Neurotransmitters and Synaptic Transmission

Synapses, specialized junctions at which a neurotransmitter is released from the presynaptic cell to signal to the postsynaptic cell. An action potential in the presynaptic neuron travels down its axon and into the presynaptic bouton, which releases neurotransmitter. That chemical binds to receptors on the postsynaptic cell and causes a change in the membrane potential of that cell.

Glutamate and gamma-aminobutyric acid (GABA). Glutamate is the main excitatory neurotransmitter in the brain: its release causes postsynaptic neurons to increase the probability of firing action potentials. GABA is the most important inhibitory neurotransmitter: a neuron releasing GABA tends to decrease the probability of the postsynaptic cell firing action potentials. These two make up the vast majority of synapses in the central nervous system.

The common belief has been: one. A given neuron releases one type, so you can have glutamatergic neurons, GABAergic, dopaminergic, etc. However, it has been discovered that co-release of more than one neurotransmitter from the same neuron can also occur in many neurons. I think for this class we just consider the simplified picture of one NT per neuron.

Acetylcholine. It is the neurotransmitter that acts at the mammalian neuromuscular junction: a motor neuron releases acetylcholine from its nerve ending, and it binds to acetylcholine receptors present on muscle fibres, forming the first step in initiation of muscle contraction. It also has an important role in regulating heart function.

Reward-based learning. Parkinson's disease involves a deficit in the regulation and release of dopamine, and loss of dopaminergic neurons. Dopaminergic dysfunction is also thought to be involved in schizophrenia. Dopamine is part of a family of neurotransmitters called the catecholamines (dopamine, NE and E).

Neurotransmitters made out of small peptide chains, which are slightly larger than the small molecules such as glutamate, GABA, acetylcholine and dopamine. There are many different neuropeptides that can function as neurotransmitters. They are only a small subset of the total number of neurotransmitters present, and act through slower biochemical signalling cascades rather than fast ligand-gated ion channels.

Met-enkephalin, composed of five different amino acids, and oxytocin, composed of nine different amino acids. These are only two of many.

1. An action potential travels down the presynaptic axon and invades the presynaptic bouton.
2. The depolarization activates voltage-gated calcium channels in the bouton, allowing an influx of calcium.
3. The elevation in calcium concentration causes the release of stored neurotransmitter from the bouton into the synaptic cleft.
4. The neurotransmitter diffuses across the synaptic cleft and binds to a ligand-gated ion channel in the postsynaptic membrane, increasing its open probability.
5. This causes an influx of ions that changes the membrane potential of the postsynaptic cell and produces a postsynaptic potential.

A protein and an ion channel very similar to the other types of ion channels, but its open probability is not gated by voltage; it is gated by the presence or absence of a neurotransmitter. Binding of the released neurotransmitter increases the open probability, causing an influx of ions that changes the postsynaptic membrane potential. They are the receptors at which fast chemical synaptic transmission occurs.

Voltage-gated calcium channels. They are voltage-gated ion channels similar to the potassium and sodium channels, except they conduct calcium ions, and they are activated by the depolarization of the action potential. Their increased open probability allows an influx of calcium into the presynaptic bouton, and that elevation in calcium concentration causes the release of stored neurotransmitter.

About 50 nanometers of extracellular space separating the presynaptic and postsynaptic membranes. The neurotransmitter released by exocytosis diffuses across this distance, something like 40 to 50 nanometers, and its diffusion is rapid. It then binds to a neurotransmitter receptor present in the postsynaptic membrane.

Synapses are typically on dendrites far from the somatic area of the neuron, so the postsynaptic current is filtered by the cable properties of the dendrites before reaching the soma and the axon initial segment, the trigger point for an action potential. That filtering is spatial, reducing amplitude, and temporal, slowing and broadening the time course. So a single synapse alone cannot depolarize the cell beyond threshold. Thus, a typical cell has many hundreds or thousands of synapses, and it is their summation at the axon initial segment that determines whether it fires.

Many hundreds or thousands of synapses, some excitatory and some inhibitory. The potentials are summated at the level of the axon initial segment, the location where action potentials are triggered.

The postsynaptic potential typically occurs with a latency of within 1 millisecond of the action potential in the presynaptic terminal. There is a short delay for getting calcium into the bouton, and the release process also takes a short delay. Synaptic transmission is therefore a very fast process, and neurons can communicate with each other on the millisecond timescale.

A graded change in the membrane potential of the postsynaptic cell, caused by ion flux through ligand-gated ion channels opened by neurotransmitter binding. An excitatory postsynaptic potential (EPSP) is a depolarization, taking the cell towards action potential threshold and increasing firing. An inhibitory postsynaptic potential (IPSP) is a hyperpolarization, taking the membrane potential more negative than threshold and reducing action potential output.

Excitatory synapses, which typically release glutamate onto ionotropic glutamate receptors, and inhibitory synapses, which release GABA onto ionotropic GABA_A receptors. Glutamate receptors are permeable to sodium and potassium with a reversal potential around 0 mV, producing depolarizing EPSPs. GABA_A receptors conduct chloride with a reversal potential around −70 mV, producing hyperpolarizing IPSPs.

They are permeable to both sodium and potassium, being a little bit more permeable to sodium, and their reversal potential is around 0 mV, which is more depolarized than the action potential threshold (−50 mV). So the excitatory postsynaptic potentials (EPSPs) made in the postsynaptic neuron tend to depolarize the cell beyond action potential threshold and increase action potential firing.

GABA_A receptors conduct chloride, which has a reversal potential of around −70 mV (more hyperpolarized than action potential threshold). Chloride is present at low concentrations intracellularly and high concentrations extracellularly, so it enters the cell carrying negative charge with it and causes hyperpolarization of the membrane. The resulting inhibitory postsynaptic potentials (IPSPs) take the membrane potential more negative than action potential threshold, causing a reduction in action potential output.

The diameter of the total synaptic contact area (the "length scale") is typically smaller than 1 micron, usually between 200-500 nm, but in class he cited 100nm to 1 micron as the full range. The presynaptic and postsynaptic membranes are separated by only 20-50 nm (this is the distance of the synaptic cleft). Specialized proteins, which are electron dense and therefore show up dark on EM, link the pre- and postsynaptic specializations and hold them tightly apposed to each other.

A small sphere of lipid bilayer membrane, around 40 nanometers in diameter, filled with a high concentration of neurotransmitter. Its membrane has a similar composition to the plasma membrane of the cell. Vesicles in the presynaptic bouton are the store of neurotransmitter that is released into the synaptic cleft.

