Tentonin 3 is a pore-forming subunit of a slow-inactivation mechanosensitive channel

doi:10.21203/rs.3.rs-2785213/v1

Download PDF

Research Article

Tentonin 3 is a pore-forming subunit of a slow-inactivation mechanosensitive channel

https://doi.org/10.21203/rs.3.rs-2785213/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 31 May, 2024

Read the published version in Cell Reports →

Version 1

posted

You are reading this latest preprint version

Mechanically activating (MA) channels transduce numerous physiological functions. Tentonin 3/TMEM150C (TTN3) confers MA currents with slow-inactivation kinetics in somato- and baro-sensory neurons. Despite the distinct currents when heterologously expressed, whether TTN3 forms a channel pore is unknown. Here, we demonstrate that purified TTN3 proteins incorporated into the lipid bilayer exhibited spontaneous and pressure-sensitive channel currents. MA currents of TTN3 with unique slow-inactivation kinetics were conserved throughout the vertebrate phyla. The activation threshold of TTN3 and its pharmacological perturbations separated TTN3 from Piezo1. Deep neural network structure prediction programmes coupled with mutagenetic analysis predicted a rectangular shaped, tetrameric structure with six transmembrane helices and a pore at the inter-subunit centre. The putative pore aligned with two helices of each subunit and had constriction sites whose mutations changed the MA currents. These results suggest that TTN3 is a pore-forming subunit of a slow-inactivation MA channel, possibly with a unique structure.

Molecular Biology

Structural Biology

Tentonin3

TMEM150C

Ion Channel

Mechanosensitive

Ion Pore

NMB-1

Structure

Mechanotransduction is a fundamental biological mechanism that is necessary for physiological functions such as somatosensation, motor coordination, baroreceptor reflex, hearing and pain reception ^1–3. Perceiving the mechanical stimuli for these functions begins with the mechanically activating (MA) channels in the sensory end organs that convert mechanical stimuli into electrical signals ⁴. In the dorsal root ganglion (DRG) neurons that detect touch, pressure, vibration and mechanical pain, there are three types of native MA currents ^3–5 namely; rapidly adapting (RA), slowly adapting (SA), and intermediate-adapting (IA) currents depending on their inactivation times. These distinct MA currents are also observed in the baroreceptor neurons of the nodose ganglia ⁶.

Many candidate genes have been identified in various animal species for their mechano-sensitivity. Piezo channels are activated by mechanical indentation with RA and possibly IA kinetics ^7,8. The Piezo channels are differentially expressed in numerous cell types and mediate the diverse physiological and cellular functions ⁹. Therefore, their dysfunction in humans causes numerous hereditary diseases ¹⁰. Cryo-electron micrograph analysis has revealed the unique molecular structure of Piezo1 which has propeller blades and a concave body that easily senses change in membrane tension ^9,11,12.

Tentonin 3/TMEM150c (TTN3) was identified to be a gene responsible for the SA-type MA currents in the DRG neurons ⁵. The heterologous expression of TTN3 gave rise to robust currents in response to mechanical step stimuli, with two inactivation phases, a rapid then slow inactivation ⁵. Genetic deletion of the Ttn3 gene in DRG and nodose ganglion neurons ablates the SA-type MA currents ^5,6. Moreover, TTN3 is expressed in muscle spindles, baroreceptors, and pancreatic β-cells, and its ablation results in a loss of muscle coordination, blood-pressure control, and insulin release, respectively ^5,13,14. However, because TTN3 MA currents were lost in Piezo1-knock-out HEK (human embryonic kidney) cells (HEK-P1KO), its mechanosensitivity was suspected ^15–17. Thus, the present study aimed to determine whether TTN3 was a pore-forming subunit of an MA channel. In addition, we utilised deep learning methods to achieve prediction accuracies greater than those of force-field-based approaches ^18–22 to gain structural insight into TTN3 action.

Mechanosensitivity of TTN3 is conserved throughout the vertebrate phyla

Since TTN3 is expressed only in the vertebrate phyla⁵, its seven representative orthologs (Supplementary Fig. 1) were tested for their mechanosensitive properties. Step mechanical indentations (6.4 ~ 8.5 µm, 600 ms) with a glass micro-probe elicited MA currents in HEK293T cells transfected with mouse (Mus musculus), zebrafish (Danio rerio), chick (Gallus gallus), human (Homo sapiens), cat (Felis catus), turtle (Chrysemys picta bellii), and frog (Xenopus tropicalis) Ttn3 (Fig. 1a, Supplementary Fig. 1). Among the orthologs, HEK293T cells transfected with zebrafish Ttn3 (drTTN3) displayed the greatest MA current amplitude, whereas cells expressing turtle and frog Ttn3 displayed the least response to the mechanical stimuli (Fig. 1b).

The five (mouse, human, zebrafish, cat, and chick) TTN3 orthologs demonstrated typical MA currents to mechanical step stimuli; a rapid activation followed by a rapid and slow inactivation, similar to the SA-type MA currents observed in the mouse DRG and nodose ganglion neurons ^4–6.

The TTN3 orthologs showed displacement-response relationships. Among the orthologs, drTTN3 was most sensitive to mechanical stimuli as the half-maximal displacement was significantly smaller than that of mouse TTN3 (mTTN3) (4.9 ± 0.6 vs 6.5 ± 0.4 µm, p < 0.05, Student’s t-test) (Fig. 1c). In addition, these TTN3 orthologs showed different sensitivity to gadolinium (Gd³⁺), which is a non-selective blocker of MA channels ²³. Mechanical step-activated currents of mTTN3 were almost completely inhibited by 100 µM Gd³⁺ (Fig. 1d,e) ⁵. However, drTTN3 showed only ~ 60% inhibition by 100 µ M Gd³⁺ (Fig. 1d,e). Thus, these results suggest that the mechanosensitivity of TTN3 was well conserved throughout the vertebrate phyla, but with different mechanical and pharmacological sensitivities.

