Tip:
Highlight text to annotate it
X
[SOUND] Okay. Let's get going.
Okay. First off, important announcement.
Hopefully you all saw the note the other day after Monday's lecture
about, the actual neurotransmitterology of the
different synapses in the autonomic nervous system.
I don't quite know who they allowed to come in the other
day to lecture, but pay no attention to, whatever it is he said.
So look at the announcement there about the mysteries, the correct mysteries of
acetylcholine and norepinephrine. Okay, so where have we gotten to here.
Finally, time to look at what has been hinted at throughout, which is
every single factoid that I have piled onto you in these five lectures.
Every single thing about the nervous system
that you've seen so far, has two characteristics.
One is enormous individual differences.
Everything you've heard about differs
in you,
in some manner, from the person sitting next to you,
and equally importantly, focus of today, it could change over time.
It shows plasticity, the ability to change, predominantly
in response to environment, in response to experience.
What we'll look at today is, the workings of
that neuroplasticity at a whole bunch of different levels.
We've already seen some examples
of it, back to the very first one.
Depending on what your estrogen levels are, and if you're a female
hamster, what's going on with your axon hillock plasticity in that sense.
We'll look today at a variety of types of
plasticity theme over and over and over, despite all
sorts of dogma for zillions of years in this
business, very little about the brain is set in stone.
This is a good thing. This is a good thing
because this plasticity, for example, is what learning is
about and that better be going on about like crazy.
So, we have, what are the mechanisms for it?
Looking at, how does learning occur?
Okay, so if you were a neuroscientist around 1890 or so, and you
were asked, what happens when you learn a new fact, everybody knew the answer.
Which was, when you learn a new fact, you get a new neuron.
A neuron codes for a new fact.
A neuron knows one thing and one thing only.
Information is contained in single neurons.
And soon people learned that was absolutely ludicrous, and those
scientists were laughed to their graves, because people realized no,
the adult brain doesn't make new neurons and this certainly
doesn't make a neuron, to correspond to a new fact.
This gave rise to the next view, which was
where is the information contained when you learn a new
fact, you form a new synapse, and that was the dominant model for decades.
Right around the time people had fancy enough of
microscopes to figure out there were these synapse things.
You formed a new connection between two neurons, and that was what memory was.
The formation of new synapses.
And soon, people got sophisticated enough to
realize that those were people were ludicrously,
grotesquely wrong, and they were laughed to their deaths also,
and by the 1950s, in came what remains the dominant model.
And this was due to the work of
this one guy, this Canadian neurobiologist, named Donald Hebb.
If you were ever going to become a neurobiologist, you are
forced to know everything about the blessed childhood of Donald Hebb.
This guy was so, you cannot get tenure unless you name your children after him.
Both of my children are named Hebb.
What you wound up seeing was this unbelievably
influential view that he brought in, totally radical, at
the time, which turns out to be true, and
is the basis for a gazillion careers since then.
Memory is not coded for in new neurons or
new dendrites or new axon terminals or new synapses.
Memory is coded in the form of strengthening
pre-existing synapses.
Strengthening the connections between two neurons that are already there.
And we can immediately translate this into the first lecture in single
neurobiology, single cell stuff. What memory is about, what plasticity is,
that for the presynaptic neuron, yelling x amount, the postsynaptic neuron is
now more likely to have an action potential.
The presynaptic neuron, has more power over events in the postsynaptic neuron.
And we can immediately also see, you
can do that plasticity at an inhibitory synapse.
A presynaptic neuron yelling just as loud of an inhibitory message as before, now
has an even more hyperpolarizing effect. The presynaptic
element has more power over the postsynaptic component.
You have strengthened the preexisting synapse.
And this has been the dominant model for everafter, this is what memory is about.
You know already from all that neural network stuff the
other week, it's not going to be one single synapse.
Memory plasticity, formation of new memories is causing
strengthening of whole networks of synapses, but
that's the basic building block of it.
A synapse that's already there becomes
more excitable, if its an excitatory synapse,
or becomes more inhibitory. The presynaptic
neuron has more control over the latter.
Okay, so what does this actually look like?
So here you have some neuron, and what you're doing
is you're sticking in an electrode in the first neuron.
And you're able to stimulate it, and you're recording from
the dendritic spine of the postsynaptic neuron, or you're just recording
from the postsynaptic neuron, and you stimulate it and you see
on the left our usual deal from the very first lecture.
You get a depolarization out of the dendritic end.
Totally routine. We know the drill on that.
Now what you do is stimulate it over and over
and over and over, jargon,
tetanic, stimulation, tetanus, repeated stimulation
at a rate of 10, 20, 30 times a second for 10 seconds, very high stimulation rate.
Now, you come back and you stimulate it once
more and what you see is lo and behold
a miracle, the postsynaptic neuron has more of a depolarization.
It has been more excited by the presynaptic neuron.
You have potentiated its response. Okay.
This is amazing enough.
What's even more amazing, is you go away now for a long weekend
or three weeks or whatever and you come back to the same exact synapse.
And you stimulate the presynaptic
neuron, and you get that same potentiated response on the other side of the synapse.
Thus the official jargon of the field, what
you are observing is, LTP, Long Term Potentiation.
And Hebb, after predicting this is exactly how it would work,
merely had to wait a decade or so before
electrophysiologists began to show in the 60s, that's exactly
how it works.
This phenomenon of LTP, and 100 billion neuroscientists since
then have spent their lives studying LTP, long term potentiation.
You stimulate the presynaptic neuron, and
you get your usual boring postsynaptic depolarization.
You stimulate it like crazy and instead
you potentiate it, you potentiate it long term.
You potentiated by the standpoint of the single
cell, for glacial periods, you've induced long term potentiation.
People loved this they went completely crazy about it.
Because, depending on what your taste was, what you
either immediately said was, this is the greatest model on
earth, as to what learning must be like in the
brain, or if you were more expansive you would say
this is learning. This is exactly what goes on in the brain.
