Synaptic transmission · Grey Matter

Synaptic transmission is how one neuron passes a signal to the next across a synapse, by turning the electrical spike into a chemical released into a narrow gap and back into an electrical change in the receiving cell; electricity inside the neuron, chemistry between neurons.

Synaptic transmission. Synaptic transmission is how one neuron passes a signal to the next across a synapse, by turning the electrical spike into a chemical released into a narrow gap and back into an electrical change in the receiving cell; electricity inside the neuron, chemistry between neurons.

A chemical synapse has three parts: the presynaptic terminal, packed with vesicles of transmitter; the cleft, a gap of about 20 nm; and the postsynaptic membrane, dense with receptors. The spike opens calcium channels in the terminal, calcium triggers exocytosis, the transmitter crosses the cleft in microseconds and binds receptors, and the receptors open channels or start chemical cascades. The result in the receiving cell is a small graded potential, excitatory (an EPSP) or inhibitory (an IPSP), which the cell adds to all its other inputs.

It is fast but not instant. At fast synapses at body temperature the receiving cell begins to respond about 0.15 ms after the spike arrives; textbook figures for the synaptic delay are 0.5 to 1 ms, and an electrical synapse through gap junctions is faster still.

It is where the brain computes and learns. Each step (how much calcium enters, how likely a vesicle is to fuse, how many receptors wait) can be adjusted, which is what plasticity and most drugs act on.

The numbers are large. The human brain holds an estimated 100 to 500 trillion synapses, about 150 trillion of them in the neocortex, and a cortical pyramidal cell receives thousands to tens of thousands.

The signal must be removed. Transmitter is cleared by reuptake, enzymes and diffusion, and how fast that happens sets how long each message lasts.

The chemical step is a delay the brain pays for control.

An electrical contact would be faster, but only a chemical synapse can amplify, invert the sign of a signal, be tuned by neuromodulators and change its strength with experience.

Questions: Why does the brain mostly use chemical synapses when electrical ones are faster? An electrical synapse passes current straight through gap junctions with almost no delay, in both directions, but the receiving cell gets only a fraction of the signal and its sign cannot change. A chemical synapse costs a delay (about 0.15 ms at the fastest mammalian synapses at body temperature, 0.5 to 1 ms in textbook figures) and in exchange it can amplify, since one vesicle opens hundreds of receptors. It can also turn excitation into inhibition by releasing GABA, be tuned by neuromodulators, and change its strength with experience. Electrical synapses are kept for jobs where speed and synchrony matter more than control, such as coupling interneurons into a common rhythm. How could two brain areas use a shared rhythm to talk to each other? In an oscillating group of neurons, inhibition opens and closes in each cycle, so the cells are most excitable during a short window of every cycle. If a sending area's spikes arrive at a receiving area during its excitable window, they have a strong effect; if they arrive during its inhibited phase, they are largely ignored. The communication-through-coherence hypothesis of Fries proposes that areas select whom to listen to by aligning the phase of their rhythms, gamma for feedforward signals and alpha-beta for feedback in the visual system. Too much synchrony would defeat the purpose, since a network in which everything fires together carries little information. What makes a neuromodulator different from a fast transmitter like glutamate? Glutamate and GABA act through ion channels at a specific synapse for milliseconds, so they carry the content of a signal: which cell, when, and with what sign. Neuromodulators such as dopamine, noradrenaline, serotonin and acetylcholine come from small nuclei with widely branching axons, are often released outside synapses, and act mostly through G protein-coupled receptors for seconds to minutes. They change how strongly neurons respond, how likely synapses are to release, and whether coincident activity will be stored. The same molecule can be both (acetylcholine is a fast transmitter at muscle and a modulator in the cortex), so the distinction is about the receptor and the wiring, more than the chemical. Where does the time go between a spike reaching a terminal and the next cell responding? Most of it goes into opening the calcium channels and fusing the vesicle; crossing the 20 nm cleft takes only microseconds, and the receptors begin to open almost as soon as the transmitter lands. At fast synapses in the rat cerebellum at body temperature, fusion lags calcium entry by about 60 microseconds and the postsynaptic current starts about 150 microseconds after the presynaptic spike begins. Textbooks give 0.5 to 1 ms for the synaptic delay, a figure from older and cooler preparations, and the postsynaptic potential then takes another millisecond or more to rise and reach the soma. The delay is short but it adds up: a signal that crosses five synapses has spent several milliseconds in transit at the synapses alone. Where do the brain's 20 watts go? Mostly into pumping ions back after signals. In the budget Attwell and Laughlin built for rodent grey matter, the largest items are the currents through glutamate receptors at synapses and the sodium entry of action potentials, both repaid by the sodium-potassium pump, with smaller shares for keeping the resting potential, recycling transmitter and clearing calcium. Housekeeping that any cell needs (making proteins and lipids, maintaining the cytoskeleton) takes a smaller part. Because demanding tasks change the whole brain's consumption by only a few per cent, most of the 20 W pays for ongoing activity, whatever the person is doing.