Structural Basis for Differential KCNQ1 Interactions with KCNE1 and KCNE2 in the Extracellular Juxtamembrane Region

Abstract

KCNE1 and KCNE2 are both expressed in human heart and can associate with KCNQ1. KCNE1 and KCNE2 share transmembrane topology and sequence homology, yet they differ in KCNQ1 modulation (KCNE1 slows KCNQ1 Closed-to-Open transition, while KCNE2 slows KCNQ1 Open-to-Closed transition) and pharmacology (KCNQ1/KCNE2 is 5 and 10 fold less sensitive than KCNQ1/KCNE1 to niflumic acid, IKs activator, and azimilide, IKs suppressor). Previous work has shown that the extracellular juxtamembrane (EJM) region of KCNE1 interacts with the extracellular surface of KCNQ1 to modulate gating kinetics and to form IKs activator binding sites. This prompts us to compare the EJMs of the two KCNE subunits in terms of KCNQ1 interaction. We apply cysteine (Cys) scanning mutagenesis to EJMs of KCNE1 and KCNE2 and analyze the patterns of functional perturbation when coexpressed with KCNQ1. We use methanethiosulfonate (MTS) reagents to probe the relationship between EJMs of KCNE subunits and KCNQ1. Finally we probe disulfide formation between Cys engineered into the EJMs of KCNE subunits and those engineered into extracellular surface of KCNQ1. The EJM of KCNE1 makes frequent contacts with KCNQ1, so that MTS modification of exposed Cys side chains can affect channel gating. KCNQ1/KCNE2 is largely indifferent to MTS modification of exposed Cys side chains engineered to EJM of KCNE2. However, MTS can slowly access 7 consecutive hydrophobic positions in the beginning of KCNE2 transmembrane domain, as if there is a crevice between KCNQ1 and KCNE2. Disulfide trapping experiments suggest that KCNE2 is leaning more toward KCNQ1 S2 than KCNE1. We propose that KCNE2 interferes with S2-S4 interactions during KCNQ1 Open-to-Closed transition and thus slows deactivation. This weakens sensitivity to niflumic acid and azimilide by affecting binding site directly or by an allosteric mechanism.