Yale Psychiatry Grand Rounds: "Circuits, Neurotransmitters, and Electrophysiological Events Through Which Sleep Could Reset Emotions and Maladaptive Sleep Could Intensify Them"

May 19, 2023

Speaker: Gina Poe, PhD, Professor, Department of Integrative Biology and Physiology, UCLA

Information

ID: 9943
To Cite: DCA Citation Guide

Download Transcript

00:00I really appreciate that.
00:02I really appreciate that
00:04wonderful introduction.
00:04Thank you so much.
00:06I am so pleased to be here.
00:09I see familiar faces and names like Violet,
00:12Kimball, Hi, Violet and you,
00:14Al and Marina and other people.
00:18I really, it's feels like I am
00:20coming and joining some friends so.
00:23I want to, in the spirit of that,
00:26encourage you to ask me questions
00:28all along while I give this talk
00:30because I I will know then that
00:33you're listening and understanding
00:35and you're still with me and so
00:37and then of course at the end,
00:38I guess we have a whole 15 minutes of time.
00:41We'll see how much time I take,
00:43but let's get started.
00:44All right, so.
00:54Okay. Hey,
00:57did it, Did it, Did it?
00:58Did it show up?
01:00What's going on here?
01:01Something's happening.
01:02The share is not working very well.
01:06Or my our point is
01:08not working very well.
01:09Here we go.
01:21better. Yes. OK.
01:23So I had a very long title.
01:26And so I said forget it.
01:27I'm not typing all that anymore.
01:30Can you see my screen?
01:32Yes. All right. All right.
01:35So let me just start by saying
01:38that we used to think it was
01:40just mammals and birds that had
01:43really good two stages of sleep.
01:46And that was just because we weren't
01:48observing closely enough and we didn't have.
01:49Necessarily all the tools or the
01:51patients to watch all the different
01:54animals sleep through the night.
01:56But but now we know that even lizards
02:00have not only great sleep but two
02:02stages of sleep which includes non REM
02:05sleep and rapid eye movement sleep.
02:07That is their Dragon's lizard's eyes
02:10will rapidly move in a stage of sleep,
02:13even Drosophila fruit flies.
02:15Seem to have two stages of sleep and
02:18the and with the reason why there's a
02:20great group in New Zealand working on
02:23this is they have the quiet stage of sleep,
02:25just like we've always observed,
02:28although very few people
02:29wanted to call it sleep.
02:30But they also have a twitching stage where
02:32their limbs twitch and they don't have many.
02:35They don't have rapid eye movements so much,
02:36but their limbs twitch just like dogs
02:39and cats do and our limbs sometimes
02:42twitch in some stages of sleep.
02:44And then now even the jellyfish.
02:47There's a great study out of Caltech
02:50a few years ago that shows the
02:53jellyfish sleeps and and probably
02:55has two stages of sleep.
02:57Or at least that's how I interpret it.
02:59We don't know about the water bear,
03:01but it looked like it was sleeping.
03:02So I I put this picture in here.
03:06So sleep has to have a really
03:09good essential function.
03:10Here's jellyfish,
03:11which doesn't even have
03:12a central nervous system.
03:14It's this is the Cassiopeia,
03:15which is an upside down jellyfish,
03:17pulsing at its waking pulse rate.
03:21And then during sleep it pulses much,
03:24much more slowly.
03:25And you can disturb a jellyfish of
03:28sleep by giving it a pulse of a jet
03:31of of water to move it and then it
03:35will wake up and be annoyed and then
03:38quickly get back down to the bottom
03:40and and try to go back to sleep again.
03:42And if you do this a lot it will actually
03:45try and make up for that lost sleep
03:49the next day by taking many more naps.
03:52We could call it jellyfish napping,
03:54but anyway, this is its pulse rate during
03:55the day and the pulse rate at night.
03:57And this is a Figure 2, I believe,
04:00of this Current Biology paper.
04:02But it doesn't even mention the
04:04fact that in for, you know,
04:06good 20 seconds at a time,
04:08it's not pulsing at all,
04:09which means it's not breathing.
04:11And that could be equivalent
04:13to our stage REM sleep,
04:16which of course they don't have.
04:17Bias it move.
04:18But we have in our REM sleep a period
04:21of time when our muscles are atonic.
04:23We are actually not able to
04:26move because we are inhibiting
04:28the reactivation of our dreams.
04:30Or maybe there's another reason
04:31for atonia that we don't know yet,
04:33but but we don't know if
04:37jellyfish are dreaming.
04:38It would be cool to see what they
04:39are dreaming about if they were,
04:41but in any case, it does appear like
04:43they have at least two stages of sleep.
04:45And here is a new paper.
04:47In science just published a few weeks
04:50ago showing that elephant seals
04:52which are you know marine mammals,
04:55they unlike other things like
04:57fur seals or dolphins or whales,
05:00they don't sleep uni hemispherically.
05:02So those animals sleep have adapted
05:04by sleeping unit hemispherically.
05:06SO1 hemisphere is awake and keeping them
05:08at the surface and breathing while the
05:10other is sleeping and then they switch.
05:13But elephant seals don't do that,
05:15nor do we.
05:16And how they've adapted is that they
05:19dive quickly down past the point where
05:22sharks and killer whales would eat them,
05:25so they dive pretty darn deep.
05:27And then when they get to that past that
05:31depth that sharks and seals would get them,
05:33then they start sleeping.
05:34And they did this by recording their EEG,
05:37outfitting them with EEG,
05:38putting them back out there and their family.
05:41And when they go into
05:42the deep slow wave sleep,
05:44you can see their big slow waves by,
05:46you know, both hemispheres at the same time.
05:49And then when they lose muscle
05:50tone and go into rim.
05:52Usually one side of their body
05:53or the other is a little more,
05:55you know, Finn down.
05:56And so they start circling
05:57and they circle down,
05:58down, down,
05:59and for a good 10 minutes
06:01they're circling down.
06:03And then when they hit the bottom or
06:06just finish their REM sleep cycle,
06:08they'll wake up and swim back
06:10up to the surface.
06:11It's a really cool paper with a really nice.
06:15Video So you can just see this
06:18happening not in a live seal,
06:20but a model of what they've recorded.
