The structural basis of the divalent cation blocking on tetrameric cation channel

doi:10.21203/rs.3.rs-2252854/v1

Download PDF

Article

The structural basis of the divalent cation blocking on tetrameric cation channel

https://doi.org/10.21203/rs.3.rs-2252854/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 15 Jul, 2023

Read the published version in Nature Communications →

Version 1

posted

You are reading this latest preprint version

Divalent cation blocking is observed in various tetrameric ion channels. For the blocking, a divalent cation is thought to stack in the ion pathway of the channel, but this has not yet been directly observed, so the blocking mechanism by these small divalent cations remains uncertain. Here, we elucidated the divalent cation blocking mechanism by reproducing the blocking effect into NavAb, a well-studied tetrameric sodium channel. Our crystal structures of NavAb mutants showed that the mutations increasing the hydrophilicity of the inner vestibule of the pore domain enable a divalent cation to stack on the ion pathway. Furthermore, molecular dynamics simulation showed that the stacking calcium ion repels the sodium ions at the bottom of the selectivity filter. These results suggest the primary mechanism of the divalent cation block in biologically essential channels.

Biological sciences/Biophysics/Computational biophysics

Biological sciences/Structural biology/X-ray crystallography/Nanocrystallography

Biological sciences/Biophysics/Membrane biophysics/Ion transport

Neural activity controls critical biological responses, such as thinking, memory formation, and muscle contraction. The transmission of neural activity occurs through various cation-selective channels on the nerve cell membrane ¹. Divalent cation blocking is observed in important tetrameric cation channels. For example, magnesium ions block the current of the NMDA receptor (NMDAR) in a voltage-dependent manner (Fig. 1a) ². AMPA receptor (AMPAR), an ionotropic receptor similar to NMDAR, alters calcium ion selectivity by RNA editing (Fig. 1a) ³. Calcium ions also block the current of voltage-dependent calcium channels and IP3 receptors, which are known as calcium facilitation ⁴. The pore domain of the tetrameric cation channels consists of two transmembrane helices, and the selectivity filter (SF) locates in the loop between them. The loop also contains one or two pore helices (P1 and P2 helix), and the SF follows the P1 helix (Fig. 1b). A simulation study suggested the direct interaction between the divalent cation and the SF causes divalent cation blocking ⁵. As the name indicates, the SF of the tetrameric cation channels is responsible for the selective permeation of cations. It also concerns the activity control of the channels, such as the C-type inactivation ⁶. For NMDAR, functional analyses and simulations suggest that the residues at the SF (Asn616 of GluN1 subunit, Asn615, Asn616, and Val618 of GluN2B subunit) are involved in magnesium inhibition (Fig. 1d) ⁵. Another study reports RNA editing Gln586 at AMPAR SF into arginine inhibits calcium permeation (Fig. 1e) ⁷. While recent studies revealed the structures of these receptors in various states ^8–10, the detailed molecular mechanism of divalent cation blocking on the selectivity filter is not fully understood.

The divalent cation blocking is not observed in any tetrameric channels including typical voltage-gated sodium channels (Navs), whose structures have been well studied. Therefore, reproducing the blocking on Nav and its structural analysis would help elucidate their molecular mechanisms. Prokaryotic Navs (BacNavs) were cloned from bacteria living in various environments and characterized ^11–13, which provided many insights into the molecular basis of sodium channels ^14–16. However, BacNavs did not show divalent cation blocking. Nevertheless, an ancestor-like channel group of BacNav (AnclNav) was recently identified, and one homolog of AnclNav had similar divalent cation blocking ¹⁷. Therefore, BacNavs are good targets for reproducing divalent cation blocking. NavAb, one of BacNav cloned from Arcobacter butzleri, is the most critical contributor to the first full-length structure of Nav at atomic resolution ^18,19. It is a helpful template for understanding disease mechanisms, toxin activity, and human drug discovery targets ^20–22. While the inserted direction of the transmembrane region of NMDAR and AMPAR is the opposite of that of other tetrameric cation channels, the structures of the pore domain of these channels are very similar (Fig. 1a and c). Superimposing NavAb on the two receptors aligned with the P1 helix, Leu176 of NavAb overlaps with the residues involved in the functional inhibition by divalent cations (Fig. 1d and e). By introducing some single mutations into Leu176 of the SF, we successfully reproduced the divalent cation-blocking effect on NavAb (Fig. 1f-h). Using these mutated NavAb, we conducted structural analysis and molecular dynamics simulation to elucidate mechanisms of the divalent cation blocking from structural point of view.

