A privileged ER compartment for posttranslational heteromeric assembly of an ion channel

Abstract

Mechanisms underlying heterotypic subunit assembly of ion channels and other oligomeric complexes are poorly understood. In the human heart, heteromeric assembly of two isoforms encoded by the human ether-à-go-go related gene (hERG) is essential for the normal function of cardiac I_Kr in ventricular repolarization, with loss of hERG1b contributing to arrhythmias associated with long QT-syndrome (LQTS). While hERG1a homomers traffic efficiently to the plasma membrane, hERG1b homomers are retained in the endoplasmic reticulum (ER). When expressed together, the two subunits avidly associate during biogenesis. Seeking rules specifying heteromeric association, we characterized the fate of hERG1b proteins using confocal and superresolution imaging in fixed and live HeLa cells. We found hERG1b sequestered in punctate intracellular structures when expressed alone in HeLa cells. These puncta, which depend on the presence of an N-terminal "RXR" ER retention signal, represent a privileged ER subcompartment distinct from that containing ER-retained, type 2 (hERG-based) LQTS mutant proteins, which were rapidly degraded by the proteasome. Introducing hERG1a to cells with preformed hERG1b puncta dissolved these puncta by rescuing extant hERG1b. Rescue occurred by association of fully translated hERG1b with 1a, a surprising finding given previous studies demonstrating cotranslational heteromeric association. We propose that sequestration limits potentially deleterious surface expression of hERG1b homomeric channels while preserving hERG1b for an alternative mode of heteromeric hERG1a/1b channel assembly posttranslationally. These findings reveal a surprising versatility of biosynthetic pathways promoting heteromeric assembly.

Keywords: arrhythmia; condensate; hERG; long QT syndrome; protein trafficking.

Fig. 1.

hERG1b is sequestered in intracellular puncta. (A) Confocal images of fixed and immunostained HeLa cells transfected with hERG1a or (B) hERG1b plasmids. (C) Quantification of the number of puncta per cell; data are mean ± SD, n = 36 cells per condition, analyzed with Student’s t test (****P < 0.0001). (D) Confocal live cell images of iPSC-CM transfected with 1a-GFP or (E) 1b-GFP. (F) Quantification of number of puncta per cell; data are mean ± SD, n = 18 cells for 1a_GFP, 12 cells for 1b_GFP, analyzed with Student’s t test (****P < 0.0001); scale bar, 10 μm in large image and 1 μm in the Inset.

Fig. 2.

hERG1b puncta are localized to ER. (A–C) STED microscopy of iPSC-CM transfected with mRNAs encoding hERG1b_SNAP labeled with JFX554 (magenta) and ER marker Halo-KDEL labeled with JFX650 (cyan). (D) Quantification of colocalization events thresholded with three different distances between the center of hERG1b puncta and the surface of the ER; data are mean ± SD, n = 16 regions of interest (Methods). (E) Cartoon of ER localization of hERG1b puncta. The scale bar is 10 μm for (A) and 500 nm for (B and C).

Fig. 3.

hERG1b is sequestered in a privileged compartment. (A) Immunostaining of HeLa cells transfected with hERG1b_HA or (B) hERG1a-Y611H_HA. (C) quantification of the number of puncta per cell; data are mean ± SD, n = 33 for hERG1b, n = 26 for 1a-Y611H, analyzed with a Student t test (****P < 0.0001). (D) experimental design for the pulse–chase (en rule) experiment. (E) HeLa cells transfected with either hERG1a_SNAP, 1b_SNAP, 1a-Y611H_SNAP, or 1b-Y271H_SNAP imaged at various time points provided in (D) and sample images provided for 0 h and 8 h after labeling. (F) Quantification of normalized corrected total fluorescence intensity (CTCF) plotted across time; hERG1a-Y611H and hERG1b-Y611H data are fitted with single exponential decay; hERG1a, hERG1b data are plotted with connected lines; data are mean ± SEM, n = 35 to 45 cells per condition; data with individual data points are provided in SI Appendix, Fig. S4. (G) Quantification of the number of puncta per cell; data are mean ± SD, n = 31 for hERG1b, n = 32 for 1a-Y611H, analyzed with a Student t test (****P < 0.0001); Inset on the Upper Right of (E) shows the Halo-tag expression (yellow) used to identify the transfected cells. Scale bar, 10 μm for large images (A–E) and 1 μm for Inset (A and B).

Fig. 4.

An ER retention motif promotes sequestration of hERG1b in ERSA. (A) Confocal images of HeLa cells expressing FLAG-tagged hERG1b alone, (B) hERG1b with N¹⁵XN mutation, and (C) coexpression of WT hERG1a and hERG1b. (D) Quantification of the number of puncta per cell; data are mean ± SD (n = 30 cells per condition) analyzed with a one-way ANOVA (****P < 0.0001); scale bar, 10 μm in large image and 1 μm in Inset.

Fig. 5.

Are hERG1b puncta are condensates? (A) Confocal images of live HeLa cells expressing hERG1b (magenta) stained with Potomac-Yellow dye (cyan) labeling membranes; scale bar, 10 μm in the large image and 2 μm in the Inset (i and ii). (B) Quantification of percentage of number of hERG1b puncta colocalizing within 250 nm with Potomac-labeled membranous puncta (n = 15 cells). (C) Confocal images of live HeLa cell expressing hERG1b_GFP with photobleaching of individual puncta imaged before and after photobleaching; Scale bar, 5 µm in large image and 1 µm in Inset. (D) Quantification of fluorescence intensity of the photobleached puncta before and after photobleaching; data are mean ± SEM; n = 10 puncta from different cells. Data are from a representative experiment repeated three times. (E) GFP tagged hERG1b amino-terminal domain (1 to 64 residues) undergoes phase separation, with fluorescence recovery after photobleaching of individual puncta. (F) Quantification of normalized fluorescence intensity after photobleaching; data are mean ± SEM; n = 13 puncta from different cells in a representative experiment repeated three times. (G) Droplet-promoting regions in hERG1b identified by Fuzdrop, which includes the N-terminal domain. (H) Confocal image of HeLa cell expressing 1bNT^N15XN_GFP; scale bar, 10 μm in large image and 1 μm in Inset. (I) Graph showing % cells with puncta in 1bNT-WT and 1bNT-N₁₅XN mutant protein-expressing cells; 1bNT-WT exemplar is shown in E; n = 3 independent wells per condition, with an average of 50 cells in each well, data are mean ± SD, analyzed with a Student’s t test (***P < 0.001).

Fig. 6.

Sequestered hERG1b proteins can be rescued by newly expressed 1a. (A) Experimental design. (B) Mock-transfected cells showing hERG1b puncta. (C) 1a-transfected cells showing hERG1b signal in or near plasma membrane in live cell imaging. (D) Quantification of the number of puncta per cell. Data are mean ± SD, n = 35 to 42 cells per condition and analyzed with a Student’s t test (****P < 0.0001). Scale bar, 10 μm in large image and 1 μm in Inset.