06:23So here are the just the basic stages
06:26of sleep that you could see from
06:28something like a Fitbit or Apple Watch.
06:31It doesn't come with EE G because
06:33you need electrodes, you know,
06:34on the skull to be able to see this,
06:35but I just overlaid some EE G to
06:37look to show you what it looks like.
06:39So here we go,
06:41from wakefulness to stage two
06:43sleep with sleep spindles.
06:44To deep,
06:45slow wave sleep with big slow waves
06:47that sweep through our brain and then
06:50back into stage 2 with sleep spindles
06:53that come and go once every 10 or
06:5620 seconds and they are 10 to 15 Hertz.
06:58These are one to three Hertz,
07:00something like that.
07:01And then in REM sleep with the rapid eye
07:03movements where we're actively dreaming,
07:05we have in our limbic system,
07:07our emotional system,
07:08which we're going to talk about
07:09a lot more today,
07:10and we have a Theta rhythm that that
07:13takes over. And it's big. It's juicy.
07:16It's even more beautiful than
07:18you see during wakefulness,
07:20when people on animals are learning and
07:22paying attention to their environment,
07:24and it's induced by acetylcholine and
07:27gabourgic neurons of the basal forebrain.
07:30Really important,
07:31we know for learning and memory.
07:32So what's its function during REM sleep?
07:34So yeah, going to concentrate.
07:37So what happens with disturbed sleep?
07:39Well, you know, it's less lovely.
07:41There's a lot more wakefulness
07:43that that's interspersed.
07:44And that happens either from
07:46exogenous stimuli like we did to
07:48that poor jellyfish or or they did
07:50a Caltech to that poor jellyfish.
07:52Or it could come from internal
07:54sources like sleep apnea.
07:56Can wake people up 500 * a night
07:58and they won't even be aware that
08:00they woke up because it's so brief.
08:02They just have to wake up to breathe and
08:04then they go back to sleep and it's but it's,
08:06as you can imagine,
08:07profoundly disturbing in terms
08:09of the functions of sleep,
08:11which needs some continuity
08:13we found to proceed.
08:14So what happens if we don't get enough sleep?
08:17Well, I don't know about you,
08:19but I feel cranky and short tempered.
08:21Inflexible, hard to handle,
08:23impulsive and accident prone, in fact.
08:26All all causes of mortality increase the
08:30further away from 7 hours of sleep you get.
08:34The six hours,
08:35five hours four hours.
08:36If you get sleep 4 hours a night you
08:38become more accident prone and the
08:40all causes mortality becomes more
08:42due to accidents, car accidents,
08:44ladder accidents, whatever it is,
08:47we also are metabolism changes.
08:49We become our.
08:50Immune system changes,
08:51We get prone to infection and illness.
08:53Our memory is not as good.
08:54So we're going to talk about
08:56that a little more.
08:57We have less insight, more pedantic,
08:59less able to abstract,
09:00and more anxious and depressed and angry.
09:03And so adolescents unfortunately
09:05have a circadian misalignment,
09:08social jet lag really,
09:09and their preferred sleep
09:11time shifts later at puberty,
09:13so that during the week
09:15because of school start times.
09:17They are getting up too early for
09:20their and and depriving themselves
09:22asleep based on relative to the
09:24time that they went to sleep.
09:25They need just as much sleep
09:27as a 10 year old does.
09:29Their brains are still developing
09:30but they're not getting it
09:32because of that social jet lag.
09:33So they sleep deprived and then
09:35they do recovery sleep on weekends
09:37and they usually feel much better
09:39and happier if you let them sleep
09:41in but they're back to social jet
09:42lag during the week and and so.
09:47These sleep deprived teenagers as as
09:51well as anyone else who's sleep deprived,
09:54actually have more difficult time
09:58with negative emotional circumstances.
10:02So those who are sleep well rested
10:07actually have better forebrain
10:10prefrontal cortex control of
10:12amygdala activity and so.
10:16And this is a great paper by Michelle
10:18Krask and others have shown this.
10:20So there's more prefrontal control
10:23over amygdala activity during the
10:26presentation of emotional stimuli
10:29when we're well rested and less
10:31prefrontal control over all of
10:33that when we're not well rested.
10:35And so, you know, several papers have
10:38shown that really a good adaptive
10:41sleep reduces also our our fear,
10:43our anger, our aggression.
10:45And it increases our sense of good judgment,
10:48our rationality and our self-control.
10:50There've been really some fun
10:52psychology experiments with sleep
10:53deprived people where they give them,
10:55you know, cake versus salad
10:57and say what would you prefer.
10:58And well rested people go
11:00for the salad a little more,
11:02but of sleep deprivation they
11:03go for the cake. So
11:07let's see. So we also feel
11:10lonelier and less desirable.
11:11This is Eddie Ben Simon's.
11:14And Ben and Nat Walker's study,
11:16which shows that sleep deprived people
11:20actually distance themselves from others,
11:23physically distance
11:24themselves from others more,
11:27which was really an interesting study.
11:30So we other studies show that you
11:34know suicidal urges are mediated
11:36through a prior night's sleep quality,
11:39our impulsivity and our suicidal
11:41urges are all mediated that way.
11:43So, so there's kind of a vicious
11:48negative cycle feedback of
11:50emotional of disturbed sleep,
11:53emotional dysregulation,
11:54distress that then further disturbs our
11:57sleep because we're so distressed and.
12:01Yeah, it's a it's a bad,
12:03it's a bad situation.
12:05So just one more paper by Eddie
12:08Van Simon showing that it seems
12:10to be that deep slow of sleep,
12:12which is interesting.
12:14That's most associated with anxiety.
12:16So the more the less slow of sleep we get,
12:21the more anxious we feel.
12:23Yeah.
12:24OK,
12:24so I'm going to talk a little bit more
12:26now about what we're doing in my lab.
12:29This is my lab over the pandemic,
12:31commuting on zoom every day for
12:33a while there because of that
12:36sense of isolation that that
12:38the pandemic really gave us.
12:41So, so the function of that deep,
12:44slow way of sleep seems really
12:47clearly to clean and restore
12:49the energy of our brain.
12:52And going there needs to be a lot
12:54more studies but the studies that
12:56are pointing to the function of
12:58slow I sleep point to point to that.
13:00And then our new memories really
13:02seem to be consolidated,
13:03consolidated in the end two
13:05stage with the sleep spindles.
