Figure 1. subunits of voltage-gated potassium channels. (a) Simplified topological model of a Kv channel subunit in the cell membrane. Six putative transmembrane segments form the core of the channel complex. Segment S4 is highly positively charged and acts as voltage sensor, the linking structure between S5 and S6 takes part in lining the pore and forming the selectivity filter. Both N- and C-termini are located in the cytosol. (b) Sections through Kv-channel models indicating the arrangements of the four subunits. The green part indicates the pore region with the selectivity filter, the red segments represent S4, which undergoes a translocation upon membrane depolarization. This voltage-gated translocation of S4 results in pore opening and subsequent outward flux of  K⁺ ions. (c) Representation of three families of voltage-gated  K⁺ channels: the classical Kv, the KCNQ and the EAG family. Also shown is a selection of subfamilies and some members. These Kv channels all share the common six-transmembrane segment motif. In this respect  Ca²⁺-activated  K⁺ channels (KCa) as well as cyclic nucleotide-gated channels (CNGC) are close relatives of Kv channels.

Figure 2. Structural data of the  K⁺ channel KscA from Streptomyces lividans. (a) Top view on a ribbon diagram of a KscA tetramer. The four subunits are distinguished by colour. (b) Side view of a ribbon diagram of two KscA subunits. The numbered residues indicate mutations in Shaker channels at homologous positions that affect pharmacological properties. The red residues ‘GYG’ denote the signature sequence of  K⁺ channels that forms the selectivity filter. (c) Molecular surface of KscA and contour of the pore with three  K⁺ ions (green spheres). Reprinted from Doyle et al. (1998). Copyright © 1998 American Association for the Advancement of Science.

Figure 3. Functional properties of Kv channels. (a) Voltage protocol and resulting ensemble  K⁺ currents mediated by delayed rectifier Kv1.5 channels, expressed in host cells. With increasing depolarization more and more channels open and conduct  K⁺ ions. (b) The peak current of the data in (a) is plotted as a function of the test voltage (open circles). This current–voltage relationship indicates that the channels start to open when the membrane is more depolarized than to –40 mV. Using a simple kinetic model of channel activation (c) this voltage-dependent channel opening can be described (continuous thick curve). From such an analysis one obtains the probability for finding a channel in the open state (P_open, green). Since the S4 segments of four subunits have to undergo transitions before channel opening one can calculate the probability for finding an individual subunit in the ‘activated’ configuration (red). (c) Simple state diagram used for a quantitative description of voltage-dependent Kv channel activation and inactivation. In this model it is assumed that four subunits are either deactivated (grey) or activated (red). Only when all of them are activated, can the channel open. Additionally indicated are transitions into inactivated channel states, entered either from closed or open states. The Greek letters indicate voltage-dependent rate constants. (d) Current recordings of Shaker B channels (top) and Shaker B channels with a truncated N-terminal domain, in response to membrane depolarization. While Shaker B channels undergo rapid N-type inactivation by means of the ‘ball-and-chain’ mechanism (e), deletion of the ‘ball’ domain removes this rapid inactivation, but unmasks the slower P/C-type inactivation. (e) Models for open and N-type-inactivated Kv channels. The scissors indicate how N-type inactivation can be abolished by N-terminal deletions.

Figure 4. Channel assembly and interactions with auxiliary proteins. (a) Transmembrane topology of a Kv subunit indicating the N-terminal domains T1A,B for subunit tetramerization (b) and T1B for interaction with Kv subunits (c). Kv channels can also interact via their C-terminal domains with scaffold proteins. This example shows a membrane-associated guanylate kinase homologue (MAGUK), harbouring three PDZ (named after three MAGUK family members), one SH3 (Src homology) and one GuK (guanylate kinase) domain. (b) Kv subunits can form homotetramers (left) or heterotetramers (others). The specificity for subunit–subunit interactions is mainly determined by the T1 domain, thus limiting the number of possible tetramer combinations (right). (c) Assembly of Kv channels with auxiliary subunits. KCNE proteins (left) can interact with some Kv channels of the KCNQ and the EAG family. Kv subunits bind within the T1B domains of Kv1 channels. Some of them provide an inactivating N-terminal ball domain to confer rapid inactivation to delayed-rectifier channels (right).