Mechanosensation allows organisms to detect and respond to external and internal mechanical forces (Haswell et al., 2011). Mechanical stimuli such as touch, gravity, and osmotic pressure are sensed by mechanically activated (MA) ion channels (Haswell et al., 2011; Ranade et al., 2015). Previously reported as putative ion channels sensitive to osmolality (Hou et al., 2014; Yuan et al., 2014; Zhao et al., 2016), a companion paper characterizes OSCA/TMEM63 family members as a novel family of MA ion channels conserved across eukaryotes (Murthy et al., 2018). Plants with mutant OSCA1 have impaired osmotic stress signaling (Yuan et al., 2014), implying that OSCAs function as sensors of osmotic-stress induced mechanical force in vivo. TMEM63 proteins are animal orthologues of OSCA that produce MA currents when expressed in naïve cells (Murthy et al., 2018), but their physiological roles remain undetermined. OSCA/TMEM63 are not homologous to other known MA ion channels, and no previous study has addressed their molecular structure. To facilitate a mechanistic understanding of how OSCAs sense force in plants, and to shed light on a newly identified family of eukaryotic MA ion channels, we conducted cryo-EM studies of OSCA1.2.

We purified Arabidopsis thaliana OSCA1.2 expressed in HEK293F cells and determined cryo-EM reconstructions in lipidic nanodiscs and Lauryl Maltose Neopentyl Glycol (LMNG) detergent micelle with cholesteryl hemisuccinate (CHS) at 3.1 and 3.5 Å resolution, respectively (Figure 1A–D, Figure 1—figure supplement 1, Figure 1—figure supplement 2). The resulting density maps allowed building of 82% of the full-length OSCA1.2 sequence (Figure 1—figure supplement 3, Figure 1—source data 1). Structures of OSCA1.2 in nanodisc and LMNG were nearly identical (RMSD = 0.81 Å, Figure 1—figure supplement 4A–C). We refer to the nanodisc structure throughout unless otherwise noted because of superior resolution and map quality. Importantly, the purified channel reconstituted in liposomes retains MA ion channel activity (Murthy et al., 2018). Thus, our structures and interpretations correspond to that of a functional OSCA1.2 protein. Additionally, we performed MD simulations at both coarse-grained (CG) and atomistic (AT) levels to model the local membrane interactions of OSCA1.2 (Figure 1—figure supplements 5–6, Video 1).

Video 1

Download asset

Figure 1 with 6 supplements see all

Download asset Open asset

Figure 1—source data 1

Data collection, processing, model refinement, and validation.

https://doi.org/10.7554/eLife.41845.009

Download elife-41845-fig1-data1-v2.docx

The structure of OSCA1.2 reveals a trapezoid-shaped homodimer with a two-fold symmetry axis perpendicular to the membrane (Figure 1A–H). The dimer is 139 Å across at its widest dimension. The bulk of the protein lies within the membrane; the transmembrane (TM) domain (TMD) of each monomer contains eleven TM helices. Linkers between TM helices and a C-terminal region come together to form an intracellular domain (ICD) that extends ~31 Å below the membrane (Figure 1A–H). The ICD contributes the sole dimeric interface in OSCA1.2, with a buried surface area of 1,331 Ų (Figure 1—figure supplement 4D) (Krissinel and Henrick, 2007), only 4% of the total surface area of the dimer. Within the membrane, inter-subunit contacts are nonexistent. Instead, a cleft with minimum width of approximately 8 Å separates one subunit’s TMD from the other (Figure 1I). Interestingly, AT-MD simulations of OSCA1.2 in a phospholipid (1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine, POPC) bilayer, consistently place lipid molecules inside the cleft (Figure 1I, Figure 1—figure supplement 5A, Video 2), indicating a role for lipids in stabilizing the dimeric assembly.

Video 2

Download asset

Figure 2A,B show the topology and domain arrangement of an OSCA1.2 subunit. The N-terminus of OSCA1.2 faces the extracellular environment while the C-terminus is intracellular. The majority of the ICD is comprised of the second intracellular linker (IL2), which is over 150 residues and contains a 4-stranded antiparallel ß-sheet (IL2ß1-IL2ß4) with well-conserved sequences across the OSCA family (Figure 2—figure supplements 1, 2 and 3), and four helices (IL2H1-IL2H4). Three additional helices, contributed by intracellular linkers 1 (IL1H2) and 4 (IL4H) and the C-terminus (CTH), constitute the rest of the intracellular domain. Other than short linkers, we did not observe structured extracellular domains, presumably due to flexibility.

Figure 2 with 4 supplements see all

Download asset Open asset

As noted previously (Hou et al., 2014), a C-terminal region of OSCA1.2 (corresponding to TM4-TM9 in our structure) has loose homology to TMEM16 proteins, which are a family of Ca²⁺-activated ion channels (Caputo et al., 2008; Chen et al., 2010; Schroeder et al., 2008; Yang et al., 2008) and lipid scramblases (Malvezzi et al., 2013; Suzuki et al., 2013; Suzuki et al., 2010) which, to our knowledge, have no mechanically gated activity describe so far. Strikingly, our structure reveals that the structural homology extends beyond the C-terminal region; ten of the eleven TM helices of OSCA1.2 closely follow the topology and organization of mouse TMEM16A (mTMEM16A) Ca²⁺-activated chloride channel (Dang et al., 2017; Paulino et al., 2017a; Paulino et al., 2017b) and Nectria haematococca TMEM16 (nhTMEM16), which is a Ca²⁺-dependent lipid scramblase (Brunner et al., 2014) and nonselective ion channel (Lee et al., 2016) (Figure 2C; Figure 2D, Figure 2—figure supplement 4A; Figure 2—figure supplement 4B). Relative to these TMEM16 structures, OSCA1.2 has an additional N-terminal TM helix positioned at the periphery of the TMD (Figure 2A). To facilitate structural comparison between OSCA and TMEM16 proteins, we refer to this N-terminal TM helix as ‘TM0’, and the remaining TM helices ‘TM1-TM10’. Though TM1-TM10 of OSCA1.2 and TMEM16A align well at the level of a single subunit, the relative orientation of one subunit to the other is different (Figure 2E). As a result of this distinct dimeric packing, the TMD of the OSCA1.2 dimer is more than 20 Å wider than mTMEM16A or nhTMEM16. Additional features distinguish the overall structure of OSCA1.2 from TMEM16 proteins. First, the intracellular domain of OSCA1.2 is mostly comprised of IL2, which connects TM2 and TM3, while the intracellular domains of mTMEM16a and nhTMEM16 are formed primarily by the N- and C- termini. Second, the regulatory Ca²⁺ binding site composed of acidic and polar residues conserved across the TMEM16 family (Brunner et al., 2014) does not have a chemical environment conducive for Ca²⁺ binding in OSCA1.2 (Figure 2—figure supplement 4C), consistent with distinct modes of regulation (mechanical activation vs. Ca²⁺ dependence).

