Regulation of human cardiac potassium channels by full-length KCNE3 and KCNE4

Abstract

Voltage-gated potassium (Kv) channels comprise pore-forming α subunits and a multiplicity of regulatory proteins, including the cardiac-expressed and cardiac arrhythmia-linked transmembrane KCNE subunits. After recently uncovering novel, N-terminally extended (L) KCNE3 and KCNE4 isoforms and detecting their transcripts in human atrium, reported here are their functional effects on human cardiac Kv channel α subunits expressed in Xenopus laevis oocytes. As previously reported for short isoforms KCNE3S and KCNE4S, KCNE3L inhibited hERG; KCNE4L inhibited Kv1.1; neither form regulated the HCN1 pacemaker channel. Unlike KCNE4S, KCNE4L was a potent inhibitor of Kv4.2 and Kv4.3; co-expression of cytosolic β subunit KChIP2, which regulates Kv4 channels in cardiac myocytes, partially relieved Kv4.3 but not Kv4.2 inhibition. Inhibition of Kv4.2 and Kv4.3 by KCNE3L was weaker, and its inhibition of Kv4.2 abolished by KChIP2. KCNE3L and KCNE4L also exhibited subunit-specific effects on Kv4 channel complex inactivation kinetics, voltage dependence and recovery. Further supporting the potential physiological significance of the robust functional effects of KCNE4L on Kv4 channels, KCNE4L protein was detected in human atrium, where it co-localized with Kv4.3. The findings establish functional effects of novel human cardiac-expressed KCNE isoforms and further contribute to our understanding of the potential mechanisms influencing cardiomyocyte repolarization.

Figure 1. Full-length human KCNE4 protein is expressed in human atrium.

(A) Predicted protein sequence for additional N-terminal residues of human KCNE3L (upper) and KCNE4L (lower) that are lacking in KCNE3S and KCNE4S. (B) Cartoon showing placement of N-terminal sequences from panel A at the N-terminal, extracellular end of the KCNE subunits (black portion). (C) Western blots probing expression of KCNE4L using an antibody raised to a peptide sequence specific to the long form of KCNE4. Bands were detected in lysates from human atria, and from CHO cells expressing KCNE4L, but not in those from CHO cells expressing KCNE4S. (D) Immunofluorescence images showing KCNE4L and KCNQ1 expression in human atrial myocytes. Upper, single-channel images; lower, merged images. Scale bar, 5 μm. Arrows, KCNQ1 with no co-localized KCNE4L. (E) Immunofluorescence images showing KCNE4L and Kv4.3 expression in human atrial myocytes. Upper, single-channel images; lower, merged images. Scale bar, 5 μm. Arrows, Kv4.3 co-localized with KCNE4L. (F) Immunofluorescence images showing negligible detection of KCNE4L and KCNQ1 in mouse ventricular myocytes. Upper, single-channel images; lower, merged images. Scale bar, 10 μm.

Figure 2. Functional effects of KCNE3L and KCNE4L on hERG, and human Kv1.1.

(A) Exemplar current traces recorded from Xenopus oocytes expressing hERG alone (n = 25) or with KCNE3L (E3L; n = 18), or KCNE4L (E4L; n = 10). Insets: left, Voltage clamp protocol; right, scale bars. Zero current level indicated by dashed line. (B) Mean ± SEM peak current/voltage relationship for currents as in panel A; n values as in panel A. ***P < 1 × 10⁻⁸. NS, P > 0.05 (C) Exemplar current traces recorded from Xenopus oocytes expressing Kv1.1 alone (n = 26) or with KCNE3L (E3L; n = 19), or KCNE4L (E4L; n = 30). Insets: left, Voltage clamp protocol; right, scale bars. Zero current level indicated by dashed lines. (D) Mean ± SEM peak current/voltage relationship for currents as in panel C; n values as in panel C. ***P < 0.005. (E) Mean ± SEM normalized G/V relationship measured at beginning of tail pulse for currents as in panel C; n values as in panel C.

Figure 3. Absence of functional effects of KCNE3L and KCNE4L on human HCN1.

(A) Exemplar current traces recorded from Xenopus oocytes expressing HCN1 alone (n = 13) or with KCNE3L (E3L; n = 17), or KCNE4L (E4L; n = 15). Insets: left, scale bars; right, Voltage clamp protocol. Zero current level indicated by dashed line. (B) Mean ± SEM instantaneous current/voltage relationship for currents as in panel A; n values as in (A). (C) Mean ± SEM peak current/voltage relationship for currents as in panel A; n values as in panel A. (D) Mean ± SEM normalized G/V relationship measured at beginning of tail pulse for currents as in panel A; n values as in panel A.

Figure 4. Effects of KCNE3L and hKCNE4L on human Kv4.2 current magnitude with/without KChIP2.

(A) Exemplar current traces recorded from Xenopus oocytes expressing human Kv4.2 alone (n = 10) or with human KCNE3L (E3L; n = 9), KCNE4L (E4L; n = 10), KCNE3S (E3S; n = 12), or KCNE4S (E4S; n = 10). Insets: left, scale bars; right, Voltage clamp protocol. Zero current level indicated by dashed line. (B) Mean ± SEM peak current/voltage relationship for currents as in panel A; n values as in panel A. See main text for statistics. (C) Exemplar current traces recorded from Xenopus oocytes expressing human Kv4.2 + KChIP2 alone (n = 12) or with human KCNE3L (E3L; n = 12) or KCNE4L (E4L; n = 10). Inset: scale bars. Zero current level indicated by dashed line. (D) Mean ± SEM peak current/voltage relationship for currents as in panel C; n values as in panel C. See main text for statistics.

