Yale Psychiatry Grand Rounds: "The NIH Brain Initiative: Accelerating Discovery Toward Cures"

December 15, 2023

Information

December 15, 2023

"The NIH Brain Initiative: Accelerating Discovery Toward Cures"

John Ngai, PhD, Director, National Institutes of Health Brain Initiative, National Institute of Neurological Disorders and Stroke.

ID11095

To CiteDCA Citation Guide

00:00Is to I got it. OK And then
00:05we got it. Got it here.
00:10All right. Sorry everybody.
00:12This is me doing this so that
00:14Doctor Nye doesn't have to.
00:16And that's important.
00:18Chemistry and biology, remember,
00:19because the one of the primary
00:21goals of the BRAIN Initiative is
00:24to develop new tools so that we can
00:27push forward the ideas with the
00:29technology that supports those ideas.
00:31And of course we don't want
00:33to be limited by technology.
00:34So the development of new
00:36technologies is a way to basically
00:38free us up to be more creative.
00:40And the
00:44his career then went on to get
00:46a PhD in biology from Caltech.
00:48He was a post doc at both Caltech
00:50and then at Columbia and then he
00:52started his faculty position at
00:54Berkeley and he there his lab was
00:58involved in olfactory neurobiology.
01:01He in his own work was pushing the boundaries
01:05particularly of molecular technologies.
01:07He had published you know way back,
01:09way back in the days and
01:12when we did microarrays,
01:14I don't know if any of you remember
01:15that he was already in this space
01:17and now of course has moved forward
01:19in that space towards the really
01:21cutting edge single cell technologies
01:22and so on that we use today.
01:24And he brings that knowledge to to his
01:27role as Director of the of the institute.
01:30He also served as director of Berkeley's
01:32Neuroscience Graduate program and of
01:34the Helen Wells Neuroscience Institute.
01:36So he comes into this with a lot
01:40of experience with students.
01:41So today he will be meeting
01:43with students for lunch.
01:44So I hope if any of you students
01:46are in the audience that are going
01:48to be having lunch with Doctor Nye
01:50after this that you take advantage
01:52of his of his experience.
01:55He's also served as Co Chair
01:58of the Brain Initiative,
02:00Cell Senses Consortium Steering Group
02:03and now he overseas the long term
02:05strategy and day-to-day operations of
02:06the NIH Brain Initiative as it strives
02:08to revolutionize our understanding of
02:10the brain and both health and disease.
02:12So I would like to welcome Doctor
02:15Nye and thank him very much
02:16for making the trip today.
02:24Thanks, Marina.
02:24I thank you all for being here.
02:27It's really a pleasure to get out of the.
02:28Get out of the house every now
02:30and then and see real folks.
02:33Yeah. Thank you, Marina,
02:34for the for organizing and John as
02:36well for the invitation to be here.
02:38It's really a delight.
02:40And just a touch screen.
02:41Do I dare? No, don't.
02:43Don't touch it.
02:44Don't touch it. OK,
02:47that so far. I had a wonderful visit.
02:49Really looking forward to meeting the
02:51students and other faculty as well
02:52later today. From what I can see,
02:54what's going on here really does align
02:56with what I'm going to tell you about what
02:58in our mission in the brain Initiative,
03:00as Marina said, to revolutionize our
03:02understanding of the human brain.
03:04So let me just see if this will work.
03:08OK, That works, right?
03:11We're good. OK, great.
03:14Do. What am I supposed to do?
03:17Oh, move my face out of the way.
03:20Is this of the mouse?
03:23You can see this is a technology
03:25initiative, and I'm the director.
03:27It's really, it's like you should
03:29all really be kind of worried.
03:34OK. How's that? Cool. Oh, but this,
03:36this this thing's still in the way.
03:38OK, yeah, great.
03:39I've only been doing this for three years.
03:41OK. So it's really a delight to be here,
03:43I think. And this audience doesn't
03:45need to be reminded of the vast
03:47complexity of the human brain and
03:49other other brains couple 100 billion
03:51cells making trillions of connections.
03:53It's the most powerful computer,
03:55I think, that we know of and certainly
03:57the most complex organ in the body,
03:59which also makes it most
04:01vulnerable to disease.
04:02So we the goal here is to develop new and
04:04better tools to understand this remarkable
04:07organ and eventually to understand
04:08how it works in health and disease.
04:11So we can start thinking about actual
04:13cures and not just ineffective treatment.
04:14So that's kind of the gist of what I'm
04:17going to get at to with you today.
04:19OK, so the mission of the
04:22US Brain Initiative,
04:23let me minimize my window here too,
04:25is to revolutionize our understanding
04:27of the human brain by accelerating
04:29the development and application
04:31of innovative technologies,
04:32Brain research through advancing
04:34innovative neuro technologies.
04:38It was announced as an initiative
04:39by the White House in 2013,
04:41with the first awards being made in 2014.
04:44And really the kind of the cool
04:45vision here was that it came at a
04:47time when there were advances and
04:49fields adjacent to biology and
04:50neuroscience that people recognize
04:52would actually help us in our quest
04:54to develop better tools to understand
04:56how this thing works and engineering,
04:59physics, chemistry,
05:00computer science and so and and
05:02and also in the social sciences.
05:04And this is really just a great a time
05:06of confluence to really leverage this,
05:08this, this huge,
05:10huge development in knowledge across
05:12across different and diverse fields.
05:15The BRAIN Initiative,
05:16the US BRAIN Initiative represents
05:17a partnership between five U.S.
05:19Federal agencies and private foundations.
05:21And our efforts at the NIH have
05:23been guided by two strategic plans.
05:24The first was the so-called
05:26BRAIN 2025 report.
05:27This was a report commissioned by then NIH,
05:31Director of Francis Collins,
05:33Advisor Counsel to the Director of
05:35Working group of that ACD we call
05:37it and that was chaired by Corey
05:39Bargman from Rockefeller University.
05:40I'm Bill Newsom at Stanford
05:41University and it really laid out
05:43the 1st 10 years of where they
05:45thought this thing could be going.
05:47An update to to the strategic plan was
05:49made in the brain two point O reports
05:51that was released in October of 2019.
05:53That kind of took a look at where
05:55things stood at the five years in and
05:57and a kind of a refreshed look at
05:58where things should go in the future.
