Molecular pathway and structural mechanism of human oncochannel TRPV6 inhibition by the phytocannabinoid tetrahydrocannabivarin

Cryo-EM analysis of TRPV6 in the presence of THCV

We employed cryo-EM to characterize TRPV6 inhibition by THCV structurally (Supplementary Figs. ¹, ²). To avoid potential interference between binding of THCV and calmodulin (CaM), we used a human TRPV6 construct, hTRPV6-CtD, with truncated CaM binding site-containing C-terminus and wild type-like function^35,38,45,46. Cryo-EM micrographs for hTRPV6-CtD reconstituted in cNW30 nanodiscs in the presence of 114 µM THCV showed evenly dispersed particles with diverse angular coverage (Supplementary Fig. ^2a). Clearly visible secondary structure elements in 2D class averages supported high quality of the collected cryo-EM data (Supplementary Fig. ^2b). The three-dimensional reconstruction of TRPV6_THCV with four-fold rotational symmetry (C4) resulted in a cryo-EM map with the overall resolution of 2.79 Å (Fig. 1d–f, Supplementary Figs. ¹, ^2c–e, Supplementary Table ¹). Consistent with the previously published structures of hTRPV6-CtD^35,38, the TRPV6_THCV reconstruction contained no signs of CaM. For each subunit of the TRPV6_THCV homotetramer, we built a model, including residues 27–638 and excluding the N-terminal (residues 1–26) and C-terminal (residues 639–666) regions that were not resolved in the cryo-EM density.

Structure of TRPV6_THCV and THCV binding sites

The TRPV6_THCV structure (Fig. 2a, b) has a similar overall architecture as the previously determined TRPV6 structures^35,36,38,47. TRPV6 is assembled of four subunits and contains a transmembrane domain (TMD) with the central ion channel pore and an intracellular skirt that is mostly built of ankyrin repeat domains connected to each other by three-stranded β-sheets, N-terminal helices, and C-terminal hooks. Amphipathic TRP helices run nearly parallel to the membrane and interact with both the TMD and the skirt. The TMD is composed of six transmembrane helices S1–S6 and a re-entrant pore loop (P-loop) between S5 and S6. A bundle of the first four transmembrane helices represents the S1–S4 domain which forms a voltage sensor in voltage-gated ion channels⁴⁸. The pore domain of each subunit includes S5, P-loop, and S6, and is packed against the S1–S4 domain of the neighboring subunit in a domain-swapped arrangement⁴⁹.

The TMD of TRPV6_THCV is surrounded by numerous auxiliary densities, the majority of which represent annular lipids (Fig. 1d–f). High-quality reconstruction of TRPV6_THCV identifies three such densities per subunit as representing cholesteryl hemisuccinate (CHS), which was added during protein purification (see Methods). With the exception of two densities per subunit, other densities were modeled well by phosphatidylcholine (PC) molecules. The two exceptions (red in Fig. 1d, f) have not been seen in previous structures of TRPV6 and have the shape and size of THCV molecules (Supplementary Fig. ^3a–d). The first one of the two corresponding THCV binding sites is formed by S5 and S6 of one subunit and S6 of the neighboring subunit and located in the portal connecting the pore and the membrane surrounding TMD (Fig. 2c, d). The second site is located next to the first one but it is more peripheral with respect to the pore, surrounded by lipids, including CHS molecules at the vanilloid site (at the interface of S5 and S6 of one subunit and S4 of the neighboring subunit) and at the interface of P-loop of one subunit and S6 of the neighboring subunit.

At both sites, putative THCV molecules do not form strong interactions with the surrounding residues (Fig. 2c, d, Supplementary Fig. ^{3e, f}). At site 1, THCV makes a number of hydrophobic (Van der Waals) contacts with the residues surrounding the side portals. In contrast, the majority of weak interactions with THCV at site 2 are with lipids. Correspondingly, site 2 is unlikely to cause strong conformational changes in the protein and we propose that site 1 is the main inhibitory site of THCV.

