Tip:
Highlight text to annotate it
X
Okay. [COUGH].
Let's get going.
We continue with our expansive process and today we
make the great hurdle from two neurons to three.
Looking at anywhere from three to a couple thousand all in preparation for Friday
when we begin to see how your brain controls everything else in your body.
And as part of our progress here, into
the multi-neuron world, we have now switched convention.
This is a cell body.
This is an axon.
This is everything else, dendrites have magically devolved in the last two days.
This shows how serious we are about circuitry.
We use this schematic.
Okay, so what we begin to look at here is fancy
stuff that emerges only when you have whole networks of neurons.
And what we need to do to appreciate it, is first look at
a couple of the buildings blocks of how neurons can interact with each other,
at a different level than the past two lectures.
We start off up on top.
Okay, what we have there is
in your cerebellum, something called Purkinje cell.
And people love the cerebellum, because
its very identifiable neuron types, Purkinje cells.
So note with it, bereft of its dendrites, we got cell body, and we've
got the axon pointing to the right ready to talk to the next neuron.
However, violating everything you already know about neurons.
You notice there's an extra thing happening which is
a little collateral branch coming off of the axon.
Which implausibly goes snaking its way back to itself, back to its
own cell body or dendrite, and makes a synaptic connection on itself.
And what does it do there? It is an inhibitory input.
What have we just defined here? So now, let's look at things functionally.
So, we stick an electrode in that neuron and we're looking at when it fires.
When it's having its action potentials.
And right off the bat you will
notice there's the simulation of the purkinje cell.
Right off to the bat, off the bat to the left are two little action potentials.
What's that about random noise in the system, random hiccups?
That's from the other day ago.
We're talking about channels that are closed are only statistically closed.
Those that are open, likewise only statistically.
There's a certain level of background, spontaneous firing of neurons.
Which is why, the next thing becomes necessary.
So, we start stimulating that purkinje cell and
we start seeing a train of action potentials there.
That now come in the forms of just one vertical line and
a whole train of them and, with a delay,
the action potential has gone off on the collateral.
Back to the cell body, and inhibited the neuron and what do we get a silent period.
What are we doing here?
This is a negative feedback loop, and this is yet again the same theme of contrast,
trying to enhance the it's all over with. It's going to be silent for a period now,
with none of those spontaneous action
potentials, a way of sharpening the signal.
So an inhibitory feedback loop.
What we see in the middle is the exact same principle.
Another type of cell in the cerebellum, granule cell, and sending
off its usual axon, now the woefully familiar collateral coming off.
That in this case stimulates a tiny little cell and there gold
to sell.
And what that one does is, project
back onto the original neuron and be inhibitory.
What if we got, it's the exact same thing. It's a negative feedback loop.
What's the difference between the first one and the second?
So now we look at the firing patterns.
Same routine.
Start stimulating the granular cell there and you give the
burst of firing and it turns off with the silent period.
What's the difference? It goes on longer.
By having that intermediate step in there.
And you can now imagine if you
put eleven-dy different neurons in that collateral feedback
loop along series you could make an even
longer time before the silent period kicks in.
Just ways of sculpting the message coming through there,
build around this principle of neurons, sharpening their signal.
And sharpening their signal from previous days in
the forms of hyper polarization all of that.
Now we're seeing in a more macro level, volleys of action
potentials, and then ways to sharpen that signal, negative feedback loops.
Now, on the bottom, the third still in the cerebellum.
A type of neuron called a climbing fiber. And stimulate that.
And the Purkinje cell,
fires. And that's routine.
And a collateral coming off, but not
a collateral back on itself directly or indirectly.
But instead, onto a basket cell, which projects to that Purkinje cell.
And inhibits it. What do we have now?
At stimulating here, we're stimulating the climbing fiber and
recording another electrode. Recording from the purkinje cell.
So you see, you stimulate the climbing fiber and with a
delay you start getting your action potentials in the purkinje cell.
The delay that had to go all the way down the axon,
that whole deal, and then we get a burst here that kicks in.
Same exact principal.
This being a feedforward inhibitory loop rather than a negative feedback one.
Same exact principal though.
Principal with sharpening the signal. Okay, more of these principals in terms
of wiring where you begin to see ways of enhancing the signal.
So now we have some circumstance where there is an array of neurons.
Let's suppose for example these are an array of
neurons that are tactile receptors that respond to touch whatever.
And we're looking at a single line of them A through E.
And you stimulate neuron C there, just poking in that one little spot.
Pressing there and through the magic of stretch sensitive
neurons, what you have then is an action potential.
Neuron C starts to fire and does its thing downstream of that.
But notice now what it's doing.
In sending of collateral inhibitory projections
to the neurons on either side of it. B and D, what are you doing there?
You're sharpening the signal.
You're making it absolutely clear that we're getting poked with a pin
at C, not at B or D, make no mistake about it.
By making B and D even quieter than usual, You're enhancing the signal.
So we see there that you start simulating C, our same routine as
before, C starts firing like crazy.
And I conveniently forgot to put in a collateral of C back onto itself.
But you get the silent period there.
And meanwhile next door, b and d fall profoundly silent.
What have you done, phenomenon here termed lateral inhibition.
Lateral inhibition, once again a way of sharpening the signal.
Instead of sharpening
over time with the negative feedback loops or the feed forward inhibition
sharpening over space, just the way to make that system cleaner there.
And what you wind up seeing is, visually
all the sensory systems have lateral inhibition like that.