05:59And everything we're doing is kind
06:01of based on these two visionary
06:04documents Now at the NIH.
06:06Our goal is to develop a new and
06:09apply new tools for understanding
06:11how neural circuits underlie complex
06:13behaviors in both health and disease.
06:15The initiative spans 10 of the
06:1827 NIH institutes or centers,
06:20I'll refer to those as ICS.
06:21And in order to achieve this goal,
06:24we feel it's important to leverage
06:26emerging technologies from across the
06:27scientific disciplines to enable new
06:29discoveries about neural circuit function,
06:31to use these discoveries as a
06:33foundation for new therapies for
06:35human brain disorders and very
06:37importantly to disseminate and
06:38democratize these technologies.
06:40We're both for discovery as well As
06:42for clinical applications and very
06:43importantly for the benefit of all.
06:45Now with the NIH BRAIN Initiative,
06:47we we've,
06:47we've mapped out what we're doing based
06:49on the BRAIN 2025 report in nine main areas.
06:54We study,
06:55we,
06:56we we fund studies looking at tools,
06:59tools that allow us to better
07:02understand cell and circuit functions,
07:04invasive and other non invasive neuro
07:06recording and modulation technologies.
07:08I'm going to touch on examples
07:09of all these as they go along.
07:10Newer neuro imaging technologies
07:13across different scales.
07:14We have a very robust portfolio
07:17in systems neuroscience where
07:19our program is designed to use,
07:21develop and use the latest tools for
07:23dissecting neuro circuit function.
07:25And very importantly,
07:26we have a robust program in human
07:28neuroscience which touches not
07:29only on developing first and human
07:32treatments for various disorders,
07:33but also to use the human brain as a
07:36model system for understanding neuro
07:38circuit function and kind of interleaves.
07:40Across these different programs
07:41we are supporting various efforts
07:43in data science and informatics,
07:45very important training,
07:46inclusion and equity.
07:47We need to make sure that this is a
07:50sustainable enterprise with bringing
07:51in new talent into the into the fold.
07:54Neuroethics, you know what we're
07:55doing matters for society,
07:56it matters for individuals and again,
07:58as I mentioned before, dissemination,
08:01democratization and commercialization.
08:04OK, just a very brief history of the
08:06funding of the BRAIN Initiative.
08:08It started in 2014 at NIH with a
08:11very modest $46 million investment.
08:13And you can see it's grown quite a bit.
08:16And the last fiscal year, FY23,
08:17which just ended in September of this year,
08:20we made it,
08:21we had $680 million to invest
08:24in in these studies.
08:25Our funding comes from 2 main sources.
08:27Well,
08:27two sources and dark blue
08:29is the base funding.
08:30So this represents funds that are allocated,
08:33appropriated by Congress to
08:35the NIH to these ten ICS.
08:37It's a line item in these ten IC budgets.
08:40On top of that in the light blue are
08:42funds from the 21st Century Cures Act,
08:43which was passed in 2016 that
08:46started in 2017.
08:46You can see it's a variable
08:48amount of funding.
08:49It runs through 2026 and it's
08:52really allowed us to pursue this
08:54robust growth of investments that
08:57really has catalyzed the field in
08:59terms of developing these tools
09:00for doing really cool stuff.
09:02OK,
09:03So what have we done since then since 2014?
09:05So these are the numbers we
09:07have since up through 2022.
09:08We funded by now actually over 1200
09:12PIS across over 230 institutions
09:14and they've they've been supported
09:16by now by over 1100 Brain Awards,
09:19they've published a bunch and
09:21they've published in journals
09:23covering different areas.
09:24So this reflects the multidisciplinary
09:26nature of what we're supporting.
09:28Here's a word cloud showing
09:31the common themes.
09:32One day I expect to see non
09:33rental cells better represented,
09:35but we're we're working on that and
09:37quite a few really nice publications. OK.
09:40So that's kind of brain by the numbers.
09:43Today I'd like to leave you
09:45with three key takeaways.
09:46The 1st is that brain funded
09:48advancements in tools and technology.
09:50They're already making their
09:51way into the clinic.
09:52So we see we're already seeing big
09:54potential to impact humans today,
09:56not just sometime in the future,
09:59but in the meantime,
10:00as I mentioned before,
10:01we really do need to understand more
10:03about the brain in order to to come
10:05up with some actual cures and preventions.
10:07And here our teams are developing
10:09new resources and technologies
10:10that are laying the foundation
10:12for these future cures.
10:13And in the process,
10:14of course generating in a lot of great
10:16information about how the brain works.
10:18And then finally,
10:19kind of part of all this is
10:21that we're creating a new
10:23way of doing science that we feel is and
10:25will accelerate the pace of discovery.
10:28OK. So I'm just going to go through
10:30each of these three points and
10:32give you some examples and little
10:34vignettes to support these claims. OK.
10:37So what's going on in the clinic today?
10:40Well, I'm sure a lot of you are aware
10:43or almost all of you are aware of deep
10:45brain stimulation that's been used for
10:47over 2 decades now to treat the symptoms,
10:50the motor symptoms of Parkinson's
10:51disease and other movement disorders.
10:54And it's kind of been great for that.
10:55It's the gold standard for treating
10:57patients with these conditions.
10:58But applying it to other arguably more
11:01complex conditions like treatment,
11:02refractory depression, OCDPTSD,
11:06things like this has been not really
11:09been going so well until very recently.
11:11And we're now we're seeing kind of a
11:14whole bumper crop of papers and studies
11:16showing the application to these more
11:18complex neuropsychiatric conditions.
11:20So what's changed?
11:21Well, a couple things has changed.
11:23One is that we are seeing in addition to
11:25these devices in addition to stimulating
11:27that can now record neural activity.
11:29So there's a possibility of recording
11:32activity in the brains of patients
11:35and identifying neural activity
11:37biomarkers for the conditions that
11:39could be stimulated back into,
11:41we might think,
11:42a better space.
11:43Another big advance has been the
11:46the development of these really cool
11:49artificial intelligence algorithms
11:50that can actually deconvolve and
11:52interpret that information to
11:54kind of give give us actionable
11:57biomarkers for the stimulation.
11:58And then finally better mapping on
12:01an individual basis of patients of
12:03their actual pathways and circuits
12:05so that the electrodes can be placed
12:07in the ideal location and record
12:10and stimulate in such a way that
12:12can give an effective treatment.
