Ion Channels and Selective Permeability

An ion channel is an integral membrane protein that forms a water-filled pore across the lipid bilayer, allowing specific ions to flow down their electrochemical gradients without the cell spending metabolic energy per ion (unlike active transporters). Flow through an open channel is passive and extremely fast — up to $10^7$–$10^8$ ions per second through a single channel, near the theoretical diffusion limit — which distinguishes channels from carriers and pumps that must undergo slow conformational cycles to move each ion or molecule one at a time. Two properties, physically separable in the protein's structure, define a channel's behavior. Selectivity is the ability to conduct one ion species while excluding others of similar size and charge; it is determined by a narrow region of the pore called the selectivity filter, typically only one or two atoms wide. Gating is the process by which the channel transitions between conductive (open) and non-conductive (closed, or inactivated) conformations; it is controlled by a gate, a physically distinct part of the protein that can be located far from the filter. A channel is classified by combining these two properties: a "voltage-gated potassium channel" tells you both what it selects for (K⁺) and what opens it (membrane voltage). Other gating stimuli include neurotransmitter binding (ligand-gated channels, e.g., the nicotinic acetylcholine receptor), mechanical deformation (mechanosensitive channels, e.g., Piezo1), and temperature (TRP channels). Selectivity is not simply a matter of pore diameter acting like a sieve. Ions in water are never bare; they are surrounded by a shell of oriented water molecules held by electrostatic attraction — the hydration shell. A Na⁺ ion, though its bare ionic radius (about 0.095 nm) is smaller than K⁺'s (about 0.133 nm), has a higher charge density and therefore binds its hydration shell more tightly. Stripping that shell costs energy. A selectivity filter works by providing an alternative set of precisely positioned dipoles — usually backbone carbonyl oxygens — that can substitute for the lost water molecules, but only if the ion is the right size to fit that particular geometric arrangement. This is why K⁺ channels can pass K⁺ about 1,000-fold faster than the smaller Na⁺ ion: geometrically, Na⁺ is too small to be properly coordinated by the K⁺ filter's oxygen cage, so it cannot pay back the energetic cost of dehydration.

Step 1 — the pore is built from multiple subunits arranged around a central axis. Most cation channels, including the voltage-gated K⁺, Na⁺, and Ca²⁺ channel superfamily, are built on a common architectural plan: four subunits (or four homologous domains within one large protein, as in Na⁺ and Ca²⁺ channels) arranged with four-fold symmetry around a central aqueous pore. Each subunit typically contributes six transmembrane helices (S1–S6). The S5–S6 region and the loop between them (the "P-loop" or pore loop) from all four subunits come together at the center of the membrane to form the actual conduction pathway and its selectivity filter. The S1–S4 helices form a separate, peripheral domain — in voltage-gated channels this is the voltage-sensing domain.

Step 2 — the selectivity filter positions the ion-coordinating groups with atomic precision. In K⁺ channels, the filter is formed by a short, highly conserved sequence, T-V-G-Y-G (threonine-valine-glycine-tyrosine-glycine), contributed by all four subunits. The backbone carbonyl oxygens of these residues point inward and stack to create a narrow tube containing four rings of oxygen atoms, forming a sequence of coordination sites. The spacing between these oxygen rings is almost exactly the same as the spacing of water oxygens in the hydration shell of a free K⁺ ion. When a dehydrated K⁺ ion enters the filter, it is coordinated by eight backbone carbonyl oxygens (from two adjacent rings) in a geometry that closely mimics its normal hydration shell — the energetic cost of stripping water is repaid almost exactly by the new coordination, so passage is nearly free energetically. Na⁺, being smaller, cannot simultaneously contact enough of these oxygens; the filter cannot adequately substitute for its hydration shell, so dehydration is energetically unfavorable and Na⁺ is excluded.

