Neuron At Resting Potential

Understanding Membrane Potential Before It Goes Anywhere

Most people think of neurons as either firing or not firing. That binary view gets you through introductory biology, but it breaks down the moment you need to actually measure or model what a neuron is doing. The resting state isn't a stable idle position. It's a dynamic equilibrium held together by ion gradients, leak channels, and a pump that's constantly burning ATP just to keep things from drifting apart. A typical mammalian neuron sits around -70 millivolts inside relative to the outside. That number isn't arbitrary. It comes from the balance between potassium leaking out through always-open K+ channels and the sodium-potassium pump working against concentration gradients. Chloride and other ions play minor roles, but in most standard models, the Goldman-Hodgkin-Katz equation is where you end up when you stop simplifying. The Nernst potential for K+ alone works out to roughly -90 mV. For Na+, it's around +60 mV. Because the membrane is far more permeable to potassium at rest, the actual resting potential lands closer to the K+ equilibrium than anything else. That's why blocking Na+/K+ pumps doesn't instantly collapse the resting potential. It takes several minutes of pump inhibition before you see the membrane drift significantly. The gradients have some buffer built into them.

I ran into this once while setting up whole-cell patch clamp recordings on cultured hippocampal neurons. The overnight bath solution had been sitting too long and the glucose was depleted. By the time I started recording at 3 PM, the resting potentials were slowly depolarizing throughout the session. Not a dramatic shift, just a drift of about 3 to 4 millivolts over forty minutes. My first pass of data looked clean until I compared the holding current traces at the beginning versus the end. The trick was adding fresh glucose to the perfusion chamber mid-recording and only using cells that showed stable baseline for at least five minutes before I locked in. If you're doing similar work, check your bath solution age and temperature. Those two variables alone account for most of the instability people blame on bad seals or damaged pipettes. Another thing beginners consistently get wrong is assuming the resting potential is the same across all neuron types. It isn't. A cortical pyramidal neuron might rest at -65 mV while a spinal motor neuron sits closer to -70. Cerebellar Purkinje cells can be anywhere from -60 to -75 depending on their recent activity history. There's no single resting potential for "a neuron." You have to measure it or look up the specific cell type you're working with.

How to Approach This If You're Modeling It

Start with the basic equivalent circuit. The membrane is a capacitor. Ion channels are resistors in parallel. The Na+/K+ pump acts as a current source. That's the Hodgkin-Huxley simplification. For most purposes, you don't need the full channel dynamics to simulate resting potential. A simple two-conductance model with fixed permeability ratios gets you within 2 mV of experimental measurements for standard neurons. If you're building a simulation from scratch, the main pitfall is neglecting the extracellular space. In tissue, the extracellular potassium concentration isn't perfectly buffered. During sustained activity, [K+]o can climb from 3 mM to 8 or 9 mM, and that alone shifts the resting potential by 10 millivolts or more. For single-cell models in artificial bathing solution, this doesn't matter. For network simulations or realistic in vitro preparations, it does. I learned that the hard way when my simplified model produced unrealistic firing patterns because the potassium accumulation wasn't accounted for. Switching to a compartmental model with explicit extracellular compartments fixed the issue without requiring any additional biological parameters I hadn't already measured. There's also a practical limitation worth noting. The resting potential you calculate from ion concentrations will often disagree with what you measure experimentally. Concentration-based predictions tend to overshoot by 5 to 10 mV in real neurons. The discrepancy comes from fixed intracellular charges, imperfect selectivity in leak channels, and the fact that pumps contribute a small direct electrogenic current beyond just maintaining gradients. Don't treat the Nernst or GHK equations as exact predictions. Treat them as starting points that explain direction and relative magnitude.

If you need a reference for the standard ionic concentrations and permeability ratios used in most textbooks, the classic Hodgkin-Huxley squid axon values are still widely cited, but they apply to an invertebrate preparation. For mammalian neurons, the values in Johnston and Wu's Foundations of Cellular Neurophysiology or Hamill and Marty's patch-clamp manuals are more appropriate. There's no downloadable spreadsheet that replaces knowing which source matches your cell type.