Neurotransmitters

What a neurotransmitter is

A neurotransmitter is a molecule that a neuron uses to pass a message to another cell. When an electrical impulse reaches the end of an axon, it cannot leap the gap to the next neuron directly. Instead it triggers the release of a chemical stored there in tiny membrane sacs called synaptic vesicles. That chemical drifts across the narrow synaptic cleft and binds to receptors on the receiving cell, changing its behaviour. The neurotransmitter is, in short, the physical carrier of the signal at the point where one cell hands off to the next.

These molecules are chemically ordinary. Many are amino acids or are built in a step or two from amino acids you eat; others are short chains of amino acids no different in principle from small proteins. What makes them special is not the chemistry itself but the machinery the neuron wraps around them: the enzymes that make them, the vesicles that store them, the triggered release, the specific receptors, and the systems that clear them away. It is that whole apparatus, not the molecule alone, that turns an ordinary chemical into a signal.

The criteria for one

Neuroscience does not call every chemical found in the brain a neurotransmitter. To earn the name, a molecule is traditionally expected to meet a set of criteria. It is worth being explicit about why the criteria exist at all, because they are usually presented as a checklist to memorise rather than as an argument.

The problem they solve is this. Brain tissue is full of chemicals. Glutamate is a neurotransmitter, but glutamate is also an ordinary metabolic amino acid found in every cell in the body, including cells that signal to nobody. Merely finding a molecule in the brain, even finding a lot of it, tells you nothing about whether it carries information. The criteria are therefore a test designed to separate a molecule that carries a signal from one that is merely present in the tissue. Each criterion asks for one piece of the apparatus that a message requires: a sender that can produce it, a trigger that releases it on cue, a receiver that reads it, and an off switch so that the message ends and a second message can be distinguished from the first. A molecule that fails any of them may still be biologically important. It simply is not talking.

1. It is made in the neuron

   The molecule, or the enzymes needed to make it, must be present in the presynaptic neuron. The cell must be able to produce its own supply of the transmitter.

2. It is stored and released

   The transmitter is stored, usually in vesicles, and released into the synaptic cleft when the neuron is stimulated, in amounts that depend on the incoming activity.

3. It acts on the target cell

   When applied to the target, the molecule reproduces the effect of the natural signal, acting through specific receptors on the receiving cell.

4. There is a way to end its action

   A mechanism exists to remove or inactivate the transmitter, by reuptake into cells or by enzymes that break it down, so the signal is brief and controllable.

The exception that proves the criteria are a description, not a law

The usual treatment mentions nitric oxide in a sentence and moves on. That is a mistake, because nitric oxide is not a marginal case that scrapes past the rules. It breaks nearly every one of them, and it is unambiguously a signal, and what it demolishes is the assumption hidden inside the checklist: that there is only one way for one neuron to speak to another.

Nitric oxide is a gas, and the consequences follow from that single fact.

1. It is not stored, and it is not released

   You cannot bottle a gas in a membrane sac. Nitric oxide is not packed into vesicles and does not leave by exocytosis. It is manufactured on demand, by an enzyme, at the moment it is needed, and it leaves by simply diffusing away. There is nothing to trigger and nothing to spill.

2. It does not bind a surface receptor

   Being small and lipid-soluble, it passes straight through cell membranes rather than docking on the outside of one. It acts on an enzyme inside the target cell, guanylyl cyclase, producing a second messenger there directly. The membrane, which is the whole basis of the receptor concept, is simply not an obstacle to it.

3. It needs no transporter and no clearing enzyme

   Nothing pumps it back in and nothing has to break it down on cue. It is chemically reactive and short-lived, so its signal ends on its own, within seconds, by chemistry rather than by machinery. The off switch is built into the molecule.

4. It travels backwards

   This is the one that matters. Nitric oxide is made in the postsynaptic cell, the one receiving the message, and it diffuses back across the cleft to the presynaptic terminal that sent it. The arrow of the synapse is reversed.

So the right conclusion is not that nitric oxide is an odd footnote to the criteria. It is that the criteria describe one mode of chemical signalling, the vesicular, receptor-mediated, forward-travelling mode, which happens to be the commonest and the best understood. It is not the only one. Gases signal by diffusion, neuropeptides signal without reuptake, and modulators released by volume transmission are not addressed to any particular cell at all. The checklist is a good description of the classical synapse and a poor definition of chemical communication.