13:07I also call it transition to REM.
13:09So that's why I double labeled this year.
13:13And then we'll get into some
13:15circuits how our memories become
13:18familiar to us through actually
13:20distal dendrites in our neurons
13:22and then the proximal dendrites can
13:25be depotentiated once the memories
13:28have been consolidated out and and
13:30that refreshes our brain of able to
13:33learn new things the next day and
13:36and in that way it's our hypothesis,
13:39our working hypothesis right now.
13:41Our sensory and emotional circuits
13:44could actually become detached
13:46from the semantic and episodic
13:49versions of our memories,
13:51so so that they can be refreshed
13:54and learn new things the next day,
13:56and so that when we're recalling things,
13:58we can remember the facts of
14:00the emotion and the facts of the
14:03sensation without reexperiencing.
14:05The emotions and the sensations.
14:07As you might imagine,
14:08that would be awful if we could
14:11remember every pain we ever
14:12experienced and when we remember it
14:14we are re experiencing that pain.
14:17So this is our working hypothesis,
14:19the circuit mechanism for that,
14:23that adaptive detachment.
14:27All right,
14:28so so here's our general overview
14:31of our functions of slow asleep.
14:33We clear the debris through
14:35our lymphatic system,
14:37probably through the pumping action
14:39of those slow waves themselves.
14:41Each slow wave is is characterized
14:44by silence.
14:45Of cortical neurons,
14:46and then activity,
14:47the simultaneous activity of a
14:49bunch of them at the same time,
14:51and neurons shrink and swell
14:55when they're inactive and active.
14:57And the group function or action
15:00of that could actually physically
15:02pump out the intracellular and
15:04extracellular space into the
15:06glymphatic system we also know.
15:09There's a ton of protein synthesis that
15:12happens 5 to 10 times faster during
15:15slow wave sleep than than during other
15:18States and and and there's a lot of.
15:24Actually, I'll talk a little bit about
15:26the the role of the norepinephrine
15:29which fires the locust surrealist
15:31fires with every slow wave,
15:33and when norepinephrine is
15:35present it can actually prevent
15:37weakening and protect our memories.
15:39During that transition to REM
15:41sleep state with sleep spindles,
15:44we actually can transfer information.
15:46There's an and I'll talk about about this.
15:49There's kind of a unique connectivity
15:51between the hippocampus and the cortex
15:53during each of these sleep spindles,
15:55where the cortex seems to be listening
15:57to the hippocampus and responding
15:59to the hippocampal reactivations.
16:01And then during REM sleep,
16:03I'll tell you this is one of
16:05my first studies.
16:06We really can weaken old connections
16:08of those proximal dendrites that
16:10I just mentioned in the last slide
16:12and and strengthen new ones because
16:14there's a ton of plasticity that
16:16can happen during that Theta state.
16:19All right,
16:20so so here are the cycles of sleep.
16:22We go from waking to non REM sleep
16:24and through the transition to REM
16:26to REM and we go back and forth
16:29until the job of sleep is done.
16:31Different things happening in these
16:33different stages and we wake up.
16:34When I was a graduate student,
16:37I heard a talk by John Listman,
16:40who came to tell us that he'd
16:42taken a slice of hippocampus,
16:44added acetycholine to it to cause
16:46the Theta rhythm to have to happen,
16:49and then when he electrically
16:50stimulated the inputs to the
16:52hippocampus at the peaks of Theta,
16:55which is where most cells fire most
16:57of their spikes during wakefulness.
17:00He was able to get Long Term
17:01Potentiation with just four spikes
17:03at the peaks of 1 Theta Cycle,
17:04which was really cool and exciting
17:07because before that LTP could only be.
17:10Induced with, you know,
17:12100 Hertz for a solid second,
17:14which is not something you ever saw
17:16the hippocampus do spontaneously.
17:18So as far as the hypothesis that LTP
17:22was the building block for synaptic
17:24strengthening and learning and memory,
17:27there was some skepticism at the
17:29time because you never really saw
17:31anything that could induce it in a
17:33physiological manner until these papers.
17:35But and then he went on to say when
17:37he stimulated the troughs of Theta,
17:38which is when.
17:39The inside of the cell is most
17:42negative and least able to respond
17:45to the external stimuli input.
17:47Then he actually got a reversal of
17:49what was previously potentiated,
17:50which was exciting because in
17:53computational modeling theory
17:54depotentiation would be really
17:56important for not saturating your brain.
17:58You can see each one of these red dots
18:01is a is a synapse on this neuron and
18:03if all of them were potentiated then
18:05any stray incoming piece of information.
18:07Or anything coming from the
18:09outside world would just cause
18:11all of their cells to fire.
18:12There's they're all connected
18:14to one another eventually and
18:16you would just get white noise.
18:17So depotentiation might be a way
18:20to sculpt the memory circuits
18:22and and he showed how to do
18:24this. How one could do this with
18:27a very physiological stimulus just
18:28at the troughs of Theta and the the
18:31neurochemical environment of the
18:32slice is I would argue more like
18:35REM sleep than any other state.
18:37Because you have an absence of
18:41some neurotransmitters that come
18:43in from from distal parts.
18:45So like locus, cerilis brings
18:47norepinephrine to the forebrain.
18:49In a hippocampus slice you
18:51don't have that input.
18:52The dorsal rafae brings serotonin to
18:54the forebrain and a hippocampus slice,
18:56you don't have that unless you add it.
18:59They they did add acetylcholine
19:01which also comes from outside and
19:03and to get that Theta and so that
19:06is neurochemically the most like a
19:08REM sleep state where you don't have
19:10those norepinephrine and and serotonin
19:12inputs but you do have lots of acetylcholine.
19:16So when I told John Lisman that I
19:18was a graduate student, I said, hey,
19:20that sounds like program, sleep,
19:22neurochemical environment, he said.
19:24That's really interesting and that
19:25I didn't have anything more to say
19:27at the time because I wasn't in a
19:29learning memory field at the time I was,
19:31but I did think it was interesting.
19:32Then for my post doc,
19:34I was able to go and actually test
19:36out whether that was important.
19:38So here's the neurotransmitum,
19:40a year of the different sleep states.
19:42So wakefulness.
