17.09.03 · mol-cell-bio / cellular-neuroscience

Synaptic transmission: neurotransmitter release, SNARE-dependent exocytosis, and receptor gating

stub3 tiersLean: nonepending prereqs

Anchor (Master): Sudhof, T. C. — Neuron 80 (2013) 675-690

Intuition Beginner

Neurons do not touch each other directly. Between the end of one neuron and the start of the next lies a tiny gap — the synaptic cleft — about 20 nanometres wide. When an action potential reaches the end of a neuron, it cannot jump this gap electrically. Instead, the electrical signal is converted into a chemical one: the neuron releases packets of signalling molecules called neurotransmitters into the cleft.

These molecules diffuse across the gap in microseconds and bind to specific receptor proteins on the surface of the next neuron. The receptors are ligand-gated ion channels: when a neurotransmitter binds, the channel opens, ions flow through, and the electrical state of the receiving neuron changes. If enough receptors open at once, the receiving neuron fires its own action potential, and the signal continues down the chain.

The whole process — electrical-to-chemical-to-electrical conversion — takes about one to five milliseconds. This is the synaptic delay. Every thought, movement, and sensation in your nervous system runs through millions of these chemical relays per second. Synapses are also where the nervous system does most of its computation: a single neuron receives thousands of synaptic inputs and integrates them to decide whether to fire.

Visual Beginner

Picture two neurons end-to-end with a narrow gap between them. The sending neuron (the presynaptic side) has a swollen terminal filled with tiny spherical packets called synaptic vesicles, each packed with thousands of neurotransmitter molecules. Voltage-gated calcium channels line the membrane at the terminal.

When an action potential arrives, calcium channels open. Calcium rushes in. The vesicles docked at the membrane fuse with it, dumping their neurotransmitter into the cleft. The molecules diffuse across the 20 nm gap and bind to receptors on the postsynaptic neuron, opening ion channels and changing the voltage.

Worked example Beginner

Consider a single synapse in the brain where the neurotransmitter is glutamate, the most common excitatory neurotransmitter.

Step 1. An action potential arrives at the presynaptic terminal, depolarising the membrane to about $+ 40$ mV.

Step 2. Voltage-gated calcium channels open. Calcium concentration outside the cell is about 1.5 mM, while inside it is roughly $1 0^{- 4}$ mM — a ten-thousand-fold gradient. Calcium rushes in through the open channels.

Step 3. The rise in intracellular calcium triggers synaptic vesicles to fuse with the presynaptic membrane. Each vesicle contains about 3,000–5,000 glutamate molecules. The fusion is mediated by proteins called SNAREs that act like a zipper, pulling the vesicle membrane and the cell membrane together until they merge.

Step 4. Glutamate molecules diffuse across the 20 nm cleft in about 50 microseconds and bind to AMPA receptors on the postsynaptic membrane. AMPA receptors are glutamate-gated sodium channels: when glutamate binds, sodium flows in, depolarising the postsynaptic cell.

Step 5. If enough AMPA channels open to depolarise the postsynaptic neuron past its threshold (about $- 55$ mV), an action potential fires in the receiving neuron. The signal has been successfully transmitted.

Check your understanding Beginner

Exercise (easy, multiple choice).

What triggers the release of neurotransmitter from the presynaptic terminal?

A. Sodium influx through voltage-gated sodium channels. B. Potassium efflux through voltage-gated potassium channels. C. Calcium influx through voltage-gated calcium channels. D. Direct electrical depolarisation of the postsynaptic membrane.

Hint

The key event is the entry of a specific ion that acts as an intracellular messenger, triggering vesicle fusion.

Answer

C. The arrival of an action potential depolarises the presynaptic terminal, opening voltage-gated calcium channels. Calcium rushes in down its steep electrochemical gradient, and this calcium influx triggers the SNARE-dependent fusion of synaptic vesicles with the presynaptic membrane, releasing neurotransmitter.

Exercise (medium, short answer).

Explain why synaptic transmission involves a chemical intermediate (neurotransmitter) rather than direct electrical coupling between neurons. What advantage does chemical transmission provide?

Hint

Think about signal amplification, integration, and the ability to produce both excitatory and inhibitory effects.

