it might even be an empathogen.
28:00But nevertheless, it just shows
28:01you how far the field has come.
28:04OK, so again,
28:05this was published several years ago.
28:06I'm going to go through it pretty rapidly.
28:10This is all work that Boris did with
28:12some others in my lab as a postdoc.
28:14So the first thing he did he
28:16had to do was develop an assay
28:20that would demonstrate that MDMA
28:23has some effect he could study.
28:25So he decided to use the three
28:28chamber social preference assay
28:29that's shown here, here.
28:30This is going to become important.
28:32He did a dose response and you can
28:35see 7 1/2 milligrams per kilogram
28:37of MDMA had a pretty robust effect
28:40on enhancing social preference in
28:42the three chamber social preference
28:44assay as did 15 migs per kig.
28:48This just shows the time
28:50course of the effect.
28:51This is one of my favorite experiments
28:54where excuse me for a second he
29:00if you just look at the graph down here,
29:03if he gives the MDMA only to the animal
29:05under the cup in the three chamber assay,
29:08he he sees an effect.
29:10It's not statistically significant,
29:11but it looks like an effect.
29:12If he gives it to the animal,
29:14this like an experimental
29:15animal that's free to explore.
29:17He sees a more robust effect.
29:19If he gives it to both animals,
29:20he really sees a robust effect.
29:23So I just think that's cute.
29:24For those of you who have
29:25taken this drug at raves,
29:26you don't have to raise your
29:28hand if you've done this,
29:30you know it's cone of silence here.
29:32OK, so we were particularly
29:35interested in finding a dose of
29:37MDMA that still had this in quotes.
29:40Pro social effect,
29:41but did not have the classic effects of a
29:45amphetamine derivative of a psychostimulant.
29:48And as most of you,
29:49many of you know classic psychostimulants
29:52like Dexter or methamphetamine or cocaine.
29:55They do two things to mice that
29:57are very robust and very reliable.
29:59They increase the locomotion
30:00of the of the animal and they
30:02cause condition place preference
30:04because they're reinforcing.
30:06So what this shows you is that at 7
30:081/2 Migs per kig the the drug that I
30:11just showed you had this pro social
30:13effect doesn't cause locomotion
30:15and this shows that it doesn't
30:17cause condition place preference
30:18where as a higher dose does.
30:20So we're going to take advantage of
30:22this dose response and then we're going
30:23to get into a little bit of mechanism.
30:25The major molecular targets of MDMA
30:28are the serotonin transporter or CERT
30:31and with lower affinity it interacts
30:33with the dopamine transporter or DAT.
30:36So we are given our work on serotonin.
30:39We are obviously our hypothesis
30:41was that the pro social effects
30:43of MDMA are primarily mediated via
30:46its interaction with CERT with
30:48the serotonin transporter.
30:50So we began to test that.
30:51So in this experiment,
30:53the first thing Boris did is he gave
30:56animals either saline or citalopram,
30:58which binds to the serotonin transporter
31:01and in a different way than MDMA.
31:04And I'm happy to address
31:06why that's important.
31:07But it blocks the ability of MDMA
31:09to interact with serotonin transport
31:10and you can see that completely
31:13blocked the effects of MDMA in
31:15the social preference asset.
31:16We then got a conditional knockout
31:18of the serotonin transporter.
31:20It was brutal to get it, but we got it.
31:23And then Boris injected with at the time.
31:26I forget why he didn't cross
31:28it with dessert cream house.
31:29So what he did is he injected A
31:31Cree virus into the dorseraft A so
31:34he could eliminate the serotonin
31:36transporter from serotonin neurons
31:38in the Dorseraf A.
31:40And what this shows is it's noisy
31:42data I admit,
31:43but on average MDMA was blocked
31:47the in these animals lack even with
31:51a reduced expression or lacking
31:53the serotonin transporter,
31:55it blocked the effects of
31:57MDMA on social preference.
31:59And most importantly that
32:00manipulation did not block the effects
32:03of MDMA at a higher dose
32:05unconditioned place preference,
32:06which is consistent with the idea that
32:09the place preference effects of MDMA are
32:11being mediated by release of dopamine.