The synaptic vesicles in the presynaptic bouton fuse their membrane with the presynaptic membrane. That fusion of the membrane causes exocytosis, and the neurotransmitter inside the synaptic vesicle is released and diffuses into the synaptic cleft. It then binds to its ligand-gated ion channel and causes a conductance.

Neurotransmitter is released in discrete packets rather than continuously, and the packet is the contents of a single synaptic vesicle (typically several thousand molecules). Exocytosis of a vesicle is all-or-none: each vesicle either releases its whole contents or it does not, and whether it does is probabilistic. So the postsynaptic response is made up of whole quanta, with no fractions of a quantum, and larger responses reflect more vesicles being released, either at one synapse or across several. The amplitude of the evoked postsynaptic potential is therefore approximately a whole-number multiple of the amplitude of the unitary postsynaptic potential produced by a single vesicle, rather than taking intermediate values.

Sir Bernard Katz, in the first measurements of synaptic transmission at the frog neuromuscular junction in the 1950s, showing that neurotransmitter occurs in quantized packages. The idea had come from electron microscopy showing synaptic vesicles, each filled with a high concentration of neurotransmitter: if transmitter is packaged this way, an action potential might release a vesicle or might not, giving a quantized, all-or-none form of communication at individual synapses.

Recordings from frog motor neurons showed spontaneous unitary postsynaptic potentials occurring without stimulation, all of roughly the same small amplitude (~0.4 mV). Extracellular calcium was then lowered to reduce release probability, and the nerve was stimulated. The evoked responses were all-or-none: sometimes nothing at all (a failure), sometimes an event the same size as a spontaneous unitary potential, and sometimes a larger response that was a whole-number multiple of it. So each stimulus released 0, 1, 2 or more quanta from the presynaptic specialization, never a fraction of one.

The evoked potentials cluster into discrete peaks at whole-number multiples of the unitary amplitude (about 0.4, 0.8, 1.2, 1.6 mV), corresponding to release of one, two, three or four vesicles, with relatively few values in between. A separate bar at 0 mV counts the failures, trials where no vesicle was released, showing that release is probabilistic. A curve predicted by assuming the number of vesicles released follows a Poisson distribution fits the data closely, including the number of failures. The peaks broaden at higher amplitudes because variability in quantal size accumulates, which is why evoked amplitudes are only approximately multiples of the unitary potential.

The inside of a synaptic vesicle is acidic, because the V-ATPase pumps in protons to drive neurotransmitter uptake, while the extracellular space is at neutral pH. A pH-sensitive fluorescent protein (e.g. pHluorin, a variant of GFP) attached to the inside face of the vesicle membrane is therefore quenched while the vesicle is intact, and becomes brightly fluorescent the moment the fusion pore opens and the lumen is exposed to the extracellular pH. Each fusion event produces a discrete step increase in fluorescence, so quantal release can be counted optically as steps of light rather than electrically as steps of voltage. Higher fluorescence = more vesicles being released.

The action potential is reliable, but synaptic transmission has failures: on some trials it fails to cause exocytosis, no neurotransmitter is released, and there is no postsynaptic response. On successful trials the postsynaptic potentials are roughly the same size, though with small variations in amplitude, because each is the response to one quantum, i.e. the contents of a single synaptic vesicle.

Because there are different amounts of neurotransmitter present, or because the neurotransmitter reaches a little further and activates more postsynaptic receptors. What is quantal is the release of the vesicle, not the exact size of the response it produces.

By making multiple synapses onto that postsynaptic cell, with its axon branching to form many different synapses on the same target. Their postsynaptic potentials summate at the axon initial segment, so the stochastic variations that occur at each of the individual synapses average out, giving rise to a statistically average, well-defined quantal amount of neurotransmitter sensed postsynaptically. This is necessary because transmission at any single synapse is unreliable, failing on some trials and varying in amplitude on others.

Chemical synapses, the most important form of communication in the mammalian brain, transmit indirectly: an action potential triggers release of neurotransmitter, which diffuses across the cleft and binds receptors on the postsynaptic cell, causing a postsynaptic potential. Electrical synapses (rare) transmit directly: the two membranes are joined by gap junctions, special proteins present on both sides through which current flows straight from one cell into the other, with no transmitter involved.

Junctions formed by gap junctions, special proteins sitting in the plasma membrane on both sides of the contact, connecting axons or dendrites of different neurons. They link the cytoplasms of the two cells, so current can flow directly between one cell and another, without the signal being converted into a chemical one. They are rare in the mature mammalian brain, though present at some junctions of specific neuron types. During early development, however, many cells are electrically coupled.

Much smaller. If cell one depolarizes by, say, 10 mV, the coupled cell might depolarize by only 1 or 0.1 mV. This is because there is quite a lot of resistance at the gap junction coupling between the two cells, so there is a big drop in potential across electrical synapses.

Part of the released neurotransmitter leaks out of the synaptic cleft into the extracellular space, where it can bind to neurotransmitter receptors and produce effects at a distance from the actual one-to-one synapse. This is called volume transmission. It is slower and weaker than point-to-point transmission, since the concentration falls as the transmitter diffuses, and it typically operates through other, slower biochemical signaling cascades. It is nonetheless an important regulator of brain function: glutamate and GABA have both a fast one-to-one component that acts on postsynaptic receptors and a diffusive volume component that acts on different types of receptors than those present in the synapse.

Dopamine. At a dopaminergic synapse or presynaptic bouton, there is no postsynaptic specialization at all. The dopamine is simply released into the extracellular space, where it can bind to different dopamine receptors present on a variety of different membranes.

Signalling that travels backwards across the synapse, from the postsynaptic neuron to the presynaptic bouton and other nearby neuronal elements. Dendrites and postsynaptic compartments release neurotransmitters such as endocannabinoids and nitric oxide, which act on receptors there to regulate the forward direction of synaptic transmission. Communication at the synapse is therefore bidirectional, though the most important signal is the one going from the presynaptic bouton to the postsynaptic cell.

In part of the olfactory bulb, through a process called dendrodendritic communication, which differs from the typical axo-dendritic synapses. Some dendrites are equipped with synaptic vesicles that can be released and act on neurotransmitter receptors in the postsynaptic dendrite, which in turn causes calcium increases there. Vesicles clustered in that second dendrite can then be released as a retrograde signal back onto the first dendrite.

