A puzzle known only by its currents
The claim at the centre of this work is a structural one: MacKinnon’s laboratory showed what a potassium channel actually looks like at the atomic level, and in doing so explained a property that had puzzled physiologists since the 1950s. Potassium channels let potassium ions cross a cell membrane rapidly while almost entirely excluding sodium ions, even though sodium ions are physically smaller. That combination is counterintuitive on size alone, and until 1998 it had been inferred only from indirect electrical measurements of current flowing through membranes. MacKinnon’s crystal structure of a bacterial potassium channel called KcsA showed a narrow region, the selectivity filter, built from a ring of backbone oxygen atoms spaced to substitute for the water molecules that normally surround a potassium ion in solution.
A bacterial channel stands in
MacKinnon’s group worked with KcsA, a potassium channel taken from the bacterium Streptomyces lividans, chosen because it could be produced and purified in the quantities crystallography requires, unlike the channels found in animal nerve cells. The purified protein was crystallised and its structure solved by X-ray crystallography at a synchrotron facility at Cornell, a method that fires X-rays through an ordered crystal and reconstructs the atomic arrangement from the pattern the rays produce. This built on decades of work that had approached the same channels only from the outside: Hodgkin and Huxley’s electrical analysis of the action potential in the early 1950s, and later the patch clamp technique developed by Erwin Neher and Bert Sakmann, which recorded the current passing through single channels without ever showing their shape.
The filter made visible
What has held up is the central mechanism: the selectivity filter’s ring of oxygen atoms is positioned to coordinate a potassium ion in much the same geometry as the water molecules it sheds to enter the channel, letting it pass through with little energetic cost, while a sodium ion, despite being smaller, cannot be coordinated the same way and is effectively excluded. This confirmed a mechanism that Bertil Hille and Clay Armstrong had proposed on theoretical grounds in the late 1960s, based on dehydration behaviour, well before anyone could see the structure responsible. The finding has since been extended: related selectivity filters and channel architectures have turned up across many other channel families, and the general principle that channel selectivity is a matter of geometric fit rather than simple pore size is now well established.
Why the smaller ion loses out
What a single crystal structure cannot do is show a channel in motion. KcsA’s structure captures one operational state, most likely close to closed, and says little on its own about the sequence of movements the channel makes as it opens, conducts, and closes again, or about how gating signals from elsewhere in the protein reach the filter. Producing the protein for crystallography also required extracting it from its membrane using detergent, which raises a standing question about how closely a purified, crystallised structure matches the channel’s shape in the fatty membrane it actually sits in. Broader questions about lipid-gated channels, the precise role auxiliary subunits play in the many channel families beyond KcsA, and the full conformational cycle of gating remain areas the structure alone does not settle.
One structure, one moment
Ion channels are not a specialist curiosity: they set the resting electrical charge of every cell, shape the nerve impulse, trigger muscle contraction, and pace the heartbeat, all by controlling which ions cross a membrane and when. Knowing the physical shape of the selectivity filter turned drug design for these channels from a matter of trial and error against an unseen target into something closer to structure-based design, relevant to local anaesthetics, antiarrhythmic drugs for the heart, and anticonvulsants for epilepsy, all of which act on specific channel types. It also gave later structural biology a template: once one channel’s atomic structure was solved, the same crystallographic approach, and later cryo-electron microscopy, could be turned on the many other channel families that regulate nerve and muscle function.
A Nobel Prize in Chemistry
Yes, particularly for anyone who has only ever encountered ion channels as arrows on a diagram of the nerve impulse. What makes this worth the time is less the biochemistry itself than the shift it represents, from decades of channels known purely by the electrical currents they produced to a channel whose actual shape could finally be examined and matched against that behaviour. It rewards patience with a genuinely satisfying piece of mechanistic reasoning, the idea that a ring of oxygen atoms can substitute for a shell of water molecules closely enough to let one ion through and stop another. It is a denser read than a discovery story with a single dramatic moment, but the payoff is a clear, checkable answer to an old question.