Transient Receptor Potential V (TRPV) channels are part of the larger TRP channel family which plays important roles in numerous physiological processes (Clapham et al., 2001). A subset of TRPV channels, including subtypes TRPV1-TRPV4, possess an intrinsic capability to sense heat and are therefore referred to as thermoTRPV channels (Cao et al., 2013a; Liu and Qin, 2016; Smith et al., 2002; Chung et al., 2003). TRPV1-TRPV4 are non-selective cation channels which play important physiological roles in sensing noxious heat (Bölcskei et al., 2010; Julius, 2013; Marwaha et al., 2016; Mitchell et al., 2014), maintaining cardiac structure (Katanosaka et al., 2014) and maintaining skin (Eytan et al., 2014; Imura et al., 2009; Kim et al., 2016), hair (Asakawa et al., 2006; Imura et al., 2007; Xiao et al., 2008) and bone physiology (Masuyama et al., 2008). A distinctive feature of TRPV1 and TRPV2 is their permeability to large organic cations (Chung et al., 2008), such as the cationic dye YO-PRO-1 and the sodium channel blocker QX-314. This feature has led to proposals to utilize these channels as conduits for delivering small molecules to intracellular targets (Puopolo et al., 2013). The non-conducting structures of TRPV1 and TRPV2 possess two restrictions, one at the selectivity filter (SF) and second one at the intracellular mouth of the pore (termed the common gate) (Liao et al., 2013; Zubcevic et al., 2016; Huynh et al., 2016). Both restrictions must open widely to accommodate the passage of large organic cations. However, the mechanism that enables such opening was long unclear. In order to study the permeation of both metal ions and large organic cations in TRPV2, we recently crystallized the rabbit resiniferatoxin (RTx)-sensitive (Zhang et al., 2016) TRPV2 channel with a truncation in the pore turret in the presence of the agonist RTx (Zubcevic et al., 2018a). This study led to the revelation that the binding of RTx leads to a two-fold symmetric (C2) opening at the selectivity filter that is wide enough to permeate YO-PRO-1. This unexpected result offered the first experimental evidence that the homotetrameric TRPV2 can adopt C2 symmetric conformations during the gating cycle. However, it was unclear if crystal contacts or the crystallization conditions (e.g. high concentration of Ca²⁺) played a role in stabilizing the C2 symmetry. In addition, the minimal TRPV2 construct used in the crystallographic study lacked the pore turret, a region that is not essential for function (Liao et al., 2013; Zubcevic et al., 2016; Zubcevic et al., 2018a; Yao et al., 2010) but had previously been shown to have a modulatory effect on gating in TRPV1 and TRPV2 (Jara-Oseguera et al., 2016; Dosey et al., 2019). It was uncertain if the absence of this region in our crystallographic study affected the symmetry of the channel.

In order to answer these questions and further study the role of two-fold symmetry in TRPV channel gating, we conducted cryo-electronmicroscopy (cryo-EM) studies of the full-length, RTx-sensitive rabbit TRPV2 (Zhang et al., 2016) channel reconstituted into nanodiscs and amphipol. We present four structures of the TRPV2/RTx complex, one obtained in nanodiscs (TRPV2_RTx-ND) and three in amphipol (TRPV2_{RTx-APOL 1-3}) determined to 3.8 Å, 2.9 Å, 3.3 Å and 4.2 Å resolution, respectively (Table 1, Figure 1). Our data shows that binding of RTx induces C2 symmetric conformations in TRPV2, but the extent of C2 symmetry depends on the environment in which the channel is reconstituted. C2 symmetry is particularly pronounced in the dataset collected from nanodisc-reconstituted TRPV2, which better approximates the physiological environment of the channel. Moreover, the data offers further insights into the allosteric coupling between the RTx-binding site and the activation gates in TRPV2, confirms the critical role of the S4-S5 linker π-helix (S4-S5_π-hinge) in ligand-dependent gating of TRPV2, and provides a glimpse of the conformational landscape of TRPV2 gating.

Figure 1 with 7 supplements see all

Download asset Open asset

In order to capture the RTx-induced gating transitions in the rabbit TRPV2 channel, we conducted cryo-EM studies of the TRPV2/RTx complex reconstituted into amphipol (TRPV2_RTx-APOL) and nanodiscs (TRPV2_RTx-ND). Amphipols (Zoonens and Popot, 2014) have been a useful tool in structural studies of membrane proteins, and especially TRP channels (Liao et al., 2013; Zubcevic et al., 2016; Cao et al., 2013b; Paulsen et al., 2015; Yoo et al., 2018; Hirschi et al., 2017; Zubcevic et al., 2018b). Indeed, Amphipol A8-35 enabled the very first structural determination of the TRPV2 channel (Zubcevic et al., 2016). Nanodiscs, on the other hand, represent the closest in vitro approximation to the native lipid membranes used in structural studies (Denisov and Sligar, 2016). The data was processed using RELION (Scheres, 2012) (Materials and methods), with no symmetry imposed during particle classification and 3D reconstruction in order to avoid obscuring any classes with lower symmetry (C1 and C2) that might exist in the sample. Symmetry was only imposed in the last step of the refinement and only if the 3D reconstructions showed clear two-fold (C2) or four-fold (C4) symmetry (Figure 1—figure supplements 1–3). Classification of the TRPV2_RTx-APOL sample revealed the presence of four classes: one low-resolution (~7 Å) class, which was excluded from further analysis, and three higher resolution classes which are representative of three different conformations. These include one C4 symmetric and two distinct C2 symmetric classes refined to 2.9 Å, 3.3 Å and 4.2 Å, respectively (Figure 1, Figure 1—figure supplement 1). By contrast, 3D classification of the TRPV2_RTx-ND converged on a single C2 symmetric conformation resolved to 3.8 Å (Figure 1, Figure 1—figure supplements 2–3). All four maps were of sufficient quality to enable placement of individual structural motifs with confidence (Figure 1—figure supplements 4–7) and the models for all four structures were built to good overall geometry (Table 1).

RTx induces a break in symmetry in TRPV2_RTx-ND

In stark contrast to the amphipol-reconstituted channel, RTx binding induces C2 symmetry in the nanodisc-reconstituted TRPV2 which extends throughout the channel. The symmetry of the TRPV2_RTx-ND map was assessed both visually, and by the Map Symmetry function in Phenix which yielded a CC = 0.9 and score of 1.28 for C2 symmetry. For comparison, C4 symmetry yielded a lower correlation coefficient (CC = 0.8). To further confirm the correctness of the symmetry assignment, we evaluated the fit of the TRPV2_RTx-ND model built into the C2 symmetric map to the non-symmetrized C1 map (Figure 1—figure supplement 3). In addition, we evaluated the fit of the TRPV2_RTx-ND model to the individually refined non-symmetrized classes 1 and 6, which constitute the TRPV2_RTx-ND map (Figure 1—figure supplement 3). All FSC curves indicate that the two-fold symmetric model fits well into the C1 maps and the density of the C1 symmetric TRPV2_RTx-ND map supports the model (Figure 1—figure supplement 3), showing that two-fold symmetry is truly present in the TRPV2_RTx-ND sample.

The pore of TRPV2_RTx-ND adopts a C2 symmetric arrangement (Figure 2A). The pore helices are arranged so that the carbonyl oxygens of the selectivity filter in subunits B and D line the entry to the pore while pore helices of subunits A and C tilt away from the permeation pathway. This arrangement creates a large C2 symmetric opening where the narrowest constriction between SF residues in diametrically opposing subunits A and C and B and D is ~11 Å and ~8 Å, respectively. This results in an SF with ample room to accommodate large organic cations (Figure 2B). A closer look at the pore helices reveals that this arrangement in the SF is achieved through a ~ 10° swivel of the subunit A pore helix, which brings the N-terminal part of the helix closer to S5 while distancing it from S6 (Figure 2C). The position of the pore helices controls the size and the shape of the SF, and appears to exert control over ion permeation in TRPV2. While the SF is widely open, the common gate adopts a putative intermediate conformation where two of the diagonally opposing subunits adopt a closed state, and the remaining two are open. In subunits A and C, the S6 helix adopts a straight α-helical, closed conformation, while the S6 of subunits B and D is bent and the common gate apparently open (Figure 2A). However, the overall functional state of the common gate is likely non-conductive as the gate residues from subunits A and C would presumably hinder ion permeation.

Figure 2 with 1 supplement see all

Download asset Open asset

In order to establish the origin of the C2 symmetry in the TRPV2_RTx-ND structure, we aligned subunits A and B (Cα R.M.S.D = 0.96) (Figure 3—figure supplement 5). Similar to our previous findings, this alignment shows that the two subunits diverge at the S4-S5 linker and the PH and indicates that rotation of subunits around the S4-S5_π-hinge appears to result in the distinct C2 symmetric arrangement observed in TRPV2_RTx-ND (Figure 3—figure supplements 4–5).