12:13So this has been going on for a while.
12:14Here's just one example of one of these
12:17studies published just this past fall
12:21from Helen Mayberg and Chris Rozell's Group,
12:23A really great collaboration between
12:25neurologist Helen Mayberg's been been
12:26been pushing this for about two decades
12:29and Chris Rizzell who's an engineer.
12:31So again embodying the ethos
12:33of the brain initiative using
12:34multidisciplinary approaches.
12:36So here they had about,
12:37I think it was 12 patients that
12:39had we're suffering,
12:40we're living with chronic and
12:44treatment resistant depression and
12:47placing the electrodes in the singlet
12:49very precisely and being able to
12:52record activity from these patients.
12:54They were actually able to derive
12:56using AI really cool AI techniques
12:58biomarkers for the patient state.
13:00So they would implant these patients,
13:02they stimulate them.
13:04I think 3/4 of them actually either
13:06showed a great response to the
13:09stimulation if not remission and they
13:11can actually use the this biomarker
13:14activity to predict how they were doing.
13:17And in fact in in in one or a few cases,
13:20they could predict a month in advance when
13:21the patient was getting into trouble.
13:23So this is a great way to not only
13:25understand what's going on over time,
13:28but actually as a way of of
13:30tuning the the therapy before
13:31the patient gets into trouble.
13:33So this is really kind of cool stuff
13:35and now the challenge with all these DBS
13:36technologies is how do you scale it.
13:38These are again our small case
13:40or small patient studies,
13:41but it really does pave the way for
13:44for thinking about how we can treat
13:47these otherwise debilitating disorders.
13:49Treatment resistant depression,
13:51OCDPTSD, Bingeing eating disorder.
13:53Eddie Chang's group at UCSF has
13:55now been able to record activity
13:57by markers associated with chronic
13:59pain and now they're really working
14:01hard to see if they can't use DBS
14:03to alleviate those those symptoms,
14:05which has great implications not
14:07only for treating the pain,
14:08but also adjacently for for avoiding the
14:12consequences of substance use disorder.
14:15OK, here's a somewhat different study.
14:18This is from Kappa Kappa GROSSA
14:20Group and Pittsburgh,
14:22where they're looking at now
14:23not deep brain stimulation,
14:24but epidural stimulation of the spinal cord.
14:28So here we have patients that
14:29suffered stroke.
14:30And in fact,
14:31the case study that I'll show you here
14:33was a woman who suffered a stroke,
14:35I believe,
14:36nine years before the study was conducted.
14:39And the idea here is that there's
14:42a stroke thinks,
14:43I think it was a thalamic stroke
14:45affecting the transmission of
14:47information down the cortical spinal
14:50tract to move the upper limbs.
14:52OK.
14:53And as it turns out,
14:54there are local circuits in the spinal
14:56cord that control this movement.
14:58So it's not just like you're you're
15:00flipping a switch that goes right
15:01to your hands or your fingers,
15:03but you're actually engaging the
15:06local surface in the spinal cord.
15:07So the idea here is that if the signal
15:10traveling to the spinal cord is diminished,
15:12maybe one can get some recovery
15:14of function by amplifying the
15:16output in the spinal cord.
15:18So here in a less invasive technique,
15:20they've laid electrodes epidurally
15:22over the spinal cord in the cervical
15:25spinal cord and stimulation was
15:27optimized for each patient.
15:29I'm not going to bother going through
15:30these these graphs because I have
15:31a much better video to show you
15:33what's going on and the really cool,
15:34there are two really cool things.
15:35One is that the benefits once
15:37they turn the stimulator on were
15:40immediate in terms of giving,
15:41functional recovery of the arm
15:43and it lasted at least four weeks
15:44after they stopped the study.
15:46So that tells us two things.
15:48One is that the circuits are there to
15:52be engaged and you just have to engage them.
15:54And the other is there's probably
15:56some plasticity going on as well.
15:58OK, so a picture is worth 1000 words.
16:01A video's worth,
16:02I think 1,000,000 here is just
16:05this patient in the clinic,
16:08and the task here is to feed herself
16:11something that most of us take for granted.
16:13And so she is supposed to pick up this,
16:16I think it's a chicken nugget.
16:19Dip it in the sauce and put it in her
16:21mouth and you can see she's struggling
16:24with this. But she's a good sport.
16:25She's really kind of working at it.
16:29This was done during COVID
16:32times, hence the masks.
16:36She can barely lift her arm up
16:40and and she needs a little
16:44help at the end to get there.
16:50OK, so the next clip,
16:54the stimulator is turned on.
16:55So just an epidural stimulator.
16:59Still fine.
17:00Motor skills are still impacted,
17:03but she's doing a bit better,
17:04A little bit of help getting
17:06her fork in her hand,
17:06but she's holding the
17:08fork pretty well. Dip
17:13and there you go.
17:14Really quite remarkable.
17:15No physical therapy involved.
17:17This is just turning the stimulator
17:19on and it happens immediately and
17:21it lasts for weeks afterwards.
17:23And this parallels other studies
17:25done by this group as well as Greg
17:28Cortine in Switzerland where they've
17:30done similar epidural stimulation
17:31in spinal cord injury patients.
17:33Again, re engaging or amplifying the
17:36signal in the lumbar spinal cord to
17:39help movement of the lower limbs.
17:41And they've actually been able
17:43to identify and map the cells and
17:45circuits that show the plasticity
17:47using various techniques.
17:48So this is really showing great promise
17:51for how to assist patients recover
17:53function after these traumatic events.
17:56OK.
17:57So these are just two examples.
17:58There's a lot of other cool stuff
18:00going on in this space,
18:01including with deeper in stimulation
18:02for motor recovery.
18:03But just thought I'd give you 2
18:06brief vignettes for how we're taking
18:09technologies and pushing them into the
18:10clinic in these first in human trials.
18:12And now,
18:13of course,
18:13the challenge is to expand these studies,
18:15to validate them in randomized trials
18:18and then to disseminate it more broadly.
18:21OK.
18:22So the second key,
18:23key take away is that our teams are
18:26developing new resources that are laying
18:28down the foundation for future cures.
18:30We need more information about
18:32how brain cell types and circuits
18:35work to underpin behavior,
18:37so here are just a couple
18:38of examples I'll give you,
18:39some of which was done locally.