TTN3 in the lipid bilayer elicits spontaneous and stretch-activated channel currents

To determine if TTN3 retains a pore-forming subunit, we purified human TTN3 (hTTN3) and drTTN3 and incorporated the proteins into the lipid bilayer to determine their spontaneous and stretch-activated single-channel currents. We isolated hTTN3 and drTTN3 proteins from Escherichia coli and stabilised them with an amphiphilic polymer, alkyl polyglycoside (APG), for better dispersion in the aqueous media (Supplementary Fig. 2a,b) ²⁴. Proteoliposomes containing hTTN3 or drTTN3 were fused and reconstituted into a model 1,2-diphytanoyl-sn-glycero-3-phosphocholine (DPhPC) membrane-spanning across an 80–100 µm aperture in a polytetrafluoroethylene film between the cis and trans chambers (Supplementary Fig. 2c) ²⁵. For the positive control, proteoliposomes containing purified bacterial MscL proteins were incorporated into the DPhPC model membrane, which exhibited typical large and Gd³⁺ reversible single-channel currents (Supplementary Fig. 2d) ²⁶. Similarly, spontaneous channel openings were observed in the membranes that were fused with the proteoliposomes containing hTTN3 or drTTN3 (Fig. 2a,b). The mean amplitudes of the unitary single-channel currents of hTTN3 or drTTN3 at 100 mV were 4.21 \(\pm\) 0.48 or 5.98 \(\pm\) 0.92 pA, respectively (Fig. 2c). The mean amplitude of the single-channel currents of drTTN3 was significantly greater than that of hTTN3 (Fig. 2c). The current and voltage (I–V) relationships between the single-channel currents of both hTTN3 and drTTN3 were linear in a symmetric 150 mM KCl solution (Fig. 2d). Consistent with the results of the whole-cell currents in HEK293T cells, drTTN3 was more active than hTTN3 as the open probability of drTTN3 was significantly (p < 0.001, Student’s t-test) greater than that of hTTN3 at 30 min after reconstitution (Fig. 2e). Spontaneous channel currents of hTTN3 were completely blocked by Gd³⁺, whereas those of drTTN3 were partially blocked (Fig. 2f). The activity of these single-channel currents increased over time and eventually became macroscopic currents (Fig. 2g). Macroscopic currents of hTTN3 were also blocked by 20 µM of Gd³⁺, whereas those of drTTN3 were partially blocked (Fig. 2f-h), thereby confirming the different Gd³⁺sensitivity of hTTN3 and drTTN3 when expressed in HEK293T cells (Fig. 1e).

Both hTTN3 and drTTN3 were sensitive to the stretches applied to the lipid bilayer. When 20 µL of the solution (equivalent to 0.42 mmH₂O) was added to a cis or trans chamber, an abrupt increase in reversible MA channel currents was observed in both hTTN3 and drTTN3-reconstituted membranes (Fig. 2i,j). These MA currents were not observed in the blank lipid bilayer (Fig. 2i,j). These results suggest that both hTTN3 and drTTN3 are pore-forming subunits which are sensitive to membrane stretch.

TTN3 mechanosensitivity is separate from Piezo1

To determine whether TTN3 and Piezo1 have a different sensitivity to mechanical stimuli, we measured the MA currents of both mTTN3 and mPiezo1 in response to different speeds (8 µm, 80 ~ 2,560 µm/sec) of a mechanical step stimulus. As shown in Fig. 3a, step stimuli with a speed less than 160 µm/s rarely activated mTTN3, whereas a speed over 80 µm/s readily activated mPiezo1 when overexpressed in HEK293T cells (Fig. 3a). The half-maximal speed of mPiezo1 was significantly lower than that of mTTN3 (Fig. 3b). In addition, the threshold (10% maximal response) of mPiezo1 was markedly lower than that of mTTN3. Thus, the sensitivity to mechanical stimuli of TTN3 appears to be lower than that of Piezo1.

A mutant of an ρ-conotoxin (r-TIA), also known as noxious mechanosensation blocker 1 (NMB-1), is known to inhibit SA but not RA MA currents in DRG neurons ²⁷. Because TTN3 confers the SA-MA currents observed in DRG neurons ⁵, NMB-1 would inhibit the TTN3 MA currents. The application of 10 µM NMB-1 strongly inhibited the MA currents in TTN3/HEK cells (Fig. 3c,d). In contrast, 10 µM NMB-1 failed to inhibit the Piezo1 MA currents (Fig. 3c,d). In addition, Yoda1 is a selective agonist of Piezo1 known to slow the inactivation of Piezo1 ^8,28−30. If the TTN3 mechanosensitivity was dependent on Piezo1, Yoda1 would modulate the kinetics of TTN3. Consistent with previous reports ^28–30, the application of 10 µM Yoda1 to Piezo1/HEK cells increased the inactivation time without affecting the amplitude of the MA currents (Fig. 3e,f). However, the application of Yoda1 to TTN3/HEK cells did not affect the TTN3 inactivation kinetics (Fig. 3g,h). These differences in pharmacological profiles further indicate that TTN3 activation is separate from Piezo1 activity.

mTTN3 tetrameric structures predicted by deep learning-based algorithms

Because state-of-the-art deep learning methods able to predict the native protein structures from their sequences with high confidence have been developed ^{19,20,22,31,32}, we sought to predict the TTN3 structure using these algorithms. First, we determined the subunit composition of the hTTN3 protein. We purified the protein, cross-linked it with bis(β-[4-azidosalicylamido]-ethyl) disulphide, and isolated on SDS-PAGE. The western blot analysis of the cross-linked hTTN3 suggested a tetrameric subunit composition (Supplementary Fig. 3).

The native structure of the mTTN3 monomer was modelled using AlphaFold2 ^20,22,32. The resulting structural model manifested six transmembrane α-helices (S1–S6) that have the highest accuracy scores calculated using the predicted local-distance difference test (pLDDT) in AlphaFold2 ²⁰ (Fig. 4a, Supplementary Fig. 4). This scoring function uses a per-residue measurement that evaluates the local-distance differences between a computational model and the experimentally determined reference structure ³³. The extracellular loop between S1 and S2 and the C-terminus also contained short α-helices (Fig. 4a).

Upon confirming a tetrameric composition of the TTN3 proteins, the optimal tetrameric conformation of the monomers was generated via the high ambiguity-driven DOCKing (HADDOCK) program ^34–36. The optimal orientation/conformation of the four monomers was determined by evaluating the tetramer’s molecular mechanical potential energy as a sum of individual energy terms such as van der Waals, electrostatic, desolvation and restraints violation energies in the presence of an explicit water model ³⁷. The obtained tetrameric structures were grouped into four representative configuration types based on the two most relevant energy terms—van der Waals and electrostatic energies (Fig. 4b,c). All four model configurations demonstrated a pore-like cavity region in the inter-subunit centre. However, the location and orientation of the transmembrane helices (S1–S6) relative to the centre axis of the tetramer were different (Fig. 4b). Each type had a most constricting residue in the putative central pore that could determine ion conduction. The Glu127, Glu33 and Cys66 of the configuration type 1–3, respectively, were positioned at the top of their pores, whereas the Trp98 residue of Type 4 was positioned at the bottom of the pore (Fig. 4b). We then substituted each of these residues with Ala and evaluated the changes in mechanosensitivity. The E127A mutant, located at the pore of the configuration type 1, resulted in a complete reduction in the amplitudes of TTN3 MA currents (Fig. 4d,e). Thus, configuration type 1 was selected as the most probable model representing the native tetrameric structure of mTTN3.