Probing of THCV binding site

To verify the identified sites 1 and 2, we made mutations of nearly all residues predicted to contribute to THCV binding (Supplementary Fig. ^{3e, f}). Most residues were mutated to alanine, while alanine A561 and leucines L490 and L568 were mutated to the bulky tryptophan. We used fluorescence-detection size-exclusion chromatography (FSEC) to confirm expression and tetrameric assembly of the wild type and mutant channels produced in HEK 293 cells (Supplementary Fig. ^4a–h) and tested their function using Fura-2 AM-based measurements of intracellular calcium. Based on the height of the FSEC tetrameric peak, the L490W mutant showed somewhat lower expression level (Supplementary Fig. ^4b). Nevertheless, Fura-2 AM fluorescent signals were strong enough to make reliable measurements of intracellular calcium. Two mutants, C494W and L568W, did not show measurable calcium uptake. Confirming that these mutants were not functional, whole-cell recordings from HEK 293 cells expressing these mutants did not detect measurable TRPV6-mediated currents, while select examples of mutants that did show calcium uptake, conducted typical TRPV6-mediated currents (Supplementary Fig. ^4i–l). As an alternative to the L568W mutation, we substituted leucine L568 with glutamine and the resulting L568Q mutant turned out to be functional.

Our calcium uptake measurements revealed that except for T567A, which showed a THCV concentration dependence similar to wild type (for example, at 30 µM THCV, the Δ(F₃₄₀/F₃₈₀) values for the mutant were not significantly different from wild type, t-Test, P = 0.33), all mutants were divided into two major groups: those that demonstrated a rightward shift in the THCV concentration dependence (M497A, A561W, L568Q) or weakening of the inhibition and those that showed a leftward shift in the concentration dependence (L490W, F493A, I564A) or increased potency of THCV correspondingly (Fig. 2e–g). Thus, replacement of the long hydrophobic side chain of M497 with the short side chain of alanine as well as substitution of hydrophobic leucine L568 with hydrophilic glutamine resulted in reduced potency of THCV (at 30 µM THCV, the Δ(F₃₄₀/F₃₈₀) values were significantly different from wild type, t-Test, P = 0.0081 for M497A and P = 0.0136 for L568Q), likely due to the loss of hydrophobic interactions. On the other hand, weakening of THCV inhibition observed for the A561W mutant (at 30 µM THCV, the Δ(F₃₄₀/F₃₈₀) values were significantly different from wild type, t-Test, P = 0.0009) most probably originated from more restricted access to the portals imposed by bulky side chain of tryptophan compared to the small side chain of alanine (Fig. 2c, d). On the contrary, smaller side chains of alanine in F493A and I564A likely made it easier for the THCV molecule to reach the deep portal site 1 or have created more room and opportunity for the inhibitor to form more optimal hydrophobic interactions with the surrounding protein, thus leading to an increased inhibitor potency (leftward shift in the THCV concentration dependence; at 30 µM THCV, the Δ(F₃₄₀/F₃₈₀) values were significantly different from wild type, t-Test, P = 0.0429 for F493A and P = 0.0394 for I564A).

Interestingly, replacement of leucine L490 that points towards sites 1 and 2 with the bulky tryptophan produces a dual effect: it increases THCV potency, likely due to increased number of hydrophobic interactions with the indole ring in one rotameric conformation, and the same time makes the inhibition ~50% incomplete, probably because the indole ring in an alternative rotameric conformation fills up the portal space and prevents binding of THCV (Fig. 2g; while at 30 µM THCV, the Δ(F₃₄₀/F₃₈₀) values for L490W were not significantly different from wild type, t-Test, P = 0.479, they were significantly different at 3 µM THCV, t-Test, P = 0.0185, and 100 µM THCV, t-Test, P = 0.0271). In this regard, a small ~20% fraction of uninhibited F493A channels at high THCV concentration may originate from an alternative positioning of the THCV molecule in the roomier site 1, where it no longer causes conformational changes that lead to channel closure (see the next section). However, given high values of basal calcium signals in experiments with L490W (see Source Data), their results should be interpreted with caution. For select mutants that showed pronounced leftward (L490W and I564A) and rightward (M497A and A561W) shifts in IC₅₀ for THCV, we confirmed the ability to conduct currents in whole-cell patch-clamp experiments (Supplementary Fig. ^4i–l), lack of dramatic changes in the concentration dependencies of calcium activation (Supplementary Fig. ^4m) and econazole inhibition (Supplementary Fig. ⁴ⁿ), and complete block of the L490W mutant by 100 µM Gd³⁺ (Supplementary Fig. ^4o), suggesting that functional properties other than THCV inhibitory potency were not strongly affected by the corresponding mutations. Interestingly, the slope of the concentration–response curve for econazole inhibition of M497A was less steep compared to wild type and other mutants (Supplementary Fig. ⁴ⁿ). Given that M497 is located on the opposite site of the S5 helix, close to W495, which makes a direct contact with the econazole molecule, it is possible that this mutation weakens the cooperativity that normally exists between four econazole binding sites (one per TRPV6 subunit). Overall, however, the changes in the concentration dependence of calcium uptake inhibition by THCV observed in mutant channels compared to wild type (Fig. 2g) strongly support THCV binding to sites 1 and 2.

