Molecular Determinants of Mg²⁺-Mediated Inhibition in RyR1: Insights from Computational Approaches

Abstract

We investigated Mg²⁺-mediated inhibition of RyR1 by analyzing solvation, permeation, and binding interactions of Mg²⁺, Ca²⁺, Na⁺, and K⁺ across three functional states: Ca²⁺-activated (opRyR1), closed (clRyR1), and Mg²⁺-inhibited (HMg²⁺RyR1). Using molecular dynamics simulations, potential of mean force (PMF) analysis, quantum mechanical calculations, and MM-GBSA binding free energy calculations, we identified the structural and energetic determinants of Mg²⁺ inhibition. Our water occupancy analysis reveals that Mg²⁺ binding at D4945 stabilizes the S6 helical arrangement within the cytoplasmic vestibule in the HMg²⁺RyR1 state, maintaining a narrowed pore and reducing water accessibility. PMF calculations show that Mg²⁺ encounters the highest energy barriers, effectively restricting its permeation. Among the studied ions, Mg²⁺ exhibits the strongest affinity at the D4945 site, particularly in the HMg²⁺RyR1 state, reinforcing its inhibitory role. Binding energy analyses reveal that Mg²⁺ in HMg²⁺RyR1 has the lowest mobility and the most favorable binding free energy, indicating a highly stable ion-protein interaction and stronger retention in the closed and inhibited states. Additionally, Mg²⁺ binding is primarily stabilized by electrostatic interactions, which dominate over nonpolar contributions. These findings provide a comprehensive understanding of Mg²⁺-mediated RyR1 inhibition and offer critical insights into ion-specific regulation within the channel, further supporting structural models that highlight Mg²⁺’s role in stabilizing the closed conformation.

Highlights

• Mg²⁺ inhibition stabilizes the S6 helices, narrowing the pore and limiting water and ion conduction.

• PMF analysis shows Mg²⁺ encounters the highest energy barriers, preventing pore permeation.

• Mg²⁺ binds strongly at D4945, forming a structural blockade in the nonconductive state.

• Computational analyses (PMF, MD, QM, and MM-GBSA) reveal Mg²⁺ inhibition mechanisms and ion-specific regulation.

1. Introduction

Ryanodine receptors (RyRs) are intracellular calcium channels that mediate the release of Ca²⁺ from the sarcoplasmic/endoplasmic reticulum (SR/ER), playing a fundamental role in excitation-contraction (EC) coupling in skeletal and cardiac muscle. , RyRs form large homotetrameric ion channels, with each protomer consisting of over 5,000 amino acids, making them among the largest known ion channels. There are three RyR isoforms: RyR1, primarily found in skeletal muscle; RyR2, essential for cardiac muscle function; and RyR3, which is expressed in various tissues. ,

The RyR1 conduction pathway spans from the SR/ER lumen to the cytoplasm and consists of several distinct structural regions that regulate ion flow, including the selectivity filter (SF), the gating region (Q4933), and the hydrophobic gate (I4937) (Figure ). The ion permeation pathway is primarily formed by the SF and the transmembrane (TM) S5–S6 helices, assembled from four identical subunits that coordinate ion transport. − The selectivity filter (SF), located at the luminal entrance, consists of a highly conserved sequence (-GGGIGDE-, residues G4894-E4900), where negatively charged residues (D4899, E4900) attract cations and facilitate their entry into the pore. − The central gating region (Q4933) acts as a regulatory checkpoint that transitions between open and closed states, , while the hydrophobic gate (I4937) forms a permeation barrier that determines channel closure. ,,,

1.

(A) Tetramer structure of the RyR1 pore domain. (B) RyR1 pore domain in three functional states: opRyR1, clRyR1, and HMg²⁺RyR1 (PDB IDs: 7TDH, 7K0T, and 7UMZ, respectively) , shown with two diagonal subunits for clarity. The z-axis provides a comparative scale and highlights regional details within the pore. (C) Key pore residues are color-coded to distinguish structural regions.

While RyR1 is primarily responsible for Ca²⁺ release, electrophysiological studies have shown that it also conducts other cations, including K⁺, Na⁺, and Mg²⁺, with minimal selectivity under bi-ionic conditions. − Computational studies have further demonstrated that multiple cations, ,, including Mg²⁺, interact with negatively charged residues in the conduction pathway, influencing RyR1’s functional state. , Mutations in RyR1 can disrupt cellular Ca²⁺ homeostasis, leading to severe skeletal muscle disorders such as Malignant Hyperthermia (MH) and Central Core Disease (CCD). Notably, the D4938 residue in rabbit RyR1, corresponding to D4939 in human RyR1, has been specifically implicated in these conditions. , Furthermore, the negatively charged residues in the cytosolic vestibule of RyR1 (D4938 and D4945) play a key role in regulating ion flux. Mutagenesis studies have shown that the D4938N and D4945N mutations lead to reduced K⁺ conductance, with D4938N also diminishing the channel’s selectivity for Ca²⁺ over K⁺. These findings highlight the significant impact of these residues in RyR1 function.

Mg²⁺ is a physiological inhibitor of RyR1, preventing spontaneous Ca²⁺ release by stabilizing the channel in a closed state. Under physiological conditions, intracellular Mg²⁺ is maintained at ∼ 8 mM, with the majority most Mg²⁺ complexed with ATP (∼7 mM ATP-Mg²⁺), while only ∼ 1 mM remains as free Mg²⁺. − A reduction in free cytosolic Mg²⁺ levels has been linked to increased RyR1 activity, , contributing to pathological conditions such as Malignant Hyperthermia (MH) and Central Core Disease (CCD). , Mg²⁺ inhibits RyR1 through two distinct mechanisms: (i) allosteric inhibition – Mg²⁺ competes with Ca²⁺ at cytoplasmic activation sites, reducing Ca²⁺ binding affinity; (ii) Mg²⁺ also competes with Ca²⁺ at cytoplasmic inhibition sites, producing a similar inhibitory effect; and (iii) pore occlusion – Mg²⁺ enters the conduction pathway and interacts with negatively charged residues, physically blocking ion flow. ,,,

Recent cryo-electron microscopy (cryo-EM) studies have revealed that D4945 in the S6 helices serves as a key Mg²⁺-binding site in the high-Mg²⁺-bound (HMg²⁺RyR1) state. This structural state shows that Mg²⁺ forms stable interactions with all four D4945 residues, stabilizing the closed conformation and blocking the permeation pathway. Additionally, D4938 and E4942 in the cytoplasmic vestibule have been implicated in ion binding and regulation, further influencing RyR1 gating. However, while cryo-EM provides structural snapshots of Mg²⁺ interactions, it does not reveal the energetic and dynamic contributions of Mg²⁺ binding across different RyR1 functional states. Key questions remain regarding how Mg²⁺ binding influences ion permeation energetics, water accessibility, and structural stability in the RyR1 conduction pathway.

