Multiubiquitination of TRPV4 reduces channel activity independent of surface localization

Abstract

Ubiquitin (Ub)-mediated regulation of plasmalemmal ion channel activity canonically occurs via stimulation of endocytosis. Whether ubiquitination can modulate channel activity by alternative mechanisms remains unknown. Here, we show that the transient receptor potential vanilloid 4 (TRPV4) cation channel is multiubiquitinated within its cytosolic N-terminal and C-terminal intrinsically disordered regions (IDRs). Mutagenizing select lysine residues to block ubiquitination of the N-terminal but not C-terminal IDR resulted in a marked elevation of TRPV4-mediated intracellular calcium influx, without increasing cell surface expression levels. Conversely, enhancing TRPV4 ubiquitination via expression of an E3 Ub ligase reduced TRPV4 channel activity but did not decrease plasma membrane abundance. These results demonstrate Ub-dependent regulation of TRPV4 channel function independent of effects on plasma membrane localization. Consistent with ubiquitination playing a key negative modulatory role of the channel, gain-of-function neuropathy-causing mutations in the TRPV4 gene led to reduced channel ubiquitination in both cellular and Drosophila models of TRPV4 neuropathy, whereas increasing mutant TRPV4 ubiquitination partially suppressed channel overactivity. Together, these data reveal a novel mechanism via which ubiquitination of an intracellular flexible IDR domain modulates ion channel function independently of endocytic trafficking and identify a contributory role for this pathway in the dysregulation of TRPV4 channel activity by neuropathy-causing mutations.

Keywords: Charcot–Marie–Tooth disease; NEDD4; TRPV4; channel activation; ion channel; plasma membrane; ubiquitination.

Figure 1

TRPV4 is monoubiquitinated/multiubiquitinated.A and B, ubiquitination of endogenous TRPV4 in HA-ubiquitin (Ub)-expressing 50B11 (A) and MDCK (B) cells, as detected by immunoprecipitation (IP) with anti-TRPV4 antibody and Western blot analysis with anti-HA antibody. C, Western blot analysis using anti-HA antibody of endogenous TRPV4 immunoprecipitated from the choroid plexus of HA-Ub mice; parallel IP from WT mice demonstrates the specificity of the anti-HA antibody. D, cell-based ubiquitination assay performed using HEK293T cells transiently transfected with TRPV4-FLAG. TRPV4 was immunoprecipitated from whole cell lysates with anti-FLAG antibody and Western blot analysis performed with anti-Ub antibody. E, cell-based ubiquitination assay utilizing HEK293T cells cotransfected with TRPV4-FLAG and either WT-HA-Ub or K0-HA-Ub, in the presence or the absence of the TRPV4-specific antagonist HC067 (0.5 μM). TRPV4 was immunoprecipitated from whole cell lysates with anti-FLAG antibody, and Western blot analysis was performed with anti-HA antibody. F, quantification of relative TRPV4 ubiquitination in the presence or the absence of HC067 from the experiment shown in E. n = 3 independent transfections. Data are presented as means ± SD. ∗p = 0.0422 (unpaired t test). G, cell-based ubiquitination assay performed using HEK293T cells transiently transfected with TRPV4-FLAG. TRPV4 was immunoprecipitated from whole cell lysates with anti-FLAG antibody, and Western blot analysis was performed with anti-Ub, anti-K48-linkage-specific polyubiquitin, or anti-K63-linkage-specific polyubiquitin antibodies. Arrows indicate ubiquitinated TRPV4 and arrowheads indicate polyubiquitinated TRPV4. HA, hemagglutinin; HC067, TRPV4-specific antagonist HC067047; HEK293T, human embryonic kidney 293T cell line; MDCK, Madin–Darby canine kidney; TRPV4, transient receptor potential vanilloid 4; TRPV4-FLAG, FLAG-tagged TRPV4.

