The phenotype of a KCNQ1 mutation depends on its KCNE partners: Is the cardiac slow delayed rectifier (I_Ks) channel more than a KCNQ1/KCNE1 complex?

I_Ks is important for cardiac action potential repolarization, especially under conditions when the β-adrenergic tone is high or when the rapid delayed rectifier (I_Kr) channels is suppressed¹. It has been well established that the I_Ks channel has at least two components: a pore-forming component (the KCNQ1 channel, also known as KvLQT1 or Kv7.1), and an auxiliary regulatory component (the KCNE1 subunit, also known as IsK or minK)^2,3. Things became more complicated when other members of the KCNE family (KCNE2 - KCNE5, also known as minK-related peptides or MiRP1 - MiRP4) were cloned⁴. In heterologous expression systems, KCNQ1 can associate with each of the KCNE2 - KCNE5 subunits. These partnerships lead to distinctively different channel phenotypes⁵. KCNQ1 association with KCNE4 or KCNE5 produces a slowly activating Kv channel with the voltage-range of activation markedly shifted in the positive direction, similar to the effect of KCNE1. On the other hand, KCNQ1 association with KCNE2 or KCNE3 produces largely constitutively active K channels that manifest little time- and voltage-dependence in activation. Furthermore, association of KCNQ1 with KCNE3 increases the current amplitude, similar to the effect of KCNE1, while KCNQ1 association with KCNE2 or KCNE4 has the opposite (current suppressing) effect. The transcripts of KCNE2 - KCNE5 subunits have been detected in human hearts^5-7, raising the possibility that cardiac myocytes may coexpress KCNQ1 with not only KCNE1, but also other KCNE family members. Under these conditions, does KCNQ1 preferentially associate with KCNE1, or does KCNQ1 randomly associate with different KCNE subunits? Can one KCNQ1 channel simultaneously associate with more than one type of KCNE subunits? If a KCNQ1 channel simultaneously binds multiple KCNE subunits that have different or opposite effects on the KCNQ1 channel function, what determines the final channel phenotype?

Among these new KCNE family members, a role of KCNE2 in human cardiac electrical activity seems most certain based on the linkage between KCNE2 mutations and LQT6^8,9. The initial proposal that KCNE2 serves as an auxiliary subunit of the I_Kr channel⁸ has been challenged¹⁰. We have explored the role of KCNE2 in cardiac I_Ks channel function, and observed the following¹¹: (a) KCNE2 is colocalized with KCNQ1 and KCNE1 in adult rat ventricular myocytes, which expressed a robust I_Ks current manifested as an HMR1556-sensitive current. (b) In both oocyte and COS-7 expression systems, coexpressing KCNQ1, KCNE1 and KCNE2 produces an I_Ks-like current that has a markedly reduced current amplitude than when KCNQ1 and KCNE1 are coexpressed in the absence of KCNE2. (c) An antibody targeting KCNE2 can coimmunoprecipitate KCNE1 with KCNE2 in the presence of KCNQ1, but not in its absence, indicating that KCNQ1 can simultaneously bind KCNE1 and KCNE2, forming a ternary complex. In conjunction with the voltage clamp data, we propose that in such a ternary complex, the gating modifying effect of KCNE2 on KCNQ1 (making it constitutively active) is masked by the opposing gating modifying effects of KCNE1 (shifting the activation curve in the positive direction and slowing activation). However, KCNE2 can still exert its current suppressing effect in the KCNQ1/KCNE1/KCNE2 channel complex. Therefore, KCNE2 may function as a kinetically silent I_Ks suppressor, reducing the I_Ks current amplitude without altering its gating kinetics.

In this issue of the Heart Rhythm journal, Lundby et al report a novel KCNQ1 mutation (Q147R) identified in an 82-year old female patient presenting permanent atrial fibrillation and marginal QT prolongation (QTc interval of 470 ms)¹². When the KCNQ1-Q147R mutant was expressed alone, it produced a Kv channel similar to that produced by the wild-type KCNQ1. However, when KCNQ1-Q147R was coexpressed with KCNE1, it produced an I_Ks current with a significantly smaller current amplitude than KCNQ1-WT/KCNE1 expressed at the same level. Interestingly, when KCNQ1-Q147R was coexpressed with KCNE2, it produced a considerably larger current than that produced by KCNQ1-WT/ KCNE2 at the same expression level. Importantly, this current-augmenting effect of Q147R was seen in the presence of KCNE1. Similar to our observations described above¹¹, Lundby et al showed that KCNQ1 coexpressed with both KCNE2 and KCNE1 produced an I_Ks current with a markedly reduced current amplitude than KCNQ1/KCNE1. It is likely that the Q147R mutation of KCNQ1 reduced the current-suppressing effect of KCNE2 in the KCNQ1/KCNE1/KCNE2 ternary complex.

Therefore, Q147R represents a ‘loss-of-function’ KCNQ1 mutation when coexpressed with KCNE1, but a ‘gain-of-function’ mutation when coexpressed with KCNE2, alone or with KCNE1. The structural basis for these differential effects is not clear. Position 147 is close to two other KCNQ1 positions where mutations have been linked to the short QT syndrome (SQT2) or familial atrial fibrillation (fAF): S140G and V141M^13,14. Neither mutations alter the KCNQ1 channel function when expressed alone, but produce constitutively active channels when coexpressed with KCNE1, i.e. the gating modifying effects of KCNE1 are drastically altered by the two mutations on KCNQ1. These observations indicate that direct or allosteric interactions between the extracellular S1-S2 linker of KCNQ1 and the associated KCNE1 and/or KCNE2 subunits are important for the KCNQ1/KCNE1 (some with additional KCNE2) channel function. Identifying the spatial relationship between the S1-S2 linker of KCNQ1 and associated KCNE1 and KCNE2 subunits will be useful for building structural models of KCNQ1/KCNEx channel complexes. These structural models can help us better understand what determine KCNQ1/KCNEx association, and what determine the final outcome when the KCNQ1 channel is simultaneously associated with multiple KCNE subunits that have different effects on the KCNQ1 function.

The study by Lundby shows us that the phenotype of KCNQ1 mutations can be very much dependent on its KCNE partner(s)¹². In turn, these observations suggest that the subunit composition in cardiac I_Ks channels may be more complicated than a KCNQ1/KCNE1 complex. This leads to new questions that need to be addressed in future experiments: (a) How are the multiple KCNE subunits expressed in different regions of the heart under normal and pathological conditions? (b) Is there a hierarchy in terms of the preference of KCNQ1 association with different KCNE subunits? (c) What determines the hierarchy in terms of dominance of KCNE effects on the KCNQ1 channel function? (d) How are the phenotypes of other KCNQ1 mutations identified in LQT1, SQT2 or fAF patients modified by the partner KCNE subunits?