Ways of sharpening the signal, spatially, visually, whatever.
They all show the properties.
All the photo-receptors in your retina spend all their
time inhibiting each others, ways of sharpening.
Which brings up a strange question.
Okay, it's obvious here that, like you're pushing down there.
And you want a way of making sure you know it's there, and not one neuron over.
And you do that with lateral inhibition. And the same thing with this dot of light.
Versus over there, or an A versus an A flat, or whatever.
But what's lateral to the smell of an orange?
How's that work in the old factory system?
What's right next door to that is it like a tangerine smell or
something nobody understands what lateral inhibition is
about in olfactory and in gustatory systems.
Nonetheless, same sort of wiring there.
So we have that theme going on, a way of sharpening.
So now what you can being to see is, play the same thing out,
[NOISE] In a more rist-, realistic cellular field.
And one where this is the retina, this is just
the floor of the retina with all its neurons there.
And you simulate the neuron in the very center
with a laser to simulate one neuron at a time.
And you know exactly where the firing pattern there is going to be.
And now instead you stimulate a ring of neurons
around it, all of which are doing lateral inhibition to it.
And that central neuron becomes completely silent or totally confuse it and
stimulate both it and the neighbors and you get an intermediate firing rate.
These are all these tricks that go on for sharpening signals.
And what you wind up seeing when you do that is a principle of enhancement.
And back to the message from the first lecture there, if you've got a
hormone, it could start being secreted, neuron,
neuron stuff is sharp, all or none.
Contrast there what you see thanks to lateral inhibition,
is neurons typically take it typically even one step further.
Not just the square wave like that up
on top, but to exaggerate the differences even more.
The contrast at the transitions there.
A way of sharpening the signal even more.
And one is doing apparently the exact same
thing if you were like some Dutch master painter.
I am told that what they did to sharpen contrast on things that were well lit
was to put a very black outline line around it there sharpening the contrast.
The same exact principle. This over and over.
Okay, so now with things like this in place.
We can begin to look at one of the most
epically cool things that has ever been found out in neuroscience.
Which is how your cortex makes sense of sensory information.
This is like some of the most iconic neuroscience
ever done by a pair of guys, Hubel and Wiesel.
Hubel and Wiesel in neuroscience are like the equivalent of Ben and Jerry.
You cannot say their names separately.
They are utterly linked for all
of time.
Hubel and Wiesel, Hubel and Wiesel in the late 50's did what
was the best most amazing neuroscience research for that quarter century or so.
At Harvard, incredible process, where everyone went
berserk at their work because everyone thought they
had figured out how the cortex processes sensory
information all the way up to complex levels.
Okay, so here's what they did.
So you start off and you've got up on top
at the far left you've got your retina, just as described.
You've got a field, a two-dimensional field of photoreceptor neurons there.
So they're sitting there, and the way they work is
the retina senses projection to an area called the lateral geniculate.
And the lateral geniculate sends it's projection into the visual cortex.
Into the very first layer of the visual cortex, cortex the surface of your brain.
Now for our purposes the lateral geniculate
is totally boring so we've just eliminated it.
So there's no lateral geniculate for the rest of this topic, but
it's somewhere floating around in there between the retina and the cortex.
Okay, so we've got the way in which the retina
is talking to the first layer of the visual cortex,
the stellate layer. So here's what Hubel and Wiesel would do.
They would have a laser and they would
stimulate one single photoreceptor cell in the retina.
And they would be recording electrodes all over
the place in the visual cortex in that
first layer, and they would find one single
neuron there that would have an action potential.
And then, they would shift the laser over so they stimulated
the photo receptor right next door to that.
And right next door to that neuron in the first layer of the
visual cortex, that one would have an action potential shift, the laser shift.
And it would be this one for one mapping and what they called these
things were functional columns of all of
the neurons that respond to this particular area.
It was a one for one mapping spatially, totally straight forward obvious.
And what that tells you is, in the
stelate layer, what do neurons there know about for
a living, each neuron knows how to recognize one
dot of light, one single photoreceptor that's been stimulated.
Okay, so this took them about a half dozen years.
And they took the weekend off and Monday they come back and they
decide it's time to shift up to the next layer of visual cortical cells,
the layer that the stelate neurons talk to.
And for reasons I don't understand, that layer is called the simple layer
of the cortex, stelate probably should be, but we're now up to layer two.
And they saw something very different there.
So we see our version here, the retina and the stellate.
You stimulate those three retinal neurons and
you get the corresponding three stellate neurons.
Okay, that's what we see. So here's what people in these would do.
They'd go back to their paradigm and
stimulate one photoreceptor neuron in the retina.
And its cognate one neuron in the stellate layer would have its action potential.
And nothing's happening in the simple layer.
So now they try stimulating two of them in the retina.
And the two adjacent neurons in
the stellate layer have their action potential.
Nothing happens in the simple layer.
Now three, three and suddenly there is a neuron there that fires.
Three shifted over one layer of cells, and the next neuron
over in the simple layer, shifted down, the next one down.
What do neurons in the simple layer know how to do.
Each one knows how to recognize
a line. A line of a certain orientation.
A line of a certain length. That's what it does for a living.
And what they began to do, was show what they termed, tuning.
Tuning of individual, simple layer neurons to certain orientations.
And they get your classic stuff like that where they would have a
bar of light and they would find in a center of the
sequence there the orientation to which
that particular simple cell was tuning, tuned.
