Profiling the TRPV4 ankyrin repeat domain interactome and its disruption by neuromuscular disease-causing mutations

Abstract

The ankyrin repeat is one of the most abundant protein-protein interaction motifs in eukaryotes yet occurs in only a small number of ion channels. These channels are all members of the transient receptor potential (TRP) superfamily and contain prominent ankyrin repeat domains (ARDs) in their cytoplasmic N termini. In transient receptor potential vanilloid 4 (TRPV4), the importance of this domain has been highlighted by the finding that gain-of-function neuromuscular disease-causing missense mutations cluster on the ARD surface. Little is known currently about the extent of the TRPV4-ARD interactome, nor how it may be altered by disease-causing mutations. Here, we utilized a human proteome microarray to profile the ARD interactomes of WT and mutant TRPV4. Probing of the microarray with TRPV4^WT-ARD revealed 78 interactors, including proteins related to ubiquitination and small GTPase signaling, such as the ubiquitin ligase NEDD4L and the RhoGEF ARHGEF10. In parallel experiments, we also identified the deubiquitinase OTUB2 as an interactor of the proximal N terminus. Comparison of the ARD interactomes of WT and mutant TRPV4 revealed 21 interactions affected by disease-causing mutations. Strikingly, one of these interactors, ARHGEF10, is also mutated in neuromuscular disease. Cell-based studies confirmed that ARHGEF10 exhibits a reduced capacity to coimmunoprecipitate with mutant TRPV4. Furthermore, calcium imaging studies demonstrated that ARHGEF10 overexpression suppressed TRPV4^WT channel activity, but that this inhibition is abrogated by disease-causing mutations. Together, these findings provide insights into the functional roles of an ion channel ARD, as well as their disruption in disease, and offer a resource for future cell-based studies.

Keywords: GEF; TRP channels; calcium imaging; neurodegenerative disease; ubiquitination.

Figure 1

Identification of candidate interactors of the TRPV4 WT and mutant ARD utilizing a human protein microarray. A, schematic of a TRPV4 protomer with the key functional domains highlighted. The PIP₂-binding domain (PBD) and proline-rich domain (PRD) reside within the N-terminal intrinsically disordered region. B, schematic of the TRPV4 homotetramer with the transmembrane regions (green) and ARD (purple) of a single subunit highlighted. Red, the R269 and R315 residues mutated in TRPV4-mediated neuromuscular disease examined in the present study; pink, additional residues mutated in TRPV4-mediated neuromuscular disease; blue, the D333 residue mutated in TRPV4-mediated skeletal dysplasia (Homo sapiens, PDB: 8FC9). C, Coomassie blue-stained polyacrylamide gel showing the purified TRPV4 N terminus and ARD constructs. The R271P-ARD construct (dagger) was not examined further in the present study. D, schematic summarizing the experimental design adopted to identify potential interactors of the TRPV4 N terminus and ARD utilizing the HuProt Microarray. ARD, ankyrin repeat domain; TRPV4, transient receptor potential vanilloid 4.

Figure 2

Protein interactome of the TRPV4^WT N terminus. A, candidate interactors of the TRPV4^WT N terminus. Two previously identified TRPV4 interactors (PSD95 and PACSIN2) are identified by red asterisks. The blue double asterisks highlight the deubiquitinase OTUB2. The inset shows the duplicate protein spots for PSD95 (white arrowheads) following immunodetection of the V5 tag (red) present on the TRPV4^WT N-terminus peptide. Plot shows mean ± SD. B, immunoblots demonstrating coimmunoprecipitation of TRPV4^WT-GFP and PSD95-FLAG from transiently cotransfected HEK293T cells. TRPV4, transient receptor potential vanilloid 4.

Figure 3

Protein interactome of the TRPV4^WT-ARD. A, top 20 candidate interactors of the TRPV4^WT-ARD based on mean z-score. The remaining interactors are listed in Table 1. Plot shows mean ± SD. B–D, Summary of the cellular localization (B), protein domains (C), and molecular function (D) of the candidate interactors of the TRPV4^WT-ARD based on UniProt and Gene Ontology (GO) annotations and the supporting literature. FDR values indicating significant enrichment for GO categories utilizing the PANTHER database (B, GO:0005829, cytosol) or UniProt keywords utilizing the STRING database (C, KW-0175, coiled coil domain) are also shown. ARD, ankyrin repeat domain; GO, Gene Ontology; TRPV4, transient receptor potential vanilloid 4.

Figure 4

Novel interactors alter TRPV4 ubiquitination levels. A, coimmunoprecipitation of TRPV4-FLAG and HA-NEDD4L in transiently cotransfected HEK293T cells (n = 3 transfections/condition). B, TRPV4 is ubiquitinated in transfected HEK293T cells with overexpression of NEDD4L, but not with the catalytically inactive mutant NEDD4L-DD. C, densitometric quantification of B (n = 3 transfections/condition). Plot shows mean ± SD; one-way ANOVA, Tukey’s multiple comparison test; ns, not significant, ∗p = 0.012, ∗∗p = 0.007. D, coimmunoprecipitation of TRPV4-FLAG and OTUB2-V5 from transiently cotransfected HEK293T cells (n = 4 transfections/condition). E, TRPV4 is multiubiquitinated in cells without changing plasma membrane localization (30). Coexpression with OTUB2 results in deubiquitination. Shown are two independent samples per condition. F, densitometric quantification of E (n = 4 transfections/condition). Plot shows mean ± SD; unpaired two-sided t test, ∗∗p = 0.004. TRPV4, transient receptor potential vanilloid 4.

Figure 5

Neuromuscular disease-causing mutations alter the TRPV4 ARD interactome. A, Venn diagram of the candidate interactors of the TRPV4^WT-ARD, TRPV4^R269C-ARD, TRPV4^R315W-ARD, and TRPV4^D333G-ARD constructs. The 21 candidate interactors of the TRPV4^WT-ARD and TRPV4^D333G-ARD constructs that were not detected for the TRPV4^R269C-ARD and TRPV4^R315W-ARD constructs are listed in Table 2. B, molecular functions of the 21 candidate interactors absent from the microarrays probed with the TRPV4^R269C-ARD and TRPV4^R315W-ARD constructs based on UniProt and Gene Ontology (GO) annotations and the supporting literature. C and D, representative immunoblot images (C) and densitometry-based quantification (D) of coimmunoprecipitation experiments showing that neuromuscular disease-causing mutations in the TRPV4 ARD (R269C, R315W) reduced interactions with ARHGEF10. Plot shows mean ± SD; Kruskal-Wallis test, Dunn’s multiple comparison test, WT versus R269C, p = 0.043; WT versus R315W, p = 0.025. E, changes in the Fura-2 ratio (emission at 340/380 nm), an indicator of alterations in cytosolic free calcium, evoked in transfected MN-1 cells by application of hypotonic saline (vertical dotted line). N = 11 to 14 wells per condition, with >30 transfected cells per well. F and G, baseline cytosolic free calcium (F) and change in cytosolic free calcium (ΔCa²⁺) following hypotonic saline application (G) for the cells show in E. Plots show mean ± SD; F, unpaired two-sided t test with Welch’s correction, ns, not significant, ∗p = 0.022. G, unpaired two-sided t test; ns, not significant, ∗p = 0.043. ARD, ankyrin repeat domain; GO, Gene Ontology; TRPV4, transient receptor potential vanilloid 4.