32:13Although we don't have proof,
32:14well actually I'll show you one
32:16experiment we did to show that.
32:18So you know, the hypothesis is that
32:22it's MDMA acting on the serotonin
32:25transporter that is necessary for its
32:28effects on our sociability assays.
32:30Is this happening in the nucleus secumbens?
32:32So we did old fashioned behavioral
32:35pharmacology, which I'm still a big fan of.
32:37You know, there's so I'm going to
32:38make a lot of editorial comments.
32:40There's this almost hysterical
32:42emphasis on innovation on new stuff,
32:46and that's obviously important.
32:48But what's more important is doing the
32:50right experiment that test your hypothesis.
32:52And if it's an old fashioned methodology,
32:55who cares?
32:56If it's the right experiment, do it.
32:58And I wish NIH would accept this rather
33:01than always looking for innovation,
33:03innovation, just do the right experiment,
33:06do good science.
33:07That's what's important.
33:08So for us,
33:09the good science was old fashioned
33:11behavioral pharmacology.
33:13So if MDMA is primarily causing
33:15release of serotonin to promote
33:17affect these sociability assays,
33:19we should be able to infuse
33:21MDMA directly into the nucleus
33:23accumbens and get the same effect.
33:25That's what's shown here.
33:26And in fact,
33:27you'll see that the effect of MDMA
33:29infusion in the accumbens is actually more,
33:32you know, pretty robust.
33:34We did the converse experiment where
33:37we gave MDMAIP and we infused a
33:40this drug citalopram that prevents
33:42MDMA from binding to the serotonin
33:44transporter and we infused that
33:46directly into the nucleus secumbens
33:48and on average it blocked the effects
33:51of MDMA And most importantly the same
33:54manipulation in the same animals.
33:57That is,
33:58administration of citalopram did
34:00not block the effects of MDMA
34:03on condition place preference,
34:05whereas A dopamine receptor antagonist.
34:09I I'm embarrassed,
34:10I can't remember if this was done
34:12in intra accumbens infusion or IP.
34:14Nevertheless I think it was into
34:16the accumbens.
34:17Nevertheless,
34:17that did on average block the the
34:21rewarding component of MDM as effects
34:23as assayed by condition place preference.
34:26So does MDMA cause the release of serotonin.
34:30Subsequently,
34:30after this paper was published,
34:33we finally got a genetically encoded
34:35sensor of serotonin that we've done
34:37a lot of work with grab serotonin,
34:39which many of you are familiar with.
34:41We expressed it in the nucleus accumbens.
34:43We did fiber photometry.
34:45And just as we expected,
34:47if you give MDMAIP and you measure
34:49serotonin release use and grab serotonin,
34:52the incumbents,
34:52you get a very you get kind of a
34:55humongous increase in serotonin.
34:56And
34:59then what else did we do?
35:00Oh yeah, then we also studied MDMI,
35:04Can't see my we also studied MDMA in a
35:07number of different mouse models of autism.
35:10And you can see MDMA caused a very
35:13robust enhancement of sociability
35:15in these mouse models of autism.
35:17Here's three different ones.
35:20We then we also did some work
35:23looking at the serotonin receptors
35:25in the accumbens mediating or
35:28necessary for the effects of MDMA.
35:31And consistent with the previous
35:33work I showed you, although I will
35:35tell you at the end of my talk,
35:36it's not the whole story.
35:38A serotonin 1B antagonist,
35:40Nas 181 infused into the incumbents
35:43blocks the effects of MDMA when given
35:47parenterally for a variety of reasons.
35:49We also started testing the enantiomers
35:51of MDMA and I can explain why
35:53again in the question and answer.
35:55But the MDMA that is being given
35:57to human beings that is mostly
35:59on used being used illegally
36:01recreationally is a mixture of R,
36:04the RNA and tumor and the S&amp;N tumor.
36:06And there's some evidence that the
36:08RNS enant tumors have different
36:10affinities for the key molecular
36:12target certain that and then we were
36:15interested in whether the R enantiomer
36:17would still have the effects in
36:19these simple assays we have done.
