The human ether-Æ`a-go-go-related gene (HERG) encodes the pore-forming subunit of a cardiac potassium (K⁺) channel involved in action potential repolarisation. HERG channel gating is characterised by very slow activation and fast C-type inactivation, which unlike other voltage-gated K⁺ channels appears to be independent of activation. Thus, for wild-type (WT) HERG the potentials for half-maximal activation (V_0.5,act) and inactivation (V_0.5,inact) are -25 ± 0.7 mV and -85 ± 6 mV, respectively. Although S4 is assumed to be the voltage sensor for both gating processes, the contribution of each charged residue to gating is unknown. The aim of this study was to mutate each positively charged residue on S4 to an uncharged cysteine and investigate how this altered the activation and inactivation properties of HERG channels.

Two-electrode voltage clamp was used to record HERG currents from isolated Xenopus oocytes perfused with a Na⁺-based solution containing 2 mM K⁺ and 2 mM Ca²⁺, pH 7.6, at room temperature. Currents were recorded 1-5 days after injection of cRNA for WT or S4 mutant channels. Mean (± S.E.M.) data are from more than five cells.

The voltage dependence of activation was determined from normalised peak tail currents measured at -70 or -140 mV, following 5 s depolarisations to potentials between -100 and +70 mV. Mutation of residues in the middle of the putative S4 domain resulted in positive shifts of V_0.5,act of 31.5 ± 0.44, 47.6 ± 0.46 and 9 ± 1.12 mV for R528C, R531C and R534C, respectively, and no shifts of inactivation, measured using a triple pulse protocol. Cysteine substitutions at the ends of S4 had effects on both activation and inactivation gating. Mutation of R537 resulted in a -16.1 ± 1 mV shift of V_0.5,act and +35 ± 6 mV shift of V_0.5,inact. The mutations K525C and K538C resulted in substantial negative shifts of activation. Furthermore, in addition to a WT-like fast component of inactivation, a slower component was also identified. Tail currents following a depolarisation to +20 mV were smaller than at -40 mV, probably because of slow recovery from inactivation.

These data suggest that all six positively charged residues in the S4 domain of HERG contribute to voltage-dependent gating. The S4 mutant channels can be divided into two groups based on changes to gating. One group in the middle of S4 have positive shifts of activation and WT-like inactivation properties, whereas the mutants at the ends of S4 have negative shifts of activation and modified inactivation properties.

The results provide further evidence that HERG channel inactivation is independent of (or weakly coupled to) activation.

This work was supported by the Medical Research Council.