In this study, we employed molecular dynamics (MD) simulations, potential of mean force (PMF) calculations, quantum mechanical (QM) calculations, and binding free energy (MM-GBSA) analysis to investigate Mg²⁺-mediated inhibition in RyR1. Using cryo-EM structures of RyR1 in three functional states: Ca²⁺-activated (opRyR1), closed (clRyR1), and Mg²⁺-inhibited (HMg²⁺RyR1), we aimed to (i) characterize Mg²⁺ binding interactions at D4945 and other key residues across RyR1 states, (ii) analyze Mg²⁺ permeation barriers using PMF calculations, determining the energetic constraints on ion conduction, (iii) quantify Mg²⁺ binding stability through MM-GBSA calculations, decomposing electrostatic versus nonpolar interactions and (iv) assess water accessibility in the pore across functional states to understand Mg²⁺-induced hydration effects. By integrating computational and structural analyses, this study provides a mechanistic understanding of Mg²⁺-mediated RyR1 inhibition, elucidating ion-specific regulation, gating mechanisms, and molecular determinants of RyR1 function.

2. Materials and Methods

2.1. Structural Models of RyR1 Pore-Forming Region

Structural models for RyR1 in three distinct functional states, including opRyR1, clRyR1 and HMg²⁺RyR1 (PDB IDs: 7K0S, 7TDH, 7UMZ, respectively), were used as the starting point for simulations. , To investigate the permeation pathway, the pore domain (residue 4820–4956) of the tetramer cryo-EM structures was extracted from each cryo-EM structure and used as the starting structure for MD simulations. For the HMg²⁺RyR1 system, the initial model of the pore domain was derived from our previous simulation study and based on the cryo-EM structure of the Mg²⁺-inhibited state (PDB: 7UMZ). In this model, a single Mg²⁺ ion was positioned at the D4945 coordination site within the cytoplasmic vestibule, consistent with the electron density observed in the experimental map. The Mg²⁺ ion was modeled in its hexahydrated form, using established parameters that accurately capture its coordination geometry and hydration behavior in atomistic simulations.

2.2. Protein–Membrane System and Simulation Setup

The methodology for constructing protein–membrane systems followed established protocols. , Briefly, missing hydrogen atoms were added to the RyR1 structure using standard procedures. The initial protein–membrane systems for MD simulations were constructed using VMD software (version 1.9.3) with TCL scripting. Protonation states of all ionizable residues in the RyR1 pore domain were assigned based on pK _a predictions at pH 7.4 using the PROPKA module in PDB 2PQR 3.4. The system was modeled as a homotetramer, and identical protonation states were applied symmetrically to all four subunits to preserve structural and electrostatic consistency. A summary of the predicted pK _a values for all titratable residues is provided in Table S1 (Supporting Information), with nonstandard protonation states highlighted. For residues with borderline pK _a values (within 7.0–7.5), assignments were further reviewed in the context of local structural environment and available literature. Notably, residues such as Asp4945 and Glu4948 in the ion coordination site were retained in their deprotonated form. To address known limitations of PROPKA, final assignments were manually inspected and validated against experimental data to ensure biological plausibility and modeling accuracy.

The protein was embedded into a pre-equilibrated lipid bilayer composed of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) and TIP3P water model. The system was neutralized by adding counterions and maintained at a physiological salt concentration of 0.15 mM KCl using VMD’s Autoionize plugin at 298 K. Periodic boundary conditions were applied, with a simulation box size of ∼125 × 125 × 160 ų. A distance cutoff of 12 Å was used for calculating nonbonded interactions. Electrostatic interactions were handled via particle mesh Ewald summation using fast Fourier transform. van der Waals interactions were included with a switching distance of 10 Å. Langevin dynamics were employed to maintain a constant temperature of 298 K, with a damping coefficient of 1 ps^–1. The pressure was maintained at 1 atm using the Nose–Hoover Langevin piston method, with a piston period of 200 fs and a damping time of 50 fs.

In all simulations, the Mg²⁺ and Ca²⁺ ions were modeled in their explicitly hydrated forms to reflect realistic coordination environments. Mg²⁺ was represented as a hexahydrated complex ([Mg­(H₂O)₆]²⁺) and Ca²⁺ as a heptahydrated species ([Ca­(H₂O)₇]²⁺), based on the parametrization developed by Yoo and Aksimentiev. These hydrated ion models were incorporated using fixed bonding topologies that preserved the first-shell water coordination throughout the simulation. In contrast, Na⁺ and K⁺ were treated as bare ions, without predefined hydration shells, allowing for dynamic coordination with surrounding water molecules and protein residues. To prevent artificial overbinding to electronegative groups such as Asp or Glu side chains, we applied NBFIX corrections to the Lennard-Jones parameters for specific ion–atom pairs involving Na⁺ and K⁺.

Energy minimization was performed to eliminate steric clashes. Restrained MD simulations were employed to relax structural strains in the model systems. Initially, restraints were applied to the protein and lipid head groups, allowing solvent and counterion to equilibrate. In the subsequent phase, constraints were gradually released, enabling full equilibration of waters, lipids, and counterions. Finally, the equilibration and production MD simulations were performed with a time step of 2 fs over 300 ns using the CHARMM36 force field parameters. Each system was repeated 3 times for statistical estimation. Harmonic constraints were applied to protein terminal ends and loop regions to maintain structural integrity. The MD simulations were performed with the program NAMD version 2.12. The MD trajectories collected during the production run were analyzed to investigate structural properties. A summary of simulation systems is provided in Table S2.

2.3. Steered Molecular Dynamics (SMD) Simulations

SMD simulations were performed to generate reaction coordinates for evaluating the interactions between residues and ions. The systems were pre-equilibrated through multiple successive steps: (1) all atoms except lipid tails were fixed to stabilize lipid–protein interactions; (2) harmonic restraints were applied to protein atoms to relax protein–water interactions; (3) all atoms were subsequently allowed to move freely. After a 2 ns pre-equilibration, the protein was well accommodated within the lipid membrane, as evidenced by the stability of the protein structure and system energy. The starting position of the target ion, including Ca²⁺, Mg²⁺, Na⁺ and K⁺, was placed at the luminal side. Based on previous studies, Mg²⁺ and Ca²⁺ were modeled in their hydrated forms to ensure accurate ion interactions using improved force field parameters and hydrated ion topology developed by Yoo and Aksimentiev. The ion pulling pathway was generated along the z axis from 40.0 Å to −59.0 Å. The probe ion was pulled along the z direction from the luminal side to the cytosolic side with force constant of 5.0 kcal/mol·Å² at a constant velocity of 1.0 Å/ns. To minimize structural perturbations in clRyR1 and HMg²⁺RyR1 during ion passage due to pore closure, harmonic restraints were applied to the C_α atoms of the protein, preserving structural integrity. A force constant of 5.0 kcal/mol·Å² was used to counterbalance the applied pulling force. Harmonic restraints were applied to the protein backbone using a force constant equal to that of the ion-pulling force, in order to minimize artificial translation or rotation of the tetrameric pore during ion displacement. These restraints were selectively applied to transmembrane helices positioned away from the ion conduction pathway to disrupting ion-binding sites or gating dynamics. To ensure the target ion remained within the permeation pathway, additional restraints of 100.0 kcal/mol·Å² were applied in the xy-plane, restricting lateral movement beyond 8.0 Å from the central axis.