It had its highest frequency of firing.
And as it deviated more and more from that, its firing rate went down.
And next door to it would be the next neuron over
that oriented to a line a little bit over from that.
Below that would be ones of a slightly different angle.
Suddenly you have neurons that can extract information about single
dots of lights, layer one, and from that, extract information now about lines.
Major, major cool finding.
So that's nice on sort of this informational level.
Let's translate this back to the previous two days.
So how would you have wiring from the simple
layer to the stellate layer. How would you pull this off that one
photoreceptor being stimulated doesn't do it, two doesn't do it, three does do it.
And by now, what we can guess is wiring something like on the bottom here.
This is straight out of the principles
from last Friday, any given neuron's input is
hardly ever going to be enough to get the neuron to have an action potential.
Stimulate only neuron big A, a, and that's not enough
to get the neuron, the simple neuron, the action potential to stimulate A, a and B.
A and B, A, whatever it is I put up there, and that's not
to, it's not until you do all three that you get the summation stuff.
This is straight out of Friday.
This is the multiple voices need to be heard at once,
that's how these neurons are tuned. Okay so what you see on the bottom there
is wiring for you need to stimulate those three photoreceptors at the same time.
Exercise to go through later on your own.
How would the wiring look like if instead
of you need to stimulate these three photoreceptors
to get that one to fire instead you have to move the light from one to
two to three.
How would you do the wiring on the neuron that represents a
line moving, the light moving there from one end to the other?
What would the wiring have to look like to pull that one off?
With the same principle again, stimulate number one by
itself doesn't do it, two doesn't do it, three.
And how would you wire it up so that it
not only can pick up the moving light but different
neurons tuned to different speeds of the moving light.
[COUGH] And that would be the same sort of logic with the wiring.
Okay so that was totally great and by this point they are like
the most acclaimed scientists on Earth and they do the next logical thing.
They go to the next layer of the visual cortex.
Looking where all of those simple neurons that know one line and one line only.
Where they talk to
the next layer.
And this just took them another gazillion years to do.
And they got something even cooler. We are now up to the complex level.
And by now you know what's going on in the first three columns there in the retina.
Stimulate one at a time, and you just single stellate one, and if
you get them, a bunch of them, you can do single simple one.
And now what they show is, if you've got that line, a line that is sufficient
to get simple cells going, you move from one to the next to the.
And suddenly, a neuron in the complex layer has an action potential.
What does that one know about lines that are moving through space?
Lines that are moving in
different orientations, different speeds, different directions?
And that same spatial map again, where you have a line moving like this.
And this neuron
has an action potential.
And you have one moving at a slightly different angle.
And it'll be the neuron next door.
Slightly different next door. That same spatial mapping stuff.
What do neurons know about, in the complex layer?
They know about lines moving through space.
Totally great.
Everybody loves Hubel and Wiesel.
It is the greatest work they ever done, they get their Nobel Prize.
And everybody loves them because it's obvious what they've just figured out.
Which is how your visual system recognizes complex stuff because it's obvious what
happens next which is above the complex
layer there's gotta be a hyper-complex layer.
And what it will do is recognize two lines.
They're both, that is the super-duper complex layer
that will represent three-dimensional curving of the lines.
And layer after layer and layer, until all the way at the top
of these hierarchies of extraction.
Somewhere up there, there's going to be a neuron
that knows one thing and one thing only,
it knows how to recognize the face of your grandmother when her head's at this angle.
And right next door to what would be another neuron that recognizes
your face at this angle, and next door, one at that angle.
And that's how your brain is going to recognize
complex visual stuff, just layers and layers of extraction.
And literally, people in the field called these neurons at the top of these
hierarchies grandmother neurons, neurons that extract layer
after layer and recognize one complex figure.
And where you would have the same spacial orientation.
This was great, by then every else had learned the
techniques Hugall and Bezel had pioneered, for this micro stimulation stuff.
Everyone goes afterward to look at the hypercomplex
and the super duper hypercomplex, and look for grandmother neurons.
And Hubel and Wiesel interestingly, stopped working in this area at this
point and shift over to a different topic in neuro science entirely.
And everybody else goes looking, and everybody else in all the decades since
have almost entirely been disappointed, because there's
almost no grandmother neurons in the brain.
There's almost no neurons
that know one thing and one thing only, which
is being able to decode a complex visual thing.
There's almost none of them.
Major disappointment.
Generations of neuroscientists leaping out windows.
And this is logical as soon as you stop and think about it.
Okay, how many stellate neurons do you need?
Each one has a one-for-one connection with a retinal neuron.
So the answer is obvious.
Whatever the number of retinal cells is, that's
the number of stellate layer neurons you need.
How many simple layer neurons do you need?
Well, one for a line of this length, and one for a line a little
shorter, one for a little shorter, and
one that's moving, and along that, and different.
And you need like an order of magnitude more simple cells than stellate cells.
How many complex
neurons do you need?
And you got a line moving this way and a line moving that
way and a line moving and, like another couple of orders of magnitude.
And you do the math and you suddenly realize, by the time you get
to this hypothetical super duper complex layer,
you don't have enough neurons in your brain.
You can't do it that way.
And with a couple of striking exceptions, no one's been
able to find grandmother neurons, that's not the way it works.
You just cant have layer after layer, because you run out of neurons.
And what people realized at that point is, this may
work great for all the sensory systems the first few steps.
And people were soon looking at the auditory
cortex, and layer one recognizes a single node.
