Reduced and PC-free mTRPV3-Y564A had the biggest Grid 13 in the Pre-S1/TRP interface for a calculated Tm 38°C
The cryo-EM structures of both closed and open states in detergent-solubilized PC-free mTRPV3-Y564A were first sampled at 37°C after heat sensitization. Therefore, it is necessary to examine if the release of the phosphatidylcholine (PC) lipid from the vanilloid site by the Y546A mutation is responsible for the lower experimental temperature threshold and sensitivity [22].
The previous chimera studies between rat TRPV1 (rTRPV1) and mTRPV3 indicated that the pre-S1 segment 358–434 plays a critical role in mediating the temperature threshold and sensitivity Q10 [30]. On the other hand, the chimera investigations between heat-sensing TRPV1 and cold-sensing TRP melastatin 8 (TRPM8) showed that the C-terminal including the TRP domain (693–710) is required for the polarity of thermal sensitivity [31]. In this regard, the segment from D396 in the pre-S1 domain to K705 in the TRP domain should be at least included as the necessary gating pathway for the temperature threshold and sensitivity and the systemic thermal instability. Along such a gating pathway, the diversity of non-covalent interactions between amino acid side chains in the closed and PC-free Y564 mutant was found after heat sensitization (Fig. 1A). They included 9 H-bonds emerged between different hydrophilic residues, twenty-six π interactions between aromatic residues and nearby residues, and 3 salt bridges between several charged pairs (Fig. 1A, Table S1).
When these non-covalent interactions formed a systematic grid-like non-covalent interaction mesh network, the total non-covalent interactions and grid sizes along the gating pathway from D396 to K705 were 38 and 77, respectively (Fig. 1A). Thus, the systemic thermal instability (Ti) was 2.03 (Table 1). Meanwhile, in addition to the smallest grid with a 0-residue size in the VSLD, the biggest Grid13 with a 13-residue size appeared in the pre-S1/TRP interface via the shortest path from D396 to Y409, R698, R696, W433, K432 and back to D396 to control the D396-K432 salt bridge (Fig. 1B-D). When 1.0 equivalent H-bond sealed this grid, the removal of the PC lipid from the vanilloid site by the Y564A mutation allowed a calculated Tm of 38°C near the experimental Tm 37°C (Table 1) [22].
The melting of the biggest Grid13 at 37°C initiated channel opening of reduced and PC-free mTRPV3-Y564A with a low Ω10 comparable to the low Q10
When the mTRPV3-Y564A mutant opened with the melting of Grid13 in the TRP/pre-S1 interface at 37°C, the disruption of the D396-K432 salt bridge triggered several changes in the systematic grid-like non-covalent interaction mesh network. In addition to one salt bridge, three H-bonds and 17 π interactions were conserved, two salt bridges, six H-bonds and eight π interactions were replaced with three new salt bridges, six new H-bonds and eight new π interactions (Tables S1 and S2). Thus, the total non-covalent interactions along the gating pathway from D396 to K705 had only a minor change from 38 to 39 (Figs. 1A and 2A).