Answer

Chemical synaptic transmission provides several advantages over direct electrical coupling. First, it allows signal amplification: a single presynaptic action potential triggers the release of thousands of neurotransmitter molecules, which can open hundreds of postsynaptic channels. Second, it enables signal integration: a postsynaptic neuron receives inputs from thousands of synapses and sums them over time and space to decide whether to fire. Third, it permits both excitation and inhibition: different neurotransmitters can either depolarise (excite) or hyperpolarise (inhibit) the postsynaptic cell. Fourth, the strength of chemical synapses can be modified — strengthened or weakened — providing the basis for learning and memory (synaptic plasticity). Direct electrical synapses (gap junctions) do exist but are limited to bidirectional, unmodifiable signal pass-through.

Formal definition Intermediate+

Synaptic transmission is the process by which an electrical signal in one neuron is converted to a chemical signal (neurotransmitter release), transmitted across the synaptic cleft, and reconverted to an electrical signal in the postsynaptic neuron through ligand-gated ion channel activation. The process decomposes into four stages: (1) presynaptic depolarisation and calcium entry, (2) vesicle priming, docking, and calcium-triggered fusion, (3) neurotransmitter diffusion and clearance, and (4) postsynaptic receptor activation.

Presynaptic calcium entry

When an action potential invades the presynaptic terminal, voltage-gated calcium channels open. The relevant channel types at most chemical synapses are P/Q-type (Cav2.1) and N-type (Cav2.2) calcium channels. The driving force on calcium is enormous: extracellular $[Ca^{2 +}] \approx 1.5$ mM, intracellular $[Ca^{2 +}] \approx 0.0001$ mM, giving a Nernst potential of about $+ 120$ mV. At a terminal depolarised to $0$ mV, the driving force $V_{m} - E_{Ca} \approx - 120$ mV, producing a large inward calcium current.

The calcium that enters does not distribute uniformly. It forms localised calcium microdomains near the mouth of each open channel, with concentrations reaching 10–100 $μ$ M within a few tens of nanometres of the channel pore, while the bulk cytoplasm remains at resting levels (~0.1 $μ$ M). Synaptic vesicles are positioned within 10–20 nm of these calcium channels, ensuring they experience the high local calcium concentration needed to trigger fusion. This tight spatial coupling is why the synaptic delay (the time from presynaptic depolarisation to postsynaptic response) can be as short as 100–500 $μ$ s.

The SNARE complex and vesicle fusion

The molecular machinery for vesicle fusion centres on the SNARE complex, a four-helix bundle formed by three proteins ^{[Sudhof 2013]}:

Synaptobrevin (VAMP) — anchored in the vesicle membrane (the v-SNARE)
Syntaxin — anchored in the plasma membrane (a t-SNARE)
SNAP-25 — also anchored in the plasma membrane (the second t-SNARE), contributing two helices to the bundle

Synaptobrevin contributes one helix, syntaxin contributes one, and SNAP-25 contributes two, making a four-helix bundle that zippers from the N-terminal end toward the C-terminal transmembrane anchors. As the bundle zippers closed, it pulls the vesicle membrane and the plasma membrane together, overcoming the energetic barrier to membrane fusion.

Before calcium arrival, the SNARE complex is partially assembled (the vesicle is primed). The calcium sensor synaptotagmin sits on the vesicle membrane, bound to the partially assembled SNARE complex and held in check by complexin. When calcium binds to synaptotagmin's C2 domains (which have a Hill coefficient of approximately 4, reflecting cooperative binding to multiple calcium ions), synaptotagmin undergoes a conformational change: it inserts into the presynaptic membrane, displaces complexin, and triggers the final zippering of the SNARE complex. This completes membrane fusion in less than 100 $μ$ s.

The quantal hypothesis

Bernard Katz and colleagues established that neurotransmitter release is quantised — it occurs in discrete packets corresponding to the fusion of individual synaptic vesicles ^{[Katz & Miledi 1965]}. At the neuromuscular junction, the postsynaptic response to nerve stimulation is always an integer multiple of the quantal size $q$ (the response to a single vesicle), with the number of quanta released following a Poisson distribution:

P (n) = \frac{m ^{n} e ^{- m}}{n !}

where $m$ is the quantal content (mean number of vesicles released per stimulus) and $n$ is the actual number released. The quantal content $m$ depends on calcium concentration, with a steep dependence described by a Hill-type relation:

m = m_{m a x} \cdot \frac{[ Ca ^{2 +} ] ^{n}}{K _{d}^{n} + [ Ca ^{2 +} ] ^{n}}

where the Hill coefficient $n \approx 3$ – $4$ , reflecting the cooperative calcium binding of synaptotagmin.

