Ion Channels and Selective Permeability

Ion channels sit at the center of modern pharmacology, and the selectivity-filter chemistry described here is the direct molecular target of major drug classes. Local anesthetics (lidocaine, procaine) block voltage-gated Na⁺ channels from the intracellular side, preventing the depolarizing upstroke of action potentials in pain-sensing neurons. Class III antiarrhythmics (amiodarone, sotalol) block cardiac K⁺ channels (hERG/Kv11.1) to prolong the cardiac action potential — the very same hERG channel is also the single most important off-target liability screened in early-stage drug development, because inadvertent hERG block by unrelated drugs causes long-QT syndrome and fatal arrhythmia; this is why virtually every new small-molecule drug candidate is now tested computationally and experimentally against hERG before advancing to clinical trials. Sulfonylurea drugs for type 2 diabetes (glibenclamide) close ATP-sensitive K⁺ channels (K_ATP) in pancreatic β-cells, depolarizing the cell and triggering insulin release — a direct pharmacological exploitation of ligand-gated channel chemistry. Cystic fibrosis is a channelopathy of a Cl⁻ channel (CFTR), and modern CFTR modulator drugs (ivacaftor, lumacaftor, elexacaftor — the "Trikafta" combination) work by directly correcting the folding and gating defects caused by the ΔF508 mutation, an achievement built on decades of structural and gating-kinetics knowledge of the exact type described here. Local snake and scorpion toxins (tetrodotoxin, saxitoxin, charybdotoxin) are classic pharmacological tools that plug selectivity filters or vestibules with near-absolute specificity, and remain indispensable for isolating specific channel subtypes in research. Mechanically gated Piezo1/Piezo2 channels, discovered only in 2010, now anchor an entire subfield connecting ion channel gating to touch, proprioception, and vascular blood-flow sensing — a frontier that connects directly outward to Signal Transduction at the Membrane, where channel opening is often just the first electrical step in a much longer intracellular signaling cascade.

Ask the tutor: How does the voltage-sensor "paddle" motion in Kv channels mechanically couple to opening the pore's bundle crossing?

An ion channel is fundamentally different from a transporter. A transporter (uniporter, symporter, antiporter, or ATPase pump) never opens a continuous path across the membrane — it alternates between conformations that expose a binding site to one side or the other, a mechanism that inherently limits turnover to roughly 10²–10⁴ ions per second. A channel, by contrast, forms a continuous water-filled pore that spans the bilayer whenever it is in the open state, so ions move by essentially unassisted electrodiffusion through that pore. This is why channel conductances (10⁶–10⁸ ions/s, i.e., picosiemens-to-nanosiemens single-channel conductances) exceed transporter turnover by four to five orders of magnitude. But an open water column, if left unregulated, would let any small ion or even water itself flood through in an undirected way — useless for signaling, and lethal if it dissipated gradients built at high metabolic cost. Two structural devices fix this: a gate, typically a bundle-crossing of transmembrane helices near the cytoplasmic or extracellular end, which physically occludes the pore in the resting/closed state, and a selectivity filter, a short, narrow segment (often just 12–15 Å long) near the extracellular mouth that chemically discriminates among ion species regardless of gate state. The selectivity filter of K⁺ channels (KcsA, Kv, Kir — all share the signature sequence T-V-G-Y-G) is the best-understood example. Four identical filter loops from the channel's four subunits present their backbone carbonyl oxygens toward the pore axis, creating a stack of four oxygen "cages," each an approximate cube of eight carbonyl oxygens (two per subunit, four subunits) capable of coordinating a dehydrated cation exactly as water molecules would in bulk solution. The K–O distances at these sites (~2.8 Å) closely match the K⁺–water oxygen distance in the hydration shell. Na⁺, being smaller (ionic radius 0.95 Å vs. K⁺'s 1.33 Å), cannot be coordinated at this same distance without an energetically costly filter collapse, so Na⁺ effectively finds no good replacement for the water it must shed — this is the physical root of selectivity.

