We've now reached a turning point in our studies
of the cellular mechanisms of brain function.
The purpose of brain function is to govern behavior.
We must now see to what extent the biophysical mechanisms
that we've studied over the last five weeks are able to account for behavior.
Let's begin by briefly reviewing what we've learned.
The brain is made of cells, each delimited by a phospholipid bilayer.
The cell membrane is impermeable to ions
and so we can have charge separation across the plasma membrane.
The plasma membrane is about five nanometers thick
and so there can be strong electric fields across this plasma membrane.
That gives rise to membrane potential and the electrical signals
that we know underlie brain function.
Charge movement across the plasma membrane
is mediated by ion channel proteins that are inserted into the plasma membrane
and they can have highly specific conductances
for one species of ion, for example.
These ion channels are not always open, but rather, they are tightly regulated,
and they have a certain open probability.
Sometimes they're closed — no current flow —
and then they open and close again.
That open probability of these ion channels
determines the net flux of ions
and therefore the change in potential across the plasma membrane.
The tight regulation of ion channels underlies the ability
of electrical signals to occur across the plasma membrane.
In particular, we considered a class of voltage-gated ion channels
that's particularly useful
in terms of making the brief action potential signals
that are an important way of coding information in the brain.
The voltage-gated sodium channel increases its open probability
steeply as the membrane potential depolarizes more positive
to around -40 millivolts.
As the membrane potential depolarizes, the open probability
of the voltage-gated sodium channel increases — that in turn causes
more depolarization and that in turn then increases the open probability
of the voltage-gated sodium channel,
generating the explosive depolarization
of the cell towards the sodium reversal potential.
The brief opening of the sodium channel, its fast inactivation,
and the delayed activation of the potassium conductance
drives repolarization and we're left with a brief all-or-none unitary signal,
the action potential, lasting about one millisecond.
Nerve cells don't operate on their own; they communicate with each other.
Through the action potential that propagates down the axon,
invades specializations, boutons, there's a release
of a neurotransmitter substance
that binds onto ligand-gated ion channels in the postsynaptic neuron,
in turn causing a change in the membrane potential
of that downstream neuron.
So the action potential in one cell, lasting about one millisecond,
conveys a longer-lasting signal, sub-threshold depolarization
or hyperpolarization lasting some tens of milliseconds
interconnected downstream neurons.
The process of synaptic transmission is highly plastic.
It depends on the previous history of activity of that neuron
and also depends upon the overall experience
of the animal.
So there's that interesting synaptic plasticity
that occurs, and that in turn appears to be the mechanism
underlying learning and memory.
So the individual neurons communicate with each other
and each neuron receives inputs from many hundreds of neurons
and in turn also delivers messages to hundreds of neurons,
some of which are widely distributed across the neuronal circuits of the brain.
We thus need to begin to divide the neurons and the types of neurons.
We know that there are excitatory glutamatergic neurons
that release glutamate and try to excite their postsynaptic targets.
There are inhibitory GABAergic neurons that release GABA
and try to inhibit their targets.
There are neuromodulatory neurons
that release neuromodulators that affect the dynamics and interactions
of the system as well as the plasticity of these neuronal circuits,
and we need to deal then with a great deal of complexity
as to where and how these neurons communicate in highly specific patterns
and how they do so during behavior.
So if our goal is to obtain a causal understanding of behavior,
we need to investigate brain function at the level of individual neurons
and their synaptic interactions — that's how we're going to get
a causal, mechanistic understanding — and we need to do that
within the complex neuronal networks
of the living mammalian brain during behavior.
We need to see the action of the brain during action.
So how are we going to do that?
The first thing we need to be able to do is to measure neuronal activity
in the living brain and to correlate it directly with ongoing behavior.
At the level of cells and synapses, there are two methods
that so far have proved to be extremely useful.
One is the electrophysiological approach,
where we introduce recording electrodes into the brain.
We've already seen the whole-cell patch-clamp recording
of membrane potential and currents
that has been applied <i>in vitro</i> in brain slices.
It turns out that this method can also be applied
with small modifications to study the brain in action —
<i>in vivo</i>, during behavior.
It also turns out to be possible to make extracellular recordings
of action potential firing.
To date we've thought of the extracellular solution
as being at zero millivolts isopotential and invariant.
In fact, that turns out not to be completely true,
and there are small changes in the extracellular potentials
that can be recorded,
and they can be recorded relatively easily.
The inference of which cell is firing action potentials
can also be made.
The extracellular recordings are technically easier
than the intracellular recordings, and so one can in parallel record
from many neurons at the same time and study network function
using extracellular recordings.
Electrophysiological measurements are certainly proving useful
in correlating neuronal activity with behavior.
