Conformational ensemble of the human TRPV3 ion channel

Transient receptor potential vanilloid channel 3 (TRPV3), a member of the thermosensitive TRP (thermoTRPV) channels, is activated by warm temperatures and serves as a key regulator of normal skin physiology through the release of pro-inflammatory messengers. Mutations in trpv3 have been identified as the cause of the congenital skin disorder, Olmsted syndrome. Unlike other members of the thermoTRPV channel family, TRPV3 sensitizes upon repeated stimulation, yet a lack of structural information about the channel precludes a molecular-level understanding of TRPV3 sensitization and gating. Here, we present the cryo-electron microscopy structures of apo and sensitized human TRPV3, as well as several structures of TRPV3 in the presence of the common thermoTRPV agonist 2-aminoethoxydiphenyl borate (2-APB). Our results show α-to-π-helix transitions in the S6 during sensitization, and suggest a critical role for the S4-S5 linker π-helix during ligand-dependent gating.

TRPV3, a thermo--sensitive TRP channel, serves as a modulator for biological events in epidermal and hair follicle keratinocytes. The authors found a mutation T69A that does not affect the function of TRPV3 but significantly stabilizes the protein for structural investigations. This tour de force allows the authors to capture the TRPV3 in the distinct states at the atomic level. In general, this manuscript presents an interesting and exciting story that illuminates the molecular mechanism of TRPV3 opening and regulation. This referee also appreciates that the authors posted this manuscript to bioRxiv (doi: https://doi.org/10.1101/395616) before it is accepted for publication. This finding seems to warrant publication in Nature Communications in this referee's opinion, since the work will serve to present an intact story and illuminate the regulatory mechanism of this important channel. This work should be accepted, after the authors address the following concerns.
Major Concerns: 1, the authors points out that N671, the only polar residue in S6, plays a key role in regulating channel activity. It is necessary to introduce a mutation on N671 to perform functional characterization of this mutant. This will convince the readers of the physiological importance of this residue.

R)
We acknowledge that a functional characterization of the N671 residue would increase the importance of this residue. However, the purpose of pointing out the change in position of N671 was merely to point out that the α--to--π transition in the S6 helix changes the electrostatic properties of the pore. Other groups have identified a similar role for the corresponding residue in TRPV1 (Kasimova et al, J Phys Chem Lett, 2018;Kasimova et al, Biorxiv, DOI https://doi.org/10.1101/310151, 2018, and we have now included references to these papers in the manuscript. 2, since this manuscript reported an opening mechanism of TRPV channels, the authors could consider comparing the opening of TRPV3 with other types of TRP channels ( This manuscript has been previously reviewed at another journal that is not operating a transparent peer review scheme. This document only contains reviewer comments and rebuttal letters for versions considered at Nature Communications. Mentions of prior referee reports have been redacted. and the presence of C2 symmetry has enabled us to dissect how distinct domains (i.e. ARD and TMD) might be coupled. 3, update the discussion to include a brief comparison with recently published details of TRPV3 (Singh et al NSMB). There are similarities and differences.

R)
We have now included a section in the discussion as well as a supplementary figure  (Supplementary Fig. 9) where we compare the mouse TRPV3 to human TRPV3 structures. Minor Points: 4, Line 72, the small--molecule inhibitor should be indicated.

R)
We have now updated the methods to include temperature conditions for protein preparation, grid freezing and electrophysiology experiments. We agree that temperature and camphor controls would be a nice addition. However, since our study focuses on ligand--dependent activation, we feel that temperature experiments are beyond its scope. In the study, we have chosen 2--APB as a ligand as it is the most potent and well--studied agonist of TRPV3. Camphor control experiment would be nice, but it is likely that camphor binds to a site (or sites) in TRPV3 different from 2--APB. In addition, the recently published NSMB paper on TRPV3 (Singh et al, NSMB, 2018) does not report temperature or camphor responses for the mutant channels presented in the paper.
APB stimulated states. The data appear of high quality, and provide novel insights into TRPV3 channel functioning. Specific points: 1) To obtain structures of the sensitized TRPV3, the authors apply a protocol with repeated incubations with 2--APB followed by washing steps to TRPV3 attached to anti--FLAG resin. Whereas the authors show that such a protocol can sensitize the channel when present in a cell membrane, this does not necessarily mean that similar sensitization occurs in a cell--free environment in a Biorad column. Is it known whether TRPV3 sensitization actually occurs in a cell--free environment (e.g. in lipid bilayers)? In the absence of such evidence, it seems premature to denote this structure as "sensitized' TRPV3, and the authors should at least discuss this in more detail.

