Emily Jacobs “Applying dense-sampling methods to reveal dynamic endocrine modulation of the nervous system”

Transcript

• 00:06Our next speaker is Professor Emily Jacobs,
• 00:10and she's joining us
• 00:12remotely from Santa Barbara.
• 00:14So you're not in the snow,
• 00:17hopefully you went surfing or something,
• 00:19and we're happy to have you.
• 00:21Thank you. Thank you so much.
• 00:25So please just interrupt if you can't hear
• 00:27me if something goes wrong with the mic.
• 00:30So I just want to start off by saying I am so
• 00:34disappointed that I can't be there in person,
• 00:37you know, thank you to the organizers,
• 00:39to Todd and the whole organizing
• 00:41committee for this conference.
• 00:42It looks incredible and I have.
• 00:45At least top three FOMO moments of my life.
• 00:48So I am shaking my fists at Air Canada for
• 00:51cancelling my flight at the last minute.
• 00:53I'm I'm grateful that my trainee,
• 00:55Laura, was able to make it and I'm just.
• 00:57I'm distraught that I
• 00:59can't be there with you.
• 01:01So with that,
• 01:02the title of this talk is applying
• 01:04dense sampling methods to reveal dynamic
• 01:06endocrine modulation of the nervous system.
• 01:12So if I ask you to close your
• 01:15eyes and imagine the human brain,
• 01:18you might imagine something
• 01:20that looks like this,
• 01:21sort of like a £3.00 squishy walnut.
• 01:25Um, or because you are. You know,
• 01:28the audience is mostly human brain imagers.
• 01:30Maybe you imagine something
• 01:32a little bit more like this.
• 01:35But as my buddy John Morrison likes to say,
• 01:39sometimes neuroscientists can be so
• 01:41taken by the unique capabilities and the
• 01:45complexities of the human brain that we
• 01:48forget something really important about it,
• 01:51which is that it's connected
• 01:53to the rest of the body.
• 01:55So if you were to ask a neuroendocrinologist
• 01:59or Evan Gordon to imagine the brain,
• 02:02you'd probably imagine
• 02:04something more like this, right?
• 02:06The brain, you know,
• 02:07in its full richness and its full
• 02:10connection to the human body.
• 02:11And the endocrine system is a
• 02:14perfect example of this kind of
• 02:16whole body or brain body integration.
• 02:18And in this talk I'm going to focus
• 02:21specifically on gonadal hormones,
• 02:23which are produced by the ovaries
• 02:26and testes and use the circulatory
• 02:28system as their superhighway.
• 02:30These are highly lipophilic hormones.
• 02:32They travel through the blood and they
• 02:36bind anywhere there are adequate receptors.
• 02:40So in short, the brain is an endocrine organ.
• 02:44We've got 20 years of animal studies
• 02:47suggesting that gonadal hormones,
• 02:49in other words sex steroid hormones,
• 02:52are critical neuromodulators
• 02:53of learning and memory.
• 02:55And this cartoon here just shows
• 02:58the hypothalamic,
• 02:59pituitary gonadal or HPG axis.
• 03:02It's more famous sister is the HP access.
• 03:06But this HPG axis represents the
• 03:09interaction between our brain and our
• 03:12reproductive endocrine system, right?
• 03:14So there's this tightly coordinated
• 03:17neuroendocrine cascade in which about
• 03:191500 neurons in the hypothalamus.
• 03:22These are GNRH,
• 03:23Eric and Atrophin releasing hormone neurons.
• 03:25They regulate the production
• 03:28via the pituitary.
• 03:29They regulate the production of gonadal
• 03:32steroids or sex steroids from your gonads.
• 03:35So.
• 03:36Estrogen, progesterone,
• 03:37testosterone from the ovaries and testes.
• 03:40And these steroid hormones,
• 03:42as I mentioned,
• 03:43they're highly lipophilic.
• 03:44They travel through the bloodstream
• 03:45to bind to receptors in different
• 03:47tissues throughout the body.
• 03:49And we know that the brain is a major
• 03:52target organ for these sex steroid hormones.
• 03:54And in particular regions of
• 03:57the medial temporal lobe,
• 03:58including the hippocampus proper
• 04:00and the surrounding tissue and
• 04:02the prefrontal cortex contains
• 04:04significant populations of these
• 04:06steroid hormone receptors,
• 04:08and they may be particularly
• 04:10responsive to this. Hormone signaling.
• 04:15Some of that work comes from John Morrison.
• 04:17So he's the director of the National Primate
• 04:19Research Center at UC Davis right now,
• 04:20and he his lab showed a number of
• 04:23years ago now that estrogen receptor
• 04:26alpha R alpha is present in about
• 04:3050% of axo spinous synapses within
• 04:33the monkey prefrontal cortex.
• 04:35And this is true in young and
• 04:38in aged animals.
• 04:39And if you look at the abundance of
• 04:41sort of if you look at individual
• 04:43differences between animals just in
• 04:45a kind of crude correlational sense,
• 04:47we know that the abundance of synaptic
• 04:49ER alpha within the prefrontal
• 04:52cortex correlates with performance
• 04:54on a delayed response task.
• 04:56So this was some of the some of the
• 04:58early evidence at least within the
• 05:00non human primate literature that's
• 05:01something about estrogen receptor
• 05:03signaling may have something to do with
• 05:05these higher order cognitive functions.
• 05:08And this comes on the heels of work in
• 05:10the 90s, which I'll get to in a minute,
• 05:12showing effects of estrogen
• 05:15in the hippocampus.
• 05:16But but really what?