To assess translational stability during SMD, we tracked the time evolution of the RyR1 pore domain center-of-mass (CoM) relative to its initial position and backbone RMSD profiles with respect to the equilibrated structure.

2.4. Calculation of Potentials of Mean Force (PMF)

Potentials of mean force (PMF) of ion transfer from the luminal to cytosolic medium were calculated for Ca²⁺, Mg²⁺, Na⁺ and K⁺ ions using the adaptive biasing force (ABF) method. , The reaction coordinates for each ion were derived from SMD simulations as described in the previous section, and represent the ion movement pathway along the pore (z-) axis of the RyR1 channel. Sampling was performed over a 99 Å range (from 40.0 Å to −59.0 Å along the z-axis), divided into 59 equidistant windows with a bin size of 2.0 Å for uniform sampling. Initial structures for each window were extracted from SMD snapshots based on the ion’s z-position. These structures were energy-minimized and equilibrated for 2.0 ns to remove artifacts introduced by SMD pulling forces. Following equilibration, PMF sampling was conducted for 6.0 ns per window, with output values recorded every 1,000 steps. Boundary potentials with a force constant of 100.0 kcal/mol·Å² were applied to confine ion movement. PMF profiles from individual windows were merged through a short simulation along the reaction coordinate. Each system was simulated in triplicate to ensure reproducibility, resulting in a total cumulative simulation time exceeding 1.38 μs. The final PMFs were obtained as the averaged profiles for the four ions. A summary of all systems, including details of SMD, ABF, and PMF calculations, is provided in Table S3.

2.5. MD Trajectory Analysis

To assess the structural stability of the RyR1 pore domain during the 300 ns production runs, root-mean-square deviation (RMSD) analyses were conducted as part of the structural validation. RMSD profiles were calculated for backbone atoms across three independent replicates for each functional state (opRyR1, clRyR1, and HMg²⁺RyR1) to monitor convergence and overall structural stability over time. To characterize the coordination environment of Mg²⁺ at the D4945 site in the RyR1 pore, radial distribution functions (RDFs, g­(r)) were calculated between the Mg²⁺ ion and oxygen atoms of water molecules (Mg²⁺–O­(H₂O)) as well as carboxylate oxygen atoms of D4945 residues (Mg²⁺–O­(D4945)). The g­(r) and the corresponding coordination numbers n­(r) were computed using a custom Tcl script in VMD, applied to trajectory frames from windows of the equilibrated simulation. A bin width of 0.05 Å and a cutoff of 15 Å were used. This analysis provided quantitative insight into the first-shell hydration structure of Mg²⁺ and second-shell interactions with protein side chains, allowing us to assess the stability and coordination mode of Mg²⁺ across different RyR1 functional states.

Comprehensive analyses of MD trajectories were performed to investigate the role of hydration, pore geometry, and solvent accessibility in ion conduction through the RyR1 pore. These analyses included quantifying the number of water molecules within the pore, and evaluating the solvent accessible surface area (SASA), both of which were calculated using Tcl scripts in VMD. The structural dimensions of the RyR1 pore were characterized using the HOLE program, which provides a quantitative profile of pore radii as a function of position along the channel axis. To assess the hydration environment, the number of water molecules within the pore was determined by counting oxygen atoms, ensuring an accurate representation of water distribution. Water molecule positions were mapped based on their z-coordinates relative to the pore axis, and their distributions were analyzed to reveal hydration patterns along the ion translocation pathway. To obtain high-resolution spatial profiles, all data were binned along the z-axis using a bin width of 1 Å, enabling a detailed examination of water density variations throughout the pore. SASA calculations were performed using the ″measure sasa″ command in VMD, focusing on residues lining the pore. Only atoms facing the interior of the channel were considered to capture variations in solvent accessibility within the ion permeation pathway. The computed SASA values provide insight into dynamic pore fluctuations and potential constriction points that may influence ion movement.

2.6. Calculations of Solvation Free Energy and Electrostatic Potentials

The solvation free energy (ΔG _solv ) was computed using a nonlinear Poisson–Boltzmann continuum model. This energy consists of two components, electrostatic (ΔG _elec ) and nonpolar (ΔG _np ), such that ΔG _solv = ΔG _elec + ΔG _np . Only residues lining the interior of the RyR1 pore were selected for these calculations totaling 60 residues (Table S4).

Parameters for ΔG _solv evaluation were adapted from previous studies. , The multigrid lengths used in the Poisson–Boltzmann calculations were determined based on the protein’s dimensions, with each dimension scaled by a factor of 1.5 to ensure sufficient grid resolution. For electrostatic solvation energy calculations, the dielectric constants for the protein, water, and vacuum were set to 2, 80, and 1, respectively. Protein charges were assigned based on predicted pK _a values. The ionic strength of the bathing solution was set to 0.15 M, with monovalent ion carrying formal charges of +1 and – 1 and an effective ion radius of 2.0 Å. A water probe radius of 1.4 Å was used to define the solvent-accessible surface. For the nonpolar solvation component, the solvent pressure was set to 0.150624 kJ·mol^–1·Å^–3, bulk solvent density to 0.033428 Å^–3, and solvent molecule radius to 1.4 Å. The optimized surface tension coefficient was 0.0209 kJ·mol^–1·Å^–2. Protein atomic charges and van der Waals radii were assigned according to the AMBER ff99 force field parameters. Electrostatic potentials were mapped onto the surface of the RyR1 pore domain to estimate charge distribution. The solvation free energy calculations and the generation of electrostatic potential surfaces were performed using the PDB 2PQR and APBS programs. These analyses provided insights into the electrostatic properties of the pore environment.

