The proteostasis interactomes of trafficking-deficient variants of the voltage-gated potassium channel KV11.1 associated with long QT syndrome

Abstract

The voltage-gated potassium ion channel K_V11.1 plays a critical role in cardiac repolarization. Genetic variants that render Kv11.1 dysfunctional cause long QT syndrome (LQTS), which is associated with fatal arrhythmias. Approximately 90% of LQTS-associated variants cause intracellular protein transport (trafficking) dysfunction, which pharmacological chaperones like E-4031 can rescue. Protein folding and trafficking decisions are regulated by chaperones, protein quality control factors, and trafficking machinery comprising the cellular proteostasis network. Here, we test whether trafficking dysfunction is associated with alterations in the proteostasis network of pathogenic Kv11.1 variants and whether pharmacological chaperones can normalize the proteostasis network of responsive variants. We used affinity-purification coupled with tandem mass tag-based quantitative mass spectrometry to assess protein interaction changes of WT K_V11.1 or trafficking-deficient channel variants in the presence or absence of E-4031. We identified 572 core K_V11.1 protein interactors. Trafficking-deficient variants K_V11.1-G601S and K_V11.1-G601S-G965^∗ had significantly increased interactions with proteins responsible for folding, trafficking, and degradation compared to WT. We confirmed previous findings that the proteasome is critical for K_V11.1 degradation. Our report provides the first comprehensive characterization of protein quality control mechanisms of K_V11.1. We find extensive interactome remodeling associated with trafficking-deficient K_V11.1 variants and with pharmacological chaperone rescue of K_V11.1 cell surface expression. The identified protein interactions could be targeted therapeutically to improve K_V11.1 trafficking and treat LQTS.

Keywords: hERG; interactome; ion channel; long QT syndrome; mass spectrometry; trafficking.

Figure 1

Identification of K_V11.1 protein interactors.A, multiplexed affinity-purification mass spectrometry workflow for mapping K_V11.1 and variant protein interactions. B, bar graph illustrating the normalized abundance of K_V11.1 in TMT-labeled channels from replicates, with bars representing mean ± SD. C, volcano plots depicting K_V11.1 interactors identified via multiplexed AP-MS proteomics in HEK293 cells expressing WT K_V11.1 and two trafficking-deficient variants. The plots highlight significantly enriched proteins (shown in gray, q < 0.05 by multiple t-tests with false discovery rate correction using two-stage linear step-up of Benjamini, Krieger, and Yekutieli), further divided using manual gene ontology terms (N = 10) and color-coded for clarity. Key proteins like K_V11.1 (KCNH2, red) and other genes are labeled. HEK, human embryonic kidney; TMT, tandem mass tag.

Figure 2

Diagram of K_V11.1 protein trafficking and interaction networks. The cellular model highlights K_V11.1 protein interactions involved in (1) nuclear import/export, (2) translation, (3) early N-linked glycosylation, (4) folding chaperones in the endoplasmic reticulum, (5) ER-Golgi transport, (6) plasma membrane transport, (7) endosomal recycling, and lysosomal degradation, and (8) proteasomal degradation. Each cellular organelle or critical process is marked in orange and underlined.

Figure 3

Comparative protein interaction dynamics in K_V11.1 and variants.A, heatmap illustrating the 572 enriched protein interactions from K_V11.1 AP-MS experiments. Log₂ fold changes are scaled to WT treated with 0.1% DMSO and ranked from highest to lowest fold change for G601S. B, the top four enriched pathway terms identified by DAVID analysis for the 572 enriched K_V11.1 protein interactors, with the count of associated proteins for each GO term displayed within the bars. C, heatmap representing log₂ fold change from WT K_V11.1 for selective interactions, categorized and assessed using DAVID clustering to illustrate their potential functional significance and interconnectivity. D–I, dot plots illustrating quantified protein enrichment for individual genes, each dot representing a distinct biological replicate. In heat maps and dot plots, log₂ fold change is scaled to WT K_V11.1. Statistical significance was calculated using one-way ANOVA with post hoc Tukey’s multiple comparisons test and p values are depicted by ∗∗∗∗ < 0.0001. Bars represent mean ± SD. AP-MS, affinity-purification coupled with quantitative mass spectrometry; DAVID, database for annotation, visualization, and integrated discovery; DMSO, dimethyl sulfoxide; GO, gene ontology.