Figure 2

TRPV4 is ubiquitinated within its N-terminal intrinsically disordered region.A, schematic of the double IP workflow utilized to purify ubiquitinated TRPV4 from HEK293T cells cotransfected with TRPV4-FLAG and HA-ubiquitin. TRPV4 was initially immunoprecipitated using anti-FLAG antibody (FLAG IP fraction), and ubiquitinated TRPV4 was then specifically isolated by IP with anti-HA antibody (HA IP fraction). The image at far right shows the two immunoprecipitated fractions as well as TRPV4 that did not immunoprecipitate with anti-HA antibody (HA unbound), on a Coomassie blue–stained gel; open arrow indicates unmodified TRPV4, solid arrow indicates ubiquitinated TRPV4. B, linear schematic of a TRPV4 monomer with functional domains highlighted: the N-terminal IDR (IDR) includes the PBD and PRR, which precedes the ARD. The CTD follows the TRP and contains the CAM. The vertical black lines show the positions of all 45 lysine residues in the TRPV4 protein sequence; green circles indicate ubiquitinated residues identified by mass spectrometry. C, representative mass spectrum demonstrating TRPV4 ubiquitination at K130. D, model of a TRPV4 tetramer, with a single monomer highlighted (Protein Data Bank ID: 6BBJ). Lysines within the N-terminal IDR are identified as spheres, with ubiquitinated residues indicated in green nonubiquitinated residues in gray. The PBD is shown in red, and the PRR is shown in purple. ARD, ankyrin repeat domain; CAM, calmodulin-binding domain; CTD, C-terminal IDR; HA-ubiquitin, hemagglutinin-tagged ubiquitin; HEK293T, human embryonic kidney 293T cell line; IDR, intrinsically disordered region; IP, immunoprecipitation; PBD, PIP₂-binding domain; PRR, proline-rich region; TRP, TRP box domain; TRPV4, transient receptor potential vanilloid 4; TRPV4-FLAG, FLAG-tagged TRPV4.

Figure 3

Genetically blocking ubiquitination of the N-terminal IDR results in increased TRPV4 channel activity.A, schematic of a TRPV4 monomer indicating the location of the V5 tag in the first extracellular loop (magenta) and C-terminal EGFP tag (green) of the TRPV4-V5-EGFP construct. B and C, representative confocal images (B) and quantification (C) of unpermeabilized HEK293T cells transfected with TRPV4-V5-EGFP (WT, 5K-IDR, and 5K-CTD), showing cell surface (magenta; anti-V5 antibody) and total (green; EGFP tag) TRPV4 as well as nuclei (blue, DAPI) in the absence or the presence of HC067 (0.5 μM). The scale bar represents 10 μm. n = 69 to 90 cells/condition representing three coverslips from three independent transfections. Data are presented as means ± SD. ∗∗∗∗p < 0.0001 (one-way ANOVA with Dunnett’s post hoc test), ####p < 0.0001 (two-way ANOVA). D, Fura-2 ratiometric calcium imaging of HEK293T cells transfected with TRPV4-V5-EGFP constructs and cultured in HC067 (0.5 μM) for 24 h prior to antagonist washout at time 0. EGFP-positive cells for each condition are shown at left, and Fura-2 ratio images (warmer colors represent more intracellular calcium) prior to and following antagonist withdrawal are shown at right. The scale bar represents 10 μm. E, change in Fura-2 ratio (340/380) over the time course of antagonist withdrawal. The vertical dotted line indicates the time point at which HC067 was removed. n = 24–36 cells/condition from three independent coverslips. Solid lines represent means, and shading SEM. F, quantification of the maximum change in intracellular calcium from baseline to 60 min after HC067 washout for the experiments shown in E. n = 24 to 36 cells/condition from three independent transfections. Data are presented as means ± SD. ∗∗∗∗p < 0.0001 (one-way ANOVA with Dunnett’s post hoc test). G, calcium imaging of transfected HEK293T cells cultured in HC067 (0.5 μM) for 24 h prior to antagonist washout at time 0 and immediate switching to calcium-free buffer (vertical dotted line at left). Calcium was then reintroduced to the imaging buffer after 1 h (vertical dotted line at right). n = 31–38 cells/condition from two independent coverslips. Data are presented as means ± SEM. H, calcium imaging of transfected HEK293T cells cultured in HC067 (0.5 μM) for 24 h prior to antagonist washout (vertical dotted line) and application of the TRPV4-specific agonist GSK101 (10 nM; horizontal black line). Solid lines represent means, and shading represents SEM. I, quantification of the maximum change in intracellular calcium following GSK101 application for the experiments shown in H. n = 50 to 103 cells/condition from three independent coverslips. Data are presented as means ± SD. ∗p = 0.0455, ∗∗∗∗p < 0.0001 (one-way ANOVA with Dunnett’s post hoc test). J and K, representative confocal images (J) and quantification (K) of TRPV4 localization prior to and following HC067 removal. Cell surface and total TRPV4 were detected as in B. n = 29 to 135 cells/condition from two to three independent coverslips. Data are presented as means ± SD. ∗∗∗p = 0.0004, ∗∗p = 0.0023 (two-way ANOVA with Dunnett’s post hoc test). The scale bar in J represents 10 μm. CTD, C-terminal IDR; DAPI, 4′,6-diamidino-2-phenylindole; EGFP, enhanced GFP; GSK101, GSK1016790A; HC067, TRPV4-specific antagonist HC067047; HEK293T, human embryonic kidney 293T cell line; IDR, intrinsically disordered region; TRPV4, transient receptor potential vanilloid 4.

