Transient receptor potential vanilloid (TRPV) channels are activated by ligands and heat and are involved in various physiological processes. In contrast to the architecturally related voltage-gated cation channels, TRPV1 and TRPV2 subtypes possess another activation gate at the selectivity filter that can open widely enough to permeate large organic cations. Despite recent structural advances, the mechanism of selectivity filter gating and permeation for both metal ions and large molecules by TRPV1 or TRPV2 is not well known. Here, we determined two crystal structures of rabbit TRPV2 in its Ca2+-bound and resiniferatoxin (RTx)- and Ca2+-bound forms, to 3.9 Å and 3.1 Å, respectively. Notably, our structures show that RTx binding leads to two-fold symmetric opening of the selectivity filter of TRPV2 that is wide enough for large organic cation permeation. Combined with functional characterizations, our studies reveal a structural basis for permeation of Ca2+ and large organic cations in TRPV2.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Data were collected at Northeastern Collaborative Access Team (NE-CAT) beamline 24-ID-C and Southeast Regional Collaborative Access Team (SER-CAT) beamline 22-ID in the Advanced Photon Source. We thank Yang Zhang for assistance with the imaging experiments and data analysis. This work was supported by NIH R35 NS097241 (S.-Y.L.), NIH R00 NS086916 (H.Y.) and NIH DP2 GM126898 (H.Y.). Beamline 24-ID-C is funded by P41GM103403 and S10 RR029205 and APS is supported by the U.S. Department of Energy under Contract No. W-31-109-Eng-38.
Integrated supplementary information
Red arrows mark N- and C-terminal deletions, and red shaded box indicates deletion in the turret for the TRPV2 construct (miTRPV2) used for structural studies. Asterisks indicate point mutations introduced to render TRPV2 sensitive to RTx. ARD, Ankyrin Repeat Domain; PH, pore helix; SF, selectivity filter; PL, pore loop; CTD, C-terminal Domain.
a,b, Representative whole-cell recordings from HEK293T cells expressing TRPV2QM (n = 14), miTRPV2QM (n = 9) and miTRPV2 (n = 3) in response to RTx (repeated voltage ramps from −60 mV to +60 mV at 5 s intervals). The blue bar indicates application of 50 nM RTx, and the red bar 200 μM ruthenium red. a stands for basal currents, b for stationary RTx-evoked currents, and c for RuR-inhibited currents. c-d, Representative whole-cell recordings of TRPV2 WT (n = 4), TRPV2QM (n = 12) and miTRPV2QM (n = 5) in response to 2-APB (c) from repeated voltage ramps from -120 mV to + 120 mV at 5 s intervals (d). The red bar indicates application of 2 mM 2-APB. Characters: a stands for basal currents, b for stationary 2-APB-induced currents, and c for currents after 2-APB washout. e-f, Representative RTx-evoked current traces of TRPV2QM and miTRPV2QM from inside-out recordings from control experiments (e) or in the presence of extracellular 200 μM RuR (pipette solution) (f). g, Quantification of inhibition of inward currents by RuR as quantified by the ratios of peak current amplitudes at -120 mV and + 120 mV. The number of patches is noted in parentheses. Data are presented as mean ± s.e.m. (****p < 0.0001, two-tailed unpaired Student’s t-test. p-values for TRPV2QM and miTRPV2QM are both p < 0.0001). h, YO-PRO-1 dye uptake by HEK293T cells expressing wtTRPV2 (n = 7 cells), TRPV2QM (n = 10 cells) and non-transfected cells (n = 8 cells). YO-PRO-1 uptake was triggered by application of 2 mM 2-APB (black arrow).
a,b, Anomalous difference Fourier maps for selenium was calculated to 5 Å and contoured to 1.5 σ. a, Selenium peaks in the tetramer. Methionine residues are shown in stick representation. b, Close-up views of the ARD. c-g, 2Fo-Fc electron density (light blue mesh, contoured at 1 σ) in the channel (blue, cartoon and stick representation). Some side chains are labelled for orientation. c, Segments of the ankyrin repeat domain. d, The TRP domain and the CTD. e, Helices S2, S3 and the S2-S3 linker. f, Pore forming helices S5, S6. g, The pore helix.