2.7. DFT Calculations for Metal Ion Binding Energies

To assess the relative stability of metal ions (M) interacting with amino acid residues (aa) in the RyR1 pore, we computed the binding energy of the [M­(H₂O) _x (aa)₄]^k ± complex using density functional theory (DFT). The [M­(H₂O) _x (aa)₄]^k ± models consist of a hydrated metal ion, M (Ca²⁺, Mg²⁺, K⁺, and Na⁺), coordinated by x water molecules within a 3.0 Å cutoff, and four identical amino acid residues (aa)₄ selected from the tetrameric RyR1 structure. These residues were extracted from specific interaction sites spanning from the SF to the cytoplasmic vestibule. Each configuration was taken from equilibrated MD snapshots to ensure accurate representation of the coordination environment observed during the simulations. The total charge of the complex, including the hydrated metal ion and the four coordinating residues, is denoted by k. To compute the binding energy, single-point DFT calculations were performed at the B3LYP level using the 6–31+G­(d,p) basis set. The solvent environment was accounted for implicitly using the conductor-like polarizable continuum model (CPCM) to approximate solvation effects. The binding energy (ΔE _bind ) were determined according to the following equation:

$$\mathrm{\Delta }{E}_{bind}=E_{AB}-(E_{\mathrm{A}}+E_B)$$ 1

where E _AB is the total energy of the [M­(H₂O) _x (aa)₄]^k ± complex, E _A is the energy of the hydrated metal ion [M­(H₂O) _x ]^m+, and E _B is the energy of the amino acid cluster [aa]₄ ^n ± . To ensure consistency, total energies of the complex (E _AB ) and its isolated system (E _A and E _B ) were computed separately. The ΔE _bind for each metal ion were computed from n independent DFT single-point energy calculations, where n corresponds to the number of distinct metal–residue complex configurations. Each configuration was extracted as a unique snapshot from equilibrated MD trajectories to ensure structural diversity and representativeness. To sample variation around each site, snapshots were selected from a spatial window within ± 2 Å of the average center of mass (CoM) of the respective residue environment. All DFT computations were performed using the Gaussian 09 software package. These computations provided quantitative insights into the strength of metal ion interactions within the RyR1 pore.

2.8. MM-GBSA Binding Free Energy Calculation of Mg²⁺ at the D4945 Site

The molecular mechanics method combined with the generalized Born and surface area continuum solvation (MM-GBSA) approach was employed to estimate the 2D-contour map of protein-Mg²⁺ binding free energy (ΔG _bind ), which governs the stabilization of the complex is defined by,

$$\mathrm{\Delta }{G}_{bind}=\mathrm{\Delta }{G}_{solv}^{complex}-(\mathrm{\Delta }{G}_{solv}^{protein}+\mathrm{\Delta }{G}_{solv}^{ion})$$ 2

where solvation energy of complex, ΔG _solv , is subtracted with solvation energies of protein and Mg²⁺ ion, ΔG _solv and ΔG _solv , respectively. In the MM-GBSA calculations, over 5,000 snapshots were extracted from equilibrated trajectories where Mg²⁺ was positioned at the D4945 site. Explicit water molecules and other ions were removed, except for the target Mg²⁺ ion. The detailed procedure and all parameters were adopted from previous studies. MM-GBSA calculations were performed using NAMD version 2.12 , with the CHARMM36 force field and refined force field parameters for Mg²⁺. To construct the 2D-contour map, the center of mass of Mg²⁺ at the D4945 site was recorded and mapped with ΔG _bind by projecting it onto the x–y plane of D4945.

Error bars are shown in figures where data were obtained from multiple independent configurations or trajectories (e.g., ΔG _bind , ΔE _bind ), while figures derived from single production runs (e.g., PMF profiles) present representative values without statistical averaging, due to computational limitations.

3. Results

RMSD plots for all three functional states of RyR1 (opRyR1, clRyR1, and HMg²⁺RyR1), show that all systems reached equilibrium within the first 100 ns and remained structurally stable throughout the 300 ns simulations. RMSD values fluctuated between approximately 0.7–2.0 Å, indicating that the simulated structures remained closely aligned with their initial experimental conformations across three independent replicates (Figure S1A–C). These results confirm that the simulations achieved conformational convergence while preserving the characteristic structural features of each functional state. Notably, the opRyR1 system exhibited the highest RMSD values, indicating greater conformational flexibility. In contrast, HMg²⁺RyR1 displayed the lowest RMSD values, reflecting a higher degree of structural rigidity, likely stabilized by Mg²⁺ binding at key sites within the cytoplasmic vestibule.

To assess whether each functional state of RyR1 evolved toward alternative conformations during the simulations, we performed cross-state structural comparisons between the MD-derived trajectories and cryo-EM reference structures. Specifically, RMSD profiles were calculated between each MD simulation and the backbone atoms of the alternate cryo-EM conformations (Figure S1D–I), and pore radius profiles were compared to the corresponding cryo-EM states (Figure S2A–F). Across all three functional states, the results showed that each system largely retained its initial structural features over the 300 ns time scale. No spontaneous convergence toward alternate conformations was observed. RMSD values remained distinct and reflective of their respective starting states, and the pore architecture preserved its original conformation throughout the trajectories. These observations were further supported by structural superpositions (Figure S2A–F), which demonstrated the preservation of key geometrical features without significant deviation toward other functional configurations. Collectively, these findings suggest that the cryo-EM-derived conformations represent stable local minima on the RyR1 energy landscape within the simulated time scale.

3.1. Exploring the RyR1 Pore Shaping and Water Occupancy

To examine the structural differences in RyR1 pore across its distinct functional states, we computed the pore profile for each conformation and analyzed the diameter surface maps along the pore axis (Figures A and S2–3). The results revealed notable variations in pore morphology among the Ca²⁺-activated open state (opRyR1), closed state (clRyR1), and Mg²⁺-inhibited closed state (HMg²⁺RyR1). A clear hourglass-like geometry was observed in all three states, characterized by a wider luminal mouth and cytoplasmic vestibule flanking a constricted central region. However, the degree of constriction varied significantly between states, influencing pore accessibility and ion conduction properties.

2.

(A) Pore shape and diameter of RyR1 in MD simulations, with red indicating narrow regions (r < 2.0 Å) and blue representing wider region (r > 3.0 Å). (B) Pore radius along the z-axis, with cryo-EM structures (blue) and MD simulations (black, shaded: SD). Water occupancy (red curve) is shown with error bars (SD). (C) MD snapshots of pore water (yellow surface) in RyR1 with pore helices (gray), key residues (licorice representation), and pore region definitions aligned to the z-axis scale.