R)
We accept the reviewer's criticism. In our revised manuscript we refer to this structure as "putative sensitized". Nevertheless, application of the 2--APB sensitization protocol resulted in a structure that is distinct from both the human apo TRPV3 and the open mouse TRPV3 structures, in that it possesses a π--helical turn in S6 but its gate is closed. Therefore, our putative sensitized structure is clearly an intermediate state that is on the path toward the open state, consistent with the sensitized states. However, we cannot say that the putative sensitized structure presented here is the fully sensitized state of TRPV3. Studies suggest (Liu et al, JGP, 2011) that TRPV3 might have multiple such sensitized states, with decreasing energy requirements for opening. In addition, structural elements other than S6 might be undergoing conformational changes during sensitization (Liu et al, PNAS, 2017). We have now included a section in the discussion where this is addressed in more detail.
2) The authors present several structures of TRPV3 in the presence of 2--APB, which no longer show fourfold symmetry. Have the authors considered the possibility that these could arise from channels in which the 2--APB ligand binding sites are only partially occupied? Channels with 1--3 2--APB molecules bound would automatically lose fourfold symmetry. Or do the authors have evidence that all four ligand binding sites are occupied in the structures? R) This is an interesting point. Unfortunately, we could not unambiguously identify 2--APB in our cryo--EM maps, so we cannot say with certainty whether the 2--APB binding sites are fully occupied, or how many 2--APB molecules might be bound per tetramer. Our experiments were performed in the presence of high 2--APB (1mM), which would make partial occupancy less likely but not impossible. Interestingly, our previous work on TRPV2 (Zubcevic et al, NSMB, 2018) showed that C2 symmetry can be achieved even when the binding sites are fully occupied. In this case, the binding mode of the ligand differs in diagonally opposing subunits, resulting in a two--fold symmetric channel. We have now included a section in the discussion that addresses this question.
3) Singh et al. recently published TRPV3 structures in the apo and agonist--bound states (NSMB, 2018). Please discuss/compare. figure  (Supplementary Fig. 9) where we compare the mouse TRPV3 to human TRPV3 structures. 4) In the introduction: Line 54--56: The statement that TRPV1--TRPV4 are involved in thermosensing is not supported by the literature. Whereas the role of TRPV1 as one of the sensors of noxious heat is well established, elimination of TRPV2, TRPV3 or TRPV4 in mice does not have any effect on thermosensation. Please reword, and/or cite more recent work or reviews on this topic R) The reviewer brings up a valid point, and it seems that we have unintentionally miscommunicated the statement on thermosensation. Our comment related to the ability of the channels to open in response to heat, which is well documented for TRPV1--TRPV4, and not their physiological role in termosensation. We have now reworded this sentence. Line 59: reference 12 seems out of place here, and references 1 and 11 are a bit outdated and do not really provide an up to date view on the role of TRPV channels.

R) We accept this and have updated the references to reflect the pharmacological and biophysical diversity in the thermoTRPV channels.
Reviewer #3 (Remarks to the Author): TRPV3 plays important roles in skin, hair and itch physiology. Gain of function, eg due to mutations in the S4--S5 linker, results skin disease called Olmsted. Loss of function, eg via inhibitors applications, may be associated with analgesic effects on inflammation and pain. TRPV3 can be activated by repeated application of heat or agonists such as 2--APB, and is regulated by lipids such as PIP2. This study reports high resolution Cryo--EM structures of human full--length TRPV3 bearing mutation T96A. This manuscript from Dr Lee's group presents high quality structural data that show apo and sensitized (or activated) states of TRPV3 channels, as well as three 2--APB bound 'intermediate' states. The finding of decreased symmetry to C2 is novel and intriguing. The authors found that the T96A mutant is optimal as compared with WT protein for structural purposes and that it has similar functional characteristics as WT channel. Major comments. 1. I agree that Fig. 1c and d and Suppl Fig. 4 show that the T96A mutation does not affect the basic channel property. However, I do not agree with the use of 'hysteresis' in line 98. Hysteresis is characterized by the presence of a so--called hysteresis loop/curve, ie, for a given x value (eg temperature, agonist concentration, or repeated # of a given agonist centration) there are two different y values, obtained from measurements carried out with increasing and decreasing the x value, respectively. Data in Fig 1c and d rather show 'use--dependence', a term previously used (PMID: 28154143). To show that mutation T96A does not affect the hysteresis, the authors should use both increasing and decreasing temperature to obtain two temperature--dependence curves that form a hysteresis type of loop (ie non--superposing curves), as reported in 2002 for TRPV3 (PMID: 12077604).