Calcium blocking on Leu176 mutants

N49K mutation of NavAb, introduced in this study, provides a stable current ²³. The current of NavAb N49K mutant was not attenuated even if the calcium concentration increased from 2.5 mM to 40 mM in all depolarizing membrane potential (Fig. 1f and Sup. Figure 1 a). To reproduce the divalent cation-blocking effect on NavAb N49K, we introduced L176N (L176N^NK) and L176Q (L176Q^NK) mutations which imitate the selectivity filter of NMDAR and AMPAR, respectively. They marked a significant decrease in sodium current according to extracellular calcium concentration. (Fig. 1g, and h). Unlike the case of NMDAR ², these calcium-blocking effects were observed in all depolarizing membrane potentials (Sup. Figure 1d, and g). The positive shift of the reversal potential was small when extracellular calcium ion concentration increased (Sup. Figure 1b, e, and h). It indicates the low calcium permeability of L176N^NK and L176Q^NK mutants, as well as the calcium ion selectivity against sodium ion (P_Ca/P_Na) of BacNavs ^13,18,24. Under higher extracellular calcium ion concentrations, the voltage dependency of activation was slightly shifted positively (Sup. Figure 1c, f, i), as in the case of NavPp ¹⁷. These results suggested that these mutations represent a developmental process of divalent cation blocking that occurred early in the evolution of the channel.

We introduced several single mutations into the 176th residue and evaluated the effect on divalent cation blocking (Fig. 2f). L176T^NK mutant, a hydrophilic mutant like L176Q^NK and L176N^NK mutants, also showed calcium blocking (Fig. 2a and Sup. Figure 2a-c). The blocking extent of calcium ions on the L176T^NK mutant is similar to those of L176Q^NK and L176N^NK mutants (Fig. 2f). Next, we investigated the effect of the side-chain size. The L176A^NK mutant showed the strongest calcium blocking (Fig. 2b and Sup. Figure 3a-c). The current of the L176A^NK mutant almost disappeared in the 40 mM extracellular calcium condition. Smaller side chain mutant, L176G^NK, showed only a half reduction of current against L176A^NK in 40 mM extracellular calcium condition (Fig. 2c and Sup. Figure 3d-f). L176V^NK showed little current reduction (Fig. 2d, and Sup. Figure 4a-c). On the other hand, a large side chain mutant, L176F^NK, showed no current reduction (Fig. 2e, and Sup. Figure 4d-f). The calcium-blocking mutants also showed the current blocking in each magnesium and strontium condition (Sup. Figure 5). Therefore, this blocking effect would be common to divalent cations.

These results suggested that the increment of the hydrophobicity and bulkiness of the 176th residue does not contribute to divalent cation blocking. Contrarily, small or hydrophilic mutations at the 176th residue introduce the divalent cation blocking into NavAb^NK. Therefore, we hypothesized these mutations cause new interactions with divalent cations at the bottom of the selective filter. To elucidate this mechanism, we proceeded structural analyses of these mutants.

The structures of L176Q^NK and L176G^NK mutants

To investigate the structural changes of the calcium-blocking mutants, we crystallized and determined the crystal structures of L176Q^NK, L176G^NK, and N49K mutants with and without calcium ions at resolutions of approximately 3Å. Because the solubilized proteins of L176N^NK, L176T^NK, and L176A^NK mutants were unstable, we could not obtain their well-diffracted crystals. The electron density fitted well to the mutated side chain of the 176th residue of L176Q^NK and L176G^NK mutants (Fig. 3a-c). NavAb has three ion interaction sites in the selectivity filter, a high-energy-field site (Site_HFS), a center site (Site_CEN), and an inner site (Site_IN) (Fig. 3d) ¹⁸. For all three mutants, the electron density of Site_IN and Site_CEN increased in the calcium condition (Fig. 3d-i: blue mesh).

For evaluating the electron density, we compared difference maps between structures of each mutant with and without calcium ions (Fig. 3e, g, and i: green mesh). In the ionic pathway of the N49K mutant, there is no differential density of the calcium ion condition subtracted with the electron density of the no calcium ion condition (Fig. 3e). In the L176Q^NK mutant, significant increases of differential density are observed in Site_Cen and Site_In. These increased densities seem to cause the current inhibition in the presence of calcium ions. The amine group of the mutated Q176 side chain forms two weak hydrogen bonds with the main chain of Val173 and Met174 of the P1 helix (Fig. 3b, f, and g). These interactions make the carboxyl group of the Q176 side chain face the center of the ion pathway (Fig. 3f and g: middle and bottom). Therefore, the Q176 side chain is responsible for the increase in electron density under calcium conditions. It is also suggested by comparing the electron density of the L176Q^NK and N49K mutants for the same ionic condition (Fig. 3f and g middle and bottom: brown mesh). In the non-calcium condition, the differential density of Site_IN of the L176Q^NK mutant was stronger than that of the NavAb N49K mutant (Fig. 3f: middle and bottom). The differential density at Site_IN of the L176Q^NK mutant seems to interact with two oxygen atoms of the carboxyl group of the Q176 side chain and Thr175 main chain (Fig. 3f). It suggested that the L176Q^NK mutant can stabilize sodium or water molecule in this site more than N49K mutant. In the calcium condition, the differential density of Site_CEN is stronger than that of the NavAb N49K mutant as well as that of Site_IN (Fig. 3g). The stabilization effect is stronger for more positively charged calcium ions than that of sodium ions. Therefore, calcium ions stacked at the center of the ion permeation pathway and blocked the current in the L176Q^NK mutant.