19:43You've got lots of acetylcholine,
19:45norepinephrine, serotonin,
19:46glutamate,
19:46all of that during slow wave sleep,
19:49the deep slow wave sleep state.
19:52The most
19:55striking feature is the
19:57lack of acetylcholine.
19:58The basal forebrain neurons that provides
20:01acetylcholine all over the brain are off.
20:04They're actively inhibited
20:05during slowing sleep and in unit
20:07hemispherically sleeping animals.
20:09It's acetylcholine that switches sides.
20:13Then during that transition to
20:14REM and two with sleep spindles,
20:16you get kind of what is seems to be
20:18to me the opposite of wakefulness,
20:21the lack of all of these neurotransmitters.
20:24No acetylcholine,
20:25norepinephrine or serotonin or
20:26levels are really, really low.
20:28And then during rapid eye movement,
20:30sleep is almost the opposite
20:31of slow way of sleep.
20:32You've got tons of acetylcholine but
20:35very little norepinephrine or serotonin.
20:38And all of these neurotransmitters have
20:41their function for learning memory,
20:44generating these these patterns that
20:46we are seeing here and and then I'm
20:50going to argue for for emotional control.
20:52So locus surrealists down there
20:55in the brainstem,
20:56these neurons don't fire during
20:58specific sleep states like I just
21:00showed you during REM sleep and
21:02that transition to REM which is
21:03also called intermediate sleep.
21:05You don't have,
21:06you don't have much firing of
21:07the local surrealist bringing
21:09norepinephrine to the forebrain.
21:11So here's the firing rate
21:13across the different states.
21:15And we don't know really about females
21:18because the ones ever studied them
21:21until we have very recently with some
21:24great preliminary data that we're
21:25about to amplify with a lot more.
21:28But anyway,
21:29this is where it exists in the brainstem,
21:31I'm sure.
21:32Doctor Al K has shown you this,
21:34but you know,
21:35in the brainstem of a rat it's
21:38here's the locus realist projecting
21:40its axons to all over the brain and
21:44in in a really beautiful fashion.
21:47And what norepinephrine does.
21:48One of the things that it does at
21:51the cell body is when it occupies
21:53the beta receptors,
21:55it causes a cascade of events that
21:58actually prevent depotentiation,
21:59So that depotentiation.
22:03Function if if cells are firing
22:05at the Theta troughs can't happen
22:07when norepinephrine is present,
22:08so the only time it can happen is
22:10during that transition to REM and REM
22:12sleep state when the slope surrealist
22:14isn't firing and not providing
22:16norepinephrine to the forebrain.
22:18All right, so I wanted to go to
22:21the University of Arizona and
22:23see what REM sleep Theta is for
22:26and firing during and REM sleep.
22:28Is it for learning and memory or
22:30for depotentiating and erasing?
22:32So we have this,
22:33you know tetrodes system where we
22:35can record from multiple cells at
22:36the same time in the hippocampus
22:38as animals are learning and running
22:41around in their environment and we
22:42can see how they fire in relation to
22:45that local field potential of Theta.
22:47Here's a task where we have rats
22:50running around on a on a maze and three
22:52of the boxes are baited with food.
22:54After a week of that we switch
22:56which boxes are baited,
22:57so they have to sort of relearn.
22:59And depotentiation becomes really
23:00important because we want them to stop
23:03checking the old boxes where food
23:05used to be and start checking the new ones.
23:08We can track which cells are firing
23:10where and see which cells are
23:12associated with encoding old box
23:15positions versus new box positions.
23:17So we can really see whether the cells
23:20are involved in encoding something new.
23:23And here is how the cells fire
23:26during wakefulness here is.
23:28Here's hippocampal cells bursts during
23:30when as it goes through a place field.
23:34So here's a place where this cell
23:37is encoding and you can see the
23:39most of the spikes are occurring
23:40at the peaks of Theta.
23:42As you can see this is the Theta phase
23:44as they run around and then they stop to
23:47eat and you can see Theta stops altogether.
23:51And then during REM sleep,
23:52the first data set I looked at,
23:53the cells are flying at the opposite
23:55phase of Theta at Theta troughs.
23:57So Francis Crick had put and Graham
24:00Mitchinson had put out a paper to say,
24:02hey, maybe REM sleep is for forgetting.
24:04And it is sort of belied decades
24:07of data where, you know,
24:09REM sleep seemed to be really
24:10important for memory consolidation.
24:12So it was kind of puzzling why are
24:16cells firing at Theta troughs?
24:18Consistent with what John Lisman had
24:20said is important for depotentiation
24:22when norepinephrine is not present,
24:25or which would be erasing memories.
24:28The next data set I looked at though,
24:30it was animals learning a new maze,
24:34and day after day they're running it
24:35Always during the learning session,
24:37the cells are firing at Theta peaks
24:39consistent with longterm potentiation,
24:41but they start firing at Theta
24:44troughs only after five or six days.
24:47Of running that novel environment,
24:50initially novel environment.
24:52And this was really cool to me
24:55because what this this time course
24:58is is consistent with the length
25:00of time it takes us to consolidate
25:02memories from
25:03the hippocampus to the neocortex.
25:05After which time you can lesion
25:07the hippocampus bilaterally after
25:09seven days and still get an animal
25:11a month later that remembers.
25:13The place you introduced it to on the
25:15first day. So that was the end to me.
25:17Like, so exciting.
25:18I thought maybe room sleep is for
25:20remembering or for forgetting,
25:21but in fact what seems to be occurring
25:23is in the first couple of days before
25:26the memories are fully consolidated.
25:28The hippocampus is still firing at
25:31Theta peaks consistent with Long
25:33Term Potentiation and only after.
25:35Enough time has passed for
25:37that consolidation to happen.
25:39Does it start firing at Theta
25:41traps consistent with erasing the
25:42memory from the hippocampus so
25:44the hipocampus can be refreshed?
25:46And learn something new the next day.
25:49So. So, yeah, so that's the idea.
25:52Temporary memory of the hippocampus is
25:54cleared in REM sleep to avoid saturation.
25:56This is my son when he was 18.
25:58Now he's 21.
26:00So do not deprive yourself of sleep.
26:02There's a really cool paper by
26:05from Antoine Adamantides' lab.
26:06Richard Boyce did it.