And it was not by chance, the hippocampus, the part
of the brain that does the most dramatic learning and memory.
The hippocampus is where LTP was demonstrated.
And the fight immediately was, is this a model for learning?
Is this what actually occurs?
Does learning ever involve tetanic stimulation at that sort of pace?
And over the years, what's become pretty clear is,
yeah, this is kind of what learning really is about.
It's not just the model system, this is what occurs.
So long term potentiation.
So of course, what immediately comes up are, the two critical questions.
What causes the potentiation of that synapse?
And what causes it to be long term?
What causes an explosive amount of depolarization
in that post-synaptic neuron, and what are the mechanisms for it, so you can
go away and come back a month later and that synapse is still doing it.
And that's where staggering amount of work went
in to figuring out the mechanisms for LTP.
Okay.
What it all revolves around is a neurotransmitter,
that's been woefully underappreciated in the class so far.
A neurotransmitter called glutamate.
Wait, you were no doubt saying. Glutamate, that's an amino acid.
What do we have here?
We have the cheapest version of all of those
themes of neurotransmitter synthesis, hormone
synthesis, from the previous lectures.
Forget get a plentiful precursor and just have a couple of biosynthetic steps.
Get a plentiful precursor and you're set.
You've already got your neurotransmitter, glutamate.
And then amino acid,
along with the structurally very close to it,
aspartate, both of them also function as neurotransmitters.
And this turned out to be a major discovery because it
turns out glutamate is the most excitatory neurotransmitter in the brain.
The majority of your brain synapses use glutamate, it is incredibly important.
And it turns out, this is the perfect neurotransmitter
for doing something as fancy as
plasticity, as learning as something like LTP.
Because glutamate works in a much more
complicated way than everything we've learned so far.
For simple reason factors more than one type of receptor for glutamate.
Okay, I've alluded to this, and in actuality
every type of nerve transmitter, serotonin has like 30
different types of receptors.
There's multiple, all the neurotransmitters have multiple receptors.
What am I talking about here?
In the same synapse, the same dendritic spine
will have more than one type of glutamate receptor.
It's got two different kinds of lying.
It's got a whole bunch of, for
our purposes there are only two different kinds.
And it's the fact that the two can show up in the same dendritic
spine, that allows you to do this amazing LTP stuff.
Because there's something very subtle power that you get information
lay, when you got two types of receptors there, okay.
Here we've got, shockingly enough, one is called the NMDA
receptor, and the other one is called the non-NMDA receptor.
NMDA, N-methyl-D-aspartate, do not memorize that.
The NMDA receptor and
the non-NMDA.
Glutamate works by way of working on both the NMDAs and the non-NMDAs.
The non-NMDAs are actually a class of half
a dozen, ignore that, these are the two types.
The non-MMDAs, on the right, are totally familiar and boring.
They're the usual sort of assembly
line, neurotransmitter receptor for an excitatory neurotransmitter.
You know exactly what it does.
Glutamate comes along and binds to the non-MMDA, and that opens
up a channel and in comes sodium, the usual, blah, blah.
On the left though is the NMDA receptor,
and there's two things about it that are different.
First off, it doesn't allow sodium to
flow, once channel opens it's calcium that flows.
Okay. Calcium already made a cameo
appearance many lectures ago.
In the axon terminal, a voltage gated calcium channel is what
opens up the very, and the last neurotransmitter to be released.
But here suddenly, we've got a different world from, at the
dendritic spine, everything is mediated by either sodium or chloride or potassium.
You know, suddenly now we've got as
the first anomaly it's calcium that's flows in.
Calcium flows
in through the channel, connected to the NMDA receptor.
The second weird thing is, calcium really doesn't flow in, because
normally, the channel is clogged up, its clogged up with magnesium.
Normally, the channel is clogged up with magnesium such that, if glutamate
comes along and binds to the NMDA receptor, nothing happens whatsoever.
So what occurs now? Here's the rule.
On the right, we have our usual deal.
Glutamate shows up, binds to the non-NMDA. Sodium flows in and the
rule is if and only if, there's enough sodium
underneath the NMDA receptor to depolarize the neighborhood there.
If you depolarize the area
just underneath the NMDA receptor, you kick out the magnesium.
The magnesium gets tossed out of the channel.
And at that point, if there's glutamate binding
to the NMDA receptor, Calcium will flow in.
This should sound like a totally bizarre combination
of a voltage gated channel, and a receptor gated channel.
And that's
exactly what it is, it's a hybrid. You've got a double if and only if,
if and only if, there's enough depolarization
just south of the receptor to kick
out the magnesium, and there's glutamate around
binding to the NMDA receptor, Calcium flows in.
So we've got this double contingency here.
We have our usual receptor gated channel on the right.
And on the left, this weird hybrid voltage, and receptor gated channel, where
you've gotta have enough sodium coming in, to open up the, the NMDA receptor.
So how does this work?
What does this look like when you put the pieces together?
You throw some glutamate onto a synapse.
Some sodium comes in. You get a little bit of depolarization.
Nothing happens on the left. You throw
in some more glutamate. More sodium comes in.
Nothing happens on the left.
You finally throw in enough sodium, you throw in enough glutamate,
that you get enough sodium coming in that finally you pass threshold.
You pass threshold, to kick out the magnesium and
suddenly a tidal wave of calcium comes pouring in.
What have we just defined here?
A non linear synapse.
Everything we've heard about already is on the right.
A simple sort of linear relationship between the amount of neurotransmitter
you throw on, and the amount of bionic changes that occurred there.
On the left, you suddenly have a non linear
threshold, you've got a system where only if you finally
get enough of yelling coming through on the right,
by way of the non NMDA's you passed threshold and
suddenly, the NMDA comes into play and
there's an explosion of calcium rushing in.
What is this?
This is somebody, some lecturer for example, says something
and it goes in one ear and out the other.
And they say it again, and it goes in one ear and out the other.
And they say it again and again. And finally, aha, you get it.