In the opRyR1 state, the pore exhibited the widest diameter, particularly at the gate (G) and hydrophobic gate (HG) regions, supporting an ion-permeable configuration. This conformation allows for efficient ion flow from the SR lumen to the cytoplasm, aligning with the expected conductive state of the channel. In contrast, the clRyR1 and HMg²⁺RyR1 states displayed progressive narrowing of the pore, particularly at the hydrophobic gate. These regions exhibited a more compact conformation, leading to a higher energy barrier for ion permeation. The HMg²⁺RyR1 state, in particular, showed the most pronounced constriction, suggesting that Mg²⁺ binding further stabilizes the closed conformation by enhancing steric hindrance and reducing pore hydration.

To quantify these differences, we compared the pore radius profiles of the cryo-EM structures with those of their MD-equilibrated counterparts. Profiles derived from the final 100 ns of each trajectory showed strong agreement with the experimental structures (Figure B), confirming that the overall pore shape and gating architecture were well preserved. Minor deviations observed in specific regions were attributed dynamic fluctuations during equilibration but did not significantly impact the overall pore geometry (Figures A-B and S2–3). These results indicate that the MD simulations accurately reproduce structural features of the cryo-EM models while capturing expected thermal relaxation, thereby supporting the robustness and validity of the computational approach. Furthermore, the key pore-lining residues (e.g., D4845, G4894, and I4937) exhibited state-dependent variations in their spatial orientation, contributing to distinct pore accessibility and water distribution patterns across the three states. The relative positioning of these residues plays a crucial role in modulating ion permeability and hydrophobic gating, reinforcing the idea that conformational transitions in RyR1 directly influence ion transport efficiency.

The averaged pore radius profiles along the z-axis for the MD-equilibrated structures are presented in Figure B, comparing the simulated pore radii (black lines) to their respective cryo-EM structures (blue lines). The shaded region represents the standard deviation (SD) of the mean, indicating fluctuations during equilibration. Overall, the pore radius profiles exhibit an hourglass-shaped structure, with a wider cytoplasmic vestibule and luminal entrance, flanking a more constricted central region. While minor deviations between MD-equilibrated and cryo-EM structures were observed, these differences likely stem from applied structural constraints and equilibration adjustments, ensuring the stability of the isolated pore domain. Among the three states, the opRyR1 conformation exhibits the largest pore radii, particularly at the hydrophobic gate and gate regions, supporting an ion-permeable state. In contrast, the clRyR1 and HMg²⁺RyR1 states display narrower pore radii, particularly in the selectivity filter and gate regions, suggesting a nonconductive conformation that impedes ion flow. The most notable constriction occurs in HMg²⁺RyR1, where the hydrophobic gate and gate regions exhibit the most severe narrowing, reinforcing the role of Mg²⁺ in stabilizing the closed state and raising the energy barrier for ion permeation.

To further assess how these structural differences influence pore hydration, we analyzed the distribution of water molecules along the pore axis and their fluctuations across different regions (Figure B, red curve). Water occupancy closely follows the pore radius profile, with regions of high hydration corresponding to wider sections, and dehydration occurring in the narrowest regions. In the opRyR1 state, the pore is fully hydrated from the luminal entrance to the cytoplasmic side (Figure C). This state supports continuous water-mediated ion conduction, aligning with the channel’s conductive function. However, at G4894 (within the selectivity filter), where the pore radius is smallest (∼1.4 Å, Figure A), a notable reduction in water occupancy is observed (Figure B). In contrast, clRyR1 and HMg²⁺RyR1 exhibit a distinct dewetting effect, particularly within the hydrophobic gate and gate regions. The average water distribution profiles (red curves in Figure B) indicate a clear depletion of water molecules in the region spanning 0 Å > z ≳ 10 Å, compared to opRyR1. This dehydration effect is particularly pronounced in HMg²⁺RyR1, where the pore closure is more extensive, leading to a broader dewetting zone (Figure C).

The hydration behavior of the gate and hydrophobic gate regions is critical for understanding RyR1 gating mechanisms. The closure of the channel at Q4933 (gate, G), combined with the narrowing at I4937 (HG region), results in the formation of a hydrophobic dewetting zone. This zone prevents water penetration (Figure B–C) and effectively blocks ion transport, reinforcing a nonconductive conformation. A fully hydrated pore is necessary for stable ion conduction, as water molecules form hydration shells around permeating cations, reducing electrostatic barriers. The reduction of water occupancy in the clRyR1 and HMg²⁺RyR1 states thus represents a structural mechanism of channel inhibition, where pore dehydration leads to increased energetic barriers for ion entry and permeation.

The pronounced dewetting observed in HMg²⁺RyR1 suggests that Mg²⁺ binding enhances pore closure by stabilizing a tightly packed, hydrophobic environment that discourages water infiltration. This aligns with experimental findings that Mg²⁺ functions as a physiological inhibitor by preventing RyR1 activation, effectively reducing Ca²⁺ flux. , Furthermore, Mg²⁺ inhibition at the pore may operate via two mechanisms: (i) direct electrostatic interactions with key pore-lining residues, reinforcing the nonconductive conformation, and (ii) indirect effects on hydration, where Mg²⁺ binding shifts the equilibrium toward a closed, dehydrated pore state, increasing the energy required for water and ion entry. These results suggest that Mg²⁺ plays a critical role in maintaining RyR1 in a functionally closed state, not only by sterically hindering ion flow but also by promoting hydrophobic dewetting, further reinforcing the nonconductive nature of the inhibited state. Together, the pore radius profiles and hydration analysis provide a structural and energetic basis for understanding how Mg²⁺ inhibits RyR1 directly at the pore. The narrowing of the hydrophobic gate and gate regions in HMg²⁺RyR1, coupled with the increased dewetting, effectively blocks ion transport. This mechanism highlights the interplay between pore geometry, hydration dynamics, and cation binding in regulating RyR1 activity.

3.2. Hydration Energetics and Solvation Barriers in RyR1 Pore

We analyzed the hydration energetics of the RyR1 pore by computing the free energy of solvation (ΔG _solv ) for pore-facing residues along the z-axis (Figure ). This energy decomposition provides insights into the thermodynamic cost of water accommodation in different pore regions and its dependence on channel conformation. The total ΔG _solv consists of two primary components: ΔG _elec and ΔG _np . Figure presents these energetic components along the pore axis for three different RyR1 states: opRyR1, clRyR1, and HMg²⁺RyR1. The total ΔG _solv is lowest at the luminal entrances and cytoplasmic vestibule in all three systems, indicating favorable hydration due to increased water accessibility. In contrast, the highest ΔG _solv values are observed at the narrowest pore regions, where significant energetic barriers to solvation exist. Notably, the HMg²⁺RyR1 system exhibits the most energetically unfavorable band, which can be attributed to the extended constriction in the hydrophobic gate and gate regions. This extended narrow section significantly increases the solvation energy barrier, making water penetration less favorable compared to the opRyR1 and clRyR1 states.

3.