Predicted molecular structure of TTN3

The selected tetrameric structure revealed a square shape monomeric arrangement with a putative pore at the centre of the complex (Fig. 5a,b). Transmembrane helices S3 and S4 formed the inner columns, whereas S1, S2, S5 and S6 formed the outer columns facing the membrane lipids (Fig. 5c; Supplementary Video 1). The central cavity extended from the top to the bottom. This hollow vase-like space has a wall surrounded by S3 and S4 (Fig. 5c,e; Supplementary Video 1). The S3 and S4 cross each other, and the top end of S4 and the bottom end of S3 comprises the upper and lower openings of the pore, respectively (Fig. 5c,e). There are four constricting sites, E127, V131, F139 and W98, along the ion-conduction pathway where the minimum cross-sectional diameter along the centre pore was estimated as 2.08 Å using the program HOLE ³⁸ (Fig. 5d). In addition to the current block with the E127A mutation (Fig. 4d,e), we further mutated the other three constricting residues to cysteine to test assess the inhibition of MA currents by a sulfhydryl-modifying agent, methanethiosulphonate (MTSET). We applied a 6.4 µm step stimulus repeatedly before and after 1 mM-MTSET treatment. As shown in Fig. 5f and 5g, the application of MTSET reduced the MA currents of V131C, F139C and W98C mutants, whereas it failed to change the MA currents of the control mTTN3 (Fig. 5f,g). These results suggest that these residues are aligned along the putative ion conducting pathway.

The structural change in response to lipid bilayer deformation during mechanical force was examined using all-atom molecular dynamics (MD) simulations, where the selected tetrameric TTN3 structure was embedded in the 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) bilayer in a 150 mM KCl solution. The bilayer was moderately pulled laterally by applying a surface tension of 5 dyn/cm parallel to the bilayer plane, while the pressure normal to the bilayer plane was set at 1 atm. The total bilayer surface area increased by 8.50% (averaged over three replicas), while the membrane thickness was reduced by 5.10% after 500 ns simulations (Supplementary Fig. 5). This lateral deformation gave rise to structural reformations of the TTN3 tetramer. While the overall structure was similar, the upper part of the outer helices S5, S6, and S1 were moved outward (Fig. 5h). The upper part of S4 was also moved outward, possibly providing a larger opening of the pore (Fig. 5i). The cross-sectional diameter at the constricting region was estimated to increase by up to 6.06 Å from the initial value of 2.08 Å (Fig. 5d), which agrees with the enhancement of its channel conductance upon activation of the protein.

N- and C-termini of TTN3 contribute to mechanosensitivity

We also postulated that the N- or C-terminus of mTTN3 might be involved in the gating process. Mechanical step stimuli evoked robust MA currents in the control mTTN3/HEK cells (Fig. 6a). In contrast, the deletion of the C-terminus resulted in a marked reduction in the MA currents (Fig. 6a,b). The C-terminus contained a short α-helix that comprises a serine-repeat region (232-SFSESLSEAS-242), which is highly conserved among the orthologs (Supplementary Fig. 1). Thus, we deleted the Ser-rich region, '232-SFSESLSE-240' (ΔSSSS mutant) from the C-terminus. The ΔSSSS mutant also showed a significant reduction in the MA currents (Fig. 6a,b). The N-terminus is a short region spanning six amino acids (1-MDGKKC-6) (Supplementary Fig. 1). The N-terminus deletion (ΔN-term) also caused a significant reduction in the MA currents (Fig. 6a,b). These results indicate that the N- and C-termini of mTTN3 are involved in the channel gating.

Three distinct types of MA currents according to their inactivation kinetics were observed in the somato- and baro-sensory neurons ^3–6. Specific genes were identified for the biophysically defined MA currents. Piezo 1 and 2 confer RA and possibly IA-type MA currents, respectively ^7,8, whereas TTN3 mediates the SA-type MA currents ⁵. The unique TTN3 MA currents with slow-inactivation kinetics were conserved throughout the vertebrate orthologs. TTN3 fulfils the four criteria required for a MA channel ^9,39; (i) genetic ablation of TTN3 markedly reduced the amplitudes and proportions of SA-MA currents in mechanosensory cells ^5,6,14, (ii) heterologous expression of TTN3 in mammalian cells produces MA channel currents (Fig. 1) ^16,40, (iii) genetic ablation of TTN3 elicits functional phenotypes in baroreceptors, proprioceptors, and pancreatic β-cells ^5,6,14, (iv) reconstitution of purified TTN3 protein into a lipid bilayer generated spontaneous and pressure-induced channel currents (Fig. 2), and (v) NMB-1 known to inhibit SA currents in DRG neurons inhibited TTN3 (Fig. 3c,d).

Because TTN3 is not activated by mechanical stimuli in the absence of Piezo1, TTN3 is considered to be a regulator of Piezo1 ^16,17. The genetic deletion of Piezo1 or Piezo2 reduced the cytoskeletal integrity ^41,42, which would result in the loss of TTN3 mechanosensitivity in Piezo1-ablated HEK cells. Indeed, mechanical stimuli failed to evoke MA currents in Piezo1-ablated HEK cells transfected with mTTN3 ⁶. However, after the pre-treatment of the cells with an F-actin nucleation enhancer, jasplakinolide, the mechanical step stimuli evoked robust MA currents with distinct inactivation kinetics ^6,43,44. Disruption of the cytoskeleton with cytochalasin D ablates the MA currents in TTN3-transfected HEK cells ⁶. The loss of cytoskeletal integrity in Piezo1-deficient cells might reduce the mechanical force necessary for gating TTN3. Furthermore, the speed and displacement thresholds for mechanical stimuli required to activate TTN3 were significantly greater than those for Piezo1 (Fig. 3a,b) ⁵. Thus, these factors account for the loss of TTN3 MA currents in Piezo1-ablated cells.

Despite the difference in structural, biophysical, and pharmacological profiles between TTN3 and Piezo1, many physiological functions of TTN3 overlap with those of Piezo channels as both are expressed in muscle spindles and aortic baroreceptors, thereby mediating muscle coordination and blood-pressure regulation, respectively ^5,6,14. The overlapping expression and function of the different molecules are not unusual, as seen in different transient receptor potential (TRP) channels for temperature or pain sensing ^45,46. The redundancy of molecular sensors provides a backup system for the loss of any key functions.

The predicted tetrameric model of TTN3 suggests a putative pore at the inter-subunit centre. The pore in the centre appears to be an ion-conduction lining because mutagenesis of the restricting residues markedly reduced the TTN3 currents. In the MD simulation, when tension was applied to the lipid bilayer, the size of the putative pore increased (Fig. 5d,h), which may explain the activation of TTN3 through membrane tension. However, because the deletion of the N- or C-termini reduced the TTN3 MA currents, we cannot overlook the contribution of the stretch of the C- or N-terminus by tethering to the subcellular structures such as cytoskeletons. Although the structure modelling provided an insight into the overall structure of TTN3, the channel activation followed by inactivation through mechanical stimuli remains unknown. Structural information of the native form of TTN3 with cryo-electron microscopy analysis coupled with large-scale MD simulations would answer the detailed molecular mechanisms underlying its activation and inactivation.

