Neurons create electrical signals by exploiting a difference in electrical charge across their outer membrane. At rest, the inside of a neuron sits at roughly −70 millivolts relative to the outside, a state called polarization. When incoming signals push this voltage past a critical threshold, a rapid, self-reinforcing wave of charge reversal fires along the cell, and that wave is the action potential, the fundamental unit of fast communication in the nervous system. The machinery behind this process involves specialized protein channels, energy-hungry pumps, and a delicate ionic balance that the brain spends more than half its fuel budget maintaining.
The Resting State
Before a neuron can send a signal, it has to be primed to send one. That priming is the resting membrane potential, a steady voltage difference created by an uneven distribution of charged atoms (ions) inside and outside the cell. Potassium ions are concentrated inside the neuron, while sodium ions are concentrated outside. This gradient does not happen by accident. A protein embedded in the membrane called the sodium-potassium pump continuously shuffles three sodium ions out of the cell and two potassium ions in for every molecule of ATP it burns.1PubMed. On the concept of resting potential–pumping ratio of the Na⁺/K⁺ pump and concentration ratios of potassium ions outside and inside the cell to sodium ions inside and outside the cell Because more positive charges leave than enter with each cycle, the pump itself contributes a small negative bias to the interior.
The bigger contributor to the resting voltage, though, is a family of potassium channels that stay open all the time. These “leak” channels, known technically as two-pore domain potassium channels, allow potassium to trickle out of the cell down its concentration gradient. As positive potassium ions drift outward, they leave behind negatively charged proteins and other molecules that cannot follow, and the inside of the cell grows more negative.2PubMed. Role of leak potassium channels in pain signaling These leak channels are a major determinant of both the resting potential and how excitable the neuron is. They are active in motor neurons, brainstem respiratory neurons, and many other cell types throughout the nervous system.3PubMed. Pharmacology of neuronal background potassium channels
The result is a cell that sits quietly at around −70 mV, like a coiled spring. The energy stored in those ion gradients is what makes fast signaling possible. Without the pump maintaining the gradient and the leak channels setting the voltage, a neuron would have no charge difference to exploit and no way to fire.
Getting to Threshold
A neuron receives input from other neurons at junctions called synapses, mostly along its branching dendrites. Some inputs are excitatory, nudging the membrane voltage in a positive direction (toward zero), and some are inhibitory, pulling it more negative. The cell constantly adds up these competing influences, a process called synaptic integration.
Not all inputs carry equal weight. Signals arriving at the tips of long dendrites, far from the cell body, might seem like they would fade before reaching the decision point, but research on cortical pyramidal neurons shows the opposite can happen. Distal inputs get amplified with high gain and are integrated over broader time windows, making them surprisingly effective at influencing whether the neuron fires.4PubMed Central. Synaptic integration gradients in single cortical pyramidal cell dendrites Inputs closer to the cell body sum more straightforwardly and require tighter timing to add up. This means a neuron is not just a passive collector of votes; where on its surface a signal arrives changes how much that signal counts.
The critical decision happens at a specialized stretch of the axon near the cell body called the axon initial segment. This region is packed with voltage-gated sodium channels at unusually high density, making it the lowest-threshold trigger zone on the entire cell.5PubMed Central. Electrogenic tuning of the axon initial segment If the combined input from all those synapses depolarizes the membrane at the initial segment past roughly −55 mV, the neuron fires an action potential. If it falls short, nothing happens. This all-or-nothing threshold is one of the defining features of how neurons communicate.
The Action Potential
Once threshold is reached, a chain reaction begins. Voltage-gated sodium channels snap open within a fraction of a millisecond, and sodium ions rush into the cell from the high-concentration exterior. This influx of positive charge drives the membrane voltage sharply upward, from −55 mV toward +30 or +40 mV. The sodium channels are the first to respond to voltage changes, and their opening is what powers the rapid upstroke of the action potential.6StatPearls Publishing. Physiology, Sodium Channels
Each sodium channel has two gates that work on different timescales. One gate opens quickly in response to voltage, letting sodium flow. The second gate closes on a slight delay, shutting the channel regardless of what the first gate is doing. This built-in timer means that each channel is open for only about a millisecond before it inactivates.7Anaesthesia & Intensive Care Medicine. Action potential: generation and propagation That brief window is enough for the next patch of membrane to be depolarized by the inflowing current, so the signal spreads forward along the axon.
As sodium channels inactivate, voltage-gated potassium channels open. Potassium ions flood out of the cell, carrying positive charge back to the exterior and driving the membrane voltage downward again. In some neurons, specific subtypes of these potassium channels, such as Kv7.2/7.3, are responsible for pulling the voltage even below the resting level temporarily, creating a brief undershoot.8Neuron. Delayed Rectification by Kv7 Channels Is a Driver of Striatal Cholinergic Interneuron Pauses The sodium-potassium pump then gradually restores the original ion concentrations, resetting the neuron for the next signal.
