Physiology
Synaptic Transmission
An arriving spike floods Ca2+ into the terminal, fusing vesicles that dump neurotransmitter across a 20 nm cleft in under a millisecond
Synaptic transmission is how one neuron talks to the next: an action potential opens voltage-gated Ca2+ channels in the presynaptic terminal, the Ca2+ spike triggers SNARE-mediated fusion of neurotransmitter vesicles within ~0.2 ms, and the transmitter diffuses across the ~20 nm cleft to open postsynaptic receptors. Bernard Katz proved the quantal, calcium-dependent mechanism (Nobel 1970); SNARE proteins were identified by Rothman, Schekman and Südhof (Nobel 2013).
- Synaptic cleft~20-40 nm wide
- Synaptic delay~0.5-1 ms
- Calcium sensorSynaptotagmin-1
- Fusion machinerySNAREs (VAMP2, syntaxin-1, SNAP-25)
- Quantum size~5,000 molecules / vesicle
- Solved byKatz 1970 · Südhof 2013
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
What synaptic transmission is
A neuron's action potential is an electrical event, but most neurons are not physically wired to each other — they are separated by a tiny gap. Synaptic transmission is the trick nature uses to jump that gap: it converts the electrical spike into a brief chemical squirt and back into electricity again. The whole conversion happens at a structure called the chemical synapse, and at a typical fast synapse it is over in about a millisecond.
The cast is small. The presynaptic terminal is the swollen end of the sending axon, packed with hundreds of membrane sacs called synaptic vesicles, each loaded with neurotransmitter. A 20-nanometer gap — the synaptic cleft — separates it from the postsynaptic membrane of the receiving cell, which bristles with receptor proteins. The handoff goes: spike arrives → calcium enters → vesicles fuse → transmitter crosses the cleft → receptors open → the receiving cell's voltage changes. Crucially, the signal flows one way (presynaptic to postsynaptic) because only the presynaptic side has vesicles and only the postsynaptic side has the receptors — chemical synapses are diodes.
How the handoff works, step by step
Walk through one transmission event in order:
- The spike arrives and depolarizes the terminal. The action potential propagating down the axon reaches the bouton and depolarizes its membrane from about -70 mV toward 0 mV.
- Voltage-gated Ca2+ channels open. Depolarization opens Cav2.1 (P/Q-type) and Cav2.2 (N-type) calcium channels clustered just tens of nanometers from the docked vesicles at the active zone. Because extracellular Ca2+ (~2 mM) is roughly 10,000-fold higher than resting cytosolic Ca2+ (~100 nM), calcium pours in.
- A calcium microdomain forms. Right at the channel mouth, local Ca2+ jumps to tens or hundreds of micromolar for a fraction of a millisecond — a steep, short-lived spike, not a uniform rise.
- Synaptotagmin-1 senses the calcium. This vesicle protein has two C2 domains that cooperatively bind 4-5 Ca2+ ions. Calcium-loaded synaptotagmin plunges into the plasma membrane and grabs the SNARE complex.
- SNAREs zipper and fuse the membranes. The vesicle's synaptobrevin (VAMP2) winds together with the target membrane's syntaxin-1 and SNAP-25 into a four-helix bundle, zippering from the far end toward the membranes. That zippering force drags the two bilayers together and opens a fusion pore.
- A quantum of transmitter is released. The vesicle empties ~5,000 transmitter molecules into the cleft within ~0.2 ms of calcium entry.
- Transmitter crosses and binds receptors. Diffusion across 20 nm is effectively instant. Transmitter binds postsynaptic ionotropic receptors (channels that open directly, e.g. nicotinic ACh, AMPA, GABA-A) or metabotropic receptors (GPCRs that act through second messengers), changing the postsynaptic voltage.
- The signal is terminated and recycled. Transmitter is cleared by reuptake, enzymatic breakdown, or diffusion; the vesicle membrane is retrieved (often via clathrin-mediated endocytosis) and refilled; NSF/alpha-SNAP pull the spent SNAREs apart for reuse.
The molecular players and conditions
- Calcium channels (Cav2.1, Cav2.2). The trigger. They open in microseconds on depolarization and sit within ~25 nm of the release sites so the calcium signal is fast and local. Cone-snail omega-conotoxins and the autoimmune disease Lambert-Eaton (antibodies against Cav2.1) block them and weaken transmission.
- Synaptotagmin-1. The calcium sensor and the timer. Knock it out and fast, synchronous release disappears while slow, asynchronous release persists — proving it is the switch that makes release fast.
