Voltage-dependent ion channels are exquisitely sensi-Seoh et al., 1996). Second, the aqueous accessibility of tive to membrane potential, so much so that a depolarcysteines or histidines substituted into S4 segments ization of only 10 mV can cause a 100-fold increase of changes in a voltage-dependent manner (reviewed by open probability (Schoppa et al., 1992; Hirschberg et Yellen, 1998; Bezanilla, 2000). The accessibility experial., 1995). This athletic feat requires the energetic couments (eg, Figure 1C) indicate that S4 segments transpling of channel opening to the movement of at least locate their positive charges from intracellular to extra-12 elementary charges across the membrane electric cellular compartments, and therefore through the electric field. These charges reside in specialized structures of field, when a channel is depolarized, as expected for a the ion channel protein known as voltage sensors, four voltage sensor. of which are found in calcium, sodium, and potassium As icing on the cake, we now have evidence for voltchannels. The main component of the voltage sensor is age-dependent movement at the extracellular ends of the S4 segment. This is the only transmembrane segthe S4 segments. Fluorophore-tagged S4 segments ment that is appreciably charged in voltage-gated ion change the intensity of fluorescence emission in a voltchannels; S4 segments contain from two to eight posiage-dependent manner (Mannuzzu et al., 1996). The kitively charged residues, typically separated from one netics and voltage dependence of this process are simianother by two neutral residues. lar to what is observed in the gating currents of the Two unanswered questions currently captivate researchers studying voltage-gated ion channels. How channels, suggestingthatconformationalchangesonor does the S4 segment move its charges across the mem- near the S4 segment underlie voltage-dependent charge brane electric field, and how is this movement coupled movement. to the gates that control access to the highly selective How Are S4 Charges Translocated? permeationpathway? Untilveryrecentlywewerealmost One of the surprising insights from the accessibility clueless about the answers to both of these questions. studies is that, although S4 segments are true trans-However, two papers in the December 16 issue of Nature membrane segments, most of their residues are acces-(Cha et al., 1999; Glauner et al., 1999) offer a provocative sible to the aqueous solutions bathing either the extraclaim in answer to the first question: S4s transfer their cellular or the cytoplasmic sides of the channel protein charges by rotating. The main purpose of this minireview(Figure 1C). This means that the hydrophobic region of is to evaluate this claim and consider alternatives. The the protein through which the S4 moves, sometimes answer to the second question remains a mystery, but called the gating pore, is considerably shorter than the I will discuss some possibilities below. thickness of the bilayer (45 A). Therefore, aqueous Evidence that S4 Is a Voltage Sensor vestibules must largely surround the S4 segment. The subunits contain all the machinery necessary for volt- short gating pore insures that only a small movement age-dependent gating, namely the three critical compo- of the S4 segment is required to produce significant nents of voltage-gated ion channels: voltage sensors, charge translocation. gates, and a selective permeation pathway. The sub- Most of the molecular models of gating assume that unit of the potassium channel has six transmembrane the S4 segment is an helix, an assumption supported segments, S1–S6 (Figure 1A). Four …