Why Ion Channels?

Voltage-gated and ligand-gated ion channels are well known for their role in electrical activity of nerve cells, but ion channels are found in many other places too. When a Paramecium bumps into an obstacle, mechanosensitive channels, along with calcium channels produce the action potential that switches the direction of motion of its cilia. The rate of beating of the human heart—roughly once per second—is determined by potassium channels that take about one second to close, and opposing cation channels that take about one second to open. Diabetes medications, novacaine, anti-epileptic drugs, snake and spider toxins all act on ion channels. Ion channel defects are responsible for disorders like Cystic Fibrosis, cardiac arrhythmias, and some kinds of hypertension and kidney disease.

In the nervous system, the processes underlying leaning and memory involve ion channels. Of particular interest are channels that act as "coincidence detectors," serving to signal the instances in which synaptic strength should be increased according to Hebb's rule. One such coincidence detector is the NMDA-type glutamate-activated channel, which passes current only when glutamate is present and when the membrane is depolarized. Another is the IP3 receptor, an intracellular calcium channel that is opened by the simultaneous presence of IP3 and calcium ions.

Some ion channels are remarkably selective. Potassium channels are orders of magnitude more permeable to potassium ions than sodium ions. Hodgkin and Keynes showed in 1955 that multiple ions pass through these channels in "single file", and work in Chris Miller's lab in the 1980s demonstrated the existence of four ion binding sites in the pore. These sites have recently been visualized in crystal structures of potassium channels from Rod MacKinnons laboratory. Similarly, the selectivity of calcium channels for divalent over monovalent ions arises from a similar "knock-on" mechanism, as was demonstrated in the 1980s in the laboratories of Wolf Almers, Richard Tsien and Peter Hess.

We are particularly interested in the way in which channels switch on and off their permeability to ions in response to electrical or chemical stimuli. This "gating" process appears to come from large mechanical motions (conformational changes) in the channel proteins. For example, Clay Armstrong demonstrated in 1962 that large ions can plug potassium channels from the intracellular side in such a way that they prevent the channel's gate from closing. Work in Rick Aldrich's lab in 1990 showed that one kind of "gate", the inactivation gate of the Shaker channel, consists of a small protein domain that acts like a tethered blocking ion. More complicated is the activation gate, which is coupled to a very sensitive "voltage sensor."

How can we determine how this molecular gating machinery works? Ideally one would want a movie of the motions of all the atoms in the channel protein as it opens and closes. At present we have, on the one hand, a few still pictures of channels obtained by X-ray crystallography, and on the other hand electrical recordings of channel activity in membranes. Of particular interest to us are methods that bridge the gap between these pieces of information. One is the single-particle cryo-EM technique for protein structure determination. This technique does not yield the resolution of X-ray crystallography but is much more flexible in providing pictures of proteins in various states. Another is the exploitation of fluorescence techniques to report, in a time-resolved fashion, the movement of particular residues or domains of a protein.