First, synaptic vesicles full of neurotransmitter dock against the presynaptic plasma membrane and are primed so that they are ready to release. Second, an action potential arrives at the presynaptic terminal (bouton) and causes the intracellular calcium concentration to rise. Third, that calcium rise makes the vesicle membrane fuse with the plasma membrane, so the neurotransmitter inside is released by exocytosis.

The presynaptic bouton has a high concentration of voltage-gated calcium channels. They are similar to voltage-gated sodium and potassium channels, except that they are specifically permeable to calcium rather than any other cation. They also have a similar voltage dependence: at hyperpolarized potentials the open probability is low, and as the membrane gets more positive than around −40 mV the calcium channels begin to open and let calcium flood into the cell.

About 1-2 mM of calcium in the extracellular space, and about 100 nM of free calcium ions inside the cytosol at resting conditions. This creates a strong gradient for calcium to rapidly enter when the open probability of the calcium channel increases at depolarized potentials.

It is smaller in amplitude than the sodium and potassium currents in the terminal, and it activates more slowly than the extremely fast voltage-gated sodium conductances, so there is a short delay before it turns on. The peak of the calcium current therefore falls on the falling phase of the action potential, and this peak is delayed a few hundred microseconds from the onset and peak of the action potential. The channels stay open for a few hundred microseconds in total, and the influx ends once repolarization takes the membrane back below their activation range.

Neurotransmitter release has an extremely steep
dependence on Ca2+ concentration, with vesicle release rates depending upon the fourth power of the cytosolic Ca2+ concentration in the presynaptic bouton. On a log-log plot of calcium concentration (x-axis: cytosolic Ca2+ in μM) against vesicle release rate (y-axis: # vesicles released per ms) the relationship is linear with a slope of around four (the slope is the Hill coefficient). This means a tenfold increase in calcium, e.g. from one to ten μM, gives a 10^4 (10,000-fold) increase in release rate, so from 0.1 to 1000 vesicles per ms. The dependence is therefore extremely steep, and small increases in calcium cause huge changes in release.

Very, very small, nearly negligible, since the resting intracellular Ca2+ concentration is only around 100 nM, but there is still a small amount of vesicles fusing spontaneously even at resting levels, but the release rate increases very rapidly in the presence of calcium.

Ca2+ concentration in the cytosol of the presynaptic bouton. Increases in calcium drive release irrespective of action potential or membrane potential, so it depends on the level of calcium, not necessarily whether or not there is an AP firing in the presynaptic neuron. Calcium drives vesicle fusion in an extremely steep dependence-->small increases in calcium cause huge changes in neurotransmitter release rates.

By calcium-sensitive binding proteins in the presynaptic bouton, such as synaptotagmin, which bind the calcium and translate it into fusion of the vesicle membrane with the plasma membrane. This is only possible because the SNARE complex has already brought the vesicle into close apposition with the plasma membrane, leaving it docked and primed. With the vesicle already release (ready and waiting only for calcium) exocytosis can follow within roughly 100 microseconds of the calcium influx.

1. Synaptobrevin, the vesicular SNARE (vSNARE)
2. Syntaxin, the target plasma membrane SNARE (tSNARE)
3. SNAP-25, which is closely associated to syntaxin.

Binding strongly to one another, these proteins gradually wind up and pull the vesicle membrane and the plasma membrane into extremely close apposition, possibly so tight that the two cytoplasmic-facing leaflets (the outer leaflet of the vesicle and the inner leaflet of the plasma membrane) are already partly joined. The vesicle is then docked and primed: held release-competent against the membrane before the action potential arrives, waiting only for calcium. ====
(note that the inner leaflet of the vesicle fusing with the outer leaflet of the plasma membrane occurs during fusion specifically because that creates a pore)

Docked means the vesicle is parked in immediate apposition to the presynaptic plasma membrane. Primed means it is release-competent: the SNARE complex is assembled and holding the two membranes tightly together, so nothing further is needed except a rise in calcium. A docked and primed vesicle waits in this state until an action potential invades the terminal.

Synaptotagmin, which is closely associated with the SNARE complex (though not part of the core complex itself) and binds calcium through its C2 domains, with five binding sites in total. When calcium binds, synaptotagmin changes conformation dramatically, affecting the way it interacts with the SNARE proteins and possibly also with the phospholipids in the immediate environment. This change triggers the vesicle and plasma membranes to fuse, opening a fusion pore that connects the inside of the vesicle to the extracellular space so the neurotransmitter can escape.

Because synaptotagmin binds multiple calcium ions through its C2 domains, and fusion may require all of those sites to be occupied. If several ions must bind before the conformational change occurs, the release rate becomes a steep power function of calcium concentration rather than a linear one. This is the likely molecular origin of the Hill coefficient of around four.

The inner leaflet of the synaptic vesicle membrane becomes continuous with the outer leaflet of the plasma membrane, creating a pore (the "fusion pore") where the contents of the synaptic vesicle are in continuum with the extracellular space. On a sub-millisecond timescale the pore then widens and the vesicle flattens progressively into the plasma membrane: what began as a deep invagination collapses outwards, expelling the neurotransmitter into the cleft as the membrane returns to flat.

Around 100 microseconds between the action potential invading the bouton and neurotransmitter release. This is possible because the vesicle is already docked and primed by the SNARE complex, with the membranes held tightly together in advance. All that remains once calcium enters is for synaptotagmin to sense it and trigger fusion, so the slow preparatory work is done before the action potential ever arrives.

1. Priming and docking: the SNARE complex holds the vesicle and plasma membranes in close apposition, leaving the vesicle release-ready.
2. Calcium influx: the action potential invades the bouton and depolarization opens voltage-gated calcium channels.
3. Fusion: calcium binds synaptotagmin, whose conformational change lets the vesicle fuse with the plasma membrane and release its neurotransmitter. The whole process occurs on the microsecond timescale.

Because fusing two lipid bilayers is energetically costly: the outer leaflet of the plasma membrane has to merge with the inner leaflet of the vesicle membrane, which does not happen spontaneously. At the same time, vesicles must be able to release on a sub-millisecond timescale after calcium arrives, so they have to be held almost ready to fuse and yet not fusing. The SNARE complex resolves this by winding up to hold the membranes in extremely close apposition, leaving synaptotagmin to trigger the final step when calcium binds.

Because calcium concentration is highest right at the mouth of the open channel and falls off steeply with distance from it. A vesicle docked close to a calcium channel is therefore exposed to a much higher local calcium concentration than one further away. Since release rate goes as the fourth power of calcium concentration, this difference in position can set very different release probabilities for individual vesicles at the same terminal.