Once inside the filter, ions do not move one at a time through an empty tube — they move as a queue. Crystallographic and molecular-dynamics evidence supports a "knock-on" mechanism: two or three K⁺ ions occupy the four filter sites simultaneously, separated by water molecules, in a roughly 1-3 or 2-4 alternating occupancy pattern. Electrostatic repulsion between adjacent ions is what actually drives conduction — when a new ion enters site S4 from the cytoplasmic cavity, its charge repels the ion queue ahead of it, "knocking" the ion at S1 out into the extracellular vestibule. This many-body mechanism explains why K⁺ channels can be both extremely selective (because each ion is still individually coordinated ideally at each site) and extremely fast (because conduction is driven by electrostatic repulsion, not by waiting for a single ion to diffuse the entire filter length).

The driving force on any permeant ion is the difference between the transmembrane voltage and that ion's own equilibrium (Nernst) potential. For a single ion species at equilibrium — zero net flux — the Nernst equation gives the potential at which electrical and chemical driving forces exactly balance:

where:

• $E_X$ — equilibrium (Nernst/reversal) potential for ion X (volts)
• $R$ — gas constant, 8.314 J mol⁻¹ K⁻¹
• $T$ — absolute temperature (K)
• $z$ — ion valence (signed: +1 for K⁺/Na⁺, +2 for Ca²⁺, −1 for Cl⁻)
• $F$ — Faraday constant, 96,485 C mol⁻¹
• $[X]_{\text{out}}, [X]_{\text{in}}$ — extracellular and intracellular concentrations

Reading: $E_X$ is the membrane voltage at which an open channel selective only for X carries zero net current — the point where the electrical push exactly cancels the chemical (diffusive) push from the concentration gradient. Any membrane potential away from $E_X$ drives net flux of X in the direction that pushes $V_m$ toward $E_X$. When a membrane is permeable to several ion species simultaneously (as real membranes always are), the Goldman–Hodgkin–Katz (GHK) voltage equation gives the actual resting or steady-state membrane potential as a permeability-weighted average of the individual Nernst potentials for monovalent ions:

where:

• $P_X$ — relative membrane permeability to ion X (proportional to the number and open probability of channels selective for X)
• Anion concentrations appear inverted (in/out swapped) relative to cations, because $z=-1$ flips the sign inside the log

Reading: $V_m$ is not any single ion's equilibrium potential — it is dragged toward whichever ion's permeability dominates. A resting neuron with $P_K \gg P_{Na}$ sits close to $E_K$ (≈ −90 mV); when voltage-gated Na⁺ channels open during an action potential, $P_{Na}$ transiently dominates and $V_m$ swings toward $E_{Na}$ (≈ +60 mV). Selectivity itself — why $P_K/P_{Na}$ can reach 1000:1 in K⁺ channels — is a thermodynamic statement about the free energy of dehydration and rebinding. The overall equilibrium for an ion entering the filter can be written as a thermodynamic cycle:

where:

• $\Delta G_{\text{dehydration}}$ — free energy cost of stripping the ion's hydration shell (large and unfavorable, but nearly equal for K⁺ and Na⁺ in magnitude terms relative to their sizes)
• $\Delta G_{\text{binding}}$ — free energy gained from carbonyl oxygen coordination in the filter (favorable, strongly dependent on ion–oxygen distance matching)

Reading: selectivity emerges only because the filter geometry makes $\Delta G_{\text{binding}}$ for K⁺ almost exactly cancel $\Delta G_{\text{dehydration}}$ for K⁺, while for Na⁺ the filter cannot approach closely enough to fully compensate — leaving a large residual unfavorable $\Delta G_{\text{filter}}$ for sodium.

Example 1 — Foundational: Nernst potentials for K⁺, Na⁺, and Cl⁻ in a typical mammalian neuron. Typical mammalian intracellular/extracellular concentrations (mM): K⁺ 140 in / 5 out; Na⁺ 12 in / 145 out; Cl⁻ 4 in / 110 out (all $z=\pm1$, at 37 °C, $RT/F = 26.7$ mV, so $RT/F \times \ln = 26.7 \times 2.303\log_{10} = 61.5\,\text{mV}\times\log_{10}$). For K⁺ ($z=+1$):

For Na⁺ ($z=+1$):

For Cl⁻ ($z=-1$, note the sign flip):

These values — $E_K \approx -89$ mV, $E_{Na} \approx +67$ mV, $E_{Cl}\approx -89$ mV — are the textbook benchmarks used throughout electrophysiology; note that $E_K$ and $E_{Cl}$ happen to be close but arise from concentration ratios that are reciprocal in valence sign, an asymmetric numerical coincidence, not a general rule.