Postsynaptic receptor activation

Neurotransmitter molecules that reach the postsynaptic membrane bind to two classes of receptors:

Ionotropic receptors (ligand-gated ion channels) open an intrinsic ion pore upon neurotransmitter binding. Key examples:

AMPA receptors (glutamate-gated $Na^{+} / K^{+}$ channels) — fast excitatory transmission, rise time ~0.5 ms
NMDA receptors (glutamate-gated $Na^{+} / K^{+} / Ca^{2 +}$ channels) — slow excitatory transmission, voltage-dependent block by $Mg^{2 +}$ , permeable to calcium, critical for synaptic plasticity
GABA-A receptors (GABA-gated $Cl^{-}$ channels) — fast inhibitory transmission
Glycine receptors (glycine-gated $Cl^{-}$ channels) — fast inhibitory transmission in spinal cord and brainstem

Metabotropic receptors (G-protein-coupled receptors, GPCRs) activate intracellular signalling cascades rather than opening an ion channel directly. Examples include metabotropic glutamate receptors (mGluRs), GABA-B receptors, and neuromodulator receptors (dopamine, serotonin, acetylcholine muscarinic).

Excitatory postsynaptic potentials (EPSPs) are produced by cation influx (primarily $Na^{+}$ through AMPA receptors), depolarising the postsynaptic cell. A single synapse typically produces an EPSP of 0.1–1 mV.

Inhibitory postsynaptic potentials (IPSPs) are produced by anion influx (primarily $Cl^{-}$ through GABA-A or glycine receptors), hyperpolarising or shunting the postsynaptic cell.

Synaptic vesicle cycle

After fusion, synaptic vesicle membrane and proteins are recovered by endocytosis and recycled into new vesicles. The cycle proceeds: (1) vesicle filling with neurotransmitter (via vesicular transporters using the proton gradient), (2) trafficking to the active zone, (3) docking at the presynaptic membrane, (4) priming (partial SNARE assembly), (5) calcium-triggered fusion, (6) endocytic recovery. Two modes of endocytosis have been characterised: clathrin-mediated endocytosis (slow, ~10–30 s) and kiss-and-run (fast, ~1 s), in which the vesicle forms a transient fusion pore without full collapse into the plasma membrane.

Counterexamples to common slips

Neurotransmitter release is not triggered directly by depolarisation. Depolarisation opens calcium channels; calcium triggers release. Blocking calcium channels (e.g., with $ω$ -conotoxin for N-type channels) abolishes transmission even though the action potential depolarises the terminal normally.
A single vesicle does not contain one neurotransmitter molecule. Each vesicle contains thousands of molecules (3,000–10,000 depending on the synapse type), and the vesicle is the unit of release, not the individual molecule.
Inhibition is not the absence of excitation. Inhibitory synapses actively hyperpolarise or shunt the postsynaptic cell by opening chloride channels (GABA-A, glycine) or potassium channels (GABA-B). Removing excitatory input and providing active inhibition are mechanistically distinct.
The synaptic cleft is not a passive void. It contains extracellular matrix proteins, adhesion molecules (neuroligins, neurexins), and enzymes that degrade or回收 neurotransmitter (e.g., acetylcholinesterase at cholinergic synapses).

Key mechanism Intermediate+

Mechanism (Calcium-triggered SNARE-dependent vesicle exocytosis). The fusion of a synaptic vesicle with the presynaptic membrane proceeds through a well-defined sequence of molecular events, each with experimentally characterised kinetics.

Docking. The vesicle is delivered to the active zone and positioned adjacent to the presynaptic membrane (~5 nm separation). Munc13 and RIM proteins organise the active zone and tether vesicles near calcium channels.
Priming. The SNARE complex begins to zipper from its N-terminal end, bringing the vesicle closer to the membrane. Complexin binds the partially assembled SNARE complex, clamping it in a metastable "ready" state. The vesicle is now release-ready (belonging to the readily releasable pool, RRP).
Calcium sensing. An action potential opens P/Q-type or N-type calcium channels. Synaptotagmin (the calcium sensor) has two C2 domains (C2A and C2B) that bind calcium cooperatively with a Hill coefficient of approximately 4. At resting calcium (~0.1 $μ$ M), synaptotagmin is inactive. When local calcium reaches 10–50 $μ$ M in the microdomain near an open channel, synaptotagmin binds calcium.
Triggering. Calcium-bound synaptotagmin inserts its C2 domains into the presynaptic membrane, displaces complexin, and catalyses the final zippering of the SNARE complex. This pulls the vesicle and plasma membranes together, forming a fusion pore within ~50–100 $μ$ s of calcium entry.
Fusion and release. The fusion pore expands, and the vesicle contents (neurotransmitter molecules) are released into the synaptic cleft. Full collapse of the vesicle into the plasma membrane completes fusion. Alternatively, in kiss-and-run mode, the pore closes after partial release and the vesicle detaches.
Endocytosis and recycling. The vesicle membrane is recovered by clathrin-mediated endocytosis (or faster kiss-and-run retrieval), refilled with neurotransmitter by vesicular transporters, and returned to the releasable pool. The full cycle takes 30–60 s for clathrin-mediated recycling; kiss-and-run can recycle a vesicle in ~1 s.