R)
We agree with the reviewer, and have therefore exchanged the term "hysteresis" for "use-dependence". Because this study focuses on ligand--dependent activation of TRPV3, we feel that temperature experiments, while a nice addition, are beyond its scope and would not change the conclusion of our paper. 2. Line 341, the sensitization protocol used for preparing sensitized proteins for structural purposes. This protocol is very different from the one used in electrophysiology recordings. One key difference between the two is with or without voltage clamp, which would lead to different channel activities and possibly also agonist binding affinities. I'm not sure whether this protocol would effectively sensitize TRPV3 channels. To prove, the authors have to mimic this protocol in electrophysiology recordings to determine whether it indeed results in similar channel sensitization. To do this, a cell under the whole--cell mode should first be recorded for apo currents, and then de--clamped to undergo repeated agonist loads/washes (as described around line 341). The cell will then be recorded for TRPV3 currents to determine whether there is a similar sensitization/activation as in Fig. 1c and d.

R)
We accept the reviewer's criticism. In our revised manuscript we refer to this structure as "putative sensitized". Nevertheless, application of the 2--APB sensitization protocol resulted in a structure that is distinct from both the human apo TRPV3 and the open mouse TRPV3 structures, in that it possesses a π--helical turn in S6 but its gate is closed. Therefore, our putative sensitized structure is clearly an intermediate state that is on the path toward the open state, consistent with the sensitized states. However, we cannot say that the putative sensitized structure presented here is the fully sensitized state of TRPV3. Studies suggest (Liu et al, JGP, 2011) that TRPV3 might have multiple such sensitized states, with decreasing energy requirements for opening. In addition, structural elements other than S6 might be undergoing conformational changes during sensitization (Liu et al, PNAS, 2017). We have now included a section in the discussion where this is addressed in more detail. 3. Line 272, "We propose that the transition from α-- to π--helix is an essential component of TRPV3 hysteresis." Need justifications for this statement. This is because the opposite π--to α--helix transition was reported for other TRPs such as V1, P2 and P3, but there is no evidence of hysteresis associated with these channels. It would be of help if the authors justify the statement through comparing with this opposite transition associated with (equivalent) channel activation. This point links to another statement (line 165), which may have to be changed: TRPV1 is also a thermoTRP but is has an apposite π--to α--helix transition.

R)
We agree with the reviewer that more discussion concerning the α--to--π transition and sensitization should be included. Unfortunately, we are not familiar with the studies that show that TRPV1 undergoes a π--to--α transition during opening. Work from Julius and Chen labs has shown that TRPV1 has a π--helix in S6 in both apo/closed and toxin bound/open states (Liao et al, Nature, 2013;Cao et al, Nature, 2013;Gao et al, Nature, 2016), suggesting that no secondary structure transitions occur in S6 of the TRPV1 channel during gating. This correlates well with the fact that, in contrast to TRPV2 and TRPV3, TRPV1 does not exhibit heat or capsaicin induced use--dependence (Liu et al, Biophys J., 2016). We have now included a section in the discussion to address this.
Indeed, a π--to--α transition has been associated with opening of the TRPP3 (PKD2L1) channel (Su et al, Nat Comms, 2018). In this study, the authors compare the closed structure of TRPP2 (PKD2), which contains a π--helix, with the open structure of TRPP3, which has an α--helical S6 and deduce that opening results from π--to--α secondary structure transitions in S6. Such π--to--α transitions have not been observed in the related TRPML channels (Zhou et al, NSMB, 2017;Schmiege et al, Nature, 2017), where the channels maintain a π--helix in both closed and open states. However, TRPP and TRPML channels differ substantially from the TRPV subfamily in both structure and function, and we feel that a including a direct comparison here may not give much structural insights into TRPV channel gating.
Minor comments.
1. As mouse TRPV3 structure was just published in Nat Struct Mol Biol, I invite the authors to cite the paper and provide a comparison discussion. figure  (Supplementary Fig. 9) where we compare the mouse TRPV3 to human TRPV3 structures.

R) We have now included a section in the discussion as well as a supplementary
2. Line 147, "the sensitized structure is bent at a 9° angle towards the inner pore when compared to the S6 in the apo structure". To me, this would mean that the pore will become smaller but in the reality it's opposite ( Fig. 4a right panel). A clearer sentence would help.

R)
We agree with the reviewer that the sentence is confusing. We have now rewritten it, and it reads: "In addition, the C--terminal half of the S6 helix in this structure is bent at a 9° angle from the inner pore when compared to the S6 in the apo structure (Fig. 4c)."