We've now learned about the complex process coupling an action potential
in a presynaptic bouton to neurotransmitter release
through complex molecular machinery.
The action potential firing rate in a neuron
makes a large difference about its synaptic transmission.
Some neurons, some types of cells in the mammalian brain,
fire action potentials at very high rates, almost continuously.
So, some cells fire at 100 Hz:
that's an action potential every 10 milliseconds,
almost continuously throughout our lives.
Other neurons fire much more sparsely and selectively.
They might be silent for minutes,
and then fire a brief burst of five action potentials
within a few milliseconds of each other,
and then remain silent again for many minutes.
There are thus diverse patterns of action potential firing
and the brain has adapted to make use, enhance, and suppress
some of the information transmitted by action potentials
in different time scales.
The amount of neurotransmitter that's released
in response to an action potential
is therefore dependent upon the history of activity in that particular axon.
If an action potential occurs immediately <i>before</i>,
then that leaves an imprint upon the synapse.
And that can change the amount of neurotransmitter
that's released in response to subsequent action potentials.
And that process of so-called <i>presynaptic plasticity</i>
is what we're gonna study in today's video.
The modulation of neurotransmitter release in response to activity of that same axon
is called presynaptic short-term plasticity,
or presynaptic dynamics.
It can thought of as occurring on three different time scales,
which also relate to three different mechanisms
that are at play in the presynaptic terminal.
We can have short-term synaptic plasticity:
that's a form of short-term dynamics
that occurs on a time scale of less than one second,
typically 10 milliseconds;
100 milliseconds is the time scale
of interaction between neighboring action potentials
when we're thinking about short-term synaptic plasticity.
It can be divided into two forms:
<i>facilitation</i>,
where subsequent action potentials <i>increase</i> the amount of neurotransmission,
and <i>depression</i>,
where consecutive action potentials
induce less and less release of neurotransmitter.
There are other forms of presynaptic plasticity
that occur on much longer time scales:
postsynaptic potentiation, that occurs on the <i>minute</i> time scale,
and long-term presynaptic plasticity,
that can last hours, days, perhaps even a lifetime
and might ultimately contribute to forming memories in the brain.
These two longer-lasting plasticities
rely upon complex signaling cascades downstream of activity, and calcium,
whereas these are <i>immediately</i> related to the calcium
and the amount of neurotransmitter vesicles available.
We will talk about bit about each of these four different forms
of presynaptic short-term plasticity in the next slides.
We'll begin by thinking about facilitation.
In the process of facilitation, we imagine a connected pair of neurons
where action potentials in the red neuron propagate down to the presynaptic bouton,
calcium influx causes neurotransmitter release,
and then postsynaptic potentials are evoked in the blue neuron.
One action potential causes calcium influx,
causes a postsynaptic potential.
Then we'll envisage a second action potential,
occurring shortly after the first one:
in this case, on the time scale of 20 milliseconds later.
The second action potential
induces a large amount of neurotransmitter release,
and therefore also a larger postsynaptic potential.
That's the process of <i>facilitation.</i>
The mechanism underlying presynaptic short-term facilitation
is thought to be an accumulation of calcium in the presynaptic terminal.
So, the first action potential induces a calcium influx
that activates the voltage-gated calcium channels;
calcium then floods into the presynaptic bouton,
and that's what causes neurotransmitter release.
There are a number of different calcium binding proteins, of course,
including the Synaptotagmin:
that's the effector for neurotransmitter release.
But there are many other proteins that also bind calcium.
They become calcium buffers.
There are also proteins that are involved
in sucking the calcium out of the presynaptic terminal:
calcium pumps that are involved in reducing the calcium concentration.
And so, in the 20 milliseconds that follow the action potential,
there's a gradual reduction in the calcium concentration
in the presynaptic terminal;
and that depends on the activity of the pumps,
the diffusion of the calcium
away from the entry site at the voltage-gated calcium channels,
and the binding of various calcium buffers.
When the second action potential arrives,
there may, then, be
a little bit of residual calcium left in the presynaptic terminal.
We may not have reached the equilibrium baseline situation.
Equally, some of the buffers may be saturated;
and when the second action potential occurs, and calcium floods in,
those buffers won't be able to bind calcium;
and there's already an elevated baseline;
and so the peak calcium concentration that's reached
in response to the second action potential
can be higher than that in response to the first.
And even if that's only, say,
a 20-percent increase in the peak calcium concentration,
that can make a major difference
to the amount of neurotransmitter that's being released.
In the last lesson we discussed that the release rate of vesicles
depended upon the fourth power of the concentration of calcium.
And so, if we have a 20-percent increase in the calcium concentration,
we put that to the fourth power:
that induces a doubling in the amount of neurotransmitter release.
So, even small changes in the amount
of peak calcium influx reached in response to an action potential
can have a large impact upon the release of neurotransmitter.