When compared to the TRPV2_APO, the TM domains of the TRPV2_RTx-ND structure appear to contract in an asymmetric manner (Figure 3A), while the ARD assembly expands by ~10 Å and rotates by 3° (Figure 3B). The TM domains and the ARDs appear to move as a single rigid body, which is evident when individual subunits from TRPV2_APO and TRPV2_RTx-ND are superposed (Cα R.M.S.D = 1.9 Å) to reveal that only the S4-S5 linker and the pore helix deviate significantly in the two structures (Figure 3C). This coupled movement of the TM and ARD indicates that RTx-binding to TRPV2 in lipid membranes induces a rigid-body rotation of the entire subunit that originates at the S4-S5_π-hinge (Figure 3D–E).

Figure 3 with 5 supplements see all

Download asset Open asset

Interestingly, the TRPV2_RTx-ND structure exhibits different degrees of reduced symmetry from the previously determined crystal structure of TRPV2 in complex with RTx (TRPV2_RTx-XTAL) (Zubcevic et al., 2018a). Compared to the TRPV2_RTx-XTAL, the TM domains of TRPV2_RTx-ND contract in an two-fold symmetric manner (Figure 4A). This conformational change, which stems from rotation of individual TRPV2_RTx-ND subunits around the S4-S5_π-hinge (Figure 4—figure supplement 1), results in an overall fold that is closer to C4 symmetry than that of the TRPV2_RTx-XTAL (Figure 4B). However, while the TRPV2_RTx-ND helices S1-S6 adopt a more C4 symmetric arrangement, the pore helices and the SF remain distinctly C2 symmetric (Figure 4C). Remarkably, the SF of TRPV2_RTx-ND is wider than that of TRPV2_RTx-XTAL, and the two structures display different C2 symmetric openings at the SF (Figure 4C). The two different conformations result from both the different arrangements of subunits and changes in the position and tilt angle of the pore helices (Figure 4D–E). In the TRPV2_RTx-XTAL structure, the pore helices of subunits B and D, which assume a widened conformation, are free of interactions with the pore domain, while a network of interactions (presumably hydrogen bonds) between Y542-T602-Y627 in subunits A and C tethers the pore helices to S5 and S6. Our previous work showed that disruption of these interactions is detrimental to the permeation of large organic cations but has no effect on permeation of metal ions (Zubcevic et al., 2018a). Interestingly, the putative hydrogen bond triad is disrupted in all four subunits of the TRPV2_RTx-ND structure (Figure 4—figure supplement 2). Nevertheless, the SF assumes a fully open state that can potentially accommodate passage of a large cation. This suggests that the putative hydrogen bond triad, while not a feature of the fully open SF, is an essential part of the transition between closed and open states of the channel.

Figure 4 with 6 supplements see all

Download asset Open asset

It is interesting to point out that extensive rearrangements around the SF and the PH during gating have thus far only been observed in structural studies of TRPV1 (Cao et al., 2013b) and TRPV2 (Zubcevic et al., 2018a) channels. In the non-conductive state, the SFs of the remaining members of the TRPV subfamily (TRPV3-TRPV6 [Zubcevic et al., 2018b; Singh et al., 2018; Deng et al., 2018; Hughes et al., 2018; McGoldrick et al., 2018]) adopt a conformation that is wide enough to accommodate a semi-hydrated cation, and do not move appreciably during channel activation. This may indicate that TRPV1 and TRPV2 are the only members of the TRPV subfamily that possess a gate at the SF, and that the coupling of structural elements necessary for activation of these channels differs from that of TRPV3-TRPV6 (Zhang et al., 2019).

As observed in our previous study (Zubcevic et al., 2018a), RTx assumes different binding poses in subunits of the C2-symmetric structures, both in the amphipol and the nanodisc samples (Figure 4—figure supplement 3) which may lead to the distinct conformations observed in these channels.

Despite the use of a full-length rabbit TRPV2 construct in this study, we were not able to confidently resolve the entire loop connecting S5 to the pore helix known as the ‘pore turret’. Interestingly, a recent structure of rat TRPV2 with the pore turret resolved showed that this region, which contains a large number of charged and polar residues, occupies the space within the membrane plane between S5 and the Voltage Sensor-Like Domain (VSLD) (Dosey et al., 2019). While the density in our cryo-EM maps was not of sufficient quality to build the entire pore turret with confidence, we do observe density following the S5 helix and preceding the pore helix. However, the direction of this density is perpendicular to the membrane and does not agree with the structure reported for rat TRPV2 (Figure 4—figure supplement 4). Indeed, the pore turret is the least conserved region amongst the TRPV2 orthologs, and the variations in its sequence might indicate that the turret adopts different conformations in TRPV2 channels from different species. Nevertheless, our study clearly shows that the omission of this region from the construct used in the crystallographic study of the TRPV2/RTx complex is not the cause of the C2 symmetry.

While both TRPV2_RTx-ND and TRPV2_RTx-XTAL structures adopt C2 symmetry, the distinct arrangement of subunits within the two channels suggests that the structures represent different functional states. We propose that TRPV2_RTx-XTAL precedes TRPV2_RTx-ND in the conformational activation trajectory based on two observations. Firstly, the common gate is fully closed in the TRPV2_RTx-XTAL while it adopts an apparently partially open conformation in TRPV2_RTx-ND (Figure 4—figure supplement 5). Secondly, our previous studies have shown that the putative hydrogen bond network between S5 and S6 and the pore helix is essential for the channel’s ability to transition to a fully open SF that can accommodate large organic cations (Zubcevic et al., 2018a). Nevertheless, in TRPV2_RTx-ND the pore helices do not interact with S5 and S6 and the SF is fully open. We therefore propose that the conformational step that requires the presence of the putative hydrogen bond triad precedes the open SF conformation seen in TRPV2_RTx-ND.

Here, we have conducted a study that reveals symmetry transitions associated with gating of the TRPV2 channel by RTx. Interestingly, our data shows that RTx induces C2 symmetric conformations of TRPV2 in both amphipol and nanodiscs, and it thereby negates the hypothetical role of crystallization artefacts and crystal packing bias in stabilising two-fold symmetry. Similarly, C2 symmetry in TRPV2 is independent of the presence or absence of the pore turret region, suggesting that this region does not play an essential role in the regulation of the SF in rabbit TRPV2. Our study, similar to a previously published study of the magnesium channel CorA (Matthies et al., 2016), also emphasizes the notion that careful inspection of the intermediate maps and conservative application of symmetry during refinement of cryo-EM data can result in valuable insights into gating transitions and intermediate states. In addition, we have also investigated how amphipols and nanodiscs affect the conformational space that can be accessed during ligand gating of TRPV2.

While both TRPV2_RTx-APOL and TRPV2_RTx-ND are C2 symmetric, the two-fold symmetry in TRPV2_RTx-APOL is confined to regions that are not bound by the amphipol polymer. This is evident in the fact that the TM domains, which are in contact with the amphipol, largely retain four-fold symmetry and the common gate and the SF remain firmly closed, while the ARD exhibit symmetry breaking, rotation and lateral expansion. These data, while adding valuable data points to the conformational landscape of TRPV2, also illustrate the potential caveats of using amphipols in studies of conformational changes in the transmembrane domains of proteins, as they appear to constrict the TM domains and stabilize low-energy pre-open states (Figure 4—figure supplement 6). However, at this time we caution against any generalized conclusions about the effect of amphipols and look forward to more systematic studies that will address these issues in the future. The TRPV2_RTx-ND dataset yielded a single, two-fold symmetric structure thus giving strong evidence that RTx stabilizes two-fold symmetric conformational states in the TRPV2 channel in lipid membranes. The ARDs in the TRPV2_RTx-ND structure echo the conformational changes observed in TRPV2_RTx-APOL. However, in nanodiscs TRPV2 is captured with its SF fully open and its common gate in an intermediate conformation where the gate is apparently open in one set of diagonally opposing subunits and closed in the other. In this structure, the opening of the SF occurs according to a mechanism previously observed in the crystallographic study of the TRPV2/RTx complex where RTx binding in the vanilloid pocket, above the S4-S5_π-hinge, induces a rigid body rotation of the entire subunit. In turn, the rotation causes a break in the interaction network between the pore helix and helices S5 and S6, allowing the pore helices to reposition and the SF to open (Zubcevic et al., 2018a).