Finally the light bulb goes on. When the NMDA finally opens up with enough
buzzing there, with enough of the non-NMDAs on the right.
That's the aha, that's the light bulb going on
in the synapse, that's where you suddenly have passed
the threshold, that's when the synapse gets it for
the first time, when you finally pass that threshold.
Now, what this winds up doing is setting things up.
For something wonderfully subtle and logical.
Okay, so on the far left, we have a circumstance.
We've got a synapse where the dendritic spine contains only a non-NMDA receptor.
And what we're measuring on top is, you're putting in increasing amounts
of glutamate and how much calcium is flowing into that dendritic spine.
And you know already what's going on here.
Which is, there's no calcium
going to flow in, because the non-NMDA receptor only allows sodium to flow.
You could put on enough glutamate that that dendritic
spine is being flattened by it, and nothing's happening.
Now in the middle, we've got one non-NMDA receptor, and one NMDA receptor.
And what you see is exactly what I just described.
You put on some glutamate, nothing happens to the NMDA and you put some on more.
And finally you pass threshold.
Boom, explosion. Kick out the magnesium.
Calcium channel opens up, calcium pours in.
And you've passed that inflection point. Aha, the light bulb has gone on.
Critically, look on the far right.
And what we have now, are two non-NMDA receptors for every NMDA receptor.
A 2 to 1 ratio. And what you see is, the same exact thing
as in the middle, except it occurs sooner.
You've lowered the threshold for getting the aha moment.
And that was a major finding.
If you want to make a synapse more excitable
in this sense, if you want to make a synapse
more likely to have that aha light bulb moment,
increase the ratio of the non-NMDAs to the NMDAs.
Because for the same amount
of glutamate pouring across, assume it's a saturating amount
of glutamate that would fill up a gazillion receptors there.
For the same amount of glutamate coming across in this
case, twice as much sodium is going to be coming in.
You're going to hit that aha threshold point twice as quickly.
That much less glutamate is needed to finally depolarize the NMDA kick
out the magnesium, calcium rushes in, and the light bulb goes on.
In other words, you could change the tuning, the threshold for a synapse,
a glutamatergic synapse, having this explosion of calcium pouring in.
You could change the threshold for it, the set
point, by changing the ratio of non-NMDAs to NMDAs.
This winds up being critical.
Okay, so what have we just accomplished here?
We have now potentiated
the synapse, we've got this aha moment, where
suddenly a huge amount of calcium pours in.
And the channel opens up like crazy, massive
amounts of calcium comes in, both electrical forces
and chemical forces, big, big influx of calcium
and this is a highly depolarizing excitatory event.
So great, this is us titanically stimulating that
synapse enough times over and over and over to
get enough sodium in there to kick out the magnesium.
And we've potentiated.
So how do you do it now, that we walk away from this
synapse for three weeks, come back, and it's still going to be potentiated?
How does the long termness come in? And this is where massive amounts of work
has been done, working out all the molecular steps involved
in turning the potentiation into long term potentiation.
And there's also, these are like, four of the most implicated mechanisms,
and people have like, stabbed each other over which is more important.
Broadly, this is how you can get the synapse to now have the aha.
And it's gone from learning something
to remembering it, maintaining that aha forever
after, making that synapse more excitable.
So a whole bunch of mechanisms, very first one.
Okay, very first one.
Everything we'll see here, these four major mechanisms, four of many,
many additional ones, All of them are built around calcium coming in.
Because calcium pouring in to this dendritic spine
is not just doing stuff to the charge.
Oh it's depolarizing this positively charged, hooray it's
kind of like sodium, but it's got double the punch.
What we'll see here is calcium is also a
trigger for all sorts of longer term changes to occur.
So step number 1.
Calcium comes pouring in, and as a result, it
activates a bunch of, reasonably enough, calcium dependent enzymes.
One of them is called nitric oxide synthase, and shockingly,
what it does, is it synthases, it synthesizes nitric oxide, this gas.
Next, there's another enzyme there, called phospholipase.
And what does it do?
It cleaves out a piece of membrane, out of a
membrane, and makes a lipid based messenger called arachidonic acid.
Why are these interesting?
At this point, nitric oxide, and
arachidonic acid, that've been made in the dendritic spine, get released
from the dendritic spine, go back to the axon terminal,
the presynaptic element, and give it a signal to start releasing more glutamate.
Yes, you should all be up in arms at
how totally bizarre that is, what I just told you.
First piece of bizarrity,
this is a messenger going from the postsynaptic element to
the presynaptic, nitric oxide and
arachidonic acid work as neurotransmitters.
What are you talking about?
Neurotransmitters go from left to right
They are what are termed retrograde neurotransmitters,
that go from the postsynaptic dendritic
spine over to the presynaptic axon terminal.
So, now, heresy!
We've got neurotransmitters going in the wrong direction.
Next bizzarity about them. Where is the vesicles?
Where's the vesicles with the little act in
piece of dental floss, and where's the exocytosis?
None of that occurs, none of that occurs because you don't need it.
Nitric oxide is a gas, it just permeates through the membrane.
Arachidonic acid gets through membrane because two seconds
ago, it was part of the membrane.
You don't have normal mechanisms, for vesicles, and exocytosis.
They just kind of drool out of the
dendritic spine, and travel retrogradely to the presynaptic neuron.
So you've got these totally bizarre, unconventional retrograde
gaseous lipid based neurotransmitters going in the wrong direction.
And insult
added to injury you even have some receptors
for the arachidonic acid stuff at the axon terminal.
And all of this going completely *** backwards.
And what does it wind up doing.
Both of them sent a message to the presynaptic neuron.
From now own put an extra glutamate into the vesicles.
So, we've gotten our first long term-ness thrown into this potentiation.
Now what have we accomplished the next time the presynaptic neuron is stimulated.
It's going to yell louder.
It's going to be releasing more glutamate.
So a first step in causing long-term potentiation to this synapse, a
postsynaptic retrograde mechanism, to get the
presynaptic neuron to yell louder forever after.