The critical quantitative feature is the cooperative calcium dependence of release. The vesicle release probability $p_{rel}$ scales approximately as $[Ca^{2 +}]^{4}$ over the physiological range. This steep dependence ensures that release is tightly coupled to the coincidence of calcium channel opening and vesicle proximity, making synaptic transmission fast and precisely timed.

Botulinum toxin (Botox) and tetanus toxin are zinc proteases that cleave specific SNARE proteins, blocking vesicle fusion. Botulinum toxin serotypes cleave synaptobrevin (BoNT/B, D, F, G), SNAP-25 (BoNT/A, E), or syntaxin (BoNT/C). Tetanus toxin cleaves synaptotagmin in the vesicle. These toxins are both lethal biological weapons and indispensable experimental tools for identifying SNARE-dependent release mechanisms.

Exercises Intermediate+

Exercise 4 (medium, symbolic).

Explain why the Hill coefficient of approximately 4 for synaptotagmin's calcium dependence is functionally important for synaptic transmission. What would happen if the calcium sensor had a Hill coefficient of 1?

Hint

Consider the steepness of the calcium-response curve and how it affects the coupling between calcium entry and vesicle release.

Answer

A Hill coefficient of 4 means vesicle release probability increases extremely steeply with calcium concentration — roughly as $[Ca^{2 +}]^{4}$ . This has two consequences. First, it makes release an ultra-sensitive switch: small increases in calcium above threshold produce large increases in release probability, ensuring that release is tightly coupled to the arrival of an action potential (which opens calcium channels briefly and locally). Second, it ensures that spontaneous calcium fluctuations at resting levels (0.1 $μ$ M) produce very little release, while the calcium microdomain near an open channel (~10–100 $μ$ M) triggers release reliably — the steep cooperativity provides high signal-to-noise.

With a Hill coefficient of 1, the calcium-response relation would be gradual rather than switch-like. Spontaneous calcium transients would trigger more release (increased "noise"), and the coupling between action potentials and release would be weaker and more variable. The steep cooperativity is essential for the temporal precision of synaptic transmission.

Exercise 5 (medium, short answer).

Botulinum toxin serotype A (BoNT/A) cleaves SNAP-25. Predict the effect on synaptic transmission and explain the molecular mechanism.

Hint

SNAP-25 contributes two helices to the SNARE four-helix bundle. What happens to zippering if SNAP-25 is truncated?

Answer

BoNT/A cleaves the C-terminal nine amino acids from SNAP-25, producing a truncated protein that can still form a partial SNARE complex but cannot complete the final zippering step. Vesicles can dock and partially prime, but calcium-triggered fusion is abolished or severely impaired. The result is complete blockade of neurotransmitter release at affected synapses, causing flaccid paralysis. The effect is long-lasting (weeks to months) because SNAP-25 must be resynthesised — this is why BoNT/A is used therapeutically (Botox injections) and why botulism is life-threatening.

Exercise 6 (hard, symbolic).

Derive the relation between extracellular calcium concentration and quantal content under the simplifying assumption that release probability per vesicle follows $p = α [Ca^{2 +}]_{local}^{4}$ , where $[Ca^{2 +}]_{local}$ is proportional to the extracellular calcium concentration (valid when intracellular buffering and extrusion are negligible on the timescale of the calcium transient). Express the mean quantal content $m$ as a function of $[Ca^{2 +}]_{o}$ and the pool size $N$ .

Hint

$m = N \times p$ , and $p = α [Ca^{2 +}]_{local}^{4}$ where $[Ca^{2 +}]_{local} = k \cdot [Ca^{2 +}]_{o}$ for some proportionality constant $k$ .