In the case of the L176G^NK mutant, a slight increase in differential density is observed in the calcium ion condition (Fig. 3i top). The differential density increases because the glycine mutation creates an extra cavity above the inner vestibule. The observed additional electron density of the extra cavity was assigned as a water molecule (Fig. 3c). As in the case of the L176Q^NK mutant, the additional water molecule forms three weak hydrogen bonds with the main chain of Val173 and Met174 (Fig. 3c and h) in the non-calcium condition. In the calcium condition, the electron density of Site_IN of the L176G^NK mutant was also more substantial than that of the NavAb N49K mutant (Fig. 3i middle and bottom). This additional electron density is connected to the electron density assigned as the additional water molecule (Fig. 3i bottom). Therefore, it is suggested that calcium ions stacked around Site_IN and blocked the current in the L176G^NK mutant.

The electron densities of calcium-blocking mutants suggested that the calcium ions corresponding to the electron density beneath the selectivity filter suppress the sodium current of these mutants. However, it is unlikely that as many as four divalent cations stack at the bottom of the selectivity filter because of the positive charge repulsion. Therefore, it is appropriate for the observed electron density to be the averaged electron density of one or two calcium ions in the inner vestibule. Magnesium and strontium ions also showed a blocking effect on the calcium-blocking mutants (Sup. Figure 5). The divalent cations might form loose and adjustable interaction networks with amino acids at the inner vestibule of the blocking mutants.

Molecular dynamics simulation of the calcium-blocking mutants

To analyze the detailed interaction between calcium ions and the mutated side chain of L176Q^NK and L176G^NK mutants, we proceeded with molecular dynamics (MD) simulation. From these simulations, we can also gain insight into the behavior of L176N^NK and L176A^NK mutants, whose structures were not determined. Furthermore, we can deduce the generality of the calcium-blocking mechanism by comparing results of these simulations.

For suitable parametrization of the nonbonded models about metal cations, we applied the electronic continuum correction (ECC) ^25,26. Without ECC, calcium ions bound to the extracellular and cytosolic entrance of the channel and inhibited the sodium ion permeation even in the wild-type channel (Sup. Figure 6). Applying the ECC, sodium ions were kept to permeate, and calcium ions became able to permeate the pore domain of the wild-type channel (Fig. 4b and c). The ionic currents of the wild-type channel slightly increased with calcium ions (Fig. 5a). The calculated current of wild-type NavAb is approximately 20 pA with symmetrical 100 mM NaCl under − 300 mV membrane potential (Fig. 5a and Sup. Table. 2). These values are consistent with the previously reported experimental data of NavAb or BacNavs ^27,28. Therefore, the simulation applying the ECC well represented the actual situation.

Similar subsequential permeations of sodium ions were observed in the calcium-blocking mutants without calcium ions (Sup. Figure 7). With calcium ions (100 mM sodium ion and 40 mM calcium ion), the sodium ion permeation is significantly decreased in the calcium-blocking mutants (Fig. 4c, 5a, and Sup. Table 2) in the MD simulation. We frequently observed a calcium ion stacking at the bottom of the selectivity filter (z = -5 ~ 0) in the calcium-blocking mutants (Fig. 4c). More than one calcium ions are never stacked in the inner vestibule simultaneously. When calcium ion was stacked at the bottom of the selectivity filter (Fig. 4c: right), sodium ions did not permeate through the pore domain (Fig. 4c: left).

Analysis of the free energy of calcium and sodium ions in the inner vestibule

We evaluated the behavior of calcium and sodium ions in the inner vestibule with calculated free energies. The two-dimensional free energy landscapes indicate that calcium ions stably locate on the intracellular side of the intracellular gate (Fig. 5b: z = -20) in the case of all channels (Sup. Figure 8). Still, these calcium ions did not disturb the sodium ion permeation, because the current of wild type channel was not decreased in the calcium ion condition.

To simplify the complex behavior, we compared the one-dimensional free energy profiles, which were calculated by integrating two-dimensional free energy landscape with respect to the radius up to 10 Å from the center of the channel pore (Fig. 5c-g, and Sup. Figures 9 and 10). In the wild type, the standard deviation of the free energy suggested that the free energy of the calcium-free sodium ion showed sufficient convergence in the inner vestibule (Fig. 5b: z = -15 Å to 0Å) and the bulk solution (Fig. 5b: z < -20 Å, z > 20 Å) (Sup. Figure 9b). On the other hand, the standard deviation for calcium ions was more widely spread in the bulk region than for sodium ions because of the small number of calcium ions in the calculation system (Sup. Figure 9c). Because the permeation of calcium ions rarely happened, the standard deviation for calcium ions was more significant in the inner vestibule. In the calcium ion condition, the sodium ions' free energy showed a well-converged standard deviation in the bulk region but broader in the inner vestibule (Sup. Figure 9d). It indicates the presence of calcium ions significantly affects the sodium ions in the inner vestibule.

