That moment when you touch a hot stove and yank your hand back before you even feel the burn? Even so, that's not reflex. That's a race — and the finish line is a very specific number Less friction, more output..
Most people think neurons fire whenever they feel like it. In practice, every single action potential in your body — every thought, every movement, every heartbeat signal — starts the same way: a tiny electrical push that either clears the bar or falls short. No "almost.But they don't. No participation trophies. " The membrane potential hits threshold or it doesn't That alone is useful..
Here's what nobody tells you in intro biology: that bar isn't fixed.
What Is the Threshold of Excitation
The threshold of excitation — also called the firing threshold or simply threshold — is the critical membrane potential a neuron must reach to trigger an action potential. In most mammalian neurons, that number sits around -55 millivolts relative to the resting potential of -70 mV Surprisingly effective..
At its core, where a lot of people lose the thread.
But "around" is doing a lot of work there Most people skip this — try not to..
Threshold isn't a single universal constant like the speed of light. It's a property of the specific neuron, its ion channel composition, its recent history, and even its temperature. Worth adding: a cortical pyramidal cell might fire at -52 mV. Here's the thing — a cerebellar Purkinje cell might wait until -48 mV. The same neuron might shift its threshold by several millivolts depending on what happened five milliseconds ago.
The voltage-gated sodium channel is the gatekeeper
Here's the mechanism in plain English: your neuron's membrane is studded with voltage-gated sodium channels. At rest, they're closed. As the membrane depolarizes — gets less negative — these channels start to twitch open. A few open. Sodium rushes in. In practice, that makes the membrane more positive. Plus, which opens more channels. Which brings in more sodium Most people skip this — try not to..
Positive feedback loop. Boom. Action potential Most people skip this — try not to..
But the loop only engages if enough channels open fast enough to overcome the leak currents trying to drag the membrane back to rest. That tipping point? That's threshold.
Resting potential vs. threshold — the gap matters
Resting potential is typically -70 mV. Threshold is typically -55 mV. Here's the thing — that's a 15-millivolt gap. Doesn't sound like much. But in the world of ion channels, 15 mV is the difference between "nothing happens" and "the signal propagates down a meter of axon at 100 meters per second And that's really what it comes down to..
The size of that gap — the threshold distance — determines how excitable the neuron is. Smaller gap = easier to fire. In real terms, larger gap = harder to fire. Neuromodulators like acetylcholine or norepinephrine can shrink that gap. Anesthetics widen it Less friction, more output..
Why It Matters / Why People Care
If you're a neuroscientist, threshold is where the magic lives. If you're a clinician, it's where the pathology hides. If you're an engineer building brain-computer interfaces, it's the spec you design around Not complicated — just consistent. But it adds up..
Excitability is tunable — and that's the point
Neurons aren't binary switches. In real terms, they're gain-adjustable amplifiers. The threshold mechanism lets the nervous system dial sensitivity up or down without rewiring Simple, but easy to overlook..
- Sensory adaptation: Your photoreceptors raise their threshold in bright light so you're not blinded. They lower it in darkness so you can see stars.
- Learning and memory: Long-term potentiation (LTP) often works by lowering the threshold for future firing. The synapse gets "easier to trigger."
- Pain sensitization: After tissue injury, inflammatory soup (prostaglandins, bradykinin, NGF) lowers the threshold of nociceptors. Light touch becomes agony. That's peripheral sensitization.
- Epilepsy: Pathologically low threshold in cortical networks. A tiny spark becomes a seizure.
The all-or-none law — and its loopholes
You've heard it: action potentials are all-or-none. Once threshold is crossed, the spike runs to completion. Amplitude doesn't code for stimulus strength. Frequency does.
But threshold itself? Threshold is analog.
A neuron receiving two subthreshold inputs 2 milliseconds apart might fire. The same inputs 20 milliseconds apart? Here's the thing — the membrane potential sums them — temporal summation. Two inputs on nearby dendrites? Spatial summation. Nothing. The threshold is the same. The probability of reaching it changes.
This is how the nervous system computes. Which means not in the spike. In the decision to spike.
How It Works (or How to Do It)
Let's walk through the biophysics. Not the textbook cartoon — the actual moving parts Which is the point..