Averaged solvation energy (ΔG _solv ) profiles along the RyR1 pore axis, calculated using the nonlinear Poisson–Boltzmann equation from MD snapshots. Electrostatic (ΔG _elec ) and nonpolar (ΔG _np ) components are shown separately. The right panel illustrates pore regions along the z-axis for reference. Note: Extreme ΔG values in the hydrophobic gate (HG) region were omitted to enhance visibility in other regions; an axis break indicates these omissions. The excluded segments correspond to ΔG _solv values ranging from ∼−52 to −600 kcal/mol and ΔG _elec values from ∼−65 to −660 kcal/mol. Complete profiles are provided in Figure S8.

A closer examination of the individual solvation energy components reveals distinct contributions from electrostatic and nonpolar interactions. The ΔG _elec is strongly negative in regions rich in charged residues, such as the cytoplasmic vestibule (CV) and selectivity filter (SF). These regions facilitate water accommodation due to favorable interactions with polar solvent molecules. However, in the hydrophobic gate and gate regions, ΔG _elec is significantly reduced, reflecting the low presence of charged groups and restricted water accessibility. On the other hand, the ΔG _np follows an opposite trend, with the most hydrophobic regions (HG and G) exhibiting high ΔG _np values, indicating an energetically unfavorable environment for water. In HMg²⁺RyR1, the ΔG _np contribution is particularly elevated in the constricted pore region, reinforcing the observation that reduced hydration in this state is linked to increased hydrophobicity and steric constraints.

The SASA serves as a direct indicator of how exposed the pore residues are to the solvent, correlating with hydration potential. The SASA values across the different RyR1 states follow the trend: opRyR1 (14706 Ų) > clRyR1 (14607 Ų) > HMg²⁺RyR1 (14526 Ų). The reduced SASA in HMg²⁺RyR1 compared to clRyR1 aligns with the observed higher ΔG _solv values and increased hydrophobicity, further confirming that water exclusion is more pronounced in the presence of high Mg²⁺ concentration. These findings highlight the critical role of pore geometry and residue composition in hydration energetics. The constricted and hydrophobic regions, particularly in HMg²⁺RyR1, create an energetically unfavorable environment for water, leading to reduced hydration and increased solvation energy barriers.

Since Mg²⁺ is known to inhibit RyR1 by stabilizing the closed state, our results suggest that this inhibition may involve both direct electrostatic interactions with pore-lining residues and an indirect effect through altered hydration dynamics. The extended hydrophobic dewetting zone observed in HMg²⁺RyR1 likely contributes to reduced water accessibility, which in turn could hinder ion conduction and stabilize the functionally closed state. This mechanism aligns with previous experimental findings indicating that Mg²⁺ binding enhances the energetic barrier for pore opening, effectively suppressing channel activation. These insights provide a quantitative perspective on Mg²⁺-mediated inhibitory effects at the RyR1 pore, emphasizing the interplay between solvation energetics, pore constriction, and ion channel gating.

3.3. Energetic Effects of Mg²⁺ and Other Ions in RyR1 Conduction

To assess translational stability during SMD, we monitored the time evolution of the center-of-mass (CoM) of the RyR1 pore domain relative to its initial position. As shown in Figure S4A, the CoM fluctuations remained minimal, within ± 0.5 Å across all trajectories, confirming that no significant drift or global displacement occurred during ion pulling. We also evaluated protein conformational stability by calculating backbone RMSD profiles for all three functional states. As presented in Figure S4B, RMSD values stabilized within 1.0–3.0 Å, indicating that no large-scale structural distortions occurred. These fluctuations fall within the expected range for membrane protein systems and confirm that the channel remained conformationally stable throughout the simulations. In addition to global RMSD stability, key ion-binding residues such as D4945 and Q4933 retained consistent spatial orientation and did not undergo structural collapse. While peripheral loops and termini displayed normal dynamic flexibility, the transmembrane core remained well-structured across all simulations.

PMF profiles were generated to identify key interaction sites and energy barriers that regulate ion transport or blockage under different functional states. All PMF profiles, shown in Figure A, enable a comparative analysis of the energetic differences across the three channel states. The profiles are plotted along the z-axis, representing the ion’s translocation pathway from the sarcoplasmic side to the cytosolic side. This coordinate maps the free energy changes associated with ion movement through the RyR1 pore, highlighting energetic barrier and favorable binding sites along the conduction pathway. Notably, the overall PMF pattern of opRyR1 is consistent with previous studies, further validating the data set and supporting the reliability of the simulation results. ,

4.

(A) Potential of mean force (PMF) profiles for Ca²⁺ (red), Mg²⁺ (blue), Na⁺ (orange), and K⁺ (green) along the RyR1 pore, with key pore regions and residues annotated. The z-axis alignment and scale are shown in the right panels. (B) Electrostatic potential distribution on the RyR1 pore surface, with portions of helices removed to illustrate the inner architecture. Key residues are highlighted (dashed black rectangles). The color scale ranges from −20 to 20 k _B T/e at T = 25 °C, with positive (blue), negative (red), and neutral (white) regions.

We examined the permeation of four ion types: Ca²⁺, Mg²⁺, K⁺, and Na⁺. Ca²⁺ was chosen as the reference ion for comparison, as besides Mg²⁺, it is the primary ion associated with RyR1 during its normal physiological function. , Moreover, previous studies have extensively investigated the Ca²⁺ permeation energy in the open state, , providing a strong benchmarking for interpreting our results. The free energy values for all ions (Ca²⁺, Mg²⁺, K⁺, and Na⁺) within the −20 Å < z < −10 Å range, corresponding to the selectivity filter (SF) region, exhibit a decreasing trend, indicating a favorable energy landscape for ion entry. Divalent cations (Ca²⁺ and Mg²⁺) show lower free energy values compared to monovalent cations (K⁺ and Na⁺), forming a characteristic energy well across all three RyR1 states (Figure A). The PMF depths for monovalent cations are shallower than those for divalent cations, reinforcing the higher affinity of RyR1 for Ca²⁺ and Mg²⁺. Interestingly, energy profiles in the SF region remain largely unchanged between different channel states. This consistency may result from the similar pore radius of the SF region across the states, and its broader geometry, which minimizes fluctuations in local electrostatic interactions (Figure B). To further explore the electrostatic environment of the RyR1 pore, we generated electrostatic potential maps (Figure B). These maps illustrate variations in surface charge distribution, revealing the electronegative nature, depicted in red. This negative charge feature aligns with the observed PMF reductions in this region, confirming a direct relationship between the electrostatic properties and ion permeation energy.