In the presence of calcium ions, a small increment of the sodium ion free energy was observed at the extracellular entrance and the intracellular gate in the wild-type channel. Calcium and sodium ions' free energy is at a minimum around the side chains of the 176th residue (z = 0), in which the presence of calcium ions causes an increase in the free energy of sodium ions (Fig. 5c). Hence, the sodium current was reduced. Still, the calcium current compensated, so the currents with or without calcium ions are the same amount as the result of the electrophysiologic measurement (Fig. 2f and Fig. 5c).

In the L176Q and L176N mutants, calcium ions' free energy is significantly decreased near the 176th carbonyl oxygen atom (z = 0). This minimum free energy is consistent with the increment of the electron density observed in the L176Q mutant crystal structure under the calcium condition (Fig. 3h and i). Following the calcium ion's presence, sodium ions' free energy is increased from the intracellular (z = -15 Å) to the whole of the inner vestibule (z = 0~-10 Å) of the L176Q mutant (Fig. 5d). It is a reason that calcium ions disturb sodium ions from entering the channel. In the case of the L176N mutant, the free energy of sodium ions is also increased at the bottom of the selectivity filter, but the increment is small in the intracellular gate (Fig. 5f). This result is consistent with the weaker current inhibition under the calcium condition in MD calculations for L176N. However, this result is not consistent with the strong effect of L176N observed in electrophysiologic measurement (Fig. 2f).

Unlike the L176Q and L176N mutants, in the L176G and L176A mutants, the calcium ion free energy decreased at z = -5 Å, which corresponds to the extra cavity created by the small side chain of 176th residue (Fig. 5e and g). MD simulation indicates that water molecules are also more likely to be present in the extra cavity of L176G and L176A mutants (z = 0 Å) (Sup. Figure 11), which is consistent with the electron density of the crystal structure of the L176G mutants (Fig. 3h and i). Thus, hydrated Ca ions can be stably present at the selectivity filter connections in the lumen of L176G and L176A. In the case of K channels, the transition of potassium ions from the cytosol to the inner vestibule triggers the ion transition from the selectivity filter to the extracellular solution ²⁹. The stagnant of calcium ions in the inner vestibule would make it more difficult for sodium ions to enter the inner vestibule.

We succeeded in introducing the divalent cation blocking to NavAb. On the other hand, from a different perspective, the presence of divalent cations defeated the effective ion permeation in these mutants. The key to effective permeation of the wild-type channel seems to be the increased sodium ion free energy at the bottom of the selectivity filter (z = -5) to avoid overstabilization at the free energy minimum, which corresponds to the Leu176 side chain. The bottom of the selectivity filter is the entrance of the inner vestibule from the selectivity filter. This higher free energy of sodium ions of wild-type channels is especially noticeable in the presence of calcium ions. The L176N mutant, contrary to the wild type, shows a flatter free energy of sodium ions at the bottom of the selectivity filter (z = -5) in the presence of calcium ions (Fig. 5f). Thus, an excessively stable environment for the permeable ions would disturb effective permeation, which would be achieved by the hydrophobicity of the side chain of Leu176.

The crystal structure of the wild-type selectivity filter showed that calcium ions increase at Site_HFS (Fig. 3e). Still, the electrophysiologic analysis and the MD simulation indicated that the wild-type channel of NavAb allows the permeation of mainly sodium ions and only a few calcium ions (Fig. 2f, 5a, and 6d). On the other hand, the crystal structures of the divalent cation blocking mutants indicated that the mutations increase the stability of calcium ions around the bottom of the selectivity filter (Fig. 3). The blocking mechanism differs between mutations to hydrophilic residues and mutations to smaller residues (Fig. 6e and f). The hydrophilic-side chain mutations provide hydrogen bonds to the calcium ion around Site_CEN of the selectivity filter and allow the calcium ion to stack in the center of the ion permeation pathway (Fig. 6e). The small side chain mutation of Leu176 creates an extra cavity for an additional water molecule beneath Site_IN of the selective filter. It enables a divalent cation to interact with the water molecule and block the entrance of the inner vestibule (Fig. 6f).

The reproduction of the divalent cation blocking effect In NavAb helps us to reveal the molecular mechanism of the blocking effect of the divalent cations. Because BacNavs belong to one of the channel groups whose structural analysis is the most advanced, the structural analysis and simulation were well organized. The magnesium blocking of the NMDA receptor is removed during the depolarization state ². In our case, the calcium blocking remains regardless of the potential. This mismatch means there is still a lack of reproduction of magnesium blocking of NMDA receptors in NavAb mutants. Further analysis could reproduce a voltage-dependent inhibition mechanism like the NMDA receptor and elucidate its mechanism. By utilizing the accumulated knowledge of structural analysis, it has become possible to know the molecular basis of the various primary mechanism of ion channels including the mechanism of divalent cation blocking.