Why Signals Travel in One Direction
After a sodium channel inactivates, it cannot reopen immediately. There is a brief window, the absolute refractory period, during which no amount of stimulation can trigger another action potential in that same patch of membrane. This happens because the inactivation gate on the sodium channel needs time to reset.9PubMed Central. Sodium channel slow inactivation normalizes firing in axons with uneven conductance distributions The practical effect is that the signal cannot turn around and travel backward. The membrane behind the advancing action potential is temporarily locked out, so the wave of depolarization moves in only one direction, from the cell body toward the axon terminals.
Following the absolute refractory period comes a relative refractory period, during which the neuron can fire again but only if it receives a stronger-than-normal stimulus. The firing threshold rises temporarily after each spike because a fraction of sodium channels remain inactivated. If the inactivation time constant is short compared to the gap between spikes, this shift acts as a brief relative refractory period; if it is long, it produces spike-frequency adaptation, where a neuron gradually slows its firing rate during sustained stimulation.10PLoS Computational Biology. Impact of Fast Sodium Channel Inactivation on Spike Threshold Dynamics and Synaptic Integration This adaptation prevents neurons from locking into runaway firing and helps encode the intensity of a stimulus rather than simply its presence.
Saltatory Conduction and Speed
In many neurons, the axon is wrapped in a fatty insulating sheath made by glial cells called myelin. The sheath is not continuous. Small gaps called nodes of Ranvier interrupt it at regular intervals, and these nodes are where ion channels cluster. Instead of creeping along the axon point by point, the action potential effectively leaps from one node to the next, a process called saltatory conduction. This dramatically increases the speed of signal transmission.
Recent work has refined our understanding of how this leaping works. Detailed recordings and modeling have revealed a conductive nanoscale space between the axon and its myelin sheath, roughly 12 nanometers wide and about three times more conductive than the core of the axon itself. This periaxonal space creates a kind of parallel electrical pathway, with rapid nodal potentials traveling ahead of slower, attenuated waves under the myelin between nodes.11PubMed Central. Saltatory Conduction along Myelinated Axons Involves a Periaxonal Nanocircuit The picture that emerges is more nuanced than the textbook “jumping from node to node,” involving a double-cable circuit where both the axon interior and the submyelin space contribute to conduction.
The Brain’s Fuel Bill
Maintaining polarization is expensive. Every time sodium enters a neuron during an action potential, the sodium-potassium pump has to push it back out, burning ATP in the process. A classic study on dogs found that sodium pumping accounts for at least 55% of the brain’s total oxygen consumption, and in an awake, active brain the true figure is probably higher.12PubMed Central. Non-signalling energy use in the brain The brain makes up only about 2% of body weight but consumes roughly 20% of the body’s energy at rest, and more than half of that energy goes to keeping ion gradients in place. This is why neurons are so vulnerable to anything that cuts off their fuel supply, even briefly.
When blood flow to a region of the brain is interrupted during a stroke, the sodium-potassium pump begins to fail within minutes. Without the pump restoring ion gradients, sodium and water flood into cells, causing them to swell. Calcium follows, triggering a cascade of toxic events. The failure of ionic homeostasis and the resulting breakdown of polarization is a central mechanism of stroke-related cell death.13PubMed Central. Ionic regulation of cell volume changes and cell death after ischemic stroke This underscores just how fundamental the pump-maintained charge separation is: it is not just for signaling; it is for survival.
When Ion Channels Go Wrong
Because the action potential depends on precisely tuned ion channels opening and closing at the right times, mutations in the genes encoding those channels can cause serious neurological disease. Epilepsy is the most studied example. Genetic studies have identified more than 700 mutations in the genes for voltage-gated sodium channels in patients with epilepsy.14PubMed Central. Role of Sodium Channels in Epilepsy Some of these mutations make channels too easy to open (gain of function), flooding neurons with sodium and causing hyperexcitability. Others make channels harder to open (loss of function), reducing excitability in inhibitory neurons and indirectly allowing seizure activity to spread. The same gene can produce opposite effects depending on the specific mutation.15PubMed Central. Voltage-gated sodium channels in genetic epilepsy: up and down of excitability
This gain-versus-loss complexity makes treatment tricky. A drug that blocks sodium channels might help a patient with a gain-of-function mutation but worsen symptoms in someone whose channels are already underperforming. Genetic testing is increasingly used to identify the specific mutation before choosing a medication, an early example of precision medicine in neurology.
Astrocytes and the Supporting Cast
Neurons do not manage their ionic environment alone. Astrocytes, star-shaped glial cells that surround synapses with fine projections, actively help regulate the concentration of potassium in the extracellular space. When neurons fire repeatedly, potassium accumulates outside the cell. If it builds up too much, the resting potential of nearby neurons shifts toward threshold, potentially causing unwanted firing. Astrocytes soak up this excess potassium using their own pumps and inward-rectifying potassium channels, then redistribute it to regions where concentrations are lower, including the blood vessels wrapped by astrocytic endfeet.16PubMed Central. Potassium buffering in the neurovascular unit: models and sensitivity analysis Modeling work suggests that overly aggressive astrocytic uptake can even cause a transient potassium undershoot in the extracellular space, temporarily making neurons harder to excite than they would be at rest.