- The SNARE complex (VAMP2 + syntaxin-1 + SNAP-25). The fusion engine. Assembled SNAREs are so stable they resist SDS and boiling. They are the substrate of the clostridial neurotoxins (see below).
- Munc18 and Munc13. Chaperones that template SNARE assembly and "prime" docked vesicles into a fusion-ready, release-competent state. Without priming a docked vesicle cannot fuse.
- Complexin. A clamp that holds primed SNAREs in a half-zippered, cocked state, ready to fire the instant calcium-synaptotagmin releases the brake — this is part of why release is so fast.
- The neurotransmitter. Glutamate (main excitatory in CNS), GABA and glycine (inhibitory), acetylcholine (neuromuscular junction and autonomic), plus dopamine, serotonin, norepinephrine. A vesicle transporter (vGLUT, vGAT, VAChT) loads each vesicle using the proton gradient from a vesicular H+-ATPase.
- Postsynaptic receptors and the density. Ionotropic receptors give millisecond responses; metabotropic receptors give slower, longer modulation. They are anchored opposite the active zone by scaffolds (PSD-95 at excitatory synapses, gephyrin at inhibitory ones) in the postsynaptic density.
Chemical vs electrical synapse
| Property | Chemical synapse | Electrical synapse (gap junction) |
|---|---|---|
| Connector | Neurotransmitter across a cleft | Connexon channels bridging cytoplasm |
| Cleft width | ~20-40 nm | ~3.5 nm |
| Synaptic delay | ~0.5-1 ms | Essentially 0 (near-instant) |
| Direction | One-way (rectifying) | Usually bidirectional |
| Sign change | Can be excitatory OR inhibitory | Passes current as-is, no inversion |
| Amplification / gain | Yes — a small spike can release a large response | No — signal attenuates |
| Plasticity / modulation | Highly plastic (LTP, LTD, neuromodulators) | Limited |
| Speed of signaling | Slower per relay but flexible | Fast synchrony (escape reflexes, heart, retina) |
Sizes, speeds and numbers
Synapses are extreme in their geometry and kinetics. The cleft is ~20-40 nm wide; a synaptic vesicle is ~40 nm in diameter and holds on the order of 5,000-10,000 transmitter molecules. A small CNS bouton may have a few hundred vesicles, of which only ~5-10 are docked and primed in the "readily releasable pool" at any instant. Per action potential, a single small synapse releases very little — release probability for one vesicle is often only 0.1-0.4 — which is exactly why neurons pool hundreds of synaptic inputs.
| Quantity | Typical value | Note |
|---|---|---|
| Synaptic cleft width | ~20 nm (CNS), ~50 nm (NMJ) | Diffusion across is < 1 µs |
| Synaptic delay | ~0.5-1 ms | Dominated by Ca2+-triggered fusion |
| Time to fusion after Ca2+ entry | ~0.2 ms | Fast, synchronous release |
| Extracellular vs resting cytosolic Ca2+ | ~2 mM vs ~100 nM | ~10,000-fold gradient drives entry |
| Local microdomain Ca2+ at fusion | tens-hundreds of µM | Brief spike at channel mouth |
| Ca2+ cooperativity of release | ~4th power | Small Ca2+ change → large release change |
| Molecules per quantum | ~5,000 (ACh) | Content of one vesicle |
| Mini endplate potential (mEPP) | ~0.5 mV (NMJ) | Single-vesicle response, the "quantum" |
| Single central EPSP | ~0.1-1 mV | Needs summation to reach threshold |
| Acetylcholinesterase turnover | ~25,000 / s | Among the fastest enzymes known |
Where it shows up — drugs, toxins and disease
- Botulinum and tetanus toxins target SNAREs directly. Both are zinc proteases. Botulinum toxin (Botox) cleaves SNAP-25 or synaptobrevin at the neuromuscular junction, blocking acetylcholine release and causing flaccid paralysis — the same mechanism used cosmetically to relax muscles. Tetanus toxin cleaves synaptobrevin in inhibitory spinal interneurons, blocking GABA/glycine release and causing spastic paralysis (lockjaw). These toxins are the most potent known to biology and were the smoking gun that SNAREs do fusion.
- Antidepressants and stimulants act on reuptake. SSRIs (fluoxetine, sertraline) block the serotonin transporter SERT; cocaine blocks the dopamine transporter DAT; amphetamine reverses it. Each leaves transmitter in the cleft longer.
