What is Dr. Zubair Khalid's research focus?

Dr. Zubair Khalid specializes in molecular virology, mRNA vaccine development, and computational biology, with a focus on avian pathogens like IBDV and Avian Reovirus.

Where is Dr. Zubair Khalid currently working?

Dr. Zubair Khalid is a Postdoctoral Research Associate at the University of Maryland (UMD), specifically within the Department of Animal and Avian Sciences.

Action Potential: Phases, Ions, and Diagram

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-09-19

An action potential is a rapid, self-propagating wave of electrical activity that travels along the surface of a nerve cell. The definition of an action potential that most students should memorize is this: it is a transient, all-or-none reversal of the membrane potential caused by the voltage-gated opening and closing of ion channels, mainly sodium (Na+) and potassium (K+), that lets a neuron send a reliable signal over long distances [1][2].

That single sentence hides a remarkable piece of engineering. Every thought, every muscle twitch, and every sensation you have ever experienced depends on cells that can flip their internal voltage from negative to positive and back again in about one to two milliseconds. The action potential is the unit of that signaling. Understanding its phases, the ions that drive them, and the way the signal regenerates itself is the foundation of neurophysiology. This guide walks through each phase in order, explains the ion permeability changes that produce them, and covers the refractory periods that follow.

Why the Action Potential Matters

Nerve cells face a physical problem. A signal that starts at one end of an axon, which can be a meter long in a human motor neuron, has to arrive at the other end without fading. Passive electrical spread decays with distance, so a purely passive system would lose the signal. The action potential solves this by regenerating itself at every patch of membrane it crosses. Hodgkin and Huxley built the mathematical model that explains how this works, treating the membrane as a capacitor and ion channels as resistors whose conductance changes with voltage and time [1]. Their framework, linking Nernst's electrochemistry to Kelvin's cable theory, is still the backbone of cellular neuroscience [1].

The action potential is also the reason clinicians care about ion channels. Genetic variants in SCN2A, which encodes the NaV1.2 sodium channel, produce a spectrum of neurodevelopmental disorders, with gain-of-function variants increasing neuronal excitability and loss-of-function variants reducing it [3]. One unusual variant, SCN2A-p.K1422E, alters ion selectivity so the channel becomes permeable to calcium as well as sodium, lowering overall conductance and conferring resistance to tetrodotoxin [3]. These examples show that the action potential is not an abstraction. It is a measurable, modifiable event with direct clinical consequences.

The Five Phases in Order

The action potential unfolds as a fixed sequence. Each phase has a characteristic membrane potential, a dominant ion, and a specific channel state. The table below summarizes the sequence, and the sections that follow explain the mechanism behind each row.

Phase	Dominant ion	Channel state	Membrane potential
Resting potential	K+ (leak)	Voltage-gated Na+ and K+ channels closed	About -70 mV
Threshold	Na+ (early influx)	Some voltage-gated Na+ channels open	About -55 mV
Depolarization	Na+ influx	Voltage-gated Na+ channels open, then inactivate	Rises toward +30 to +40 mV
Repolarization	K+ efflux	Na+ channels inactivated, voltage-gated K+ channels open	Falls back toward -70 mV
Hyperpolarization	K+ efflux	K+ channels slow to close	Dips below -70 mV, then returns

Resting Potential (About -70 mV)

A neuron that is not signaling sits at its resting membrane potential, typically around -70 millivolts. This value is not zero because the membrane is selectively permeable. At rest, the dominant open channels are potassium leak channels, which let K+ move down its concentration gradient out of the cell. The result is a negative interior relative to the outside.

Two features maintain this state. First, the sodium-potassium pump, also called the Na+/K+ ATPase, actively transports three Na+ ions out of the cell for every two K+ ions it brings in, using ATP as fuel. This pump establishes the concentration gradients that the action potential later exploits. Second, the membrane at rest has far more open K+ channels than open Na+ channels, so the resting potential sits close to the potassium equilibrium potential.

Resting potential is not a fixed constant across all neurons. Circadian disruption of voltage-gated ion channels can shift the membrane potential toward -45 mV in some tissues, with neuronal simulations showing a drift from -63 to -50 mV and firing rates rising from 2.5 to 6.7 Hz [4]. That kind of drift matters because it moves a cell closer to threshold.

Threshold (About -55 mV)

Threshold is the voltage at which an action potential becomes inevitable. When the membrane potential reaches about -55 mV, enough voltage-gated sodium channels open to trigger a positive feedback loop. Sodium enters, which depolarizes the membrane further, which opens more sodium channels, which lets in more sodium.