Site-Directed mutagenesis and construction of NavAb mutants

The NavAb mutated DNAs were subcloned into the modified pBiEX-1 vector (Novagen) that was modified by replacing the fragment from the NdeI site to the SalI site in multi cloning site with that of the previously described modified pET 15b (Novagen) vector ³⁰ between BamHI and SalI. The polymerase chain reaction accomplished site-directed mutagenesis using PrimeSTAR® Max DNA Polymerase (Takara bio.). All clones were confirmed by DNA sequencing.

Protein Expression And Purification

Proteins were expressed in the E. coli KRX strain (Promega). Cells were grown at 37ºC to an OD₆₀₀ of 0.6, induced with 0.2% rhamnose (Wako), and grown for 16 h at 25ºC. The cells were suspended in TBS buffer (20 mM Tris-HCl pH 8.0, 150 mM NaCl) and lysed using a French Press XXXLMA MINCOO) at 12,000 psi. Cell debris was removed by low-speed centrifugation (12,000×g, 30 min, 4°C). Membranes were collected by centrifugation (100,000×g, 1 h, 4°C) and solubilized by homogenization in TBS buffer containing 30 mM n-dodecyl-β-D-maltoside (DDM, Anatrace). After centrifugation (40,000×g, 30 min, 4°C), the supernatant was loaded onto a HIS-Select® Cobalt Affinity Gel column (Sigma). The protein bound to the cobalt affinity column was washed with 10 mM imidazole in TBS buffer containing 0.05% lauryl maltose neopentyl glycol (LMNG, Anatrace) instead of DDM. After washing, the protein was eluted with 300 mM imidazole, and His tag was removed by thrombin digestion (overnight, 4°C). Eluted protein was purified on a Superdex-200 column (Cytiva) in TBS buffer containing 0.05%LMNG.

Crystallization And Structural Determination

Before crystallization, the purified protein was concentrated to ~ 20mg ml^− 1 and reconstituted into a bicelle solution ³¹, containing a 10% bicelle mixture at 2.8:1 (1,2-dimyristoyl-sn-glycero-3-phosphorylcholine [DMPC, Anatrace]: 3-[(3-cholamidopropyl) dimethylammonio]-2-hydroxypropanesulfonate [CHAPSO, DOJINDO]). The NavAb-bicelle preparation was mixed in a 10:1 ratio. Prepared proteins were crystallized by sitting-drop vapor diffusion at 20°C by mixing 300-nl volumes of the protein solution (8–18 mg/ml) and the reservoir solution (9–11% polyethylene glycol monomethyl ether [PEG MME] 2000, 100 mM sodium chloride, 100 mM magnesium nitrate, 25mM cadmium nitrate and 100 mM Tris-HCl, pH 8.4) with mosquito LCP (TTP Labtech). The crystals were grown for 1 to 3 weeks, and after growing, the crystals were transferred into the reservoir solution without cadmium nitrate, which was replaced with the following cryoprotectant solutions. The cryoprotectant solution of the non-calcium condition contains 11% PEG MME 2000, 100 mM Tris-HCl pH 8.4, 2.5 M sodium chloride and, 20% (v/v) DMSO. The cryoprotectant solution of the calcium condition additionally contains 100 mM calcium nitrate.

All data were collected at BL41XU, BL45XU, and BL32XU of sPring-8 and merged using the automatic data processing system KAMO ³² with XDS ³³. The data sets of NavAb N49K and L176G^NK were obtained from a single crystal in both the calcium condition and the non-calcium condition. The data collections of NavAb L176Q^NK were performed automatically using the ZOO system ³⁴, and the data sets were obtained from four crystals in both calcium and non-calcium condition. Analyses of the data with the University of California Los Angeles anisotropy server ³⁵ revealed severely anisotropic crystals. Therefore, the data sets were ellipsoidally truncated and rescaled to minimize the inclusion of poor diffraction data.

A molecular replacement method with PHASER ³⁶ provided the initial phase. The final model was constructed in COOT ³⁷ and refined in refmac5 ³⁸ and Phenix ³⁹. Supplementary Table 1 summarizes Data collection and refinement statistics of all crystals.