Example 2 — Applied: resting potential from the GHK equation with unequal, non-round permeabilities. A resting neuron has permeability ratios (unitless, relative) $P_K : P_{Na} : P_{Cl} = 1 : 0.04 : 0.45$ (typical resting-state values, since most Na⁺ channels are closed and some Cl⁻ channels are open). Using the same concentrations as Example 1:

Numerator: $1(5) + 0.04(145) + 0.45(4) = 5 + 5.8 + 1.8 = 12.6$ Denominator: $1(140) + 0.04(12) + 0.45(110) = 140 + 0.48 + 49.5 = 189.98$

Compare to $E_K = -89.0$ mV alone: the resting potential of $-72.5$ mV sits between $E_K$ and $E_{Cl}$, pulled away from pure $E_K$ by the non-zero $P_{Na}$ and $P_{Cl}$ contributions — exactly the qualitative behavior GHK predicts, and a good check that the equation correctly interpolates rather than picking one ion exclusively. As a limiting-case check: if $P_{Na}\to 0$ and $P_{Cl}\to 0$ (pure K⁺ permeability), the GHK expression collapses algebraically to the Nernst equation for K⁺ alone, recovering $V_m \to E_K = -89.0$ mV.

Example 3 — Exam/competition level: single-channel current, conductance, and estimating ion flux rate through an open KcsA-like channel. A single open K⁺ channel has a measured single-channel conductance $\gamma = 20$ pS (picosiemens) — a typical value for large-conductance K⁺ channels — held at a membrane potential of $V_m = -80$ mV, with $E_K = -90$ mV for this cell. Ohmic behavior for the open channel gives:

Convert this current to an ion flux rate. Current is charge per time, and each K⁺ carries charge $e = 1.602\times10^{-19}$ C:

This is a modest driving force (only 10 mV from $E_K$); at larger driving forces (e.g., during an action potential's repolarizing phase, where $V_m - E_K$ can reach 50–60 mV), the same channel would pass roughly 6–8×10⁶ ions/s, consistent with textbook values of 10⁷–10⁸ ions/s reported for K⁺ channels under maximal physiological driving force. As a limiting-case check: if $V_m = E_K$ exactly, the driving force vanishes and $i \to 0$ — correctly predicting no net current at the reversal potential, regardless of how open the channel is (open probability affects whether current can flow, not its reversal point).

Roderick MacKinnon began his career as a physician-scientist studying vascular smooth muscle before turning fully to structural biology of ion channels in the early 1990s. Working first with electrophysiological mapping of the Shaker K⁺ channel pore, MacKinnon's group used systematic cysteine-scanning mutagenesis and toxin-binding footprinting to infer the approximate architecture of the pore and selectivity filter years before any crystal structure existed — a triumph of indirect, quantitative biochemical reasoning. In 1998, MacKinnon's laboratory solved the first X-ray crystal structure of a K⁺ channel, KcsA from Streptomyces lividans, at 3.2 Å resolution, directly visualizing the TVGYG selectivity filter's backbone-carbonyl cage and the dehydrated K⁺ ions sitting within it — instantly explaining, in structural terms, decades of purely functional selectivity data. He shared the 2003 Nobel Prize in Chemistry with Peter Agre (aquaporins) for this work. MacKinnon's subsequent structures of voltage-gated K⁺ channels (KvAP) and the mechanistic elucidation of the "paddle" voltage-sensor motion extended the same rigorous structure-to-function logic to gating, cementing ion channel structural biology as a discipline built on the interplay of electrophysiology, crystallography, and physical-chemical reasoning about ions in confined, chemically patterned spaces.