Closed pore of TRPV6_THCV

In contrast to the apo-state structures of hTRPV6, which have an open conducting pore^35,38,47, the pore in TRPV6_THCV is in a fully closed conformation, hydrophobically sealed by the side chains of residues L574 and M578 (Fig. 3a). Compared to the open-state structure with the π bulge in the middle of S6, the entire S6 in TRPV6_THCV is α-helical, typical for the closed-state structures of TRPV6, including structures in complex with inhibitors ruthenium red (RR) and econazole^{35,36,38,46,47,49} (Fig. 3b). Comparison of the open-state hTRPV6 structure to the structure solved in the presence of THCV demonstrates that the open-to-closed state transformation of the lower pore includes a ~100° rotation and bending away from the pore axis of the intracellular part of S6, consistent with previous observations^36,38,47. This drastic conformational change in S6 starts below the gating hinge alanine A566, without altering the upper pore of hTRPV6 and thereby leaving the selectivity filter unaffected.

Within the hTRPV6 regions previously implicated in gating, two hydrogen bonds per subunit stabilize the open state of the channel: one bond between Q473 in the S4–S5 elbow and R589 in the TRP helix and another one between D489 in the S5 helix and T581 in the S6 helix⁴⁷. Neither interaction is present in the closed-state structure of hTRPV6 in complex with THCV, suggesting that binding of THCV leads to disruption of these bonds. The closed pore of TRPV6_THCV therefore confirms the inhibitory action of THCV. To further validate that this inhibitory action occurs through binding of THCV in the ion channel side portals and to provide an insight into the molecular mechanism of this inhibition, we performed molecular dynamics (MD) simulations.

Orientation of THCV in site 1 probed with MD simulations

There are two possible orientations of THCV at site 1: (1) with the tricyclic ring of THCV oriented towards the central pore (forward orientation, Fig. 4a) and (2) with the tricyclic ring oriented towards the membrane (backward orientation, Fig. 4b). Since the cryo-EM map provides no clear density for the acyl chain of THCV due to its flexibility, both orientations are feasible. To determine a more favorable positioning of the ligand, we performed two sets of MD simulations, one for each orientation. In each set, we carried out four simulations, with the initial orientations of the ligand differed by 90° rotations around the long THCV axis. As a measure of ligand stability in the binding pocket, we calculated the average spatial distribution of the center of the THCV molecule determined as a center of mass of heavy atoms over the two central rings of THCV. The ligand center appeared to overlap well with the experimental cryo-EM density for THCV in the forward orientation (Fig. 4a), while it was shifted towards the central pore or membrane environment for THCV in the backward orientation (Fig. 4b). Compared to the latter, the forward orientation also showed a smaller displacement of the ligand center relative to the cryo-EM model over the simulation time (Fig. 4c). According to the MD data, the only specific interaction that can stabilize the forward orientation of THCV is the hydrogen bond between the hydroxyl group of THCV and the carbonyl oxygen of L490 (Fig. 4a). This bond helps to create and maintain a kink in the S5 helix and prevents bond formation between the adjacent D489 and T581 in S6 that stabilizes the open state of the channel (Supplementary Fig. ^5a). By contrast, formation of the hydrogen bond between THCV and L490 in the backward orientation in our simulations leads to dissociation of the ligand from the binding pocket.

THCV pathway to site 1 probed by constant velocity steered MD simulations

Since conventional MD simulations are too slow to model ligand binding and dissociation in such complex systems as TRPV6_THCV, we pulled the forward-oriented THCVs out of the binding pockets using the constant velocity steered MD simulations⁵⁰. Four independently pulled ligands, one per subunit of TRPV6 tetramer, revealed two possible routes of leaving their binding sites. The first one, “horizontal”, runs along the S4–S5-linker helix through a groove formed by the side chains of L490 and L568 to the region of the THCV site 2 (Fig. 4d). The second route, “vertical”, runs along the pore helix substituting the adjacent annular CHS (lipid 803 in the cryo-EM model, Supplementary Fig. ^5b).