2Fo-Fc electron density is contoured at 1 σ and shown in light blue mesh. Some bulky residues are labelled for orientation. a, Ankyrin repeats 1-6. b, The TRP domain and CTD. c, S6 and S5 helices of subunit A and B. d, Pore helices of subunit A and B. e, Unassigned density in the cavity of . The 2Fo-Fc density contoured at 1.5 σ and the is shown in red cartoon and the side chains are displayed in stick representation. The putative calcium ions are shown as green spheres. Side view of the channel cavity with subunits A and C removed for ease of viewing.
Supplementary Figure 5 π-helices in the S4-S5 linker determine the quaternary structure of the subunits and the conformation of the pore helices.
a, A rotation of the subunit A (blue) around its π-helix hinge places it in alignment with the subunit A of (red) (Cα r.m.s.d. 0.99 Å). Conformational changes are identified in the S4-S5 linker (right panel, boxed region). To illustrate the conformational change an insert shows a 90° rotated view of the region. b, A side view of the overlay of subunits A (red) and B (pale cyan) (Cα r.m.s.d. 0.51 Å) illustrates the presence of local changes in the S4-S5 linker and a difference in the angle of the pore helices (dashed line boxes). c, Front view of the overlay showing the pore forming S5, S6 and pore helices. A ~5° change in the tilt of the pore helix was measured. d, RTx bound to subunit A is shown in maroon stick representation, and RTx bound to subunit B is shown in slate blue sticks. e,f, Simulated Annealing Omit Fo-Fc map of RTx. SA Fo-Fc omit map was calculated at full resolution (45-3.1 Å) with an annealing temperature of 3000 K, contoured at 2.8 σ and displayed as green mesh. e, RTx molecule (magenta sticks) bound to subunit A. f, RTx molecule (magenta sticks) bound to subunit B.
a, Top view, alignment illustrates the contraction of the channel along the y-axis (subunits B and D), and widening along the x-axis (subunits A and C). The differences in distance between subunits B and D and A and C of and miTRPV2EM were measured between Cα atoms of N429 residues in S2 (not shown). b, Alignment of subunits B of (red) and miTRPV2EM (green) (Cα r.m.s.d. 0.61 Å). While the VSLD align well, there is a conformational difference in the S4-S5 linker and the pore helix (dashed boxes). c, A close-up view of the S4-S5 linkers. d, A front view of the aligned subunits, showing S5, S6 and the pore helix. e, Bottom-up view of the common gate. TRPV2EM (green) and (red). For simplicity, only S6 helices are shown. Compared to TRPV2EM, S6 helices of splayed open slightly more.
a, Side view of the overlay of the (marine) and (red) subunits C and B, with RTx shown in stick and sphere representation and colored in magenta. The RTx molecule bound in this configuration is not able to exert force on the π-helix hinge, and therefore cannot cause the subunit B to rotate. b, Hydrogen bond networks around the pore helix in subunit B of . A network of hydrogen bonds is present at the Y542-T602-Y627 triad. c, Hydrogen bonds around the miTRPV2EM pore helix. No hydrogen bonds are present between the triad residues Y542-T602-Y627.
Supplementary Figure 8 The mutation T602A reduced the permeability of TRPV2 channel towards the large cation NMDG+.
a,b, Representative inside-out recordings of TRPV2QM (a) and TRPV2QM T602A (b) in symmetrical 150 mM NaCl (blue trace) or in 150 mM external NaCl/150 mM internal NMDG-Cl (red trace). Currents were evoked by intracellular application of 50 nM RTx. c, The shift in Erev induced by replacing internal NaCl with equimolar NMDG-Cl was calculated by subtracting the Erev of NMDG+ by the Erev of symmetrical Na+. For each recording, the Erev of NMDG+ was calculated from the average of five NMDG+ current traces after the shift in Erev stabilized. d, The permeability ratios PNMDG/PNa of TRPV2QM and TRPV2QM T602A were calculated from the Erev shifts determined in c. The number of recordings is noted in parentheses. Data are presented as mean ± s.e.m. p** < 0.01, two-tailed unpaired Student’s t-test. p-values: c, 0.0057; d, 0.0094.