Cell culture and transfection

HEK293T (ATCC, CRL-3216) and Piezo1-knocked out HEK293T (HEK-P1KO; Gift from Ardem Patapoutian) cell lines were maintained in Dulbecco’s modified eagle medium (Gibco, 11–995) supplemented with 10% foetal bovine serum (sh30919.03), and 1% penicillin-streptomycin (Gibco, 15140). Both cell lines were maintained and cultured in a humidified incubator at 37°C with 5% CO₂.

Cells were seeded in 35 mm dishes and cultured until approximately 50% cell confluency was reached. At a 1:3 ratio of DNA to transfection reagent ratio, 1 µg of DNA was transfected with 3 µL of FuGENE HD transfection reagent (Promega, E2311) for 15 min. TTN3 ortholog constructs, murine TTN3 single mutants and murine Piezo1 DNA constructs were subcloned into pIRES2-acGFP vectors and were utilised for transfection procedures. After 24 h, the transfected cells were collected and plated into 35 mm dishes containing glass coverslips, which were used for whole-cell patch-clamp recording the following day.

All basic chemicals were purchased from Sigma Aldrich, unless specified otherwise.

Gene preparation and mutagenesis

Tentonin 3/TMEM150C (TTN3) ortholog genes (mouse (Genscript (OMu08247)), zebrafish (Genscript (Oda15731)), frog (Genscript (OXa04470)), chicken (Genscript (OGa35266)), turtle (Genscript (Och161608)), cat (Genscript (OFb05814)), and human (Genscript (OHu05487))) were subcloned into pIRES2-acGFP vectors ^5,6. TTN3 target constructs were amplified with Dreamtaq Green Polymerase (Thermo Fisher Scientific) through PCR. DNA fragments separated in agarose gels were extracted and purified with a FavorPrep™ Gel/PCR Purification Kit (Favorgen). The fragments underwent digestion with restriction enzymes (Thermo Fisher Scientific) and were ligated with T4 ligase (Thermo Fisher Scientific) using different molar insert-to-vector ratios; 1:1, 3:1, and 5:1. Transformation of the ligated products was performed with DH5α-chemically competent E. Coli (Enzynomics).

The primers used for subcloning are as follows:

zebrafish TTN3: forward primer 5′-CGCCCTTCCGCGGGAGTTGGTCAGAC-3′, reverse primer 5′-GTCTGACCAACTCCCGCGGAAGGGCG-3′,

frog TTN3: forward primer 5′-TAATACGACTCACTATAGGG-3′, reverse primer 5′-TAGAAGGCACAGTCGAGG-3′,

chicken TTN3: forward primer 5′-CTCGAGGCCACCATGGACGGGAA-3′, reverse primer 5′-GTCGACCTACACCTGATCCGTCTGGTA-3′,

turtle TTN3: forward primer 5′-GCTAGCGCCACCATGGATGGGAAGAAAT-3′, reverse primer 5′-AGTATCAAACAGATCAGGTGTAGCTCGAG-3′,

cat TTN3: forward primer 5′-CTCGAGGCCACCATGGATGGGAAG-3′, and reverse primer 5′-GTCGACTTACACCTGGTCGGTTTGATA-3′,

human TTN3: forward primer 5`-TAGGTGAACTCGAGGCCACCATGGATGGGAAGA-3′, and reverse primer 5′-CGATAACCGCGGTACCttACACCTGGTCAG-3′

Mutants of mTTN3 were generated using either a Muta-Direct™ Site-directed Mutagenesis Kit (Intron, 15071) or EZchange™ Site-Directed Mutagenesis Kit (Enzynomics, EZ004S). Primers used for site-directed mutagenesis were specifically generated for each mutant type. The mutation products underwent transformation in DH5α-chemically competent E. coli cells and later cultured on Luria-Bertani (LB) agar plates. The TTN3 mutations in the constructs were confirmed through sequencing.

Whole cell current recording

Whole-cell currents of HEK293T or N2A cells were recorded through the voltage-clamp technique. After forming a gigaseal with a glass pipette, the membrane patch under the glass pipette was ruptured with gentle suction to make a whole cell. Tip resistance of the glass pipettes was 2–3 MΩ. After forming a whole cell, capacitive transients were cancelled. The holding potential was set at -60 mV. Whole cell currents were recorded with an Axopatch 200B amplifier (Molecular Devices). The amplifier output was filtered at 1 kHz and digitised with Digidata 1550B (Molecular Devices) at a sampling rate of 5 kHz and stored on a PC for later analysis. Whole cell currents were analysed with Clampfit 10.0 software (Molecular Devices). The inactivation curves were fitted to two exponentials. The intracellular (pipette) solution (mM) contained 130 CsCl, 2 MgCl₂, 4 Mg-ATP, 0.4 Na-GTP, 25 D-Mannitol and 10 HEPES adjusted to pH 7.2 with CsOH. The extracellular (bath) solution (mM) contained 130 NaCl, 5 KCl, 1 CaCl₂, 2 MgCl₂, 10 D-Mannitol and 10 HEPES adjusted to pH 7.2 with NaOH. All solution osmolarities were further adjusted to 295 mOsm with D-Mannitol. For the pharmacological experiments, GdCl₃ (Sigma), Yoda1 (Tocris Bioscience), or NMB-1 (Anygen, Korea) were added to the bath solution.

Mechanical stimulation

Whole cells of HEK293T cells transfected with various TTN3 orthologs or Piezo1 were physically stimulated with a glass pipette as previously described ^5,47. Briefly, after a glass pipette was pulled, the tip was fire-polished to produce a blunt tip 2 ~ 3 µm in diameter. The glass probe was fixed to the end of a Nano-controller NC4 micromanipulator (MM3A-LS, Kleindiek Nanotechnik, Reutlingen, Germany). The movement of the glass probe was controlled with a joystick or via a computer program provided by Kleindiek. The nano-controller was a Piezo-driven small motor that had two modes of operation; fine and coarse movement mode. We calibrated the actual movements of the glass probe. In the coarse mode, the glass probe moved on average of 0.266 µm per step. Distances of the mechanical step stimuli were set at 24, 28 and 32 steps for 6.4, 7.4, and 8.5 µm displacement, respectively. We generally used the coarse 8 modes for mechanical stimulation, where the piezoelectric device vibrated at 2.7 kHz corresponding to the speed of 718 µm/s except for the speed-current relationship experiments. For the speed-current relationship experiments, the frequencies of the nano-controller motion were set at 0.3, 0.6, 1.2, 2.4, 4.8, and 9.6 kHz, which corresponded to the velocity being approximately 80, 160, 319, 638, 1277, and 2550 µm/s, respectively. The displacement of the probe was fixed at 8 µm for a 600 ms duration. Before physical indentation, the tip of the mechanical probe was placed on top of the cell surface at a 50° angle and withdrawn by 1 µm, which was regarded as an initial point for the mechanical step.

