Crosstalk between regulatory elements in disordered TRPV4 N-terminus modulates lipid-dependent channel activity

Abstract

Intrinsically disordered regions (IDRs) are essential for membrane receptor regulation but often remain unresolved in structural studies. TRPV4, a member of the TRP vanilloid channel family involved in thermo- and osmosensation, has a large N-terminal IDR of approximately 150 amino acids. With an integrated structural biology approach, we analyze the structural ensemble of the TRPV4 IDR and the network of antagonistic regulatory elements it encodes. These modulate channel activity in a hierarchical lipid-dependent manner through transient long-range interactions. A highly conserved autoinhibitory patch acts as a master regulator by competing with PIP₂ binding to attenuate channel activity. Molecular dynamics simulations show that loss of the interaction between the PIP₂-binding site and the membrane reduces the force exerted by the IDR on the structured core of TRPV4. This work demonstrates that IDR structural dynamics are coupled to TRPV4 activity and highlights the importance of IDRs for TRP channel function and regulation.

Fig. 1. Structural ensemble of the TRPV4 N-terminal domain.

a TRPV4 N-terminal constructs used for structural analyses. b–d Purified TRPV4 N-terminal constructs analyzed by Coomassie-stained SDS-PAGE b, SEC-MALS c, and CD spectroscopy d. SDS-PAGE in b comparing all constructs side by side was carried out once to evaluate sample purity and respective molecular weight. e [¹H, ¹⁵N]-TROSY-HSQC NMR spectrum of ¹⁵N-labeled TRPV4-IDR (see Supplementary Fig. 2 for backbone assignments). f, g SAXS pair-distance-distribution f and SAXS EOM (Ensemble Optimization Method) g, both in arbitrary units (arb. units), of TRPV4 N-terminal constructs (Supplementary Fig. 3). The real-space distance distribution yields a radius of gyration of R_g = 3.4 nm with a maximal particle dimension of D_max = 14.0 nm for the IDR, R_g = 4.1 nm and D_max = 19 nm for the NTD as well as R_g = 2.5 nm and a D_max = 11.5 nm for the ARD. Every protein exhibits levels of conformational heterogeneity and the p(r) profiles should be interpreted as the summed volume-fraction weighted contribution within the sample population, and not as single-particle distributions. The statistical analyses of the fit in g was carried out using the reduced χ² method (one-tailed distribution) and CorMap (one-tail Schilling distribution) test methods. The determined χ² and CorMap p values are indicated in the corresponding graph. h NTD ensemble refined by EOM (Ensemble Optimization Method)^,. Using a chain of dummy residues for the IDR and the X-ray structure of the TRPV4 ARD (PDB: 3W9G) as templates, a library of 10,000 NTD structures was generated and refined against the experimental data, allowing the comparison of the fitted versus the random pool and selecting a sub-set of ensemble-states representing the experimental data. Ten IDR conformers best representing the experimental scattering profile are depicted. Source data are provided as a Source Data file.

Fig. 2. Structural dynamics of the TRPV4 ARD.

a HDX of TRPV4 NTD and its isolated subdomains. Low (blue) to high (red) HDX shown for four time points. Areas without HDX assignment are colored white. For the ARD, HDX was visualized on the available X-ray structure of the G. gallus TRPV4 ARD (PDB: 3W9G). The six ankyrin repeats (AR) are indicated on top of the heat map diagram. b, c Root-mean-square fluctuations (RMSF) obtained from atomistic molecular dynamics (MD) simulations of the isolated G. gallus TRPV4 ARD in solution. RMSF at 42 °C mapped onto the ARD X-ray structure (PDB: 3W9G) b and RMSF per residue in simulations at increasing temperatures c. For comparison, HDX profiles after 10²s from a are displayed in the plot background. d, e RMSF of the ARD with respect to the central TRPV4 axis obtained from 1 µs long MD simulations of the complete TRPV4 core (see also Supplementary Movie 1). d Schematic depiction of the MD simulation setup. The channel principal axis (defined as z) is indicated as a dashed vertical line. The RMSF was calculated as the square root of the variance of the motion along this axis (∆z). e RMSF of TRPV4 residues 134–450 comprising the ARD. The solid line represents the average RMSF from all four protomers, the light area indicates the standard error of the mean. Source data are provided as a Source Data file.