36:21And what this shows is that RMDMA and a
36:23very small number of animals did enhance
36:26social preference in the three chamber assay,
36:28but at the dose we use
36:30did not 'cause condition,
36:31place preference.
36:32And then this is just a,
36:35you know, why do we do this work?
36:39Well in in the last 10 years of my career,
36:43maybe three years,
36:44maybe 15 years,
36:45however long I last,
36:47I've decided that rather than staying
36:50in the background and criticizing
36:54the lack of progress in academic
36:56psychiatry and biological psychiatry,
36:58I should actually do something about it.
37:00So I would argue one way of doing something
37:03about it is the type of work I'm doing,
37:05which I would like to believe is
37:07reasonably translational and see if
37:09any of the findings we're getting
37:11with drugs in mice actually might have
37:14therapeutic benefit in human beings.
37:16And I'm pleased to say two companies
37:18I advise MAP Light is due,
37:20is in the midst of a phase two
37:22trial of the serotonin 1D agonist
37:24and autism spectrum disorder.
37:26Truth be told,
37:27I'm pessimistic it's going to work
37:29because we may need other serotonin
37:31receptors to be activated in order
37:33to have a therapeutic effect.
37:35We didn't know that at the time,
37:37and another company that I advised
37:39based on my suggestion is going
37:41to be pursuing our MDMA as a
37:43potential treatment in individuals
37:45with autism Spectrum disorder.
37:47Whether these work or not, who knows.
37:49I mean, it's shots on goal,
37:51but at least there's a
37:53rationale for pursuing them.
37:54So what does this have to
37:55do with empathy in mice?
37:57And so in the last 10 minutes,
37:587 minutes,
37:59I want to tell you about some work
38:01that I mean this is a far cry and
38:03I'm name dropping now from working
38:05with Rodger Nickel and Tom Sutoff,
38:08two of the hardest core molecular scientist
38:12I know who I worked with individually,
38:15my God, for 25 years.
38:18And I think they would be rolling
38:19their eyes if they could hear
38:21me talk about empathy in mice.
38:22But nevertheless, I think it's an
38:27important topic because I would argue,
38:30given the state of our world,
38:31not only in our country,
38:33but I mean it's kind of obvious when you
38:36look at the world, what are we lacking?
38:39We're lacking empathy and compassion.
38:41We look at other people's in our
38:43societies based on the color of the skin,
38:45based on their religion,
38:46and we see them as different.
38:48Whereas, in fact,
38:49we all know as neurobiologist,
38:52as biologists, that's kind of ridiculous.
38:55And I would argue,
38:56I actually really passionately believe this.
38:58What is more important for neuroscientists
39:00to be studying than neural mechanisms
39:03of empathy and perhaps even compassion?
39:06Because maybe if we understood the
39:09neurobiological underpinnings of these
39:11important human these important phenomena,
39:14maybe we could get the politicians and
39:17certain other ones to pay attention.
39:19At the very least,
39:20maybe we can develop drugs and therapies
39:23that will enhance that and put it in
39:25the water of our political leader.
39:27I've actually thought about this Freshman.
39:29I actually have thought about this.
39:31OK, so how do you study empathy in mice?
39:34I use the term.
39:35I have 5 minutes left.
39:37I use the term loosely.
39:39When when we published this paper,
39:43the major criticism we got,
39:44it was the one of the easiest
39:46papers I've ever published.
39:47I was amazed.
39:48But how easy it got into science.
39:51As you know,
39:51usually you get your reviews back,
39:53you have to do 17 years more work.
39:56They want you to do all these
39:58impossible experiments.
39:59The review of this paper,
40:00oh, great work.
40:01But we don't like how you use the
40:04term empathy because it's loaded
40:06and I just use it as a shorthand.
40:09And my death, you know,
40:10because to always say social transfer of
40:12pain, social transfer of this or that,
40:15it's a lot of words.
40:16And, you know, in mice,
40:17you have to operationalize,
40:19you have to have a behavioral asset.
40:21So I just use the term very simply
40:24as a term that means one member of
40:27sets of a species is displaying
40:30a behavior that indicates it is
40:33being influenced by the affective
40:35or emotional state of another
40:37member of its species.