Answer

If the local calcium concentration at the release site is proportional to extracellular calcium, $[Ca^{2 +}]_{local} = k \cdot [Ca^{2 +}]_{o}$ , then the release probability per vesicle is:

p = α (k \cdot [Ca^{2 +}]_{o})^{4} = α k^{4} [Ca^{2 +}]_{o}^{4}

For a pool of $N$ release-ready vesicles, each releasing independently with probability $p$ , the mean quantal content is:

m = N \cdot p = N α k^{4} [Ca^{2 +}]_{o}^{4}

This predicts that quantal content should scale as the fourth power of extracellular calcium concentration, which is approximately confirmed experimentally at many synapses (the measured exponent ranges from 3 to 4, reflecting contributions from calcium buffering, cooperativity, and non-uniform calcium domains). The steep power law explains why small changes in calcium channel density or open probability produce large changes in synaptic strength — a biophysical substrate for synaptic plasticity.

Exercise 7 (hard, symbolic).

The NMDA receptor is blocked by $Mg^{2 +}$ in a voltage-dependent manner: at hyperpolarised potentials ( $V_{m} \approx - 70$ mV), $Mg^{2 +}$ occludes the pore; at depolarised potentials ( $V_{m} > - 30$ mV), the block is relieved. Explain why this property makes the NMDA receptor a coincidence detector and why this is important for synaptic plasticity.

Hint

What two conditions must be simultaneously satisfied for calcium to enter through NMDA receptors?

Answer

Calcium entry through NMDA receptors requires two conditions to be satisfied simultaneously: (1) glutamate must be bound (presynaptic activity — neurotransmitter release), and (2) the postsynaptic membrane must be sufficiently depolarised to relieve the $Mg^{2 +}$ block (postsynaptic activity). Neither condition alone produces significant calcium influx.

This makes the NMDA receptor a molecular coincidence detector: it gates calcium entry only when presynaptic and postsynaptic neurons are active at the same time. This property is the basis of Hebbian plasticity — the principle that "neurons that fire together, wire together." When a weak input (which alone produces only AMPA-mediated depolarisation too small to relieve the $Mg^{2 +}$ block) is paired with a strong input that depolarises the postsynaptic cell, the NMDA receptors at the weak synapse open, calcium enters, and calcium-dependent signalling cascades (CaMKII, PKC) strengthen the weak synapse (long-term potentiation, LTP). Without coincidence detection, every presynaptic input would be equally strengthened regardless of postsynaptic state, eliminating the selectivity required for learning.

Synaptic plasticity and advanced dynamics Master

Short-term plasticity

Synaptic strength is not fixed. On timescales of milliseconds to minutes, synapses exhibit short-term plasticity — activity-dependent changes in release probability and postsynaptic response.

Facilitation is a progressive increase in postsynaptic response during repetitive stimulation, lasting tens to hundreds of milliseconds. The dominant mechanism is residual calcium accumulation: each action potential leaves a small amount of calcium in the terminal that adds to the calcium influx of the next spike, increasing release probability. Formally, if resting release probability is $p_{1}$ and residual calcium increments release probability by $Δ p$ per spike, then the release probability on the $n$ -th spike in a train is approximately $p_{n} = p_{1} + (n - 1) Δ p$ , saturating at 1. Facilitation is prominent at synapses with initially low release probability (e.g., hippocampal mossy fibre synapses onto CA3 pyramidal cells, where $p_{1} \approx 0.05$ and $p_{10} \approx 0.6$ ).

Depression is a progressive decrease in response during repetitive stimulation, caused by depletion of the readily releasable pool (RRP). If the RRP has $N$ vesicles and each stimulus releases a fraction $p$ , the pool after $n$ stimuli is approximately $N (1 - p)^{n}$ . The recovery time constant is set by the vesicle refill rate ( $τ_{recovery} \approx 2$ – $10$ s). Depression dominates at high release probability synapses (e.g., the calyx of Held, $p \approx 0.6$ – $0.9$ ).

Post-tetanic potentiation (PTP) is an enhancement lasting seconds to minutes after a high-frequency train, mediated by calcium-dependent activation of protein kinase C (PKC) and the phosphorylation of Munc18 and SNAP-25, which increases priming efficiency and release probability.

The interplay of facilitation and depression determines the frequency dependence of synaptic transmission: low- $p$ synapses tend to facilitate and preferentially transmit high-frequency signals (high-pass filters), while high- $p$ synapses tend to depress and preferentially transmit onset responses (low-pass filters).