The main chemical classes

Neurotransmitters are grouped first by size. On one side sit the small-molecule transmitters, compact chemicals made locally and recycled quickly. On the other sit the neuropeptides, larger chains of amino acids that mostly fine-tune rather than command. The small-molecule group is then split further by chemical family.

The two groups differ in how they are handled as much as in their size. Small-molecule transmitters are made in the axon terminal and recycled on the spot, ready to fire again in milliseconds. Neuropeptides are manufactured in the cell body, shipped down the axon, and, once used, are not recovered but broken down, so they act more sparingly and take longer to replenish.

A guide to the major ones

A small number of transmitters account for most of what the brain does. The table below is a spine: it names each major molecule, its class, its principal role, and whether its usual effect is to excite or to inhibit the target cell. Read the last column with care. Excitatory and inhibitory are properties of the receptor, not of the molecule, and the section on what makes a receptor excitatory or inhibitory gives the rule that decides it. The word "usual" in the heading is doing real work.

The three most talked-about transmitters deserve a closer look, because each is more subtle than its popular reputation suggests.

Synthesis, release, and clearance

Every transmitter runs through the same life cycle: it is built, packed away, released on cue, and then cleared so the synapse can reset. This cycle is where much of pharmacology acts, because a drug that changes any one step changes the signal.

1. Synthesis

   Enzymes assemble the transmitter from precursors. Dopamine, for example, is made from the amino acid tyrosine in two steps; serotonin from tryptophan. Peptides are instead read out from genes and cut from larger proteins.

2. Storage

   The finished transmitter is pumped into synaptic vesicles, small membrane sacs clustered near the release site. Packaging protects the molecule and readies a concentrated dose for instant release.

3. Release

   When an impulse arrives, calcium floods into the terminal and vesicles fuse with the membrane, spilling their contents into the cleft. This step is fast, calcium-dependent, and finely graded to the incoming activity.

4. Receptor binding

   The transmitter crosses the cleft and binds receptors on the target cell, opening channels or triggering signals inside it. This is the moment the message is read.

5. Clearance

   The signal is ended in one of two ways. In reuptake, transporter proteins pull the transmitter back into cells for reuse, as happens with serotonin and dopamine. Alternatively, enzymes break it down: acetylcholinesterase, for instance, splits acetylcholine in the cleft within milliseconds.

Clearance matters as much as release. Without a way to switch the signal off, a transmitter would linger and the message would blur. Many well-known drugs work precisely here: they block reuptake or inhibit the clearing enzymes, so the transmitter stays in the cleft longer and its effect is amplified.

Co-transmission: one terminal, more than one message

The account so far has assumed one terminal, one transmitter. That assumption is a simplification, and dropping it changes the picture in a way worth having.

A typical axon terminal contains two kinds of vesicle. There are the small, clear vesicles clustered tight against the release site, packed with a small-molecule transmitter such as glutamate, GABA, or acetylcholine. And there are large, dense-core vesicles, holding a neuropeptide, sitting further back from the membrane. Both are in the same terminal, in the same neuron, waiting on the same arriving impulse. But they do not answer to it in the same way.

The consequence is that the message a neuron sends is not fixed. It changes with how hard the neuron is working. Fire it gently and it releases its fast transmitter alone, and the target cell receives a brief, ordinary synaptic signal. Drive it hard and it releases the fast transmitter and a peptide, and the peptide, acting on metabotropic receptors through the slow amplifying cascade described below, produces a longer, deeper change in the target. The same synapse says one thing when whispering and another thing when shouting, and the two things are not merely different in volume. They are chemically different messages read by different receptors on different timescales.

This makes firing rate a genuine variable in the content of a signal, not just its strength, and it is one reason the older textbook principle that each neuron releases a single transmitter is no longer taught as a rule. The rule was drawn from the commonest case. Co-transmission is the general one.

Receptor types

A transmitter has no effect until it meets a receptor, and the receptor, not the transmitter, decides what happens next. The same molecule can excite one cell and inhibit another simply because they carry different receptors. Broadly, receptors fall into two families that differ in speed and mechanism.