Fig. 3. Long-range intra- and interdomain interactions of the TRPV4 NTD.

a, b Cross-linking mass spectrometry was used to probe interactions within and between IDR and ARD using either a the entire NTD or b isolated IDR (gray) and ARD (cyan) in a 1:1 ratio. Lysine residues are indicated by black tick marks, the PIP₂-binding site is marked light blue. Intradomain crosslinks are shown by curved lines in dark gray, interdomain crosslinks in light gray. c Heat map of C_ɑ-C_ɑ distances for an NTD conformational ensemble consisting of 15 EOM-refined conformers based on SAXS data of the NTD (Fig. 1h). Crosslinks are highlighted by white squares (NTD), black crosses (equimolar ARD:IDR mixture) or white squares filled with black crosses (both experimental set-ups). d TRPV4 N-terminal constructs used for tryptophan fluorescence (PBS: PIP₂-binding site, PRR: proline rich region). e, f Tryptophan fluorescence spectroscopy of TRPV4 N-terminal constructs (IDR, NTD, NTD^ΔN54 and NTD^ΔN97 lacking the first 54 or 97 amino acids, respectively, and IDR^ΔN97 (comprising PIP₂ binding site, surrounding basic residues and proline rich region) or isolated amino acid in buffer (Trp). Residue W109 in the PIP₂ binding site is the sole tryptophan residue in the entire NTD. Fluorescence intensity is presented in counts per second (cps). Bars represent the intensity weighted fluorescence emission wavelength <λ > (left axis). Data are presented as the mean value ± SEM from n = 3 individual experiments. The fluorescence emission maximum λ_max is shown by black circles connected through dotted lines (right axis). g, h ¹H chemical shift differences of W109 sidechain amide between IDR and IDR^ΔN97 as well as their respective counterparts harboring the PIP₂ binding site (¹⁰⁷KRWRR¹¹¹) mutation to ¹⁰⁷AAWAA¹¹¹. Source data are provided as a Source Data file.

Fig. 4. The PIP₂-binding site promotes compact IDR conformations.

a Constructs used in SEC-SAXS experiments. b Real-space pair-distance distribution functions, or p(r) profiles, in arbitrary units (arb. units), calculated for IDR and IDR^AAWAA (gray curves) as well as NTD and NTD^AAWAA (blue curves). p(r) functions were scaled to an area under the curve of 1. The real-space distance distribution of IDR^AAWAA yields a radius of gyration of R_g = 3.5 nm with a maximal particle dimension of D_max = 14.5 nm (native IDR: R_g = 3.4 nm, D_max = 14.0 nm). NTD^AAWAA has a R_g = 4.5 nm and a D_max = 19.5 nm (native NTD: R_g = 4.1 nm, D_max = 19.0 nm). c Fit between EOM-refined IDR and NTD models and experimental scattering data, in arbitrary units (arb. units). The statistical analyses of the fits were carried out using the reduced χ² method (one-tailed distribution) and CorMap (one-tail Schilling distribution) test methods. The determined χ² and CorMap p values are indicated in the corresponding graphs. d Comparison between R_g values of IDR and NTD variants between random pool structure library (solid area) and EOM refined models (dotted line). Source data are provided as a Source Data file.

Fig. 5. The distal N-terminus affects the structural NTD ensemble and TRPV4 channel activity.