Table 1 The grid thermodynamic model-based new parameters of the mTRPV3 bio-thermometer along the gating pathway from D396 to K705
Construct
|
WT mTRPV3
|
mTRPV3-Y564A
|
PDB ID
|
6LGP
|
7MIO
|
7MIN
|
6PVP
|
6PVO
|
Lipid PC at the vanilloid site
|
bound
|
free
|
bound
|
free
|
free
|
Redox state
|
reduced
|
oxidized
|
oxidized
|
reduced
|
reduced
|
Lipid environment
|
MSP2N2
|
cNW11
|
cNW11
|
detergent
|
Sampling temperature, °C
|
4
|
42
|
42
|
37
|
37
|
Gating state
|
Closed ↔ Open ↔ Sensitized
|
Open ↔Sensitized
|
# of the biggest grid
|
Grid12
|
Grid9
|
Grid17
|
Grid11
|
Grid13
|
Biggest grid size (Smax)
|
12
|
9
|
17
|
11
|
13
|
Equivalent H-bonds in Smax
|
2.0
|
2.0
|
2.0
|
2.0
|
1.0
|
Total non-covalent interactions
|
55
|
52
|
59
|
39
|
38
|
Total grid sizes, a.a
|
96
|
65
|
72
|
74
|
77
|
Calculated Tm, °C
|
50
|
56
|
40
|
52
|
38
|
Measured Tm, °C
|
|
|
42
|
|
37
|
Measured Tth, °C
|
52
|
57
|
32-39
|
|
|
Systemic thermal instability (Ti)
|
1.75
|
1.25
|
1.22
|
1.90
|
2.03
|
Calculated Ω10, min at Emin=0.5 kJ/mol
|
8.76
|
|
1.88
|
|
0.76
|
Calculated Ω10, mean at Emean=1.0 kJ/mol
|
18.3
|
|
4.12
|
|
1.48
|
Calculated Ω 10, max at Emax=3.0 kJ/mol
|
58.5
|
|
14.3
|
|
4.30
|
Measured Q 10,
|
16.4-22.6
|
|
1.9-3.1
|
|
1.21
|
Ref. for measured Tth or Q10
|
[19]
|
[19]
|
[19]
|
|
[22]
|
On the other hand, the smallest Grid0 was conserved with a zero-residue size via the shortest path from F441 to Y565, Y448, F449, F445 and back to F441 (Figs. 1A and 2A). Thus, it may serve as a thermostable anchor against which two smaller grids may favor channel opening. One was Grid4 with a 4-residue size via the shortest path from D519 to W521, V525, F626, Y565, R567, Q695, R698 and back to D519 in the VSLD/TRP interface (Figs. 2A-B, E); the other was the biggest Grid11 in the TRP/VSLD/pre-S1 interfaces to control the H417-E689 π interaction. It had an 11-residue size via the shortest path from T411 to H417, E689, W692, R696, R698, D519, S515, and back to T411 (Figs. 2A, 2C-E). Once 2 equivalent H-bonds sealed the grid, the calculated Tm was about 52°C (Table 1).
In any way, the disruption of the D396-K432 salt bridge in the biggest Grid13 induced a global conformational change from the pre-S1 domain to the VSLD, the TRP domain, the S4-55 linker and the pore domain (Figs. 1A and 2A). However, the grid sizes along the gating pathway from D396 to K705 had only a minor change from 77 to 74 (Figs. 1A and 2A). In this case, the systemic thermal instability (Ti) was 1.90, and the calculated Ω10 was in a range from 0.76 to 4.30 and with a mean value 1.48, which was comparable to the experimental Q10 (~ 1.21) (Table 1) [22]. In other words, the removal of the PC lipid from the vanilloid site by the Y564A mutation allowed reduced mTRPV3 to have the very low structural and functional thermo-sensitivities.
On the other hand, oxidized mTRPV3 with a disulfide bond between C612 and C619 in the outer pore has also been reported to open from a PC-bound closed state at a lower threshold 42°C after repeated heat sensitization from 25°C to 40°C [23]. Therefore, it is exciting to test if oxidation also allows a low structural temperature sensitivity Ω10 to be responsible for the measured functional temperature sensitivity Q10 (1.9–3.1) [19].