• 05:17What was groundbreaking about
• 05:19this was that if you crack open
• 05:21a neuroendocrinology textbook?
• 05:22From the 80s or even early 90s,
• 05:24all of it would be describing estrogen
• 05:27action within the hypothalamus or
• 05:29these really kind of routine um
• 05:31reproductive behaviors like lordosis.
• 05:33But we now know that these steroid
• 05:36hormones are acting in extra
• 05:38hypothalamic sites in the cortex.
• 05:42Umm, so these kinds of careful studies
• 05:44have been carried out in animal models.
• 05:46In humans, we know a lot less.
• 05:49So down here at UCSB, my lab is trying to
• 05:52get a toehold on how gay Natal hormones
• 05:55shape aspects of the human brain.
• 05:57And I don't have time to talk
• 05:59about most of what we're up to.
• 06:00So in this talk,
• 06:01I'm really just focusing on some
• 06:03of our dense sampling studies that
• 06:05are trying to map the brain across
• 06:07major in our endocrine transitions.
• 06:09For example,
• 06:11across the menstrual cycle and now pregnancy.
• 06:15So if you know nothing else about
• 06:18the endocrine system,
• 06:19know this hormone secretion varies over time,
• 06:23right. It's not static.
• 06:24There's there are these sort of pulsatile
• 06:27releases over different time scales.
• 06:29And one of these time scales in which
• 06:31we see these very sort of canonical
• 06:34changes in hormone production is
• 06:36over the menstrual cycle shown here,
• 06:38right, where there are these,
• 06:39again, this is just a cartoon,
• 06:40right,
• 06:40but there are these sort of increases in
• 06:43estrogen and these very sort of metronomic.
• 06:45Rhythms followed by an increase
• 06:47in progesterone.
• 06:49In the world of a rat or a mouse,
• 06:51as shown here,
• 06:52they don't have a menstrual cycle.
• 06:54They have something similar called
• 06:55an estrous cycle.
• 06:56This is happening over a four to
• 06:59five day period and really beautiful
• 07:01foundational studies conducted in
• 07:04the 1990s by Catherine Woolley
• 07:07was the first to link changes
• 07:10in hippocampal spine density to
• 07:13a rodent estrus cycle stage so
• 07:16the stage can be broken up into.
• 07:19Diestrus.
• 07:20When estrogen levels are very low,
• 07:22this is the start of the cycle and you
• 07:25can see fairly sparse spines within
• 07:27this is a C1 neuron within the hippocampus.
• 07:3024 hours later during proestrus
• 07:32when estradiol peaks.
• 07:34You don't even need statistics,
• 07:35right?
• 07:35Like just the the the graphics graph itself
• 07:38kind of tells you what you need to know.
• 07:40You see this proliferation of spines,
• 07:43and then the next day, following ovulation,
• 07:46estrogen levels fall and you see
• 07:48these spines again appear sparse.
• 07:52Now.
• 07:53One limitation of Catherine's
• 07:55experiments which is no fault of her
• 07:57own because they were done in the 90s,
• 07:59so she was limited to the
• 08:02technology of its time.
• 08:03One limitation is that these
• 08:05are based on Histology,
• 08:06so they are necessarily cross-sectional,
• 08:08right?
• 08:08You're sacking the animal,
• 08:09you're you're, you know, doing stains.
• 08:12These are camera lucida drawings.
• 08:14OK,
• 08:14so an open question is whether
• 08:17these morphological changes
• 08:18are evident across the estrous
• 08:21cycle in awake behaving animals.
• 08:27And so to answer that question,
• 08:28right now my lab is collaborating
• 08:31with Michael Gord's group at UCSB
• 08:33because they've developed super
• 08:35cool techniques for high resolution
• 08:37imaging of hippocampal neurons using
• 08:39two photon microscopy and awake mice.
• 08:42So they use this microprism so it's
• 08:46implanted kind of like a submarine
• 08:49Periscope flipped upside down,
• 08:51and they can image the hippocampal
• 08:53circuit at cellular resolution.
• 08:58So in a series of experiments
• 08:59led by Nora Wolcott,
• 09:01a graduate student in Mike's group,
• 09:04she imaged female mice every
• 09:0612 hours for eight days.
• 09:09This corresponds to two
• 09:11complete estrus cycles.
• 09:12And at each imaging session,
• 09:14she's performing vaginal
• 09:15cytology to stage the mice.
• 09:18This is all done blind,
• 09:19so she actually developed a machine
• 09:21learning classifier called Estrus net.
• 09:23So if anybody's doing it
• 09:25works for mice and rodents.
• 09:27If you want to stage animals
• 09:29and you don't know how,
• 09:30I I highly recommend this tool.
• 09:34So with longitudinal 2P imaging,
• 09:36she's able to track spines on
• 09:39apical dendrites of CA one
• 09:41neurons over several weeks.
• 09:43And I'll just note that these
• 09:45are the same segments of
• 09:47the dendrite every time.
• 09:49So it's not different sections,
• 09:50it's literally the same segment.
• 09:51So you can see new spines come and go.
• 09:57And consistent with
• 09:59Catharine's Classic result,
• 10:00Nora finds this increase in spine density
• 10:04during proestrus or this P period here.
• 10:08An increased pruning during estrus.
• 10:11So this band shown here,
• 10:13E and this ebbs and flows right?
• 10:15She sees it across the two cycles.
• 10:18Overall, there's about a 15% change in
• 10:21total spines from proestrus to estrus,
• 10:23which is, you know,
• 10:25potentially billions of synaptic connections.