a Topology of NTD truncations showing the charge distribution z (www.bioinformatics.nl/cgi-bin/emboss/charge) and sequence conservation (ConSurf) along the IDR. b Overall charge (z) in the IDR at physiological pH (7.4) depending on the IDR length (values determined with ProtPi, www.protpi.ch). c Normalized real-space distance distribution p(r), in arbitrary units (arb. units), of NTD and NTD deletion constructs. d Dimensionless Kratky plot of NTD and NTD deletion mutants. e Radius of gyration (R_g) and Stokes radius (R_S) determined from the real-space distance distribution in a and the SEC analysis (Supplementary Fig. 5c), respectively, plotted versus the number of IDR residues the NTD constructs. The maximum particle dimension (D_max) is plotted on the right y-axis. f N-terminal deletion mutants in the in vitro (G. gallus) and in cellulo (H. sapiens) systems. g Activation of TRPV4 constructs with the synthetic agonist GSK101 shows plasma membrane targeting and structural integrity of all constructs. Data are presented as mean values ± SEM from n = 6 biologically independent experiments, each with 10–30 cells per field of view. h Basal Ca²⁺ levels in MN-1 cells expressing different TRPV4 constructs. Data are presented as mean values ± SEM from n = 12 biologically independent experiments, each with 10–30 cells per field of view. The **** indicates a p value of p < 0.0001 (one-way ANOVA with Dunnett’s multiple comparison test). i Stimulation of Ca²⁺ flux by hypotonic saline at t = 20 s in MN-1 cells expressing different TRPV4 constructs. Data are presented as mean values ± SEM from n = 6 biologically independent experiments, each with 10–30 cells per field of view. Source data are provided as a Source Data file.

Fig. 6. A highly conserved patch in the N-terminal TRPV4 IDR transiently interacts with the C-terminal PIP₂ binding site and autoinhibits TRPV4 function.

a Chemical shift differences at high and low salt between ¹⁵N-labeled native IDR and IDR^AAWAA with a mutated PIP₂ binding site (PBS) shows that mutagenesis of the PBS leads to chemical shift changes in the conserved N-terminal patch. At the higher salt concentration (light gray), these chemical shift perturbations are significantly reduced. b Degree of conservation in TRPV4 IDR determined with ConSurf (Supplementary Fig. 11). c Chemical shift differences at high and low salt between ¹⁵N-labeled IDR and IDR^Patch with a mutation in the conserved N-terminal patch shows that mutagenesis leads to chemical shift changes in the PBS. At the higher salt concentration (light gray), these chemical shift perturbations are significantly reduced. d Relative peak intensity of IDR, IDR^AAWAA and IDR^Patch residues in the isolated IDR or in context of the ARD (i.e., NTD, NTD^AAWAA or NTD^Patch). All protein concentrations used were 100 µM. A value of 0.5 indicates that peak intensities for a respective IDR residue are halved when the ARD is present, a value of zero represents complete line broadening in the context of the NTD. Accordingly, lower values are indicative of IDR/ARD interactions. e, f Ca²⁺ imaging of hsTRPV4 variants expressed in MN-1 cells. e Basal Ca²⁺ and f hypotonic treatment at t = 20 s show increased activity of the patch mutant. For better comparison, data for TRPV4^ΔN68 are replotted from Fig. 5h, i. Data in e are presented as mean values ± SEM from n = 13 (TRPV4), 11 (TRPV4^ΔN68), 14 (TRPV4^Patch) and in f from n = 12 (TRPV4, TRPV4^ΔN68, and TRPV4^Patch) biologically independent experiments, each with 10–30 cells per field of view. Source data are provided as a Source Data file.

Fig. 7. Extensive lipid binding in the IDR is negatively affected by the distal N-terminus.

a Topology of N-terminal deletion mutants used for liposome sedimentation assay. b Protein distribution between pellet (“bound protein”) or supernatant fraction after centrifugation, quantified via densitometry of SDS-PAGE protein bands using imageJ. Data are presented as the mean value ± SEM from n = 3 individual experiments. c Distribution of charged and hydrophobic residues in the TRPV4-IDR shows a gradient of a consecutively more basic and hydrophobic protein from N- to C-terminus. Plotted with the PepCalc tool (https://pepcalc.com/). d, e NMR signal intensity differences for ¹⁵N-labeled IDR variants (100 µM) in the absence and presence of POPC (light gray circles) or POPC-POPG containing liposomes at low (filled yellow circles) or high salt concentration (open yellow circles). Higher values are indicative of lipid binding. Source data are provided as a Source Data file.