In the mTMEM16A dimer, there are two pores; one within each subunit (Dang et al., 2017; Paulino et al., 2017a). Topological similarities with mTMEM16A (Figure 2) suggest that OSCA1.2 also has two pores and could conduct ions through a structurally analogous pathway lined by TM3-TM7. The existence of two pores per OSCA1.2 dimer is consistent with the presence of a single subconductance state in stretch-activated single-channel currents (Murthy et al., 2018). To visualize the dimensions of the putative ion permeation pathway, we used the HOLE program (Smart et al., 1996) (Figure 3A–C, Figure 3—figure supplement 1, Video 3). Towards the extracellular side, this putative pore has an opening greater than 12 Å wide, which narrows into a ‘neck’ approximately 15 Å down the conduction pathway (Figure 3A–C). The neck extends for over 10 Å and reaches a minimum van der Waals radius of 0.5 Å, closing the channel. Mostly hydrophobic residues (I393, V396, A400, L434, L438, F471, V476, and F515) line the pore’s constriction, as well as several charged and polar residues (Q397, S480, Y468, K512, and D523). Interestingly, a pair of π-helical turns (in TM5 and TM6a) are nearby the pore’s neck region (Figure 3C). Because π-helices are energetically unstable, we speculate that π-to-α transitions at these positions could be associated with gating and, potentially, channel opening. In support of this idea, there are π-helical turns at highly similar locations of TM6 in mTMEM16A (Paulino et al., 2017a) and nhTMEM16 (Brunner et al., 2014), and a α-to-π transition in TM6 underlies Ca²⁺-dependent activation in mTMEM16A (Paulino et al., 2017a) (Figure 2—figure supplement 4D). Similarly, the pore-lining helices of the epithelial calcium channel TRPV6 (McGoldrick et al., 2018) undergo π-to-α helix transitions during gating. Directly below the neck, the pore widens, and the side chains of E531, R572, and T568 create a hydrophilic environment that could stabilize hydrated ions (Figure 3C). Further below, the pore widens into membrane-exposed vestibule and is surrounded by IL2 and IL4 before exiting into the cytosol.

Video 3

Download asset

Figure 3 with 1 supplement see all

Download asset Open asset

If TM3-TM7 indeed form the permeation pathway of OSCA1.2, mutation of residues that line the pathway should alter the channel’s permeation or conductance properties. We chose to examine E531, which is the only pore-facing acidic residue conserved in this region across the OSCA/TMEM63 families and thus could contribute to conductance (Figure 3C, Figure 2—figure supplements 1, 2 and 3, Figure 2—figure supplement 4C). We recorded and characterized stretch-activated currents in the cell-attached patch clamp mode from cells expressing wildtype or mutant (E531A) channels (Figure 3D). E531A had maximal current responses comparable to wildtype channels (Figure 3D,E), and modestly faster inactivation kinetics (Figure 3E). Strikingly, the mutation decreased the stretch-activated single-channel conductance by 1.6-fold, demonstrating that E531 contributes to the channel’s ion permeation pathway (Figure 3F,G), and could play a role in binding or sequestering cations. Alternatively, the mutation could alter pore structure and dynamics owing to the interactions of the E531 side chain with R572 and Y605 (Figure 3—figure supplement 1B). Nonetheless, combining these mutagenesis data with structural homology to mTMEM16A (Paulino et al., 2017a; Paulino et al., 2017b), we conclude that TM3-TM7 most likely form the pore. This conclusion is further supported by MD simulations showing hydration of the pore pathway being consistent with the pore radius profile from HOLE (Figure 1—figure supplement 5B, Figure 3—figure supplement 1C, Video 4, Video 5).

Video 4

Download asset

Video 5

Download asset

Our structural study of OSCA1.2 elucidates a novel architecture for MA ion channel and reveals unanticipated similarities between OSCA1.2 and TMEM16 proteins, including pore structure. Which structural elements in OSCA1.2 could be involved in membrane tension-sensing and gating? Two notable features present in OSCA1.2, but not TMEM16s, are a hook-shaped loop that enters the membrane that intervenes IL2H2 and IL2H3 (Figure 2A, Figure 4A), and a horizontal amphipathic helix in IL1 (IL1H1) that slightly deforms the membrane lower leaflet in MD simulations (Figure 4B, Video 1). Their association with the cytoplasmic bilayer leaflet suggests these two features could be sensitive to membrane tension, akin to the N-terminal amphipathic helix in MscL (Bavi et al., 2016), and their movement could be coupled to channel conformation through their interactions with the TMD (Figure 4A,B). Another notable aspect of the OSCA1.2 structure, similar to TMEM16s, is that the lower region of the pore is exposed to the membrane and thus could permit lipid entry (Figure 3B, Figure 3—figure supplement 1A,C). Indeed, in MD simulations, lipids occupy and occlude the cytoplasmic half of the pore, with their phosphate head groups interacting with four lysine residues on TM4 and TM6b (Figure 4C, Figure 1—figure supplement 5C, Video 6, Video 7). It is possible that membrane tension promotes gating by affecting the lipid occupancy in this region, analogous to proposed mechanistic models for TRAAK (Brohawn et al., 2014) and MscS (Pliotas et al., 2015). Finally, a general principle of MA ion channels is that their cross-sectional area should increase in response to membrane tension (Guo and MacKinnon, 2017; Haswell et al., 2011; Phillips et al., 2009). Qualitatively, the absence of a dimeric protein-protein interface between membrane domains (Figure 1I) could allow OSCA1.2 to expand its cross-sectional area significantly in response to physiological lateral tension without the energetic penalty of breaking inter-subunit interactions within the membrane. A comparison of nanodisc-embedded and LMNG-solubilized structures hints at the positional flexibility of the subunits relative to each other while in the closed state (Figure 1—figure supplement 4B,C, Video 8). Larger movements will be expected under membrane tension and channel opening. The features of OSCA1.2 outlined above might act in concert to modulate channel gating in response to mechanical stimuli (Figure 4D).