Expression and purification of hTTN3 in E. coli

The P9 sequence was attached to the N-terminus of hTTN3 (or drTTN3) to facilitate the expression of hTTN3 in the membrane fraction of E. coli ²⁴. For the mass production of the P9-TTN3 fusion protein, BL21 (DE3) star-pRARE cells harbouring the P9* expression vector, were grown in 5 mL LB containing 0.05 mg/mL carbenicillin and 0.05 mg/mL chloramphenicol for 16 h at 37°C. One millilitre of the pre-culture was inoculated into 50 mL of YTN (Yeast extract 1%, Bactotryptone 2%, NaCl 2%) medium containing 0.05 mg/mL carbenicillin and 0.05 mg/mL chloramphenicol and cultured at 37°C until the OD_600nm reached 1.0. One-third of the culture was inoculated into 150 mL of the YTN medium containing 0.05 mg/mL carbenicillin and 0.05 mg/mL chloramphenicol and cultured at 37°C until the OD_600nm reached between 1.0 and 2.0. The culture was then inoculated into a fermenter (Marado-PDA, Korea) containing 2 L culture medium (K₂HPO₄ 0.3%, KH₂PO₄ 0.5%, Yeast extract 2%, glucose 2%, MgSO₄·7H₂O 0.06%) containing 0.05 mg/mL carbenicillin and 0.05 mg/mL chloramphenicol. The cells were grown at 37°C, pH 6.7–6.8 and P_O2 of 30–40%. The expression of P9-TTN3 was induced by the addition of 1 mM IPTG for 3 h at 37°C when the OD_600nm of the culture was 46. The pH of the culture in the fermenter was maintained at 6.8 by the addition of a small volume of the feeding medium (yeast extract 27.4%, (NH₄)₂SO₄ 0.15%, glucose 21.1%, MgSO₄·7H₂O 0.1%). The cells were harvested by centrifugation for 20 min at 3,000 × g and stored at -80°C until further use.

Approximately 30 g of E. coli cells (wet weight) were resuspended in buffer A (50 mM Tris-HCl, pH 8.5, 1 mM EDTA) containing 1 mM phenylmethylsulphonyl fluoride (Merck, Germany) and 100 µg protease inhibitor cocktail (genDEPOT), and lysed using a microfluidizer M-110P (Microfluidics). The lysate was centrifuged at 15,000 × g for 30 min, and the resulting supernatant was further centrifuged at 100,000 × g for 1 h. The membrane fraction in the pellet was solubilised in treatment with buffer A containing 0.5% sarkosyl (Sigma Aldrich) for 2 h at 4°C. After removing the insoluble materials by centrifugation at 30,000 × g for 30 min, the recombinant hTTN3 in the solubilised membrane fraction was purified using a Ni-NTA agarose (Qiagen, Germany) column. The fractions containing hTTN3 were mixed with 4 mg APG for 4 h at 4°C. The hTTN3 complexed with APG was further purified by size exclusion chromatography using a Superdex^™ 200 Increase 10/300 GL column (GE Healthcare).

Preparation of proteoliposome

Each 40 mg/mL of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) and 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine (POPE) stock solution was dissolved in chloroform at a 1:1 ratio. The mixture was dried on a glass vial surface spread out as wide as possible under gentle N₂ gas blowing for 30 min. The dried lipid film was rehydrated over 12 h in 140 mM KCl (Junsei, Japan) and 10 mM HEPES buffer to a final concentration of 10 mg/mL. The liposome solution was sonicated for 10 min with a 40% amplitude using a probe tip sonicator (Q125, Qsonica). Impurities in the liposomal solution were removed by centrifugation at 10,000 × g for 1 min. Clear milky liposome solution was mixed with the solution containing purified hTTN3, drTTN3 or MscL proteins at a 2:1 ratio, and incubated for more than 12 h at 22^oC. The final lipid-to-protein ratio was 1:20,000. To remove residual APG, the proteoliposome solution was filtered using 10 k molecular weight cut-off Amicon centrifugal filters (Merck Millipore) at 15,400 × g for 15 min. All prepared proteoliposomes were used within 12 h.

Formation of the lipid bilayer incorporated with hTTN3, drTTN3 or MscL

The lipid bilayer membrane was formed with 3% DPhPC dissolved in n-decane (MP Biomedicals) using the painting method. An aperture of 80–100 µm was created using a spark generator (EM-09A, DAEDALON) on a 10 µm thick polytetrafluoroethylene film. The polytetrafluoroethylene film was clamped between two 300-µL chambers covered with transparent glass on one side. Both chambers were then filled with physiological buffer (140 mM KCl, 10 mM HEPES) and the Ag/AgCl electrodes were placed in both chambers separately. Then 0.2 µL of the DPhPC solution was painted on the aperture using a gel loading pipette (Corning). The formation of the bilayer was confirmed by microscope and capacitance measurement. After the formation of the lipid bilayer, 10–20 µL of the prepared proteoliposome was added to the cis chamber. Purified hTTN3, drTTN3 or MscL proteins in the proteoliposome were successfully reconstituted into the spanned bilayer lipid membrane by fusion.

Single-channel currents of the lipid bilayers were recorded using a patch-clamp amplifier (Axopatch 200B; molecular device). The acquired signals were processed with Digidata 1440A (molecular device). Signal recording conditions included a 10 kHz software filter, 5 kHz lowpass filter and 50 kHz sampling rate with the appropriate voltage bias in all experiments.

Protein structure prediction

The monomeric structure of mTTN3 (UniProt ID: Q8C8S3) was predicted using the AlphaFold2 program (v2.1.2) ²⁰. The standard algorithm that was used consisted of two transformer-based modules for the sequence-residue and residue-residue pairwise features. The sequence-residue graph was built using the multiple sequence alignments (MSA) embedding with UniProt Reference Clusters 90 (UniRef90), Big Fantastic Database, Uniclust30, and MGnify databases (updated on 27 Nov 2021). The residue-residue graph contains information regarding the relationship between the residues. The MSA-based representation updates the residue-residue pair information through an element-wise outer product and sum over the MSA sequence dimension in the Evoformer block. Five initial models were created using different random seeds—while two models were generated using template structures, the others were generated without templates ²⁰. The quality of the predicted monomer structures was evaluated using the pLDDT metric ²⁰ applied to the Cα atoms’ distance differences between the predicted and target structures ³³ (Supplementary Fig. 4).

