High throughput screening identifies a trafficking corrector for long-QT syndrome associated KCNQ1 variants

Figures

• 

  Figure 1. A high-throughput screen to identify compounds that alter KCNQ1 expression and/or trafficking. (A) KCNQ1 construct (mycKCNQ1-mEGFP) used for high-throughput screening. (Created in BioRender: Clowes K, 2026, https://BioRender.com/jqaizjl.) (B) Workflow for high-throughput screening. (Created in BioRender: Clowes K, 2026, https://BioRender.com/ctyod9v.) (C) Representative data showing quantifications of cell surface KCNQ1 (left), total cellular KCNQ1 (right), and KCNQ1 trafficking ratio (bottom) for a screened 384-well plate. Each point represents the average value for all quantified cells in one well. Black, compound-treated well; teal, DMSO-treated negative control well; magenta, E115G mycKCNQ1-mEGFP positive control well. E115G is not shown in the trafficking ratio plot because of skewing of values from near background-level surface and total levels. FIU, fluorescence intensity units. (D) Screening funnel summarizing the steps of the process for narrowing from about 24,000 screened compounds to the top hit VU0494372.

• 

  Figure 2. VU0494372 increases cell surface and total KCNQ1 in a dose-dependent fashion. (A) Chemical structure of VU0494372. (B) Images of total (left) or cell surface (middle) mycKCNQ1-mEGFP in LLP-int cells treated with either 0.1% DMSO (top) as a control or 10 μM VU0494372 (bottom). Scale bars: 40 μm; scale is the same in all images. (C) Expression and trafficking of mycKCNQ1 in stable cell line, as determined by flow cytometry–based trafficking assay. Values were normalized to 0.1% DMSO control within each replicate. P values were determined with Wilcoxon’s matched-pairs signed rank test. N = 8. Error bars represent standard deviation. (D) Dose-response curves for mycKCNQ1 cell surface levels (left), total levels (middle), and trafficking efficiency (right), in VU0494372-treated stable cell line. Values were normalized to 0.2% DMSO control in each replicate. The best-fit curve was generated and EC50 values were determined with GraphPad Prism. The curve in the “total” plot was constrained to have a maximum of 130%, and the curve in the “trafficking” plot was constrained to have a maximum of 145% owing to large variability at high concentrations. N = 3–5. Error bars represent standard…

• 

  Figure 3. Treatment with VU0494372 increases cell surface levels and trafficking efficiency of LQT1-associated variants. Cell surface (A), total (B), and trafficking efficiency (C) data for WT, G179S, G189E, and V207M mycKCNQ1 in “population” stable cell lines, as determined by flow cytometry–based trafficking assays, after treatment with 0.2% DMSO or 20 μM VU0494372 for 16 hours. N = 4. Error bars represent standard deviation, and P values were determined with 2-tailed unpaired t tests.

• 

  Figure 4. VU0494372 effects do not involve modulation of KCNQ1 degradation, transcription, thermal stability, or cell toxicity. (A) Fold change in KCNQ1 mRNA levels, as determined by qPCR, relative to samples treated with 0.1% DMSO. N = 4. P values were calculated with Kruskal-Wallis tests with Dunn’s multiple-comparison test as follow-up. Error bars represent standard deviation. (B) Cycloheximide chase assay to determine the rate of KCNQ1 degradation. Left: Remaining KCNQ1 was detected with Western blotting. Representative full blots are shown in Supplemental Figure 7. Right: KCNQ1 band intensities were normalized to β-actin loading control bands and plotted as a percentage remaining from the 0-hour time point. N = 3 replicates, for which error bars represent standard deviation. Lines of best fit were generated with GraphPad Prism. (C) Left: Thermal aggregation curve from cellular thermal shift assay (CETSA) experiments. Bands corresponding to KCNQ1 were quantified and Tagg values were calculated for N = 3 (VU0494372-treated) or 5 (DMSO-treated) replicates. Curves of best fit were modeled with GraphPad Prism. Right: Tagg quantifications were calculated based on the inflection point of the…

• 

  Figure 5. Treatment with VU0494372 leads to an increase in WT and V207M KCNQ1-KCNE1 activity. (A–C) KCNQ1 channel function quantified in CHO-K1 cells stably expressing KCNE1 (CHO-KCNE1) and transfected with WT or variant KCNQ1 using automated patch clamp electrophysiology (n = 10–60 cells per group). Error bars represent standard deviation, and P values were determined with 2-tailed unpaired t tests. (A) Peak current density (left) recorded 1,900 milliseconds after the start of the voltage pulse, and V1/2 of activation (right) of WT KCNQ1 after 16-hour (chronic) treatment with 0.2% DMSO or 20 μM VU0494372. Compound was removed before recording (n = 18–35 cells were quantified from 2 separate experimental replicates). (B) Peak channel current density (top) and V1/2 of activation (bottom) for CHO-KCNE1 cells transiently transfected with WT or LQT1 variant G179S, G189E, or V207M KCNQ1 after chronic treatment with 0.2% DMSO or 20 μM VU0494372. Compound or DMSO was washed off before recording (n = 15–60 cells per group). N.D., not determined, due to low current amplitude. (C) Peak current density (left) and V1/2 of activation (right), as determined for CHO-KCNE1 cells transfected with WT KCNQ1.…

• 

  Figure 6.. Computational docking, electrophysiology data, and competition assays provide support for possible direct interaction between VU0494372 and KCNQ1. (A) Overlay of PDB 7XNK showing ML277 (magenta) bound to voltage sensor–active, pore-open KCNQ1 and the top 10 docked poses of VU0494372 (assorted colors) ranked by RosettaLigand interface score. VSD, voltage-sensing domain; PD, pore domain; CaM, calmodulin. (B) Zoomed and rotated view of A. Docked VU0494372 poses overlap with bound ML277 in KCNQ1. Labeled residues line the predicted pocket and interact with VU0494372. (C) Four poses from B of the most frequently observed orientation of VU0494372 docked to KCNQ1. (D) Competition between ML277 and VU0494372 as determined with a flow cytometry–based trafficking assay. Cell surface (left) and total (right) mycKCNQ1 levels in LLP-int cells were determined with and without ML277 and VU0494372. N = 6. P values were determined with Kruskal-Wallis non-parametric test with Dunn’s multiple-comparison test for follow-up. (E) Rotated and zoomed view of C to illustrate the orientation of VU0494372. (F) Cell surface levels (left), total levels (middle), and trafficking efficiency (right) of WT KCNQ1…