Between one cell and the next
— twenty nanometres, and half a millisecond.

The minimum high-yield framework

1. Definition. A synapse is the junction at which a neuron communicates with another neuron, a muscle cell or a gland cell.
2. Classification — say this first. Synapses are chemical or electrical. Chemical synapses are further divided by the receptor into ionotropic (ligand-gated ion channel) and metabotropic (G-protein-coupled, second messenger).
3. Six steps at a chemical synapse: action potential → voltage-gated Ca²⁺ entry → vesicle fusion and exocytosis → diffusion across the cleft → receptor binding → postsynaptic potential change. Then removal of the transmitter.
4. Calcium is the trigger. No calcium entry, no release. This is the step magnesium and aminoglycosides interfere with.
5. EPSP — depolarising, from increased Na⁺ and K⁺ conductance, net Na⁺ influx. IPSP — hyperpolarising, from increased Cl⁻ influx or K⁺ efflux.
6. One EPSP is 0.5-1 mV and threshold needs 10-20 mV. Firing therefore requires summation: temporal (same synapse, repeated) or spatial (different synapses, simultaneous). Integration happens at the axon hillock.
7. Synaptic delay is about 0.5 ms, most of it in transmitter release.
8. Inhibition is presynaptic or postsynaptic. Presynaptic acts through an axo-axonic synapse to reduce calcium entry and therefore transmitter release; postsynaptic hyperpolarises the target cell.

Previously examinedOctober 2021 — the types of synapse, the physiological advantage of each, and how the postsynaptic impulse brings about skeletal muscle contraction. Worked answers in the library

Anatomical naming, which is a separate question

A synapse can also be named by where on the postsynaptic cell it lands: axodendritic (onto a dendrite, the commonest), axospinous (onto a dendritic spine), axosomatic (onto the cell body, and typically inhibitory because it is close to the trigger zone), and axo-axonic (onto another axon terminal, which is the anatomical substrate of presynaptic inhibition). This is a useful supplementary classification but it is not the fundamental one, and offering it alone does not answer the question.

An electrical synapse is a gap junction. Six connexin subunits form a connexon hemichannel in each of the two apposed membranes; two connexons dock end to end to form a continuous aqueous pore about 1.5 nm wide, and the two cytoplasms become continuous for ions and for small molecules up to about 1 kDa. The intercellular gap is about 3 nm, against roughly 20 nm at a chemical synapse.

Because current passes directly, the postsynaptic response begins at essentially the same instant as the presynaptic one — there is no measurable synaptic delay, and no transmitter, vesicle or receptor is involved. Transmission is bidirectional, and because it is passive it cannot fatigue.

The price is that it can do nothing but pass the signal on. It cannot amplify, cannot invert the sign of the signal, and offers almost no scope for modulation. A nervous system built only from electrical synapses could conduct but could not compute.

The structural elements are the same in every chemical synapse: a presynaptic terminal containing vesicles, mitochondria and voltage-gated calcium channels; a synaptic cleft of about 20 nm; and a postsynaptic membrane bearing receptors. What varies is the receptor.

Ionotropic receptors

The receptor and the ion channel are the same molecule. Transmitter binding produces a conformational change that opens the pore directly, so the ionic current begins within a millisecond of binding and stops as soon as the transmitter leaves. Examples you must be able to give:

• Nicotinic acetylcholine receptor — five subunits around a central cation pore. Two acetylcholine molecules must bind the two α subunits before it opens. Non-selective for cations, so net sodium influx and depolarisation.
• GABA_A receptor — a chloride channel. Opening admits chloride, hyperpolarising the mature neuron. The target of benzodiazepines, barbiturates, propofol and etomidate.
• Glycine receptor — also a chloride channel; the dominant fast inhibitory receptor in the spinal cord and brainstem.
• NMDA and AMPA glutamate receptors — cation channels. The NMDA receptor also admits calcium and is blocked by magnesium at resting potential, which is what makes it the coincidence detector behind wind-up and long-term potentiation.
• 5-HT₃ receptor — the only ionotropic serotonin receptor, and the target of ondansetron.

Metabotropic receptors

The receptor is a seven-transmembrane G-protein-coupled receptor. Transmitter binding activates a heterotrimeric G protein, whose α subunit exchanges GDP for GTP and dissociates to act on an effector — adenylyl cyclase, phospholipase C, or an ion channel opened at a distance. The response takes hundreds of milliseconds to seconds and can last far longer than the transmitter is present.

The gain is amplification: one occupied receptor activates many G proteins, each of which activates an effector that generates many second-messenger molecules. Examples: muscarinic acetylcholine receptors, adrenergic receptors, opioid receptors, GABA_B, and the metabotropic glutamate receptors on astrocytes that contribute to neurovascular coupling.

1. The action potential invades the terminal and depolarises it.
2. Voltage-gated calcium channels open — P/Q-type at the neuromuscular junction, N-type at many central terminals. Calcium enters down a gradient of roughly ten thousand-fold, and intracellular calcium rises transiently in microdomains beside the channel mouths.
3. Vesicles dock and fuse. Calcium binds synaptotagmin, the calcium sensor, which triggers the SNARE complex — synaptobrevin on the vesicle with syntaxin and SNAP-25 on the terminal membrane — to zipper together and fuse the membranes. Transmitter is released by exocytosis in multimolecular quanta.
4. Transmitter diffuses across the cleft, a distance of about 20 nm, which takes microseconds.
5. It binds postsynaptic receptors and changes membrane permeability, producing an EPSP or an IPSP.
6. It is removed, by one of three routes: diffusion out of the cleft, enzymatic breakdown (acetylcholinesterase for acetylcholine), or reuptake into the presynaptic terminal or into astrocytes (noradrenaline, glutamate, GABA, serotonin).

Properties of synapses

A typical central EPSP is 0.5 to 1 mV. Reaching threshold from rest requires 10 to 20 mV of depolarisation. It follows that no single excitatory synapse in the central nervous system can fire a neuron, and that firing always requires the arithmetic of many inputs.

Both forms of summation are algebraic: EPSPs and IPSPs arriving together subtract. The result is a continuously varying membrane potential at the axon hillock, and the cell fires whenever that sum crosses threshold. Because the axon hillock has the lowest threshold in the cell, it is the site at which the arithmetic is done — which is why the axon hillock is described as the trigger zone and why an inhibitory synapse placed on the soma is more powerful than one placed on a distal dendrite.

Two further terms belong here. Reciprocal (reciprocal innervation) inhibition is the arrangement in which the afferent that excites a motor neuron also inhibits, through an interneuron, the motor neuron of the antagonist — the reason a stretch reflex produces movement rather than co-contraction. Recurrent inhibition is a motor neuron inhibiting itself and its neighbours through the Renshaw cell, a negative feedback that limits and sharpens motor output. Both reappear in lesson 15.

Reading that table the other way round is a good revision exercise: given a step in the release sequence, name a drug that acts there. If you can do it in both directions, you can answer any pharmacology question that starts “describe the mechanism of action of” a centrally acting drug.