The monomer model (i.e., the one with the highest pLDDT value) was subsequently used as input into the multi-body interface implemented in the HADDOCK program (version 2.4) ^34,35,48 to generate the tetrameric structure. The computations were performed using the standard protocol where the centre of mass and non-crystallographic symmetry restraints were applied. The multi-body protocol consists of two main steps. In the first step, a set of 10,000 tetrameric models were generated via rigid-body docking followed by a refinement iteration of the 400 best-scored models introducing semi-flexible conditions. In the second step, only the 200 best-scored models obtained in the semi-flexible docking iteration were refined in the presence of an explicit water model. In this step the molecular mechanics potential energy as the sum of the individual energy terms such as van der Waals, electrostatic, desolvation, and restraints violation energies were used to evaluate the optimal orientation/conformation of the tetrameric structures. The obtained tetrameric structures were grouped into four types (types 1–4) according to the conformational similarity and two most relevant energy terms; van der Waals and electrostatic energies (Fig. 4b,c).

The model of the Type 1 group was further assessed with the ProQM program ⁴⁹. ProQM analysis showed that the highest accuracy for the model was the residues located at the six transmembrane helices (averaged ProQM score/residue: 0.695 ± 0.10) whereas the lowest accuracy was shown for the residues located at the C-terminus (0.282 ± 0.18). The score for the transmembrane α-helices in the tetrameric complex was significantly higher than those of the monomer (0.572 ± 0.10) indicating that the tetrameric configuration might be preferred over the monomeric.

MD simulations

The mTTN3 tetramer incorporated into the explicit bilayer model was simulated at the atomic level to obtain the molecular conformations of mTTN3 in the lipid membrane environment. The protonation states of the amino-acid residues of the mTTN3 tetramer were titrated at pH = 7.4 using PROPKA 3.1 ⁵⁰, and the hydrogen bond network of the tetrameric structure was optimised via PDB2PQR 2.1 ⁵¹. The protein was then embedded into the 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) bilayer using CHARMM-GUI ⁵². Each leaflet of the bilayer contained 350 DOPC. The system was fully hydrated with 130 TIP3P water per lipid ⁵³. Neutralising cations (four K⁺ ions) for the negative charge of the protein and additional KCl salt were added to achieve a 150 mM KCl bulk concentration. A molecular snapshot of the MD simulation system is shown in the Supplementary Fig. 5a. Each system was simulated for three replicas (R1, R2, R3) with different random seeds for the initial velocities.

Simulations were conducted using the OpenMM (version 7.6.0) ⁵⁴ program. The coordinates of all the atoms in the MD system were adjusted to minimise the energy and heated to 293.15 K over 125 ps followed by system equilibration for 1.75 ns. The production simulation was run using a Langevin thermostat at 298.15 K with a collision frequency of 1/ps. A Monte Carlo barostat ⁵⁵ specifically designed for membrane simulations was used. The pressure, normal to the membrane plane, was 1 atm, corresponding to the normal laboratory conditions, and the surface tension applied parallel to the plane at the water-lipid headgroup interface was set to 5 dyn/cm. This surface tension corresponded to the negative lateral pressure that is capable of activating mechanosensitive ion channels, such as MscL, by laterally stretching the bilayer ^56–58. The van der Waals term used a standard 6–12 LJ form with a force-switching function between 8 and 12 Å. Particle mesh Ewald was used for long-range electrostatic measurements beyond 12 Å. The integration time step equals 2 fs. Coordinate sets were saved for every 5 ps. The CHARMM C36 lipid ⁵⁹ and protein ^60–62 force fields were used. Three replicas were simulated with different random seeds for the initial velocities to ensure the convergence of the simulations.

Statistical analysis

All data presented in the figures represent the mean ± standard error of the mean. Student’s t-test was used to compare two means. One-way analysis of variance (ANOVA) followed by Tukey test was used to compare multiple means. P < 0.05 was considered as significant.

Acknowledgements

This research was supported by the National Research Foundation of Korea (NRF- 2020R1A3A300192913; UO, NRF-2020R1C1C101024513 and NRF-2022M3E5E801739512; G-SH, Mid-Career Research Program 2020R1A2C1101174; YY, and 2020R1A2C210036313 TK). UO and KH were funded by internal grant of Korea Institute of Science and Technology (2E32231 and 2Z06588, respectively). CF-F was funded by the NIDCD Intramural Research Funds (DC000039 to Thomas B. Friedman) and by the Program in Computational Brain Science and Health funded by the Palm Health Foundation.

Author contributions

S.P., T.L.N, S.L, and G-.S.H. performed TTN3 orthologs cloning, mutations of TTN3, and patch-clamp experiments; S.M.K. and Y.G.Y. performed TTN3 protein purification; H.R. performed electrophysiology in the lipid bilayer; T.S.K. designed and data analysis of electrophysiology; J.W., C.K., C.F-.F., and K.H used deep-learning algorithms for predicting TTN3 structure; J.N.W. and M.O.L. supervised pharmacology experiments; U.O. designed and supervised experiments and wrote the manuscript.

Competing interests: The authors declare no competing interests.

Materials & Correspondence: Correspondence and requests for materials should be addressed to Gyu-Sang Hong, Kyungreem Han, Tae Song Kim, and Uhtaek Oh.