26:08And what they did is they reduced the
26:10amplitude of Theta by silencing the
26:13gabbergic cells in the basal forebrain
26:15that projected the hippocampus.
26:16And you can see Theta goes from big
26:19and lovely to about half amplitude.
26:21Here's the five to 10 Hertz
26:23frequency range of Theta.
26:25And you can see that when they
26:27did optogenetic inhibition
26:28of these Gabourgic neurons,
26:30you got Theta that was half
26:33amplitude at best.
26:34And when they did this,
26:35the animals couldn't learn
26:38object place memory task,
26:40which is hippocampus dependent.
26:42And they also couldn't do
26:44contextual fear memory.
26:45And that was just inhibiting Theta,
26:48and only during REM sleep in these animals,
26:51these rats,
26:52after introducing them to this new things.
26:56So,
26:58so yeah,
26:59so let's concentrate for a moment,
27:02But back from Theta to that transition
27:04to REM sleep with sleep spindles.
27:06And here are some papers by Ryzowski and
27:10Ceapus from Ceapus's lab at Caltech.
27:13And what he shows is that the
27:15the more the hippocampus fires
27:17in a burst mode during sleep,
27:19so these are the the more cells
27:22that are involved in in giving
27:25a burst that they're recording,
27:27the more the prefrontal cortex responds.
27:31And so this is the amount of response
27:33to the prefrontal cortical neurons and
27:35the time lag between one response to
27:38the next is the spindle frequency, so.
27:41The more hippocampus hippocampus fires,
27:45the more the prefrontal cortex
27:48responds with spindle frequency.
27:52Activity.
27:53So you can see that the spindles
27:55that occur in the prefrontal
27:58cortex are linked and responding
28:00to hippocampal activity.
28:01Here is another paper.
28:03Now we're going to get into dendrites again.
28:06So here's a pyramidal cells of the neocortex.
28:09This is a beautiful paper by Julie Seed
28:12and then a review by her and Perash.
28:15And what they show is that when
28:17animals are in that spindle state,
28:20which is an intermediate state
28:21of sleep transition to REM
28:23and two stage.
28:24And the more you have spindle activity
28:27which is the 9 to 16 Hertz activity,
28:30the more you have signs that calcium,
28:33lots of calcium is entering
28:35these distal dendrites.
28:36So we know if calcium entry comes the
28:39ability to have longterm potentiation, so.
28:42It seems like that during
28:44these nonrem states of sleep,
28:46when you have lots of sleep
28:48spindles which is in two state,
28:50you can have longterm potentiation
28:53with the calcium entry that's going on.
28:56And it's really specifically
28:58out here at distal dendrites.
29:00At the proximal dendrites there's
29:02practically nothing happening
29:03in terms of calcium activity
29:05and and also in the cell body.
29:09So that's.
29:09So it might be a time when the hippocampus,
29:13for example,
29:14can consolidate memories to the distal
29:17dendrites of the cortical neurons.
29:19And it's the distal dendrites
29:20that house the sort of cortical
29:23cortical information that and and
29:26modification of of our perceptions
29:29and and our actions that it's a place
29:33where I loosely called schema are
29:35formed out in the distal dendrites.
29:38So,
29:38so another thing that happens
29:40specifically at distal dendrites,
29:42both in the cortex and the hippocampus.
29:44This is both of these slides are true
29:46in the hippocampus as well as the
29:48probably true in the hippocampus.
29:50Here we know it isn't true in
29:52the hippocampus that there's
29:53something called these P waves,
29:54which are big glutamaturgic surges
29:56that come from the brainstem all the
29:59way through the thalamus and the
30:01and the cortex and the hippocampus.
30:03And these P waves provide tons of
30:06glutamate also specifically to the
30:09distal dendrites of these pyramidal cells.
30:12So the and the P waves happen
30:14also in the N2 state and then they
30:17happen in spades in rems,
30:19like they're bursting all
30:21the time in REM sleep,
30:23particularly the active phase of REM sleep.
30:24And this big glutamaturgic surge combined
30:27during N2 state with with these big
30:31calcium inputs could really cause.
30:33A beautiful longterm potentiation out
30:35here that I'm going to argue is not
30:39as readily possible during wakefulness.
30:41So here's the idea.
30:43In the hippocampus during our
30:45waking and coding period,
30:47the novelty pathway,
30:49which is the trisynaptic pathway
30:51that comes from layer two of the
30:55antarainal cortex to the dentate
30:57gyrus to CA-3 to CA-1 all impacts the
31:01proximal dendrites here close to the.
31:03To the cell body and can cause
31:06beautiful longterm potentiation there
31:07in the mill year of wakefulness which
31:10includes high norepinephrine which
31:11helps us to learn and helps longterm
31:14potentiation but prevents depotentiation.
31:17And the when we're learning
31:20something brand new,
31:21the familiarity encoding circuit
31:24which was identified by Olga
31:27Vinogradova in Russia and think she
31:30published her last paper in 2001.
31:32She called this from from lots
31:34of her research,
31:36this is the familiarity coding circuit
31:38coming from enteranocortex layer three
31:40directly to the CA-1 distal dendrites.
31:43That doesn't it.
31:44It's not that active because LTP is
31:46much more difficult to get out here.
31:49That's something that Aaron
31:50Schumann showed at LTP is very
31:52difficult to get out here.
31:54But during that transition
31:55to REM sleep when we've got
31:59that those P waves.
32:02And and sleep spindles.
32:04You could actually get beautiful
32:07longterm potentiation out here
32:08And then during REM sleep when
32:10you also have no norepinephrine.
32:12That potentiated circuit out here
32:16which is at a different phase of Theta
32:18inputs are coming at the opposite
32:20phase of Theta than they are here.
32:22Can that now potentiated circuit could
32:25actually cause a dendritic spike to
32:27cause the cell to fire at the Theta through?
32:30If you're measuring the Theta
32:32trough intracellular here,
32:33and that could cause depotentiation here,
32:38because here all of these inputs
32:40are not arriving when the cells are
32:43firing with their dendritic spike
32:46causing the whole cell to fire.
32:48These inputs are not arriving and so that
32:50would cause heterosynaptic depotentiation,
32:53particularly in the absence of neuropaneph.
32:56Sorry, let me get it.
33:00There's my dog. Hello dog.