Long-term potentiation and depression

Long-term potentiation (LTP) is a persistent (hours to days) increase in synaptic strength following specific patterns of activity. At hippocampal CA3→CA1 Schaffer collateral synapses, the induction protocol is typically high-frequency stimulation (100 Hz for 1 s) or theta-burst stimulation (bursts of 4 pulses at 100 Hz, repeated at 5 Hz). The molecular mechanism requires NMDA receptor activation and calcium influx, followed by activation of CaMKII (calcium/calmodulin-dependent protein kinase II), which autophosphorylates and remains active after calcium returns to baseline. The expression of LTP involves the insertion of additional AMPA receptors into the postsynaptic membrane (AMPA receptor trafficking) via PKC-dependent exocytosis, increasing the postsynaptic response to each quantum of glutamate.

Long-term depression (LTD) is a persistent decrease in synaptic strength, induced by prolonged low-frequency stimulation (1–5 Hz for 10–15 minutes). The same NMDA receptor calcium entry underlies both LTP and LTD, but the amplitude and duration of the calcium signal determine the direction: large, rapid calcium transients activate CaMKII and trigger LTP; smaller, sustained calcium elevations preferentially activate protein phosphatases (PP1, PP2A, calcineurin) and trigger LTD by internalising AMPA receptors.

Spike-timing-dependent plasticity (STDP) refines the Hebbian rule by making the sign of plasticity depend on the precise temporal order of pre- and postsynaptic spikes. If the presynaptic spike precedes the postsynaptic spike by less than ~20 ms, the synapse is strengthened (timing-dependent LTP); if the postsynaptic spike precedes the presynaptic spike, the synapse is weakened (timing-dependent LTD). The mechanism exploits the back-propagating action potential that depolarises the dendrite: when presynaptic glutamate release slightly precedes the back-propagating spike, NMDA receptors experience both glutamate binding and postsynaptic depolarisation simultaneously, producing the large calcium transient that triggers LTP.

Spontaneous release and minis

In the absence of presynaptic action potentials, synaptic vesicles fuse at a low basal rate (~0.1–1 events per second per synapse), producing miniature postsynaptic potentials (minis). These events were first recorded by Fatt and Katz (1952) at the neuromuscular junction. Minis have the same quantal amplitude as evoked release, confirming that they represent single vesicle fusion events. Their frequency is calcium-dependent (increased by raising extracellular calcium or depolarising the terminal) but persists at a reduced rate even when voltage-gated calcium channels are blocked, suggesting a distinct molecular trigger — possibly involving spontaneous calcium release from intracellular stores or calcium-independent fusion pathways mediated by Doc2 and other proteins.

The physiological role of minis is debated. Proposed functions include: maintaining postsynaptic receptor clustering (preventing synaptic silencing), regulating dendritic protein synthesis, and contributing to synaptic homeostasis (the homeostatic scaling of synaptic strength to maintain stable activity levels).

Synaptic vesicle protein composition

A synaptic vesicle contains a defined set of membrane proteins that constitute approximately 60% of the vesicle's mass. The core vesicle proteins include:

Synaptobrevin/VAMP — the vesicle SNARE (v-SNARE); cleaved by botulinum toxin B/D/F/G and tetanus toxin
Synaptotagmin — the calcium sensor; C2 domains bind calcium cooperatively and insert into the target membrane
Synaptophysin — a tetraspan vesicle membrane protein; function not fully resolved, possibly regulating vesicle fusion kinetics
SV2 (synaptic vesicle protein 2) — a twelve-transmembrane-domain protein with structural similarity to sugar transporters; involved in vesicle priming and seizure prevention (SV2A knockout mice have lethal seizures; the antiepileptic drug levetiracetam binds SV2A)
Synapsin — a peripheral membrane protein that tethers vesicles to the actin cytoskeleton in the reserve pool; phosphorylation by CaMKII or PKA releases vesicles from the reserve pool, making them available for release
VGLUT (vesicular glutamate transporter), VGAT (vesicular GABA/glycine transporter), VAChT (vesicular acetylcholine transporter), VMAT (vesicular monoamine transporter) — vesicular neurotransmitter transporters that fill vesicles using the proton gradient maintained by the V-ATPase

Kiss-and-run vs full-collapse fusion

Two modes of vesicle fusion have been identified by capacitance recording and amperometry. In full-collapse fusion, the vesicle membrane completely merges with the plasma membrane, and the vesicle is recovered by clathrin-mediated endocytosis over 10–30 seconds. In kiss-and-run fusion, the vesicle forms a transient fusion pore (~~1 nm diameter) through which neurotransmitter partially escapes, and the pore closes within milliseconds without full collapse. Kiss-and-run preserves vesicle identity and enables rapid recycling (~~1 s), but may release only a fraction of the vesicle's neurotransmitter content (partial release).