40:38That's all I mean by empathy.
40:42And so several years ago,
40:43we published a paper beginning to do
40:46a circuit dissection of what I found
40:49really interesting behavioral phenomena,
40:51which are called the social
40:53transfer of pain and analgesia.
40:55And I'm going to go through this fast,
40:58I apologize.
40:59This is published, the, the,
41:01the assay is actually pretty simple,
41:03and it was introduced to me by a postdoc,
41:05Monique Smith, who worked on it for her PhD.
41:08And when she applied to my lab,
41:09I brought her in, started,
41:11and she started telling me about this.
41:13And it was like, wow, OK,
41:15you can come work with me as
41:16long as we study this phenomenon
41:18because I knew nothing about it.
41:20The phenomenon, as the title implies,
41:23is very simple.
41:24You take a mouse and you put it in pain.
41:27We use injection of a inflammatory
41:29agent into a hind paw,
41:31very standard pain model.
41:32And then you take another mouse
41:35that has no physical injury,
41:36and you let it just hang
41:38out with the mouse in pain.
41:40And as shown here in these
41:43purple lines in here,
41:45the bystander mouse will show behaviors
41:48indicating it's in pain that last
41:51anywhere from 4 to 24 hours just
41:53by hanging out with the mouse in
41:56pain and all of these other assays.
41:58I'm not going to tell you it's
42:00just really nailing down that
42:02this is actually the truth.
42:04And I don't have time to go through this,
42:05but I think what's really interesting
42:07is this is this graph here,
42:09one of one interesting component.
42:11So here you can see the CFA was
42:15put into the right hind paw of
42:18the mouse experiencing pain.
42:20And you can see when we do this
42:23mechanical threshold test using Von Fray,
42:25here's it's only the the the
42:27paw that was injected that shows
42:30an increased sensitivity,
42:32shows that it's that paw that's in pain.
42:34In the Bystander mouse, it's both paws,
42:37which is what you would expect
42:39because the bystander mouse,
42:40unless it's brilliant, it doesn't
42:41recognize which paw the other mouse is.
42:44So it means there's some central
42:46mechanism going on.
42:47And then we did, what do we do here?
42:50Oh,
42:54so then we developed.
42:55I will take credit for this.
42:57My postdocs keep reminding
42:58me this was my idea.
42:59I keep saying, whose idea was this?
43:01It was my idea.
43:02So here what we did is,
43:04can you get a transfer of
43:06the opposite experience,
43:07the transfer of analgesia.
43:09So here what we did is both mice are in pain.
43:13They both have gotten the CFA
43:15injection in their hind paws.
43:16You then give one mouse
43:19morphine so it's analgesic,
43:21and then you let them hang out for an hour.
43:23And then the bystander mouse who
43:26has never experienced morphine
43:28just by hanging out with its buddy
43:30who's experiencing pain relief,
43:32shows a transient analgesic effect.
43:36And this is just the quantification of that.
43:39So the mechanic with it's a different.
43:41In this case it's not important.
43:44So we were able to show that there is
43:46social transfer of pain and analgesia
43:48and these are my last few slides.
43:51We then we then did a bunch of
43:54experiments that were consistent with
43:56the hypothesis that excitatory inputs
43:59from the anterior cingulate cortex,
44:02which has been implicated in playing
44:04a role in empathy in human beings
44:07from brain imaging experiments,
44:09that those inputs into the accumbens,
44:12for reasons we don't understand,
44:14are important for the mediation of
44:17this transfer of pain and analgesia.
44:19So finally, just just recently,
44:22this work literally just got accepted
44:24to I think it's science advances,
44:26whatever that is last week.
44:29And the obvious question is
44:31there's been a lot, you know,
44:33the question is,
44:34is MDMA,
44:35does it actually enhance empathy
44:37in human beings?
44:38People are very cautious about
44:40using that term appropriately.
44:42So there's no question that it enhances one's
44:49motivation to socially
44:50interact with another member,
44:52another human being in a non aggressive way.
44:56It seems to make one more interested in
45:00that other human beings experiences.
45:02But is it really enhancing empathy?