Most transmitters act on both kinds. Glutamate, for instance, opens fast ionotropic channels to carry moment-to-moment signals, but also engages metabotropic receptors that tune how the synapse responds over time, a mechanism central to learning. Acetylcholine has fast ionotropic receptors (the nicotinic ones) and slow metabotropic receptors (the muscarinic ones). The pairing of a fast channel with a slow modulator lets one molecule do two jobs at once.

Inside the metabotropic cascade

To say that a metabotropic receptor "triggers a cascade" and stop there is to leave the most important machinery in the brain as a black box. The cascade has named parts, and once you can name them, two properties of metabotropic signalling that otherwise look like arbitrary facts, that it is slow and that it is powerful, follow directly from the mechanism.

Here is the chain, in order.

1. The transmitter binds

   The transmitter docks on the outside of the receptor. The receptor changes shape, and that shape change is transmitted through the membrane to its inner face, where a G protein is waiting.

2. The G protein is switched on

   The activated receptor makes the G protein drop its GDP and pick up a GTP. That is the switch thrown from OFF to ON. The G protein then splits into subunits that come free of the receptor and move along the inner face of the membrane.

3. An enzyme is activated

   A liberated subunit binds an effector enzyme, most famously adenylyl cyclase, and turns it on. Note what has just happened: a single receptor has activated an enzyme, and an enzyme is a catalyst.

4. The enzyme manufactures many second messengers

   Adenylyl cyclase does not make one molecule of cyclic AMP. It makes cyclic AMP continuously for as long as it stays switched on, filling the local cytoplasm with the second messenger. One enzyme, many hundreds of product molecules.

5. The second messenger activates a kinase

   Cyclic AMP binds and activates protein kinase A. A kinase is an enzyme that attaches a phosphate group to another protein, and attaching a phosphate changes that protein's shape, and therefore its behaviour. Each kinase molecule, once switched on, can phosphorylate target after target.

6. The kinase rewrites the cell

   The kinase phosphorylates ion channels, changing how readily they open; it phosphorylates receptors, changing how strongly the cell listens; it phosphorylates enzymes, changing what the cell makes. And if the cascade runs far enough, it phosphorylates transcription factors in the nucleus, changing which genes the cell reads out.

Now the two consequences, which are the whole point of following the chain.

The amplification also explains why cascades are the favourite target of the drugs, toxins, and hormones that produce large effects from tiny doses. When a system multiplies, a small nudge at the top is a large shove at the bottom. For what this machinery does at the scale of whole circuits, see neuromodulation.

What makes a receptor excitatory or inhibitory

The table above quietly assigns each transmitter a "usual effect", and the caveat attached to it, that the sign is a property of the receptor rather than the molecule, is true but useless until you say which property of the receptor decides it. The answer is short, and it is entirely mechanical.

An ionotropic receptor is a channel, and a channel is selective: it passes some ions and not others. Whether the cell is pushed towards firing or away from it depends on nothing more than which ion the channel it opens will let through.

Acetylcholine makes the point better than any argument, because the same molecule does opposite things in two tissues. At the neuromuscular junction it binds a nicotinic receptor, which is an ionotropic channel that passes sodium; the muscle fibre depolarises and contracts. At the heart it binds a muscarinic receptor, which is metabotropic and whose G protein opens a potassium channel; potassium leaves, the pacemaker cells are held further from threshold, and the heart slows. One transmitter, two receptors, two ions, two opposite effects. The molecule carries no sign of its own. It only ever asks a question, and the receptor decides what the answer will be.

Metabotropic receptors reach the same fork by the longer route described above. A cascade can end by phosphorylating a potassium channel open (inhibitory) or by phosphorylating a potassium channel closed (excitatory, since blocking the outward leak of positive charge lets the cell drift upward). The principle is unchanged: follow the ion, and you have the sign.

One chemical, one feeling?

It is tempting to file each transmitter under a single mood: dopamine for pleasure, serotonin for happiness, adrenaline for fear. This shorthand is memorable, and it is wrong. A transmitter is not a feeling; it is a molecule released at particular synapses, and its meaning depends entirely on where it is released and which receptor reads it.

This is why a biochemical account, the kind on this page, stops at the molecule and its machinery. To understand what a transmitter does for behaviour, you have to follow it into the specific circuits that use it, which is the province of systems neuroscience rather than chemistry alone.

Related topics