One of the grails at the junction of neurobiology and structural biology has been the physical basis of gating in voltage-gated ion channels. A recent paper seems to have cracked the problem for one of them, the potassium (K+) channel.
Nerves conduct impulses quite differently than do electrical wires. There are places where nerves are heavily wrapped with insulation (myelin) and electric pulses race along almost like they do through a wire, but those segments are very short, and the pulse still consists of a small voltage difference across the membrane, between inside the cell and outside, not a difference in voltage between one end of the wire and the other end, as in copper wires.
All the action is at the membrane, and it is driven by an interesting cast of ions and ion channels. Some channels pump ions against their natural gradients (using ATP) to maintain the resting state of the nerve cell. This state is high K+ inside and high sodium (Na+) outside (see animation here). Other channels then use this primed rest condition to propagate the transient nerve impulse, also called the action potential or spike. This begins with a rapid opening of Na+ channels to create a reversal in voltage, then closing of those channels, followed by rapid opening of K+ channels to bring the voltage back to the original state, then a brief refractory period, and then back to rest, ready to fire again.
The voltage reversal, or "spike" of an action potential induces nearby Na+ and K+ channels to do the same thing, unless they have just fired, in which case they are in their refractory period. It is the properties of these channels that gives the nerve impulse its form and forward direction. The nearby Na+ and then K+ channels fire because they are "voltage-gated", meaning that they are turned on by a small change of membrane voltage (another crude animation). In this respect, they are similar to transistors, which use a small voltage to regulate a larger current running through the main conduit.
Biologically, this is a very special type of ion channel control, different from other means of controlling channels, such as binding specific chemicals, temperature, light, or touching structures on other cells. Once open, the channel is passive, letting Na+ or K+ through as fast as it can, but also selective, having just the internal shape and electrical fields to allow that one ion through and no others.
When the membrane potential changes again, due the channel's own effects (high voltage flips another voltage gate to inactivate the Na+ channel while flipping the K+ channel on; then the low voltage re-established by the K+ channel closes it as well). All of which brings the local voltage back down a little below its resting state, with the Na+ and K+ channels closed again.
The current paper describes the physical mechanism of voltage gating of the K+ channel. Naively, one would think that a little charged lid might cover it at one end, but that isn't the case. What happens is that a long protein helix sitting next to the channel slides up through the membrane, levering an attached helix against the bottom of the channel, allowing it to open. The positive charge of this helix gives it its sensitivity. The authors begin by solving an X-ray crystal structure of the K+ channel with its voltage sensor in place, in the open position.
Let's start with images of the channel core in its open and closed positions. Imagine the membrane going horizontally through the picture, with the channel peeking out at top (outside of the cell) and bottom (inside).
The open form (O) is the structure they solved for this K+ channel protein from rat. The closed form (C) is a structure they model based on structures from similar channels (but solved without the voltage gating mechanism) in their closed forms. The potassium ions (tiny green dots) are in the center channel arriving from the outside, whether the bottom end is open or closed. It is only the bottom of the channel that changes shape substantially when closed- tightening into a sort of pucker. This structure can be seen in three dimensions at the protein databank (use the Jmol view tool, which gives full 3-D control and other animation/drawing options [closed form]).
The red helix at the bottom is a key piece of the voltage sensor, but not the whole thing. The sensor occurs in four copies around the channel, which is why there are four helices. Four proteins make up this structure, each contributing one-fourth of the core channel structure, but each contributing one voltage sensor on the periphery. The next structure shows the whole thing they solved- voltage sensors (two shown in red and gray) plus the open channel (blue).
You can see that the voltage sensor (S4) links to the channel bottom through the small helix S4-S5. Leverage up or down in the red sensor communicates to the channel through this linkage. The helices S3b and S4 are the most active parts of the sensor, which the authors deduce must descend dramatically through the membrane to physically push the S4-S5 helix to the other orientation and close the channel.
Then next image is the voltage sensor up close, showing the details of how these helixes fit together. Note that these structures are shows as "ribbon diagrams", only schematically showing the path of the protein backbone. There are many other ways of visualizing atomic structures, based on what one is interested in seeing.
Here they use the amino acid codes R for arginine, K for lysine (sorry about the confusion, chemists!), F for phenylalanine (green), E for glutamic acid and D for aspartic acid. The charges are critically important- lysine and arginine are both positively charged (blue- R and K), while the acids are negatively charged (red- E and D). In this open channel conformation, lysine 5 (K5) is parked comfortably opposite the negatively charged E and D and tucked under the green phenylalanine (F). Since this structure is otherwise enclosed by the oily membrane (gray bars on the sides), any charge on the protein likes to stick to an opposite charge if it can, and this F-pocket is particularly influential in allowing the positive charges R1 through K5 to ratchet down through the membrane, positive step by positive step.
The model that the authors present is that in response to membrane voltage changing, the string of positively charged amino acids from K5 to R1 descend, twist, and bind sequentially in the E+D+F pocket. This naturally wrenches the S4-S5 helix around and forces the channel closed.
The crux of the work is that the authors look at the structural model (along with decades of prior work in the field), and infer the dynamic action as outlined above. To provide some further evidence, they make a series of mutations of the R1 and K5 amino acids as well as the F position, and measure their electrical effects- what voltage it takes to open the channel, to close the channel, and even voltage associated with movement of the gating charges themselves (the "gating current"). Moving the series of positively charged R and K residues across the membrane reveals itself to extremely sensitive methods while the channel pore itself is chemically blocked.
They find that lysine (K) binds better to the E+D+F pocket than does R, so putting R at position 5 makes the channel easier to close (to move out of the open position), while putting K at position 1 makes the channel easier to keep closed. When both positions are K, the channel is harder both to open and to close, as shown by about 40 ms extra time for each process in the K1K5 mutant (D, below), compared to the R1R5 mutant (A).
Gating currents are measured for selected mutants in the 1 and 5 sites, showing that K on either end makes either opening (first half of trace) or closing (second half of trace) slower, presumably due to enhanced stability of binding in the F-pocket and reluctance to be dislodged.
It is beautiful work, and I can only touch on the surface here. The authors integrate structural and functional evidence in classic fashion, supporting the presented theory of how the voltage gate works in the K+ channel. Since this gating is one of the many pieces of machinery that lie at the heart of nerve function, of brain function, and thus of the mind, it is a worthy quest that has taken decades, journeying from awe and mystery to understanding.