It is refilled with neurotransmitter, and then docked and primed again. This whole loop of vesicle fusion, refilling, and re-docking is called the synaptic vesicle cycle (or synaptic vesicle recycling), a process which takes roughly tens to hundreds of ms.

Through an ATP-dependent process. The vacuolar proton pump (V-ATPase) uses ATP to pump protons (H+ ions) into the vesicle, so the inside becomes acidic and positively charged. The combination of the acidification and this electrical potential across the vesicle membrane gives energy to the vesicular neurotransmitter transporters (e.g. VGlut for glutamate), which bring neurotransmitter inside. Once refilled, the vesicle moves to a release site and interacts with the SNARE complex to be docked and primed again.

1. Kiss and run: the fusion pore opens only briefly, the neurotransmitter escapes, then the pore closes again. The vesicle detaches, moves a few nanometres away from the plasma membrane, refills with neurotransmitter and re-docks.
2. Kiss and stay: the fusion pore opens briefly and the neurotransmitter escapes, then the vesicle reseals. The vesicle docked and primed against the plasma membrane the whole time, so it refills without ever detaching.
3. Full fusion: the fusion pore widens until the vesicle membrane is gradually absorbed into the plasma membrane. The vesicle no longer exists as a separate structure, and a new one must be pulled back out of the membrane by clathrin-dependent endocytosis to re-enter the vesicle cycle (where clathrin proteins assemble into a cage on the cytoplasmic face of the membrane, bending a patch inwards into a coated pit, which is then pinched off as a coated vesicle; the coat is shed and the bare vesicle is refilled by its transporters and re-docks)

Kiss and run and kiss and stay, which occur on the roughly 50 to 100 ms timescale. Because the whole synaptic vesicle cycle can happen this rapidly, they help sustain fast rates of neurotransmitter release when necessary.

Around 10%, depending on the precise circumstances, since exocytosis is a stochastic process. This means synaptic transmission is unreliable. It also means the process can be highly regulated: the efficacy of synaptic transmission can be increased or decreased, for example by changing the amount of calcium influx into the terminal (=neurotransmitter release regulation).

1. The amount of calcium entering the terminal, changed either by increasing the duration of the action potential or by acting on the presynaptic calcium channels themselves.
2. The amount of neurotransmitter inside each vesicle, set by the proton pumps and vesicular transporters, and by vesicle size.
3. The release probability of a docked vesicle, set by which protein isoforms are present, by phosphorylation, and by how close the calcium channels sit to the vesicles.
4. The number of vesicles docked in the release-ready state.

Calcium influx occurs during the repolarization of the action potential and comes to an end once the membrane potential is hyperpolarized, so increasing the duration of the action potential increases the amount of calcium influx and thus the amount of neurotransmitter released. This is done by regulating potassium channels, which govern the repolarization phase of the AP. Reducing the potassium current delays repolarization, thereby broadening the action potential and increasing calcium influx into the presynaptic terminal.

By modifying the presynaptic calcium channels themselves, so that more calcium flows through each one. For example, raising their open probability increases influx directly. Or, shifting their voltage dependence (moving the open-probability sigmoid curve leftward along the voltage axis so the channels open at more hyperpolarized potentials) means more of them are open during the action potential. This is used heavily in the nervous system to regulate neurotransmitter release.

Around 10 μM, up from a resting concentration of about 100 nM. Because the release rate depends on the fourth power of the calcium concentration, this rise takes the terminal from a nearly negligible rate of vesicle release to a very high one, and small changes around this peak level cause very large changes in the amount of neurotransmitter released.

By changing the efficacy of the vacuolar proton pump (V-ATPase), which builds the acidic and electrical gradients that drive uptake, or by changing the efficacy of the vesicular neurotransmitter transporters themselves since they control neurotransmitter uptake. Both are used to some extent. Vesicle size also varies around the typical 40 nanometers, with some bigger and some smaller vesicles, and presumably a big vesicle has more neurotransmitter content than a small one.

By changing the composition of the release machinery at the docked vesicle: the SNARE proteins (synaptobrevin, syntaxin, SNAP-25), the calcium sensor synaptotagmin, and the nearby voltage-gated calcium channels.

How?
1. There are different isoforms of these proteins (for example, synaptotagmin is encoded by ~15 different genes), and these different isoforms are expressed differentially at different synapses.
2. Phosphorylation events can also change the structure and function of these proteins.
3. The physical arrangement matters too, e.g. whether the calcium channels sit close to the synaptic vesicles or far away makes a big difference to whether calcium drives vesicle release at high probability or a lower probability.

Each docked, release-ready vesicle has roughly a 10% chance of being released when an action potential invades (due to the stochastic nature of exocytosis), so docking more vesicles close to calcium channels raises the chance of release. The more vesicles in the release-ready state, the more likely release is to occur and the larger the amount released when an action potential invades the presynaptic terminal.

Presynaptic plasticity (also called presynaptic dynamics) refers the modulation of neurotransmitter release in response to previous activity of that same axon, thereby changing the strength of communication between neurons at a synapse. It can occur on three different time scales, each relating to a different mechanism of modulation in the presynaptic terminal.

The amount of NT released in response to an action potential depends on the history of activity in that particular axon: a preceding action potential leaves an imprint on the synapse that changes how much NT is released by subsequent APs. Vesicle release probability can be either enhanced or suppressed, depending on the pattern of firing.

Firing patterns are diverse. Some cells fire at very high rates almost continuously, for example at 100 Hz (1 AP every 10 ms) kept up throughout the animal's life. Others fire sparsely and selectively: silent for minutes, then a brief burst of a few action potentials within a few ms of each other, then silent again for many minutes. Synapses respond differently to these patterns, amplifying transmission for some and weakening it for others.

1. Short-term synaptic plasticity, on a timescale of less than 1 second (typically 10-100 ms), which has two forms: facilitation (subsequent APs increase NT release) and depression (subsequent APs induce less NT release).
2. Post-tetanic potentiation, on the minute timescale.
3. Presynaptic long-term potentiation, which can last hours, days or perhaps a lifetime, and is implicated in memory formation.

Short-term dynamics affect nearly all synapses in the brain, whereas the longer-lasting forms occur only at certain types of synapse.

=======
--> The two longer lasting forms of plasticity rely on complex calcium-activated signalling cascades that are triggered downstream of presynaptic action potential firing, whereas short-term facilitation and depression are immediately related to the calcium in the presynaptic terminal and the number of neurotransmitter vesicles available.