18:43One of the goals of systems
18:46in circus neuroscience is to
18:49understand how activity patterns Dr.
18:51downstream,
18:52other downstream units or
18:55ultimately behavior.
18:56We've seen a lot of progress
18:58in studying this using optical
19:00methods with genetically encoded
19:02calcium sensor sensors.
19:03But calcium is just a proxy for
19:06neural activity in most cases.
19:08And what what the field has
19:10really been driving at is to get
19:12direct sensors of voltage membrane
19:13voltage because that's really
19:15the currency in in most cases.
19:17Sorry for the pun of how
19:18circuits are functioning, right.
19:19So there have been a number
19:21of genetically encoded voltage
19:23sensors out there and also some
19:25chemical voltage sensors,
19:26but they've they've suffered
19:27from a couple of key things.
19:29One is that you don't,
19:32you don't get a very high signal,
19:33so you really have to put a lot of energy,
19:36a lot of light energy into the cells,
19:38into your tissue in order to
19:40get a reasonable signal out.
19:42So this is a problem in terms
19:44of tissue damage.
19:44Another is that up until recently,
19:47all these probes,
19:48they'd respond to changes in
19:49voltage into to depolarizations by
19:52decreasing their signal intensity,
19:53decreasing the fluorescence.
19:55So I think you can imagine
19:57intuitively that this this is becomes
20:00problematic for small changes.
20:01You're looking at a decrease in
20:03signal so you get into problems
20:04with Singleton noise.
20:05So sensitivity and you really want
20:07sensitivity is an issue to be solved
20:09and you also would rather have what
20:11we call a positive going sensor.
20:13So here are the really cool
20:15developments along these lines.
20:16So the Pure Bone and Chen groups.
20:20Developed a new spike detection through
20:22these so-called positive going sensors.
20:24These ASAP generally encoded
20:26voltage indicators called spiky,
20:28spiky and spiky 2 and they've also
20:31in parallel developed so that's the
20:33the chemistry of the actual sensor.
20:36But they've also developed low power
20:38two photon imaging that actually
20:40allows them to detect the smaller
20:42signals and you can see here on the
20:44right these little spike signals
20:45that they can detect this way in
20:48response to changes of voltage.
20:50So Michael Lynn's team who developed
20:52the so-called who developed the
20:55original ASAP sensors also developed
20:57these ultra fast sensors that
20:59now are also positive going.
21:01The upper trace here shows responses
21:04of the cell to to to calcium these
21:08you can see these fairly slow slow
21:11wave forms that in relation to the
21:12bottom where you can see the actual
21:14spiking of the cells really quite
21:16remarkable and it's it's sustainable.
21:17So the this gets to the issue of not frying
21:20the cell before your experiment is over.
21:22OK.
21:23And then finally Adam Collins Group utilized
21:26a novel screening strategy to identify
21:28these new archaeohodopsin based sensors.
21:30And here you can see two cells
21:33in a preparation that are
21:35that are electric connected,
21:37you can see them spiking together.
21:38So this now can be applied more broadly.
21:41These three types of new Gabbys can
21:44be applied more broadly to look at
21:46activity patterns across areas in vivo,
21:50so really to push the boundaries
21:52of understanding the function in
21:55neural circuits. So one example.
21:58So another example I'm showing you here
22:00is from Lynn Tian's lab at UC Davis
22:02and now she's about about to or just
22:04moved to the Mount Splunk in Florida.
22:06So the back story of this is,
22:07if we look at drugs used to treat,
22:09for example, depression,
22:10it's been about 5.
22:12It's been literally 5 decades since the
22:15introduction of philosophy or Prozac.
22:17And since then there hasn't been a
22:19whole lot done in terms of finding new
22:22classes of antidepressants until recently
22:24with this research interest in psychedelics,
22:27right.
22:27And the psychedelics fall
22:29into different classes.
22:30There's the ketamine,
22:31there's ketamine,
22:31but there's also these five HT,
22:342C serotonin receptor agonist.
22:36Now they seem to be quite promising,
22:38but of course one of the issues is
22:40the side effect of hallucinations.
22:42So what Linton's lab has done is
22:45they've developed this biosensor
22:47for serotonin based on the
22:5155 HT 2A2A. Sorry about that.
22:53Did I say 2C? And I meant 2A,
22:55which actually it gives different
22:58readouts based on conformational
23:00changes and they can deliver this
23:02into a mouse using AAV techniques to
23:06look at to monitor 5 HT release and
23:10expression and what was really cool here.
23:13Using Cyclite in a vivo screening platform,
23:16they could identify various 5 HT 2A
23:19agonist and actually parse them out
23:22depending on whether they might or
23:24might not have hallucinogenetic,
23:25hallucinogenic activity and actually
23:27use this potentially as a screen for
23:30identifying new drugs that could
23:31be used to treat depression but
23:33without these unwanted side effects.
23:35So this is just and there's a lot
23:37of work going on in this space,
23:38but it's just showing you the idea
23:42that with these new techniques,
23:44these new technologies,
23:45screening technologies,
23:46we might have a way of getting at
23:48a new class of therapeutics in
23:50the in the pharmacologic domain.
23:52OK. So the third key,
23:55key take away I'd like to leave you
23:57with is that we're creating a new way
24:00of doing science that I think will is
24:02and will accelerate the pace of discovery.
24:05So in 2022, we launched what
24:07we call the BRAIN initiative,
24:08transformative projects,
24:09kind of a bold statement.
24:11And the idea here is that in
24:13order to fully understand how
24:15Brain Circus actually work,
24:17we need ground truth information
24:19about the cell types and the
24:20connections they make with each other,
24:22and also a way of interrogating
24:24this information to test,
24:26to develop and test hypothesis about their
24:29roles in different types of behavior.
24:31So the first project I'll tell you
24:33about is the Brain Initiative,
24:34Cell Atlas Network or Bican.
24:36And the goal here is to map,
24:38map out all the brain cell types and
24:40circuits across multiple species,
24:42with an emphasis on big brands,
24:43particularly in humans.
24:45So that's the parts list.
24:47The second is the wiring diagram
24:49and we just launched this what
24:50we call the BRAIN initiative,
24:51connectivity across scales or
24:53brain Connects program,
24:54which will provide the tools that
24:56we'll need to develop a wiring
24:59diagram for entire mammalian brains,
25:01whole brains.