Electrophysiological Measurement in Insect Cells

The recordings were performed using SF-9 cells. SF-9 cells (ATCC catalog number CRL-1711) were grown in Sf-900™ II medium (Gibco) complemented with 0.5% 100× Antibiotic-Antimycotic (Gibco) at 27°C. Cells were transfected with target channel-cloned pBiEX vectors and enhanced green fluorescent protein (EGFP)-cloned pBiEX vectors using Fugene HD transfection reagent (Promega). First, the channel-cloned vector (1.5 µg) was mixed with 0.5 µg of the EGFP-cloned vector in 100µL of the culture medium. Next, 3 µL Fugene HD reagent was added, and the mixture was incubated for 10 min before the transfection mixture was gently dropped onto cultured cells. After incubation for 16–48 h, the cells were used for electrophysiological measurements.

For measurement of the calcium blocking, pipette solution (140 mM CsF, 10 mM NaCl, 10 mM EGTA, and 10 mM HEPES [pH 7.4 adjusted by CsOH]) was used for the current measurement in the same way as the measurement of the calcium blocking of NavPp ¹⁷. As a bath solution, calcium blocking starting solution (30 mM NaCl, 120 mM NMDG-Cl, 2.5 mM CaCl₂, 10 mM HEPES [pH 7.4 adjusted by NaOH] and 10 mM glucose) was used for the 2.5 mM Ca²⁺ blocking condition. In each calcium condition, 10 mM CaCl₂ was replaced per 15 mM NMDG-Cl. To measure the magnesium and strontium blocking, starting solution (30 mM NaCl, 120 mM NMDG-Cl, 2.5 mM CaCl₂, 10 mM HEPES [pH 7.4 adjusted by NaOH], and 10 mM glucose) was used for the 0 mM magnesium or strontium blocking condition. For the magnesium and strontium blocking, 10 mM MgCl₂ and SrCl₂ were replaced per 15 mM NMDG-Cl. Cancellation of the capacitance transients and leak subtraction were performed using a programmed P/10 protocol delivered at -140 mV. The bath solution was changed using the Dynaflow® Resolve system. All experiments were conducted at 25 ± 2°C. All sample numbers represent the number of individual cells used for each measurement. Cells that had a smaller leak current than 0.5nA were used for measurement. When any outliers were encountered, these outliers were excluded if any abnormalities were found in other measurement environments and were included if no abnormalities were found. All results are presented as mean ± standard error.

Molecular Dynamics Calculation

All simulations were performed using the MD program NAMD (version 2.9) ⁴⁰. The 5YUC structure was obtained from the RBSC PDB databank and used as the pore domain structure of NavAb, which was truncated to residues Met130-Val213. The side chains of Glu177 of all monomers were deprotonated. Using the NAMD program and membrane plugin version 1.1 of VMD (https://www.ks.uiuc.edu/Research/vmd/plugins/membrane/), 1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) lipids were arranged to 98.0 × 98.0 Å² lipid bilayer, and water molecules were added into the lipid bilayer and formed as 98.0 × 98.0 × 75.0 Å³ box system (APL = 70.1). The pore domain was embedded into the POPC membrane using membrane plugin version 1.1 of VMD. After embedding, the system was equibilized to 91.0 × 91.0 × 86.0 Å³ box system by 100ns calculation. Twenty-five sodium ions, 10 calcium ions, and 29 chloride ions were added into the system as the condition containing 100 mM NaCl and 40 mM CaCl₂ using Autoionize Plugin, Version 1.5 VMD (http://www.ks.uiuc.edu/Research/vmd/plugins/autoionize/). Therefore, the wild-type NavAb system containing 100mM NaCl and 40mM CaCl₂ was composed of one channel, 210 POPC lipids, 25 sodium ions, 10 calcium ions, 29 chloride ions, and 13684 water molecules (see Table S2 for other concentrations). A periodic boundary condition of 91.0 × 91.0 × 86.0 Å³ was imposed.

The force fields of CHARMM36 ⁴¹ were used for the protein and lipid, and TIP3P ⁴²for the water. The parameters of sodium, calcium, and chloride were applied with the electronic continuum correction ^43,44. Referring to the electronic continuum correction ²⁵, we applied the electric charges of sodium, calcium, and chloride ions as 0.75, 1.5, and − 0.75, respectively.

After 1000 steps minimization, MD simulations were carried out without any applied electric field for 10 ns to equilibrate the systems in each condition. A time step was 2 fs per step. The bonds, including H atoms, were constrained using the SHAKE algorithm ⁴⁵. The anisotropic NPT ensemble was employed to keep the temperature at 300 K and pressure at 1 atm with the Langevin thermostat (damping coefficient of 1/ps) ⁴⁶, and the Nosé-Hoover Langevin piston barostat (piston period of 200 fs and piston decay time of 50 fs) ⁴⁷. Electrostatic interactions were calculated using the particle mesh Ewald method ⁴⁸ with a 1.2 nm cut-off in real space. The cut-off of the Lennard-Jones interaction was 1.2 nm.