Video 6

Download asset

Video 7

Download asset

Video 8

Download asset

Figure 4

Download asset Open asset

Intriguingly, transmembrane-channel-like (TMC) proteins, which are pore-forming subunits of the hair cell mechanoelectrical transduction (MET) apparatus (Fettiplace, 2016; Kawashima et al., 2011; Pan et al., 2013), are distantly related to the TMEM16 family and proposed to have the same topology (Ballesteros et al., 2018; Hahn et al., 2009; Medrano-Soto et al., 2018; Pan et al., 2018). Although the molecular identity of the MET ion channel remains incompletely understood (Fettiplace, 2016), the shared TM topology between more distantly related TMEM16s and OSCA1.2 suggests a similar fold for TMCs. It is therefore likely that OSCA/TMEM63, TMEM16, and TMC all utilize a similar architectural scaffold to carry out various functions at the membrane. While the details underlying such functional diversity await discovery, our study establishes a framework to understand how OSCAs exploited this common fold, and evolved specific features, to serve the role of MA ion channels.

We note that cryo-EM structures of two other members of the OSCA/TMEM63 family, OSCA1.1 and OSCA3.1 (Zhang et al., 2018), were recently published after our initial BioRxiv deposition (Jojoa-Cruz et al., 2018).

Cryo-EM maps of OSCA1.2 in nanodiscs and LMNG have been deposited to the Electron Microscopy Data Bank under accession codes 9112 and 9113. Atomic coordinates of OSCA1.2 in nanodiscs and LMNG have been deposited to the PDB under IDs 6MGV and 6MGW. Due to their large size (300Gb+), the raw data files are available upon request to the corresponding author(s).

1. 1. Jojoa-Cruz S
   2. Saotome K
   3. Patapoutian A
   4. Ward AB

   (2018) Electron Microscopy Data Bank

   ID 9112. Cryo-EM map of mechanically activated ion channel OSCA1.2 in nanodisc.

   https://www.ebi.ac.uk/pdbe/entry/emdb/EMD-9112
2. 1. Jojoa-Cruz S
   2. Saotome K
   3. Patapoutian A
   4. Ward AB

   (2018) Electron Microscopy Data Bank

   ID 9113. Cryo-EM map of mechanically activated ion channel OSCA1.2 in LMNG.

   https://www.ebi.ac.uk/pdbe/entry/emdb/EMD-9113
3. 1. Jojoa-Cruz S
   2. Saotome K
   3. Patapoutian A
   4. Ward AB

   (2018) RCSB Protein Data Bank

   ID 6MGV. Structure of mechanically activated ion channel OSCA1.2 in nanodisc.

   https://www.rcsb.org/structure/6MGV
4. 1. Jojoa-Cruz S
   2. Saotome K
   3. Patapoutian A
   4. Ward AB
   5. Jojoa-Cruz S

   (2018) RCSB Protein Data Bank

   ID 6MGW. Structure of mechanically activated ion channel OSCA1.2 in LMNG.

   https://www.rcsb.org/structure/6MGW

1. 1. Abraham MJ
   2. Murtola T
   3. Schulz R
   4. Páll S
   5. Smith JC
   6. Hess B
   7. Lindahl E

   (2015) GROMACS: High performance molecular simulations through multi-level parallelism from laptops to supercomputers

   SoftwareX 1-2:19–25.

   https://doi.org/10.1016/j.softx.2015.06.001

   • Google Scholar
2. 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
3. 1. Afonine PV
   2. Klaholz BP
   3. Moriarty NW
   4. Poon BK
   5. Sobolev OV
   6. Terwilliger TC
   7. Adams PD
   8. Urzhumtsev A

   (2018) New tools for the analysis and validation of cryo-EM maps and atomic models

   Acta Crystallographica Section D Structural Biology 74:814–840.

   https://doi.org/10.1107/S2059798318009324

   • Google Scholar
4. 1. Ballesteros A
   2. Fenollar-Ferrer C
   3. Swartz KJ

   (2018) Structural relationship between the putative hair cell mechanotransduction channel TMC1 and TMEM16 proteins

   eLife 7:e38433.

   https://doi.org/10.7554/eLife.38433

   • PubMed
   • Google Scholar
5. 1. Barad BA
   2. Echols N
   3. Wang RY
   4. Cheng Y
   5. DiMaio F
   6. Adams PD
   7. Fraser JS

   (2015) EMRinger: side chain-directed model and map validation for 3D cryo-electron microscopy

   Nature Methods 12:943–946.

   https://doi.org/10.1038/nmeth.3541

   • PubMed
   • Google Scholar
6. 1. Bavi N
   2. Cortes DM
   3. Cox CD
   4. Rohde PR
   5. Liu W
   6. Deitmer JW
   7. Bavi O
   8. Strop P
   9. Hill AP
   10. Rees D
   11. Corry B
   12. Perozo E
   13. Martinac B

   (2016) The role of MscL amphipathic N terminus indicates a blueprint for bilayer-mediated gating of mechanosensitive channels

   Nature Communications 7:11984.

   https://doi.org/10.1038/ncomms11984

   • PubMed
   • Google Scholar
7. 1. Berndsen Z
   2. Bowman C
   3. Jang H
   4. Ward AB

   (2017) EMHP: an accurate automated hole masking algorithm for single-particle cryo-EM image processing

   Bioinformatics 33:3824–3826.