25:02And then finally,
25:03we're developing an armamentarium
25:05or just a big toolkit for precision
25:08brain cell access,
25:09which will leverage the information
25:10coming from the other two big
25:12projects to allow researchers to
25:15test hypothesis about the roles of
25:18specific cell types and circuits
25:20in underpinning behavior both
25:22in health and disease.
25:24And ultimately,
25:24I see these three projects working in
25:26a mutually reinforcing way that will
25:29lead us to precision circuit therapies.
25:31And I think the blue sky scenario in
25:33my mind is developing precision gene
25:35therapies for human brain disorders.
25:37So a tall order,
25:39but I I hope to to share with you
25:41some evidence that I think we're
25:43going along on the right track.
25:45OK.
25:45So let's start with the cell
25:48atlasing project.
25:49This effort actually started back in 2014.
25:52It was one of the first
25:53programs launched by the
25:55BRAIN initiative in 2014.
25:56I was actually a member
25:58of this BRAIN initiative,
25:59Cell Sensors Consortium and
26:00here the idea was to identify,
26:02validate scalable technologies that
26:05could be used to create an inventory of
26:08all the cell types in a mammalian brain.
26:10So that was back in 2014.
26:13It was rapidly scaled up in 2017 into
26:17a larger group known as the BRAIN
26:19Initiative Cell Sensors Network.
26:21And here the idea was to actually
26:24implement these tools to no small order,
26:27no small order to revolutionize
26:29our ability to classify brain cell
26:31types based on a multimodal or an
26:34integrated analysis of their molecular,
26:36anatomical and physiological property.
26:37So not just looking at transcriptomics,
26:39but really getting a full picture
26:41of what constitutes a cell type.
26:43So this program meant for five years,
26:44we're just kind of wrapping it up right now.
26:47And it serves as a basis for BICAN,
26:49the BRAIN Initiative, Cell Atlas Network,
26:51which now is placing our emphasis
26:53on the human brain.
26:55OK.
26:56So let me just go over with you what
26:58we've accomplished with the BRAIN
27:00Initiative Cell Senses Network.
27:02It started out with the BICCC and the BICCN.
27:07In October of 2021,
27:09the group and I was still involved
27:12with this group published 17 papers
27:14in Nature and 10 papers in Nature's
27:16sister journals characterizing the cell
27:19types in the primary motor cortex of mice,
27:23non human primates in humans.
27:25This was started out as a pilot project.
27:27It's kind of hysterical as a small
27:29project just to make sure we could
27:31all kind of come up with a common
27:33picture across this very large group
27:35that can integrate information
27:37across these different techniques.
27:38It wasn't a given that you could do it.
27:40In fact,
27:40many of the techniques to integrate
27:42the information were were developed
27:44through this project just published 2
27:46days ago or were series of 10 papers in
27:49nature characterizing the entire mouse brain.
27:52And I'll go over that with you in a
27:54moment really I think a monumental effort
27:56and a really a landmark series of studies.
27:58And then just back in October the BICCN
28:01non human primate and human groups
28:03published 21 papers and three science
28:06journals giving us a draft cell Atlas of
28:09the human brain and non human primate brain.
28:11So let me just go through
28:13with these with you in order.
28:15So the the primary motor cortex
28:17paper gave us a multimodal census and
28:20Atlas of the primary motor cortex.
28:22As I as I said,
28:23across 33 species,
28:24it really was a triumph of team science.
28:28This really hadn't been done before
28:30neuroscience to my knowledge.
28:31There were over 250 of us from over
28:3445 institutions across 3 continents.
28:37It wasn't easy in the beginning,
28:38but we learned how to work
28:40together collaboratively.
28:41A lot of great science was done.
28:42A lot of junior faculty careers
28:44were launched through these efforts
28:46and it's really set the stage for
28:48the larger and more comprehensive
28:50atlases of both mice as well as the
28:52big brains.
28:55So again, just published 2 days ago
28:58this huge effort to map the entire mouse
29:01brain over 32,000,000 cells were analyzed
29:04here with a variety of techniques.
29:08The group came up with over 5000
29:10neuronal and non neuronal cell types
29:13that were identified and it really
29:15started giving us insights into the
29:17organizational principles underlying the
29:18diversification of cell types in the brain.
29:21Let me just go through this with you very,
29:23very quickly.
29:23So the bedrock of all these studies
29:25has been transcriptomics.
29:27So single cell transcriptomics
29:29identified over 5000 clusters that
29:31one could identify statistically.
29:34What was really cool was throughout
29:35this effort there was this revolution,
29:37mini revolution in spatial
29:39transcriptomics that was a what?
29:41That allowed researchers to actually
29:44ask whether a cell type identified
29:46by transcriptomics could actually,
29:48did it actually exist distinct
29:49from say a cluster next door?
29:51And the answer is yes,
29:52in many cases we could the the,
29:54the group could see distinct
29:57spatially restricted cell types that
29:59corresponded to what was identified
30:02with single cell transcriptomics.
30:04So really really quite remarkable and very,
30:07very importantly the group
30:09used this information,
30:10this information plus epigenomic profiling
30:12to really look at the regulation of
30:14what actually gives you a cell type.
30:17And what they showed was that one
30:18could classify these cell types
30:20pretty well based just based on
30:22the transcription factors alone.
30:23And that together with the epigenetic
30:25profile really started giving us an idea
30:28of what what happens developmentally
30:29as well as over aging as well as evil
30:31and evolutionary terms in terms of
30:33what makes a cell type A cell type.
30:35So this is really quite,
30:37quite beautiful work.
30:37It gives us a basis now for
30:39functionalizing this information to
30:41really get deeper into what how you
30:44construct A neural circuit but also
30:45how these neural circuits function.
30:47OK.
30:49And then back in October again these
30:5221 joint publications across 3 science
30:54family journals were published.
30:56Really just a draft Atlas of the human
30:58brain and non human primate brains.
31:00And this is this effort really was
31:04is unprecedented and it paves the way
31:06for a greater understanding of the
31:08human brain at the cellular level and
31:10gives us tools for understanding disease
31:12processes using this information.
31:15Let me just walk you through this
31:17very quickly.
31:17The group characterized these many
31:20thousands identified cell types across space,
31:22again using single cell techniques as
31:25well as spatial techniques across species.
31:28So their group looked at these maps,
31:31these inventories across different
31:32non human primate species.