Closed PC-bound mTRPV3 with the disulfide bond in the outer pore had the biggest Grid17 in the Pre-S1/VSLD interface for a calculated Tm 40°C after heat sensitization
Regarding the PC-bound closed state of oxidized mTRPV3 after heat sensitization, much more non covalent interactions than those in the PC-free closed state of reduced mTRPV3-Y564A shaped a distinct systematic fluidic grid-like non-covalent interaction mesh network (Figs. 1A and 3A). In the presence of the C612-C619 disulfide bond in the pore domain, four salt bridges (E610-K614 was merged into the C612-C619 disulfide bond), fifteen H-bonds and forty π interactions were identified between D396 and K705 (Fig. 3A, Table S3). Since the total non-covalent interactions and grid sizes along the gating pathway from D396 in the pre S1 domain to K705 in the TRP domain were 59 and 72, respectively (Fig. 3A), the grid-based systemic thermal instability (Ti) was about 1.22 (Table 1). Despite several smallest grids with a zero-residue size, the biggest Grid17 with a 17-residue size was outstanding in the VSLD/pre-S1 interface to control the D519-R416 salt bridge (Fig. 3B-D). It started with D519 and went through W521, F522, Y564, Y565, F441, W433 and ended with R416 (Fig. 3E). When two equivalent H-bonds sealed the grid, the predicted Tm was about 40°C (Table 1), which was close to the measured Tm 42°C. [23]
The melting of the biggest Grid17 at 42°C drove oxidized mTRPV3 opening with a low Ω10 comparable to the low Q10
In the heat-activated open state, following the melting of the R416-D517 salt bridge in the biggest Grid17 at 42°C as predicted (Fig. 3D) [23], although one salt bridge, five H-bonds, and 35 π interactions were conserved, three salt bridges, ten H-bonds, and six π interactions were substituted by two new salt bridges, eight new H-bonds, and one new π interaction (Figs. 3A and 4A, Tables S3 and S4). However, two smallest Grid0 with a zero-residue size were still conserved as anchors near the R416-D519 salt bridge: one via the shortest path from F445 to Y565, Y448, F449, and back to F445, and the other via the shortest path from Y448 to Y565, Y564, F526 and back to Y448 (Figs. 3A and 4A). Therefore, the following gating pathway against these two anchors was proposed.
First, in the VSLD/pre-S1/TRP interfaces, the D519-R416 salt bridge was substituted by the T411-R416 and D519-R567 H-bonds. As a result, the T397-E704 H-bond and the D396-K432 salt bridge were broken with the formation of H417/E418-R690 and E423-T427 H-bonds. In addition, the K432-E704 salt bridge became an H-bond (Fig. 4A).
Second, when R567 H-bonded with D519 in the VSLD, a smaller Grid2 with a 2-residue size appeared via the shortest path from D519 to W521, F522, Y564, Y565, R567 and back to D519. Consequently, the V528-F524 π interaction was disconnected (Fig. 4A).
Third, when the conformational wave extended to the S4-S5/TRP interface, the R567-T699 H-bond was disrupted and the E689-R693 H-bond changed to a salt bridge (Fig. 4A).
Fourth, when this conformational wave continued to the pore domain, the F597-F601 π interaction and the N647-E610/K614 and E682-K686 H-bonds and the E610-K614 salt bridge were disconnected. In the meanwhile, the E631-K634 H-bond and the F633-I637 π interaction were present, and the H-bond moved from S621-Q646 to S620-Q646 (Fig. 4A).
Taking all these changes into account, after the biggest Grid17 in the VSLD/pre-S1 interface melted above the predicted 40°C, the PC lipid was released from nearby W521 and Q695 and thus the new biggest Grid9 with a 9-residue size was created in the S5-S6 interface, which may be required for channel opening (Fig. 4B-C). When two equivalent H-bonds sealed Grid9 via the shortest path from D586 to F590 and L673 and T680 and back to D586 (Figs. 4C, 4E), the calculated Tm was about 56°C (Table 1). That may be why the temperature limit is 57°C for stable efficacy [19]. Since Grid9, together with Grid7 with a 7-residue size via the shortest path from F590 to Y594, T636, Y661, T665, L673 and back to F590, was conserved in both closed and open states of oxidized mTRPV3 (Figs. 3A and 4A), they may act as thermostable anchors to secure channel activity. In the meanwhile, a smaller Grid3 with a 3-residue size in the pre-S1/VSLD/S4-S5 linker/TRP/pre-S1 interfaces may be required to stimulate the lower state of the channel. It linked multiple active residues together including W433, F441, Y565, Y564, F522, W521, D519, R567, Q570, W692, and R696 (Figs. 4D, 4E).