To determine the energy landscapes for both pathways, we performed the umbrella sampling MD simulations⁵¹ and calculated the potential of mean force (PMF) along the pulling trajectories. In both cases, the THCV position at the portal site was stabilized by the hydrogen bond between the ligand and L490 and corresponded to the energy minimum along the selected pathway (M1 region in Fig. 4e and Supplementary Fig. ^5c). There was another preferable position of the ligand along the vertical route, where THCV was involved in a π-stacking interaction with F534 near the pore helix (M3 region in Supplementary Fig. ^5c). However, the transition from M1 to M3 that followed the vertical route was associated with overcoming the energy barrier of 6–7 kcal/mol caused by significant protein-ligand clashes. In contrast, the horizontal route ran through the M1-M2-M3 energy minima separated by energy barriers of only 3–4 kcal/mol (Fig. 4e).

To check the ligand binding in the same pockets of the TRPV6 open state, we performed one more conventional MD simulation with the forward-oriented THCVs (like in the cryo-EM model) embedded in site 1 of apo TRPV6 (PDB ID 7S88)³⁸. The simulation showed that THCV molecules lost the hydrogen bond with L490 and left the pockets right after removing the positional restraints applied to the ligands. Three out of four ligands moved out of the pockets along the horizontal route, now formed by L490 and L569 (Supplementary Fig. ^5a). We therefore concluded that the horizontal route connecting THCV sites 1 and 2 is the most likely pathway for the ligand to approach or leave its portal binding pocket both in the TRPV6 inhibited and open states.

Consistent with the nonpolar character of the THCV molecule, the horizontal route also has pronounced hydrophobic properties. This reciprocity can be clearly illustrated by using the molecular hydrophobicity potential (MHP) method^52,53. Thus, with the exception of the polar OH group, which is capable of forming hydrogen bonds with the protein, the surface of THCV is hydrophobic and characterized by high MHP values (Supplementary Fig. ^5d). A similar pattern is observed for the surrounding protein surface (Supplementary Fig. ^5e). Such a complementarity in physicochemical properties between the protein and the ligand (contacts of their hydrophobic surfaces are energetically favorable) facilitates the process of ligand migration between its binding sites in TRPV6.

To rationalize the effects of mutation L490W in THCV site 1 (Fig. 2g), we performed MD simulations of the corresponding mutant TRPV6_THCV channels with the forward-oriented THCV molecules embedded in site 1 like in cryo-EM model. MD simulations showed that the indole ring of tryptophan W490 can adapt two possible conformations, with the side chain oriented inward and outward the pocket (Supplementary Fig. ^6a). In both conformations, W490 forms a parallel π-stacking interaction with the tricyclic ring of THCV (Supplementary Fig. ^{6b, c}). The improved stability of the hydrogen bond between THCV and W490 as well as additional π-stacking interaction (Supplementary Fig. ^6d) explain the increased affinity of the ligand observed for L490W in calcium uptake experiments (Fig. 2g). The significantly reduced ability of THCV to inhibit calcium entry through the L490W mutant is likely due to limited access of the ligand to the portal site because of the block of the horizontal route by the bulky side chain of tryptophan. Our MD analysis, therefore, provides a rationale for the results observed in calcium uptake experiments and reinforces the importance of site 1 for the mechanism of TRPV6 inhibition by THCV.

