Effect of introducing charged residues at the internal entrance of Shaker-WT channels. (Left) Leak- and capacitance-subtracted single-channel traces of (A) Shaker-WT, (B) V476D, and (C) S479D in a wide range of symmetrical [K⁺] (100–3,000 mM for Shaker-WT and 30–1,000 mM for V476D and Shaker-S479D). (Right) I-V relations obtained from open-channel amplitudes. Data correspond to n ≥ 10 traces from three separated patches. Data were fitted by a third-degree polynomial function having no theoretical meaning, and the slope of the fit traces at zero voltage was used to obtain the conductance values. (D) Zero-voltage slope conductance of the unitary currents displayed in the right panels of A–C. Data were fitted to a Langmuir isotherm for comparative purposes. Fit parameters: Shaker-WT, Kd = 250 ± 38 mM and g_max = 59 ± 10 pS; V476D, Kd = 3.6 ± 2.6 mM and g_max = 61 ± 3 pS; S479D, Kd = 4.7 ± 1.6 mM and g_max = 75 ± 3 pS. The discrepancy with the maximum conductance of ∼45 pS measured for Shaker-WT (Heginbotham and MacKinnon, 1993) is possibly caused by the absence of added Mg²⁺ in our solutions. Mg²⁺ could induce a low affinity blockade that reduces single-channel conductance (Heginbotham and MacKinnon, 1993; Harris et al., 1998; Moscoso et al., 2012).