   https://doi.org/10.1093/bioinformatics/btx500

   • PubMed
   • Google Scholar
8. 1. Brohawn SG
   2. Campbell EB
   3. MacKinnon R

   (2014) Physical mechanism for gating and mechanosensitivity of the human TRAAK K+ channel

   Nature 516:126–130.

   https://doi.org/10.1038/nature14013

   • PubMed
   • Google Scholar
9. 1. Brunner JD
   2. Lim NK
   3. Schenck S
   4. Duerst A
   5. Dutzler R

   (2014) X-ray structure of a calcium-activated TMEM16 lipid scramblase

   Nature 516:207–212.

   https://doi.org/10.1038/nature13984

   • PubMed
   • Google Scholar
10. 1. Bussi G
    2. Donadio D
    3. Parrinello M

    (2007) Canonical sampling through velocity rescaling

    The Journal of Chemical Physics 126:014101.

    https://doi.org/10.1063/1.2408420

    • PubMed
    • Google Scholar
11. 1. Caputo A
    2. Caci E
    3. Ferrera L
    4. Pedemonte N
    5. Barsanti C
    6. Sondo E
    7. Pfeffer U
    8. Ravazzolo R
    9. Zegarra-Moran O
    10. Galietta LJ

    (2008) TMEM16A, a membrane protein associated with calcium-dependent chloride channel activity

    Science 322:590–594.

    https://doi.org/10.1126/science.1163518

    • PubMed
    • Google Scholar
12. 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
13. 1. Dang S
    2. Feng S
    3. Tien J
    4. Peters CJ
    5. Bulkley D
    6. Lolicato M
    7. Zhao J
    8. Zuberbühler K
    9. Ye W
    10. Qi L
    11. Chen T
    12. Craik CS
    13. Jan YN
    14. Minor DL
    15. Cheng Y
    16. Jan LY

    (2017) Cryo-EM structures of the TMEM16A calcium-activated chloride channel

    Nature 552:426–429.

    https://doi.org/10.1038/nature25024

    • PubMed
    • Google Scholar
14. 1. de Jong DH
    2. Singh G
    3. Bennett WF
    4. Arnarez C
    5. Wassenaar TA
    6. Schäfer LV
    7. Periole X
    8. Tieleman DP
    9. Marrink SJ

    (2013) Improved Parameters for the Martini Coarse-Grained Protein Force Field

    Journal of Chemical Theory and Computation 9:687–697.

    https://doi.org/10.1021/ct300646g

    • PubMed
    • Google Scholar
15. 1. Dubin AE
    2. Murthy S
    3. Lewis AH
    4. Brosse L
    5. Cahalan SM
    6. Grandl J
    7. Coste B
    8. Patapoutian A

    (2017) ENdogenous piezo1 can confound mechanically activated channel identification and characterization

    Neuron 94:266–270.

    https://doi.org/10.1016/j.neuron.2017.03.039

    • PubMed
    • Google Scholar
16. 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
17. 1. Essmann U
    2. Perera L
    3. Berkowitz ML
    4. Darden T
    5. Lee H
    6. Pedersen LG

    (1995) A smooth particle mesh Ewald method

    The Journal of Chemical Physics 103:8577–8593.

    https://doi.org/10.1063/1.470117

    • Google Scholar
18. 1. Fettiplace R

    (2016) Is TMC1 the hair cell mechanotransducer channel?

    Biophysical Journal 111:3–9.

    https://doi.org/10.1016/j.bpj.2016.05.032

    • PubMed
    • Google Scholar
19. 1. Guo YR
    2. MacKinnon R

    (2017) Structure-based membrane dome mechanism for Piezo mechanosensitivity

    eLife 6:e33660.

    https://doi.org/10.7554/eLife.33660

    • PubMed
    • Google Scholar
20. 1. Hahn Y
    2. Kim DS
    3. Pastan IH
    4. Lee B

    (2009) Anoctamin and transmembrane channel-like proteins are evolutionarily related

    International Journal of Molecular Medicine 24:51–55.

    https://doi.org/10.3892/ijmm_00000205

    • PubMed
    • Google Scholar
21. 1. Haswell ES
    2. Phillips R
    3. Rees DC

    (2011) Mechanosensitive channels: what can they do and how do they do it?

    Structure 19:1356–1369.

    https://doi.org/10.1016/j.str.2011.09.005

    • PubMed
    • Google Scholar
22. 1. Hess B

    (2008) P-LINCS: A parallel linear constraint solver for molecular simulation

    Journal of Chemical Theory and Computation 4:116–122.

    https://doi.org/10.1021/ct700200b

    • PubMed
    • Google Scholar
23. 1. Hou C
    2. Tian W
    3. Kleist T
    4. He K
    5. Garcia V
    6. Bai F
    7. Hao Y
    8. Luan S
    9. Li L

    (2014) DUF221 proteins are a family of osmosensitive calcium-permeable cation channels conserved across eukaryotes

    Cell Research 24:632–635.

    https://doi.org/10.1038/cr.2014.14

    • PubMed
    • Google Scholar
24. 1. Humphrey W
    2. Dalke A
    3. Schulten K

    (1996) VMD: visual molecular dynamics

    Journal of Molecular Graphics 14:.

    https://doi.org/10.1016/0263-7855(96)00018-5

    • PubMed
    • Google Scholar
25. 1. Jojoa-Cruz S
    2. Saotome K
    3. Murthy SE
    4. Tsui CCA
    5. Sansom MS
    6. Patapoutian A
    7. Ward AB

    (2018) Cryo-EM structure of the mechanically activated ion channel OSCA1.2

    eLife 7:e41845.

    https://doi.org/10.7554/eLife.41845

    • PubMed
    • Google Scholar
26. 1. Kawashima Y
    2. Géléoc GS
    3. Kurima K
    4. Labay V
    5. Lelli A
    6. Asai Y
    7. Makishima T
    8. Wu DK
    9. Della Santina CC
    10. Holt JR
    11. Griffith AJ

    (2011) Mechanotransduction in mouse inner ear hair cells requires transmembrane channel-like genes

    Journal of Clinical Investigation 121:4796–4809.