And in the process of short-term facilitation,
it is thought to be increased calcium signals
that give rise to increased amounts of neurotransmitter release.
And that's the process of short-term facilitation,
that typically lasts less than a second,
and is typically, sort of, studied on the time scale
of action potentials that are 10 milliseconds,
50 milliseconds, or 100 milliseconds apart.
There's another process of short-term synaptic plasticity
that does the converse: <i>depression.</i>
In the process of depression,
the second action potential gives rise to a smaller postsynaptic response
because of less release of neurotransmitter
compared to the first action potential,
when those action potentials are separated by a few tens of milliseconds.
In the process of synaptic depression,
we imagine that the first action potential is quite efficacious,
and it releases a large amount of neurotransmitter
from the presynaptic nerve terminal.
After that neurotransmitter has been released,
the vesicles need to replenish.
And in the time space of 20 milliseconds, perhaps,
there's very little time for the vesicles to replenish;
and so, in response to the second action potential,
less neurotransmitter is released.
We can envisage the situation here:
in response to the first action potential,
there might, for example, be two release-ready vesicles.
The action potential induces calcium influx;
one of these is released and causes a postsynaptic potential.
In response to the second action potential
that first vesicle is now gone;
It's probably somewhere in the recycling process,
trying to load up with neurotransmitter.
So there's only one vesicle, now, that's left;
and the action potential causes the same calcium influx;
and, in this case,
there's, then, a reduced probability of releasing a neurotransmitter.
It's also possible to modulate presynaptic release on longer time scales;
and this typically involves
high-frequency stimulation of the presynaptic nerve terminal
and activation of additional signaling pathways.
These sort of longer-lasting forms of presynaptic plasticity
only occur at certain types of synapses.
So, most types of synapses
have the short-term forms of synaptic plasticity,
and only certain ones have, for example, <i>post-tetanic potentiation,</i>
or long-term potentiation, that we'll discuss next.
And so, at synapses that express post-tetanic potentiation,
we can imagine having a high-frequency burst of action potential
where we might fire the presynaptic cell at 100 Hz for one second:
so, that's 100 action potentials
and each one separated by a 10-millisecond time scale.
And we can have a baseline, where we stimulate, perhaps,
once every 10 seconds, or once every 20 seconds.
That gives us our baseline postsynaptic potential.
We can plot the amplitude of that,
in the form of these dots here, across time.
So we stimulate, perhaps, every 10 seconds:
we see some variation in the postsynaptic potential;
then, we give our high-frequency tetanic stimulation, the 100 Hz.
That causes a massive influx of calcium
through the voltage-gated calcium channels.
That rise in calcium activates an enzyme called protein kinase C
that depends upon calcium to activate it:
so, there's a large calcium influx
during this one second of high-frequency stimulation
that activates protein kinase C;
and one of the targets of protein kinase C
is part of the molecular machinery involved in neurotransmitter release.
It's a protein called Munc-18;
and if that becomes phosphorylated by protein kinase C,
then the release rate at any given calcium concentration is increased.
And so, if our action potential
induces, say, 3 micromolar calcium in the presynaptic cell
under baseline conditions,
after the post-tetanic potentiation -- this 100 Hz stimulus;
after the activation of protein kinase C in the phosphorylation of Munc-18,
three micromole of calcium will now induce a much higher release rate;
and that's, then, what we see here.
After the post-tetanic potentiation stimulation,
we go back to our stimulation of action potentials every 10 seconds or so,
and we now see that the release of neurotransmitter is enhanced;
and it can be enhanced for a time of many seconds.
So, here we're in seconds, so this is one minute.
And, after one minute, there's still a substantial enhancement
in the release from this synapse, compared to the baseline.
So, a brief period of high-frequency activity
can induce a long-lasting change here over a period of about a minute,
where there's a substantial increase in neurotransmitter release.
And that can, of course, affect brain function
in important and interesting ways,
where one can imagine holding a short-term memory of activity
at the synapse for a period of a minute.
We can also have much longer terms of presynaptic plasticity,
in so-called long-term potentiation.
Again, this only occurs at <i>some</i> types of synapses in the brain.
Again, we have a high-frequency train of stimuli in the presynaptic cell
that induce the long-term potentiation;
we have baseline PSPs that come in;
we stimulate every now and then, and we see
that there's a relatively-reliable postsynaptic potential that comes in.
We give our high-frequency tetanization, the amplitude of the PSP,
that, we recall, postsynaptically increases
because of an enhanced release of neurotransmitter.
And molecularly, what's thought to be happening
at synapses that express long-term potentiation
is that the calcium influx during the high-frequency train
activates an enzyme called adenylate cyclase.
There are calcium-dependent forms of adenylate cyclases,
and so the calcium that floods in
during the presynaptic train of action potentials
activates adenylate cyclase;
that, in turn, turns ATP into cyclic AMP.
That's the job of adenylate cyclase.