Expression and purification

hTRPV6-CtD was expressed and purified based on our previously established protocols^{38,40,59–63}. Bacmids and baculoviruses were produced using standard procedures. Baculovirus was made in Sf9 cells for ~72 h (Thermo Fisher Scientific, mycoplasma test negative, GIBCO #12659017) and was added to suspension-adapted HEK 293S cells lacking N-acetyl-glucosaminyltransferase I (GnTI^–, mycoplasma test negative, ATCC #CRL-3022) that were maintained in Freestyle 293 media (Gibco-Life Technologies #12338-018) supplemented with 2% FBS at 37 °C and 5% CO₂. Twenty-four hours after transduction, 10 mM sodium butyrate was added to enhance protein expression, and the temperature was reduced to 30 °C. Seventy-two hours after transduction, cells were harvested by centrifugation at 5471 × g for 15 min using a Sorvall Evolution RC centrifuge (Thermo Fisher Scientific), washed in phosphate-buffered saline pH 8.0, and pelleted by centrifugation at 3202 × g for 10 min using an Eppendorf 5810 centrifuge. The cell pellet was solubilized under constant stirring for 2 h at 4 °C in ice-cold lysis buffer containing 1% (w/v) n-dodecyl β-D-maltoside, 0.1% (w/v) CHS, 20 mM Tris-HCl pH 8.0, 150 mM NaCl, 0.8 μM aprotinin, 4.3 μM leupeptin, 2 μM pepstatin A, 1 mM phenylmethylsulfonyl fluoride, and 1 mM β-mercaptoethanol (βME). The non-solubilized material was pelleted in the Eppendorf 5810 centrifuge at 3202 × g and 4 °C for 10 min. The supernatant was subjected to ultracentrifugation in a Beckman Coulter ultracentrifuge using a Beckman Coulter Type 45Ti rotor at 186,000 × g and 4 °C for 1 h to further clean up the solubilized protein. The supernatant was added to strep resin and rotated for 1 h at 4 °C. The resin was washed with 10 column volumes of wash buffer containing 20 mM Tris-HCl pH 8.0, 150 mM NaCl, 1 mM βME, 0.01% (w/v) GDN, and 0.001% (w/v) CHS, and the protein was eluted with the same buffer supplemented with 2.5 mM D-desthiobiotin. The eluted protein was concentrated using a 100 kDa NMWL centrifugal filter (MilliporeSigma Amicon) to 0.5 ml and then centrifuged in a Sorvall MTX 150 Micro-Ultracentrifuge (Thermo Fisher Scientific) using an S100AT4 rotor for 30 min at 66,000 × g and 4 °C before injection into a size-exclusion chromatography (SEC) column. All constructs were further purified using a Superose™ 6 10/300 GL SEC column attached to an AKTA FPLC (GE Healthcare) and equilibrated in 150 mM NaCl, 20 mM Tris-HCl pH 8.0, 1 mM βME, 0.01% GDN, and 0.001% CHS. The tetrameric peak fractions were pooled and concentrated using 100 kDa NMWL centrifugal filter to 3.5 mg/ml.

Circularized cNW30 nanodiscs were prepared according to the standard protocol⁴⁸ and stored (~2–3 mg/ml) before usage at –80 °C in 20 mM Tris-HCl (pH 8.0) and 150 mM NaCl. For nanodisc reconstitution, the purified protein was mixed with circularized cNW30 nanodiscs and soybean lipids (Soy polar extract, Avanti Polar Lipids, USA) at a molar ratio of 1:3:166 (TRPV6 monomer:cNW30:lipid). The lipids were dissolved to a concentration of 100 mg/ml in 150 mM NaCl and 20 mM Tris-HCl (pH 8.0) and subjected to 5–10 cycles of freezing in liquid nitrogen and thawing in a water bath sonicator. The nanodisc mixture (500 µl) was rocked at room temperature for 1 h. Subsequently, 20 mg of Bio-beads SM2 (Bio-Rad), pre-wet in a buffer containing 20 mM Tris-HCl pH 8.0, 150 mM NaCl and 1 mM βME, were added to the nanodisc mixture, which was then rotated for 1 h at 4 °C. Additional 20 mg of Bio-beads SM2 were added, and the resulting mixture was rotated at 4 °C for another ~14–20 h. Bio-beads SM2 were then removed by pipetting and nanodisc-reconstituted hTRPV6-CtD was purified from empty nanodiscs by SEC using Superose™ 6 10/300 GL SEC column equilibrated in 150 mM NaCl, 20 mM Tris-HCl (pH 8.0), and 1 mM βME. Fractions of nanodisc-reconstituted hTRPV6-CtD were pooled and concentrated using 100 kDa NMWL centrifugal filter to 2.58 mg/ml. 114 µM THCV (from a 5 mM stock in 100% DMSO) was added to the TRPV6 monomer:cNW30:lipid mixture and the specimen was incubated for 1 h at room temperature before grid preparation.

Cryo-EM sample preparation and data collection

UltrAuFoil R 1.2/1.3, Au 300 grids were used for plunge-freezing. Prior to sample application, grids were plasma treated in a PELCO easiGlow glow discharge cleaning system (0.39 mBar, 15 mA, “glow” 25 s, “hold” 10 s). A Mark IV Vitrobot (Thermo Fisher Scientific) set to 100% humidity at 4 °C was used to plunge-freeze the grids in liquid ethane after applying 3 µl of protein sample to their gold-coated side using a blot time of 5 s, a blot force of 5, and a wait time of 15 s. The grids were stored in liquid nitrogen before imaging.