However, in the electrophysiology, we need to introduce
recording electrodes into the brain.
These recording electrodes must have some size,
and that in turn perturbs the brain.
It's interesting to think about other techniques
that might be less invasive, and introducing photons
into the brain is one of the most interesting techniques
that is being developed at the moment
for studying brain function at high resolution.
So optical imaging techniques, and in particular fluorescence microscopy,
is turning out to be extremely interesting
in terms of getting high-resolution structural and functional measurements
of the brain, and at this time this is possible to do
in the living brain and correlate it with behavior.
In later videos of this week, we'll study how we can apply
electrophysiological measurements in video 6.3 and how we can apply
optical methods for investigating
neuronal activity in video 6.4.
So we're beginning to have methods
for measuring and correlating neuronal activity with behavior.
That's absolutely essential as a starting point.
But correlations don't necessarily imply causality,
and so if we want to investigate the causal impact
of a specific neuronal activity,
we need to specifically perturb that activity.
We need to be able to take control of the neurons and the neuronal circuits
and see what effect that has upon behavior.
That's turned out to be an extremely difficult problem
to have highly specific interventions in the brain.
So typically, one has relied upon lesions,
brain damages, diseases, stimulation or pharmacological manipulations,
and all of these lack specificity in space and time.
So it's been difficult to get a real detailed information
about the causal impact of specific types of neuronal activity.
Over the last years, there's been a revolution in neuroscience
that's been driven by <i>optogenetics</i>,
a way in which we can control neuronal activity by light.
This occurs through expressing specific light-sensitive proteins
that can then affect the activity of specific nerve cells,
we can use genetics to put these optogentic actuators
in specific cell types, and we can then shine light
on specific neurons or parts of neurons gaining high spatiotemporal resolution.
So there's a great deal of hope now that we'll be able to make
highly specific perturbations during behavior,
and then we'll be able to study the impact of that.
So if we're able to make measurements and perturbations
of highly specific neuronal activities,
the next step is to try and see
how those measurements and perturbations fit together.
In order to do that, we need to quantitatively model the phenomenon
and that then will give us causal mechanistic insight
and allow us to test specific hypotheses.
Unfortunately, we currently have a very incomplete understanding
of neuronal circuits in the mammalian brain.
We're very far from having a quantitative mechanistic understanding
of even the very simplest behaviors.
So the starting point is far from where we want to be.
Of course, the brain is highly interconnected,
so it's an extremely complex question
as to how we can quantitatively model the neuronal networks of the brain.
There's a simplification that we might be able to make.
Although many brain regions are highly interconnected,
there's also a high degree of modularity.
So there are so-called small world networks where neurons near to each other
might be highly interconnected with each other but they will only have
sparse connections to another part of the brain.
So we can think about local subnetworks and it may be that we can get
detailed cellular and synaptic modeling of these selected subnetworks
and model that,
and that is now becoming feasible with the increases in computing power.
So there's some hope that over the next decades —
perhaps in our lifetimes — we might be able to get
quantitative, causal, mechanistic understanding of at least very simple forms
of brain function and their link to behavior.
As a cautionary note, I think it's important to consider the limitations
of the scientific method.
These were clearly stated by Karl Popper in his famous treatise
of 1934, <i>The logic of scientific discovery.</i>
He pointed out that we can falsify hypotheses,
but we cannot prove them, and so scientific progress
is at best a continual refinement of our ideas as we make
more and more detailed measurements and perturbations.
Here we've begun to think about how we might put together
our biophysical knowledge of how the components of the brain work
in the context of behavior, and most of the videos of this week
will deal with methods as to how we might approach the problem
of linking the biophysics of brain function
with animal behavior.
We've come across three important steps that we need to make.
First we need to measure brain function,
and we need to do so at the level of cells and synapses
and how the membrane potential changes drives further action potential activity
in those neurons that we've been recording from.
So we need to make detailed measurements of brain function at the level
of cells and synapses.
These measurements will allow us
to correlate neuronal activity with behavior.
But in order to reach causality, we need to perturb brain function
and see what types of effects that causes.
We need to take control of the brain and see what happens
if we stimulate and inhibit specific neurons within the brain
in a highly controlled spatiotemporal pattern.
If we can take control of these neurons and measure the effects of the brain
and at the same time see the behavioral consequences,
then we're getting to a position where it becomes extremely interesting
to try to quantitatively model and see to what extent
we've really got the right variables
and that we've made the correct measurements
and that we really have a causal and mechanistic understanding
of brain function, including its link to behavior.
So in this week's videos, we're going to take a look
at the methodology as to how one might obtain a causal
and mechanistic insight into brain function during behavior.