The players: Nav channels, leak currents, and capacitance
Voltage-gated sodium channels (Nav) are the engine. They have three states: closed (resting), open (conducting), and inactivated (refractory). The transition from closed to open is voltage-dependent. The transition from open to inactivated is time-dependent.
Leak potassium channels are the brakes. They're always open. They pull the membrane toward -90 mV (EK). At rest, they balance the tiny sodium leak. During a depolarizing input, they fight the Nav channels.
Membrane capacitance is the inertia. The membrane is a capacitor. You can't change its voltage instantly — you have to charge it. Current = capacitance × dV/dt. A larger neuron (more membrane area) has more capacitance. Needs more current to reach threshold at the same speed That's the part that actually makes a difference..
The threshold equation (simplified)
Threshold occurs when:
Inward Na+ current > Outward leak current + Capacitive current
Or more precisely, when the net inward current becomes regenerative — when each additional open Nav channel recruits more than one additional channel on average It's one of those things that adds up..
Mathematically, this is where the slope conductance of the I-V curve goes negative. But you don't need the math. You need the intuition:
- More Nav channels → lower threshold (easier to fire)
- More leak channels → higher threshold (harder to fire)
- Larger capacitance → slower rise, but same threshold voltage
- Higher temperature → faster channel kinetics → slightly lower effective threshold
Axon initial segment: the real trigger zone
Here's what textbooks often gloss over: the action potential doesn't usually start at the soma. It starts at the axon initial segment (AIS) — a specialized 20-60 μm zone right after the axon hillock Practical, not theoretical..
The AIS has 3-5x the Nav channel density of the soma. Which means they're just voting. Think about it: " All those synaptic inputs on dendrites and soma? Faster kinetics. It's the neuron's "decision point.Worth adding: lower threshold. The AIS casts the final ballot.
And the AIS isn't fixed. Chronic depolarization shifts it farther from the soma. Still, it can move. Day to day, the neuron literally rewires its own trigger zone to maintain stable firing rates. Chronic silencing shifts it closer. Homeostatic plasticity in action Simple as that..
Myelination changes the game
In myelinated axons, the action potential jumps between nodes of Ranvier. Each node is packed with Nav channels. The internodal membrane is insulated — high resistance, low capacitance.
Threshold at the node? Here's the thing — myelination doesn't change threshold voltage. Still ~-55 mV. But the current required to reach it drops dramatically because the capacitive load is tiny. It changes threshold current It's one of those things that adds up..
This is why demyelinating diseases (MS, Guillain-Barré) cause conduction block. The voltage threshold is fine. The current available at the next node isn't enough to charge the exposed membrane
— and the spike fails to propagate.
What makes this failure insidious is that it is not an all-or-none loss from the start. As myelin wears away, the first symptom is often delay: the capacitive load at the node creeps up, the local depolarization arrives late, and the timing code the brain relies on begins to smear. That said, only later, when enough charge leaks sideways through the exposed axolemma, does the inward Na⁺ current fall below the outward leak plus capacitive demand, and the axon goes silent. The threshold voltage never moved. The neuron simply ran out of current.
No fluff here — just what actually works.
Why threshold is a process, not a point
We speak of "threshold" as if it were a line drawn on the membrane. It is not. It is an emergent property of competing conductances, geometry, and history. Here's the thing — a neuron that fired a second ago has Nav channels still partially inactivated — its threshold is temporarily higher. In real terms, a neuron held near rest for minutes has Nav channels fully available and may fire to the slightest whisper of input. Threshold is therefore a state, negotiated continuously by the cell.
This reframing matters clinically and computationally. Because of that, drugs that mildly block Nav channels do not erase action potentials; they raise threshold and slow kinetics, preferentially silencing hyperexcitable cells while sparing quiet ones. Neuromodulators that shift AIS position or leak conductance retune the entire input–output curve without altering a single synapse No workaround needed..
Conclusion
The action potential threshold is not a fixed voltage etched into neural tissue. It is the moment when inward regenerative current overtakes outward and capacitive loss — a balance set by channel density, membrane area, temperature, and the cell's own recent behavior. And the axon initial segment acts as the privileged arena where this balance is struck, and myelination determines how cheaply that balance can be reached. To understand when a neuron fires, we must stop asking "what is the threshold?" and start asking "what is the threshold right now, for this cell, given its history?" That question — not a number on a graph — is the true gateway to the spike.