45:06And we still don't have the answer to that.
45:07But at least using these assays we can
45:10say it enhances the social transfer
45:12of pain and analgesia and that's all.
45:15And this is and then I'm done.
45:17So this just shows the experiment.
45:18This is the threshold of the Von Fry
45:22testing in the mouse that's in pain.
45:24And you know the lower the
45:26mechanical threshold,
45:27it just means that hind paws
45:29in pain if we give.
45:31So what we've done here now is with
45:33shorten the social interaction to 10
45:35minutes because remember I said if
45:37you let them hang out for an hour,
45:39the bystander mouse will manifest
45:42pain behaviors.
45:43So we wanted to reduce that so
45:45we could see an effect of MDMA.
45:47So I said let's shorten the social
45:50interaction to a time at least where
45:52there's a very modest transfer of pain.
45:54So this is the bystander mouse
45:56that got saline and then this is
45:59the bystander mouse that got MDMA.
46:02So in in a 10 minute interaction,
46:04if you give your bystander mouse
46:06MDMA in ways we do not understand,
46:09it's somehow a sensitized that mouse
46:12to the painful experience of its body.
46:14And obviously,
46:15we tested whether MDMA itself effects
46:19pain thresholds and it doesn't.
46:21And if we activate up to genetically
46:25serotonin inputs in the accumbens,
46:28that enhances the social transfer of pain,
46:30suggesting that MDMA is working
46:33via the release of serotonin.
46:36If we infuse MDMA into the
46:39nucleus accumbens directly,
46:41that enhances the social transfer of pain.
46:44I know I'm going through fast
46:46because I want to finish up and
46:49then it also enhances the social
46:53transfer of analgesia given IP.
46:56What is this showing me?
46:57And then again,
46:59if we infuse MDMA into the accumbens,
47:02it enhances the social transfer of analgesia.
47:04So again,
47:05the story is incredibly simple.
47:07Serotonin release in the incumbents
47:10is doing something magical that we
47:12don't understand how it's working.
47:14And as a mechanistically driven neuroscience,
47:17it drives me insane that we don't understand.
47:21It's because this is all phenomenology,
47:24this is all just mystical
47:27experiences with serotonin release.
47:30And then all of this just shows is that if
47:34we take a different mouse model of autism,
47:37in this case the Shank
47:393 heterozygous knockout,
47:41it shows a deficit in the
47:44social transfer of pain.
47:45And we can rescue that with MDMA,
47:48again saying there's some and similarly
47:51with the transfer of analgesia.
47:54So these mouse models of autism and
47:56this is manifest deficits in these
48:00behavioral assays or behavioral
48:03antecedents of end quotes, empathy,
48:06and we can rescue those with MDMA.
48:08And I'll, I'll finish up there.
48:10So what I hope I've convinced you of
48:13is that MDMA is a worthwhile topic
48:16of rigorous and ethical scientific
48:18investigation, at least in mice.
48:21It promotes pro social behaviors and
48:23perhaps in quotes, empathy, in part at
48:26least by causing release of serotonin.
48:33This is where it gets complicated.
48:35So truth be told, serotonin 1B receptor
48:39antagonist do not block the effects
48:42of MDMA on these assays of empathy,
48:46on social transfer of pain and analgesia.
48:50So there's something more going on that
48:52we don't understand and obviously,
48:53we're pursuing that now.
48:55And I would, I actually do believe this,
48:58that understanding these detailed
49:00mechanisms may really be a path towards
49:03the development of normal therapeutics.
49:05So finally, I only went 3 minutes over,
49:07which for me, I only talked for 47 minutes.
49:11So for me that's amazing.
49:14Obviously I didn't do any
49:16of this work myself.
49:17The initial serotonin work was done by
49:19Jess Walsh with some help from others.
49:22The MDMA work was started out by Boris
49:25Heifetz with some help from others.
49:27Matt Pomerance is still in the lab.
49:29He helps with everything.
49:30This the the social transfer
49:32of pain and animal.
49:33Jeezy and the MDA work were
49:35done by Monique and Ben.
49:36So thank you for your attention and
49:38I'm happy to answer any questions.