    https://doi.org/10.1172/JCI60405

    • PubMed
    • Google Scholar
27. 1. Kimanius D
    2. Forsberg BO
    3. Scheres SH
    4. Lindahl E

    (2016) Accelerated cryo-EM structure determination with parallelisation using GPUs in RELION-2

    eLife 5:e18722.

    https://doi.org/10.7554/eLife.18722

    • PubMed
    • Google Scholar
28. 1. Kirchhofer A
    2. Helma J
    3. Schmidthals K
    4. Frauer C
    5. Cui S
    6. Karcher A
    7. Pellis M
    8. Muyldermans S
    9. Casas-Delucchi CS
    10. Cardoso MC
    11. Leonhardt H
    12. Hopfner KP
    13. Rothbauer U

    (2010) Modulation of protein properties in living cells using nanobodies

    Nature Structural & Molecular Biology 17:133–138.

    https://doi.org/10.1038/nsmb.1727

    • PubMed
    • Google Scholar
29. 1. Krissinel E
    2. Henrick K

    (2007) Protein interfaces, surfaces and assemblies service PISA at European Bioinformatics Institute

    Journal of Molecular Biology 372:774–797.

    https://doi.org/10.1016/j.jmb.2007.05.022

    • Google Scholar
30. 1. Lee BC
    2. Menon AK
    3. Accardi A

    (2016) The nhTMEM16 Scramblase Is Also a Nonselective Ion Channel

    Biophysical Journal 111:1919–1924.

    https://doi.org/10.1016/j.bpj.2016.09.032

    • PubMed
    • Google Scholar
31. 1. Lukacs V
    2. Mathur J
    3. Mao R
    4. Bayrak-Toydemir P
    5. Procter M
    6. Cahalan SM
    7. Kim HJ
    8. Bandell M
    9. Longo N
    10. Day RW
    11. Stevenson DA
    12. Patapoutian A
    13. Krock BL

    (2015) Impaired PIEZO1 function in patients with a novel autosomal recessive congenital lymphatic dysplasia

    Nature Communications 6:8329.

    https://doi.org/10.1038/ncomms9329

    • PubMed
    • Google Scholar
32. 1. Malvezzi M
    2. Chalat M
    3. Janjusevic R
    4. Picollo A
    5. Terashima H
    6. Menon AK
    7. Accardi A

    (2013) Ca2+-dependent phospholipid scrambling by a reconstituted TMEM16 ion channel

    Nature Communications 4:2367.

    https://doi.org/10.1038/ncomms3367

    • PubMed
    • Google Scholar
33. 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
34. 1. Medrano-Soto A
    2. Moreno-Hagelsieb G
    3. McLaughlin D
    4. Ye ZS
    5. Hendargo KJ
    6. Saier MH

    (2018) Bioinformatic characterization of the Anoctamin Superfamily of Ca2+-activated ion channels and lipid scramblases

    Plos One 13:e0192851.

    https://doi.org/10.1371/journal.pone.0192851

    • PubMed
    • Google Scholar
35. 1. Murthy SE
    2. Dubin AE
    3. Whitwam T
    4. Jojoa Cruz S
    5. Cahalan SM
    6. Mosavi SA
    7. Ward AB
    8. Patapoutian A

    (2018) OSCA/TMEM63 are an evolutionarily conserved family of mechanically activated ion channels

    eLife 7:e41844.

    https://doi.org/10.7554/eLife.41844

    • PubMed
    • Google Scholar
36. 1. Neumann M

    (1986) Dielectric relaxation in water. Computer simulations with the TIP4P potential

    The Journal of Chemical Physics 85:1567–1580.

    https://doi.org/10.1063/1.451198

    • Google Scholar
37. 1. Pan B
    2. Géléoc GS
    3. Asai Y
    4. Horwitz GC
    5. Kurima K
    6. Ishikawa K
    7. Kawashima Y
    8. Griffith AJ
    9. Holt JR

    (2013) TMC1 and TMC2 are components of the mechanotransduction channel in hair cells of the mammalian inner ear

    Neuron 79:504–515.

    https://doi.org/10.1016/j.neuron.2013.06.019

    • PubMed
    • Google Scholar
38. 1. Pan B
    2. Akyuz N
    3. Liu XP
    4. Asai Y
    5. Nist-Lund C
    6. Kurima K
    7. Derfler BH
    8. György B
    9. Limapichat W
    10. Walujkar S
    11. Wimalasena LN
    12. Sotomayor M
    13. Corey DP
    14. Holt JR

    (2018) TMC1 forms the pore of mechanosensory transduction channels in vertebrate inner ear hair cells

    Neuron 99:736–753.

    https://doi.org/10.1016/j.neuron.2018.07.033

    • PubMed
    • Google Scholar
39. 1. Paulino C
    2. Kalienkova V
    3. Lam AKM
    4. Neldner Y
    5. Dutzler R

    (2017a) Activation mechanism of the calcium-activated chloride channel TMEM16A revealed by cryo-EM

    Nature 552:421–425.

    https://doi.org/10.1038/nature24652

    • PubMed
    • Google Scholar
40. 1. Paulino C
    2. Neldner Y
    3. Lam AK
    4. Kalienkova V
    5. Brunner JD
    6. Schenck S
    7. Dutzler R

    (2017b) Structural basis for anion conduction in the calcium-activated chloride channel TMEM16A

    eLife 6:e26232.

    https://doi.org/10.7554/eLife.26232

    • PubMed
    • Google Scholar
41. 1. Periole X
    2. Cavalli M
    3. Marrink SJ
    4. Ceruso MA

    (2009) Combining an elastic network with a coarse-grained molecular force field: Structure, dynamics, and intermolecular recognition

    Journal of Chemical Theory and Computation 5:2531–2543.

    https://doi.org/10.1021/ct9002114

    • PubMed
    • Google Scholar
42. 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
43. 1. Phillips R
    2. Ursell T
    3. Wiggins P
    4. Sens P

    (2009) Emerging roles for lipids in shaping membrane-protein function

    Nature 459:379–385.