Interestingly, however, the TRPV2_RTx-ND structure differs from the previously obtained TRPV2_RTx-XTAL. While both structures assume C2 symmetric conformations, the TRPV2_RTx-ND channel appears to make a return toward C4 symmetry. Because the SF in TRPV2_RTx-ND is fully open, and the common gate adopts an apparently partially open conformation, we reason that TRPV2_RTx-ND follows the TRPV2_RTx-XTAL structure in the conformational trajectory of the channel. Therefore, it is possible that TRPV2, as it travels toward the final open state where both the SF and the common gate are fully open, would adopt further conformations that increasingly approximate C4 symmetry (Figure 5). However, it is interesting to note that while the overall fold of TRPV2_RTx-ND indeed is more C4 symmetric than that of TRPV2_RTx-XTAL, the extent of C2 symmetry is not diminished in its SF. Because the symmetry of the SF does not appear to be dictated by the symmetry of the overall channel, we cannot exclude the possibility that the final open state might indeed possess a C2 symmetric SF while otherwise adopting a nearly C4 symmetric conformation. Our previous functional studies have suggested that C2 symmetric states play a role in the permeation of large organic cations and consequently in the full opening of the SF (Zubcevic et al., 2018a). Hence, the channel might be utilizing C2 symmetric states as means to achieve full opening in a step-wise manner. Similar C2 symmetric states elicited by ligand binding have been observed in TRPV3 (Zubcevic et al., 2018b) and TRPM2 (Yin et al., 2018) channels, which opens up the possibility that C2 symmetry might be widely associated with gating in members of the TRP channel superfamily. Intriguingly, a recent cryo-EM study of the human BK channel reconstituted in liposomes showed that this channel also enters C2 symmetric states (Tonggu and Wang, 2018), suggesting that two-fold symmetry might also play a role in the molecular mechanisms of other tetrameric ion channels.

Figure 5

Download asset Open asset

Two-fold symmetry is a well-stablished feature of mammalian Na⁺ selective Two Pore Channels (TPCs) and Voltage Gated Sodium channels (Na_V) (She et al., 2018; Shen et al., 2017; Pan et al., 2018; Shen et al., 2018). Interestingly, the arrangement of pore helices in TRPV2_RTx-ND resembles that observed in TPC and Na_V (Figure 6) and the selectivity filters in all three channels form a 'coin-slot' (Hille, 1971) opening. However, while the selectivity filters of TPC and Na_V remain static during channel gating in order to maintain the structure necessary for Na⁺ selectivity, the SF of TRPV2 displays a large degree of plasticity. Moreover, the two-fold symmetry observed in TRPV2 is unique in that it arises in response to conformational changes in the TM domains induced by ligand binding. By contrast, the two-fold symmetry in TPC and Na_V stems from the arrangement of their respective homologous tandem repeats.

Figure 6

Download asset Open asset

The EM maps and atomic models have been deposited with the Electron Microscopy Data Bank (accession numbers EMD-20143, EMD-20145, EMD-20146, and EMD-20148) and the Protein Data Back (entry codes 6OO3, 6OO4, 6OO5, and 6OO7), respectively.

1. 1. Zubcevic L
   2. Hsu AL
   3. Borgnia MJ
   4. Lee S-Y

   (2019) Electron Microscopy Data Bank

   ID EMD-20143. Cryo-EM structure of the C4-symmetric TRPV2/RTx complex in amphipol resolved to 2.9 A.

   http://www.ebi.ac.uk/pdbe/entry/emdb/EMD-20143
2. 1. Zubcevic L
   2. Hsu AL
   3. Borgnia MJ
   4. Lee S-Y

   (2019) Electron Microscopy Data Bank

   ID EMD-20145. Cryo-EM structure of the C2-symmetric TRPV2/RTx complex in amphipol resolved to 3.3 A.

   http://www.ebi.ac.uk/pdbe/entry/emdb/EMD-20145
3. 1. Zubcevic L
   2. Hsu AL
   3. Borgnia MJ
   4. Lee S-Y

   (2019) Electron Microscopy Data Bank

   ID EMD-20146. Cryo-EM structure of the C2-symmetric TRPV2/RTx complex in amphipol resolved to 4.2 A.

   http://www.ebi.ac.uk/pdbe/entry/emdb/EMD-20146
4. 1. Zubcevic L
   2. Hsu AL
   3. Borgnia MJ
   4. Lee S-Y

   (2019) Electron Microscopy Data Bank

   ID EMD-20148. Cryo-EM structure of the C2-symmetric TRPV2/RTx complex in nanodiscs.

   http://www.ebi.ac.uk/pdbe/entry/emdb/EMD-20148
5. 1. Zubcevic L
   2. Hsu AL
   3. Borgnia MJ
   4. Lee S-Y

   (2019) Protein Data Bank

   ID 6OO3. Cryo-EM structure of the C4-symmetric TRPV2/RTx complex in amphipol resolved to 2.9 A.

   http://www.rcsb.org/structure/6OO3
6. 1. Zubcevic L
   2. Hsu AL
   3. Borgnia MJ
   4. Lee S-Y

   (2019) Protein Data Bank

   ID 6OO4. Cryo-EM structure of the C2-symmetric TRPV2/RTx complex in amphipol resolved to 3.3 A.

   http://www.rcsb.org/structure/6OO4
7. 1. Zubcevic L
   2. Hsu AL
   3. Borgnia MJ
   4. Lee S-Y

   (2019) Protein Data Bank

   ID 6OO5. Cryo-EM structure of the C2-symmetric TRPV2/RTx complex in amphipol resolved to 4.2 A.

   http://www.rcsb.org/structure/6OO5
8. 1. Zubcevic L
   2. Hsu AL
   3. Borgnia MJ
   4. Lee S-Y

   (2019) Protein Data Bank

   ID 6OO7. Cryo-EM structure of the C2-symmetric TRPV2/RTx complex in nanodiscs.

   http://www.rcsb.org/structure/6OO7

1. 1. Adams PD
   2. Afonine PV
   3. Bunkóczi G
   4. Chen VB
   5. Davis IW
   6. Echols N
   7. Headd JJ
   8. Hung LW
   9. Kapral GJ
   10. Grosse-Kunstleve RW
   11. McCoy AJ
   12. Moriarty NW
   13. Oeffner R
   14. Read RJ
   15. Richardson DC
   16. Richardson JS
   17. Terwilliger TC
   18. Zwart PH

   (2010) PHENIX: a comprehensive Python-based system for macromolecular structure solution

   Acta Crystallographica Section D Biological Crystallography 66:213–221.

   https://doi.org/10.1107/S0907444909052925

   • PubMed
   • Google Scholar
2. 1. Asakawa M
   2. Yoshioka T
   3. Matsutani T
   4. Hikita I
   5. Suzuki M
   6. Oshima I
   7. Tsukahara K
   8. Arimura A
   9. Horikawa T
   10. Hirasawa T
   11. Sakata T

   (2006) Association of a mutation in TRPV3 with defective hair growth in rodents

   Journal of Investigative Dermatology 126:2664–2672.

   https://doi.org/10.1038/sj.jid.5700468

   • PubMed
   • Google Scholar
3. 1. Bölcskei K
   2. Tékus V
   3. Dézsi L
   4. Szolcsányi J
   5. Petho G

   (2010) Antinociceptive desensitizing actions of TRPV1 receptor agonists capsaicin, resiniferatoxin and N-oleoyldopamine as measured by determination of the noxious heat and cold thresholds in the rat

   European Journal of Pain 14:480–486.

   https://doi.org/10.1016/j.ejpain.2009.08.005

   • PubMed
   • Google Scholar
4. 1. Cao E
   2. Cordero-Morales JF
   3. Liu B
   4. Qin F
   5. Julius D

   (2013a) TRPV1 channels are intrinsically heat sensitive and negatively regulated by phosphoinositide lipids

   Neuron 77:667–679.

   https://doi.org/10.1016/j.neuron.2012.12.016

   • PubMed
   • Google Scholar
5. 1. Cao E
   2. Liao M
   3. Cheng Y
   4. Julius D

   (2013b) TRPV1 structures in distinct conformations reveal activation mechanisms

   Nature 504:113–118.

   https://doi.org/10.1038/nature12823

   • PubMed
   • Google Scholar
6. 1. Chen VB
   2. Arendall WB
   3. Headd JJ
   4. Keedy DA
   5. Immormino RM
   6. Kapral GJ
   7. Murray LW
   8. Richardson JS
   9. Richardson DC

   (2010) MolProbity: all-atom structure validation for macromolecular crystallography

   Acta Crystallographica Section D Biological Crystallography 66:12–21.

   https://doi.org/10.1107/S0907444909042073

   • PubMed
   • Google Scholar
7. 1. Chen S
   2. McMullan G
   3. Faruqi AR
   4. Murshudov GN
   5. Short JM
   6. Scheres SH
   7. Henderson R

   (2013) High-resolution noise substitution to measure overfitting and validate resolution in 3D structure determination by single particle electron cryomicroscopy

   Ultramicroscopy 135:24–35.