Post-tetanic potentiation and LTP rely on complex downstream signalling cascades that are triggered by calcium influx. These cascades phosphorylate the release machinery, increasing the release rate at any given calcium concentration.

Short-term plasticity is more immediately related to calcium and vesicle availability. With facilitation, residual intracellular calcium from the previous AP adds to the next influx, so the peak calcium is higher and more vesicles are released (a 20% rise in calcium roughly doubles release, since release goes as the fourth power of calcium). With depression, the previous AP emptied part of the readily-releasable pool of vesicles, so fewer are docked and primed ready to release.

An increase in neurotransmitter release caused by a preceding action potential in the same axon. If a second action potential arrives shortly after the first, for example 20 milliseconds later, it induces a larger amount of neurotransmitter release and therefore a larger postsynaptic potential. It is typically studied with action potentials 10, 50 or 100 ms apart and lasts less than a second.

An accumulation of calcium in the presynaptic terminal. The first action potential causes calcium influx, and the calcium concentration falls only gradually, so if another action potential is fired within about 10-100 ms, there may still be residual calcium present in the presynaptic terminal. Calcium buffers (proteins that bind free calcium and lower its concentration) may also be partly saturated by the first AP, so they cannot bind as much of the new influx. Starting from an already elevated baseline, and with less of the incoming calcium being buffered, the peak calcium concentration reached by the second action potential is higher than that reached by the first, and more neurotransmitter is released

--> and even a small rise in peak calcium makes a large difference, since vesicle release depends on the fourth power of the calcium concentration.

1. The activity of calcium pumps, which pump calcium out of the presynaptic terminal.
2. The diffusion of calcium away from its entry site at the voltage-gated calcium channels, spreading out into the rest of the bouton so that the local concentration near the vesicles falls.
3. The binding of calcium by various calcium buffers which bind free calcium and lower its intracellular concentration. (note: these proteins are distinct from effector proteins such as synaptotagmin, which binds calcium to trigger neurotransmitter release rather than to lower its concentration).
4. Uptake into intracellular organelles, in particular the endoplasmic reticulum and mitochondria, which sequester calcium from the cytosol.

It roughly doubles it. The release rate depends on the fourth power of the calcium concentration (the Hill coefficient of ~4 from the calcium dependence of release), so 1.2 to the power of four is approximately 2. This is why even small changes in the peak calcium reached in response to an action potential can have a large impact on the amount of neurotransmitter released.

A decrease in neurotransmitter release caused by a preceding action potential in the same axon. When action potentials are separated by a few tens of milliseconds, the second releases less neurotransmitter than the first, giving a smaller postsynaptic response. It arises because the first action potential is quite efficacious and uses up docked, release-ready vesicles faster than they can be replaced.

Depletion of the readily-releasable pool of vesicles. The first action potential releases a large amount of neurotransmitter, which reduces the number of docked, release-ready vesicles, and replenishing them takes time, about 10-100 ms (the vesicle has to undergo synaptic vesicle recycling). If a second action potential arrives before the pool has been refilled, fewer vesicles are available and the probability of release is reduced. For example, if two release-ready vesicles are present and the first action potential releases one of them, only one remains for the second action potential: the calcium influx is the same, but with each vesicle having only about a 10% chance of being released, fewer vesicles means a lower chance of any release at all.

Both processes operate at once after an action potential: residual calcium raises release probability while depletion of the readily-releasable pool lowers the number of vesicles available. What is observed is whichever effect dominates, and different synapses display facilitation and depression to varying degrees. Which one dominates at a given synapse is determined experimentally by firing two action potentials a few tens of milliseconds apart and recording the postsynaptic potentials: a larger second PSP indicates facilitation, a smaller one depression.

An increase in the probability of presynaptic neurotransmitter release lasting about a minute, found only at some types of synapse. After a brief period of high-frequency firing in the presynaptic cell, single action potentials release substantially more neurotransmitter than they did before the stimulation. The enhancement is measured once the high-frequency stimulation has stopped, and persists for around a minute because the calcium rise during the tetanus leaves the release machinery chemically modified (i.e. Munc-18 is phosphorylated by PKC--> see card after next)

By tetanic stimulation: a sustained high-frequency train of action potentials in the presynaptic axon, for example firing the presynaptic cell at 100 Hz for one second (100 action potentials each 10 ms apart). Baseline postsynaptic potentials are measured by stimulating only once every 10 or 20 seconds, and after the tetanus these are substantially larger for around a minute. So you stimulate once, measure PSP, then apply the tetanic stimulation, then stop the tetanic stimulation, then apply another single stimulation and measure the PSP, which is transiently larger than the baseline PSP.

The high-frequency burst causes a massive influx of calcium through the voltage-gated calcium channels, which activates the calcium-activated enzyme protein kinase C. One of the targets of protein kinase C is Munc-18, a protein that is part of the molecular machinery involved in neurotransmitter release. When Munc-18 is phosphorylated by PKC, the vesicle release rate at any given calcium concentration is increased.

So an action potential producing, e.g. 3 μM calcium under baseline conditions will drive much more vesicle release after the tetanus than it did before, even though the calcium influx itself is unchanged.

A long-lasting enhancement of neurotransmitter release at the presynaptic terminal, found only at some types of synapse. It is induced experimentally the same way as post-tetanic potentiation :by a high-frequency train of stimuli in the presynaptic cell, after which the amplitude of the postsynaptic potential increases because more neurotransmitter is released. It differs from post-tetanic potentiation in the signalling cascade involved, which makes it last far longer: many minutes to hours, with a large increase in synaptic efficacy still present one or two hours later, and possibly going on for days. It is implicated in the formation of memory.

The calcium influx during the high-frequency stimulation train activates a calcium-dependent form of adenylate cyclase (aka adenylyl cyclase), which transforms ATP into cyclic AMP. Cyclic AMP activates protein kinase A, which phosphorylates the release machinery protein synaptotagmin-12 (a different isoform from the standard synaptotagmin, which is the one responsible for calcium-dependent neurotransmitter release). Phosphorylation of synaptotagmin-12 causes the long-lasting (many hours) enhancement of neurotransmitter release.

With exponential time courses (be able to draw both).

Short-term depression is modelled as a readily-releasable pool of vesicles: an action potential uses some of them up through exocytosis, and the pool then refills with an exponential recovery time constant (left of image).