    https://doi.org/10.1038/nature08147

    • PubMed
    • Google Scholar
44. 1. Pliotas C
    2. Dahl AC
    3. Rasmussen T
    4. Mahendran KR
    5. Smith TK
    6. Marius P
    7. Gault J
    8. Banda T
    9. Rasmussen A
    10. Miller S
    11. Robinson CV
    12. Bayley H
    13. Sansom MS
    14. Booth IR
    15. Naismith JH

    (2015) The role of lipids in mechanosensation

    Nature Structural & Molecular Biology 22:991–998.

    https://doi.org/10.1038/nsmb.3120

    • PubMed
    • Google Scholar
45. 1. Punjani A
    2. Rubinstein JL
    3. Fleet DJ
    4. Brubaker MA

    (2017) cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination

    Nature Methods 14:290–296.

    https://doi.org/10.1038/nmeth.4169

    • PubMed
    • Google Scholar
46. 1. Ranade SS
    2. Syeda R
    3. Patapoutian A

    (2015) Mechanically activated ion channels

    Neuron 87:1162–1179.

    https://doi.org/10.1016/j.neuron.2015.08.032

    • PubMed
    • Google Scholar
47. 1. Robert X
    2. Gouet P

    (2014) Deciphering key features in protein structures with the new ENDscript server

    Nucleic Acids Research 42:W320–W324.

    https://doi.org/10.1093/nar/gku316

    • PubMed
    • Google Scholar
48. 1. Robertson MJ
    2. Tirado-Rives J
    3. Jorgensen WL

    (2015) Improved Peptide and Protein Torsional Energetics with the OPLSAA Force Field

    Journal of Chemical Theory and Computation 11:3499–3509.

    https://doi.org/10.1021/acs.jctc.5b00356

    • PubMed
    • Google Scholar
49. Book

    1. Schrodinger L

    (2015)

    The PyMOL Molecular Graphics System

    Scientific Research.

    • Google Scholar
50. 1. Schroeder BC
    2. Cheng T
    3. Jan YN
    4. Jan LY

    (2008) Expression cloning of TMEM16A as a calcium-activated chloride channel subunit

    Cell 134:1019–1029.

    https://doi.org/10.1016/j.cell.2008.09.003

    • PubMed
    • Google Scholar
51. 1. Sievers F
    2. Wilm A
    3. Dineen D
    4. Gibson TJ
    5. Karplus K
    6. Li W
    7. Lopez R
    8. McWilliam H
    9. Remmert M
    10. Söding J
    11. Thompson JD
    12. Higgins DG

    (2011) Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega

    Molecular Systems Biology 7:539.

    https://doi.org/10.1038/msb.2011.75

    • PubMed
    • Google Scholar
52. 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 14:354–360.

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

    • PubMed
    • Google Scholar
53. 1. Stansfeld PJ
    2. Goose JE
    3. Caffrey M
    4. Carpenter EP
    5. Parker JL
    6. Newstead S
    7. Sansom MS

    (2015) MemProtMD: automated insertion of membrane protein structures into explicit lipid membranes

    Structure 23:1350–1361.

    https://doi.org/10.1016/j.str.2015.05.006

    • PubMed
    • Google Scholar
54. 1. Stansfeld PJ
    2. Sansom MS

    (2011) From coarse grained to atomistic: A serial multiscale approach to membrane protein simulations

    Journal of Chemical Theory and Computation 7:1157–1166.

    https://doi.org/10.1021/ct100569y

    • PubMed
    • Google Scholar
55. 1. Suloway C
    2. Pulokas J
    3. Fellmann D
    4. Cheng A
    5. Guerra F
    6. Quispe J
    7. Stagg S
    8. Potter CS
    9. Carragher B

    (2005) Automated molecular microscopy: the new Leginon system

    Journal of Structural Biology 151:41–60.

    https://doi.org/10.1016/j.jsb.2005.03.010

    • PubMed
    • Google Scholar
56. 1. Suzuki J
    2. Umeda M
    3. Sims PJ
    4. Nagata S

    (2010) Calcium-dependent phospholipid scrambling by TMEM16F

    Nature 468:834–838.

    https://doi.org/10.1038/nature09583

    • PubMed
    • Google Scholar
57. 1. Suzuki J
    2. Fujii T
    3. Imao T
    4. Ishihara K
    5. Kuba H
    6. Nagata S

    (2013) Calcium-dependent phospholipid scramblase activity of TMEM16 protein family members

    Journal of Biological Chemistry 288:13305–13316.

    https://doi.org/10.1074/jbc.M113.457937

    • PubMed
    • Google Scholar
58. 1. Wang RY
    2. Song Y
    3. Barad BA
    4. Cheng Y
    5. Fraser JS
    6. DiMaio F

    (2016) Automated structure refinement of macromolecular assemblies from cryo-EM maps using Rosetta

    eLife 5:e17219.

    https://doi.org/10.7554/eLife.17219

    • PubMed
    • Google Scholar
59. 1. Yang YD
    2. Cho H
    3. Koo JY
    4. Tak MH
    5. Cho Y
    6. Shim WS
    7. Park SP
    8. Lee J
    9. Lee B
    10. Kim BM
    11. Raouf R
    12. Shin YK
    13. Oh U

    (2008) TMEM16A confers receptor-activated calcium-dependent chloride conductance

    Nature 455:1210–1215.

    https://doi.org/10.1038/nature07313

    • PubMed
    • Google Scholar
60. 1. Yuan F
    2. Yang H
    3. Xue Y
    4. Kong D
    5. Ye R
    6. Li C
    7. Zhang J
    8. Theprungsirikul L
    9. Shrift T
    10. Krichilsky B
    11. Johnson DM
    12. Swift GB
    13. He Y
    14. Siedow JN
    15. Pei ZM

    (2014) OSCA1 mediates osmotic-stress-evoked Ca2+ increases vital for osmosensing in Arabidopsis

    Nature 514:367–371.

    https://doi.org/10.1038/nature13593

    • PubMed
    • Google Scholar
61. 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
62. 1. Zhang M
    2. Wang D
    3. Kang Y
    4. Wu JX
    5. Yao F
    6. Pan C
    7. Yan Z
    8. Song C
    9. Chen L

    (2018) Structure of the mechanosensitive OSCA channels

    Nature Structural & Molecular Biology 25:850–858.

    https://doi.org/10.1038/s41594-018-0117-6

    • PubMed
    • Google Scholar
63. 1. Zhao X
    2. Yan X
    3. Liu Y
    4. Zhang P
    5. Ni X

    (2016) Co-expression of mouse TMEM63A, TMEM63B and TMEM63C confers hyperosmolarity activated ion currents in HEK293 cells

    Cell Biochemistry and Function 34:238–241.

    https://doi.org/10.1002/cbf.3185

    • PubMed
    • Google Scholar
64. 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

Congratulations, we are pleased to inform you that your article, "Cryo-EM structure of the mechanically activated ion channel OSCA1.2", has been accepted for publication in eLife.