33:07OK, so all right,
33:11so spontaneously spindles increase after
33:14learning hippocampus dependent learning.
33:16So in humans, during a declarative
33:20task which involves the hippocampus,
33:23spindles increase. In animals,
33:26during a digging task where they
33:28have to dig in a particular place
33:29and associate that dig with an
33:31odor that's in a particular place,
33:33or a nose poke task where they have to
33:35poke their nose in a particular place,
33:36sleep spindles really increase.
33:40And then there's just been study after
33:42study showing the importance of sleep
33:44spindles for memory consolidation.
33:46So my student Michelle Frazier,
33:49who's almost finished with her dissertation,
33:51I'm going to be sad to see her go.
33:52She's absolutely brilliant,
33:54is able to test.
33:55Are kind of working hypothesis that
33:58that input to the distal dendrites
34:01versus the proximal dendrites is is
34:04really important for the sense of
34:07familiarity and she's looking at is the
34:09interneurons that specifically inhibit
34:12activity at the distal dendrites,
34:15they're called OLM interneurons.
34:17So again,
34:19the idea is that during wakefulness
34:21you're able to strengthen quickly
34:24through a long term potentiation the
34:26proximal dendrites of the hippocampus,
34:28which encodes novel information and then and.
34:32But nothing much is happening here.
34:34And then during the late consolidation phase,
34:38after potentiation has happened
34:40here at the distal dendrites,
34:42you can get dendritic spikes
34:45occurring at to force.
34:46The cell to fire at what is
34:49locally fatal troughs.
34:51Sorry to cause deep potentiation there.
34:54Sleep spindles are the thing,
34:56and PGO waves to strengthen these dendrites.
35:00OK,
35:03So what does all this have to do with
35:06REM sleep, dreaming, and emotions?
35:08Let's get back to that. All right, so.
35:13What's happening during REM sleep is
35:14you have these P waves and they come
35:16from what's called the sub cyrillus,
35:17an area just beneath the local cyrillus.
35:20We know that the local cyrillus and
35:22dorsal rafae nucleus are not firing,
35:24so you not providing those two
35:26neurotransmitters and if I've skipped
35:28something you please feel free to
35:31interrupt me if you're like, wait a minute,
35:33what does this have to do with this?
35:35Please just feel free to interrupt me that.
35:38So areas of the brain that are really active
35:42during REM sleep are the limbic areas,
35:45including the anterior cingulate cortex,
35:48the secondary visual areas probably
35:50responsible for the visual content
35:52of our dreams.
35:53But there are whole swaths of our
35:55brain that are actually not very
35:57active at all if you look at pet pet
36:00images like our prefrontal cortex.
36:03Where judgments and decision making
36:06happen probably the reason why in
36:10our dreams we have do things that we
36:13wouldn't necessarily do when we're
36:14awake and things happen that are not
36:17necessarily logic logical and we don't
36:19really question them because you know,
36:21our prefrontal cortex is really
36:23fairly inactive and these are
36:25various studies that show that so.
36:28So what we think this is important
36:30for is again that heterosynaptic
36:33depotentiation idea.
36:35So when norepinephrine is not present,
36:37you can get depotentiation.
36:38When some areas of the brain are
36:40super active and other areas of
36:42the brain are super inactive,
36:44you can actually get a weakening
36:47of synapses between those areas.
36:49So our frontal cortex is really
36:52relatively inactive during REM sleep,
36:55and this is a PET scan.
36:58And also inactive relative to slow a sleep.
37:02But REM sleep is has a time when our
37:06limbics areas are very very active
37:08and probably responsible for the
37:10emotional content of our dreams.
37:12And it is our idea that without
37:15norepinephrine there to cause potentiation
37:17and to block deep potentiation,
37:20we could actually get a separation
37:24between this activated emotional circuit.
37:27And the prefrontal cortex.
37:29So during wakefulness it's all,
37:31you know,
37:32being knit together and the
37:34emotionality and the facts are all
37:36coming in together into our brain and
37:38causing lovely longterm potentiation.
37:41Because the locus Cerulis is providing
37:43neuropidephrine all over the place,
37:44letting everything be knit together.
37:46But normally during that transition to REM,
37:49the the information can be
37:52transferred toward distal dendrites.
37:55And then during REM sleep we can
37:58actually erase them from our that
38:00information from our proximal dendrites
38:03and reduce then the immediacy,
38:05the novelty of all of those emotional,
38:11emotional memories as at once
38:13memories have been consolidated okay.
38:15So they also the the cyclicity of sleep
38:18is probably really important here.
38:21So during nonrems,
38:23we need one stage to happen after
38:27the next or or you know,
38:29we might end up depotentiating before
38:32we've potentiated and consolidated.
38:36So really the timing of sleep is,
38:37is important and that's probably
38:40why disturbed sleep is so bad
38:42because when we go back to sleep,
38:44we don't necessarily go
38:45back into the same state.
38:46We could start back up, but you know,
38:48wakefulness and then end to and then
38:50try and get into deep slow sleep.
38:51We might miss our slowly
38:53sleep stayed altogether.
38:54We might go, you know,
38:57just our REM sleep might be disturbed.
38:59And actually with insomnia it's been
39:03shown that that the locus surrealis
39:05is overly active awakening us and
39:08probably also preventing some of that
39:11depotentiation function from happening.
39:13So here is our idea.
39:16This is true of some mice that that
39:19the neurodinergic locus realis is off
39:21during all the states of sleep except
39:23for and then on during wakefulness.
39:25But at least we know the rats and cats
39:28and and other monkeys and probably humans,
39:32that this is the pattern of locus realis
39:34activity across the sleep waking states.
39:35And so there's been a lot of
39:37sort of anecdotal, not anecdotal,
39:39but secondary evidence in people
39:41with post traumatic stress.
39:42Disorder that the locus cerulis
39:44actually doesn't shut off during
39:45REM sleep like it should.
39:47Sorry, these things are a little
39:48shifted and it might be doing.
39:50Depression is also,
39:51you know,
39:52a difference in terms of the way
39:55things happen during during sleep.
39:56So if the locus cerulis isn't shutting off,
39:59what would that do?
40:01One of the things it could do is
40:04instead of erasing the OR weakening the
40:08proximal synapses associated with novelty.