Images of frozen-hydrated particles of cNW30-reconstituted hTRPV6_THCV were collected on a Titan Krios transmission electron microscope (Thermo Fisher Scientific) operating at 300 kV and equipped with a post-column GIF Quantum energy filter and a Gatan K3 Summit direct electron detection camera (Gatan, Pleasanton, CA, USA) using SerialEM 4.0. 6699 micrographs were collected in super-resolution counting mode with raw image pixel size of 0.4125 Å across the defocus range of −0.8 to −2.0 µm. The total dose of ~60 e⁻Å⁻² was attained by using the dose rate of ~16 e⁻pixel⁻¹s⁻¹ across 50 frames during the 2.0-s exposure time.

Image processing and 3D reconstruction

Data were processed in cryoSPARC⁶⁴. Movie frames were aligned using the patch motion correction. Contrast transfer function (CTF) estimation was performed on non-dose-weighted micrographs using the patch CTF estimation. Subsequent data processing was done on dose-weighted micrographs. Following CTF estimation, micrographs were manually inspected and those with outliers in defocus values, ice thickness, and astigmatism as well as micrographs with lower predicted CTF-correlated resolution (higher than 5 Å) were excluded from further processing (individually assessed for each parameter relative to the overall distribution). After several rounds of selection through 2D classification, particles were further 3D classified (heterogeneous refinement) into four classes. Particles representing the best class were re-extracted without binning (256-pixel box size) and further 3D classified. The final set of 213,249 particles for hTRPV6_THCV representing the best class was subjected to homogenous refinement. The reported resolution of 2.79 Å was estimated using the gold standard Fourier shell correlation (GSFSC). The local resolution was calculated with the resolution range estimated using the FSC = 0.143 criterion. Cryo-EM density visualization was done in UCSF Chimera⁵¹ and UCSF ChimeraX⁶⁵.

Model building

To build the model of TRPV6_THCV in Coot⁶⁶, we used the previously published cryo-EM structure of closed-state TRPV6³⁸ (PDB ID: 7S8C) as a guide. The model was tested for overfitting by shifting its coordinates by 0.5 Å (using Shake) in Phenix⁶⁷, refining the shaken model against the corresponding unfiltered half map, and generating densities from the resulting models in UCSF Chimera. TRPV6_THCV and other structures were visualized and figures were prepared in UCSF Chimera, UCSF ChimeraX⁶⁵, and Pymol⁶⁸. The pore radius was calculated using HOLE⁶⁹.

MD simulations

To determine a suitable THCV structure for MD simulations, we first explored the conformational ensemble of this ligand using quantum chemical calculations. Since the most significant changes in the THCV conformation are associated with the rotation around the C₁₇-C₀₄ bond (see atom labels in Supplementary Fig. ^{3e, f}), the dihedral angle C₁₄-C₁₇-C₀₄-C₀₃ was scanned to determine possible energy minima. As a result, two stable states were found in the energy landscape, with the dihedral angle values of 170° (conformation 1) and 91° (conformation 2). Equilibrium geometry (minimum energy conformation) of the isolated THCV and energy profiles for the states generated by means of the dihedral drive protocol were calculated in vacuo using the Hartree-Fock method with the 6-31G(d) basis in GAUSSIAN09, Revision A.01 software⁷⁰. Geometry optimization for the calculated models was performed without restraints. The Charmm36^71,72 topology for THCV was created by providing the possibility of spontaneous conformational transitions between THCV conformations 1 and 2. The topology was successfully tested via MD simulations of the isolated THCV in vacuo. Since the cryo-EM model of THCV in the pocket is very close to conformation 2, we used this conformation in subsequent MD simulations with the channel.

Structural model of TRPV6_THCV (residues 27–638) was inserted into a hydrated lipid bilayer with the molecular composition of 50% palmitoyloleoylphosphatidylcholine (POPC), 25% palmitoyloleoylphosphatidylethanolamine (POPE), and 25% cholesterol (about 900 molecules in the membrane). Na⁺ and Cl⁻ ions were added to ensure zero net charge at 0.15 M ionic concentration. Four THCV molecules were embedded with the same orientation into the portal pockets of the protein. Eight replicas of the system with the wild-type (WT) protein and different orientations of THCV (see text and Supplementary Table ² for details), one replica with the L490W mutation and one replica of the system with THCV molecules embedded into the portal pockets of the open TRPV6³⁸ (PDB ID 7S88) were constructed this way.