Ranade, S. S., Syeda, R. & Patapoutian, A. Mechanically Activated Ion Channels. Neuron 87, 1162-1179, doi:10.1016/j.neuron.2015.08.032 (2015).
Wu, J., Lewis, A. H. & Grandl, J. Touch, Tension, and Transduction - The Function and Regulation of Piezo Ion Channels. Trends Biochem Sci 42, 57-71, doi:10.1016/j.tibs.2016.09.004 (2017).
Delmas, P., Hao, J. & Rodat-Despoix, L. Molecular mechanisms of mechanotransduction in mammalian sensory neurons. Nat Rev Neurosci 12, 139-153, doi:10.1038/nrn2993 (2011).
Hao, J. & Delmas, P. Multiple desensitization mechanisms of mechanotransducer channels shape firing of mechanosensory neurons. J Neurosci 30, 13384-13395, doi:10.1523/jneurosci.2926-10.2010 (2010).
Hong, G. S. et al. Tentonin 3/TMEM150c Confers Distinct Mechanosensitive Currents in Dorsal-Root Ganglion Neurons with Proprioceptive Function. Neuron 91, 708-710, doi:10.1016/j.neuron.2016.07.019 (2016).
Lu, H. J. et al. Tentonin 3/TMEM150C senses blood pressure changes in the aortic arch. J Clin Invest 130, 3671-3683, doi:10.1172/jci133798 (2020).
Coste, B. et al. Piezo1 and Piezo2 are essential components of distinct mechanically activated cation channels. Science 330, 55-60, doi:10.1126/science.1193270 (2010).
Hill, R. Z., Loud, M. C., Dubin, A. E., Peet, B. & Patapoutian, A. PIEZO1 transduces mechanical itch in mice. Nature 607, 104-110, doi:10.1038/s41586-022-04860-5 (2022).
Kefauver, J. M., Ward, A. B. & Patapoutian, A. Discoveries in structure and physiology of mechanically activated ion channels. Nature 587, 567-576, doi:10.1038/s41586-020-2933-1 (2020).
Delmas, P., Parpaite, T. & Coste, B. PIEZO channels and newcomers in the mammalian mechanosensitive ion channel family. Neuron 110, 2713-2727, doi:10.1016/j.neuron.2022.07.001 (2022).
Zhao, Q. et al. Structure and mechanogating mechanism of the Piezo1 channel. Nature 554, 487-492, doi:10.1038/nature25743 (2018).
Saotome, K. et al. Structure of the mechanically activated ion channel Piezo1. Nature 554, 481-486, doi:10.1038/nature25453 (2018).
Lu, Y. et al. The ion channel ASIC2 is required for baroreceptor and autonomic control of the circulation. Neuron 64, 885-897, doi:10.1016/j.neuron.2009.11.007 (2009).
Wee, J. et al. Tentonin 3/TMEM150C regulates glucose-stimulated insulin secretion in pancreatic β-cells. Cell Rep 37, 110067, doi:10.1016/j.celrep.2021.110067 (2021).
Ojeda-Alonso, J. et al. Lack of evidence for participation of TMEM150C in sensory mechanotransduction. J Gen Physiol 154, doi:10.1085/jgp.202213098 (2022).
Dubin, A. E. et al. Endogenous Piezo1 Can Confound Mechanically Activated Channel Identification and Characterization. Neuron 94, 266-270 e263, doi:10.1016/j.neuron.2017.03.039 (2017).
Anderson, E. O., Schneider, E. R., Matson, J. D., Gracheva, E. O. & Bagriantsev, S. N. TMEM150C/Tentonin3 Is a Regulator of Mechano-gated Ion Channels. Cell Rep 23, 701-708, doi:10.1016/j.celrep.2018.03.094 (2018).
Baek, M. & Baker, D. Deep learning and protein structure modeling. Nat Methods 19, 13-14, doi:10.1038/s41592-021-01360-8 (2022).
Callaway, E. 'The entire protein universe': AI predicts shape of nearly every known protein. Nature 608, 15-16, doi:10.1038/d41586-022-02083-2 (2022).
Jumper, J. et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583-589, doi:10.1038/s41586-021-03819-2 (2021).
Baek, M. et al. Accurate prediction of protein structures and interactions using a three-track neural network. Science 373, 871-876, doi:10.1126/science.abj8754 (2021).
Tunyasuvunakool, K. et al. Highly accurate protein structure prediction for the human proteome. Nature 596, 590-596, doi:10.1038/s41586-021-03828-1 (2021).
McCarter, G. C., Reichling, D. B. & Levine, J. D. Mechanical transduction by rat dorsal root ganglion neurons in vitro. Neurosci Lett 273, 179-182, doi:10.1016/s0304-3940(99)00665-5 (1999).
Han, S. G. et al. Preparation of functional human lysophosphatidic acid receptor 2 using a P9(∗) expression system and an amphipathic polymer and investigation of its in vitro binding preference to G(α) proteins. Biochem Biophys Res Commun 487, 103-108, doi:10.1016/j.bbrc.2017.04.025 (2017).
Choi, S., Yoon, S., Ryu, H., Kim, S. M. & Jeon, T. J. Automated Lipid Bilayer Membrane Formation Using a Polydimethylsiloxane Thin Film. J Vis Exp, 54258, doi:10.3791/54258 (2016).
Häse, C. C., Le Dain, A. C. & Martinac, B. Purification and functional reconstitution of the recombinant large mechanosensitive ion channel (MscL) of Escherichia coli. J Biol Chem 270, 18329-18334, doi:10.1074/jbc.270.31.18329 (1995).
Drew, L. J. et al. High-threshold mechanosensitive ion channels blocked by a novel conopeptide mediate pressure-evoked pain. PLoS One 2, e515, doi:10.1371/journal.pone.0000515 (2007).
Wang, Y. et al. A lever-like transduction pathway for long-distance chemical- and mechano-gating of the mechanosensitive Piezo1 channel. Nat Commun 9, 1300, doi:10.1038/s41467-018-03570-9 (2018).
Syeda, R. et al. Chemical activation of the mechanotransduction channel Piezo1. Elife 4, doi:10.7554/eLife.07369 (2015).
Botello-Smith, W. M. et al. A mechanism for the activation of the mechanosensitive Piezo1 channel by the small molecule Yoda1. Nat Commun 10, 4503, doi:10.1038/s41467-019-12501-1 (2019).
Pereira, J. et al. High-accuracy protein structure prediction in CASP14. Proteins 89, 1687-1699, doi:10.1002/prot.26171 (2021).
Del Alamo, D., Sala, D., McHaourab, H. S. & Meiler, J. Sampling alternative conformational states of transporters and receptors with AlphaFold2. Elife 11, doi:10.7554/eLife.75751 (2022).
Mariani, V., Biasini, M., Barbato, A. & Schwede, T. lDDT: a local superposition-free score for comparing protein structures and models using distance difference tests. Bioinformatics 29, 2722-2728, doi:10.1093/bioinformatics/btt473 (2013).
Honorato, R. V. et al. Structural Biology in the Clouds: The WeNMR-EOSC Ecosystem. Front Mol Biosci 8, 729513, doi:10.3389/fmolb.2021.729513 (2021).
van Zundert, G. C. P. et al. The HADDOCK2.2 Web Server: User-Friendly Integrative Modeling of Biomolecular Complexes. J Mol Biol 428, 720-725, doi:10.1016/j.jmb.2015.09.014 (2016).
Karaca, E., Melquiond, A. S. J., de Vries, S. J., Kastritis, P. L. & Bonvin, A. M. J. J. Building Macromolecular Assemblies by Information-driven Docking. Mol Cell Proteom 9, 1784-1794, doi:https://doi.org/10.1074/mcp.M000051-MCP201 (2010).