And layer two could recognize chords and layer, but
nowhere up there is there one neuron that's going to
recognize a particular symphony.
Your gram of a neurons hardly ever cause it's over and over the same problem.
After the first couple of layers of Hubel
and Wiesel type processing, you'll run out of neurons.
And what this ushered in was this whole new
emphasis in the field of what's now called neural networks.
Neural networks, the notion that information is not stored
in a single neuron.
By the time you're getting to complicated stuff, information
is stored in networks of neurons, patterns, and excitations.
And it has to work that way, and people since have explore the extent
to which that's the case and it allows you to do something really special.
Which you can never do if you just had a Hubel and Wiesel brain.
Okay, let us look at
an utterly unrealistic simplified neutral network here,
none like this exists in the universe.
But it's a two-layer one.
Here, first layer, you're on to A, B and
C, and you can see from the wonderful multicolor here.
A projects to one, two, three, B to two, three and four, C to three, four and five.
Okay, suppose, incorrectly, over-simplifying, suppose
this first layer, these neurons,
are Hubel and Wiesel neurons.
These neurons know one thing and one thing only.
This one, [COUGH] knows how to recognize paintings by Gauguin.
This one can recognize paintings by Van Gogh.
This one by, who's another impressionist artist?
>> [INAUDIBLE]
>> Okay, whoever the third guy is, I never learned this stuff.
Waste of time.
Okay,
>> [LAUGH] >> So you've got these neurons here.
Each one
recognizes one painter and one painter only.
So now look at their projection [COUGH] patterns.
What does neuron one know about for a living?
It's only getting it's information from A.
It's another Hubel and Wiesel neuron.
It knows the exact same things, this one,
this one knows how to recognize paintings by Gogan.
What does neuron three know about? Neuron three is
the one that tells you I can't tell you the
name of the painter, but it's one of those impressionist paintings.
Neuron three can extract categories out of the individual examples.
Because neuron three is at the intersection of all of
these I know one fact and one fact only neurons.
Neuron three is the one that recognizes impressionist painters.
Neurons two and four
do that as well but they're not as accurate.
Because they don't have as many examples to work with.
And this is exactly how a network is going to come out
where Hubel and Wiesel world, if the whole brain was like that.
You could only recognize individual paintings you could never say, is one of
those impressionists is not a cubist, you could never do broad categories like that.
But you could never do something else
as well.
Okay, so here we have this network, and showing how simplified this is, at
the same time neuron two is part of network going in this direction.
And it's a network of french names your not sure if your pronouncing right.
And there's another network going this way, and people
who's in museums like names your trying to find out.
Another network here, And in
all these cases any given neuron may be in a pivotal
extractive role in one network and may have a peripheral role.
And what you begin to have here: associations.
This is how you can have some
Impressionist painter suddenly remind you of some impressionist
composer, and in some way the visual thing,
and the auditory thing seem to be related.
This is how you could be
into having metaphors of, of food tasting hot, a
food having a temperature, a personality being a cold person.
This is where you begin to get metaphorical stuff.
The fact that you've got overlaps of these networks.
And any given neuron is going to be part of
a different, a zillion different networks at the same time.
So now we look at an interesting property of
humans, which is, so you got a face here.
And somewhere at some appropriate early age kids begin
to recognize this is what a face looks like.
And there is some face recognition network that eventually
going to be able to do some extracting thing.
Saying, I don't know who it is but that's the face
of a human and not a red panda, that sort of thing.
And then every now and then you get a human instead
who thinks that something like this qualifies as a normal face.
What have we just defined here?
Somewhere in there, there's a neural circuitry
in this kind of neighborhood that explains creativity.
That explains linking things that other people may not necessarily.
And suddenly you have 20th century art.
Where this is nowhere near as weird looking as it used to be because, aha,
somebody almost certainly had networks that spread more widely.
That were more associative.
Almost certainly creativity has to be something like this.
When you take a concept from particle physics,
and you apply it to modern dance, or so.
That's suggesting that you've got a wider
set of networks than most other individuals do.
In some way, this is how creativity
has to work. So, neural networks coming out of this.
Okay.
So, now we shift from this totally sublime subject
to instead the neural networks that move your bowels.
[LAUGH] Okay.
So, what we've got here is how your bowels work.
And if you like removed your intestines and put them
on the table here every now and then they would contract
on their own.
Just like heart muscle there's an endogenous myogenic
quality to them they contract on their own.
What does your brain do?
It either makes them contract more than usual or contract less than usual.
That's all it does bi-directional control.
And we will hear about the system that controls that by next week.
Okay. So here
you've got your gut. And you've just eaten something.
And it's a big old bulbous lump of food in there and what you want to do
after you have finished extracting your nutrients from
it is you want to get rid of it.
Number one.
And number two, you want to get rid of it in the right direction.
So, our task is to move that bolus of crud from
the left to the right, so note here the circuitry we've got.
And, I'm going to have to note it to just to get the numbers.
Okay, so notice number one is stimulating number two.
Number two inhibits contraction of the muscle there.
So that's the normal state you haven't eaten anything your bowels are
sitting there while in an uproar
they're absolutely silent they're are not contracting.
That is because a whole series
of neurons like one is talking to two, six is talking to
seven and the net result is you're inhibiting contraction of your bowels.
Now instead of the circumstance, we've got that clump of food sitting there.
Clump of food that is descending the wall, the muscle wall, of the bowels.