So that's the first step. Second step.
Second step,
is built around exactly the point I just made,
which is so you want to make this post-synaptic
element, this dendritic spine, you want to make it
more excitable for the rest of time, long term potentiation.
Let's get some more glutamate receptors.
Which kind do you want?
This was from five minutes ago.
Do you want to stick more NMDA receptors on the surface or more non-NMDAs.
You want to stick more non-NMDAs.
So, what does calcium do? Second mechanism, it goes to intracellular
mothballed stored sites for non-NMDA receptors and little vesicles
sitting there just under the dendritic spine, and a whole signalling cascade.
And it pulls more of the non-NMDAs onto the cell surface.
So what have we accomplished here? Perfectly logical.
The presynaptic neuron,
thanks to step number 1, is yelling louder,
and the postsynaptic neuron has more ears now.
It's got more non-NMDA receptors And that's going to
bias towards a lower threshold for the aha moment.
So that's great.
Step number 3.
Let's go to step number 4 first, step number
4, previously known as number 4, now number 3.
The next logical thing,
which is doing one of those, let's find some switch to
turn on and off that will cause some long-term functional change.
You phosphorylate the non-NMDA receptor, and as
a result, the sodium channel stays open longer.
And how do you that? Calcium just through a totally textbook
calcium-dependent kinase pathway does that, so now you've got more non-NMDAs
And once they're stimulated by the larger amounts of glutamate
now coming out of the axon terminal just to your left.
They're staying open longer, potentiating the signal
more, making it ever more likely that you're
going to pass threshold, kick out the magnesium and
the light bulb will be forever kept on.
So that's great.
Fourth step is ironically the very first one
that was ever discovered and it was discovered
because all you needed to find it was a fancy microscope.
So, you got your dendrite spine, which by definition, is sort of spine shape,
and all there is this membrane biophysics stuff that goes on where when you have
got the wave of depolarization, our pebble and the pond, when it's coming around the
curve of the dendritic spine, that tends to impede the flow of the charge.
What happens is, at this point, what calcium also does, is
it activates a bunch of calcium-dependent proteases, and what do they do?
They cleave apart the cytoskeleton, in the
dendritic spine, and rebuild the shape of it.
The spine flattens, and as a result of
it flattening, the wave of depolarization spreads further.
You do this major
little renovation project there, in this dendritic spine.
You cleave the cytoskeleton there, and then you rebuild it with a new shape, and
you've got a dendritic spine now, whose depolarization spreads further.
And this was the first one discovered because it was just simply visual,
ooh, notice the spines flatten out at that point, this whole restructuring process.
So this is great, look what
you've accomplished here. With these four steps.
So, on top we've got the pre LTP
universe of stimulate it and get your usual depolarization.
And on the bottom, thanks to these four
steps, bigger arrow coming out of the axon terminal.
You're releasing more glutamate.
And there's more receptors and the key receptors that
you really need for the plasticity on the postsynaptic tide.
More non MDAs, which thanks to their being phosphorylated,
they stay open longer and thanks to flattening the dendrite, the
charge spreads further and what have you got, you've potentiated the synapse.
And this is totally great.
Okay, so, immediately, there are some complications, though.
First one is, there's long term and then there's long term.
Long term potentiation. Hooray, look at this!
You've got more non-NMDA receptors
there, so they're going to stick around longer and you've got
more of them with a half-life of, half-life of a
receptor of a protein like that is, I don't know,
a couple hours, a couple of days, a couple of weeks.
If you've learned something, you may remember it for the next 80 years or so.
There's long-term potentiation and then there's what memory is really like, which
is a hundred gazillion times longer than the lifetime of any of the molecules
we just talked about.
How do you turn long term
potentiation, which would keep a cell biologist
happy, into long term memory, which we're
talking about quarter centuries. In a throw,
how do you get it really, really long term.
And this was a big puzzle for the field.
Because there had to be a way to
make these markers last longer than the lifetime
of the individual molecules involved. Because you're constantly recycling
your molecules as they get old and smelly, and get oxidative damage.
And you've got a half life on every single molecule there.
So hurray for you, you have stuck more of these non-NMDAs onto the surface.
And you phosphorylated them, and that
phosphorylation mark may not last forever.
And at
the very least that molecule isn't going to last forever, cause it's going to
get recycled at some point and replaced with a new copy of the non-NMDA receptor.
Which knows nothing about what occurred two hours ago.
And doesn't know it's supposed to be phosphorylated.
How can you make this potentiation last longer
than the lifetime of the individual molecules involved?
And, during this period,
scientist Mary Kennedy at Caltech came up with a
speculative model for how this would have to work.
And that turns out to be absolutely how it works.
So, you have got this kinase for example.
The kinase, which is calcium activated, and what it
does is, is phosphorylates the non-NMDA receptor, that's terrific.
What she speculated was an additional piece of that, that had to be
happening, which is it also had to autophosphorylate.
It had to phosphorylate itself.
It had to phosphorylate both itself, and other copies of itself.
And this is the critical step. So, you've got this kinase, which thanks
to calcium, has been phosphorylated. And what it now does is,
phosphorylates any non-NMDA receptors it bumps into.
So it's doing its task.
And it's about to run out of its expiration date and be
recycled and trashed, and here comes a new copy of the kinase.
And because it is an autophosphorylator, what it does is as the new copy shows up.
It autophosphorylates it.
It phosphorylates it, and thus it continues to phosphorylate
the new non NMDAs as they come online.
And this autophosphorylation step is the means by which you can
turn memory three weeks old into memory three quarters of a century old.
This was a critical demonstration, jumping from long term from a
cell biologist's standpoint, to long term from a human lifetime standpoint.
Critical, critical finding. Okay, so this was totally beautiful, and
these were four mechanisms there's many, many more on top of that.
For example, one is looking at GABA.
Gaba inhibitory projections onto a synapse, we remember that if, then clause.
