Abstract 12145: Cell Surface Trafficking and Functional Deep Mutational Scans of the Potassium Channel Gene KCNE1

Abstract

Introduction: Loss of function variants in KCNE1 cause type 5 Long QT Syndrome. Deep mutational scanning is a highly multiplexed method to assay the function of thousands of genetic variants. Hypothesis: A deep mutational scan of KCNE1 will identify variants that disrupt KCNE1 function and may lead to Long QT Syndrome. Methods: A comprehensive library of nearly all possible KCNE1 variants was generated by inverse PCR. The library was integrated in a one-variant-per-cell format into HEK293 landing pad cells expressing the KCNE1 partner KCNQ1 . To assess surface trafficking, cells expressing an HA-tagged KCNE1 variant library were stained with an anti-HA antibody, FACS sorted into 4 equal bins, and deep sequenced. For each variant, a weighted average of variant counts across the 4 bins was calculated. To assess channel function, the KCNE1 library was introduced into landing pad cells expressing a gain of function S140G KCNQ1 variant, which causes KCNE1-dependent K+ efflux and cell death, for 12 days. Functional scores for each variant were calculated as a normalized ratio of initial and day 12 counts. Results: The KCNE1 library had 88% of 2,451 possible missense variants. The results of the 2 assays indicate that the KCNE1 C-terminus is dispensable for trafficking but not function. Missense variant trafficking scores (n=1,886) were normally distributed (0.93+/-0.37, mean+/-SD); 277 loss- and 183 gain-of-trafficking variants were identified. Missense variant functional scores were bimodally distributed. The functional assay identified loss of function variants with trafficking-deficient or trafficking-normal scores (likely gating variants). The assays identified structurally important regions, such as an N-terminal salt bridge, a transmembrane helix at residues 47-59 resistant to hydrophilic variants, and a cytoplasmic amphipathic helix at residues 90-114 resistant to hydrophobic variants. Manual patch clamp results for 4 previously studied and 4 novel variants correlated with deep mutational scanning scores. Conclusions: Two deep mutational scans of KCNE1 identified variants that disrupt channel or function. This dataset elucidates structure-function relationships and may help predict LQT5 variant pathogenicity.