   https://doi.org/10.1016/j.ultramic.2013.06.004

   • PubMed
   • Google Scholar
8. 1. Chung MK
   2. Lee H
   3. Caterina MJ

   (2003) Warm temperatures activate TRPV4 in mouse 308 keratinocytes

   Journal of Biological Chemistry 278:32037–32046.

   https://doi.org/10.1074/jbc.M303251200

   • PubMed
   • Google Scholar
9. 1. Chung MK
   2. Güler AD
   3. Caterina MJ

   (2008) TRPV1 shows dynamic ionic selectivity during agonist stimulation

   Nature Neuroscience 11:555–564.

   https://doi.org/10.1038/nn.2102

   • PubMed
   • Google Scholar
10. 1. Clapham DE
    2. Runnels LW
    3. Strübing C

    (2001) The TRP ion channel family

    Nature Reviews Neuroscience 2:387–396.

    https://doi.org/10.1038/35077544

    • PubMed
    • Google Scholar
11. 1. Deng Z
    2. Paknejad N
    3. Maksaev G
    4. Sala-Rabanal M
    5. Nichols CG
    6. Hite RK
    7. Yuan P

    (2018) Cryo-EM and X-ray structures of TRPV4 reveal insight into ion permeation and gating mechanisms

    Nature Structural & Molecular Biology 25:252–260.

    https://doi.org/10.1038/s41594-018-0037-5

    • PubMed
    • Google Scholar
12. 1. Denisov IG
    2. Sligar SG

    (2016) Nanodiscs for structural and functional studies of membrane proteins

    Nature Structural & Molecular Biology 23:481–486.

    https://doi.org/10.1038/nsmb.3195

    • PubMed
    • Google Scholar
13. 1. Dosey TL
    2. Wang Z
    3. Fan G
    4. Zhang Z
    5. Serysheva II
    6. Chiu W
    7. Wensel TG

    (2019) Structures of TRPV2 in distinct conformations provide insight into role of the pore turret

    Nature Structural & Molecular Biology 26:40–49.

    https://doi.org/10.1038/s41594-018-0168-8

    • PubMed
    • Google Scholar
14. 1. Emsley P
    2. Cowtan K

    (2004) Coot: model-building tools for molecular graphics

    Acta Crystallographica Section D Biological Crystallography 60:2126–2132.

    https://doi.org/10.1107/S0907444904019158

    • PubMed
    • Google Scholar
15. 1. Eytan O
    2. Fuchs-Telem D
    3. Mevorach B
    4. Indelman M
    5. Bergman R
    6. Sarig O
    7. Goldberg I
    8. Adir N
    9. Sprecher E

    (2014) Olmsted syndrome caused by a homozygous recessive mutation in TRPV3

    Journal of Investigative Dermatology 134:1752–1754.

    https://doi.org/10.1038/jid.2014.37

    • PubMed
    • Google Scholar
16. 1. Hille B

    (1971) The permeability of the sodium channel to organic cations in myelinated nerve

    The Journal of General Physiology 58:599–619.

    https://doi.org/10.1085/jgp.58.6.599

    • PubMed
    • Google Scholar
17. 1. Hirschi M
    2. Herzik MA
    3. Wie J
    4. Suo Y
    5. Borschel WF
    6. Ren D
    7. Lander GC
    8. Lee SY

    (2017) Cryo-electron microscopy structure of the lysosomal calcium-permeable channel TRPML3

    Nature 550:411–414.

    https://doi.org/10.1038/nature24055

    • PubMed
    • Google Scholar
18. 1. Hughes TET
    2. Pumroy RA
    3. Yazici AT
    4. Kasimova MA
    5. Fluck EC
    6. Huynh KW
    7. Samanta A
    8. Molugu SK
    9. Zhou ZH
    10. Carnevale V
    11. Rohacs T
    12. Moiseenkova-Bell VY

    (2018) Structural insights on TRPV5 gating by endogenous modulators

    Nature Communications 9:4198.

    https://doi.org/10.1038/s41467-018-06753-6

    • PubMed
    • Google Scholar
19. 1. Huynh KW
    2. Cohen MR
    3. Jiang J
    4. Samanta A
    5. Lodowski DT
    6. Zhou ZH
    7. Moiseenkova-Bell VY

    (2016) Structure of the full-length TRPV2 channel by cryo-EM

    Nature Communications 7:11130.

    https://doi.org/10.1038/ncomms11130

    • PubMed
    • Google Scholar
20. 1. Imura K
    2. Yoshioka T
    3. Hikita I
    4. Tsukahara K
    5. Hirasawa T
    6. Higashino K
    7. Gahara Y
    8. Arimura A
    9. Sakata T

    (2007) Influence of TRPV3 mutation on hair growth cycle in mice

    Biochemical and Biophysical Research Communications 363:479–483.

    https://doi.org/10.1016/j.bbrc.2007.08.170

    • PubMed
    • Google Scholar
21. 1. Imura K
    2. Yoshioka T
    3. Hirasawa T
    4. Sakata T

    (2009) Role of TRPV3 in immune response to development of dermatitis

    Journal of Inflammation 6:17.

    https://doi.org/10.1186/1476-9255-6-17

    • PubMed
    • Google Scholar
22. 1. Jara-Oseguera A
    2. Bae C
    3. Swartz KJ

    (2016) An external sodium ion binding site controls allosteric gating in TRPV1 channels

    eLife 5:e13356.

    https://doi.org/10.7554/eLife.13356

    • PubMed
    • Google Scholar
23. 1. Julius D

    (2013) TRP channels and pain

    Annual Review of Cell and Developmental Biology 29:355–384.

    https://doi.org/10.1146/annurev-cellbio-101011-155833

    • PubMed
    • Google Scholar
24. 1. Katanosaka Y
    2. Iwasaki K
    3. Ujihara Y
    4. Takatsu S
    5. Nishitsuji K
    6. Kanagawa M
    7. Sudo A
    8. Toda T
    9. Katanosaka K
    10. Mohri S
    11. Naruse K

    (2014) TRPV2 is critical for the maintenance of cardiac structure and function in mice

    Nature Communications 5:3932.

    https://doi.org/10.1038/ncomms4932

    • PubMed
    • Google Scholar
25. 1. Kim HO
    2. Cho YS
    3. Park SY
    4. Kwak IS
    5. Choi MG
    6. Chung BY
    7. Park CW
    8. Lee JY

    (2016) Increased activity of TRPV3 in keratinocytes in hypertrophic burn scars with postburn pruritus

    Wound Repair and Regeneration 24:841–850.

    https://doi.org/10.1111/wrr.12469

    • PubMed
    • Google Scholar
26. 1. Liao M
    2. Cao E
    3. Julius D
    4. Cheng Y

    (2013) Structure of the TRPV1 ion channel determined by electron cryo-microscopy

    Nature 504:107–112.

    https://doi.org/10.1038/nature12822

    • PubMed
    • Google Scholar
27. 1. Liu B
    2. Qin F

    (2016) Use Dependence of Heat Sensitivity of Vanilloid Receptor TRPV2

    Biophysical Journal 110:1523–1537.

    https://doi.org/10.1016/j.bpj.2016.03.005

    • PubMed
    • Google Scholar
28. 1. Marwaha L
    2. Bansal Y
    3. Singh R
    4. Saroj P
    5. Bhandari R
    6. Kuhad A

    (2016) TRP channels: potential drug target for neuropathic pain

    Inflammopharmacology 24:305–317.

    https://doi.org/10.1007/s10787-016-0288-x

    • PubMed
    • Google Scholar
29. 1. Masuyama R
    2. Vriens J
    3. Voets T
    4. Karashima Y
    5. Owsianik G
    6. Vennekens R
    7. Lieben L
    8. Torrekens S
    9. Moermans K
    10. Vanden Bosch A
    11. Bouillon R
    12. Nilius B
    13. Carmeliet G

    (2008) TRPV4-mediated calcium influx regulates terminal differentiation of osteoclasts

    Cell Metabolism 8:257–265.

    https://doi.org/10.1016/j.cmet.2008.08.002

    • PubMed
    • Google Scholar
30. 1. Matthies D
    2. Dalmas O
    3. Borgnia MJ
    4. Dominik PK
    5. Merk A
    6. Rao P
    7. Reddy BG
    8. Islam S
    9. Bartesaghi A
    10. Perozo E
    11. Subramaniam S

    (2016) Cryo-EM Structures of the Magnesium Channel CorA Reveal Symmetry Break upon Gating

    Cell 164:747–756.

    https://doi.org/10.1016/j.cell.2015.12.055

    • PubMed
    • Google Scholar
31. 1. McGoldrick LL
    2. Singh AK
    3. Saotome K
    4. Yelshanskaya MV
    5. Twomey EC
    6. Grassucci RA
    7. Sobolevsky AI

    (2018) Opening of the human epithelial calcium channel TRPV6

    Nature 553:233–237.