The concept of threshold is where the all-or-none law comes from. Below threshold, a small depolarizing stimulus produces only a graded potential, a local change in voltage that decays with distance and does not regenerate. At or above threshold, the neuron fires a full action potential. There is no such thing as a half action potential in a single neuron. The signal either happens at full amplitude or it does not happen at all.

The all-or-none law applies to the action potential in a single neuron. It does not apply to graded potentials, which are the small, variable, decremental voltage changes produced by synaptic input or sensory stimulation. Graded potentials summate. They can be larger or smaller depending on stimulus strength. They are the analog part of neuronal signaling, while the action potential is the digital part.

Depolarization

Depolarization is the rising phase, and it is driven by sodium influx. Once threshold is crossed, voltage-gated Na+ channels open rapidly. Sodium ions rush into the cell along their electrochemical gradient, carrying positive charge with them. The membrane potential climbs steeply, crossing zero and reaching a peak of roughly +30 to +40 mV.

The channels that carry this current are not all identical. In rat visceral sensory neurons, tetrodotoxin-sensitive (TTX-S) and tetrodotoxin-resistant (TTX-R) sodium channels activate in sequence, producing a biphasic upstroke. The TTX-R channels kick in after the TTX-S channels, and their activation voltage correlates inversely with the maximal upstroke velocity [5]. This kind of fine-tuning shows that depolarization is not a single uniform event in every cell type.

Depolarization ends because the sodium channels inactivate. A short segment of the channel protein acts as a plug that blocks the pore from the inside, a process that happens within a millisecond or two of opening. Inactivated channels cannot reopen until the membrane repolarizes. This inactivation is the molecular basis of the absolute refractory period.

Repolarization

Repolarization is the falling phase, and it is driven by potassium efflux. As sodium channels inactivate, voltage-gated K+ channels open. Potassium ions leave the cell down their concentration gradient, carrying positive charge out and pulling the membrane potential back toward negative values.

The timing matters. Sodium inactivation and potassium activation overlap, which is why the action potential is so brief. If potassium channels opened too late, the cell would stay depolarized. If they opened too early, the peak would be blunted. The Hodgkin-Huxley model captures this interplay with separate gating variables for sodium activation, sodium inactivation, and potassium activation [1][6].

In some neurons, a second wave of sodium current contributes to the repolarization phase. In C-type nodose neurons, reactivation of TTX-R channels generates a repolarization hump, a small secondary depolarization during the falling phase [5]. This is a reminder that the textbook five-phase description is a template, and real neurons add variations.

Hyperpolarization

Hyperpolarization is the undershoot. After the membrane potential returns to about -70 mV, it often dips below that value, sometimes to -75 or -80 mV, before settling back to rest. The cause is that voltage-gated K+ channels are slow to close. They remain open for a few milliseconds after the membrane has repolarized, so potassium continues to leave the cell and the interior becomes more negative than usual.

This undershoot has functional consequences. It makes the neuron temporarily harder to excite, which contributes to the relative refractory period. It also helps prevent runaway excitation by giving the sodium channel inactivation gates time to reset.

The depth and duration of hyperpolarization vary by cell type. Some neurons show a pronounced afterhyperpolarization that shapes their firing pattern, while others recover quickly. The high depolarization baseline offset seen in some pyramidal neurons during trains of action potentials depends on factors including fast potassium delayed rectifier kinetics and appropriate transient sodium current kinetics [7]. In other words, the details of the recovery phase help determine whether a neuron fires slowly or in bursts.

Ion Permeability Changes During the Action Potential

The action potential is best understood as a scheduled change in which ions the membrane is permeable to. The table below lays out the permeability shift phase by phase.

Phase	Na+ permeability	K+ permeability	Net current direction
Resting	Low	Moderate (leak)	K+ outward
Threshold	Rising	Moderate	Na+ inward begins
Depolarization	Very high	Moderate	Na+ inward dominant
Repolarization	Falling (inactivated)	High	K+ outward dominant
Hyperpolarization	Low (reset)	High, then falling	K+ outward continues
Return to rest	Low	Moderate	Na+/K+ ATPase restores gradients

The Na+/K+ ATPase is the long-term reset button. During an action potential, sodium leaks in and potassium leaks out, which slightly degrades the concentration gradients. The pump runs continuously, extruding three sodium ions and importing two potassium ions per ATP hydrolyzed, restoring the gradients so the next action potential can fire. Without the pump, a neuron would eventually lose its ability to signal.