The simulated systems were first equilibrated in several stages: 5 × 10³ steps of steepest descent minimization followed by heating from 5 to 310 K during a 100-ps MD-run, then 10 ns of MD run with fixed positions of the protein and THCV heavy atoms, 10 ns of MD with fixed positions of the protein backbone and THCV heavy atoms, 20 ns of MD with fixed positions of the protein C_α atoms to permit membrane and THCV relaxation after insertion. Then, 50 ns MD-production runs were carried out for the WT protein and open TRPV6 systems, 200 ns MD-production runs were carried out for the mutated and one of the WT systems with THCV oriented like in the cryo-EM model.

MD simulations were carried out using GROMACS 2021.4 package⁷³, CHARMM36 force field^74–79 and the TIP3P water model⁸⁰. Simulations were carried out with an integration time of 2 fs, constrained hydrogen-containing bond lengths by the LINCS algorithm⁸¹, imposed 3D periodic boundary conditions, constant temperature (310 K) and pressure (1 bar). Cutoff distance of 1.2 nm was used for evaluation of nonbonded interactions and the particle-mesh Ewald method⁸² employed for treatment of long-range electrostatics.

The WT system with forward-oriented THCV like in the cryo-EM model was used for pulling and umbrella sampling MD-simulations. After equilibration, position restraints were imposed on the C_α atoms of S6-helices (residues 553–583) to preserve the protein structure. Each THCV was pulled independently from the portal pockets along the directions that were defined as a line connecting the center of the pore gate (center of mass of C_δ1 and C_δ2 atoms of L574 of four protein subunits) and the center of a ligand (center of mass of heavy atoms of two central rings of THCV: C₀₄, C₀₅, C₀₆, C₀₇, C₀₈, C₁₂, C₁₃, O₀₂, C₁₄, see atom labels in Supplementary Fig. ^{3e, f}). The pulling rate was set to 0.05 nm/ns and the force constant of 1000 kJ/(mol × nm²), because at these parameters the ligand moves along the protein surface and has enough mobility to adapt to the surface relief. As a result of 50 ns simulation run, we got MD trajectories describing dissociation of THCV, which were suitable for the potential of mean force (PMF) calculation with the umbrella sampling method. Then, 32 snapshots were taken from the pulling trajectory to generate starting configurations for umbrella sampling^83,84 windows along a collective variable (distance between the center of the pore gate and the ligand center). 5 ns MD-simulation with 1000 kJ/(mol × nm²) force constant constraining along the collective variable were performed for each window. The results were processed with the weighted histogram analysis method (WHAM)⁸⁵.

The approach of Molecular Hydrophobicity Potential (MHP) is a method that enables quantitative estimation of the spatial distribution of hydrophobic/hydrophilic properties, e.g., on a molecular surface. The formalism of MHP assumes that each atom in the molecular system is attributed with its specific hydrophobicity constant and the sum of atomic contributions is calculated at the molecular surface^52,86. In this study, we used the atomic hydrophobicity constants obtained by Ghose et al.⁸⁶. MHP calculations were performed with PLATINUM software⁸⁷.

Fura-2 AM measurements

Full-length wild-type or mutant human TRPV6 was expressed in HEK 293S cells as described above. Forty-eight hours after transduction, cells were harvested by centrifugation at 550 × g for 5 min. The cells were resuspended in prewarmed HEPES-buffered saline (HBS: 118 mM NaCl, 4.8 mM KCl, 1 mM MgCl₂, 5 mM D-glucose, 10 mM HEPES pH 7.4) containing 5 µg/ml of Fura-2 AM (Life Technologies) and incubated at 37 °C for 45 min. The loaded cells were then centrifuged for 5 min at 550 × g, resuspended again in prewarmed HBS, and incubated at 37 °C for 30 min in the dark. The cells were subsequently pelleted and washed twice, then resuspended in HBS for experiments. The cells were kept on ice in the dark for a maximum of ~2 h before fluorescence measurements, which were conducted using spectrofluorometer QuantaMaster 40 (Photon Technology International) at room temperature in a quartz cuvette under constant stirring. Intracellular Ca²⁺ was measured by taking the ratio of fluorescence measurements at two excitation wavelengths (340 and 380 nm) and one emission wavelength (510 nm). The excitation wavelength was switched at 200-ms intervals. In experiments with THCV, every batch of HEK293 was split into two halves: half of cells were transfected with TRPV6 and the other half remained non-transfected. Measurements were repeated 4 to 8 times at each different concentration of THCV, which was added up to 100 µM concentration (from a 50 mM stock in 100% DMSO; higher than 100 µM concentrations of THCV were not used due to non-specific effects of THCV), and then the changes in fluorescence (F₃₄₀/F₃₈₀ ratios) recorded from non-transfected cells were subtracted from the changes in fluorescence recorded from transfected cells. For any single experiment with THCV, the antagonist was added to cells resuspended in HBS buffer 200 seconds prior to the addition of 10 mM calcium for activation in the recording (the last 50 seconds of the 200-s incubation were recorded as the baseline for each experimental recording trace). The normalized average of the corrected fluorescent signals was fitted with the logistic equation to determine the value of IC₅₀ (Fig. 2g). Data analysis was performed using Origin 9.1.0 (OriginLab Corp.).