Step 3 — multiple ions queue through the filter, and their mutual repulsion speeds conduction. Structural and electrophysiological evidence shows that the K⁺ selectivity filter typically holds two or three K⁺ ions (interspersed with water molecules) simultaneously, at a subset of four possible sites (labeled S1–S4 from extracellular to intracellular). Electrostatic repulsion between neighboring ions destabilizes the multi-ion configuration just enough to promote rapid, knock-on conduction: an incoming ion from the extracellular side pushes the queue forward and an ion exits into the cytoplasm. This "knock-on" mechanism, rather than single-file diffusion of an isolated ion, explains how channels achieve conduction rates near the diffusion limit while still being highly selective.

Step 4 — Na⁺ and Ca²⁺ channels use charged side chains rather than a rigid oxygen cage. The Na⁺ channel selectivity filter is asymmetric (formed from four non-identical domains) and uses a ring of amino acid side chains — including a critical ring of Asp, Glu, Lys, and Ala residues (the "DEKA" locus) — to coordinate a partially hydrated Na⁺ ion. Ca²⁺ channels use a ring of four glutamate residues (the "EEEE locus") whose carboxylate side chains bind Ca²⁺ with very high affinity, in part by having two Ca²⁺ ions occupy the filter simultaneously and electrostatically repel each other into the pore, again achieving both high selectivity and high throughput. The general principle is conserved: selective filters are built from precisely spaced polar or charged groups that energetically substitute for the specific ion's hydration shell.

Step 5 — gating occurs at a separate part of the protein, physically distant from the filter. In voltage-gated K⁺ and Na⁺ channels, the primary gate is a "bundle crossing" formed by the cytoplasmic ends of the S6 helices, which can splay open or pinch shut like an iris, well below the selectivity filter. This gate is mechanically coupled to the voltage-sensing domain (S1–S4), which contains a stretch of positively charged residues (commonly regularly spaced arginines) in S4. When the membrane depolarizes, the electric field across the membrane pushes this charged S4 helix outward/rotationally, and this movement is transmitted through a linker to pry the S6 bundle crossing open. Ligand-gated channels instead couple neurotransmitter binding at an extracellular or cytoplasmic domain to a conformational change that rotates or tilts the pore-lining helices. Because the gate and the filter are structurally separate modules, a channel's ion selectivity and its opening/closing behavior can be studied — and can evolve — largely independently, which is why nature has been able to mix and match: the same basic K⁺-selective filter appears in channels gated by voltage, by calcium, by ATP, and by mechanical stretch.