    https://doi.org/10.1038/nature25182

    • PubMed
    • Google Scholar
32. 1. Mitchell K
    2. Lebovitz EE
    3. Keller JM
    4. Mannes AJ
    5. Nemenov MI
    6. Iadarola MJ

    (2014) Nociception and inflammatory hyperalgesia evaluated in rodents using infrared laser stimulation after Trpv1 gene knockout or resiniferatoxin lesion

    Pain 155:733–745.

    https://doi.org/10.1016/j.pain.2014.01.007

    • PubMed
    • Google Scholar
33. 1. Pan X
    2. Li Z
    3. Zhou Q
    4. Shen H
    5. Wu K
    6. Huang X
    7. Chen J
    8. Zhang J
    9. Zhu X
    10. Lei J
    11. Xiong W
    12. Gong H
    13. Xiao B
    14. Yan N

    (2018) Structure of the human voltage-gated sodium channel Na_v1.4 in complex with β1

    Science 362:eaau2486.

    https://doi.org/10.1126/science.aau2486

    • PubMed
    • Google Scholar
34. 1. Paulsen CE
    2. Armache JP
    3. Gao Y
    4. Cheng Y
    5. Julius D

    (2015) Structure of the TRPA1 ion channel suggests regulatory mechanisms

    Nature 520:511–517.

    https://doi.org/10.1038/nature14367

    • PubMed
    • Google Scholar
35. 1. Pettersen EF
    2. Goddard TD
    3. Huang CC
    4. Couch GS
    5. Greenblatt DM
    6. Meng EC
    7. Ferrin TE

    (2004) UCSF Chimera--a visualization system for exploratory research and analysis

    Journal of Computational Chemistry 25:1605–1612.

    https://doi.org/10.1002/jcc.20084

    • PubMed
    • Google Scholar
36. 1. Puopolo M
    2. Binshtok AM
    3. Yao GL
    4. Oh SB
    5. Woolf CJ
    6. Bean BP

    (2013) Permeation and block of TRPV1 channels by the cationic lidocaine derivative QX-314

    Journal of Neurophysiology 109:1704–1712.

    https://doi.org/10.1152/jn.00012.2013

    • PubMed
    • Google Scholar
37. 1. Ritchie TK
    2. Grinkova YV
    3. Bayburt TH
    4. Denisov IG
    5. Zolnerciks JK
    6. Atkins WM
    7. Sligar SG

    (2009) Chapter 11 - Reconstitution of membrane proteins in phospholipid bilayer nanodiscs

    Methods in Enzymology 464:211–231.

    https://doi.org/10.1016/S0076-6879(09)64011-8

    • PubMed
    • Google Scholar
38. 1. Scheres SH

    (2012) RELION: implementation of a Bayesian approach to cryo-EM structure determination

    Journal of Structural Biology 180:519–530.

    https://doi.org/10.1016/j.jsb.2012.09.006

    • PubMed
    • Google Scholar
39. 1. Scheres SH
    2. Chen S

    (2012) Prevention of overfitting in cryo-EM structure determination

    Nature Methods 9:853–854.

    https://doi.org/10.1038/nmeth.2115

    • PubMed
    • Google Scholar
40. 1. She J
    2. Guo J
    3. Chen Q
    4. Zeng W
    5. Jiang Y
    6. Bai XC

    (2018) Structural insights into the voltage and phospholipid activation of the mammalian TPC1 channel

    Nature 556:130–134.

    https://doi.org/10.1038/nature26139

    • PubMed
    • Google Scholar
41. 1. Shen H
    2. Zhou Q
    3. Pan X
    4. Li Z
    5. Wu J
    6. Yan N

    (2017) Structure of a eukaryotic voltage-gated sodium channel at near-atomic resolution

    Science 355:eaal4326.

    https://doi.org/10.1126/science.aal4326

    • PubMed
    • Google Scholar
42. 1. Shen H
    2. Li Z
    3. Jiang Y
    4. Pan X
    5. Wu J
    6. Cristofori-Armstrong B
    7. Smith JJ
    8. Chin YKY
    9. Lei J
    10. Zhou Q
    11. King GF
    12. Yan N

    (2018) Structural basis for the modulation of voltage-gated sodium channels by animal toxins

    Science 362:eaau2596.

    https://doi.org/10.1126/science.aau2596

    • PubMed
    • Google Scholar
43. 1. Singh AK
    2. McGoldrick LL
    3. Sobolevsky AI

    (2018) Structure and gating mechanism of the transient receptor potential channel TRPV3

    Nature Structural & Molecular Biology 25:805–813.

    https://doi.org/10.1038/s41594-018-0108-7

    • PubMed
    • Google Scholar
44. 1. Smart OS
    2. Neduvelil JG
    3. Wang X
    4. Wallace BA
    5. Sansom MS

    (1996) HOLE: a program for the analysis of the pore dimensions of ion channel structural models

    Journal of Molecular Graphics 14:360–376.

    https://doi.org/10.1016/S0263-7855(97)00009-X

    • PubMed
    • Google Scholar
45. 1. Smith GD
    2. Gunthorpe MJ
    3. Kelsell RE
    4. Hayes PD
    5. Reilly P
    6. Facer P
    7. Wright JE
    8. Jerman JC
    9. Walhin JP
    10. Ooi L
    11. Egerton J
    12. Charles KJ
    13. Smart D
    14. Randall AD
    15. Anand P
    16. Davis JB

    (2002) TRPV3 is a temperature-sensitive vanilloid receptor-like protein

    Nature 418:186–190.

    https://doi.org/10.1038/nature00894

    • PubMed
    • Google Scholar
46. Preprint

    1. Tonggu L
    2. Wang L

    (2018) Broken symmetry in the human BK channel

    bioRxiv.

    https://doi.org/10.1101/494385

    • Google Scholar
47. 1. Xiao R
    2. Tian J
    3. Tang J
    4. Zhu MX

    (2008) The TRPV3 mutation associated with the hairless phenotype in rodents is constitutively active

    Cell Calcium 43:334–343.

    https://doi.org/10.1016/j.ceca.2007.06.004

    • PubMed
    • Google Scholar
48. 1. Yao J
    2. Liu B
    3. Qin F

    (2010) Pore turret of thermal TRP channels is not essential for temperature sensing

    PNAS 107:E125.

    https://doi.org/10.1073/pnas.1008272107

    • Google Scholar
49. Preprint

    1. Yin Y
    2. Wu M
    3. Hsu A
    4. Borschel WF
    5. Borgnia MJ
    6. Lander GC
    7. Lee SY

    (2018) Visualizing structural transitions of ligand-dependent gating of the TRPM2 channel

    bioRxiv.

    https://doi.org/10.1101/516468

    • Google Scholar
50. 1. Yoo J
    2. Wu M
    3. Yin Y
    4. Herzik MA
    5. Lander GC
    6. Lee S-Y

    (2018) Cryo-EM structure of a mitochondrial calcium uniporter

    Science 361:eaar4056.

    https://doi.org/10.1126/science.aar4056

    • Google Scholar
51. 1. Zhang F
    2. Hanson SM
    3. Jara-Oseguera A
    4. Krepkiy D
    5. Bae C
    6. Pearce LV
    7. Blumberg PM
    8. Newstead S
    9. Swartz KJ

    (2016) Engineering vanilloid-sensitivity into the rat TRPV2 channel

    eLife 5:e16409.

    https://doi.org/10.7554/eLife.16409

    • PubMed
    • Google Scholar
52. 1. Zhang K

    (2016) Gctf: Real-time CTF determination and correction

    Journal of Structural Biology 193:1–12.

    https://doi.org/10.1016/j.jsb.2015.11.003

    • PubMed
    • Google Scholar
53. 1. Zhang F
    2. Swartz KJ
    3. Jara-Oseguera A

    (2019) Conserved allosteric pathways for activation of TRPV3 revealed through engineering vanilloid-sensitivity

    eLife 8:e42756.

    https://doi.org/10.7554/eLife.42756

    • PubMed
    • Google Scholar
54. 1. Zheng SQ
    2. Palovcak E
    3. Armache JP
    4. Verba KA
    5. Cheng Y
    6. Agard DA

    (2017) MotionCor2: anisotropic correction of beam-induced motion for improved cryo-electron microscopy

    Nature Methods 14:331–332.

    https://doi.org/10.1038/nmeth.4193

    • PubMed
    • Google Scholar
55. 1. Zivanov J
    2. Nakane T
    3. Forsberg BO
    4. Kimanius D
    5. Hagen WJ
    6. Lindahl E
    7. Scheres SH

    (2018) New tools for automated high-resolution cryo-EM structure determination in RELION-3

    eLife 7:e42166.