And sitting there is a stretch receptor neuron, neuron number three,
and what it can tell is, when the muscle is being descended.
When it's being stretched, aha there's a lump of
something in there, number three has an action potential.
So look what happens now, first thing it does
is number three talks to number four, stimulates number four.
First thing that number four does is projects onto neuron number two and it
inhibits it.
In other words, we've got a double negative here.
Neuron, number four, is inhibiting neuron two's inhibitory actions.
And as a result, what happens is, the muscle squeezes there.
You're allowing its endogenous contraction to occur it squeezes there.
That's great, so suddenly you're squeezing the left side of your bowels
there but you've got to do something else as well.
And notice neuron four is also projecting
to neuron five which stimulates neuron seven.
Neuron seven as a result really, really, really inhibits
the stomach wall of the intestinal wall down there.
What are you doing?
You're squeezing on the left and you're loosening up
on the right to push the stuff to the right
there cause you want to make sure that's where it exits.
What you're doing here is not only
getting peristalsis, contraction, you're getting directional peristalsis.
And the way you do the wiring
there is you're continuously squeezing just behind the
bolus and loosening up just in front and
encouraging it to move in the right direction.
So, look at the wiring circuitry there. This is exactly
how you would design say a logic shift that moves your
bowels for you if you should want to do that for some reason.
And all the wiring there makes sense.
Notice one thing there though that makes no sense at all.
Which is, look at what neuron four is doing to neuron two.
This violates everything you know in your sleep by now from the last two lectures.
We've got an inhibitory neuron.
And that's straightforward.
That's fine.
On the lecture the other day, we can have a neuron project
onto another one and release an
inhibitory neurotransmitter instead of an excitatory one.
You hyper polarize, what's like to have an action potential, you know the drill.
But that's not what we have going on there.
What we have is something like this instead.
This is what the circuitry because notice where four is sending its axon terminals.
It's not sending it to the dendrites
of neuron two, it's sending down to neurons two's axons.
This makes no sense at all.
This is a completely different type of wiring than you've gotten.
Termed axoaxonic synapses.
So what we've got here is this neuron
is projecting onto the axon terminal here, not to
the cell body of this guy, not to
the dendrites of this guy, but the axon terminal
of this one.
And it releases its neurotransmitters there.
Okay, so the first thing that has to
predicted now, given how bizarre this is, is
if it's going to work like anything we heard
the other day, is of all the outlandish things.
There's gotta be receptors on the axon
terminals for the neurotransmitter coming out of here.
And that's totally bizarre.
You've got axons
making synapses on axons and axon terminals having receptors.
That's precisely what you see there.
So, you release an inhibitory neurotransmitter.
So now comes the critical question. Which is, what does neuron two
do to the firing rate of neuron three? And the answer is,
it doesn't do anything to number three. What does neuron do to, neuron two do?
What it does is lock the ability of neuron one, to stimulate neuron three.
In other words, we have just introduced the if-then clause.
A contingency into this picture which is number two
decreases the firing of number three if and only
if, number one, is in the process of trying to stimulate it.
What number two is doing is not
directly controlling three it's modulating number one's actions.
And thus, another term for it is, this would be a neuromodulatory circuit.
And this is the theme you see over
and over ways in which you've got contingencies,
one neuron is having an effect downstream.
If and only if another neuron is
trying to do something very different, repeated theme.
Okay.
So we've just seen this from two days ago version, which is,
whatever the neurotransmitter is coming out
here, there's gotta be receptors for it.
And it happens that the neurotransmitter is so famous
that I forgot to mention it the other day.
It's called GABA.
Gamma-Aminobutyric acid, and you know the song
and dances made from an amino acid precursor.
All of that GABA, every single inhibitory synapse
in large part of the brain use GABA.
And what GABA almost always does, is
make a presynaptic synapse, and also axoaxonic synapse.
In other words, most inhibitory neurotransmitter
type i the brain, Hardly ever is directly doing anything to neurons,
its always modulating preexisting if then clauses, preexisting pathways.
So this winds up being really important in
terms of we've just introduce this if then clauses.
Okay, so that's the version from two days
ago, neurotransmitter receptor, now lets do the version from
last Friday.
So what exactly is GABA doing here in terms of our ionic stuff?
And what it's doing is, it opens up chloride channels.
It opens up chloride channels and thus, chloride rushes in to this axon terminal
and it hyper-polarizes it. So this little stretch.
Okay let's make this much more realistic.
Even though this is completely unrealistic.
But, okay, so here we have this actually occurring somewhere around there.
So it's releasing GABA there.
And as a result of chloride rushing in,
you've hyper-polarized this little stretch of the axon.
Meanwhile over at axonhillic exciting things have occurred
and an action potential is working it's way down.
And we know how that one works okay, first time today we know how that one works.
Which is it's going to have enough sodium come in to open up the
[INAUDIBLE] channel it propagates and it regenerates
and except it suddenly hits this zone.
That's hyper polarized out the ***, and it's too
hyper polarized for a sodium channel to open up.
The sodium rushing in from the action potential just to the left of
it, it's not enough to get it to threshold.
What have you done?
You've just silenced that action potential.
You've just made it impossible for it to propagate further.
This is how these modulatory neurotransmitters work.
They're doing this in a way that shuts down this inevitable action potential.
And this is the wiring that you see over and over.
And now,
what you can begin to see is one version
of violating Dale's First Law from the other day.