Cyclic AMP is an important signaling molecule
that activates another protein called protein kinase A;
and that phosphorylates one of our favorite proteins here
in the release machinery: a Synaptotagmin.
Not the standard Synaptotagmin
that's responsible for calcium-dependent release,
but another one, another isoform, Synaptotagmin-12,
at least in some cases where it's been studied.
And the phosphorylation of Synaptotagmin-12
causes this long-lasting enhancement of neurotransmitter release,
known as presynaptic long-term potentiation.
And this is something that can take place
on many minutes', on the hour time scale.
So one hour, two hours later, there's still a large increase
in the efficacy of synaptic transmission at the synapse,
and that can maybe go on for days.
So the amount of neurotransmitter that's being released at a synapse
depends on a large number of things;
and one of them is the recent history,
the activity of that individual bouton in its past.
If it just received another action potential before,
then the calcium concentration will rise to higher levels
with the next action potential,
and that can induce presynaptic facilitation:
something that occurs on the 10- to 100-millisecond time scale.
Equally, the release of a lot of vesicles from the presynaptic terminal
leads to a process of depletion.
The vesicles need to dock and refill with neurotransmitter
before they can be released again;
and so there's also a process where we need to refill and reload vesicles.
And again, that process of replenishment
of the vesicles that are depleted by a previous action potential
can take place on the time scale of 10 to 100 milliseconds.
So there are these short-term facilitation and depression dynamics
that affect nearly all synapses in the brain.
Other synapses are able to be modified on longer time scales
through signaling processes involving kinases
that are calcium-sensitive, for the most part.
And so, they can sense the activity of the presynaptic terminal;
and, through phosphorylating different enzymes
or changing the amount -- availability -- of vesicles,
can then change the efficacy of neurotransmitter release.
And some of those modifications can last for minutes or hours.
In order to begin thinking
about how these modulations of neurotransmitter release
can occur in the brain and at different synapses
-- and, in particular, when we start thinking
about how <i>many</i> neurons interact with each other --
it's useful to come up with some computational models,
some simple models that, nonetheless, provide reasonable descriptions
about the types of changes
that occur in synaptic transmission on different time scales.
And so one useful way of thinking about it
is by bringing in exponential time courses
that will help model different events.
And so, if we imagine that there's a readily-releasable pool of vesicles,
and they, then, need to be replenished,
and that will take some sort of time course, here,
that we can give an exponential recovery rate
to a readily-releasable pool:
So, we have a given number of resources
and when an action potential is fired, some of those resources are used up;
there's an exocytosis, a release of vesicles;
and then those vesicles need to recover.
And that can be thought of as occurring
on some exponential recovery time constant.
Equally, we know that there's a process of facilitation,
so that when calcium floods into the presynaptic nerve terminal,
then the release probability in response to an action potential goes up.
And so, if we have an action potential
that then causes an increase in release probability,
which we can think of as an increase in the calcium concentration
that, again, lasts for a given amount of time.
So, we can think of a facilitation time constant.
And so, we have two different time constants.
We can also have, of course, different amplitudes for these effects;
and by putting, simply, two different,
exponentially-dependent time constants together
in some sort of releasable, pool fashion,
where action potentials then convert resources
into released neurotransmitter, and some of those get used up,
we can then generate very complex patterns of interactions, with...
we can imagine modeling spike trains with quite complicated dynamics.
And it turns out that simple models, with just a few variables in them,
can actually recover quite a lot of the dynamics
that occur at real synapses in the brain.
And so, there are simplifications
to the complexity that we see at the molecular level.
We can have empirical, much simpler models
that do a relatively good job of modeling synaptic transmission,
at least on short time scales.
And of course, in future work, more detailed models need to be generated
in order to fully understand the synaptic process of transmission.
And so, we've learned that not all action potentials are equal.
An action potential will certainly cause calcium influx
into the presynaptic bouton, and neurotransmitter release;
but the amount of neurotransmitter release and the amount of calcium influx
depends upon the recent history of action potential firing in that axon.
Action potentials occurring
within some tens or hundreds of milliseconds of each other
can induce facilitation through summation
of calcium concentrations in the presynaptic terminal,
and they can induce depression through the depletion of synaptic vesicles
and a reduction in the readily-releasable pool.
On the longer time scales,
the calcium influx that occurs, particularly during high-frequency firing,
can induce changes in calcium-activated signaling cascades.
Calcium activation of protein kinase C can cause post-tetanic potentiation
lasting for a minute or so, whereas calcium activation
of adenylate cyclase and cyclic AMP signaling pathways
can cause long-lasting forms of potentiation
at the presynaptic nerve terminal
that probably contributes to the formation of memories.
As a result, changes in the efficacy
of presynaptic neurotransmitter release are prominent,
and form an interesting way in which the brain can auto-regulate
its amount of neurotransmitter release from its synaptic terminals.