Which was the gaba is making the axo axonic
projection there, and we've got the if, then clause
gaba inhibits the postsynaptic neuron if and only if
the presynaptic one is trying to talk to it.
You remember that bit of circuitry.
It looks as if, in some circumstances, the
plasticity doesn't occur in the excitatory glutamate synapse.
It occurs in the GABA synapse, impinging on the axon terminal.
Plasticity is about the GABA becoming more quiet.
And thus, this becomes more excitable.
And that has spawned another thousand generations in neuro
scientist named Hebb, studying what are the plastic changes
that will make a gaba ergic synapse more quiet, so
the glutamatergic one just below it becomes more excitable,
you could begin to see layers and layers of this.
It is a huge amount of work with all these different mechanisms.
Huge amounts of conflict, what's the most important step?
Is it the phosphorylations, is it the increased glutamate released.
People who study LTP all
fall into these different tribes based on
which mechanism they think is the most important.
But then, then came the biggest emotional crisis for the LTP field
which, as I said from day one, is all about the hippocampus.
The hippocampus, the hippocampus is the place
in the brain for learning and memory.
You want to have a hippocampus it is the part
of the brain which is blown out of the water
in Alzheimer's disease and alcoholic dementia
and a whole bunch of other disorders.
It does learning and memory for you, it is the most wonderful part of the brain.
I've lived in the hippocampus for the last 30 years.
Everyone who studies the hippocampus is basically insufferable
snots because it's the best, best part of
the brain because it does LTP it does
learning, it does SATs for you all of that.
And what the crisis
was, was the discovery that oh, all sorts of
far less classy parts of the brain do LTP also.
This was this huge crisis for the LTP hippocampus sort of elite.
The discovery that LTP is a phenomenon all over the brain.
For example, next door, in the amygdala.
We've already heard, amygdala is about fear.
It's about anxiety. It's about aggression.
Suppose you have a traumatic event.
Suppose you are a rat, and you get a
shock now and then, and it involves some learning.
Just before each shock, a little bell rings, and
a la Pavlov and all of that after awhile,
the rat will learn to freak out not just
at the shocks, but at the sounds of the bell.
You hit the bell,
and the amygdala activates like crazy.
And stress hormone levels go up, and the
animal freezes, and it has undergone fear conditioning, jargon.
In other words, it has learned something about the
world out there built around things to be afraid of.
And what do you know, it turned out that requires LTP in the amygdala.
Fear conditioning involves LTP going on there
as you learn something new.
And what fear deconditioning is about, is how do you unlearn processes like that?
Okay, so we got LTP going on the amygdala.
Meanwhile over in another county in the brain,
you've got dopamine, having to do with reward and
reinforcement and all of that, and you are
abusing some drug that causes dopamine release, *** for
example, and you always use it in the same location.
You always buy it on the same street corner, you always,
you have a context that you associate with using that drug.
And what you begin to do is learn that context, so that
you finally stop taking the drug, for a year, for ten years, whatever.
And you come back to that same street corner 50 years
later, where you haven't taken that drug in half a century
and get back into that same context, and you'll feel a craving.
You have learned to associate that setting, that street corner, that stuffy
armchair in the men's club where you used to get ***, whatever.
That context you associate with that drug and
the glutamate neurons projecting onto those dopamine neurons
will have undergone LTP.
So this is the part of the brain having everything to do with addiction.
You get a chronic pain syndrome.
You get an injury, a certain type of burn or a back injury where you know some of
the circuitry involved, and one of the things that also happens is those C fibers
the chronic throbbing ones, they can get LTP going on in your spine.
LTP occurs all
over the place. This was a gigantic crisis for the field,
all the hippocampal LTP elite having to come to terms with, it's not
just about the hippocampus, learning like this is occurring all over the brain.
Okay.
And, what's important there is a contrast between
some types of learning we know are good.
Some types are not so good.
The amygdala can be very wise in learning you
to make, to make you to be afraid of certain things.
Or it can be totally irrational in other cases of LTPing your way into a new fear.
And we call those neuroses and phobias and things of that sort.
So the contrast there between classical hypocampal type
learning and all these other forms of learning.
One other interesting thing that came out of this literature.
Okay, so just limiting
yourself to the world of hippocampus LTP, if you're
into the hippocampus and you're into learning and memory.
Glutamate is your friend.
The NMDA receptor is the most important receptor in the universe, the more
non NMDA's the better, its totally
wonderful, glutamate is about learning in memory.
Meanwhile, at the conference that you're at,
you've gotten the rooms confused, and you
go into the room next door by accident, and there's all the neurologists sitting
there, talking about stroke.
And talking about seizure and talking about concussive head trauma
and all they're about is how much glutamate is the biggest
villain on earth because all of the neurological insults like
that involve neurons, glutamatergic neurons,
glutamatergic synapses, spiraling out of control,
where you wind up releasing too much glutamate.
For example, during a titanic seizure, titanic on a level 100,000
times more than what we've been talking about here, or the presynaptic
neuron loses control over glutamate, because it's running out of energy
and just dumps all of its vesicles, because you've had a stroke.
There's no oxygen, glucose, whatever.
In those cases, you've got calcium coming into the postsynaptic neuron.
Calcium to an extent that winds up killing the neuron.
LTP, learning and memory.
You get enough calcium coming in to cause the cytoskeleton to get cleaved
in that dendritic spine and change the shape of it and isn't that wonderful?
And you've got a new factoid under your belt.
Have the glutamate come out that occurs during the stroke, and
that's a hundred gazillion times more, and now enough calcium comes in,
not just to cleave the cytoskeleton in that dendrite
spine, but cleave the cytoskeleton in the entire damn neuron.
And it collapses and dies on you.
That's one of the mechanisms by which neurons die after
that, and at an extreme people in the room you've
wandered into by accident, the neurologists rather than the neuro
biologists. For the neurologist, glutamate is not an excitatory neurotransmitter,
their jargon is its an excitotoxin.