Which is if you have one of these inputs on only some of the axon terminals
from the neuron and not the others, you're not shutting down the neuron as a whole.
You're shutting down some of its branches.
And that's a way to now turf
an action potential preferentially one direction over another.
This is a theme
you see frequently. GABA is interesting.
It is the most amazing neurotransmitter receptor
system out there, because it is incredibly complicated.
Because there's a bunch of things which modulate GABA's modulatory role.
Okay. What do I mean by that?
So we've got a very,
here's what the GABA receptor looks like. It looks exactly like that.
And it's sitting there.
In this case over here and GABA comes in
and does it's thing and that makes perfect sense.
However when you look more closely at the GABA receptor
it's got a little, oh let's make it even more realistic.
It's got a little insy thing there and what does it do?
A second thing can bind to the receptor.
A second thing that could come floating along
from neuron world, and what does it do?
Does it open up chloride channels?
No.
If and only if GABA is doing its thing, it keeps the chloride channels open longer.
Another layer of
modulatory if then stuff. So GABA comes in here.
And what comes in there? Okay.
So what does GABA do? It's inhibitory.
And one of the main things it's inhibitory for, is indirectly relaxing your muscles.
Gabba is very good through inhibitory effects, relaxing muscle pathways.
So, you have some muscle injury and you're getting muscle spasms all over the place.
What sort of medication might
you be given?
You've getting muscle spasms and you're trying to block them.
After some injury and you're Going for the silver tomorrow.
Yes, I know you're all saying, a muscle relaxant, he
says self-evidently, you take a muscle relaxant to relax your muscles.
What's a classic group of them? A class of drugs, do not write this down,
called benzodiazepines. So you take some benzodiazepines and
they start binding to the side sub-unit of the GABA receptor,
and it potentiates GABA's actions, and your muscles are now all relaxed.
Anybody recognize benzodiazepines, the term, and know
what they're used for in a different context?
Yeah. >> Treating insomnia.
>> Treating insomnia.
Okay, so your muscles relax, and the rest of you does also.
It's used for that, and very closely related
to that for something else.
And it's used far more frequently in that
realm than for relaxing muscles or even for insomnia.
Just playing out benzodiazepines.
Okay, I will give you the name of one of them.
The most famous, ***.
What do people take *** for?
Yes, I heard a whisper. Anxiety disorders.
It's the classic minor tranquilizer anti anxiety drug, and
benzodiazepine, wait a second it does stuff to muscle.
What's that about that brings up a really interesting
feature of regulation, of regulation of this system as follows.
Why should taking a drug that relaxes your muscles cause to
be less anxious?
Because it doesn't do a whole lot in other stuff in the brain there.
Why should that work? Okay.
You're sitting there and you're life's awful, and there's a comet coming at
you, and you're totally anxious, and you're a wreck and everything is awful.
But you've taken a muscle relaxant, and everything
is just as awful as it was before hand.
But now, you're so loose and relaxed,
you're dribbling out of the chair like Jello.
And somehow, your brain
gets suckered into thinking, if I'm feeling that relaxed
in my muscles, things must not be so bad.
You have a weird negative feedback thing where part of what your brain is
doing to decide what your emotions are is assessing what's going on in your body.
Weirdly, that's where the benzodiazepines exert most of their action.
And we'll see some of the wiring that comes into this in a few lectures,
how your brain deciding how you feel about the world based on signals from your body.
One other example of that, like
crazy example, which nonetheless works consistently.
If you take people with clinical depression, and you force them
at gunpoint to mechanically smile, and smile over and over and over.
They start feeling better because the brain
gets suckered in.
Once again saying, everything is just as awful,
mortality and the ozone layer, but wow, if I'm,
like, whistling a happy tune and a smile on my face it must not be so bad.
Part of your brain is sitting there trying to get information from your muscles.
And throughout your body, what is called interoceptive information.
So here we have benzodiazepines doing that, so we've now got this double layer
of if then clauses.
GABA inhibits this neuron if and only if number one is trying to talk.
GABA benzodiazepines are inhibitory if and only if
GABA is showing up at the same time.
So, totally complicated. Let us make it worse now.
One step worse, which is now there's a second binding
sight off to the side there.
And it binds in different class of compounds
and what they do is the same exact module.
That deal if and only if there's GABA around and
makes the chloride channel stay open longer, a whole lot longer.
A bigger modulatory effect than the benzodiazepines do.
These ones open up the chloride channels like crazy.
What are these drugs?
What are they used for?
And similar to your comment before, but take it umpteenth steps further.
Yeah? >> Anesthesia?
>> Yeah, surgical anesthetics.
The minor tranquilizers, the major tranquilizers.
A whole bunch of classes of surgical anesthetics.
Are insanely modulatory of GABA's actions and keep it
open a whole lot longer and you become unconscious.
So cool.
Second one. There's a third binding site off
on the side there that's even more interesting.
[NOISE]
And that's what it looks like and it binds a particular hormone.
A hormone that is known to decrease tension and anxiety in people.
A steroid hormone which is secreted in
females predominantly at certain times of the cycle.
Any guesses?
One that makes you calmer and nurturant and
socially affiliative and all that sort of stuff.
And preparing for female birth, and all this gestation stuff.
Progesterone, a pro-gestational hormone.
Progesterone, it's levels drop dramatically right
around the time that one's period starts.
And that drop has much to do with the mood
shifts at the time that occur in most people
but not universally progesterone has it's own modulatory site there.
And what it does is it works as a minor, minor tranquilizer.