Figure 4

TRPV4 is ubiquitinated by NEDD4 family E3 ubiquitin ligases.A, coimmunoprecipitation of TRPV4-FLAG with Myc-NEDD4 and Myc-ITCH as well as their catalytically inactive (CI) mutants, in transiently transfected HEK293T cells. B, TRPV4 ubiquitination (red arrows) in transfected HEK293T cells with coexpression of NEDD4 or ITCH or their CI mutants. C, densitometry-based quantification of B. n = four transfections/condition. Data are presented as means ± SD. ∗∗∗∗p < 0.0001, ∗∗p = 0.003 (one-way ANOVA with Dunnett’s post hoc test). D, quantification of TRPV4 abundance in the presence or the absence of E3 ligase coexpression. n = 4 to 8 independent experiments. Data are presented as means ± SD. ns (one-way ANOVA Dunnett’s multiple comparison test). HEK293T, human embryonic kidney 293T cell line; ns, not significant; TRPV4, transient receptor potential vanilloid 4; TRPV4-FLAG, FLAG-tagged TRPV4.

Figure 5

Enhancing N-terminal IDR ubiquitination with NEDD4 overexpression reduces stimulated TRPV4 channel activity.A, calcium imaging of HEK293T cells expressing either empty vector (EV), WT TRPV4 alone, or with either NEDD4 or NEDD4-CI, showing responses to stimulation with GSK101 (10 nM; applied at the vertical dotted line). n = 153–189 cells/condition representing 9–10 individual coverslips. Solid lines represent means, and shading represents SEM. B, quantification of the maximum change in intracellular calcium with GSK101 application for the experiments shown in A. n = 153 to 189 cells/condition. Data are presented as means ± SD. ∗∗∗∗p < 0.0001, ∗∗p = 0.0022 (one-way ANOVA with Dunnett’s post hoc test). C, latency of the calcium response to GSK101, as defined by the first time point at which Fura-2 ratio is greater than 5 SD above baseline levels. n = 76 to 186 cells/condition. Data are presented as means ± SD. ∗∗∗∗p < 0.0001 (one-way ANOVA with Dunnett’s post hoc test). D, percent of cells responding to GSK101 application with an increase in Fura-2 ratio greater than 5 SD above baseline levels. n = 9 to 10 coverslips/condition. Data are presented as means ± SD. ∗∗∗∗p < 0.0001, ∗p = 0.0161 (one-way ANOVA). E, quantification of the maximum change in intracellular calcium levels following stimulation with hypotonic saline. Data are presented as means ± SD. n = 42–56 cells/condition from three independent coverslips. ∗∗∗p = 0.0004, ∗p = 0.0315, ∗∗∗p = 0.0006. F, calcium imaging of 50B11 cells cotransfected with GFP and either NEDD4 or NEDD4-CI. GFP-positive cells for each condition are shown at left and indicated by red arrows. Fura-2 ratio images at time 0 and 5 min after GSK101 (10 nM) stimulation are shown at right. The scale bar represents 10 μm. G, time course of 50B11 calcium responses following GSK101 application (10 nM; applied at vertical dotted line). Solid lines represent means, and shading represents SEM. H, quantification of the change in intracellular calcium following GSK101 application for the experiments shown in G. n = 28 to 46 cells/condition from four coverslips representing four independent transfections. Data are presented as means ± SD. ∗p = 0.0137 (one-way ANOVA with Dunnett’s post hoc test). I–N, WT, 5K-IDR, or 5K-CTD TRPV4 was transfected into HEK293T cells without or with NEDD4 coexpression. Cells were cultured in HC067 (0.5 μM) for 24 h prior to antagonist washout and immediate switching to media containing GSK101 (10 nM; vertical dotted line). I and J, show transfection of WT TRPV4. n = 72 to 92 cells/condition, ∗∗p = 0.0035 (two-tailed unpaired t test). K and L, show transfection of 5K-IDR TRPV4. n = 74 to 101 cells/condition, ns (two-tailed unpaired t test). M and N, show 5K-CTD TRPV4-transfected cells. n = 84 to 85 cells/condition, ∗p = 0.0311 (two-tailed, unpaired t test). All histograms are presented as means ± SD. CI, catalytically inactive; CTD, C-terminal IDR; GSK101, GSK1016790A; HC067, TRPV4-specific antagonist HC067047; HEK293T, human embryonic kidney 293T cell line; IDR, intrinsically disordered region; ns, not significant; TRPV4, transient receptor potential vanilloid 4.