In the region between −10 Å < z < 0 Å, where ions transition from the selectivity filter into the central cavity, the free energy values increase significantly, reaching a peak at z ∼ −5 Å (∼3 kcal/mol in most ions). This rise in energy is likely influenced by the constriction site at G4894 (Figure ), where the negative electrostatic character diminishes (Figure B). As ions proceed through the pore, a PMF basin emerges near the gate formed by Q4933 at z ∼ 5 Å, where the energy profile exhibits a slight decrease, suggesting a more favorable ion binding site. Key residues including the negatively charged E4900 and D4899 in the SF, along with G4894 and Q4933, collectively shape the PMF profiles of all ions in a similar manner across different RyR1 conformational states. However, at the narrowest point of the RyR1 pore (z = 10 Å), a significant energy barrier was observed (Figure A), primarily attribute to the hydrophobic gate formed by I4937. In the opRyR1 state, the free energy of Mg²⁺ reached the highest value (∼6 kcal/mol) among the tested ions, indicating that Mg²⁺ encounters the most substantial energetic resistance in the open conformation. This finding aligns with the known lower permeability of Mg²⁺ through RyR1, reinforcing the channel’s selective preference for Ca²⁺ under physiological conditions.

The PMF analysis further reveals distinct energetic profiles for Ca²⁺, Mg²⁺, K⁺, and Na⁺, with markedly different patterns in the HMg²⁺RyR1 and clRyR1 states compared to opRyR1. In these nonconductive states, the PMF exhibited a sharp increase at the I4937 region (Figure A), which is consistent with the formation of the hydrophobic gate. This pronounced energy barrier serves as a signature feature of this region, reinforcing the structural and functional consistency of the HG in blocking ion permeation. The elevated free energy values observed in the clRyR1 and HMg²⁺RyR1 states can be attributed to the nonconductive conformation of the channel, characterized by pore narrowing and a more hydrophobic environment. This observation correlates directly with the reduced pore radius and altered electrostatic properties in these states, as seen in Figure and B. The more restrictive pore geometry and enhanced hydrophobic gating in HMg²⁺RyR1 further suggest that Mg²⁺ binding stabilizes a closed conformation, effectively preventing ion passage by increasing the energetic barrier for conduction.

The PMF analysis along the RyR1 conduction pathway reveals significant energy variations, particularly in the region spanning from the hydrophobic gate to the cytoplasmic vestibule (z = 7.5 Å to 35 Å, Figure A). This section, located just upstream of the hydrophobic constriction zone, demonstrates a noticeable reduction in energy in the open state (opRyR1), suggesting a more favorable ion conduction environment. This decrease is likely influenced by the negatively charged acidic residues lining the vestibule (Figure B), where facilitate strong interactions with cations as they traverse the pore. Among these residues, carbonyl groups are believed to play a role in stabilizing ion interactions during conduction. A particularly striking feature of the PMF profiles is the behavior of Mg²⁺ which exhibits a deeper energy well (∼ – 5 kcal/mol) compared to other ions, emphasizing its unique interaction with this region. Notably, at the D4945 site (z = 22.5 Å), the energy well becomes significantly deeper, in Mg²⁺-bound inhibited state (HMg²⁺RyR1). The energy for Mg²⁺ drops to approximately – 9 kcal/mol, which is considerably lower than that observed for Ca²⁺ (−6.3 kcal/mol). In contrast, monovalent cations (K⁺ and Na⁺) exhibit only modest energy reductions (∼ – 3 kcal/mol), indicating weaker interactions at this site (Figure A). This pronounced energy minimum strongly suggests that Mg²⁺ becomes effectively trapped at D4945, which aligns with experimental findings by Nayak et al., identifying this residue as the key Mg²⁺-binding site responsible for pore blockage.

Both clRyR1 and HMg²⁺RyR1 exhibit consistently lower energy wells for all ions (Figure A), primarily attributed to a combination of reduced pore radius and enhanced electrostatic attraction between ions and acidic residues (Figures B and B). These structural and electrostatic constraints impose greater restrictions on ion transport, supporting the nonconductive nature of these states. However, the HMg²⁺RyR1 state presents a particularly distinctive energy profile, characterized by a sharper and deeper energy well specifically for divalent ions. This finding supports the hypothesis that Mg²⁺ plays a crucial role in stabilizing the closed state and inhibiting ion conduction by strengthening electrostatic interactions and increasing the energetic barrier for ion permeation. The results clearly illustrate key insights into how ions interact with the RyR1 pore as they traverse its conduction pathway. The type of ion, along with local pore characteristics, such as pore size and residue composition, strongly influences the energy landscape of ion permeation. These effects are further modulated by the conformational state of the protein, highlighting the dynamic interplay between pore structure, electrostatic interactions, and ion conduction efficiency. These findings underscore the critical role of localized residue interactions in shaping the conduction energetically at each key region of the pore, ultimately governing the gating behavior of RyR1.

3.4. Pore-Lining Residues and Their Contributions to Ion Binding

Building upon the PMF results, which provided insights into the energy landscape of ion permeation, we conducted a quantitative assessment of ion binding at key pore-lining residues using quantum mechanical (QM) calculations with density functional theory (DFT). The binding energy (ΔE _bind ) for Ca²⁺, Mg²⁺, K⁺, and Na⁺ was evaluated at selected residues spanning from the selectivity filter (SF) to the cytoplasmic vestibule (Figure and Table S5). These residues including D4899, G4894, Q4933, I4937, D4938, and D4945, were identified as functionally significant based on insights from the PMF profiles and structural characterization of the RyR1 pore.

5.

DFT binding energy (ΔE _bind ) of metal ions (Ca²⁺, Mg²⁺, Na⁺, and K⁺) at key pore-lining residues in RyR1 across three functional states. Error bars represent standard deviation from n independent DFT single-point energy calculations, each based on distinct metal–residue complex configurations extracted from MD trajectories. The corresponding n values for each ion-site combination are listed in Table S6. Representative input geometries used in these calculations are shown in Figure S6. The right panel illustrates the pore regions where each binding site is located.

To complement the energetic analysis and provide structural context for the DFT models, we generated representative visualizations of ion coordination at the D4945 site across all three functional states (Figure S5). These images illustrate the spatial positioning of each ion, nearby D4945 side chains, and coordinating water molecules, as extracted from the MD trajectories used in the DFT calculations. For Ca²⁺ and Mg²⁺, a predefined hydrated model was used, retaining 7 and 6 water molecules in the first solvation shell, respectively. In contrast, for K⁺ and Na⁺ the number of coordinating water molecules was determined based on a 3.0 Å distance cutoff from the ion. The visualizations clarify the variation in coordination geometry and distances that contribute to the observed ΔE _bind values and enhance interpretability of the structural assumptions underlying the QM models.