The equilibrated systems were used for MD simulations in the presence of -147 mV membrane potential via a constant applied electric field [-0.04 kcal/ (mol ×Å×e)] applied across the Z-dimension of the simulation cell based on the equation: V(mV) = Ez×Lz(Å)×43.37 where Ez as the z-axis electric field, and Lz as the z-axis length of the simulation cell. The numerical value of 43.37 was the force conversion coefficient used by NAMD. MD simulations were performed for four initial configurations at all conditions. The calculation time of all simulations was summarized in Supplementary table 2. The coordinates were visualized with VMD ⁴⁹.

Acknowledgments

The synchrotron radiation experiments were performed at BL41XU and BL32XU in sPring-8 with the approval of the Japan Synchrotron Radiation Research Institute (JASRI) (Proposal number 2016B2721, 2017B2735, and 2018B2710). We thank the beamline staff for their excellent facilities and support. The computation was performed using Research Center for Computational Science, Okazaki, Japan (Project: 22-IMS-C114 and 22-IMS-C132). This research was partially supported by Platform Project for Supporting Drug Discovery and Life Science Research (Basis for Supporting Innovative Drug Discovery and Life Science Research [BINDS]) from AMED under Grant Number JP20am0101070. This work was supported by Grants-in-Aid for Scientific Research (17K17795, 20H00451 and 20K09193) and the Japan Agency for Medical Research and Development (AMED), SEI Group CSR Foundation, Takeda Science Foundation, Institute for Fermentation, and World Premier International Research Center Initiative (WPI), MEXT, Japan.