In their manuscript, the authors describe the structure of the mechanically activated channel OSCA1.2 determined by cryo-EM in detergent and lipid nanodiscs. The data is of very high quality and both structures appear to be well refined. Whereas the bulk of the functional data supporting the conclusions are presented in an accompanying manuscript, the work also includes the functional characterization of a putative pore-lining mutant and data from molecular dynamics simulations that both support the putative location of the pore. Although this pore is discontinuous, as expected for a stable closed conformation in the absence of force, the proposal is credible. In summary, I think that this is a very exciting study, which describes a novel architecture of a mechanically activated membrane protein. The recently published structure of a homolog does in no way compromise the general interest of this work.

Minor comments are listed below:

1) Please specify which PEI was used in large-scale transfections, in Materials and methods.

2) There are a couple of Ramachandran outliers that should be fixed (one lysine per chain in the membrane hook).

3) Please provide a bit more detail (in the Materials and methods) on how Rosetta was used in model refinement.

4) Figure 1—figure supplement 1A, H: is the y-axis trp fluorescence or A280?

5) Figure 1—figure supplement 1C, J: were the apparently junk classes included in refinement? For example, panel J, row 3, column 1?

6) Figure 1—figure supplement 2: at the final step, can you please elaborate a bit more on what is meant by Relion 3D autorefine x2? Detail added in the legend or in the Materials and methods would be fine.

7) I am generally curious about confidence in positioning of anionic side chains proposed to be important in conformational transitions or permeation. Specifically, please provide a figure panel or supplement panel that shows EM density quality for side chains: Figure 2—figure supplement 2C, in particular the E531 and its electrostatic interactions; D523 vicinity in Figure 3C. If density for these residues is clearly displayed already in Figure 1—figure supplement 3, please just indicate them with an arrow or asterisk. The equivalent residue to E531 appears to have been modeled quite differently in the recent NSMB paper (Zhang et al., 2018) on OSCA channels – it may be that the density is simply not clear in one or both cases.

8) It should be mentioned whether the data shows bound lipids in the vicinity of the protein, particularly at the region between the two subunits. It would also be interesting to learn more about features of the lipid environment surrounding the dimeric protein from a low-pass filtered map of the protein in nanodiscs. Is there any evidence that protein induces distortions of the membrane environment as observed in MD simulations?

9) Since the MD simulations are an important part of the work, they might deserve better coverage in the Results. It can be expected that the gap between subunits would be filled with lipids and the wide entrances to the pore region would be occupied by water. Is this not already the case in the initial setups used for simulation? Was the equilibration of the system the only purpose of the atomistic simulations? In Figure 1—figure supplement 6, the authors show snapshot after 50 ns of simulations. It would be interesting to know what has changed in the solvent, protein and membrane region compared to the starting structure.

10) The authors speculate that gap between two helices at the intracellular part of the pore leading to its exposure to the membrane might play a role during channel activation. It could be mentioned that a similar opening is found in TMEM16A, which has so far not been described to be mechanically gated.

11) The authors emphasize the similarity between detergent and nanodisc structures in the Results but propose that differences between both structures might indicate their movements during activation in the Discussion. This is confusing. I suggest that the small differences between both structures could already be better described in the Results.

12) Results, fifth paragraph: Ca²⁺-activation in TMEM16A involves an α-to-π transition.

13) It would be helpful to show Figure 3—figure supplement 1A as stereo figure.

Minor comments are listed below:

1) Please specify which PEI was used in large-scale transfections, in Materials and methods.

PEI MAX 40K was used for large scale transfections. This detail has been added in the Materials and methods section.

2) There are a couple of Ramachandran outliers that should be fixed (one lysine per chain in the membrane hook).

We have decided not to fix these outliers due to this region being flexible and the lowest resolution region of our maps. Our current models represent our best attempt at fitting the residues of the membrane hook inside the density, which resulted in one outlier in this region.

3) Please provide a bit more detail (in the Materials and methods) on how Rosetta was used in model refinement.

More detail has been added in the Materials and methods section.

4) Figure 1—figure supplement 1A, H: is the y-axis trp fluorescence or A280?

Changes have been made on Figure 1—figure supplement 1 to show that the measurement corresponds to UV absorbance (A280).

5) Figure 1—figure supplement 1C, J: were the apparently junk classes included in refinement? For example, panel J, row 3, column 1?

For LMNG-solubilized OSCA1.2 (Figure 1—figure supplement 1J), no filtering of particles was done in the 2D classification stage. Data were processed with and without filtering for the best 2D classes. The final reconstruction did not differ greatly between both alternatives. Furthermore, a 2D classification of the final 3.5 Å model shows that particles that made up the junk classes were filtered out during ab-initio reconstruction. For nanodisc-embedded OSCA1.2 structure (Figure 1—figure supplement 1C), 2D classification was performed using cryoSPARC and only the best classes were used for subsequent analysis, as explained in the Materials and methods.

6) Figure 1—figure supplement 2: at the final step, can you please elaborate a bit more on what is meant by Relion 3D autorefine x2? Detail added in the legend or in the Materials and methods would be fine.

“Relion 3D autorefine x2” means two subsequent rounds of 3D autorefine using relion. The second round uses the results of the previous one as input. This has been made clearer in the Materials and methods section.