Electrophysiology

DNA encoding wild-type or mutant human TRPV6 was introduced into a plasmid for expression in eukaryotic cells that was engineered to produce GFP via a downstream internal ribosome entry site⁸⁸. HEK 293 cells (ATCC #CRL-1573) grown on glass coverslips in 35-mm dishes were transiently transfected with 1-5 µg of plasmid DNA using Lipofectamine 2000 reagent (Invitrogen). Recordings were made at room temperature, 36 to 72 h after transfection. Currents from whole cells, typically held at a 0 or −60 mV membrane potential, were recorded using Axopatch 200B amplifier (Molecular Devices, LLC), filtered at 5 kHz, and digitized at 10 kHz using low-noise data acquisition system Digidata 1440 A and pCLAMP software (Molecular Devices, LLC). The external solution contained 140 mM NaCl, 6 mM CsCl, 1 mM MgCl₂, 10 mM HEPES pH 7.4 and 10 mM glucose. To evoke monovalent currents, 1 mM EGTA was added to the external solution. The pipette solution contained (in mM): 100 CsAsp, 20 CsF, 10 EGTA, 3 MgCl₂, 4 NaATP and 20 HEPES pH 7.2. TRPV6 currents were recorded in response to 50-ms voltage ramps from −100 to 70 mV. Data analysis was performed using Origin 9.1.0 (OriginLab Corp.).

Fluorescence-detection size-exclusion chromatography

The experimental setup to conduct FSEC experiments in order to evaluate protein expression levels and tetrameric assembly of the channel was previously described in detail⁵⁹. Full-length wild-type or mutant human TRPV6 was expressed in HEK cells as described above. Forty-eight hours after transduction, cells were harvested by centrifugation at 550 × g for 5 min. 1-ml aliquots of cells were solubilized under constant stirring for 1 h at 4 °C in 0.5 ml ice-cold lysis buffer containing 1% (w/v) n-dodecyl β-D-maltoside, 0.1% (w/v) CHS, 20 mM Tris-HCl pH 8.0, 150 mM NaCl, 0.8 μM aprotinin, 4.3 μM leupeptin, 2 μM pepstatin A, 1 mM phenylmethylsulfonyl fluoride, and 1 mM βME. The non-solubilized material was pelleted in a Sorvall MTX 150 Micro-Ultracentrifuge (Thermo Fisher Scientific) using an S100AT4 rotor for 30 min at 66,000 × g and 4 °C before 100 µl of the supernatant was injected into a Superose^TM 6 10/300 GL SEC column mounted onto a Shimadzu HPLC system that included RF-10AXL fluorescence detector (Shimadzu). The SEC column was equilibrated with a buffer containing 1 mM n-dodecyl β-D-maltoside, 20 mM Tris-HCl pH 8.0, 150 mM NaCl, and 1 mM βME.

We thank Sean Mulligan (Pacific Northwest Center for Cryo-EM) for help with microscope operation and data collection. A portion of this research was supported by NIH grant U24GM129547 and performed at the PNCC at OHSU and accessed through EMSL (grid.436923.9), a DOE Office of Science User Facility sponsored by the Office of Biological and Environmental Research. A.N. is a Walter Benjamin Fellow funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation)–464295817. MD simulations were supported by the RSF (23-14-00313). A.I.S. was supported by the NIH (R01 AR078814, R01 CA206573, R01 NS083660, R01 NS107253) and NSF (1818086).