Figure 6

NEDD4 overexpression does not reduce TRPV4 surface localization.A, representative confocal images of HEK293T cells cotransfected with TRPV4-V5-EGFP and either Myc-NEDD4 or Myc-NEDD4-CI. Unpermeabilized cells were first stained with V5 antibody to identify surface-localized TRPV4 (magenta) and then fixed and permeabilized to stain for intracellular Myc-tagged NEDD4 (red). The scale bar represents 10 μm. B, quantification of relative surface fluorescence of TRPV4 from experiments shown in A. n = 51 to 86 cells/condition representing three independent experiments. Data are presented as means ± SD (one-way ANOVA with Dunnett’s post hoc test). C and D, Western blot analysis of TRPV4 cell surface biotinylation assays in transfected HEK293T cells showing changes in TRPV4 surface localization after GSK101 (10 nM) stimulation in the absence (C) or presence (D) of NEDD4 coexpression. E–H, TIRF imaging of HEK293T cells transfected with either WT or M680K (pore-inactive) TRPV4-EGFP and stimulated by GSK101 (10–100 nM) application. Cells were loaded with the calcium indicator Cal590 prior to imaging to allow for simultaneous assessments of TRPV4 surface localization and intracellular calcium levels. Representative TIRF images prior to and following GSK101 application are shown in E; warmer colors represent increased levels of intracellular calcium (Cal590) or increased TRPV4 localization within the TIRF evanescent (TRPV4-EGFP). The scale bar represents 10 μm. Quantification of the fold changes (ΔF/F) in intracellular calcium and TRPV4 surface localization following GSK101 application are shown in F and G, respectively. Solid lines represent means, and shading represent SEM. n = 9–13 cells/condition from four independent experiments. H, quantification of surface TRPV4 fluorescence in the TIRF evanescent field 10 min after treatment with GSK101. Data are represented as means ± SD. n = 9 to 13 cells/condition from four independent experiments. ∗∗∗∗p < 0.0001 (one-way ANOVA with Dunnett’s multiple comparison test). I–L, TIRF imaging of HEK293T cells transfected with WT TRPV4-EGFP alone or with NEDD4 and stimulated by GSK101 (10 nM) application. Representative TIRF images prior to and following GSK101 application are shown in I. The scale bar represents 10 μm. Quantification of the fold changes (ΔF/F) in intracellular calcium and TRPV4 surface localization following GSK101 treatment are shown in J and K, respectively. Solid lines represent means, and shading represent SEM. n = 7 cells/condition from four independent experiments. L, quantification of the changes in TRPV4-EGFP TIRF signal for the experiments shown in K, examining the effects of NEDD4 coexpression on WT TRPV4 surface localization in response to 10 nM GSK101 treatment. Data are presented as means ± SD. n = 7–8 cells/condition from four independent experiments. Unpaired two-tailed t testK-. M, representative images of HEK293T cells expressing either WT or 5IDR TRPV4 and their AAWAA counterparts. Unpermeabilized cells were stained with V5 antibody to identify surface-localized TRPV4 (magenta), and total TRPV4 was visualized by the EGFP tag (green). The scale bar represents 10 μm. N, quantification of relative surface fluorescence of TRPV4 from the experiments shown in M. n = 51 to 76 cells/condition representing two independent experiments. Data are presented as means ± SD. ∗p = 0.0106, ∗∗p = 0.0032, ∗∗∗∗p < 0.0001 (one-way ANOVA with Sidak’s multiple comparison test). O, change in Fura-2 ratio (340/380) over the time course of antagonist withdrawal. The vertical dotted line indicates the time point at which HC067 was removed. n = 51–76 cells/condition from three independent coverslips. Solid lines represent means, and shading represents SEM. P, quantification of the maximum change in intracellular calcium from baseline to 60 min after HC067 removal for the experiments shown in O. n = 51 to 76 cells/condition from four independent transfections. Data are presented as means ± SD. ∗∗∗∗p < 0.0001, ∗∗∗p = 0.0006, ∗p = 0.0483 (one-way ANOVA with Sidak’s post hoc test). EGFP, enhanced GFP; GSK101, GSK1016790A; HC067, TRPV4-specific antagonist HC067047; HEK293T, human embryonic kidney 293T cell line; TIRF, total internal reflection fluorescence; TRPV4, transient receptor potential vanilloid 4.