    https://doi.org/10.7554/eLife.42166

    • PubMed
    • Google Scholar
56. 1. Zoonens M
    2. Popot JL

    (2014) Amphipols for each season

    The Journal of Membrane Biology 247:759–796.

    https://doi.org/10.1007/s00232-014-9666-8

    • PubMed
    • Google Scholar
57. 1. Zubcevic L
    2. Herzik MA
    3. Chung BC
    4. Liu Z
    5. Lander GC
    6. Lee SY

    (2016) Cryo-electron microscopy structure of the TRPV2 ion channel

    Nature Structural & Molecular Biology 23:180–186.

    https://doi.org/10.1038/nsmb.3159

    • PubMed
    • Google Scholar
58. 1. Zubcevic L
    2. Le S
    3. Yang H
    4. Lee SY

    (2018a) Conformational plasticity in the selectivity filter of the TRPV2 ion channel

    Nature Structural & Molecular Biology 25:405–415.

    https://doi.org/10.1038/s41594-018-0059-z

    • PubMed
    • Google Scholar
59. 1. Zubcevic L
    2. Herzik MA
    3. Wu M
    4. Borschel WF
    5. Hirschi M
    6. Song AS
    7. Lander GC
    8. Lee SY

    (2018b) Conformational ensemble of the human TRPV3 ion channel

    Nature Communications 9:4773.

    https://doi.org/10.1038/s41467-018-07117-w

    • PubMed
    • Google Scholar

Thank you for submitting your article "Symmetry transitions during gating of the TRPV2 ion channel in lipid membranes" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by Kenton Swartz as the Reviewing Editor and Richard Aldrich as the Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Richard K Hite (Reviewer #3); Raimund Dutzler (Reviewer #4).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

In this report by Zubcevic and colleagues, single-particle cryo-EM structures of TRPV2 in amphipols and lipid nanodiscs in complex with the agonist resiniferatoxin (RTx) are presented. Three-dimensional classification of the amphipol-reconstituted channels reveal three distinct conformations, two of which display C2 symmetry and a third which displays C4 symmetry. Classification of the nanodisc-reconstituted channels reveal just a single conformation which displays C2 symmetry. Previously, the same group had reported a ligand-free, closed cryo-EM structure of TRPV2 which displayed C4 symmetry and a RTx-bound C2 symmetric crystal structure with an open selectivity filter gate that displayed two-fold symmetry and a closed common gate that displayed four-fold symmetry. Comparison of the ampiphol-reconstituted structures with the previous structures indicated that in all three conformations the TM domain adopts a nearly four-fold symmetric structure that resembles the ligand-free, closed state, while the ARD is flexible and adopts a range of different conformations. In contrast, large conformational changes are seen throughout the nanodisc-reconstituted structure. The selectivity filter gate is opened in a 2-fold symmetric fashion by alternating rotations of the pore helices in adjacent subunits. The common gate also adopts a 2-fold symmetric conformation, with two of the subunits adopting a closed state and two adopting an open state. Comparison of the nanodisc-reconstituted sample with the previous RTx-bound crystal structure suggested that they represent distinct conformational states with the crystal structure which contained a fully closed common gate potentially representing an earlier state in the transition from closed to an open state. This work provides evidence that RTx binding to TRPV2 induces two-fold symmetric pre-activated states in a lipid bilayer environment and will be of general interest. We request that you address the following issues in revision.

Essential revisions:

1) Some of the conformational differences observed between the structures or between subunits from a single structure are small (see Figure 3—figure supplement 1C, D; Figure 3—figure supplement 2B, C), or reflect changes in the patterns of hydrogen bonding such as the α- to π-helix transitions. Given that the resolution of the structures is not very high, the authors should provide structural comparisons including the experimental density maps to support their proposed conformational differences, rather than simply using the structural models. The density maps should be used to provide stronger evidence that the pattern of main-chain hydrogen bonds has indeed been disrupted in the S4-S5 and S6 helices to give rise to π-helical bulges. As symmetry is a major feature of this work, it would helpful to compare the model refined using the symmetrized reconstructions in density maps calculated without symmetry. This is particularly important for the selectivity filter and common gates of the four subunits in the nanodisc-reconstituted reconstruction.

2) Why is map-to-model fit so poor for TRPV2_RTx-APOL1? The map-to-model fit has a 0.5 correlation at approximately 4Å while the FSC between half-maps is approximately 2.9 Å. How was this model refined into the density map? What does nominal resolution mean? Why is it different from FSC=0.143 in Table 1 and why are the unmasked resolution estimates higher than the masked resolution estimates? Any masks used for classification or refinement should be displayed in the figure supplements.

3) The authors suggest that the ampiphol restrict the flexibility of the TM domain. It would be helpful to compare the belt of ampiphol-reconstituted samples with those of the nanodisc-reconstituted samples to see if sufficient space is available for the channel to adopt the conformation of the nanodisc-reconstituted channels. Were lipid molecules resolved in the nanodisc structures that may provide the flexibility required to adopt the different conformation?

4) It would be interesting to know how many particles contribute to the different 3D classes displayed Figure 1—figure supplement 2. Do the particles in class 1 shown in Figure 1—figure supplement 1 really display and unsymmetrical (C1) channel and can we learn anything from this low-resolution structure. While it is possible that the X-ray structure of TRPV2-RTx complex could be an intermediate on the activation pathway, the authors might also consider the possibility that the pronounced C2-symmetric features observed in this structure would reflect the conditions in the crystalline environment and might thus not be an intermediate on the activation path.

5) Even though RTx activates both TRPV1 and TRPV2 with an extremely high affinity and efficacy, the kinetics of activation are distinctly slow, even when compared to other vanilloids with lower affinity such as capsaicin in the case of TRPV1. This raises the possibility that RTx can interact with the pocket in multiple orientations, with only some of them leading to channel activation. At present, it is not possible to determine whether the structures observed in this study are indeed reflecting RTx-channel complexes that are components of the activation pathway, or whether they reflect off-pathway intermediates that are not conducing to activation. It is intriguing that no open state has been determined even though RTx activates TRPV2-QM with very high affinity and efficacy. Are the binding poses of RTx inferred from the densities different in all structures? Are they different between subunits in a given structure?

6) The text makes constant reference to a selectivity filter activation gate in these channels. However, there is currently no functional evidence indicating that the selectivity filter does indeed function as an activation gate. This should be reflected in the text.

7) In several points throughout the manuscript (subsection “RTx induces a break in symmetry in TRPV2_RTx-ND”, last paragraph), the authors mention that their previous work demonstrated that a network of hydrogen bonds between the pore helix, the S5 and the S6 is essential for activation. However, the results in the previous publication do not provide conclusive evidence for the involvement of hydrogen bonds, but instead show that alanine mutations in that region affect cation permeation through an unknown mechanism that could involve a network of hydrogen bonds. This should be clearly stated in the manuscript.

8) In the third paragraph of the Discussion, the authors mention that their previous work has shown that C2 symmetric states are critical for permeation of large organic cations. We think it is more accurate to state that their data suggests that C2 symmetry could be involved in channel activation by RTx, but it does not provide conclusive evidence that C2 symmetries are directly connected to channel activation or to the mechanism of cation permeation in an open channel. No experiments were done in that study to directly probe the functional consequences of the proposed asymmetries in the tetrameric channel.

Essential revisions:

1) Some of the conformational differences observed between the structures or between subunits from a single structure are small (see Figure 3—figure supplement 1C, D; Figure 3—figure supplement 2B, C), or reflect changes in the patterns of hydrogen bonding such as the α- to π-helix transitions. Given that the resolution of the structures is not very high, the authors should provide structural comparisons including the experimental density maps to support their proposed conformational differences, rather than simply using the structural models. The density maps should be used to provide stronger evidence that the pattern of main-chain hydrogen bonds has indeed been disrupted in the S4-S5 and S6 helices to give rise to π-helical bulges.

We have now included figure supplements showing the density around the π-helices in all structures (Figure 3—figure supplement 4). However, we agree with the reviewer that the resolution of the TRPV2_RTx-ND of the map is not very high enough to define π-helix on S6 unambiguously and thus we have removed references in the text and figures to the S6 π-helix in structure throughout the text, and refer them to as bent helices instead.

As symmetry is a major feature of this work, it would helpful to compare the model refined using the symmetrized reconstructions in density maps calculated without symmetry. This is particularly important for the selectivity filter and common gates of the four subunits in the nanodisc-reconstituted reconstruction.