• 10:28So it's clear. As sort of proof
• 10:31of concept that that you know,
• 10:33there's this modulation of spine
• 10:35density across the estrous cycle.
• 10:37But why? Right? So, So what?
• 10:40What? What are the consequences?
• 10:42Why is the brain organized in this way?
• 10:44Why are these rhythms happening?
• 10:48So umm, and don't hate me for
• 10:50showing all the rodent work
• 10:52because I'm so excited about this.
• 10:55It just came out.
• 10:55So bear with me and then I'll
• 10:57get to the human stuff.
• 10:58So the next step is really to
• 11:00see how these spine changes may
• 11:02influence functional properties
• 11:04of the hippocampal neurons.
• 11:06And so Nora just wrapped up experiments
• 11:08where she's imaging calcium activity
• 11:10and transgenic gcamp mice while they
• 11:13navigate in this floating circular chamber.
• 11:16So she can, so they're moving right?
• 11:19So she can map the response properties of C,
• 11:21A1 place fields overtime.
• 11:23So here's just a rendering of these chambers,
• 11:26and there's two different environments,
• 11:30let's call them A&B.
• 11:33So what you can see here in this panel,
• 11:37there are multiple place fields that respond
• 11:40to a particular part of the circular track,
• 11:43and here they've just been aligned to
• 11:45their preferred response in track a here.
• 11:47So you can see that these same
• 11:51cells respond very differently
• 11:53and track B this orange coating,
• 11:56but similarly when presented in track again.
• 12:00So you can look at 2 measures.
• 12:01So one is this measure of stability.
• 12:03So it's the similarity of the population
• 12:07response between A and a prime.
• 12:10Which we're coding here in dark blue,
• 12:12and you can also look at the remapping
• 12:14of place cells between two environments,
• 12:17so between A&B.
• 12:19And when estrogen is peaking in proestrus,
• 12:23so this panel here.
• 12:26We see great the greatest stability
• 12:30and the greatest remapping.
• 12:33Suggesting that these spine changes are are,
• 12:36you know,
• 12:37really having an effect on the
• 12:39functional output within this system.
• 12:42OK, what about in humans?
• 12:44You can we see, oh,
• 12:46and I'll just say, you know,
• 12:47we're following this up in collaboration
• 12:50with a group here at UCSB to do
• 12:53CRISPR to decrease the expression
• 12:55of specific sex hormone receptors to
• 12:57see if we can eliminate the effect
• 12:59and then and then we'll be done.
• 13:01OK. But what about in humans?
• 13:02So can we see morphological and functional
• 13:05changes across the menstrual cycle?
• 13:08So the lab had this crazy idea to image a
• 13:11person's brain every 24 hours for 30 days,
• 13:14collecting blood and brain scans and
• 13:16and mood scores at each time point.
• 13:21Because this is an informal conference,
• 13:22I'll show that, one editor later wrote.
• 13:25They're like old school NASA test pilots.
• 13:27They put themselves in an MRI
• 13:29scanner and drew their own
• 13:31blood every day for two months.
• 13:33That was a little colorful because
• 13:35we did have hire a certified
• 13:38phlebotomist and the first author,
• 13:40the test subject and the pioneer
• 13:42of this project, Laura Pritchett,
• 13:44is there in the audience.
• 13:46So big round of applause for her for giving
• 13:49herself over to science in a really.
• 13:51Dramatic way.
• 13:55OK, so why did we try to
• 13:57pull off this crazy study?
• 14:00Basically because cross-sectional
• 14:01studies that sample the brain
• 14:03at one time point can't tell
• 14:06us anything about how the brain
• 14:08changes day-to-day or week to week.
• 14:10And as I mentioned earlier,
• 14:12one central feature of the mammalian
• 14:15endocrine system is that hormone
• 14:17secretion varies over time.
• 14:19So during the average human
• 14:21menstrual cycle shown here,
• 14:22these are data from Laura.
• 14:25So we can see there's about an 8
• 14:28fold increase in estrogen which
• 14:30is plotted in this teal color.
• 14:32And there's actually about an 80
• 14:35fold increase in progesterone,
• 14:3780 fold change in progesterone
• 14:39plotted in blue.
• 14:40Note that these are on different scales.
• 14:43And the kinds of cross-sectional
• 14:45studies that are typically used in
• 14:47human brain imaging studies can't
• 14:49capture these dynamic endocrine changes.
• 14:51So we were really inspired by the
• 14:53dense sampling work pioneered by folks
• 14:55like Poldrack and the Midnight Scan Club,
• 14:58because that kind of intensely
• 15:01longitudinal imaging lends itself
• 15:03beautifully to our problem,
• 15:05which is trying to understand
• 15:07how these endocrine rhythms
• 15:09may be influencing the brain.
• 15:11And I think it's worth noting that
• 15:13for most women, for most of our lives,
• 15:15this ebb and flow is as steady as the tides.
• 15:19This pulse is almost like a vital sign,
• 15:21right?
• 15:22We know that these rhythmic changes Dr.
• 15:24Physiological functions.
• 15:25But nobody really knew how
• 15:27they shaped the brain,
• 15:29so the 28 me project was
• 15:32designed to to determine that.
• 15:34OK,
• 15:34so as I mentioned,
• 15:35the participant had daily blood draws,
• 15:39time locked and MRI every 24
• 15:42hours for 30 consecutive days
• 15:45across one complete cycle.
• 15:46And then she did the whole thing
• 15:48again a second time one year later.
• 15:54So the data that I'm going to show you first
• 15:57are from a 10 minute resting state scan.
• 16:00I'll note that she had very little motion,
• 16:04less than 130 microns on average each day.