Figure 7

Neuropathy-causing TRPV4 mutants exhibit reduced ubiquitination and a gain-of-channel function, which can both be partially rescued by NEDD4 overexpression.A and B, representative Western blot analysis (A) and quantification (B) demonstrating reduced ubiquitination of R269C TRPV4 relative to WT channel in transfected HEK293T cells. Ubiquitinated TRPV4 was isolated as outlined in Figure 2A. n = 4 experiments/condition. Data are presented as means ± SD. ∗∗p = 0.0003 (unpaired two-tailed t test). C, tandem mass tags (TMTs) were used to quantify relative ubiquitination of specific lysine residues on either WT or R269C TRPV4 in transfected HEK293T cells. n = 4 biological replicates/condition. Data are represented as means ± SD. ∗∗∗p < 0.0009 (two-way ANOVA with Šídák's multiple comparisons test). D, cell-based ubiquitination assay utilizing T-Rex-TRPV4 cells in which TRPV4 expression was induced over 18 h, demonstrating increased ubiquitination of the R269C mutant in the presence of HC067 (0.5 μM). E and F, intracellular calcium levels of induced T-Rex-TRPV4 cells cultured in HC067 (0.5 μM) for 18 h prior to antagonist washout at time 0 (vertical dotted line in F). n = 10 independent coverslips per time point. Data are presented as means ± SD. ∗p < 0.05, ∗∗∗p = 0.0007 (two-way ANOVA with Šídák's multiple comparisons test). G, In vivo ubiquitination assay demonstrating reduced ubiquitination of R269C TRPV4 in transgenic Drosophila. Lysates from nontransgenic Drosophila (W1118) were processed in parallel as controls. H, densitometry-based quantification of G. n = 3 independent experiments. Data are presented as means ± SD. ∗∗p = 0.0089 (one-way ANOVA with Dunnett’s post hoc test). I, cell-based ubiquitination assay performed using transfected HEK293T cells, demonstrating that NEDD4 coexpression increases ubiquitination of R269C TRPV4. J, intracellular calcium levels measured for 1 h following withdrawal of HC067 (0.5 μM; vertical dotted line) from R269C TRPV4-expressing cells transfected with or without NEDD4. n = 76 cells/condition representing three independent experiments. Solid lines represent means, and shading represents SEM. K, quantification of the increase in intracellular calcium from baseline to 60 min after HC067 removal for the experiments shown in J. Data are presented as means ± SD. ∗∗p = 0.0069 (unpaired two-tailed t test). L, time to response quantified from J. n = 71 to 74 cells/condition. ∗∗∗∗p < 0.0001 (unpaired two-tailed t test). M, calcium imaging of HEK293T cells expressing R269C TRPV4 without or with NEDD4 coexpression, showing responses to application of GSK101 (10 nM) immediately following HC067 removal (vertical dotted line). n = 51–57 cells/condition from three independent transfections. Solid lines represent means, and shading represents SEM. N and O, quantification of the increase in intracellular calcium from baseline to 5 min after GSK101 treatment (N) and time to response (O) for the experiment shown in M. n = 46 to 52 cells/condition. Data are presented as means ± SD. ∗∗∗∗p < 0.0001 (unpaired two-tailed t test). GSK101, GSK1016790A; HC067, TRPV4-specific antagonist HC067047; HEK293T, human embryonic kidney 293T cell line; TRPV4, transient receptor potential vanilloid 4.