Fig. 8. The conserved N-terminal patch modulates lipid binding to the IDR.

a, b, c Chemical shift perturbation of ¹⁵N-labeled a IDR, b IDR^AAWAA, and c IDR^Patch titrated with short-chain PIP₂. Mutated regions are indicated in gray boxes, chemical shift changes are depicted by colored spheres, residues showing line broadening are highlighted by gray bars. d Average number of membrane contacts for each residue of the native IDR (dark gray), the IDR^AAWAA mutant (light gray) and the IDR^Patch mutant (mauve) on a lipid bilayer composed of POPC (69%), CHOL (20%), DOPS (10%), PIP₂ (1%) (Supplementary Table 4). The location of the N-terminal patch and the PIP₂-binding site (PBS) are highlighted by gray boxes. In the upper panel, contacts for all lipids, in lower panel, only contacts with PIP₂ are shown. Four replicate simulations per IDR sequence were carried out for 38 µs and contact averages were calculated from the last ~28 µs of each simulation. e ³¹P NMR spectra of diC₈-PIP₂ (light blue) with increasing amounts of IDR, IDR^ΔN97 or IDR^Patch. Chemical shift changes are indicated by arrows. f Chemical shift perturbations of P4 and P5 lipid headgroup resonances upon addition of IDR (gray), IDR^ΔN97 (blue), or IDR^Patch (mauve). Source data are provided as a Source Data file.

Fig. 9. PIP₂ binding to the TRPV4 IDR’s PIP₂-binding site exerts a pulling force on the ARD.

a Coarse-grained MD simulation system setup using a lipid bilayer membrane consisting of PIP₂ (1%, dark orange) as well as POPC (69%, dark gray), DOPS (10%, light gray) and cholesterol (20%, white). Headgroup phosphates are shown as orange spheres), the IDR (pink liquorice) was kept at defined distances from the membrane midplane by its most C-terminal residue V134 (blue sphere) to emulate anchoring by the ARD. The PIP₂-binding site (PBS) is highlighted in cyan. b Force displacement curves from restrained simulations of TRPV4 IDR, IDR^AAWAA and IDR^Patch. Four 38 µs replicate simulations were carried out for each condition. The mean restraint force is plotted against the mean distance between residue V134 and the membrane midplane. Dotted lines show linear fits of the force contribution of the PIP₂-binding site. Averages were calculated from the last ~28 µs of each of the 4 replicate simulations per IDR genotype and per height restraint. Error bars show the standard errors of the mean (SEM) of the replicate simulations. c Number of membrane lipid contacts for each residue of the native IDR at a given height restraint (for results with IDR^AAWAA and IDR^Patch, see Supplementary Fig. 10c). Averages are calculated from the last 28 µs of each of the four replicate simulations. d Composite figure of a structure of the native IDR (from an MD simulation at a restraint distance of 7 nm) and an AlphaFold model of the transmembrane core of the G. gallus TRPV4 tetramer. The force displacement curves in b indicate that the interaction of the PIP₂-binding site with the membrane exerts a pull force on the ARD N-terminus (solid arrow). Source data are provided as a Source Data file.

Fig. 10. The TRPV4 IDR forms an extensive ‘belt’ along the membrane plane and encodes a hierarchy of antagonistic regulatory modules.

a, b Superimposed IDR conformations from coarse-grained MD simulations (pink licorice) integrated into a full-length TRPV4 tetramer (AlphaFold prediction of G. gallus TRPV4 transmembrane core (gray) and ARDs (cyan)) viewed from the side a and from the intracellular side b. For better visualization of the extent of the IDR ‘belt’, the IDR conformations of the front facing TRPV4 monomer have been deleted. For a view of the IDR conformations on a single TRPV4 subunit, see Supplementary Fig. 13 as well as Supplementary Movies 2 and 3. c The TRPV4 N-terminus encodes multiple antagonistic regulatory elements that regulate TRPV4 function through ligand, protein, lipid or intra-domain contacts. d PIP₂-binding site interactions with the membrane exert a pull force on the IDR C-terminus. Likewise, pulling on the IDR can lead to membrane deformation. Crosstalk between the PIP₂-binding site and the autoinhibitory patch modulates PIP₂ binding and thus IDR membrane interactions, thereby influencing channel activity.