• 16:08And to start out with so,
• 16:11so here our cortical parcellation INS,
• 16:13we're defined from the 415
• 16:15node Schafer Atlas plus the 15
• 16:18subcortical Harvard Oxford Atlas.
• 16:20And to begin with,
• 16:22we wanted to just test this
• 16:23straightforward hypothesis that
• 16:25whole brain resting state functional
• 16:28connectivity may be associated with
• 16:30intrinsic fluctuations in estradiol and
• 16:33progesterone in a time synchronous,
• 16:35so just day by day fashion.
• 16:40And sure enough that, you know,
• 16:41the first thing we observed is
• 16:43that these endocrine changes
• 16:44impact widespread patterns of
• 16:46connectivity in the human brain.
• 16:48So here we see increases in estradiol
• 16:50over time are associated with
• 16:52greater functional connectivity
• 16:53across much of the cortical mantle.
• 16:58Here I'm showing that same finding,
• 17:00just a slightly more stringent threshold.
• 17:03So hotter colors here indicate increased
• 17:07coherence with higher hormone concentrations.
• 17:10Cooler colors indicate the reverse.
• 17:13So one thing you might notice is that
• 17:15you know in contrast to estradiol
• 17:17sort of proliferative effects,
• 17:19we see that progesterone is associated with
• 17:22reduced coherence across the whole brain.
• 17:24And this although it's on a totally
• 17:27different scale fits really beautifully
• 17:29with the rodent work where we know that
• 17:32estradiol tends to be proliferative and
• 17:33progesterone tends to be sort of inhibitory.
• 17:36So we're seeing that on a completely
• 17:38different scale but but the sort
• 17:40of direction of the patterns are
• 17:41it's interesting to note those.
• 17:43Salaries.
• 17:45So next Laura calculated mean Nodal
• 17:48association strengths by intrinsic networks.
• 17:52So here positive just refers to the average
• 17:55magnitude of those positive associations,
• 17:58so stronger coherence with higher estradiol
• 18:01and negative refers to the average
• 18:04magnitude of the inverse association,
• 18:07so weak or coherence with higher estradiol.
• 18:10So you can see that these associations are
• 18:13are evident throughout the major networks,
• 18:16perhaps particularly in dorsal
• 18:20attention and temporoparietal.
• 18:24So we dug in a little bit more.
• 18:27So next we used time lag methods
• 18:29to start to discern the temporal
• 18:32directionality of these associations.
• 18:34So is it estradiol that's driving
• 18:36changes in brain state or might
• 18:39it be the other way around?
• 18:41And so in an analysis led by Tyler Santander,
• 18:44we used vector Autoregression or VAR models
• 18:47to just solve for two basic equations.
• 18:49So is like today's resting
• 18:52state functional connectivity.
• 18:53Um, you know,
• 18:54best.
• 18:55Determined by the pattern of
• 18:57yesterday's functional connectivity.
• 18:59Yesterday's estradiol levels were then
• 19:01you can go sort of to a lag of two back.
• 19:04And similarly we can solve for estradiol
• 19:06in the same way to try to get I'm
• 19:08not going to use the word causal,
• 19:10but to try to understand how these
• 19:13relationships are directed in time.
• 19:16And overwhelmingly we find evidence
• 19:18that it is estradiol.
• 19:20Previous states of estradiol driving
• 19:22current states of the brain,
• 19:24so driving this increase in
• 19:26functional coherence.
• 19:27And here we start to see really
• 19:30default mode control dorsal
• 19:31attention networks being the most
• 19:34strongly associated with estradiol.
• 19:38You know, we can use common
• 19:40graph theory metrics to start to
• 19:41characterize the network properties.
• 19:43So here I'm just going to show
• 19:45one example looking at within
• 19:46network connectivity or efficiency.
• 19:51And we find.
• 19:52So just to Orient you here,
• 19:55estradiol is being plotted
• 19:56by day of experiments.
• 19:58So to the best of her abilities,
• 19:59Laura did this blind.
• 20:00So she started randomly at
• 20:01a certain part of the cycle.
• 20:03So we're just, I'm plotting estrogen
• 20:06based on day of experiment,
• 20:07not day of cycle.
• 20:08And what you can see here is a
• 20:11measure of within the default mode,
• 20:13this measure of global efficiency.
• 20:15And what you see is that when estradiol
• 20:18levels peak right around ovulation,
• 20:21you see this increase in the
• 20:22efficiency of default mode network.
• 20:24When estradiol levels plummet,
• 20:26so too do you see this decrease
• 20:29in the efficiency in this network.
• 20:31And these data,
• 20:32I don't have time to get into it,
• 20:34but I referenced the paper down here.
• 20:36So that replicated in the follow-up
• 20:39study conducted one year later.
• 20:41And just a quick plug,
• 20:43I don't have time to do the data justice,
• 20:45but Rick Betzel's group did a
• 20:47really awesome Edge Time series
• 20:49analysis of this data set,
• 20:51finding that the frequency of
• 20:53high amplitude network states
• 20:55are associated with various HPG
• 20:57access hormones as well.
• 20:58And so here's a plug for that paper.
• 21:04And I think I'm going to skip over
• 21:06this stuff in the interest of time.
• 21:11Because I want to make this point.
• 21:14Intrinsic fluctuations in sex
• 21:16hormones are associated with patterns
• 21:19of brain coherence in both sexes.
• 21:21So I do not want you to walk away
• 21:24from this talk thinking that the
• 21:27menstrual cycle is, you know,
• 21:29pick up a newspaper and you'll hear things
• 21:31like wreaking havoc on the female brain.