- Acetylcholinesterase inhibitors — from war to Alzheimer's. Nerve agents (sarin, VX) and the insecticide class organophosphates irreversibly block acetylcholinesterase, flooding synapses with ACh until muscles and lungs fail. The same target, gently inhibited by donepezil and rivastigmine, modestly boosts cholinergic signaling in Alzheimer's disease.
- Myasthenia gravis and Lambert-Eaton. In myasthenia gravis, autoantibodies destroy postsynaptic nicotinic ACh receptors at the neuromuscular junction, causing fatigable weakness. In Lambert-Eaton syndrome, antibodies attack the presynaptic Cav2.1 calcium channels, so less ACh is released.
- Glutamate excitotoxicity. When glial EAAT transporters fail to clear glutamate — during stroke, trauma or in ALS — excess glutamate over-activates NMDA receptors, floods neurons with Ca2+, and kills them. Memantine, an NMDA-receptor blocker, is used in Alzheimer's to blunt this.
- Synapses are where learning lives. Long-term potentiation (LTP) and depression (LTD) — lasting increases or decreases in synaptic strength, largely via inserting or removing AMPA receptors — are the cellular substrate of memory, first described in the rabbit hippocampus by Bliss and Lømo in 1973.
Quantal release: how Katz read the synapse
The single most important conceptual result about the synapse is that transmitter comes in fixed packets. Bernard Katz and José del Castillo, recording at the frog neuromuscular junction in the 1950s, noticed tiny spontaneous voltage blips even with no nerve stimulation — the miniature endplate potentials (minis / mEPPs), each ~0.5 mV. When they evoked a real response in low-calcium saline, the responses came in clean integer multiples of the mini size: 1×, 2×, 3×. The conclusion: each "quantum" is one vesicle's worth of transmitter, and the synapse sets its strength by how many vesicles fuse, not by how full each one is. Katz won the 1970 Nobel Prize for this. Decades later, James Rothman, Randy Schekman and Thomas Südhof worked out the actual molecular machinery — SNAREs, synaptotagmin, the priming proteins — and shared the 2013 Nobel Prize. The dependence of release on calcium to roughly the fourth power, also from Katz's lab (with Ricardo Miledi), explains why synapses are such sensitive amplifiers.
Common misconceptions
- "The action potential jumps across the synapse." It does not. At a chemical synapse the electrical signal stops at the presynaptic membrane; a chemical messenger crosses the gap, and a fresh electrical signal is generated postsynaptically. Only at the much rarer electrical synapse does current actually flow across.
- "More calcium just means a bigger signal, linearly." Release scales with roughly the fourth power of calcium, so it is steeply nonlinear — a hallmark that synaptotagmin must bind several Ca2+ ions cooperatively. Halving calcium can cut release by an order of magnitude.
- "Neurotransmitter is either excitatory or inhibitory." The effect is set by the receptor, not the transmitter. Acetylcholine excites skeletal muscle (nicotinic) but slows the heart (muscarinic, via K+ channels). The same molecule can do opposite things at different receptors.
- "Synaptotagmin causes fusion." The SNAREs supply the fusion energy; synaptotagmin and complexin act as the trigger and clamp that make fusion fast and calcium-gated. Without SNAREs there is no fusion at all; without synaptotagmin fusion still happens but loses its tight timing.
- "One spike reliably fires the next neuron." At most central synapses a single vesicle has only a 10-40% chance of releasing, and one EPSP is ~0.1-1 mV — far below the ~15 mV needed to reach threshold. Firing requires summation of many inputs in space and time.
- "The vesicle dissolves when it releases." The vesicle membrane fuses with and is later retrieved from the plasma membrane, then refilled and reused. A busy terminal recycles its vesicle pool continuously rather than manufacturing new ones for each spike.
Frequently asked questions
How does an action potential trigger neurotransmitter release?
The action potential depolarizes the presynaptic terminal membrane, which opens voltage-gated Ca2+ channels (mainly Cav2.1 / P/Q-type and Cav2.2 / N-type) clustered at the active zone. Calcium is about 10,000 times more concentrated outside the cell (~2 mM) than inside (~100 nM at rest), so it floods in, creating a brief, steep microdomain of high Ca2+ (tens to hundreds of micromolar) right next to the docked vesicles. Synaptotagmin-1, a vesicle protein with two C2 domains, binds 4-5 of those calcium ions cooperatively. That binding pulls synaptotagmin into the plasma membrane and into the SNARE complex, removing the clamp that holds the vesicle back and triggering fusion. Because release depends on Ca2+ to roughly the fourth power, a small change in calcium entry causes a large change in release — which is why blocking these channels (as omega-conotoxins from cone snails do) silences synapses.