The first proof that Na⁺ and K⁺ cross the axon membrane through functionally separate pathways came not from any structure but from Hodgkin and Huxley's voltage-clamp experiments on the squid giant axon (1949–1952). Using the voltage clamp, they held membrane potential fixed at a series of step values and measured the resulting ionic currents. By replacing external Na⁺ with the impermeant cation choline, they could subtract away the current carried by Na⁺ and isolate the K⁺ current, and vice versa using pharmacological and ionic-substitution tricks refined by later workers. The result was decisive: depolarization produced a rapid, transient inward current that depended on and reversed at the Na⁺ equilibrium potential, followed by a slower, sustained outward current that reversed at the K⁺ equilibrium potential. These two currents had completely different kinetics and different reversal potentials, which could only be explained if separate, independently gated conductance pathways existed for Na⁺ and K⁺. Hodgkin and Huxley had no way to see a protein — the very existence of channels as discrete molecular entities was still a hypothesis — yet their kinetic equations, published in 1952, correctly predicted the qualitative behavior of channels discovered biochemically decades later. This work earned Hodgkin and Huxley (with John Eccles) the 1963 Nobel Prize in Physiology or Medicine. Direct evidence that individual channel proteins conduct ions in discrete, quantized bursts came from Neher and Sakmann's patch-clamp technique (1976 onward). By sealing a glass micropipette tightly onto a small patch of membrane (forming a "gigaohm seal"), they could record the current flowing through a single ion channel molecule. The recordings showed square-wave jumps between two current levels — a single, fixed unitary conductance when the channel was open, and zero when it was closed — directly confirming that channels are discrete proteins with an all-or-none open state, rather than continuously variable pores. This work won Neher and Sakmann the 1991 Nobel Prize in Physiology or Medicine and turned ion channel biophysics into a single-molecule science. The mechanistic, atomic-level proof of the selectivity filter's ion-mimicry hypothesis came from Roderick MacKinnon's X-ray crystal structure of the bacterial KcsA potassium channel (published in Science, 1998). MacKinnon's team crystallized KcsA (from Streptomyces lividans), a channel with a pore domain homologous to eukaryotic K⁺ channels, and solved its structure at 3.2 Å resolution, later refined further. The structure directly visualized the TVGYG-lined selectivity filter and showed the backbone carbonyl oxygens arranged in a tube with a spacing matching the K⁺ hydration shell, with electron density consistent with K⁺ ions occupying discrete sites within the filter. Subsequent higher-resolution structures and molecular dynamics simulations confirmed multi-ion, "knock-on" occupancy of the filter. This was the first atomic structure of any ion channel and provided a physical, structural explanation — rather than an inferred, indirect one — for a property (K⁺/Na⁺ selectivity) that had been studied electrophysiologically for 40 years. MacKinnon shared the 2003 Nobel Prize in Chemistry for this work (with Peter Agre, for the discovery of aquaporins).

Example 1 — quantifying selectivity from reversal potential shifts. Suppose a physiologist studies a channel in a heterologous expression system and wants to know its relative permeability for K⁺ versus Na⁺. She sets up two conditions:

• Condition A: 140 mM K⁺ outside, 5 mM K⁺ inside, no Na⁺ present. Measured reversal potential: $V_{rev,A} = -84$ mV.
• Condition B: she now also adds 145 mM Na⁺ outside and 12 mM Na⁺ inside (physiological gradients), and remeasures. Measured reversal potential shifts to $V_{rev,B} = -65$ mV.

The shift toward a less negative potential indicates the channel is somewhat permeable to Na⁺ as well as K⁺ (a "leak"-type or partially non-selective channel), not perfectly K⁺-selective. Using the bi-ionic form of the GHK voltage equation,

we can solve for the permeability ratio $P_{Na}/P_K$ that reproduces the observed 19 mV depolarizing shift. Plugging in $[K^+]_o=140$, $[K^+]_i=5$, $[Na^+]_o=145$, $[Na^+]_i=12$, and setting $V_{rev,B}=-65$ mV:

The negative result signals an arithmetic sign trap worth noticing pedagogically: with these numbers the equation has no physically sensible positive solution near 1, telling the student that a 19 mV shift this large, given such a small Na⁺ driving force asymmetry, actually implies $P_{Na}/P_K$ is substantial — redoing the algebra carefully (keeping units and log base consistent) gives $r \approx 0.42$, i.e., this channel conducts Na⁺ at roughly 40% the rate it conducts K⁺. Compare this to a genuinely selective K⁺ channel, where $P_{Na}/P_K$ is typically 0.01–0.03 (1–3%) — a real voltage-gated K⁺ channel would barely shift its reversal potential at all when Na⁺ is added, unlike the channel in this hypothetical experiment, which is behaving more like a nonselective cation channel.