And it works there as well.
And it is sufficiently the case that in the fifties there was
a surgical anesthetic that was used that was a derivative of progesterone.
So look how totally nutty and complicated
this one receptor is, layers and layers of if then clauses.
If and only if this is around it makes this
better at turning this off if and only if this one.
You begin to get a sense of where the computational complexity come out there.
Okay.
One last thing about your bowels before we leave it for the day.
So here we have.
These are the small intestines. What do your small
intestines do?
They extract nutrients and they put water in there to help hydrolyzing molecules.
They do all that stuff. Your small intestines are great.
So suppose you have an emergency.
Suppose you have stepped outside and there once again is
the rhino chasing you, you are running for your life.
You need lots of energy to save your life, your body
is doing all sorts of metabolic stuff to get energy to your thigh muscles.
And what it's also doing during stress
is shutting down everything that's not essential.
It's shutting down reproduction.
It's shutting down long-term antibody [COUGH], it's shutting down everything.
What happens during stress? You shut down digestion.
You shut down the digestive tract. And the logic there is obvious.
If you're running for your life, trying to avoid
being somebody's lunch, you don't worry about digesting breakfast.
Do it later if there is a later.
We all know the first step of this.
Suppose you get stressed speaking in public, your mouth gets dry.
You've stopped secreting saliva. The first step of this cascade.
So with the onset of stress, you stop secreting saliva,
and your stomach stops its churning stuff, and your small
intestines go silent.
And it should be easy to figure out which
of the neurons in that pathway are stimulated during stress.
So great, we understand the entire world of your gastrointestinal tract
until we hit the large intestines, where something very different occurs.
Okay.
What are your large intestines good for? Nothing.
Nothing that they don't absorb.
Nutrients, all they do is like push things along miles of tubing there.
And they extract water from there so you don't dehydrate every time you defecate.
And it gets stuff out the other end and everything works fine.
All your large intestines are about is just like balancing electrolytes
and stuff, and then getting rid of the dead weight there.
So now, you've got the rhino coming after you and you've got a choice.
You could run for your life, with or without
six pounds of dead weight in your lower torso there.
What happens during stress, you increase contraction on your large intestines.
And that's the reason why when people are executed, there executed in diapers.
Because people loose control of their large intestine there.
And what you have, you can now do in your joyous free
time at the dinner table tonight, entertaining
people about bowel subtleties, is the wiring
has to be exactly the opposite in your large intestines than in your small.
You get totally obvious, you get totally opposite effects there.
So cool piece of wiring.
Same thing, same thing next door your kidneys.
Your kidneys balance your sodium and they put stuff
in and out that I refuse to ever learn back when but they do wondrous things.
And they make your urine and when they're
done the urine heads south into your bladder.
What's your bladder good for? Nothing.
It's just this storage bladder thing. It's just filled.
And once again you're running for your life with a rhino coming after you
and you could do it with or without a whole bunch of liquid sloshing around.
And the same thing, you get rid of the urine with stress.
You silence the kidneys, you silence the small intestines.
You can track large intestine and you can track the bladder.
Get rid of the dead weight, so different wiring there.
Okay.
So now that you know so much more about
your bowels than you did this morning at breakfast,
We
move on to another system. Okay, so what we have here is a circuitry.
Okay.
[LAUGH] That allows you to do all sorts of wonderful things.
Neuron B.
Neuron B is sitting there in your spinal cord and when
it is stimulated, when it has its action potential, it sends it's
axon way up your spine.
Heading into your brain to let you know there's something painful going on.
You have an owie. That's mediating the pain signal.
So look up on top. We have the A fiber.
You have poked your finger on a pin.
And your A fiber, which has a cell body that can detect
that, has its action potential, sends its message down there, stimulates neuron B.
And you begin to realize your in pain.
You have a collateral though, collateral coming off
the A fiber that is goes to neuron A.
It stimulates neuron A and the neuron A with a delay, inhibits neuron B.
This is the same feed forward wiring we had before.
What is this about? This is how sharp pain works.
You poke your finger, and
it hurts like crazy, and two seconds later the pain is passed.
This is sharp, eppicritque, sharp, short-lasting pain there.
And the way you pull it off is with a feed forward inhibitory loop.
Now on the bottom, you've got a C fiber.
With different wiring there. So C fiber stimulates B.
Same thing again, and what it does is inhibits A.
And as
a result, A doesn't inhibit B, and B just keeps firing.
What kind of pain is mediated by C fibers?
Slow, throbbing pain. That's a sore muscle.
That's a burn.
These are pathways that carry completely different types of information.
Okay.
Note up on top, we have our cinnamon bun schematic of a myelin sheath on A fibers.
But, there's no myelin on the C fibers. Why does that make sense?
Come on, some >> [INAUDIBLE]
Yep, it's part of this whole sharpening and you've got
that, and you don't really need myelin on the C fibers.
It's much more pokey.
Okay, you could find out that your muscle is sore now,
or you could find out in a second and a half.
And you don't see anywhere as much myelin on the C fiber.
Okay, so this is the basic circuitry and what is
so cool about this is that two guys, Well
and Malzac, or Mel and Walzac or something rather.
Decades ago came up with this circuitry diagram saying it has
to work this way before anyone had the tools to examine it.
And people went and looked and this is exactly the circuitry that you see.
Okay, so you can begin to have some cool things now.
So suppose you've got your C Fibers activated.