40:11It just continues to reinforce
40:13and and strengthen those proximal
40:16synapses associated with novelty,
40:18thereby disabling people from putting
40:21the past in the past and making
40:24these emotional memories always feel
40:26like they're happening right now.
40:28Or just happened, you know,
40:30that same day.
40:31So there's a really great series
40:33of papers by Rick Wassing,
40:35who's got now got an independent
40:37position in Australia.
40:38I'm working with OS van Summeren
40:40in in the Netherlands and what
40:42they're showing in humans,
40:43what they've showed in humans is that
40:46that people with insomnia disorder
40:49have very disturbed sleep and what
40:52seems to be disturbed most are those
40:54sleep spindles of the end to stage
40:57of sleep and REM sleep itself.
41:00So they have reduced sleep spindles.
41:03They have many more arousals from REM
41:05sleep and that transition to REM sleep.
41:07They have heightened sympathetic Dr.
41:09heightened fight or flights sympathetic Dr.
41:13and the loathe.
41:14The Syrillis never seems to really
41:17rest and be silent during and to and
41:19REM sleep and it is associated with
41:21depression and other anxiety related
41:23disorders and what they showed with.
41:25Brain scans of people is that novel
41:28experience in normal sleepers
41:31are encoded
41:37initially, but then they are reduced
41:41and after after sleep you can see
41:43the activity in these brain areas
41:45that are involved in encoding.
41:47These emotional memories are
41:50less activated when they recall.
41:53But in insomnia disorder,
41:55the recall of this emotional experiences
41:58is if anything much stronger in
42:01all of these emotional areas.
42:03So here's the relived experiences
42:05in normal sleepers.
42:06You can see the you know these areas are
42:08not they are able to recall them fine,
42:11but the emotionality of it the.
42:14You know, galvanic skin responses,
42:16the heart rate, all of that,
42:17doesn't get invoked again when they're
42:19recalling an old emotional memory.
42:21But people with insomnia do have a
42:24reactivation of these emotional areas.
42:26Plus, you know,
42:27all of the external signs and
42:29that the emotionality of the
42:32memory is still being involved.
42:34So here's just dorsal.
42:38Anterior singular cortex in
42:40insomnia disorder is is activated
42:42almost as though it had happened
42:44that same day instead of in the
42:47past through a sleep period.
42:49So these people with insomnia
42:52are kind of haunted by the past,
42:55overdriven by the present,
42:57so and and probably have
43:00this dysfunctional sleep.
43:02We do know that they have dysfunctional
43:05sleep activity that could lead to
43:08a very dire and drastic actions.
43:12So here's the case of Post Traumatic
43:15Stress Disorder that that Al K
43:18had mentioned in his introduction.
43:21Here is someone in the theater of war
43:24learning that a helicopter could bring
43:26bombs and bullets and you should avoid them,
43:29but when you come home.
43:31The helicopter is probably associated
43:33with safety and people without PTSD,
43:36which is the majority of people, thankfully,
43:38who encounter A traumatic experience.
43:41They don't have PTSD.
43:42They can reassociate the sound and sight
43:45of a helicopter with safety of home,
43:48that context of home,
43:49but with PTSD it's more difficult.
43:52They they their war experience is,
43:56or whatever the traumatic experience
43:58is a lot more immediate to them.
44:01And so,
44:01yes,
44:02they could learn that the helicopters
44:03can also be news helicopters.
44:05But whether what's immediately triggered
44:08when they see a helicopter or hear
44:11one approaching is more strongly
44:13associated with that old fearful
44:16memory and not with the new memory.
44:19This is a picture of my uncle,
44:21my mother's brother, of my uncle,
44:23daddy, my favorite uncle.
44:25He's a sweet guy who played the flute
44:28and taught us how to ride bicycles.
44:31But he was drafted to go to
44:34Vietnam in the 70s and when he
44:36came home a couple years later,
44:39his wife had joined the the movement
44:42antiwar movement and he wasn't as
44:44welcomed at home and she left him and
44:46took their daughter with them and.
44:50And he,
44:51you know,
44:52drove off a freeway at 70 miles an
44:55hour after closing his bank account.
44:56So the effects of this are are very
45:00drastic and direly left his whole
45:02family behind and is missed to this day.
45:06So what we're thinking is that we have
45:07too much norepinephrine in sleep.
45:09We can't ever depotentiate so.
45:12So we also know with too much norepinephrine
45:15you have lower REM sleep Theta activity.
45:19Insomnia and disturbed sleep happen
45:21with too much norepinephrine,
45:23too much Lococils activity.
45:25And people with PTSD have nightmares,
45:27and we know they have a
45:29heightened sympathetic drive.
45:30So the idea is that they're
45:32kind of stuck in the past.
45:34That novelty and coding circuit can never,
45:36can never be depotentiated.
45:39So your hippocampus becomes
45:41saturated with that traumatic
45:44memory you can't contextualize.
45:47The fear or the shame or the guilt.
45:50You can't detach from the
45:52emotionality of these memories,
45:53and the main memories stay salient and novel.
45:57So this is just the idea.
45:58Novel information,
45:59normally with good healthy sleep,
46:01you know,
46:02with lovely sleep spindles can
46:04be incorporated into our schema
46:06or distal dendritic schema.
46:08Through those sleep spindles
46:09and then during REM sleep with
46:11the absence of norepinephrine.
46:13And I didn't really mention
46:14much about serotonin,
46:15but as part of the circuit you can get a
46:19rearasure of synapses that no longer service,
46:23like the novelty encoding circuit.
46:25But if you have maladaptive sleep,
46:27too much norepinephrine or
46:29not good sleep spindles,
46:30you can't really incorporate that new
46:33information like the context of home.
46:35And then,
46:35if REM sleep is too much norepinephrine,
46:37you can't ever depotentiate.
46:38In fact, you just keep repotentiating
46:42those familiar or those novel circuits,
46:46and you can't ever get away
46:48from from that those memories.
46:51All right, so recipe,
46:52what would be the recipe
46:54for changing your mind?
46:55Well, you got to have good slowly sleep
46:58where you can wash and replenish the energy,
47:01the milieu of your brain.
47:04You need and two state with lovely,
47:06rich sleep spindles where you can reduce
47:08the amount of norepinephrine and serotonin,
47:11allowing those sleep spindles to appear.