Refractory Periods

After an action potential, the neuron cannot immediately fire another one at full strength. This period of reduced excitability is divided into two phases.

Absolute Refractory Period

During the absolute refractory period, no stimulus of any strength can trigger a second action potential. The reason is sodium channel inactivation. Once the inactivation gate closes, the channel cannot open again until the membrane repolarizes enough to reset the gate. This period lasts roughly one to two milliseconds in typical mammalian neurons.

The absolute refractory period sets the maximum firing rate of a neuron. It also ensures that the action potential travels in one direction. Because the membrane behind the advancing signal is refractory, the signal cannot double back on itself.

Relative Refractory Period

During the relative refractory period, a stronger-than-normal stimulus can trigger an action potential, but the resulting spike is smaller and the threshold is elevated. Two factors cause this. First, some sodium channels are still inactivated. Second, voltage-gated potassium channels are still open, producing the hyperpolarized undershoot. As the potassium channels close and the sodium channels reset, the neuron returns to its normal excitability.

Refractory periods are not just a safety mechanism. They shape how neurons encode information. A neuron that fires at 200 Hz is signaling something different from one firing at 50 Hz, and the refractory period sets the ceiling on that code.

How the Signal Travels: Propagation and Myelin

An action potential does not stay in one place. It propagates along the axon by depolarizing the adjacent patch of membrane to threshold, which triggers a new action potential there, and so on down the length of the cell. This is the self-propagating property that Hodgkin and Huxley described [1].

Conduction speed depends on two main factors: axon diameter and myelination. Myelin is a fatty insulating sheath wrapped around segments of the axon by glial cells. It increases membrane resistance and decreases capacitance, which allows the depolarization to spread passively and quickly between the gaps in the sheath, called nodes of Ranvier. The action potential then regenerates at each node. This jumping pattern is called saltatory conduction, from the Latin saltare, to leap.

Saltatory conduction is much faster than continuous conduction in an unmyelinated axon of the same diameter. It is also more energy-efficient, because the Na+/K+ ATPase only needs to restore gradients at the nodes rather than along the entire membrane. This is why myelinated axons can carry signals at speeds above 100 meters per second in large mammals, while thin unmyelinated fibers conduct at less than 1 meter per second.

Myelin is not just insulation against ion leak. One hypothesis proposed that myelin's main purpose is to insulate against proton permeability, but this idea has been challenged on the grounds that it conflicts with measured chloride gradients and the effects of changing external ion concentrations [8]. The standard explanation, that myelin reduces capacitance and enables saltatory conduction, remains the consensus.

How the Action Potential Is Observed and Tested

Electrophysiologists record action potentials using patch clamp or sharp microelectrode techniques. In current clamp mode, the experimenter injects a controlled current into the cell and records the resulting voltage change. A brief suprathreshold current pulse produces a single action potential. A sustained current produces a train, and the firing rate reveals the refractory period and adaptation properties.

Voltage clamp is the complementary technique. The experimenter holds the membrane potential at a chosen value and measures the current that flows. This is how Hodgkin and Huxley separated the sodium and potassium currents that underlie the action potential [1]. Modern variants include dynamic current clamp, which injects computed currents to test how specific channel populations shape firing [5].

Optical methods add spatial resolution. Voltage-sensitive dyes and genetically encoded voltage indicators let researchers watch action potentials travel along axons and dendrites. Ultrafast sodium and calcium imaging can track ion flux at the axon initial segment, the region where action potentials typically begin [9]. These tools have shown that the axon initial segment is not a passive trigger zone but an active computational compartment.

Clinical recording uses the same principles at a larger scale. Evoked compound action potentials (ECAPs) are the summed electrical activity of many axons firing together. In spinal cord stimulation, ECAP-controlled closed-loop systems use this signal as real-time feedback to adjust stimulation delivery, with one case report describing approximately 24.2 million automated adjustments over a five-day trial [10]. This is a direct clinical application of the physiology described in this guide.

Comparative and Clinical Relevance

The action potential is conserved across animals, but the details differ. Zebrafish ventricular cardiomyocytes have action potentials that resemble human ones, and a mathematical model of the zebrafish cardiomyocyte has highlighted that T-type calcium current contributes to the upstroke while L-type calcium current shapes the plateau [11]. These comparative models help researchers study cardiac electrophysiology in a tractable organism.

In the heart, action potential duration is a critical variable. Drugs that prolong action potential duration by blocking a single potassium channel can increase the dispersion of repolarization, which raises the risk of arrhythmia. A modeling study identified strategies to prolong action potential duration without increasing transmural dispersion, including blocking multiple depolarizing and repolarizing channels together [12]. This illustrates a general principle: the action potential is a system property, and perturbing one component can have non-obvious effects on the whole.