This is a synapse, which because it is using this big scary fast tracked
dual receptor system that can have light bulbs go on for you, the calcium
turns out to be really a potentially very damaging mediator there.
All of the acute neurological insults to
the brain, involves synapses losing control over glutamate.
So that is another version of too much of a good thing.
Okay, so at this point, people were learning enough about LTP,
and how it works to begin to appreciate things that regulate LTP,
regulate as in things outside the little provincial world of that synapse alone.
So right off the bat we have a
compound which is extremely effective at screwing up LTP.
It blocks a whole bunch of mechanisms of LTP in
the hippocampus, a compound, a little old chemical called Ethanol.
ethanol.
Alcohol totally shuts down LTP in the hippocampus.
What's that about?
That's a, I remember not a damn thing from last night at all.
And that's at an extreme, the building blocks of alcoholic dementia, you need
to kill neurons there, which alcohol is
remarkably effective at, at remarkably low concentrations.
Okay so we've got one external regulator of
the LTP going on there. Ethanol disrupts LTP.
Something else that disrupts LTP in the hippocampus,
which makes perfect sense, which is, glucocorticoids, stress.
Get yourself good and stressed and you're not going to remember anything because
you're too frazzled, by what's going on there to be undergoing LTP.
And that's the whole process
by which severe stress can block memory formation.
Glucocorticoids block a whole bunch of the steps
that I just showed you in the hippocampus.
Next door though, in the amygdala, glucocorticoids enhance LTP.
What's that about?
That's why stress can increase the risk of an anxiety disorder.
Stress makes it easier
to become afraid of things that are not really justifiably
worth being afraid of, while disrupting memory formation in the hippocampus.
When you put those two pieces together,
you get a really, really interesting phenomenon.
So, you have somebody who underwent some unspeakable trauma,
years ago perhaps, and they find themselves in a circumstance.
They're in a crowded
room at a party, whatever, and they have absolutely
no idea why, but suddenly they're in a complete panic.
Suddenly their heart is racing.
Suddenly they are frantic.
Suddenly they are having flashbacks to that horrible trauma.
What is this about?
And it's not until later they realize that person standing behind
me, his voice, reminded me of the voice of that person,
the one who did that thing to me, where there's not a conscious memory of it.
What is gone on during severe trauma, with enough glucocorticoids
and stress going on, the hippocampus goes offline during that time.
So you do not file away a conscious memory of the events,
but the amygdala sure as hell remembers, because you're enhancing LTP there.
What is that? That is the free-floating
anxiety that you get after certain types of traumas.
So, the effects there absolutely opposite effects
on LTP in these two different brain regions.
Okay.
Shown here is another really interesting regulator of LTP.
So, up on top, we have a synapse, a glutamatergic synapse
in the hippocampus, going about its business, and it's doing some LTP.
It's learning some new facts and fortunately,
that organism has had a big healthy, breakfast that day, whatever.
It's got lots of energy on board.
The postsynaptic neuron has a ton of energy.
And it's got so much energy, that it does something bizarrely strange.
It does something straight out of like conspicuous consumption in economics.
What does it do? It drools out some ATP.
From its dendritic spine. How bizarre
is that?
ATP comes out of the dendritic spine, and goes floating backwards and
even more bizarre, binds to ATP receptors on the axon terminal.
We've got the weirdest retrograde neurotransmitter in the
entire universe, we've got this retrograde ATP signal.
And what does that signal mean?
By definition, if the postsynaptic neuron can
afford to be drooling ATP out of its dendrites, it means it's got so much
energy onboard that it's saying to the presynaptic
neuron, increase the amount of glutamate you're releasing.
Bring it on, I can handle it.
All the pumps are crazy.
We can handle calcium wave after calcium wave.
ATP is a presynaptic signal, is a retrograde signal to
bias towards LTP.
Meanwhile, at the bottom here, we have the opposite scenario.
Supposed the postsynaptic neuron is running out of energy.
What is that going to look like on a molecular level?
We know the ATP is going to all get turned to the ADP and AMP,
and if its a really really really bad, the AMP will eventually become adenosine.
And adenosine at
that point travels retrogradely, where there
are adenosine receptors on the presynaptic neuron.
And what does it do?
It decreases the amount of glutamate being released.
It's a negative feedback signal.
It's the postsynaptic neuron saying, I can't take it anymore.
I don't care how many facts you need to learn for a final.
I'm severely hypoglycemic.
I'm not going to remember. This is why if
you haven't eaten for six straight days you probably
shouldn't try to take a final exam at that point.
Synapses have this regulatory loop, they're built around energy availability.
This is why energy enhances memory
formation, because it's an incredibly expensive process.
And there's lots of reasons to think
that epileptic seizures, sustained seizures finally stop when
the adenosine levels build up enough to
retrogradely finally shut down the glutamate release.
So totally logical regulation going on here.
Okay.
So, this is great, this is great and
that this is this entire world of plasticity built
around Hebb's universe of changing the strength of
pre-existing synapses and that was perfectly fine for everybody.
Let's see if I can put this up. Aha, okay.
It's the last time it stabs me.
What you have got was total
explanation, incredibly elegant work being done, figuring
out all the different versions of LTP in different parts of the brain.
How does the opposite occur? How does forgetting every occur?
What are the mechanisms are at different brain regions?
What are the regulators are, et cetera, et cetera.
All of it built around this revolution from Hebb in the 1950's.
What plasticity is about in the brain
is changing the strength of pre existing synapses.
And that was perfectly fine, until around
the 1990's when the next revolution came in.
And there was one built around techniques getting
good enough for people to suddenly see something amazing.
Something amazing
[SOUND] along precisely these lines.
Which is every now and then a memory
formation is not about strengthening a preexisting synapse.
It's back to 1910 all over again. It's about the formation of a new synapse.
Incredibly elegant techniques came in where you could
observe the same little stretch of a dendrite branch
on the same neuron in an awake, behaving animal.
And watching it over time.