We thank reviewers for the great suggestion. During processing, we merged classes 1 and 6, which displayed a high overall similarity (CC=0.98) and then proceeded with refinement with C1 imposed. Because visual inspection indicated that this reconstruction possesses two-fold symmetry, we further refined the combined particles with C2 symmetry imposed. We have now also calculated the symmetry of the map combining classes 1 and 6 using the “Map Symmetry” function implemented in the Phenix suite. The best solution is C2 symmetry (score 1.28, CC=0.90). (Score reflects the square root of the number of elements in the symmetry multiplied by the map correlation for that symmetry, so the max score for C2 symmetry is 1.414 and 2 is the max score for C4 symmetry). We also calculated the score for C4 symmetry (score=1.6, CC=0.80) which indicates that C2 symmetry is more appropriate for this map. We have now also made a remark in the manuscript describing this analysis (subsection “RTx induces a break in symmetry in TRPV2_RTx-ND”, first paragraph).

For the purpose of this review, we have now included an FSC curve which estimates the fit of the model, originally built into the C2 map, into the C1 reconstruction. In addition, we have now also performed separate refinements of classes 1 and 6 with no symmetry imposed and estimated the fit of the TRPV2_RTx-ND atomic model, built into the C2 symmetry, to these non-symmetrized maps. This data is shown in Figure 1—figure supplement 3. Figure 1—figure supplement 3 now also shows the density around the critical regions (pore helices and S6 helices) in the class 1+class 6 non-symmetrized map. A mention of this is now included in the manuscript (see the second paragraph of the aforementioned subsection).

2) Why is map-to-model fit so poor for TRPV2_RTx-APOL1? The map-to-model fit has a 0.5 correlation at approximately 4Å while the FSC between half-maps is approximately 2.9 Å. How was this model refined into the density map? What does nominal resolution mean? Why is it different from FSC=0.143 in Table 1 and why are the unmasked resolution estimates higher than the masked resolution estimates? Any masks used for classification or refinement should be displayed in the figure supplements.

We thank the reviewers for pointing this out. A mistake was made in the plotting of the APOL FSC curves during our initial manuscript preparation. This has now been corrected and the new FSC curves placed in Figure 1—figure supplement 1.

3) The authors suggest that the ampiphol restrict the flexibility of the TM domain. It would be helpful to compare the belt of ampiphol-reconstituted samples with those of the nanodisc-reconstituted samples to see if sufficient space is available for the channel to adopt the conformation of the nanodisc-reconstituted channels. Were lipid molecules resolved in the nanodisc structures that may provide the flexibility required to adopt the different conformation?

We have now included a figure (Figure 4—figure supplement 6) showing the amphipol and nanodisc clouds surrounding the channels in the two different reconstitution conditions. We used the common nanodisc scaffold protein MSP2N2, which has been successfully used in many structural studies of TRP channels. MSP2N2 has an estimated diameter of ~17 nm and is large enough to accommodate TRPV2. The size of nanodisc is largely defined by the scaffold protein while amphipol polymers surround membrane proteins in a way that is similar to detergents. The quality of EM density for the nanodisc cloud is low and neither the scaffold protein nor the bound lipid molecules are well resolved. It is therefore difficult to say whether the lipids stabilize the captured conformation. However, a side-by-side comparison of the amphipol and nanodisc structures indicates that the nanodisc adopts a distinctly oval shape to accommodate the TRPV2-RTx complex, while amphipol retains a squarer form.

4) It would be interesting to know how many particles contribute to the different 3D classes displayed Figure 1—figure supplement 2.

This has now been included in the figure.

Do the particles in class 1 shown in Figure 1—figure supplement 1 really display and unsymmetrical (C1) channel and can we learn anything from this low-resolution structure.

We did indeed try to refine this class, but the resolution never got past ~7 Å. We therefore did not feel comfortable presenting this data or commenting on its symmetry and conformational state. We have now included a sentence stating this in the manuscript.

While it is possible that the X-ray structure of TRPV2-RTx complex could be an intermediate on the activation pathway, the authors might also consider the possibility that the pronounced C2-symmetric features observed in this structure would reflect the conditions in the crystalline environment and might thus not be an intermediate on the activation path.

This point was addressed in the paper published in NSMB (Zubcevic et al., 2018). In brief, we conducted electrophysiological experiments to probe the validity of the structure and found that the interactions that were uniquely found in the TRPV2-RTx crystal structure were of critical importance for channel function and that the structure was therefore not an artifact of crystallization. We also discussed the possibility of crystallization artifacts in the Introduction of the current manuscript.

5) Even though RTx activates both TRPV1 and TRPV2 with an extremely high affinity and efficacy, the kinetics of activation are distinctly slow, even when compared to other vanilloids with lower affinity such as capsaicin in the case of TRPV1. This raises the possibility that RTx can interact with the pocket in multiple orientations, with only some of them leading to channel activation. At present, it is not possible to determine whether the structures observed in this study are indeed reflecting RTx-channel complexes that are components of the activation pathway, or whether they reflect off-pathway intermediates that are not conducing to activation. It is intriguing that no open state has been determined even though RTx activates TRPV2-QM with very high affinity and efficacy. Are the binding poses of RTx inferred from the densities different in all structures? Are they different between subunits in a given structure?

We have observed that RTx can bind to TRPV2 in different modes, which we believe contributes significantly to the observed C2 symmetry. We have addressed this point in our previous publication as well as this study, but our clarification may have been somewhat insufficient. We have now included a figure supplement to show the binding poses of RTx in the different structures as well as in the subunits of different structures. In brief, in the C4 TRPV2_{RTx-APOL 1} structure, the RTx binding poses are the same in all subunits. They are also nearly identical in TRPV2_{RTx-APOL 2}. However, they differ from the ones captured in the C4 structure. In both TRPV2_{RTx-APOL 2} andTRPV2_RTx-ND,the binding poses are different within the subunits and they also differ from the poses observed in both TRPV2_{RTx-APOL 1} and TRPV2_{RTx-APOL 2}. Nevertheless, given the resolution of TRPV2_{RTx-APOL 3} and TRPV2_RTx-ND, the position of RTx in these structures should be taken with caution.

The point concerning the slow activation kinetics of RTx is excellent. In order to allow ample time between adding RTx to freezing the grids, we included RTx in our buffers from the very beginning of the purification. This is in contrast to the Julius-Cheng TRPV1/RTx/DkTx complex where the toxin was added to the sample ~30 minutes before freezing.

The point questioning whether the RTx-channel complexes observed here are components of the activation pathway is equally valid. The Swartz lab study that originally transplanted RTx-sensitivity into TRPV2 (Zhang et al., 2016) conducted recordings of single RTx-activated TRPV2 channels and found that these can achieve a high (~0.9) open probability. We hypothesize that the absence of a fully open state may point to the absence of unknown factors in our experimental conditions (e.g. the presence of the membrane bilayer rather than nanodisc). This is in line with observations made by many groups that agonist-bound ion channel structures often do not adopt fully open states.

However, we believe that the absence of the open state does not dismiss the rest of our findings as “off-pathway” conformations – the conformational changes that we observe in both the amphipol and nanodisc structures are in line with conformations associated with activation of closely related channels: the ARD rotation seen in TRPV1 (Cao et al., 2013; Liao et al., 2013) and the α-to-π transition in the S4-S5 linker in TRPV2 (Zubcevic et al., 2018). Notably, voltage-gated cation channels can assume multiple closed states before opening, and it is possible that TRPV2 channels also traverse a number of intermediate non-conducting states before entering the open conformation. It’s also worth pointing out that all of our data on the TRPV2/RTx complex, both crystallographic and cryo-EM, contains the same hot spots for conformational change: the S4-S5 linker and the pore helix. By contrast, pore of related channels TRPV3 and TRPV6 do not exhibit any change in the conformation of the pore helix during gating. And in addition, the pore of TRPV6 is preserved even in the structure of the point mutant which induces a non-domain swapped conformation in the channel, indicating that in TRPV6 the pore is a stable unit- as expected for a highly selective channel. In light of this, our TRPV2/RTx data points to the inherent flexibility of the pore helix in this channel.

Furthermore, we believe it unlikely that two-fold symmetry is an experimental artifact, as it’s only been observed in a few select TRP channels. Crystal and cryo-EM structures of K⁺ channels have never observed two-fold symmetry in the channel pores, and even the recent cryo-EM study of the K_V channel in nanodiscs (Matthies et al., 2018) is identical to the four-fold symmetric crystal structure solved by the MacKinnon lab. Along the same lines, the structures of Ca²⁺-selective TRPV5 and TRPV6, which have been solved in many different conditions (e.g. crystallization, cryo-EM, agonist and antagonist-bound, detergent, nanodisc, amphipol, mutations) have never been captured in non-C4 symmetric arrangements. These, like K⁺ channels, possess seemingly rigid pore structures (meaning extensive interactions between pore helices and the rest of pore), which is consistent with their observed strong preference for four-fold symmetry.

We think that the ability of TRPV2 to enter different two- and four-fold symmetric states is a reflection of the intrinsic flexibility in the TRPV2 channel pore and especially the pore helix. As we discussed in our previous paper, we think that this flexibility might be unique to TRPV1 and TRPV2 amongst TRPV channels and that it may endow them with the ability to permeate large organic cations.