What are SNARE proteins and what do they do?
SNARE proteins are the core fusion machinery of the synapse. The vesicle (v-SNARE) carries synaptobrevin, also called VAMP2; the target plasma membrane (t-SNAREs) carries syntaxin-1 and SNAP-25. Their helical SNARE motifs zipper together from the membrane-distal end toward the membranes, forming an extremely stable four-helix coiled-coil. That zippering pulls the vesicle and plasma membranes into close apposition and supplies the energy to overcome the electrostatic repulsion and fuse the two lipid bilayers into one, opening a fusion pore. After fusion, the ATPase NSF and its adaptor alpha-SNAP pry the spent SNARE bundle apart so the components can be recycled. Clostridial neurotoxins prove how central SNAREs are: botulinum and tetanus toxins are proteases that cleave exactly these three proteins, abolishing transmission.
How wide is the synaptic cleft and how fast does transmission happen?
The synaptic cleft is only about 20-40 nm wide (around 20 nm at a typical CNS chemical synapse, ~50 nm at the neuromuscular junction). Diffusion across such a tiny gap is essentially instantaneous: a neurotransmitter molecule crosses 20 nm in well under a microsecond. The rate-limiting step is the synaptic delay — the time from the presynaptic spike to the start of the postsynaptic response — which is about 0.5-1 ms, almost all of which is the calcium-triggered fusion step itself. Vesicle fusion begins roughly 0.2 ms after calcium enters. By contrast, electrical synapses (gap junctions) have essentially zero delay because current flows directly between cells, but they cannot amplify, invert sign, or be modulated the way chemical synapses can.
What is quantal release and what does 'quantum' mean at the synapse?
Quantal release means neurotransmitter is released in fixed packets, not continuously. Each quantum is the content of a single synaptic vesicle — roughly 5,000 to 10,000 transmitter molecules (about 5,000 acetylcholine molecules in a vesicle ~40 nm across). Bernard Katz and Jose del Castillo discovered this in the 1950s by recording tiny spontaneous voltage blips at the frog neuromuscular junction, the miniature endplate potentials (minis or MEPPs), each about 0.5 mV. They found that evoked responses came in integer multiples of the mini size: 1, 2, 3 quanta. So a synapse grades its strength by changing the probability that a vesicle fuses and the number of release-ready vesicles, not by changing the size of each packet. Katz received the Nobel Prize in 1970 for this quantal theory.
How is the neurotransmitter signal switched off?
Three mechanisms clear transmitter from the cleft so the synapse can reset. First, reuptake: transporters pump transmitter back into the presynaptic terminal or surrounding glia — the serotonin transporter (SERT), dopamine transporter (DAT) and the GABA and glutamate transporters do this, and they are the targets of SSRIs, cocaine and amphetamine. Second, enzymatic degradation: at cholinergic synapses, acetylcholinesterase splits acetylcholine into acetate and choline within ~1 ms, one of the fastest enzymes known (turnover ~25,000 molecules per second); nerve agents like sarin and drugs like donepezil inhibit it. Third, simple diffusion out of the cleft. Glutamate, the main excitatory transmitter, is mostly cleared by glial EAAT transporters; failure to clear it causes excitotoxicity, a major driver of stroke and neurodegeneration damage.
What is the difference between an EPSP and an IPSP?
An excitatory postsynaptic potential (EPSP) is a small depolarization that pushes the postsynaptic membrane toward the firing threshold; it is produced when an excitatory transmitter like glutamate opens cation channels (AMPA and NMDA receptors) letting Na+ and Ca2+ in. An inhibitory postsynaptic potential (IPSP) is a hyperpolarization or shunt that moves the membrane away from threshold; it is produced when an inhibitory transmitter like GABA or glycine opens Cl- channels (GABA-A receptor) or, via metabotropic receptors, K+ channels. A single EPSP at a central synapse is tiny — roughly 0.1-1 mV — so a neuron must integrate dozens to hundreds of EPSPs across space (spatial summation) and time (temporal summation) before the axon hillock reaches threshold and fires its own action potential. Whether a neuron fires is the running sum of all its excitatory and inhibitory inputs.