Genetic channelopathies provide the clearest clinical examples. Long QT syndrome types 1 and 2 arise from impaired delayed rectifier potassium channels, and beta-adrenergic stimulation shortens action potential duration and effective refractory period in these conditions [13]. In the nervous system, SCN2A variants alter sodium channel permeability and produce developmental delay, infantile spasms, and features of autism spectrum disorder [3]. Each of these conditions traces back to a specific ion channel and a specific phase of the action potential.

Common Mistakes and Limitations

Students and researchers make predictable errors when reasoning about action potentials. Here are the ones worth watching for.

Confusing threshold with the peak. Threshold is the voltage at which the action potential begins, about -55 mV. The peak is the highest voltage reached, about +30 to +40 mV. They are different numbers with different mechanisms.

Thinking the sodium-potassium pump generates the action potential. The pump maintains the gradients that make the action potential possible, but it does not produce the rapid voltage changes. Those come from passive ion flow through voltage-gated channels. The pump works on a timescale of seconds to minutes, while the action potential lasts milliseconds.

Applying the all-or-none law to graded potentials. Graded potentials are variable and decremental. They are not all-or-none. Only the action potential, once threshold is crossed, is all-or-none in a single neuron.

Assuming all sodium channels are the same. TTX-S and TTX-R channels have different kinetics and activation voltages, and they contribute differently to the upstroke and to repetitive firing [5]. In nociceptors, Nav1.8 channels carry the majority of sodium entry during action potentials and nearly all of the sodium current supporting repetitive firing, even though Nav1.7 channels activate faster [14].

Ignoring the role of potassium channel diversity. Not all potassium channels are the same. K2P channels help maintain resting potential and can be modulated by pH, stretch, and pharmacology [15]. Delayed rectifier channels drive repolarization. The specific complement of channels in a cell determines its firing behavior.

Overgeneralizing from one cell type. The five-phase template is a starting point. Cardiac action potentials have a plateau phase that neuronal action potentials lack. Some neurons have biphasic upstrokes. Some show pronounced afterhyperpolarization. Always check the specific cell type.

A limitation of any general guide is that individual neurons and individual patients vary. A physician or veterinarian should evaluate any clinical concern about nerve or muscle function, because the underlying cause could be a channelopathy, a toxin, a metabolic problem, or something else entirely.

Quick Review

Resting potential is about -70 mV, maintained by potassium leak channels and the Na+/K+ ATPase.
Threshold is about -55 mV. Crossing it triggers the all-or-none action potential.
Depolarization is driven by Na+ influx through voltage-gated sodium channels.
Repolarization is driven by K+ efflux through voltage-gated potassium channels.
Hyperpolarization is the undershoot caused by slow potassium channel closure.
The absolute refractory period is set by sodium channel inactivation. The relative refractory period is set by potassium channels still being open.
Myelin speeds conduction by enabling saltatory conduction, in which the action potential jumps between nodes of Ranvier.

Frequently Asked Questions

What is the definition of an action potential?

An action potential is a rapid, self-propagating, all-or-none reversal of the membrane potential that travels along the surface of a nerve or muscle cell. It is produced by the voltage-gated opening and closing of sodium and potassium channels [1][2].

What are the five phases of an action potential in order?

The five phases are resting potential (about -70 mV), threshold (about -55 mV), depolarization, repolarization, and hyperpolarization. Some descriptions fold threshold into the depolarization phase, but treating it separately makes the all-or-none property clearer.

Which ion drives depolarization?

Sodium drives depolarization. Voltage-gated Na+ channels open when threshold is reached, and sodium flows into the cell along its electrochemical gradient, carrying positive charge and raising the membrane potential toward +30 to +40 mV [1].

Which ion drives repolarization?

Potassium drives repolarization. Voltage-gated K+ channels open as sodium channels inactivate, and potassium flows out of the cell, returning the membrane potential toward negative values [1].

What causes the refractory period?

The absolute refractory period is caused by sodium channel inactivation, which prevents the channels from reopening. The relative refractory period is caused by potassium channels remaining open, which keeps the membrane hyperpolarized and raises the threshold for a new spike.

Does the all-or-none law apply to all electrical signals in neurons?

No. The all-or-none law applies to the action potential in a single neuron. Graded potentials, which are local voltage changes produced by synaptic input or sensory stimuli, vary in amplitude and decay with distance. They do not follow the all-or-none rule.