• 21:33That couldn't be farther from the truth.
• 21:34Hormones are a feature of males and females,
• 21:38and these, you know,
• 21:40rhythmic pulses exist in both sexes.
• 21:42So here we ran a follow-up study.
• 21:45The densely sampled male and this
• 21:47time the participant underwent
• 21:49brain imaging and venepuncture to do
• 21:52serology every 12 to 24 hours across
• 21:5530 days so that we could target these
• 21:58diurnal changes in steroid hormones.
• 22:00So during this AM to PM diurnal cycle,
• 22:03we see about a 60% change
• 22:05in testosterone production,
• 22:07a 40% change in estradiol.
• 22:12And just like before,
• 22:13we found that increases in testosterone
• 22:16and estradiol in the male brain is
• 22:18associated with increases in whole
• 22:21brain functional coherence overtime.
• 22:23And the overall magnitude of these
• 22:26relationships in the densely
• 22:28sampled male was comparable to
• 22:30the densely sampled female.
• 22:34So, you know, as Rebecca
• 22:37Shansky wrote in an editorial,
• 22:39this is about the animal literature,
• 22:41but it applies here. You know,
• 22:43our hormones a female problem for research?
• 22:46Absolutely not. And I can get into that
• 22:48later in the discussion if you want.
• 22:50But you know, these are are features
• 22:53of of all genders and all sexes.
• 22:56OK. Next we shifted gears
• 22:58to look at brain morphology.
• 22:59So can we see dynamic changes in
• 23:02brain volume across the cycle?
• 23:04So this analysis is led by
• 23:05Caitlin Taylor of my group.
• 23:07So she used high resolution T2
• 23:09imaging to look at the volume
• 23:11of different sub regions of the
• 23:14hippocampus and the surrounding
• 23:16tissue across the menstrual cycle.
• 23:18And then she reran the experiment.
• 23:21But this time the participant was on
• 23:23a drug that chronically suppressed
• 23:26progesterone levels by 97%.
• 23:28So notice that this increase in
• 23:31progesterone across the spontaneous
• 23:33or natural menstrual cycle is
• 23:35absent in the second study,
• 23:37whilst estrogen dynamics were the same.
• 23:42And.
• 23:44If anybody,
• 23:44I don't have the comments turned on,
• 23:46but if anybody can guess what.
• 23:49Medication she was on.
• 23:50What drug she was on.
• 23:51It's the most common form of
• 23:53the oral hormonal contraceptive
• 23:55on the market today.
• 23:59So what do we find across
• 24:01the menstrual cycle?
• 24:02We see that intrinsic fluctuations
• 24:04in progesterone are associated with
• 24:06volumetric changes in several regions.
• 24:09Here I'm just showing two.
• 24:11So when progesterone concentrations
• 24:13are elevated in the luteal phase,
• 24:16here we see increased or an
• 24:18expansion of Gray matter volume in
• 24:21hippocampal C23 and parahippocampal
• 24:22cortex and then if you chronically
• 24:25suppress progesterone as we did.
• 24:27In the second experiment,
• 24:29you abolish this effect.
• 24:34That's just the same thing.
• 24:37So, you know, I hope that this
• 24:40begins to convince you that these
• 24:42neuroendocrine transition states,
• 24:44you know, are having fairly rapid or
• 24:46dynamic effects on the human brain,
• 24:48both structural and functional level.
• 24:50But you know, the endocrine changes
• 24:53that we see across the menstrual cycle
• 24:56are dwarfed by the kinds of endocrine
• 24:59changes that happen during pregnancy.
• 25:01So this really LED us to
• 25:04ask the next question,
• 25:06which is how is the brain reorganizing
• 25:10or changing across this utterly
• 25:13fascinating window of time?
• 25:15And we were really inspired by
• 25:18there's a great paper by Hexana Elson,
• 25:21hexamer and colleagues and a group
• 25:24in Spain looking at comparing the
• 25:26brain preconception and then post
• 25:28delivery and finding all of these
• 25:30interesting Gray matter volume.
• 25:32Differences,
• 25:33specifically within the default mode network.
• 25:37But we were interested in in sort
• 25:39of mapping on the sort of endocrine
• 25:41drivers and the time course with
• 25:42which these changes were happening,
• 25:44which really means we have to
• 25:45apply this dense sampling lens.
• 25:47So this is an ongoing study,
• 25:50but we recently wrapped data
• 25:51collection on on a single individual.
• 25:53So I'm going to show those data here.
• 25:55So we followed a woman starting
• 25:58preconception with MRI's and
• 26:00serological evaluations about
• 26:01every two weeks across the complete
• 26:04gestational window and we have
• 26:06follow-up scans up to two years.
• 26:08Postpartum and I'm I'm gonna save
• 26:10those data for a little bit later,
• 26:12but you can see that sex hormone
• 26:15concentrations increase across
• 26:16the course of pregnancy.
• 26:18So we're putting estrogen and
• 26:20progesterone here and then there's this
• 26:23precipitous drop after the delivery.
• 26:26And across pregnancy,
• 26:27we see changes in medial temporal
• 26:30lobe morphology here plotting
• 26:33parahippocampal cortex with sort of
• 26:36decrease in volume tied both to this
• 26:38change in estrogen and progesterone.
• 26:40Also you can plot it by just
• 26:43gestational week.
• 26:44Suggesting that this might be
• 26:46this period in which there are
• 26:48organizational effects of sex hormones
• 26:51across this gestational period.
• 26:53So again,
• 26:53I hope you're beginning to get a sense of,
• 26:55you know,
• 26:56how responsive the brain appears to be
• 26:58to these major endocrine transitions.