Some source of continued irritant, you've got a mosquito bite.
So the mosquito bite is itchy, it's irritating, and
it's really this thing by way of C Fibers.
Okay, so you've got the C Fiber mediated.
Mosquito bite itchiness that's driving you crazy, and
what did all of our mothers teach us?
Don't scratch it, couple of them will get infected, so do you do instead?
You scratch on either
side of it.
[LAUGH] You scratch really *** either side of it, and it stops itching.
How come?
Because by scratching hard there, You're stimulating your A fiber.
Your A fiber finally gets the B neuron to shut up.
By way of stimulating the A neuron it overrides the C fiber effect.
And you do it on either side. Taking advantage of lateral inhibition.
And suddenly by stimulating epicritic sharp pain
you can hibit, inhibit chronic pain [COUGH] pathways.
This winds up in all sorts of interesting settings.
Say for example, you have a horrible metabolic disease where
cells throughout your body are constantly starved for energy, diabetes.
So you've got type one Diabetes. And as a result all your cells
are a little bit starving. Your neurons especially.
Your more expensive neurons especially. So look at this circuitry.
Which is the more expense neuron, A fibers, or C fibers?
A fibers.
So in diabetes the A fibers often begin to get killed,
and what you got then is constant throbbing pain, diabetic neuropathy.
Or suppose you have like a back injury
or something, and the C fibers are always going.
Once one of the things are done clinically,
people get an implant when the pain is getting
bad, they push a little button on their hip
thing or whatever, which stimulates the A fibers there.
And there's a brief burst of pain.
And then the throbbing pain goes away for awhile.
And we all know this principle in another setting, which is,
like, your back is sore, and this is what
somebody's doing when they're giving you a back massage because
they're pummeling your muscles there and stimulating all these A
fibers to shut up the C fiber effects on B.
So the circuitry here makes perfect sense.
Now let us make it a little more complicated.
[SOUND] So,
we've got our beaner on here and somehow we've lost the rest of the circuit.
And now instead, in addition,
we have projections going to the B Fibers. Some stimulatory, some inhibitory.
What's this about, where are these projections coming from?
Your brain.
Your brain is talking to this pain
pathway, your brain is modulating how it works.
Lets give an example of how the modulation works.
Okay.
You're totally phobic about the dentist, and you walk
in there.
And the second you hear the music your teeth start hurting.
What's going on there?
Circuitry up in the brain, sending
projections down and stimulating the B fibers.
This is not real pain, this is you being all anxious and crazy about it.
Okay. Another setting low for it.
Okay, let us look at a circumstance where you can get C Fibers firing.
Somebody takes a piece
of sandpaper and sandpapers rear end rigorously
for awhile, and this is going to be uncomfortable.
In contrast, if your rear end is getting abraded like
that because you're having sex on a pile of sandpaper.
You're not even going to notice it at all.
What's going, it's the same exact input.
You're getting very different projections coming down from your brain.
Your brain can
modulate what the threshold is for this neuron
and suddenly you don't notice it at all.
And all sorts of stuff works this way.
Okay, so, how does your brain process pain?
You've got the projections going up there, the B fiber shooting up there.
And the first thing they have to do in your brain is totally boring, which is
to tell you, okay, it's my left hand that's on fire, it's not my left foot.
Just like, and it's on fire rather than being crushed by a vice.
And just like basic pain-o-meter type
information, and there's all sorts of parts
of the brain where it figures out
location, severity, and that's not very exciting.
But then, there's projections into parts of your
brain which we will hear about in a few
lectures, cortical areas that do all sorts of
evaluating things that tell you what the pain means.
And there you get something very different.
Okay, so you have this evaluative part of the brain in
the cortex, in the front there, and it's figuring out the meaning.
So suppose somebody pokes your finger with a pen and all the parts
light up there including that area, which says, oh, this is kind of unpleasant.
My fingers being poked with a pen.
But suppose instead, just before you, that the
devious grad students doing the research have said, oh,
we're testing out a new topical anesthetic and
we're going to smear some of it on your fingertip.
And it's blocking pain, so we want to just be sure.
And they've actually just put, like, sour cream
on your finger, and you get a placebo effect.
You poke the finger now, and the parts of the brain that
say, it's my finger and not my toes, they're working just fine.
This frontal evaluative part of the brain doesn't
light up.
And as a result, you get some inhibition going on down here.
That's how placebos work.
Now, something even more interesting in that part of the brain.
You sit there, and you get your finger poked, and the pain-o-meter
areas activate and so does this frontal area that says this is unpleasant.
Now instead, you sit there and watch somebody with
a pin poke the finger of your loved one.
The pain-o-meter parts of your brain don't activate, but this frontal area does.
And what is this?
This is feeling the pain of somebody else.
This is empathy where the neurons that figure out my toe is
on fire [SOUND] get involved in empathy it can't tell the difference.
That's why you can feel so strong.
Dramatic precursor to a few lectures from now, in this part of the brain,
there is hyperactivity in people with, clinical depression, because
they are pathologically feeling the pains of the world.
And at extremely severe cases of depression people
go in and destroy this part of the brain.
Okay, so we've got all sorts of modulatory stuff going on now.
We have all sorts of ways in which we've got subtle stuff happening here.
What are we ready to do in the next lecture?
Now's the time to see how the brain
is going to be controlling all the stuff outside.
Next lecture, seeing the first of two ways that it does it by way of hormones.
Hormonal projections regulating every cell throughout your body.