This buffering role means that polarization is not purely a neuron-internal affair. The extracellular environment, actively maintained by glia, sets the baseline conditions that determine how easily neurons fire. Diseases that damage astrocytes, or conditions like spreading cortical depression in migraine, involve disruptions to this extracellular potassium balance.
Neurons That Do Not Spike
The action potential gets most of the attention, but not every neuron communicates with all-or-nothing spikes. Some neurons, particularly local interneurons in invertebrate nervous systems, use graded potentials instead. In the locust nervous system, for example, certain interneurons influence the membrane potential of downstream motor neurons without ever producing a spike. When current is injected to depolarize these nonspiking interneurons, the connected motor neurons are smoothly depolarized or hyperpolarized, and the size of the response scales with the amount of input, not with a binary on/off event.17PubMed Central. Graded synaptic transmission between local interneurones and motor neurones in the metathoracic ganglion of the locust
Graded transmission also occurs at the synaptic level even in spiking neurons. Chemical transmitter release can be driven by subthreshold changes in membrane potential, not just by full action potentials.18PubMed Central. Getting Graded: Teaching Principles of Chemical Synaptic Transmission Without Action Potentials The reliability of these graded synapses has been directly measured in insect visual pathways, where they transmit small voltage changes between identified neurons.19PubMed Central. The performance of synapses that convey discrete graded potentials in an insect visual pathway The existence of graded signaling is a reminder that the binary spike is not the only way to use membrane polarization for communication; it is the most prominent one, especially in vertebrate brains, but the underlying principle of charge difference across a membrane is far more versatile.
Dendrites as Computational Devices
Even within spiking neurons, the dendrites are not passive cables just funneling signals to the cell body. They can generate their own local electrical events, called dendritic spikes, that serve a computational purpose. In the visual cortex, dendritic spikes have been shown to sharpen a neuron’s selectivity for the orientation of a visual stimulus. Without dendritic excitability, the neuron responds to a broader range of orientations; with it, responses become more selective and more useful for perception.20PubMed Central. Dendritic spikes enhance stimulus selectivity in cortical neurons in vivo This means that individual neurons are doing some of their processing before the signal ever reaches the axon initial segment, expanding the computational power of each cell beyond what a simple “add up the inputs and fire or don’t” model would predict.
Neuromodulators Reshape Excitability
The polarization state of a neuron is not fixed between action potentials. Chemical signals called neuromodulators, such as acetylcholine, dopamine, and serotonin, can shift the baseline excitability of neurons by opening or closing leak channels. Acetylcholine, for instance, lowers the effective resistance of the membrane by adjusting the balance of non-voltage-gated potassium and sodium leak channels, making a neuron easier to push to threshold.21bioRxiv. Neuromodulators control neuronal dynamics through feature space reshaping This provides the brain with a slower, more diffuse way of tuning neural circuits: rather than flipping individual neurons on and off with synaptic inputs, a neuromodulator can raise or lower the gain on an entire population of cells. It is why your alertness, mood, and attention level change the way your brain processes identical sensory information at different times of day.
Poisons That Exploit the System
The dependence of the action potential on sodium channels makes them a natural target for toxins. Tetrodotoxin, found in pufferfish and several other marine animals, binds directly to voltage-gated sodium channels and physically blocks the flow of sodium ions through the pore. This prevents action potentials from being generated or propagated.22PubMed. Interaction between voltage-gated sodium channels and the neurotoxin, tetrodotoxin The result, at sufficient dose, is paralysis and potentially death by respiratory failure, because the motor neurons that drive breathing simply stop firing. Local anesthetics like lidocaine work on a related principle: they block sodium channels in sensory neurons, preventing pain signals from reaching the brain. The difference between a deadly poison and a useful drug, in this case, is selectivity and dose.
Other toxins target different parts of the signaling chain. Some scorpion venoms prevent sodium channel inactivation, keeping channels locked open so the neuron fires uncontrollably. Certain spider venoms block potassium channels, slowing repolarization and prolonging each action potential. The diversity of natural toxins that have evolved to interfere with ion channels is a testament to how central these proteins are to animal nervous systems, and it has made these toxins invaluable research tools for studying channel function.
Sodium Channels Before Nervous Systems Existed
Voltage-gated sodium channels are so tightly associated with nerve impulses that you might assume they evolved alongside the first nervous systems. They did not. Genomic searches have found sodium-channel-like genes in choanoflagellates, single-celled organisms that are the closest living relatives of animals and have no neurons at all. The channels in these organisms have ion selectivity filters that appear intermediate between calcium and sodium channels, suggesting that sodium channels evolved from an ancestral calcium channel before multicellular animals existed. Similar genes have been found even in placozoans, tiny animals that lack anything resembling a nervous system.23PubMed Central. Evolution of sodium channels predates the origin of nervous systems in animals What the channels were doing in these organisms before nervous systems arose is still debated, but their presence means that the molecular toolkit for electrical signaling was available long before any animal needed to think.