7) I am generally curious about confidence in positioning of anionic side chains proposed to be important in conformational transitions or permeation. Specifically, please provide a figure panel or supplement panel that shows EM density quality for side chains: Figure 2—figure supplement 2C, in particular the E531 and its electrostatic interactions; D523 vicinity in Figure 3C. If density for these residues is clearly displayed already in Figure 1—figure supplement 3, please just indicate them with an arrow or asterisk. The equivalent residue to E531 appears to have been modeled quite differently in the recent NSMB paper (Zhang et al., 2018) on OSCA channels – it may be that the density is simply not clear in one or both cases.

As requested, residues D523 and E531 have been labeled in Figure 1—figure supplement 3. Although the density for E531 only reaches the β-carbon and it is possible that the side chain has a different rotamer, the fact that E531 is facing the pore remains the same.

8) It should be mentioned whether the data shows bound lipids in the vicinity of the protein, particularly at the region between the two subunits. It would also be interesting to learn more about features of the lipid environment surrounding the dimeric protein from a low-pass filtered map of the protein in nanodiscs. Is there any evidence that protein induces distortions of the membrane environment as observed in MD simulations?

There are in fact small densities in the inter-subunit cleft. Although these densities could correspond to lipids or cholesterol, there is also the chance that they represent detergent carried over from the purification or refined noise during processing. Thus, we have decided not to explicitly assign these densities as bound lipids. Regarding the lipid environment, in our current nanodisc map we do not observe clear evidence for the protein causing distortion of the membrane environment.

9) Since the MD simulations are an important part of the work, they might deserve better coverage in the Results. It can be expected that the gap between subunits would be filled with lipids and the wide entrances to the pore region would be occupied by water. Is this not already the case in the initial setups used for simulation? Was the equilibration of the system the only purpose of the atomistic simulations? In Figure 1—figure supplement 6, the authors show snapshot after 50 ns of simulations. It would be interesting to know what has changed in the solvent, protein and membrane region compared to the starting structure.

Videos are now added to reflect the arrangement of lipid and water molecules during the simulation. Indeed, the initial lipid self-assembly already placed lipids in the inter-subunit cleft, and waters occupying the entrance into the pore. While the equilibration of the system is one of the purposes of the atomistic simulation, this also allowed us to probe the dynamic interactions of waters and lipids with the protein. To this end, we have extended the original AT simulation for an additional 200 ns, without any restraints. This allows us to observe how the mechanosensing features of OSCA1.2 evolve when the system is allowed to relax. The changes of the protein, lipids and waters in the pore are now displayed in the videos and specific changes/lack of changes are described in the accompanying captions.

10) The authors speculate that gap between two helices at the intracellular part of the pore leading to its exposure to the membrane might play a role during channel activation. It could be mentioned that a similar opening is found in TMEM16A, which has so far not been described to be mechanically gated.

We have added in the text that the opening is similar to the one found in TMEM16 structures.

11) The authors emphasize the similarity between detergent and nanodisc structures in the Results but propose that differences between both structures might indicate their movements during activation in the Discussion. This is confusing. I suggest that the small differences between both structures could already be better described in the Results.

Our initial statement compared the LMNG-solubilized and nanodisc-embedded OSCA1.2 structures to highlight which regions may be more flexible in the closed state. However, we do not claim that these movements are the ones that lead to activation of the channel. We have slightly modified the statement in hopes to make it clearer. Given that the changes between the two structures are better represented on Video 8 than what we could do in words, we have decided not to further discuss it in the text.

12) Results, fifth paragraph: Ca²⁺-activation in TMEM16A involves an α-to-π transition.

The reviewer is correct. This has been corrected in the text. We thank the reviewer for catching this mistake.

13) It would be helpful to show Figure 3—figure supplement 1A as stereo figure.

In order to incorporate the reviewer’s suggestion, but also make this view accessible to everyone (not everyone is trained in seeing stereo views), we have decided to keep Figure 3—figure supplement 1 as is and we have added Video 3, which shows a 360° view of the pore.

Author details

1. Sebastian Jojoa-Cruz

   Department of Integrative Structural and Computational Biology, The Scripps Research Institute, California, United States

   Contribution

   Conceptualization, Formal analysis, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing

   Contributed equally with

   Kei Saotome

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-4392-3898

2. Kei Saotome

   1. Department of Integrative Structural and Computational Biology, The Scripps Research Institute, California, United States
   2. Department of Neuroscience, Dorris Neuroscience Center, Howard Hughes Medical Institute, The Scripps Research Institute, California, United States

   Contribution

   Conceptualization, Formal analysis, Funding acquisition, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing

   Contributed equally with

   Sebastian Jojoa-Cruz

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-4135-5356

3. Swetha E Murthy

   Department of Neuroscience, Dorris Neuroscience Center, Howard Hughes Medical Institute, The Scripps Research Institute, California, United States

   Contribution

   Conceptualization, Formal analysis, Investigation, Visualization, Methodology, Writing—review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-9580-3380

4. Che Chun Alex Tsui

   1. Department of Integrative Structural and Computational Biology, The Scripps Research Institute, California, United States
   2. Department of Biochemistry, University of Oxford, Oxford, United Kingdom

   Contribution

   Formal analysis, Funding acquisition, Investigation, Visualization, Methodology, Writing—review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-4886-9824

5. Mark SP Sansom

   Department of Biochemistry, University of Oxford, Oxford, United Kingdom

   Contribution

   Resources, Supervision, Funding acquisition, Project administration, Writing—review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-6360-7959

6. Ardem Patapoutian

   Department of Neuroscience, Dorris Neuroscience Center, Howard Hughes Medical Institute, The Scripps Research Institute, California, United States

   Contribution

   Conceptualization, Resources, Supervision, Funding acquisition, Project administration, Writing—review and editing

   For correspondence

   ardem@scripps.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-0726-7034

7. Andrew B Ward

   Department of Integrative Structural and Computational Biology, The Scripps Research Institute, California, United States

   Contribution

   Conceptualization, Resources, Supervision, Funding acquisition, Project administration, Writing—review and editing

   For correspondence

   andrew@scripps.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-7153-3769

• 7,405

  Page views

• 1,119

  Downloads

• 86

  Citations

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