Figure 5. Effects of KCNE3L and hKCNE4L on human Kv4.3 current magnitude with/without KChIP2.

(A) Exemplar current traces recorded from Xenopus oocytes expressing human Kv4.3 alone (n = 14) or with human KCNE3L (E3L; n = 13), KCNE4L (E4L; n = 13), KCNE3S (E3S; n = 11), or KCNE4S (E4S; n = 9) (5 ng cRNA per oocyte of each subunit injected) Insets: left, scale bars; right, Voltage clamp protocol. Zero current level indicated by dashed line. (B) Mean ± SEM peak current/voltage relationship for currents as in panel A; n values as in panel A. See main text for statistics. (C) Exemplar current traces recorded from Xenopus oocytes expressing human Kv4.3 + KChIP2 alone (n = 12) or with human KCNE3L (E3L; n = 12) or KCNE4L (E4L; n = 11) (5 ng cRNA per oocyte of each subunit injected). Inset: scale bars. Zero current level indicated by dashed line. (D) Mean ± SEM peak current/voltage relationship for currents as in panel C; n values as in panel C. See main text for statistics. Inset, Kv4.3 current augmentation at 24 h and 48 h expression when 1 ng Kv4.3 cRNA was co-injected with 1 ng KChIP2 cRNA (n = 10–13). **p < 0.01; ***p < 0.001.

Figure 6. Effects of KCNE3L and hKCNE4L on human Kv4.2/3 inactivation kinetics with/without KChIP2.

(A) Box plots showing individual and mean ± SEM values for τ of fast and slow components, and relative amplitude, of fast inactivation (double exponential fit) of Kv4.2 for currents recorded as in Fig. 4A; n = 6–12. *P < 0.05; **P < 0.01; by ANOVA followed by Tukey’s HSD test. Absence of brackets indicates P value versus all other groups. (B) Box plots showing individual and mean ± SEM values for τ of Kv4.2-KChIP2 inactivation (single exponential fit) for currents recorded as in Fig. 4C; n = 10–12. ***P < 0.0005. (C) Box plots showing individual and mean ± SEM values for τ of fast and slow components, and relative amplitude, of fast inactivation (double exponential fit) of Kv4.3 for currents recorded as in Fig. 5A; n = 6–11. *P < 0.05; **P < 0.01; by ANOVA followed by Tukey’s HSD test. Absence of brackets indicates P value versus all other groups. (D) Box plots showing individual and mean ± SEM values for τ of Kv4.3-KChIP2 inactivation (single exponential fit) for currents recorded as in Fig. 5C; n = 11–12. ***P < 1 × 10⁻⁷.

Figure 7. Effects of hKCNE3L and hKCNE4L on hKv4.2/3 steady-state inactivation and inactivation recovery with/without KChIP2.

(A) Upper, exemplar current trace recorded from Xenopus oocyte expressing hKv4.2+ KChIP2, using steady-state inactivation protocol (left inset). Tail current time-point used for quantifying available channels indicated by arrow. Dashed line indicates zero current level. Lower, mean ± SEM values for voltage dependence of steady-state inactivation (plotted as fraction of channels available at arrow in panel A versus voltage) quantified using Voltage protocol as in panel A, for currents generated in Xenopus oocytes expressing hKv4.2 alone (n = 6) or with hKCNE3L (E3L; n = 4) or hKCNE4L (E4L; n = 6); or Kv4.2+ KChIP2 alone (n = 6), or with hKCNE3L (E3L; n = 5) or hKCNE4L (E4L; n = 6). (B) Upper, exemplar current trace recorded from Xenopus oocyte expressing hKv4.3+ KChIP2, using steady-state inactivation protocol as in panel A. Tail current time-point used for quantifying available channels indicated by arrow. Dashed line indicates zero current level. Lower, mean ± SEM values for voltage dependence of steady-state inactivation (fraction of channels available at arrow in panel A versus voltage) quantified using Voltage protocol as in panel A, for currents generated in Xenopus oocytes expressing hKv4.3 alone (n = 6) or with hKCNE3L (E3L; n = 5); or hKv4.3 + KChIP2 alone (n = 9), or with hKCNE3L (E3L; n = 7). (C) Upper, exemplar current traces recorded using “short” and “long” inactivation recovery protocols (inset) from a Xenopus oocyte expressing hKv4.2 + KChIP2. Dashed line indicates zero current level. (D) Mean ± SEM values for the time course of recovery from inactivation for currents generated as in panel C in Xenopus oocytes expressing hKv4.2 alone (n = 6) or with hKCNE3L (E3L; n = 5) or hKCNE4L (E4L; n = 6); or Kv4.2 + KChIP2 alone (n = 6), or with hKCNE3L (E3L; n = 6) or hKCNE4L (E4L; n = 6). Left, close-up of first 1s of recovery; right, full 5s recovery period.