47:13You can reactivate those
47:15memories with hippocampus,
47:16sharp ways,
47:17ripples,
47:17and that couples with long sleep
47:21spindles and helps consolidate
47:23those memories into your brain.
47:26And then during REM sleep you really need.
47:29High acetylcholine for good plasticity,
47:31high glutamate from those PGO waves,
47:34no norepinephrine to allow depotentiation,
47:37no serotonin to allow the familiar
47:40reconsolidation and novel depotentiation.
47:43Again, we didn't talk about serotonin.
47:45I'll just briefly say what serotonin does,
47:47and one of the things it does is it it
47:50shunts activity from those distal dendrites.
47:53So it doesn't reach the Axon hill.
47:55It can cause the cell to fire.
47:57So when it's present,
47:58which is during wakefulness,
47:59the familiar is The sensory inputs
48:04are more guided by what's novel.
48:07What's new about this?
48:08What can I learn from it
48:09rather than what's familiar?
48:10But during REM sleep,
48:12when you don't have norepinephrine
48:14with serotonin,
48:15the familiar can take over and cause
48:18the depotentiation of the novel.
48:20All right.
48:20So really need this whole cycle
48:22and the structure of sleep
48:24to really change your mind.
48:26And we've talked about how this
48:28happens on a micro circuit basis.
48:30All right.
48:31So what about in the last two minutes
48:35of this talk here? What about? PTSD.
48:39So we've started testing this in rats.
48:43We give them the worst day of their
48:45lives where they're bound for two hours,
48:47they're swimming together,
48:48and without a way out,
48:50they are put into a jar with ether,
48:52which is a direct activator of the HBA axis.
48:56And then they're isolated for a week.
48:57And that is also interestingly important for
49:00setting up the PTSD phenotype in animals.
49:02If you give them the ability
49:04to comfort one another,
49:05or if you even interact with them a lot,
49:07they're much less likely to get PTSD.
49:10And they get PTSD phenotypes,
49:12which include the inability to extinguish
49:16fear as about as often as a as a
49:20human does in a normal circumstance
49:23where they're socially connected.
49:25But if you socially I isolate them,
49:27they're much more likely to get it.
49:29And what happens with an animal with PTSD
49:32is during REM sleep or PTSD phenotype.
49:35During REM sleep,
49:35the look of syrillis is really
49:38continuing to fire,
49:39unlike the silence it happens
49:41during during normal REM sleep.
49:43So and sleep spindles are also changed.
49:46So in animals that are resilient,
49:49this is the day versus night,
49:52this is sleep phase versus waking phase.
49:55Amount of sleep spindles and you can see
49:57that that really doesn't change very much.
50:00Actually this is single prolonged
50:02stress that I just showed you.
50:04Initially they go up and
50:06then they normalize again.
50:08But in animals that are susceptible to PTSD,
50:11the sleep spindles don't rise
50:14after the single prolonged stress.
50:17And over the course of that week,
50:19during the consolidation of
50:21that traumatic memory,
50:22the number of sleeve spindles goes down,
50:24goes down instead of staying normal.
50:27What happens with the estrus cycle?
50:29Well,
50:29animals during the high estrus phase,
50:32which in humans here is about the week before
50:36our periods during the high estrus phase,
50:39this, there's a,
50:42let's see if I'm trying to say whatever,
50:43I'm starting here.
50:46We actually have as little activity
50:50in the locus cyrillus during REM
50:52sleep as this is rats as as males do.
50:57But at high at low hormonal
50:59phases the locus cyrillus remains
51:01active and what like this?
51:03Do this might actually
51:07give make females during low hormonal phases,
51:12no? Estrogen phases more susceptible to PTSD
51:16because melocosuris doesn't stop firing even
51:19on normal even without trauma exposures.
51:24All right, So we also interestingly
51:26in these high estrogen phases,
51:28we sleep a lot less,
51:30a lot less REM sleep and slow wave sleep.
51:32But when we do sleep, we have more rich
51:37sleep spindles and this high hormonal phase.
51:40So that this is the number of
51:43spindles per minute in females.
51:44In the Proestres phase,
51:45you'll see they're just much,
51:47much, much higher. And
51:51so it might be that that even
51:53though we're sleeping less,
51:54we're sleeping more efficiently.
51:56And so can is it estrogen?
51:59Well, there's been a study showing
52:01that if you give women the morning
52:03after pill in an emergency room.
52:06That contains estrogen.
52:08They're much less likely to get PTSD
52:11than than women given a morning
52:12after pill without estrogen.
52:13So there's probably something
52:15to do with with estrogen and
52:17the locus surrealis in the way
52:19it fires that has not yet been.
52:21Thoroughly investigated and it
52:22would be interesting to see.
52:23So I think I'm going to stop here
52:26because there's lots more to talk
52:28about and and I don't really
52:30necessarily have the time for it right now.
52:32So I just want to say sleep is important.
52:35It's important for memory and for
52:37erasure or weakening of at least
52:40certain aspects of memory like
52:41the novelty of it.
52:43Our work brain is working really
52:44hard and this is the threat that's
52:47been funding this research since.
52:48The year 2000,
52:49and this is a current picture
52:51of my laboratory,
52:53and I want to thank all of my
52:54students for all of the work that
52:56they've been doing to gather all
52:57these data that I've shown you,
52:58and then I could leave you with a while.
53:01We do a Q&amp;A.
53:02I'll leave you with the video of
53:05these elephant seals during wakefulness,
53:08diving down past the place where
53:14you know the sharks can get them,
53:16and then starting to glide. And sleep.
53:19And you can see this is the sleep
53:22frequency going slow to slow wave
53:25sleep as they continue to dive.
53:27I don't know why they're continuing to dive.
53:29I you know I guess they must not have
53:32enough body fat to keep them floating.
53:35But yeah here they are
53:38and fairly soon this one is
53:40going to go into it gets upside
53:42down now it's definitely asleep.
53:44Wouldn't it be freaky to be in
53:47a submarine and. Looking out.
53:48I don't know if some brains have windows.
53:50I guess they do.
53:52Looking out and seeing this seal
53:54diving and diving in a spiral
53:57fashion as it goes into REM sleep.
54:00All right.
54:00So thank you very much for your attention.
54:02Do you have any questions?