Short-term facilitation is modelled as a pool of residual calcium in the bouton: an action potential adds to it, raising vesicle release probability for a subsequent action potential, and the calcium then decays away with an exponential decay time constant (right of image).

Different synapses show depression and facilitation to varying degrees, so the two exponentials are given different weights and amplitudes depending on which synapse is being modelled, e.g. a strongly depressing synapse gets a larger depression weight.

Well, on short timescales. Running the two exponentials over a train of action potentials reproduces the way a real synapse releases more or less transmitter at each spike depending on the preceding ones, so a simple model with just a few variables recovers much of the behaviour of real synapses. These are empirical simplifications of the true molecular complexity, and while they do capture the dynamics relatively well on short timescales, more detailed models will be needed to fully understand synaptic transmission.

About 20%. Neurotransmitters can therefore diffuse relatively long distances and act on membranes far from the synapse that released them. Presynaptic boutons express receptors (most relevantly metabotropic receptors), so a neurotransmitter diffusing from one bouton can bind to receptors on a second bouton and affect neurotransmitter release from that second bouton. This is called presynaptic modulation.

Transmembrane proteins built from alpha helices, with seven parts crossing the plasma membrane (7-transmembrane receptors). Unlike ionotropic receptors (which are ligand-gated ion channels), they do not form an ion channel and do not directly change ionic conductances. Instead, when an agonist binds, the receptor changes conformation and acts on intracellular G-proteins, signalling through biochemical pathways. They are therefore also called G-protein coupled receptors (GPCRs).

The receptor changes conformation and acts upon a membrane-bound GTP-binding protein (G-protein) on the intracellular side. The G-protein binds GTP and becomes activated, then splits into two parts: an alpha part and a betaamma part. These two parts then diffuse along the inner face of the plasma membrane (i.e. in a membrane-bound fashion), and can then act directly on ion channels or they can act on other enzymes inside the cell which in turn activate downstream signalling cascades.

1. Phospholipase C, which makes inositol trisphosphate (IP3) and diacylglycerol. IP3 goes on to release calcium from intracellular calcium stores, while diacylglycerol activates protein kinase C.
2. Adenylate cyclase (aka adenylyl cyclase), which increases the level of cyclic AMP (which then as you recall activates PKA--the same cascade that, at synapses expressing presynaptic LTP, phosphorylates synaptotagmin-12).

These are only two of many such pathways: different types of G-proteins activate different enzymes, and the resulting second messengers affect not just ion channels, but also the structure and function of nerve cells and their gene expression.

Neurotransmitter binding to a ligand-gated ion channel can induce a conductance within a millisecond, whereas the process downstream of a metabotropic receptor requires many tens of milliseconds and usually more than 100 ms. The difference is the number of steps in the signalling process: the receptor must be activated, which activates a G-protein, which must diffuse, which may then activate secondary messenger cascades. The effects are also longer-lasting: a metabotropic response can persist for many seconds and in some cases hours, whereas ionotropic effects end once the channel closes.
====
(metabotropic signalling also often operates on longer length scales than ionotropic, because the neurotransmitter diffuses through the extracellular space to reach receptors far from its release site)

A ligand binding to a single G-protein coupled receptor can activate many G-proteins sequentially, and each activated G-protein can act upon many targets. This means one binding event can produce a large, amplified downstream effect.

1. The potassium channels that drive repolarization: inactivating potassium conductances broadens the action potential waveform (a longer, less steep repolarization phase), which increases calcium influx and thereby increases neurotransmitter release.
2. The voltage-gated calcium channels directly, altering how much calcium enters the presynaptic terminal.
3. The release machinery itself (i.e. the SNARE proteins, synaptotagmin, Munc-18), altering how readily a docked vesicle fuses in response to calcium.
4. The proteins governing neurotransmitter uptake, recycling, docking and replenishment of the readily-releasable pool of vesicles, altering how many filled vesicles are available to release.

=========
note: in the video he says the target for #1 is the AP waveform itself, but I've made it more precise since technically the target would be the potassium conductances not the AP- but the point is that the G protein is acting to change the AP waveform.

A reduction in the probability of neurotransmitter release caused by activation of metabotropic receptors on the presynaptic bouton. It is one important form of presynaptic modulation and occurs at almost all synapses in the brain. The receptors sit near the site where synaptic vesicles fuse, but not within it.

1. Inhibition of calcium influx: on activation of the receptor, the associated G-protein splits into its alpha and betaamma subunits and the betaamma subunit blocks the entry of calcium through voltage-gated calcium channels, inhibiting the calcium influx evoked by action potentials.

2. Inhibition of the release machinery: the G-protein signalling directly affects the proteins of the release machinery (e.g. SNARE proteins, Munc-18, synaptotagmin) to inhibit release of synaptic vesicles in response to a given calcium rise.

These mechanisms, reduction of calcium entry and reduction in vesicle release rates in response to calcium rises, can occur separately or together (hence independent)

An axon that releases a neuromodulator diffusely into the extracellular space rather than onto a specific postsynaptic partner. Very often there is no postsynaptic specialization at all. These axons travel long distances and innervate large parts of the brain. Acetylcholine, dopamine, norepinephrine, and serotonin (so-called "neuromodulators") all typically work this way. e.g., a single axon from a cholinergic neuron in the basal forebrain can innervate a large expanse of neocortical territory.

The neuromodulatory bouton belongs to a separate axon that does not form part of the synapse being modulated; it sits some distance away, typically around 5 μm from the presynaptic terminal. When the neuromodulatory axon fires, the neuromodulator (e.g. DA, 5-HT, NE, ACh) is released and diffuses through the extracellular space to bind metabotropic receptors on that presynaptic bouton of the synapse. The G-proteins then either inhibit Ca2+ influx or inhibit the release machinery, so less neurotransmitter is released in response to the presynaptic action potential and the postsynaptic potential is smaller than when no neuromodulator was present.

Without the neuromodulator, the first action potential releases a lot of neurotransmitter and empties much of the readily-releasable pool of docked, primed vesicles. Fewer vesicles remain for the second action potential, so it releases less neurotransmitter and produces a smaller postsynaptic potential. With the neuromodulator present, vesicle release probability is reduced, so the first action potential releases only a few vesicles and leaves the pool largely intact. When the second AP occurs, there are therefore plenty of release-ready vesicles still available, and the residual calcium from the first also raises release probability, so the second AP releases more neurotransmitter than the first AP did. So presynaptic inhibition can switch a synapse from depression to facilitation.