• 27:01And if you're at SRCD in a couple of
• 27:04weeks or CNS a couple days after that,
• 27:07stay tuned because the lab will
• 27:09present some more of the sort of
• 27:12multimodal findings from this project.
• 27:14Including crazy results looking
• 27:16at cortical thickness changes and
• 27:19just the functional connectome
• 27:21changes across this period of time.
• 27:26And you know I'll just say quickly
• 27:30that they're going to menopause later
• 27:32in life is another period of you
• 27:35know major and and neuroendocrine
• 27:37transition where steroid hormone
• 27:40production of estrogen progesterone
• 27:42levels declines by about 90% and you
• 27:45see this increase in the gonadotropin
• 27:48follicle stimulating hormone which
• 27:50increases fairly dramatically
• 27:52during this period of time we see.
• 27:55Seasonal. Let me move, boish.
• 28:00Does anybody hear that?
• 28:07OK, I'm going to keep going,
• 28:08but I heard somebody talking.
• 28:10So I'm just going to end by saying that,
• 28:12um, so, you know.
• 28:14We see changes in these same subfield
• 28:17regions as a function of endocrine aging,
• 28:21and I'll save those data for
• 28:23later and let me just end on this
• 28:26little bit of a soapbox point.
• 28:28This is from an opening of a
• 28:30special issue of Frontiers and their
• 28:33endocrinology that Lisa Gallia and
• 28:36Annmarie Delange and I put together.
• 28:39And we write that, you know,
• 28:41neuroscience has overlooked
• 28:42aspects of the human condition,
• 28:44whether talking about the menstrual cycle,
• 28:45the pill, pregnancy, menopause.
• 28:47That's relevant to half
• 28:49of the world's population.
• 28:51And half of the US tax base and
• 28:54we've got to correct course
• 28:56us for the field to advance.
• 28:58And just to put some some data to
• 29:01that as Nora Imager as you've all
• 29:04witnessed this stunning growth in
• 29:05neuroimaging studies of the human
• 29:07brain over the last 30 years,
• 29:09let's call it the blue wave.
• 29:11In yellow as a reference point,
• 29:13you can see the number of brain
• 29:16imaging publications on depression.
• 29:17And then if you squint really hard,
• 29:20you might be able to detect the
• 29:22black line at the bottom.
• 29:24This is the totality of articles
• 29:27covering whole suite of factors,
• 29:29including menopause, pregnancy,
• 29:30hormonal birth control,
• 29:31the menstrual cycle, and much, much more.
• 29:34So altogether,
• 29:35studies on these Women's Health
• 29:37factors constitute less than half of
• 29:391% of the total brain imaging literature.
• 29:43And I think by ignoring these factors,
• 29:45we risk making two major mistakes.
• 29:47First, the brain is an endocrine organ,
• 29:49and yet we have almost no insight
• 29:51into how dynamic changes in sex
• 29:53hormones shape the human brain at,
• 29:55you know,
• 29:56the mesoscopic and macroscopic scales
• 29:58discernible through brain imaging.
• 30:00And second,
• 30:01we're missing critical clues about
• 30:03why these endocrine transition
• 30:05states may be periods of resiliency
• 30:07or vulnerability, right?
• 30:09Think about postpartum depression,
• 30:10perimenopausal depression,
• 30:11the fact that women make up
• 30:122/3 of the Alzheimer's disease.
• 30:14Population.
• 30:15You know,
• 30:15right now I think the human brain
• 30:17imaging community is underserving women.
• 30:20Or to put it in a more positive light,
• 30:22let's say progress in neuroscience
• 30:24will flourish when when the health
• 30:26of men and women are valued equally.
• 30:29So I'm a firm believer in Katarinas's
• 30:32model of kind of go big or go home.
• 30:36And so we're taking that to heart.
• 30:38My lab does a lot of these sort of,
• 30:40you know,
• 30:41small and focused studies to map the
• 30:43brain across these transition states.
• 30:45But we also want to to take
• 30:49the other approach.
• 30:50And last month,
• 30:51we secured funding to launch the and
• 30:53Bowers women's Brain Health initiative.
• 30:55And the idea is to take
• 30:57Women's Health prime time.
• 30:58And at the University of California,
• 30:59I think we can do that through
• 31:02deeply collaborative science.
• 31:03So across the University
• 31:04of California system,
• 31:05we have eight brain imaging
• 31:07centers dedicated for research.
• 31:08These centers generate data from
• 31:10thousands of participants annually,
• 31:11and so we're capitalizing on that
• 31:14activity by creating a population level.
• 31:16Open Access brain imaging database
• 31:19designed unabashedly and specifically to
• 31:21strengthening Women's Health research,
• 31:23so pooling data collected
• 31:25across each of these sites.
• 31:28And Russ has graciously agreed to
• 31:30serve as our data coordinating center.
• 31:33So this is all quite new and and in
• 31:35fact I will probably be reaching out
• 31:37to a number of you in the audience
• 31:39to be part of our Executive Advisory
• 31:41Board as we as we grow this baby,
• 31:44but it is now happening and I'm thrilled.
• 31:47And with that,
• 31:48I'll just give thanks to all of the
• 31:51people who actually did this work,
• 31:53including the biggest shoutout
• 31:54to the one in the middle,
• 31:55Laura Pritchett for leading
• 31:57many of these studies.
• 31:59And with that,
• 32:00I will take questions.
• 32:07Thank you so much. Do we have any questions?
• 32:11Yeah.
• 32:14So great talk. Thank you.
• 32:15Very super interesting.