de Vries, S. J. et al. HADDOCK versus HADDOCK: new features and performance of HADDOCK2.0 on the CAPRI targets. Proteins 69, 726-733, doi:10.1002/prot.21723 (2007).
Smart, O. S., Neduvelil, J. G., Wang, X., Wallace, B. A. & Sansom, M. S. HOLE: a program for the analysis of the pore dimensions of ion channel structural models. J Mol Graph 14, 354-360, 376, doi:10.1016/s0263-7855(97)00009-x (1996).
Beaulieu-Laroche, L. et al. TACAN Is an Ion Channel Involved in Sensing Mechanical Pain. Cell 180, 956-967 e917, doi:10.1016/j.cell.2020.01.033 (2020).
Hong, G. S., Lee, B. & Oh, U. Evidence for Mechanosensitive Channel Activity of Tentonin 3/TMEM150C. Neuron 94, 271-273.e272, doi:10.1016/j.neuron.2017.03.038 (2017).
McHugh, B. J. et al. Integrin activation by Fam38A uses a novel mechanism of R-Ras targeting to the endoplasmic reticulum. J Cell Sci 123, 51-61, doi:10.1242/jcs.056424 (2010).
Ranade, S. S. et al. Piezo1, a mechanically activated ion channel, is required for vascular development in mice. Proc Natl Acad Sci U S A 111, 10347-10352, doi:10.1073/pnas.1409233111 (2014).
Holzinger, A. & Meindl, U. Jasplakinolide, a novel actin targeting peptide, inhibits cell growth and induces actin filament polymerization in the green alga Micrasterias. Cell Motil Cytoskeleton 38, 365-372, doi:10.1002/(sici)1097-0169(1997)38:4<365::Aid-cm6>3.0.Co;2-2 (1997).
Bubb, M. R., Senderowicz, A. M., Sausville, E. A., Duncan, K. L. & Korn, E. D. Jasplakinolide, a cytotoxic natural product, induces actin polymerization and competitively inhibits the binding of phalloidin to F-actin. J Biol Chem 269, 14869-14871 (1994).
Diver, M. M., Lin King, J. V., Julius, D. & Cheng, Y. Sensory TRP Channels in Three Dimensions. Annu Rev Biochem 91, 629-649, doi:10.1146/annurev-biochem-032620-105738 (2022).
Tominaga, M. in TRP Ion Channel Function in Sensory Transduction and Cellular Signaling Cascades (eds W. B. Liedtke & S. Heller) (CRC Press/Taylor & Francis Copyright © 2007, Taylor & Francis Group, LLC., 2007).
Hu, J. & Lewin, G. R. Mechanosensitive currents in the neurites of cultured mouse sensory neurones. J Physiol 577, 815-828, doi:10.1113/jphysiol.2006.117648 (2006).
Karaca, E., Melquiond, A. S., de Vries, S. J., Kastritis, P. L. & Bonvin, A. M. Building macromolecular assemblies by information-driven docking: introducing the HADDOCK multibody docking server. Mol Cell Proteomics 9, 1784-1794, doi:10.1074/mcp.M000051-MCP201 (2010).
Ray, A., Lindahl, E. & Wallner, B. Model quality assessment for membrane proteins. Bioinformatics 26, 3067-3074, doi:10.1093/bioinformatics/btq581 (2010).
Olsson, M. H. M., Søndergaard, C. R., Rostkowski, M. & Jensen, J. H. PROPKA3: Consistent Treatment of Internal and Surface Residues in Empirical pKa Predictions. J Chem Theor and Comput 7, 525-537, doi:10.1021/ct100578z (2011).
Dolinsky, T. J. et al. PDB2PQR: expanding and upgrading automated preparation of biomolecular structures for molecular simulations. Nuc Acids Res 35, W522-W525, doi:10.1093/nar/gkm276 (2007).
Jo, S., Kim, T., Iyer, V. G. & Im, W. CHARMM-GUI: A web-based graphical user interface for CHARMM. J Comput Chem 29, 1859-1865, doi:https://doi.org/10.1002/jcc.20945 (2008).
Jorgensen, W. L., Chandrasekhar, J., Madura, J. D., Impey, R. W. & Klein, M. L. Comparison of simple potential functions for simulating liquid water. J Chem Physics 79, 926-935, doi:10.1063/1.445869 (1983).
Eastman, P. et al. OpenMM 7: Rapid development of high performance algorithms for molecular dynamics. PLOS Comput Biol 13, e1005659, doi:10.1371/journal.pcbi.1005659 (2017).
Chow, K.-H. & Ferguson, D. M. Isothermal-isobaric molecular dynamics simulations with Monte Carlo volume sampling. Comp Phys Coms 91, 283-289, doi:https://doi.org/10.1016/0010-4655(95)00059-O (1995).
Sukharev, S. I., Blount, P., Martinac, B., Blattner, F. R. & Kung, C. A large-conductance mechanosensitive channel in E. coli encoded by mscL alone. Nature 368, 265-268, doi:10.1038/368265a0 (1994).
Roux, B. Commentary: surface tension of biomembranes. Biophys J 71, 1346-1347, doi:10.1016/s0006-3495(96)79335-9 (1996).
Perozo, E., Cortes, D. M., Sompornpisut, P., Kloda, A. & Martinac, B. Open channel structure of MscL and the gating mechanism of mechanosensitive channels. Nature 418, 942-948, doi:10.1038/nature00992 (2002).
Klauda, J. B. et al. Update of the CHARMM All-Atom Additive Force Field for Lipids: Validation on Six Lipid Types. J Phys Chem B 114, 7830-7843, doi:10.1021/jp101759q (2010).
Huang, J. & MacKerell Jr, A. D. CHARMM36 all-atom additive protein force field: Validation based on comparison to NMR data. J Comput Chem 34, 2135-2145, doi:10.1002/jcc.23354 (2013).
Best, R. B. et al. Optimization of the Additive CHARMM All-Atom Protein Force Field Targeting Improved Sampling of the Backbone ϕ, ψ and Side-Chain χ1 and χ2 Dihedral Angles. J Chem Theor Comput 8, 3257-3273, doi:10.1021/ct300400x (2012).
Huang, J. et al. CHARMM36m: an improved force field for folded and intrinsically disordered proteins. Nat Methods 14, 71-73, doi:10.1038/nmeth.4067 (2017).

SupplementaryVideo1.mp4
Movie showing the tetrameric structure of mTTN3.
SupplementaryData.docx

Download PDF

Journal Publication

published 31 May, 2024

Read the published version in Cell Reports →

Version 1

posted

You are reading this latest preprint version

Tentonin 3 is a pore-forming subunit of a slow-inactivation mechanosensitive channel

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Results

Mechanosensitivity of TTN3 is conserved throughout the vertebrate phyla

TTN3 in the lipid bilayer elicits spontaneous and stretch-activated channel currents

TTN3 mechanosensitivity is separate from Piezo1

mTTN3 tetrameric structures predicted by deep learning-based algorithms

Predicted molecular structure of TTN3

N- and C-termini of TTN3 contribute to mechanosensitivity

Discussion

Methods

Cell culture and transfection

Gene preparation and mutagenesis

Whole cell current recording

Mechanical stimulation

Preparation of proteoliposome

Formation of the lipid bilayer incorporated with hTTN3, drTTN3 or MscL

Protein structure prediction

MD simulations

Statistical analysis

Declarations

References

Supplementary Files

Status:

Journal Publication

Version 1