Many of them do. Glutamate and GABA are especially important since the majority of the synapses in the brain are glutamatergic and GABAergic, so the ECF in the brain is thought to have a high concentration of glutamate and GABA escaping from their synaptic release sites. Many other neurotransmitters act as neuromodulators including ACh, DA, 5-HT, NE, etc.

1. Fast synaptic transmission: they act directly on ionotropic receptors in point-to-point transmission across the 50 nm cleft, producing postsynaptic potentials on the millisecond timescale.
2. Neuromodulation (slower): they also diffuse out of the cleft into the extrasynaptic space and travel long distances, reaching relatively high concentrations there. They bind metabotropic glutamate and metabotropic GABA receptors, driving slow neuromodulatory events including presynaptic inhibition, on the hundreds of ms to second timescale.

The existence of metabotropic glutamate and metabotropic GABA receptors, because they give the GABA and glutamate that escapes the synaptic cleft something to act on. These receptors are just as important for modulating brain function as many of the other metabotropic receptors, and form an important complement to the ionotropic glutamate and GABA transmission that work on the millisecond timescale for fast synaptic transmission.

Their axons travel long distances and innervate large parts of the brain, e.g. a single axon from a cholinergic neuron in the basal forebrain can innervate a large expanse of neocortical territory. Very often, for these neuromodulatory neurotransmitters, there is no postsynaptic specialization: the boutons diffusely release neurotransmitter into the extracellular space, where it diffuses and can bind different receptors on various targets. This is called volume transmission. It lacks the point-to-point specificity of direct synaptic transmission and also acts on longer length and time scales.

The neuromodulator is released with some temporal and spatial specificity, but much less than for fast synaptic transmitters, so the specificity of its action comes mainly from:

1. Which cells express the receptors for that neuromodulator.
2. Where those receptors are located on the cell: i.e. on axons or dendrites, on presynaptic or postsynaptic specializations. In some cases, an axon that contacts several types of target cell expresses the receptors only on the boutons that contact one of those target cell types, so the neuromodulator affects release at those boutons while the same axon's other boutons are unaffected.
3. The specific receptor isoforms and G-proteins involved.

Metabotropic receptors are also expressed postsynaptically, on the postsynaptic specialization and along the dendrite leading to the soma, where they affect several cellular processes. Neuromodulation therefore has both presynaptic and postsynaptic effects (presynaptic and postsynaptic modulation), and both contribute equally importantly to the neuromodulation of brain function.

1. GPCRs on the postsynaptic membrane block voltage-gated calcium channels on the postsynaptic dendrites via G-proteins, decreasing the excitability of the dendrite and forming a type of postsynaptic inhibition because there is less inward depolarizing current.

2. They also strongly regulate potassium channels, in particular they activate G-protein-coupled inwardly-rectifying potassium channels (GIRK channels, also known as Kir channels). With these channels open, K+ leaves the cell down its electrochemical gradient, hyperpolarizing the membrane and producing a slow inhibitory post synaptic potential.

A channel that conducts ions inward much more readily than outward. For a potassium channel (e.g. GIRK channels), inward flow at potentials below E_K passes easily, while outward flow at potentials above E_K is partly blocked. The name describes the conductance, not the direction ions actually move: at a resting potential of around −70 mV, above E_K, the electrochemical gradient still drives potassium out of the cell, so opening these channels hyperpolarizes the membrane, it is just harder for K+ to leave the cell through these channels than other K+ channels.

Inhibitory postsynaptic potentials that are produced when a neuromodulator activates postsynaptic metabotropic receptors, which, through G-protein action, open GIRK potassium channels that hyperpolarize the membrane. They take place on the hundreds-of-ms timescale, much slower than the 10-20 ms postsynaptic potential produced via ionotropic receptors, because the signal must pass through several steps (receptor activation, G-protein activation and diffusion, then channel opening) rather than the transmitter opening the channel directly.

E.g. a GABAergic input can produce both a fast IPSP via ionotropic GABA receptors and a slow one via this neuromodulatory G-protein pathway.

Because neuromodulatory receptors are expressed in highly specific manners, both in terms of which cells express which receptor subtypes and where in the nerve cell those receptors are located (see card about what determines the specificity of neuromodulator action).

For example, some cells express one type of dopamine receptor and others a different type, and the same is true for serotonin, adrenergic and oxytocin receptors. A drug binding a specific metabotropic receptor may therefore have a very specific action, since that receptor subtype might only be expressed in a specific cell type in a specific brain region. A systemically administered drug can thus act on one population of cells while leaving the rest of the brain largely untouched.

Neuromodulators also act slowly and without much spatial precision, which matches how a systemically administered drug acts, so there is a good fit between what is possible pharmacologically and how these systems work physiologically.

They are selective serotonin re-uptake blockers, used against depression and anxiety. Serotonergic axons release serotonin (5-HT) into the extracellular space with no postsynaptic specialization, and its concentration there is regulated largely by serotonin re-uptake proteins that take it back up into the axon. By blocking this re-uptake machinery, these drugs allow 5-HT to diffuse further and reach higher concentrations. So these SSRIs increase 5-HT in ECF.

It blocks specific subtypes of dopamine, serotonin and adrenergic receptors, and acts as an antipsychotic used to help treat patients with schizophrenia. It acts at many different neuromodulatory receptors, so its specificity comes partly from which receptors it leaves alone: it binds muscarinic acetylcholine receptors more than a thousand-fold more weakly than its targets, sparing the cholinergic system and the side effects that blocking it would cause.

A form of microscopy that uses a beam of electrons rather than light to image a sample. Because electrons have a far shorter wavelength than photons, it resolves much finer structures than a light microscope, down to the nanometre scale. In neuroscience this allows the detailed structure of fixed brain tissue to be imaged: axons, dendrites and synapses are a few hundred nanometres across, while synaptic vesicles, organelles and microtubules are smaller still.

1. Fixation with aldehydes (paraformaldehyde and glutaraldehyde), delivered rapidly by perfusion.
2. Cutting slices on a vibratome to reach the brain region of interest.
3. Staining with heavy metals.
4. Dehydration (removing water) through increasing concentrations of ethanol and embedding in resin polymerized at 60°C to get a hard block of plastic with tissue embedded in it.
5. Cutting the region of the brain out of the block using a dissecting microscope and then cutting ultrathin sections (~50-90 nm thick) with a diamond knife in an ultramicrotome. These sections are collected onto a grid for imaging.