And people were flabbergasted to see synapses were forming all over the place.
Synapses were coming apart all over the place, an
incredibly dynamic picture going on there, where you would have,
for example, over the course of 20 minutes or
so a new dendritic spine would emerge, sort of mechanism.
[SOUND] Whoa.
That's wonderful.
There's no gravity here. Okay.
So what we see on the left is just happening to happen here.
So we've got. Oh, screw it.
You saw the picture, if you don't know it by now.
Okay, so what. [LAUGH]
>> Okay.
That's just going to give away the next punchline.
[NOISE]
That worked!
Okay. >> [LAUGH]
>> So. [LAUGH]
>> I don't think the physics class on the
other side, they got the good side of the board.
That doesn't happen.
Okay, so, what we see here is, if you've got our typical synapse doing
its thing here, and we've got a stretch of dendrite, with a dendritic spine.
And we've got the usual enough sodium coming in,
and all you need to have is another version
of this motif of, if enough sodium is getting in to start depolarizing
around here, that's a signal, for the formation of a new dendritic spine.
And that can occur literally over the course of minutes, and these totally cool
studies where you can show time lapse photos and here comes this new dendritic
spine, and at that point, its not hard to come up with a mechanism
by which it starts releasing some trophic factor, which causes some
axon terminal to wander its way over, and form a synapse there.
New synapses can form over the course of minutes.
And this was a major discovery.
And the discovery being that some types of learning require synapse formation.
So, there goes that whole Hebbian revolution of, it's only about
pre-existing synapses.
Sometimes plasticity takes the form of forming a new synapse.
Here's a great example of that.
If you would like to get a whole lot more dendritic spines, and thus synapses in
your hippocampus, probably the single best thing you
could do is to marinate your hippocampus in estrogen.
Estrogen promotes the formation of new
dendritic spines and new hippocampal synapses.
Hippocampuses get better at the sort of learning and
memory that they do when estrogen levels are elevated.
It is a great endocrine signal for accomplishing that.
Meanwhile, you can absolutely guess who the villain is going to be
that causes dendritic spines to retract
and disconnect synapses in the hippocampus.
Glucocorticoids. Stress causes the retraction of them.
Estrogen causes the enhancement of them.
And if you're a lab rat, and you ovulate once every four days or so,
you could look at the number of spines in the hippocampus of that female rat.
And it goes up around ovulation, and down two days later, and up two
days after that, and a constant dynamic changing of the number of synapses there.
And along with that, the sort of spatial memory
that the rat hippocampus specializes in.
Gets better around the time of ovulation, goes down,
and the exact same thing occurs in humans as well.
Spatial memory in women get better when estrogen levels are higher.
They tank at the other end of the cycle.
Cyclicity along those lines, averaging out to around median where males are
with testosterone levels.
So there we have changing the number
of synapses ever two days or so in the hippocampus if you're a rat.
Incredibly dynamic system in that regard.
Okay so that was a very exciting discovery.
Then people began to find the next layer of plasticity.
Not only can plasticity take the form of strengthening or weakening
a pre-existing synapse or changing the number of synapses, but now you
can also change entire processes, entire branches, of dendrites and axons.
That was the next discovery that came, that you can get massive remapping
of parts of the brain, in terms of what is projecting to what.
Neurons can, fairly rapidly, grow new branches
of their axons, new branches of their dendrites.
And conversely,
under opposite circumstances, they will retract them.
You can get dramatic remapping going on.
Okay, so it can occur in two general forms.
So, we have a circuit here, and those neurons
are talking to the next neurons in line, and
the first scenario on top, thanks to an injury,
a stroke for example, this neuron has been killed.
So you've lost a target
neuron in your circuitry there.
And what occurs over the subsequent weeks to months is, the neuron
projecting to it and obviously in this case, the stroke is going to
take out thousands of neurons and those thousands of neurons projecting, much
larger numbers but what you begin
to see is something called compensatory sprouting.
This neuron will begin to grow processes to
any place around where it can fight its way in and form some synapses.
What's that about, that's defining that
people months to years after stroke damage.
Will be slowly getting some degree of recovery of function.
You get rewiring going on there. Converse, on the bottom, opposite
type of injury here, thanks to an injury you've lost the projection neuron.
The neuron talking to an array of target neurons.
And in this case, you get exactly the opposite.
You get compensatory sprouting, the neurons on
either side suddenly see here's some unoccupied territory.
Nature abhors a vacuum, all of that.
And you get expansion of the projection fields, so now this
neuron, and this neuron, have a larger area of target neurons
that they're talking to, that they're influencing.
Two amazing examples of that, that have been shown a whole bunch of times.
What's this about?
First example, you get somebody who's congenitally deaf
thanks to something in the ear, rather than the
auditory cortex, and what you get over time is
atrophy of all the projections from the auditory system.
And, thus, what you have are a
whole bunch of neurons that are having no place to go.
There are neurons that have extra possible target sites.
And what you see is remapping of those neurons, such that people who are deaf,
when they look at you and you
just wiggle your fingers, their visual cortex activates.
But when they look at you, and instead,
you're using American Sign Language, their auditory cortex activates.
Those visual neurons saying, I'm seeing this person wiggling this finger,
if they have learned that this is, instead, informational, those neurons have
sent projections that have taken them 11 counties over, to God knows
where, and those visual relay neurons wind up in the auditory cortex.
Somebody who knows American Sign Language, when they
are watching somebody signing, their auditory cortex activates.
Totally amazing.
And then you see the exact equivalent in someone who is blind.
And they are adept at reading braille.
What you get there is, if they just touch their finger and you're asking
them, how many bumps are there here, you get activation in the sensory cortex.
The tactile cortex telling you two bumps or three bumps.
If instead they're reading braille,
they get activation in the part of the visual
cortex that a sighted person has when they're reading.
Somehow these neurons and this circuitry have learned
to go wandering over and take over that area.
And remarkably, a remarkable case report a few years ago a woman
who was congenitally blind and completely fluent in reading braille, had a stroke.