At its core, the question posed here applies universally to all structural data: how well do cryo-EM or crystallization conditions sample the conformational landscape of proteins in their native environments? In this manuscript, we have evaluated our cryo-EM structures based our structural and functional data as well as the available literature and proposed a hypothetical sequence of events that may lead to TRPV2 channel opening.

6) The text makes constant reference to a selectivity filter activation gate in these channels. However, there is currently no functional evidence indicating that the selectivity filter does indeed function as an activation gate. This should be reflected in the text.

That is an excellent point for discussion. To make our long answer short, we think that the SF acts as an activation gate in TRPV1 and TRPV2, but not in other TRPV members (TRPV3-TRPV6). The references to the “selectivity filter gate” originated with the very first TRPV1 structure which showed that the channel possesses two restrictions in the non-conductive state: the methionine in the SF which seals the selectivity filter (dubbed the “upper gate”) and the isoleucine in S6 which creates a hydrophobic seal at the common gate (dubbed the “lower gate”). In the RTx/DkTx-bound TRPV1, both of these restrictions are removed: the common gate opens via a bend in the S6 and the SF via a tilt of the pore helix. While it is possible that the SF opening is due to DkTx binding or simply reflects a conformational change that is coupled to the opening of the “lower gate”, the structural data points to the existence of the SF gate. We agree with the reviewer that the functional evidence for the SF activation gate is lacking. From a structural standpoint, we reason that TRPV2, like TRPV1, possesses a SF activation gate: the architecture of TRPV2 is very similar to TRPV1 and shows two restrictions in the closed state and binding of RTx induces an open conformation of the SF gate (that the pore helices tilt significantly and move away from the ion permeation pathway).

However, we think that other TRPV channels (TRPV3-6) only possess one restriction in their pores: the common gate. Structural studies of these channels have not observed significant movement around the PH or the SF during gating (e.g. TRPV3), which is consistent with the absence of a SF gate. We have now included a paragraph (subsection “RTx induces a break in symmetry in TRPV2_RTx-ND”, last paragraph) to address this point.

7) In several points throughout the manuscript (subsection “RTx induces a break in symmetry in TRPV2_RTx-ND”, last paragraph), the authors mention that their previous work demonstrated that a network of hydrogen bonds between the pore helix, the S5 and the S6 is essential for activation. However, the results in the previous publication do not provide conclusive evidence for the involvement of hydrogen bonds, but instead show that alanine mutations in that region affect cation permeation through an unknown mechanism that could involve a network of hydrogen bonds. This should be clearly stated in the manuscript.

We believe that our reference to the hydrogen bond network between the threonine in the pore helix and tyrosine residues in S5 and S6 is chemically sound. The donor/acceptor are within 3Å of each other in our 3.1 Å crystal structure, and therefore classified as a hydrogen bond of moderate strength (like most hydrogen bonds in proteins). While it is true that hydrogens are not resolved at ~3 Å, this assumption is widely used in structural biology (i.e. the hydrogen bonding in α-helices).

No other type of major interaction can exist between threonine and tyrosine side chains that are within 3Å beside hydrogen bonding. It is true that we performed alanine mutagenesis studies our previous studies (Zubcevic et al., 2018) instead of more subtle mutations such as threonine to valine (isosteric mutation) or tyrosine to phenylalanine (hydroxyl group removal). However, if we apply the same strict standards, introducing above-mentioned mutations to probe H-bonding would also be insufficient evidence for the involvement of H-bonding, and one would at least need to conduct an unnatural amino acid mutagenesis study of TRPV2 to properly probe the H-bonds. In response to the reviewer’s request, throughout the text, we have clarified this point by stating that the interaction between the Thr and Tyr residues (which includes H-bonds), is important for activation.

8) In the third paragraph of the Discussion, the authors mention that their previous work has shown that C2 symmetric states are critical for permeation of large organic cations. We think it is more accurate to state that their data suggests that C2 symmetry could be involved in channel activation by RTx, but it does not provide conclusive evidence that C2 symmetries are directly connected to channel activation or to the mechanism of cation permeation in an open channel. No experiments were done in that study to directly probe the functional consequences of the proposed asymmetries in the tetrameric channel.

This is a great point to bring up, as it allows us to explain our reasoning in more detail. In our NSMB paper we performed a series of experiments to address the role of two-fold symmetry. Our structures showed that RTx-bound TRPV2 adopts a two-fold symmetric state, where the SF takes on two distinct conformations: a narrow conformation, which coordinates Ca²⁺ ions, and a wide conformation, which is wide enough to permeate YO-PRO-1. In this structure, an interaction exits between the pore helix residue Thr602 and S5 helix residue Tyr542 in the subunits with the narrow SF conformation, but it is absent in the subunits in the wide conformation. This interaction is also absent in the ligand-free cryo-EM structure of TRPV2 in C4 symmetry, which adopts a closed conformation of the SF gate. We showed that eliminating the Thr602-Tyr542 interaction, which exists only in the 2-fold symmetric structure, results in channels that have normal Na⁺ currents but are unable to conduct YO-PRO-1. In addition, mutations Thr602A and Tyr542A both decrease the NMDG⁺/Na⁺ permeability ratio (these experiments were done in inside-out patches to avoid ion accumulation artifacts). This was also shown to be true for the corresponding residues in TRPV1. Interestingly, the Thr602-Tyr542 interaction is absent from all four subunits of the RTx/DkTx-bound TRPV1, consistent with fact that the pore of RTX/DkTx-bound TRPV1, due to its narrow profile, is incompatible with permeation of large organic cations.

These experiments showed that:

1) Disruption of the interactions between the SF and the S5 preferentially reduce the permeation of large organic cations;

2) This interaction has only been observed in the channel with C2-symmetry;

3) Intriguingly, this interaction exists in the contracted subunit, not in the widened subunit, further supporting the involvement of reduced symmetry in the ability of the channel to permeate large organic cations.

Therefore, based on our data we believe that reduced symmetry is involved in large organic cation permeation. In our manuscript, we have now changed the text to state that “our previous studies suggest that C2 symmetric states play an important role for permeation of large organic cations”.

Two-fold symmetry has now been observed in a number of TRP channels in a variety of different experimental conditions, but almost always in the presence of agonists: TRPV3 (Zubcevic et al., 2018), TRPV2 (Zubcevic et al., 2018; Pumroy et al., 2019) and TRPM2 (Yin et al., 2019). We believe that further exploration of these experimentally observed symmetry transitions may help us refine our understanding of TRP(V) channel activation and physiology.

References:

Matthies D, Bae C, Toombes GES, Fox T, Bartesaghi A, Subramaniam S, Swartz KJ. Single-particle cryo-EM structure of a voltage-activated potassium channel in lipid nanodiscs. eLife. 2018. doi: 10.7554/eLife.37558.

Pumroy RA, Samanta A, Liu Y, Hughes TET, Zhao S, Yudi Y, Huynh KW, Zhou ZH, Rohacs T, Han S, Moiseenkova-Bell VY. Molecular mechanism of TRPV2 channel modulation by cannabidiol. bioRxiv, May 24, 2019. Doi: 10.1101/521880

Yin Y, Wu M, Hsu AL, Borschel WF, Borgnia MJ, Lander GC, Lee SY. Visualizing structural transitions of ligand-dependent gating of the TRPM2 channel. bioRxiv, Jan 9, 2019. Doi: 10.1101/516468.

Author details

1. Lejla Zubcevic

   Department of Biochemistry, Duke University School of Medicine, Durham, United States

   Contribution

   Conceptualization, Data curation, Validation, Writing—original draft, Writing—review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-1884-9235

2. Allen L Hsu

   Genome Integrity and Structural Biology Laboratory, National Institute of Environmental Health Sciences, National Institutes of Health, Department of Health and Human Services, Research Triangle Park, United States

   Contribution

   Data curation, Validation, Writing—review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-2065-3802

3. Mario J Borgnia

   1. Department of Biochemistry, Duke University School of Medicine, Durham, United States
   2. Genome Integrity and Structural Biology Laboratory, National Institute of Environmental Health Sciences, National Institutes of Health, Department of Health and Human Services, Research Triangle Park, United States

   Contribution

   Supervision, Funding acquisition, Validation, Writing—review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-9159-1413

4. Seok-Yong Lee

   Department of Biochemistry, Duke University School of Medicine, Durham, United States

   Contribution

   Conceptualization, Supervision, Funding acquisition, Validation, Writing—original draft, Writing—review and editing

   For correspondence

   seok-yong.lee@duke.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-0662-9921

• 2,365

  Page views

• 496

  Downloads

• 37

  Citations

Article citation count generated by polling the highest count across the following sources: Scopus, Crossref, PubMed Central.