The preservation of freshly killed material in a state that most closely resembles the living structure. Aldehydes are used: paraformaldehyde and glutaraldehyde. They cross-link (polymerize) proteins by forming methylene bridges between amine groups, linking proteins into long chains and huge complexes that trap lipids, carbohydrates and nucleic acids, rendering everything immobile. Glutaraldehyde gives stronger fixation because it has an aldehyde group at each end and so forms many more cross-links.

Biological structures are mostly made of low molecular weight atoms (carbon, hydrogen, oxygen, nitrogen), which interact only weakly with the electron beam. Heavy atoms interact much more strongly, so staining is what provides contrast.

Osmium tetroxide, which heavily stains membranes and proteins. Uranyl acetate, a very basic stain that reveals DNA and RNA. Lead citrate, which heavily stains glycoproteins such as starch and glycogen. They are applied in that order, with the lead staining often done after sectioning.

Because the sample goes into the high vacuum of the electron microscope, and must also withstand the intense heat of the beam. The section is dehydrated through increasing concentrations of ethanol, then once at 100% ethanol, it is infiltrated with a resin such as epoxy resin, which is polymerized in an oven at 60°C to give a hard block of plastic with the tissue embedded in it.

Because electrons are scattered by gas molecules so they can travel greater distances in a vacuum. In air the beam would be deflected and absorbed before reaching the sample or the detector, so no image could form. The column is therefore evacuated, which is also why the sample must be dehydrated since water would boil off in the vacuum and destroy the tissue structure.

It is very similar to a light microscope, but uses electrons instead of photons, which pass through the sample and are focused by electromagnetic lenses rather than glass ones. An electron gun produces the electrons, which pass through condenser lenses, then the objective lenses where the sample sits, then projection lenses that magnify the image onto a screen or CCD camera. Heavily stained structures such as membranes and mitochondria appear black, while lightly stained cytoplasm appears much lighter.

Because the heavy metal stains bind unevenly. Where heavy atoms are concentrated, they scatter electrons out of the beam so fewer reach the screen, and those regions appear black: membranes, mitochondria and proteins stained by osmium tetroxide. Lightly stained regions such as cytoplasm let most electrons through and appear much lighter.

A thick postsynaptic density and a thinner presynaptic density, separated by the synaptic cleft. This asymmetry is typical of an asymmetric or glutamatergic excitatory connection. Round, clear synaptic vesicles are visible in the presynaptic terminal, accumulating around the active zone.

Instead of passing electrons through a thin section, a fine beam is scanned across the sample by scanning coils, and electrons coming off the sample are captured by detectors to build up an image. Secondary electrons ejected from the surface give information about surface structure, while backscattered electrons reflected out of the sample give information from just below the surface. The number of backscattered electrons depends on the density and heavy atom content of the sample, so heavily stained regions give a high yield and unstained resin gives very few.

Secondary electrons are ejected from the surface of the sample when the incident beam hits it, along with other particles such as photons. They give information about surface structure, for example the general shape of a cultured neuron and its processes.

Backscattered electrons are electrons from the incident beam itself, reflected back out of the sample. They give information from just below the surface of the embedded material. Their yield depends on the nature of the sample: the denser it is and the higher the concentration of heavy atoms, the more are produced, so heavily stained regions appear bright while empty resin or lightly stained cytoplasm appears dark.

It alternates imaging and milling: the electron beam collects an image from the face of the block, then an ion beam mills away a layer a few nanometres thick, then another image is collected. This is automated and repeated many thousands of times to give a serial stack through the sample. With ~5 nm isotropic voxels, structures such as microtubules, synapses and individual vesicles can be resolved and reconstructed in 3D.
Recall the approximate values for:
-Synaptic cleft width, synapse size, vesicle diameter, synapses per postsynaptic cell
-AP duration in the bouton, synaptic delay, exocytosis timescale, Ca²⁺ current peak delay
-Reversal potentials of ionotropic glutamate and GABA_A receptors; voltage drop across an electrical synapse
-Extracellular, resting and peak intracellular Ca²⁺; voltage at which presynaptic Ca²⁺ channels open
-Hill coefficient; effect of a 10× and a 20% rise in Ca²⁺
-Synaptotagmin Ca²⁺ sites and gene number; release probability of a docked vesicle; kiss and run/stay timescale
-Short-term plasticity timescale; PTP protocol and duration; LTP duration; vesicle replenishment time
-Extracellular space fraction; neuromodulatory bouton distance; metabotropic vs ionotropic timescales; slow vs fast IPSP

Structure
Cleft ≈50 nm; synapse <1 µm, usually a few hundred nm; vesicle ≈40 nm
Hundreds to thousands of synapses per postsynaptic cell
Timing
AP in bouton ≈500 µs to 1 ms; PSP latency <1 ms; exocytosis ≈100 µs
Ca²⁺ current peaks on the falling phase, delayed a few hundred µs
Potentials
Ionotropic glutamate ≈0 mV; GABA_A (Cl⁻) ≈−70 mV; AP threshold ≈−50 mV
Electrical synapse: 10 mV in cell 1 → 0.1 to 1 mV in cell 2
Calcium
Extracellular ≈2 mM; resting ≈100 nM; peak during AP ≈10 µM; channels open above ≈−40 mV
Release ∝ [Ca²⁺]⁴ (Hill ≈ 4): 1→10 µM raises release from 0.1 to 1,000 vesicles/ms (10,000×); a 20% rise gives 1.2⁴ ≈ 2, doubling release
Release machinery
Synaptotagmin: 5 Ca²⁺ sites, ≈15 genes; docked vesicle release probability ≈10%
Kiss and run / kiss and stay: 50 to 100 ms
Plasticity
Short-term: <1 s, typically 10 to 100 ms (facilitation example: 20 ms apart)
Replenishment: 10 to 100 ms (video) or hundreds of ms (notes)
PTP: 100 Hz for 1 s (100 APs), lasts ≈1 min, baseline stim every 10 to 20 s
LTP: many minutes to hours, still present 1 to 2 h later, maybe days
Neuromodulation
Extracellular space ≈20%; neuromodulatory bouton ≈5 µm away
Ionotropic conductance <1 ms, PSP 10 to 20 ms; metabotropic tens of ms, usually >100 ms, effects seconds to hours, slow IPSP hundreds of ms
EM
Vibratome 50 to a few hundred µm; ultrathin sections ≈40 nm (range 40 to 200 nm)
Resin polymerized at 60°C for ≈24 h; FIBSEM voxels 5×5×5 nm