Had a stroke in her visual cortex, in the exact sort of region where, an
asighted person that's associated with loss of the ability to read, to decode letters,
and this woman was perfectly capable of saying this one is two dots, this
one is three dots, but she lost the capacity to be able to read Braille.
A visual incapacity of reading in someone
who's visual cortex was being activated by tactile information.
Incredible potential for plasticity.
What's most amazing is this sort of remapping stuff, isn't just restricted to
circumstances of disease, congenital, sort of circumstances, injury, whatever.
It occurs in everybody all the time. Great example.
You look at serious musicians,
take some serious, oboe player who's
abusing their propranolol and salivating like crazy.
And in addition to that going on that's going to make them
kind of socially unsightly, the additional thing that's happened there is,
you stick' em in a brain scanner and you play various
sounds in their auditory cortex, which happens to sit around here,
goes lighting up normally, but what you see is a greater percentage
of this auditory cortex, is devoted to responding to the sound,
not just of music, but to the sound of an oboe.
The cortex is remapped in that individual so that
more space is devoted to that person's musical instrument.
Another example of this, this was this amazing study in nature about 20
years ago looking at the fact that the hippocampus, as I noted, not only
specializes in memory but it's very specialized in spatial memory.
Remembering spatial maps.
And what the people looked at in the study
were individuals who professionally have to have spectacular spatial maps.
It was published by a group at University
College London and what they looked at were London
cab drivers, who have incredible spatial maps and
what they showed was people who had been driving
cabs for 10 years or more, their hippocampus had gotten bigger.
And it was bigger as a function of enhanced spatial mapping.
Spend the summer learning how to juggle and someone will expand
the motor cortex devoted to that hand stuff and then they stop.
And you come back six months later, it has shrunk again.
Take a female rat when she's nursing and the amount of tactile cortex
responding to *** stimulation will have expanded.
Take a rat and put it in an enriched environment and its
hippocampus gets bigger, all these forms of plasticity going on all the time.
Now this led the way for what is arguably the biggest
revolution in all of the neuroscience in the last quarter century.
Back to 1900, the stupidest neanderthal idea on Earth, which is a new memory is
coded in the form of a new neuron.
Everybody has learned ever since that the adult brain doesn't make new neurons.
Massive revolution that had its first sort of squeals
in the 1960s but wasn't really appreciated til the 90s.
The adult brain makes new neurons.
The hottest topic in the whole field, adult neurogenesis.
It occurs all over the brain, but the two
hot spots are, the hippocampus, and weirdly, the olfactory system.
The olfactory system just because every time
you smell, take a whiff of ammonia, and
you'll have just killed like half of your
olfactory receptor neurons, and you gotta replace them.
That's kind of boring stuff, but the hippocampus is interesting.
There's a population of neural stem cells sitting there, and it turns
out that the adult brain makes new neurons.
90-year-old people, if you put them in a
more stimulating environment will make a new hippocampal neurons.
And this has been this massive revolution.
And it's this big dramatic epic story where two guys who started the
whole field in the 60s, had their careers destroyed, because nobody believed them.
And this is well recognized.
And all sorts of drama, and egos, and
backstabbings and such.
And this is the hottest topic around, adult neurogenesis.
What stimulates it? All the logical stuff.
Learning, an enriched environment, exercise, estrogen.
What inhibits neurogenesis in the adult hippocampus?
You can give me a list right now also. Stress,
glucocorticoids, inflammation, and at this point, what's
been seen is, with all these different
pieces of plasticity put together, different disorders
will change the size of different brain regions.
People who are clinically depressed for a
decade or more, their hippocampus shrinks in part
due to loss of new neurons and part due to retraction of dendritic processes and
loss of synapses, in part due to death of neurons.
You change the size of somebody's hippocampus.
You look at somebody with post traumatic
stress disorder, and their amygdala is bigger.
Stick them in a brain scanner, and look how big the amygdala is.
And because they went through some unspeakable hell back when,
you've changed the size of this part of the brain.
Tremendous amounts of plasticity.
One last piece of it. Which is in the lecture notes there.
Okay have you guys had transposable genetic elements yet?
Transposons.
Uh-oh, some very puzzled looking faces and half yes's or no's.
Have you?
Okay, well now and then bits of DNA can get copied and jump
around and get inserted into other places in the genome, as I like
simplified this to horrifying extent, they're called jumping genes
transposable genetic elements, transposons, amazing epic story this discovery.
And people used to think, this was what plants
do because they can't run away from stressors, so
they juggle their DNA to try to come up
with something, and this is what path it's used to.
It turns out mammals have transposable events, where
they will copy little stretches of DNA and
insert new pieces in there.
When is the highest rate of transposable genetic events ever in
a mammal, just at the time you make a new neuron.
When you make a new neuron, you induce transposable events
in there, and most dramatically when you make frontal cortical neurons.
In other words, the same theme that's from two days
ago, when you are making the most important cell type in
your body.
And when you were making the fanciest type
in your whole brain in the frontal cortex, whatever
code of codes, holy grail of genetic instruction
and determinism you inherited at birth, isn't good enough.
Time to juggle your DNA a bit and induce some variability there.
Even your genetic code in your neurons is more plastic than in any other part
of your body. Okay.
So what have we gotten to by now.
We started with one single neuron.
We've now seen how to take over the whole body.
We've seen hopefully a theme from today.
There is nothing about the nervous system that should get stamped with inevitability.
Especially when it comes to neurobiology of
behavior, and the neurobiology of social behavior.
There is very little, by the time you get a brain as
fancy as ours, where you have to conclude
there's nothing that can be done to change things.
What you should also have as the take home
message is, as you now hear subsequent lectures, and in
fact for the rest of your life as you
hear people going on about how amazing their spleen is.
Or they're bursa in their bones or whatever.
You know in fact what the most important part of the body is.
But that will be our little secret.
Okay,
so that's it for the nervous system. [SOUND]