As a result, the total non-covalent interactions and grid sizes along the gating pathway from D396 to K705 decreased from 59 and 72 to 52 and 65, respectively (Figs. 3A & 4A). Such a decrease produced a low systemic thermal instability (Ti) value as 1.25. More importantly, the systematic structural thermo-sensitivity Ω10 was in a range from1.88 to 14.3 and with a mean value 4.12 (Table 1), which was close to the experimental Q10 (1.9–3.1) [19]. Therefore, even if the PC lipid at the corresponding vanilloid site was not released, the presence of the C612-C619 disulfide bond in the outer pore may be adequate for mTRPV3 to open with both low Tm and Ω10 to match the measured Tth and Q10 in response to the second heat stimulation [19]. In that regard, it is attractive to test if the disruption of the C612-C619 disulfide bond can increase the Tm and the Ω10 upon channel opening from reduced mTRPV3 to meet the requirement of the higher Tth (> 50 ⁰C) and Ω10 (16.4–22.6) [19].
The melting of the biggest Grid12 at the vanilloid PC site above Tm 50°C was required to release PC from reduced mTRPV3 for channel opening with a high Ω10
In the absence of the C612-C619 disulfide bond, reduced mTRPV3 had some different noncovalent interactions to form the systematic grid-like non-covalent interaction mesh network at 4°C when compared with the closed and oxidized one at 42°C after heat sensitization (Figs. 4A & 5A, Table S5) [32]. In the pore domain, after the E610-K614 salt bridge and the E610-N647 and S621-Q646 and E682-K686 H-bonds were disrupted, the F625-V629 and F633-I637 and T649-Y650 π interactions and the Y594-Y661 H-bond emerged (Fig. 5A).
When this conformational change extended to the S4-S5 linker/TRP interface, the R567-T699 H-bond and the R698-E702 salt bridge were broken but the Q570-E689 H-bond replaced the Q570-W692 CH-π interaction. As a consequence, along with the T566-S576 H-bond in the S4-S5 linker/VSLD interface, D519 H-bonded with T411 in the VSLD/pre S1 interface and formed an additional salt bridge with R698 in the VSLD/TRP interface (Fig. 5A).
When this conformational change extended to the VSLD, the T456-W559 H-bond was broken, the H471-Y540/Y547 π interaction network changed to the Y448/Y551-Q529 H-bonds, the π interaction moved from F542-Y544 to Y540-Y547, and the H-bond shifted from K500-E501 to Q514-S518. When the PC bridge moved from W521-PC-Q695 to W521-PC-F524/R567 (Figs. 4B & 5B), in addition to a salt bridge between R567 and the PC lipid and the CH-π interactions of the PC lipid with W521 and F524, Y564 formed a CH-π interaction with R567 (Fig. 5A).
By all account, the disruption of the C612-C619 disulfide bond brought about the biggest Grid12 at the vanilloid PC site (Fig. 5B). When two equivalent H-bonds governed the 12-atom path from W521 to PC to F524 (Figs. 5C-D), the predicted melting temperature was about 50°C (Table 1), which was close to the initial experimental Tth 52°C for TRPV3 opening [19]. On the other hand, when compared with oxidized mTRPV3 in both closed and open states, only four H-bonds and 26 π interactions were conserved, and two new salt bridges and four new H-bonds and six new π interactions were added (Tables S3, S4 and S5). As the total non-covalent interactions and grid sizes along the gating pathway from D396 to K705 were 55 and 96, respectively (Fig. 5A), the systemic thermal instability (Ti) was 1.75 (Table 1). When the same open state as shown in the oxidized and PC-free mTRPV3 was employed (Fig. 4A), the melting of the W521-PC-F524 bridge in Grid12 would produce the calculated Ω10 ranging from 8.76 to 58.5 with a mean value 18.3, which was approximate with the experimental Q10 (16.4–22.6) (Table 1) [19]. Thus, the initial high Tth and Q10 of mTRPV3 upon the brief heat stimulation may result from the gating transition from the reduced and closed state to the open and oxidized one.