31:34Not too surprisingly,
31:35the cell types identified on humans are
31:37most similar to the ones in the great apes.
31:39What was really cool is that the
31:42gene showing the highest differential
31:44expression between chimpanzees and
31:46us tended to localize in areas known
31:48to be in the so-called accelerated
31:51genomic regions,
31:52but also very importantly to
31:54encode proteins in the synapse.
31:56So this really started to tell
31:58us what makes us different from
32:00our nearest relatives.
32:01And then also across time,
32:03very importantly, looking at studies
32:05during development as well as aging
32:08and again giving us insights into
32:10what it takes to construct a neural
32:12circuit based on the cell types.
32:15OK. So this really sets the stage
32:17for this larger brain initiative,
32:19Cell Atlas Network,
32:20which now we'll actually dive in and
32:22get a deep dive into all the cell types
32:25in the human brain and also starting
32:27to look at the the variability in these
32:31parameters across multiple individuals.
32:34So our goal is not just to get
32:35information on one or a few individuals,
32:37but really to understand the basis of
32:40variation among humans at the cellular level.
32:42OK, so we've come a long way
32:46in the last number of years.
32:47And now the key to is to disseminate
32:50and democratize this information.
32:52The group is working very,
32:53very hard to develop knowledge bases that
32:56can be queried by researchers anywhere
32:58in the world who are interested in
33:00their particular cell type or circuit.
33:02And to put this in the context
33:03of a larger map or inventory.
33:05And another purpose here is
33:07to use this information,
33:09as I said,
33:10to understand the cellular basis of
33:11disease and disease progression.
33:13Already we're seeing advances in this area.
33:15This is just one figure from the Seattle
33:18Alzheimer's Disease Brain Cell Atlas.
33:20It's a group led by Ed Lean at the
33:22Allen Institute and Dirk Keane at
33:24University of Washington together as a
33:26collaboration with with Kaiser Permanente.
33:29And here they've looked at I think 90
33:31some odd human brain samples from mid,
33:34mid,
33:35temporal gyrus from patients at
33:39different stages of Alzheimer's.
33:41And what they're seeing is either
33:42under representation or over
33:44representation of different cell types.
33:45And this might start giving us some
33:47really good clues about what's
33:48going on not just at the very end
33:50stage of the disease but what what
33:51could be going on in the early
33:54stages to identify the drivers from
33:56a cell biological basis that could
33:58give us mechanistic insight about
34:00how to actually intervene with
34:02with prevention or cures.
34:03So we're just getting started on this
34:05but again the idea here is to lay the
34:07groundwork for future therapies and cures.
34:09So the second project here is how
34:12to develop tools to to give us a
34:15wiring diagram of organisms brains.
34:17Now as Biga left us,
34:21creating these human cell type atlases
34:24has been the connectivity analysis is
34:26orders of magnitude I think more complex.
34:28We don't yet have the tools to do
34:30this for home mammalian brains and
34:32one of the problems is of scale
34:34and resolution and frankly just
34:36handling all the data.
34:37So there's been a lot of work and
34:39model systems that we feel that are
34:42helping us get there step by step.
34:43So a lot has been done in the fruit fly
34:46starting with the larval connectome.
34:49Here a collaboration between various groups,
34:51Hopkins,
34:51Hughes and Cambridge has led to the
34:54generation of a full connectomer for
34:56larval fly that contains just 3000
34:58neurons with about half a million synapses
35:01and they can be categorized as input,
35:03output or inter neurons and cluster.
35:05You can cluster these based based on
35:08their connectivity into about 93 types.
35:10And what was one cool insight here
35:13is that they could they identified
35:16connections in about 40% of the
35:19neurons that were were recurrent.
35:20And these happened in areas
35:22of learning and action.
35:24So we we hear a lot about
35:26recurrent neural networks that
35:27are being used in AI algorithms.
35:28Mother evolution has figured
35:29a lot of this out.
35:31We can learn a lot about how
35:32to develop better algorithms,
35:34better computers,
35:35based on looking at how
35:37neural circuits functions.
35:38So this is from the larval fruit fly.
35:41Just very recently the the whole adult
35:44Drosophila brain has been characterized
35:47the connectivity has been characterized
35:49by the flywire group led by Sebastian
35:51Son and Mala Murthy at Princeton.
35:53They used citizen science,
35:54many,
35:55many hundreds of researchers
35:56to proofread these maps.
35:58They've they've launched
35:59this out on public websites.
36:00Really cool stuff.
36:01And now maps like this allow
36:03researchers interested in the
36:05neuro circuit basis of behavior
36:07to formulate their hypothesis
36:08based on the ground truth of the
36:11connectivity diagrams they can do.
36:12You can do your experiments,
36:14you can go back and either falsify
36:17or truthify your your results
36:19based on what is or isn't there,
36:21at least in terms of the anatomical
36:23basis of these circuitry.
36:25So it really opens up a really big
36:28domain where of investigation that can
36:31really be constrained by biological reality.
36:35So this is a family brain.
36:37One has to scale this up 1000 fold,
36:403 hours of magnitude to get similar
36:43information from a mouse brain.
36:45So here's a paper from 2015
36:47Bobby Kasturi when he was in Jeff
36:49Lichtman's lab and here they analyzed.
36:51They reconstructed all the connections
36:53in a 1500 cubic Micron volume
36:56of the mouse neocortex.
36:58Since then it's been the
36:59effort's been scaled up.
37:00Here's some images from a preprint
37:03from Jeff Lichtman's lab in a
37:05cubic millimeter of human cortex.
37:07So they've scaled this up 500,000 fold.
37:11So lots of orders of magnitude,
37:135 orders of magnitude,
37:14almost almost six orders of magnitude.
37:17And they've reconstructed about 50,000
37:19cells with 130 million synapses.
37:21This data set,
37:23which gives us beautiful pictures like this,
37:26takes up a petabyte of data.
37:29And to scale up to whole mouse
37:30brain is another 500 fold.
37:32Let's call it 1000 fold.
37:33So that means that a synapse level
37:36reconstructed entire mouse brain is
37:38going to occupy an exabyte of data.
37:40That's a lot,
37:41OK.
37:41Similar studies have been done by
37:43the IRPA Fund and Microns project,
37:46where they've also looked at about
37:48a cubic millimeter of cortex in
37:49the mouse brain, in visual cortex.
