How a nerve fires

A nerve cell signals with a brief electrical pulse, the action potential, and every pulse from a given cell is the same size. Press harder on your skin and the pulses in the nerve do not get bigger. They come faster. The whole trick is a threshold. Below it nothing happens, and once it is crossed the membrane runs through the same fixed sequence every time.

What is being claimed

The voltage across a membrane is measured inside relative to outside, in millivolts (mV). OpenStax's Anatomy and Physiology gives the textbook numbers for a neuron. At rest the inside sits at about −70 mV. The exact value varies between cells, and −70 is the one most commonly used. If a stimulus lifts it to about −55 mV, the threshold, the membrane fires. It shoots up to about +30 mV, falls back, dips briefly below rest, and recovers, all in about 2 milliseconds. The book is blunt about the rule. Action potentials are “all or none”. They all peak at the same voltage, and a stronger stimulus starts more of them more quickly, but no single one is bigger.

Why it is worth knowing

OpenStax calls the action potential the basis of how neurons communicate. Because each pulse is the same size, a nerve cannot report “how much” by the size of its pulses. It reports it by how often they arrive. OpenStax spells out what follows: you do not feel greater pain, and a muscle does not contract harder, because of the size of an action potential, since they are not different sizes.

Two kinds of door

An ion is an atom that carries an electric charge. At rest, sodium ions (Na⁺) are about ten times more concentrated outside the cell than inside, and potassium ions (K⁺) are more concentrated inside. Both can cross the membrane only through channels, which are proteins that form pores. Some of those channels are gated by voltage: they open when the inside becomes less negative.

A stimulus lets a little sodium in, and the inside becomes less negative. This is depolarisation. If it gets only partway, to −60 mV say, the membrane simply drifts back to rest. At −55 mV the voltage-gated sodium channels open. Now sodium floods in and makes the inside still more positive, which opens more sodium channels. The loop feeds itself, and it carries the membrane all the way to about +30 mV whatever the stimulus was. That is a 100 mV swing from rest.

Two things end it. Each sodium channel has a second gate that shuts on its own after a fraction of a millisecond. Meanwhile the voltage-gated potassium channels, which respond more slowly, open. Potassium leaves the cell, taking positive charge out, and the voltage falls. The potassium channels are slow to close, so the voltage overshoots below −70 mV before it settles back.

Why a harder push means more spikes

For a moment after a spike the sodium channels' second gate is still shut, and no stimulus can fire the cell. This is the absolute refractory period. A little later a spike is possible but needs a stronger stimulus than usual (the relative refractory period). So a cell that keeps being pushed fires, recovers and fires again, and a harder push reaches threshold sooner after each recovery. Size stays fixed and spacing shrinks. The widget above is a cartoon of this: fixed spike, fixed threshold, a restart from rest after each spike; it leaves out the relative refractory period. Its threshold rule and its fixed spike shape come from the textbook numbers, and its rates are our arithmetic for the cartoon, not measurements. The refractory period is also why a pulse travels only one way down an axon: the patch behind it cannot fire again straight away. The pulse is not current running down a wire. Each patch regenerates it for the next. On an axon wrapped in myelin it is regenerated only at gaps called nodes, and the book calls this jumping saltatory conduction, from saltare, “to leap”.

The sodium–potassium pump is not what makes the spike. The channels do that. The pump uses energy to keep the concentration differences in place over time.

Where the numbers first came from

In 1952 A. L. Hodgkin and A. F. Huxley fitted equations to currents measured across the membrane of the squid's giant nerve fibre. They modelled the membrane as a capacitor in parallel with sodium, potassium and small “leakage” pathways, and then asked whether those equations alone would produce a nerve impulse. They did. Their equations are written for 6.3 °C, and at that temperature the model had a definite threshold. When a push was only just strong enough to fire it, the voltage first sagged to a lowest point about 6 mV above rest before taking off. In the real fibre that point was about 8 mV. Their Fig. 12 shows computed spikes after starting displacements of 90, 15 and 7 mV, and a 6 mV one, just below threshold, that fizzles out. They judged the agreement with recorded spikes good “as regards amplitude, form and time-scale”. Their model also predicted a travelling speed of 18.8 m/s on a fibre whose measured speed was 21.2 m/s.

Two cautions keep their numbers in their own frame. They measured voltage as a displacement from rest, not as the inside-minus-outside figure used above. Their squid numbers are not the textbook −70, −55 and +30. And they counted a depolarisation as negative, so their figures put −V on the vertical axis to make it point upward. Read their raw V as a modern membrane voltage and every sign comes out backwards. They also did not claim to have seen channels. They wrote that their physical interpretation was “unlikely to provide a correct picture of the membrane”. What they showed was that the conductance changes they had measured were enough to account for the spike.

In short

A neuron sits at about −70 mV. Push it to about −55 mV and sodium channels open in a self-feeding rush that carries it to about +30 mV. Then the sodium gates shut and potassium pulls it back, all within about 2 ms. Every spike is the same, so a stronger stimulus shows up only as more spikes per second.