The choice between kiss-and-run and full-collapse fusion depends on calcium concentration, stimulation frequency, and the molecular composition of the release machinery. At the calyx of Held synapse, kiss-and-run predominates during low-frequency stimulation, while full-collapse fusion dominates during high-frequency trains — suggesting that the synapse switches between modes to balance speed of recycling against completeness of release.

Neuromodulation

Neuromodulators are neurotransmitters (primarily dopamine, serotonin, norepinephrine, acetylcholine, and various neuropeptides) that do not directly excite or inhibit postsynaptic neurons but instead modulate their excitability, synaptic strength, and plasticity through metabotropic receptors (GPCRs). Neuromodulatory systems originate from small brainstem or basal forebrain nuclei and project diffusely throughout the brain:

Dopamine (ventral tegmental area, substantia nigra) — D1/D2 receptors modulate cAMP signalling, affecting working memory, reward learning, and motor control
Serotonin (raphe nuclei) — 5-HT receptors (14 subtypes) modulate mood, sleep, appetite, and pain
Acetylcholine (basal forebrain, brainstem) — muscarinic receptors modulate attention, learning, and cortical plasticity; nicotinic receptors are ionotropic and mediate fast excitation
Norepinephrine (locus coeruleus) — alpha and beta adrenergic receptors modulate arousal, attention, and stress responses

Neuromodulators alter synaptic transmission by (1) changing presynaptic release probability (via modulation of calcium channels or the release machinery), (2) changing postsynaptic responsiveness (via phosphorylation of receptors), and (3) altering intrinsic excitability (via modulation of voltage-gated channels). The net effect is a gain control that adjusts the sensitivity and plasticity of entire circuits without carrying specific information themselves.

Connections Master

Resting membrane potential and ion channels 17.09.01. The electrochemical gradients established by the Na+/K+ pump provide the driving force for all postsynaptic currents. At a resting potential of $- 70$ mV, the sodium driving force is ~137 mV inward, making glutamate-gated AMPA receptor opening strongly depolarising. The resting potential stores the energy that synaptic transmission releases.
The action potential 17.09.02. The action potential is the presynaptic signal that opens voltage-gated calcium channels and triggers vesicle fusion. The frequency of action potentials encodes information at the single-neuron level, and synaptic transmission converts that frequency code into postsynaptic depolarisation (or hyperpolarisation).
Vesicle trafficking 17.02.04 pending. Synaptic vesicle exocytosis is a specialised form of the general vesicle trafficking machinery used throughout the secretory pathway. The SNARE proteins, Rab GTPases, and Sec1/Munc18 family proteins are conserved from yeast to humans. Synaptic transmission adds speed (sub-millisecond fusion) and calcium sensitivity (synaptotagmin) to the generic trafficking toolkit.
Cell signalling — GPCRs 17.07.01. Metabotropic neurotransmitter receptors (mGluRs, GABA-B, muscarinic ACh receptors, dopamine receptors, serotonin receptors) are GPCRs. The signalling cascades they activate (cAMP, IP3/DAG, MAPK) modulate synaptic strength and neuronal excitability on timescales of seconds to hours.
Pharmacology of ion channels 17.09.04 pending. Many drugs target synaptic transmission: local anaesthetics block voltage-gated sodium channels (preventing the presynaptic action potential), calcium channel blockers reduce neurotransmitter release, benzodiazepines enhance GABA-A receptor function (increasing inhibition), and SSRIs block serotonin reuptake. The receptor gating mechanisms described here are the direct targets of ion channel pharmacology.
Membrane proteins 17.02.03 pending. Ligand-gated ion channels (AMPA, NMDA, GABA-A, glycine receptors) and voltage-gated calcium channels are membrane proteins whose structure and function build on the general principles of membrane protein biology: transmembrane helices, extracellular ligand-binding domains, and conformational gating transitions.

Historical notes Master

The chemical nature of synaptic transmission was established in a series of experiments spanning the early twentieth century. In 1921, Otto Loewi demonstrated that vagus nerve stimulation releases a soluble substance ("Vagusstoff," later identified as acetylcholine) that slows the heart, providing the first direct evidence for chemical neurotransmission. Henry Dale extended this work to the nervous system, establishing the Dale's principle (now refined) that a neuron releases the same neurotransmitter at all of its synapses. Loewi and Dale shared the 1936 Nobel Prize.

