External electronegative surface of human KCNQ1 plays a minor role in the K ⁺ _o sensitivity of the channel. (A) Alignment of human and Xenopus KCNQ1 sequences along with several other K⁺ channels in S5—pore helix region: Xenopus KCNQ1 (xKCNQ1; NCBI ID NP_001116347.1), human KCNQ1 (NCBI ID NM_000218.2). Red letters indicate negatively charged residues making up the externally charged surface. Diagram above the sequences corresponds to secondary structure elements corresponding to the human KCNQ1 structure (Sun and MacKinnon, 2020). Below is a representation of the surface potential of wild type KCNQ1 channels and that of the modeled human triple mutant scaled from −5 KT/e to 5 KT/e (top view). (B) Representative current traces and analysis of [K⁺]_o-dependent inhibition of the triple mutant compared to wild type. Parameters of the linear fits are as follows: WT, 0.2 mM K⁺ = 0.98 ± 0.02 and 100 mM K⁺_o = 0.47 ± 0.02; 290A/291A/295A mutant, 0.2 mM K⁺ = 0.98 ± 0.02 and 100 mM K⁺ = 0.38 ± 0.03, n = 8, error bars represent ±SEM; ***P = 0.0007 ANCOVA. (C) Sample current traces of single mutants at five different [K⁺]_o conditions as indicated. (D) Peak current amplitudes in variable [K⁺]_o normalized to the first point of recording at 0.2 mM [K⁺]_o for E290A, S291A, and E295A channels, n = 5–9, error bars represent ±SEM. (E and F) Concentration–response curves for mutants after correction for theoretical reduction (Ith) assuming [K⁺]_in = 108 mM (Weber, 1999). Solid lines are fits of the Hill equation to data. Dashed line represents fit of wild type data. Parameters of fits are shown in Table 1. (G) Dependency of maximal inhibition from side-chain volume at E290 position. Values of maximal inhibition are taken from Table 1. n = 4–6, error bars represent ±SEM.