37:53In addition,
37:53they've actually acquired
37:55functional data on these, on these,
37:57on the sample, which kind of it.
37:59It's the beginnings of functionalizing
38:01the static anatomical studies.
38:03But again that we're
38:04looking at over an exabyte
38:06of data. So this is the big
38:07challenge for the field.
38:08How do we develop,
38:10develop tools where we can scale this up.
38:12The bad news is we we have 1000 fold
38:15scale up to accomplish the good news,
38:17it's only 1000 fold.
38:19OK, so we've launched this
38:22brain connects project.
38:23The first five year phase is to
38:25develop tools where we can do
38:26this not on one mouse brain but
38:28on multiple mouse brains and also
38:30develop tools for looking at
38:32big brains at less resolution.
38:33We just funded 11 grants from 40
38:36universities and research institutions
38:38across the globe and over the next
38:41five years they're going to help us
38:43identify scalable technologies that
38:45can do in the connectomics world
38:47what the Cell Cell Census Group has
38:49done in the cell Atlassian world.
38:52And the projects fall into three
38:54complementary core technologies.
38:56One is using high throughput
38:57electron microscopy.
38:58Here's just an example reconstruction
39:00from the Microns group using
39:02pretty cool molecular sequencing
39:04tools to map out on a large scale
39:07connectivity patterns as well as
39:09optical and X-ray tomography to map
39:11out the these broader connections.
39:14So we're very excited about this.
39:15The project just launched this past
39:17year and we're looking forward to
39:19seeing progress that will get us
39:20to the second phase where we could
39:22actually Start learning about whole brains.
39:24But in the meantime,
39:25the groups are charged with getting
39:28us some good biological information
39:30about cell significant cell volumes
39:32of the brain.
39:33And right now one of the key
39:35projects is on hippocampus and
39:36another one is on basal ganglia.
39:37So stay tuned for that.
39:40OK.
39:40So once we start having in hand
39:44this inventory of cell types
39:46and their connectivity patterns,
39:47we want to be able to test hypothesis of
39:50what these cells and circuits are doing.
39:52And this is where the armamentarium
39:54or toolkit for accessing these
39:56cell types comes in.
39:58The main the workhorse for these studies
40:00so far has been Adeno associated virus
40:02which has been developed for years.
40:04It's already been used in gene
40:05therapies that we've been reading
40:07about and right now it's it's
40:09kind of the the the standard and
40:12we're hoping to develop develop
40:14this with other viral vectors as
40:15well as non non viral methods.
40:18The the strategy here is there
40:20are two complementary strategies
40:21for gaining precision access
40:23to cell types in the brain.
40:25One is to leverage the information
40:26coming out of the cell census projects
40:29that have identified cell type
40:31specific expression of genes as well
40:33as candidate enhancer regions that
40:35might be driving that specificity.
40:38So and the other is going to
40:39be on capsule development.
40:40So I'll go through this in turn.
40:42So Basilica,
40:43Tasik and colleagues of the Allen
40:45suit have gone through and screened
40:47this information from the cell census
40:50projects and the different ways
40:52of identifying putative enhancers.
40:54They can show that they actually do
40:56give us good restricted activity,
40:58and one can start combining this
41:00in combinatorial arrays
41:02to give more targeted,
41:04more targeted expression of payloads
41:06that are carried by the AAB in
41:09molecularly identified cell types.
41:12Now, work led by Viviana Grotonaro
41:14in collaboration with various folks
41:16at UCSD and the Allen Institute,
41:18have been busy engineering the capsules
41:20to tune the tropism of these viruses.
41:22One of the big what advance this has
41:25been that Viviano's group and others
41:27others have also identified capsid
41:29variations using directed evolution that
41:31actually cross the blood brain barrier.
41:33Someone can actually deliver these
41:35vectors to the two brain cell types not
41:37not just by injecting into the brain
41:40itself but by injecting bloodstream
41:42and they've actually being able to
41:44find variants that target get across a
41:47blood brain barrier and target neurons
41:49in various non human primate models.
41:52It shows enhanced delivery to the brain
41:54a very importantly reduced delivery
41:56to the to the liver and hepatotoxicity
41:58is one of the big bugaboos of human
42:00gene therapy and they can they can
42:02observe expression mostly neurons.
42:04They also they have other variants
42:05that also target non neural cells
42:07and it facilitates studies and non
42:09human primates here as well as
42:11uncultured human neurons.
42:12So these are just some of the efforts
42:14that are underway geared toward
42:16getting better and more precise
42:18access to cell types in in post Natal
42:22brains with the goal of interrogating
42:25circuits to develop better models
42:26for neuro circuit function.
42:28But I think these are going to serve
42:30as the progenitors for precision gene
42:33therapies for human brain disorders.
42:36So these are the three projects that we've
42:38launched over the last couple of years.
42:40We think they will mutually reinforce
42:42each other and provide researchers with
42:44new tools to investigate neuro circuit
42:46function in a way that we couldn't have
42:48imagined just a few 5 plus years ago.
42:52More recently,
42:53we just launched what we call the
42:54Brain Behavior Quantification and
42:56Synchronization Program or BBQS program.
42:58And here the goal is to develop and
43:01validate new tools for analyzing and
43:03precisely quantifying complex behaviors.
43:06That's a BBQ part.
43:07The S part is now then to synchronize
43:10this information with high resolution
43:12neural activity mapping to really start
43:15getting at causality between neural
43:17circuit activity and behavioral output.
43:20So we just launched a bunch of RF as
43:22or now we call them Notice of funding
43:25opportunities or Nofos covering non human,
43:27primarily non human models,
43:30human and clinical models.
43:33The efforts here are going to be
43:35coordinated through a data center
43:36and AI center.
43:37You could imagine that there's going
43:39to be a lot of information here
43:41that's going to require ever more
43:43sophisticated computational techniques.
43:45We had just had a workshop on
43:47developing on identifying opportunities
43:48and new sensor technologies.
43:50And we have an ongoing program throughout
43:53the Brain initiative in terms of how to
43:55store access and analyse these data.
43:57And together we hope to build a
43:59consortium watching that we've
44:00done for these other projects,
44:01for really developing and deploying
44:04new tools for understanding the
44:06circuit basis of behavior that
44:08that will complement and add to
44:10our current circuits program.
44:12So here's just an example of what
44:13one might be interested in doing.