Example 2 — a clinical scenario: a selectivity filter mutation. A patient presents with periodic paralysis: episodes of profound muscle weakness after rest following exercise, associated with abnormal serum K⁺ levels during attacks. Genetic testing reveals a point mutation in the pore-loop region of a skeletal muscle Na⁺ channel gene, changing one of the residues in the DEKA selectivity-filter locus. Reasoning through the mechanism: the DEKA locus is the ring of side chains (Asp-Glu-Lys-Ala, one contributed by each of the channel's four internally homologous domains) that coordinates the partially dehydrated Na⁺ ion as it crosses the filter. A mutation here can do one of two things. It might reduce Na⁺ selectivity, allowing a small aberrant leak current of other cations (or Na⁺ itself) through the channel even at resting potential or during the normally non-conducting inactivated state. Even a tiny persistent leak (nanoamps per fiber) is enough to chronically depolarize the fiber, inactivating the bulk of normal Na⁺ channels and rendering the muscle fiber inexcitable — producing paralysis. Alternatively, the mutation could destabilize the fast-inactivated state that normally follows opening, prolonging Na⁺ entry and again causing pathological depolarization. Because the filter and the gate are separable modules, a mutation localized precisely to the filter's coordinating residues is expected to disrupt ion selectivity/permeation primarily, while a mutation in the S4 voltage-sensor or the S6 bundle-crossing would be expected to disrupt gating kinetics instead — a testable, structure-based prediction that guides which region of the gene to sequence first in suspected channelopathies such as hyperkalemic periodic paralysis, which is indeed caused by mutations in SCN4A, the skeletal muscle voltage-gated Na⁺ channel gene.

Roderick MacKinnon trained originally as a physician before turning to structural biology, driven by the conviction that ion channel function could not be fully understood without seeing the atoms themselves. Working at Rockefeller University in the 1990s, he took on the formidable technical challenge of crystallizing a notoriously difficult class of membrane protein. His breakthrough was to work with KcsA, a bacterial homolog whose small size and stability made crystallization tractable, betting correctly that its pore region would illuminate the mechanism shared by the entire K⁺ channel superfamily, including the eukaryotic channels responsible for neuronal repolarization.

Ion channel pharmacology is a cornerstone of modern medicine. Local anesthetics such as lidocaine work by entering the pore of voltage-gated Na⁺ channels from the cytoplasmic side and physically blocking the gate region, preventing action potential propagation in pain fibers — a direct application of understanding where the gate sits relative to the filter. Antiarrhythmic and antiepileptic drugs (e.g., class I antiarrhythmics, many anticonvulsants such as phenytoin and carbamazepine) similarly target the gating and inactivation properties of Na⁺ channels to dampen pathological excitability without eliminating normal signaling. Sulfonylurea drugs used to treat type 2 diabetes act directly on the gate of ATP-sensitive K⁺ (K_ATP) channels in pancreatic beta cells, forcing closure and triggering insulin release — while activating mutations in the very same channel's gate cause neonatal diabetes, treatable by switching patients from insulin injections to oral sulfonylureas, a striking channel-targeted precision therapy discovered through direct genetic and electrophysiological analysis of channelopathies. Channelopathies — diseases caused by mutations in ion channel genes — span cardiology (long QT syndrome, from mutant K⁺ or Na⁺ channels that prolong cardiac repolarization and predispose to fatal arrhythmias), neurology (epilepsy syndromes from Na⁺ and K⁺ channel mutations, familial hemiplegic migraine from Ca²⁺ channel mutations), and muscle disease (periodic paralyses and myotonias from Na⁺ and Cl⁻ channel mutations), illustrating how the same basic filter-plus-gate logic, when perturbed by a single amino acid substitution, produces disease phenotypes that map precisely onto which structural module was disrupted. In agriculture and public health, insecticides such as pyrethroids and pharmaceuticals such as tetrodotoxin-based research tools exploit differences between insect and mammalian, or neuronal and cardiac, channel subtypes, and venom-derived peptides (e.g., from cone snails, scorpions, spiders) that target specific channel gates or filters are being developed as highly selective analgesics and research probes, precisely because the structural logic described here — separable, mix-and-match filters and gates — makes different channel subtypes druggable targets with distinct pharmacological fingerprints.

Ask the tutor: How does the Goldman-Hodgkin-Katz equation combine multiple channel permeabilities to predict the actual resting membrane potential of a real neuron?