Figure 4

Cluster Analysis of trafficking-deficient K_V11.1 interactomes.A, hierarchical clustering of 572 protein interactors arranged by intensity similarity. Abundances for G601S and G965^∗ variants in relation to WT K_V11.1 were transformed into an Euclidian distance matrix and clustered using Ward’s minimum variance method. Each row represents an interactor with color indicating log₂ fold change normalized to WT. B, swarm plots for all clusters depicting the log₂ fold change normalized to WT for each variant and cluster. Statistical significance was calculated by ANOVA with post hoc Tukey’s multiple comparisons test and p values are depicted by ∗∗∗∗ < 0.0001. Bars represent mean ± SD. C–H, bar charts showing fold change compared to WT for individual genes within the top enriched protein GO term for each cluster. All samples were treated with vehicle (0.1%) for 24 h before sample collection. GO, gene ontology.

Figure 5

Impact of E-4031 on K_V11.1 trafficking and function.A, Western blot depicting K_V11.1 and variant protein expression in HEK293 cells in the absence or presence of E-4031 for 24 h. The truncated G965^∗ has a lower molecular weight of ∼100 kD than 135 kD of full-length Kv11.1. B, bar graph illustrating K_V11.1 trafficking efficiency quantified by the ratio of trafficked 155 kD glycosylated protein to total K_V11.1 protein, including untrafficked 135 kD K_V11.1 protein. Statistical significance was calculated using one-way ANOVA with post hoc Tukey’s multiple comparisons test and p values are depicted by ∗∗∗∗ < 0.0001. All points are from biological replicates performed in two blots. C, representative K_V11.1 current traces at −120 mV step from voltage-clamp experiments after 24-h treatment and washout. Inset: voltage-clamp protocol. K_V11.1 current was quantified as the peak inward current upon repolarization to −120 mV (gray box). D, bar graph comparing WT K_V11.1 currents after 24-h treatment and washout of vehicle or E-4031. Data analyzed with unpaired student’s t test. Each data point represents individual cells. E, bar graphs comparing WT K_V11.1 current and two trafficking-deficient variants after 24-h treatment with washout. All experiments follow a 24-h treatment with vehicle (0.1% DMSO) or E-4031 (10 μM). Patch clamp experiments included a 30-min washout to remove residual E-4031. Data are presented as mean ± SD. Statistical significance was calculated by Kruskal–Wallis with post hoc Dunn’s multiple comparisons test and p values are depicted by ∗∗∗∗ < 0.0001, ∗∗∗ = 0.0001. DMSO, dimethyl sulfoxide; HEK, human embryonic kidney.

Figure 6

Impact of chaperone treatment on protein interactions in trafficking-deficient K_V11.1 variants.A, heatmap and (B) violin plots showing comparison of log₂ fold changes over WT of 572 enriched protein interactions in trafficking-deficient variants G601S and G965^∗, with or without 24-h E-4031 treatment. Statistical significance was calculated by paired two-tailed t test, and p values are depicted by ∗∗∗∗ < 0.0001. C and D, correlation of proteasomal degradation protein interactions for (C) G601S and (D) G965^∗ in absence (x-axis) or presence (y-axis) of E-4031 (all log₂ fold changes normalized to WT with DMSO). The black dotted line indicates the normal line with a slope of 1 and intercept at the origin. The orange error bars from each individual protein represent the SEM for vehicle (0.1% DMSO, horizontal) and E-4031 (10 μM, vertical). E, heatmap showing the top five interactions from ten manually annotated GO terms sorted by largest difference in log₂ fold change/WT in absence or presence of E-4031 for 24 h. Only three proteins responsible for autophagy were enriched for K_V11.1 interaction in our dataset. Cells were treated with either vehicle (0.1% DMSO) or E-4031 (10 μM) for 24 h before experiments. DMSO, dimethyl sulfoxide; GO, gene ontology.