Bernard Katz and Ricardo Miledi at University College London performed the decisive experiments on the quantal nature of release in the 1950s and 1960s ^{[Katz & Miledi 1965]}. Using extracellular recordings at the frog neuromuscular junction, they showed that the postsynaptic end-plate potential fluctuates in discrete steps, each a multiple of the miniature end-plate potential (mevt) observed spontaneously. This led to the quantal hypothesis: neurotransmitter is released in packets (vesicles), and the number of quanta released follows Poisson statistics. Katz received the 1970 Nobel Prize.

The SNARE hypothesis was proposed by James Rothman and colleagues in 1993 based on biochemical reconstitution of vesicle fusion in cell-free systems. Thomas Sudhof at Stanford identified synaptotagmin as the calcium sensor in 1992–1994 and elucidated the triggering mechanism through a series of knockout and reconstitution experiments. The structural basis of SNARE zippering was resolved by Axel Brunger and colleagues using X-ray crystallography of the neuronal SNARE complex at 2.4 A resolution in 1998. Rothman and Sudhof shared the 2013 Nobel Prize with Randy Schekman for their work on vesicle trafficking.

The NMDA receptor's role as a coincidence detector for Hebbian plasticity was demonstrated experimentally by Mark Bear, Robert Malenka, and Roger Nicoll in the late 1980s and early 1990s, building on the theoretical framework proposed by Donald Hebb in 1949 ("The Organization of Behavior"). The discovery of spike-timing-dependent plasticity by Henry Markram, Sen Song, and others in the mid-1990s, and its quantitative characterisation by Guo-qiang Bi and Mu-ming Poo in 1998, provided a temporally precise instantiation of the Hebbian rule.

Botulinum toxin has been known as a cause of fatal food poisoning since the early nineteenth century (Justinus Kerner's descriptions of "sausage poisoning," 1817–1822). The molecular target was identified only in the 1990s, when it was shown that BoNT serotypes cleave SNARE proteins, providing some of the strongest causal evidence that SNARE-mediated exocytosis is necessary for neurotransmitter release.

Bibliography Master

Primary literature.

Loewi, O., "Uber humorale Ubertragbarkeit der Herznervenwirkung", Pflugers Arch. 189 (1921), 239–242.
Fatt, P. & Katz, B., "Spontaneous subthreshold activity at motor nerve endings", J. Physiol. 117 (1952), 109–128.
del Castillo, J. & Katz, B., "Quantal components of the end-plate potential", J. Physiol. 124 (1954), 560–573.
Katz, B. & Miledi, R., "The measurement of synaptic delay", J. Physiol. 181 (1965), 656–670.
Hebb, D. O., The Organization of Behavior: A Neuropsychological Theory (Wiley, 1949).
Sollner, T. et al., "SNAP receptors implicated in vesicle targeting and fusion", Nature 362 (1993), 318–324.
Sudhof, T. C. & Rothman, J. E., "Membrane fusion: grappling with SNARE and SM proteins", Science 323 (2009), 474–477.
Sudhof, T. C., "The synaptic vesicle cycle", Annu. Rev. Neurosci. 27 (2004), 509–547.
Sudhof, T. C., "Neurotransmitter release: the last millisecond in the life of a synaptic vesicle", Neuron 80 (2013), 675–690.
Bi, G.-q. & Poo, M.-m., "Synaptic modifications in cultured hippocampal neurons: dependence on spike timing, synaptic strength, and postsynaptic cell type", J. Neurosci. 18 (1998), 10464–10472.
Bliss, T. V. P. & Collingridge, G. L., "A synaptic model of memory: long-term potentiation in the hippocampus", Nature 361 (1993), 31–39.
Sutton, R. B. et al., "Crystal structure of a SNARE complex involved in synaptic exocytosis at 2.4 A resolution", Nature 395 (1998), 347–353.

Textbook and monograph.

Hille, B., Ion Channels of Excitable Membranes, 3rd ed. (Sinauer, 2001).
Kandel, E. R. et al., Principles of Neural Science, 5th ed. (McGraw-Hill, 2013).
Alberts, B. et al., Molecular Biology of the Cell, 7th ed. (Garland Science, 2022).
Lodish, H. et al., Molecular Cell Biology, 9th ed. (W. H. Freeman, 2021).
Katz, B., Nerve, Muscle, and Synapse (McGraw-Hill, 1966).