• 32:18I had a question not about Women's
• 32:21Health perceive but more about the time
• 32:24dependency of connectivity in other words.
• 32:27My tendency mean people get scanned
• 32:29in the morning versus afternoon and
• 32:31I have not seen that controlled
• 32:33course of compound any kind of
• 32:35any study that I've ever reading.
• 32:36Do you think that difference is
• 32:39significant enough for you to start
• 32:41worrying about it looking at it closely?
• 32:44Thomas's group had you know,
• 32:45time of day paper out,
• 32:47so you should pick his brain about that.
• 32:49So yeah, I think.
• 32:53I think I would certainly consider
• 32:55it as a factor and make sure that
• 32:57it is not systematically biased by
• 33:00the factor that you're interested
• 33:02in or by your groups.
• 33:04You know,
• 33:04I think one of the reasons you know
• 33:08that this question came up was,
• 33:11you know,
• 33:11doesn't matter where a woman is in
• 33:13her menstrual cycle when you scan her.
• 33:15And you know by chance you're going
• 33:18to have hopefully some kind of random
• 33:20distribution which may wash out your effect.
• 33:23But I think what these data suggests
• 33:25is that to a certain extent,
• 33:27during certain epochs,
• 33:28certainly during that ovulatory period,
• 33:30we are seeing pretty striking changes,
• 33:33you know?
• 33:34Happening of either the morphological
• 33:36and the functional level and.
• 33:38Here I'm just trying to find this slide.
• 33:43You know in the the 20 and he data set is,
• 33:46you know, suggesting too that at least
• 33:48whether it's time of day driven by
• 33:50something else or whether it's, you know,
• 33:53these clear diurnal changes in hormones
• 33:55that these are also affecting properties
• 33:58of the functional connectome so.
• 34:01You know, we always time lock our scans,
• 34:04but that's because, you know,
• 34:05we're interested in these
• 34:06changing endocrine functions.
• 34:07So I don't know whether I want
• 34:09to advise you to do the same,
• 34:10but it's certainly something that I
• 34:11would pay attention to in your data set.
• 34:17Sorry, I can't see who's asking questions.
• 34:20OK. I have a question 2 actually.
• 34:22One is did you collect any
• 34:25behavioral data in the 28 and me?
• 34:27And then the second question,
• 34:29now that you're going you know,
• 34:31more large scale and how are
• 34:33you going to control or maybe
• 34:36look at depending on you know,
• 34:38your perspective of the individual variation
• 34:40and you know hormonal fluctuations
• 34:42when you are now looking at you know?
• 34:45Thousands of people.
• 34:47Yeah, great.
• 34:48Great questions.
• 34:49So in the first in Laura's 28 and me study,
• 34:53we got that out into the field very
• 34:58quickly and did not spend the time
• 35:02necessary to think about how we could
• 35:06collect behavioral data that we were
• 35:08interested in or that we thought
• 35:10could be robust to 30 days of testing.
• 35:13We could have done it,
• 35:14we just didn't.
• 35:15So we have if you look at.
• 35:17Her paper we have.
• 35:20Characterize sort of mood changes
• 35:23and you know sleep etcetera.
• 35:25So just state changes in those,
• 35:27but we don't have very rich assessments of
• 35:31changes in in cognitive behavior over time.
• 35:35We do have data.
• 35:39So that said, we do have.
• 35:43Data from a selective attention task
• 35:45that she performed in the scanner.
• 35:48So her performance is pretty
• 35:50much near ceiling the whole time.
• 35:52It's not fluctuating,
• 35:53but we do and we haven't actually
• 35:54looked at this data yet,
• 35:55but we do have the ability to
• 35:58look at kind of.
• 35:59The brains response to
• 36:01that paradigm overtime so.
• 36:04We're just swimming in data and
• 36:06haven't gotten a chance to do that yet.
• 36:08And then the second question.
• 36:09Yeah how in the big data set,
• 36:12how do you control for these
• 36:13hormones questions?
• 36:14This is this is the biggest conundrum
• 36:16that we're having right now.
• 36:17I you know, we're we're exploring a lot
• 36:21of different options including just.
• 36:24You know,
• 36:25at UCSB we have a phlebotomist
• 36:26who just lives at the brain
• 36:28imaging center and draws blood on
• 36:29and everybody that we bring in,
• 36:31whether we can do that across
• 36:33all centers is TBD,
• 36:34but we're there's no option
• 36:36that's off the table.
• 36:38So in the pooling of the brain imaging data,
• 36:42we pair that with deidentified
• 36:43health information that looks
• 36:45across the life course to get
• 36:47information about women's.
• 36:50Whether they're natural cycling,
• 36:52whether they're on some kind of hormone based
• 36:55medication features of pregnancy et cetera.
• 36:57So we have those data and now we're
• 37:01we're thinking about ways of pairing
• 37:03it with time locked hormone assessments
• 37:06and saliva unfortunately is no good.
• 37:08So that would be the easiest solution,
• 37:10but it's, it's not great.
• 37:13For what we're interested in.
• 37:16That terrific talk with my
• 37:18friend Lisa Galea was here.
• 37:20She'd be joining you on your
• 37:23soapbox and plotting.
• 37:24That was fantastic.
• 37:25I've got a sort of mechanistic question,
• 37:28I guess.
• 37:30So if you think that the human variation
• 37:36across the cycle in synaptic density,
• 37:39for example, is similar to the mouse
• 37:42variation within the estrous cycle,
• 37:45which is like.
• 37:47We're doing something to something.
• 37:49That that's a lot, yeah.