44:14This is from Cory Miller's group at UCSD
44:18and he's looking at what what a marmoset
44:22does while it's capturing its prey.
44:24And they could break this down
44:26into three main behaviors,
44:27either capture and flight stalk,
44:29paws and lunge and mouth capture.
44:32And they can do this with markerless
44:35motion tracking pretty much in the
44:36in a kind of a naturalistic setting.
44:38And being able to quantitatively
44:40characterize behavior in this
44:41way is going to be critical for
44:43understanding the neural basis behavior.
44:45So there's just one example of what
44:461 can be thinking about with better
44:49tools for quantifying behavior.
44:52Similar efforts are going on to
44:55understand what I call awake behaving humans.
44:58So this is a group from the market,
44:59but this is our work from the Markovich
45:01and Sedana groups at UCLA where
45:03they've developed a miniaturized
45:05device that can be worn by the patient
45:07and this is called a neuro stack
45:10device to both record and stimulate.
45:12These are patients that are in the
45:14epilepsy monitoring unit and it's
45:15really great to be able to record
45:17from patients who are willing
45:19to participate in studies,
45:20but even better if they can get out
45:22of bed and start walking around and
45:24navigate the space that they're in.
45:26It's a wireless mode that can record
45:28single neuron activity and also
45:30combine this with eye tracking and
45:32other parameters having to do with
45:35movement and and other behavioral
45:37and physiological outputs.
45:38They have this onboard processing unit
45:40that enables real time inferences
45:42that can be used in closed lip
45:44stimulation for the treatment as well.
45:46So this is very exciting stuff.
45:48These kinds of studies are what
45:50we're hoping to be supporting as
45:53we launch the BBQS program.
45:55And this gets us to the window
45:56of opportunity we have for
45:58understanding human neuroscience.
45:59The human brain,
46:02for patients who are undergoing
46:03treatments for other disorders
46:05and who are willing to participate
46:07in research with us is really a
46:09window into the the human brain
46:11to understand processes like a
46:13memory and emotion and so forth,
46:16which kind of gets us into some interesting
46:18space about the ethics of what we're doing.
46:20We'd have to be aware not
46:22only of issues of safety,
46:23but also privacy agency and so forth.
46:26We take this very seriously.
46:27The BRAIN Initiative,
46:28we have a neuroethics working group
46:30that helps advise us about upcoming
46:32as well as current challenges.
46:34We have a set of new ethics
46:35guiding principles.
46:36We hold workshops each year to examine
46:39these issues to hopefully stay
46:41ahead of any potential issues that
46:44might arise based on getting into
46:46a frankly really new new territory.
46:48So we take this very,
46:49very seriously.
46:51Another issue that we're very,
46:53very much interested in supporting
46:55is is data science and informatics.
46:57The legacy of any scientific
46:59project really is going to
47:01be in the data and resources
47:03that we leave behind.
47:05I think you can appreciate that
47:06from the studies I just showed you.
47:08We're generating a ton of data and
47:11we've set up eight brain data archives.
47:13I'm just showing 6 here.
47:15And our our mission is really to
47:17organize them so they can be findable,
47:19accessible, interoperable and reusable.
47:21It's a huge challenge.
47:23I imagine many of you are vexed by the new
47:27NIH's data management and sharing policy,
47:29and we will try to work very hard
47:31with you to make this happen,
47:32but we're at least a year ahead
47:34of the curve for the rest of NIH.
47:35But again,
47:36this is something we're taking
47:38very seriously and we really try.
47:39We need to be able to make these
47:42data not only accessible but
47:45also interoperable between these
47:47different modalities.
47:48And finally,
47:49one of the very important things that
47:51we're getting into is building a
47:53stronger and more sustainable workforce.
47:55Numerous studies have shown
47:56that diversity of thought leads
47:58to better outcomes,
47:59more creative solutions to tough problems,
48:01and I think the human brain is the
48:03toughest problem we can work with.
48:05We have a couple of different approaches,
48:08funding,
48:08funding mechanisms for
48:10for developing a stronger,
48:13more diverse workforce.
48:14We're doing capacity building with
48:16various programs and very importantly,
48:18a couple of years ago we introduced
48:20what's called the Plan for Enhancing
48:22Diverse Perspectives where we require
48:24applicants to tell us how they're
48:26going to use diverse perspectives in
48:28their team to to do better science.
48:31We have a number of funding opportunities,
48:33general funding opportunities at F32
48:35post doctoral awards as well as a
48:38Clinician Scientist Mentored Career award.
48:40Down below in the blue are
48:42diversity focus mechanisms.
48:43We just signed on to the Blueprint and
48:46Door program for getting kids in kids
48:49folks in at the undergraduate level
48:53F99K00 and K99R00 transition awards.
48:56The the first one is for late stage
48:58graduate students going out to do
49:00postdocs and the second one is similar
49:03to the Parent Pathway to Independence
49:05award for senior postdocs looking
49:07to transition to faculty careers.
49:10Again, these are diversity focused.
49:12In the case of the K9 and Nine award,
49:15it is restricted for folks from
49:17underrepresented groups and unfortunately
49:18in the case of BRAIN initiative,
49:20this also includes women,
49:21but we're really trying to address
49:23that through these awards.
49:24We also issue administrative supplements
49:26to get folks into the pipeline who
49:29hopefully can be competitive for these
49:31these types of funding mechanisms
49:33additional sets of opportunities,
49:34I I'm not going to go through all these here,
49:36but there's AQR code and you can ask us,
49:38we can send you the information
49:40on that really to bring in a more
49:43diverse and strong workforce.
49:44So I left you I think,
49:45with three key takeaways.
49:46I lied, there's actually 4 takeaways.
49:48And that is it's really important
49:50for us all to build,
49:51continue building the momentum to
49:54bring cures for devastating human
49:57brain disorders within our lifetime.
49:59You can keep up to date on what we
50:01do in the BRAIN Initiative if you
50:03don't mind a few emails and your inbox
50:05through these different channels.
50:07And the most important channel for
50:09getting information if you're interested
50:11in BRAIN Initiative funding is to
50:13find us to find a program officer who
50:15overseas that program and to call.
50:17Call them because we really do enjoy
50:19talking to folks and seeing how
50:21we can align your interests with
50:23our funding opportunities.
50:25And you can find us@rain.gov and I
50:28will stop there and take any questions.