Hille, B. Ion Channels of Excitable Membranes, Third Edtition (Sunderland, MA: Sinauer Associates Inc). (2001).
Mayer, M. L., Westbrook, G. L. & Guthrie, P. B. Voltage-dependent block by Mg2+ of NMDA responses in spinal cord neurones. Nature 309, 261–263 (1984).
Chen, S. & Gouaux, E. Structure and mechanism of AMPA receptor — auxiliary protein complexes. Curr Opin Struct Biol 54, 104–111 (2019).
Hamada, K. & Mikoshiba, K. IP3 Receptor Plasticity Underlying Diverse Functions. Annu Rev Physiol 82, 151–176 (2020).
Fedele, L. et al. Disease-associated missense mutations in GluN2B subunit alter NMDA receptor ligand binding and ion channel properties. Nat Commun 9, (2018).
Cuello, L. G. et al. Structural basis for the coupling between activation and inactivation gates in K+channels. Nature 466, 272–275 (2010).
Wright, A. & Vissel, B. The essential role of AMPA receptor GluA2 subunit RNA editing in the normal and diseased brain. Front Mol Neurosci 5, 1–13 (2012).
Sobolevsky, A. I., Rosconi, M. P. & Gouaux, E. X-ray structure, symmetry and mechanism of an AMPA-subtype glutamate receptor. Nature 462, 745–756 (2009).
Lü, W., Du, J., Goehring, A. & Gouaux, E. Cryo-EM structures of the triheteromeric NMDA receptor and its allosteric modulation. Science (1979) 355, (2017).
Zhu, S. et al. Mechanism of NMDA Receptor Inhibition and Activation. Cell 165, 704–714 (2016).
Ren, D. et al. A prokaryotic voltage-gated sodium channel. Science (1979) 294, 2372–2375 (2001).
Irie, K. et al. Comparative study of the gating motif and C-type inactivation in prokaryotic voltage-gated sodium channels. J Biol Chem 285, 3685–94 (2010).
Koishi, R., Xu, H., Ren, D. & Navarro, B. A superfamily of voltage-gated sodium channels in bacteria. Journal of Biological Chemistry 279, 9532–9538 (2004).
Payandeh, J., Gamal El-Din, T. M., Scheuer, T., Zheng, N. & Catterall, W. A. Crystal structure of a voltage-gated sodium channel in two potentially inactivated states. Nature 486, 135–9 (2012).
Tsai, C. J. et al. Two alternative conformations of a voltage-gated sodium channel. J Mol Biol 425, 4074–4088 (2013).
Irie, K., Shimomura, T. & Fujiyoshi, Y. The C-terminal helical bundle of the tetrameric prokaryotic sodium channel accelerates the inactivation rate. Nat Commun 3, 793 (2012).
Shimomura, T. et al. A native prokaryotic voltage-dependent calcium channel with a novel selectivity filter sequence. Elife (2020) doi:10.7554/eLife.52828.
Payandeh, J., Scheuer, T., Zheng, N. & Catterall, W. A. The crystal structure of a voltage-gated sodium channel. Nature 475, 353–359 (2011).
Lenaeus, M. J. et al. Structures of closed and open states of a voltage-gated sodium channel. Proceedings of the National Academy of Sciences 114, E3051–E3060 (2017).
Ahuja, S. et al. Structural basis of Nav1.7 inhibition by an isoform-selective small-molecule antagonist. Science (1979) 350, (2015).
Jiang, D. et al. Structural basis for gating pore current in periodic paralysi. Nature 557, 590–594 (2018).
Xu, H. et al. Structural Basis of Nav1.7 Inhibition by a Gating-Modifier Spider Toxin. Cell 176, 702-715.e14 (2019).
Gamal El-Din, T. M., Martinez, G. Q., Payandeh, J., Scheuer, T. & Catterall, W. A. A gating charge interaction required for late slow inactivation of the bacterial sodium channel NavAb. J Gen Physiol 142, 181–190 (2013).
Tang, L. et al. Structural basis for Ca2+selectivity of a voltage-gated calcium channel. Nature 505, 56–61 (2014).
Nikitin, A. & Del Frate, G. Development of Nonbonded Models for Metal Cations Using the Electronic Continuum Correction. J Comput Chem 40, 2464–2472 (2019).
Phan, L. X., Lynch, C. I., Crain, J., Sansom, M. S. P. & Tucker, S. J. Influence of effective polarization on ion and water interactions within a biomimetic nanopore. Biophys J 121, 2014–2026 (2022).
Shaya, D. et al. Voltage-gated sodium channel (NaV) protein dissection creates a set of functional pore-only proteins. Proceedings of the National Academy of Sciences 108, 12313–12318 (2011).
Preisler, S. S. et al. Saccharomyces cerevisiae as a superior host for overproduction of prokaryotic integral membrane proteins. Curr Res Struct Biol 3, 51–71 (2021).
Sumikama, T. & Oiki, S. Digitalized K+ Occupancy in the Nanocavity Holds and Releases Queues of K+ in a Channel. J Am Chem Soc 138, 10284–10292 (2016).
Irie, K., Haga, Y., Shimomura, T. & Fujiyoshi, Y. Optimized expression and purification of NavAb provide the structural insight into the voltage dependence. FEBS Lett 592, 274–283 (2018).
Faham, S. Crystallization of bacteriorhodopsin from bicelle formulations at room temperature. Protein Science 14, (2005).
Yamashita, K., Hirata, K. & Yamamoto, M. KAMO: towards automated data processing for microcrystals. Acta Crystallographica Section D 74, 441–449 (2018).
Kabsch, W. XDS. Acta Crystallographica Section D 66, 125–132 (2010).
Hirata, K. et al. ZOO: an automatic data-collection system for high-throughput structure analysis in protein microcrystallography. Acta Crystallographica Section D 75, 138–150 (2019).
Strong, M. et al. Toward the structural genomics of complexes: Crystal structure of a PE/PPE protein complex from Mycobacterium tuberculosis. Proceedings of the National Academy of Sciences 103, 8060–8065 (2006).
McCoy, A. J. et al. Phaser crystallographic software. J Appl Crystallogr 40, (2007).
Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. Features and development of Coot. Acta Crystallogr D Biol Crystallogr 66, (2010).
Murshudov, G. N., Vagin, A. A. & Dodson, E. J. Refinement of Macromolecular Structures by the Maximum-Likelihood Method. Acta Crystallogr D Biol Crystallogr 53, (1997).
Adams, P. D. et al. PHENIX : a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr D Biol Crystallogr 66, (2010).
Phillips, J. C. et al. Scalable molecular dynamics on CPU and GPU architectures with NAMD. J Chem Phys 153, 44130 (2020).
Klauda, J. B. et al. Update of the CHARMM All-Atom Additive Force Field for Lipids: Validation on Six Lipid Types. Journal of Physical Chemistry B 114, 7830–7843 (2010).
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 Phys 79, 926–935 (1983).
Kohagen, M., Mason, P. E. & Jungwirth, P. Accurate description of calcium solvation in concentrated aqueous solutions. Journal of Physical Chemistry B 118, 7902–7909 (2014).
Nguyen, M. T. H., Tichacek, O., Martinez-Seara, H., Mason, P. E. & Jungwirth, P. Resolving the Equal Number Density Puzzle: Molecular Picture from Simulations of LiCl(aq) and NaCl(aq). Journal of Physical Chemistry B 125, 3153–3162 (2021).
Ryckaert, J.-P., Ciccotti, G. & Berendsen, H. J. C. Numerical integration of the cartesian equations of motion of a system with constraints: molecular dynamics of n-alkanes. J Comput Phys 23, 327–341 (1977).
Phillips, J. C. et al. Scalable molecular dynamics with NAMD. J Comput Chem 26, 1781–1802 (2005).
Feller, S. E., Zhang, Y., Pastor, R. W. & Brooks, B. R. Constant pressure molecular dynamics simulation: The Langevin piston method. J Chem Phys 103, 4613–4621 (1995).
Essmann, U. et al. A smooth particle mesh Ewald method. J Chem Phys 103, 8577–8593 (1995).
Humphrey, W., Dalke, A. & Schulten, K. VMD -- Visual Molecular Dynamics. J Mol Graph 14, 33–38 (1996).

There is NO Competing Interest.

SupplementaryTablesAndFigures.docx

Download PDF