• 37:51So that's a lot of spine density changes,
• 37:56right.
• 37:58Which means, you know,
• 37:58if you think about C1 or
• 38:00C2 or something like that,
• 38:01that that's a lot of human experience
• 38:04that is being dissolved and then
• 38:07coming back and yet you know we are
• 38:09who we are and we don't change that.
• 38:12You know,
• 38:12there may be behavioral changes but
• 38:14you know our long term memories
• 38:16and so on don't change.
• 38:17So what, what do you think these
• 38:20changes actually subserve?
• 38:21It's kind of a philosophical question
• 38:24but that's a lot of pruning and growth.
• 38:27Without a.
• 38:29Obvious reason for it.
• 38:32And who are you again?
• 38:37Oh, you can't see me. I'm Ravi Menon.
• 38:39I'm a professor at Western University.
• 38:42Yeah. OK. Hi. Yes.
• 38:44Lisa is a dear friend.
• 38:46And and we are. We think about
• 38:48these things a lot together.
• 38:54You know, yeah. So this is sort of
• 38:56like ultimate level of causation,
• 38:58like why are these things happening?
• 39:02You know. It's the $1,000,000 question.
• 39:06I think it it's easier to describe
• 39:09why this might be happening
• 39:11in the rodent models, right?
• 39:13So maybe this is vestigial,
• 39:14maybe this is sort of carryover
• 39:15from just sort of mammalian brain.
• 39:17But, you know, in a rodent and
• 39:20certainly as evidenced in, you know,
• 39:22in some of Norah's work looking at
• 39:24the functional advantages of these,
• 39:26you know this.
• 39:28Affecting place field stability
• 39:31and let's say flexibility,
• 39:34you know I'm,
• 39:35we know that there are brain
• 39:37changes if you are in,
• 39:39you know in seasonally breeding animals
• 39:41across the seasons in Prairie voles.
• 39:44We know that there are these
• 39:46changes during mating seasons
• 39:48because they can traverse then a
• 39:51greater space in the environment.
• 39:53And you see these sort of growth
• 39:55in either sort of you know growth
• 39:57hippocampal volume or spine agenesis.
• 39:59Um,
• 39:59so there it's it clearly has this sort of,
• 40:02you know, functional role where you know,
• 40:05the animals either sourcing food,
• 40:08sourcing mates and they need a
• 40:10neuronal system that can support that.
• 40:13You know why we still see these
• 40:15ebbs and flows and in humans,
• 40:17um, you know that's less clear.
• 40:21But clearly these hormones are
• 40:23responsive not just across the cycle
• 40:26but as I mentioned very briefly,
• 40:29you know across these other endocrine
• 40:31transition states like pregnancy and
• 40:33there I start I think we can start to
• 40:36think about and and I'll just give a plug to.
• 40:39To CNS.
• 40:39If anybody's there,
• 40:40go and take a look at.
• 40:43At the rest of these data.
• 40:44But, you know,
• 40:45here's a period where the brain is preparing,
• 40:49you know, if you take an alibris woman
• 40:51who's never experienced pregnancy before,
• 40:54the brain is preparing for this completely,
• 40:57you know,
• 40:58pretty radical change in function.
• 41:01And what we're seeing is a lot
• 41:03of sculpting and remodeling
• 41:04within the default mode network.
• 41:06This echoes earlier work,
• 41:08as I mentioned from Ellen Hexamers group.
• 41:12So, you know,
• 41:12maybe this is this sort of fine tuning of
• 41:15those sort of theory of mind capabilities,
• 41:17because what more do you need to do
• 41:19as a parent than have the ability
• 41:22to empathize with this hungry,
• 41:24screaming,
• 41:24crying,
• 41:24you know thing and and to try
• 41:27to understand those mental
• 41:29states for the first time.
• 41:33Just a guess, but we'll we'll
• 41:36keep exploring those questions as
• 41:38we dig more and more into this.
• 41:40Data. No, thank you.
• 41:43I also have a question related to this.
• 41:46I'm interested particularly in
• 41:49neuroendocrine neuromodulation.
• 41:50So the results you see with the the
• 41:54direct imaging on the the rats,
• 41:56would they be replicated if we just
• 41:59have like neuronal organoids in vitro
• 42:02and you change the concentration of,
• 42:04I don't know, progesterone and estradiol,
• 42:06would we see this increase in spine
• 42:10numbers you know? This is the very basic.
• 42:15Response cell to changes in the
• 42:18concentration of the modulator?
• 42:20Or is it? I don't know.
• 42:24Great question.
• 42:24I don't know, I, you know I can.
• 42:27Say with some confidence.
• 42:28I don't think those experiments have
• 42:31ever been run in an organoid setup.
• 42:33It would be really,
• 42:35really interesting to do that.
• 42:38Yeah, thank you.
• 42:39Because yeah,
• 42:39there are no another looks.
• 42:43Yeah, I think that's a I think that's
• 42:46a great you know I again I think.
• 42:49These are pretty specific,
• 42:50so it's not like we're seeing,
• 42:52at least in this model we're not seeing.
• 42:56Every cell, right,
• 42:57every pyramidal neuron show these effects.
• 42:59You know these in these sets of
• 43:02experiments we're looking at,
• 43:04you know, CA one neurons, so.
• 43:07TBD on the the sort of extent to which
• 43:10we can see this remapping play in other,
• 43:14you know, regions.
• 43:15OK, at this stage it was only
• 43:17verified in C3 criminal neurons.
• 43:20It wasn't verified.
• 43:21It happens or not in other neurons.
• 43:23That's the thing.
• 43:29OK. Thank you so much.