The ΔE _bind values across three conformational states (opRyR1, clRyR1, and HMg²⁺RyR1) revealed distinct variations in ion-residue interactions, reflecting the influence of pore structure on binding stability (Figure ). Generally, divalent cations (Ca²⁺ and Mg²⁺) exhibited significantly stronger binding than monovalent cations (K⁺ and Na⁺), likely due to greater electrostatic attraction. D4899 at the SF pore mouth exhibited the strongest binding affinity for Mg²⁺, with relatively weaker binding for Ca²⁺ and significantly lower interactions for K⁺ and Na⁺. In some cases, water molecules in the solvation shell of monovalent cations were displaced by carboxylate oxygen atom(s) from D4899, a behavior previously reported for Ca²⁺ in RyR1. G4894 exhibited consistently weak binding for all ions across all conformational states, aligning with the PMF data, which indicate low energy stabilization at this site due to the highly electronegative environment of this region.

Q4933, a key gating residue, demonstrated varying binding affinities, with divalent cations (Ca²⁺ and Mg²⁺) interacting more strongly than monovalent cations (K⁺ and Na⁺) (Figure ). Ion binding was notably stronger in clRyR1 and HMg²⁺RyR1, whereas opRyR1 exhibited weaker interactions, except at D4899, where ion retention remained substantial.

The hydrophobic gate (HG) residue I4937 exhibited positive ΔE _bind values for all ions, indicating unfavorable interactions across all RyR1 states (Figure ). This aligns with the hydrophobic nature of I4937, which contributes to electrostatic repulsion and pore occlusion, resulting in the high energy barrier observed in PMF profiles (Figure A). Conversely, D4938 and D4945 exhibited much stronger binding than I4937 (Figure ), forming a negatively charged environment that enhances cation retention. The wider pore radius in this region, relative to I4937, increases ion accessibility and stabilizes ion-residue complexes. D4945, in particular, exhibited a strong preference for Mg²⁺ over Ca²⁺, a pattern consistent with previous studies and cryo-EM data.

Our findings indicate that localized residue interactions critically shape conduction energetics within the RyR1 pore, influencing ion permeation, and gating. The observed stronger binding of divalent cations (Mg²⁺ > Ca²⁺) compared to monovalent cations (K⁺ and Na⁺) aligns with RyR1’s physiological preference for Ca²⁺ while also supporting Mg²⁺’s inhibitory role. A key observation is the distinct behavior of Mg²⁺ at D4945. The highly negative ΔE _bind values for Mg²⁺ at D4945 suggest that it becomes effectively trapped, consistent with the deep energy well in PMF profiles (Figure A). The strong affinity of Mg²⁺ for D4945 supports its role in blocking RyR1 ion conduction, likely through the formation of a stable coordination environment involving a hexahydrated Mg²⁺ ion stabilized by interaction with the four D4945 residues in the pore. Further insights into pore-region-specific ion interactions emerge when considering Q4933, I4937, and D4938/D4945. The weaker binding of ions at Q4933 in opRyR1 is likely due to a wider pore radius, reducing local electrostatic attraction (Figure ). I4937, forming the hydrophobic gate (HG), predictably exhibits weak cation binding, reinforcing its role as a high-energy barrier to ion permeation (Figure A). In contrast, D4938 and D4945 provide strong electrostatic stabilization for divalent cations, further enhancing Mg²⁺ retention and pore blockage in HMg²⁺RyR1. This preference, combined with cryo-EM studies and experimental data, strongly supports the role of D4945 in Mg²⁺-mediated RyR1 inhibition. In the HMg²⁺RyR1 state, Mg²⁺ binding likely occludes the permeation pathway, acting as a structural blockade to ion flux. This is further reinforced by the PMF results, which show low Mg²⁺ potential energy in the HMg²⁺RyR1 state, suggesting that Mg²⁺ is energetically trapped in a stable complex. This conformational and electrostatic trapping mechanism provides a structural basis for Mg²⁺-mediated inhibition, as previously suggested by cryo-EM studies and functional assays.

3.5. Mg²⁺ Binding at the D4945 Site

To further investigate ion-residue interactions in the RyR1 pore, we analyzed the equilibrated configurations of Mg²⁺ bound at D4945 across three functional states: opRyR1, clRyR1, and HMg²⁺RyR1 states. Binding free energy calculations using the MM-GBSA method were performed to construct the binding free energy (ΔG _bind ) landscape of Mg²⁺ binding at D4945, projected onto the x–y plane corresponding to Mg²⁺ positions (Figure A). These contour maps provide insights into the stability and spatial distribution of Mg²⁺ within the D4945 site. These maps represent the free energy surface associated with Mg²⁺ binding at D4945, providing insights into both binding stability and characteristics.

6.

(A) 2D contour maps of Mg²⁺ binding free energy (ΔG _bind ) referenced to Mg²⁺ at D4945 as shown in the right panel with S6 helices (gray) and D4945 (red beads). (B) Binding free energy comparisons for opRyR1, clRyR1, and HMg²⁺RyR1, with electrostatic (ΔG _bind ) and nonpolar (ΔG _bind ) contributions. Boxes indicate the interquartile range, median, mean, and the 1st–99th percentile range.

As shown in Figure A, the binding landscape varies significantly between RyR1 states. In the opRyR1 system, the energy surface is broad and shallow, indicating weak Mg²⁺-D4945 interactions that allow greater ion mobility. This results in a diffuse binding region with a larger spread of Mg²⁺ positions and relatively higher ΔG _bind values. In contrast, the clRyR1 and HMg²⁺RyR1 states display a more localized and deeper energy surface, reflecting stronger Mg²⁺-D4945 interactions that restrict ion movement. The HMg²⁺RyR1 state exhibits the lowest Mg²⁺ dynamics and the most favorable binding free energy, indicating a more stable ion-protein interaction, consistent with previous findings by Nayak et al. The average ΔG _bind follow the trend: opRyR1 > clRyR1 > HMg²⁺RyR1 (Figure B), indicating progressively stronger Mg²⁺ retention in the closed and inhibited states. Decomposition of ΔG _bind into electrostatic (ΔG _bind ) and nonpolar (ΔG _bind ) contributions reveals that electrostatic interactions are the dominant stabilizing force to Mg²⁺ binding at D4945. Specifically, the ΔG _bind component is significantly more negative in the HMg²⁺RyR1 states, suggesting that electrostatic attraction between Mg²⁺ and D4945 is the primary determinant of binding stability, while the nonpolar component (ΔG _bind ) remains relatively unchanged across all states.

To evaluate the stability and coordination environment of the hydrated Mg²⁺ ion in the HMg²⁺RyR1 system, we analyzed the RDFs between Mg²⁺ and surrounding water oxygen atoms as well as the carboxylate oxygens of D4945 residues (Figure S7). The Mg²⁺–O­(H₂O) RDF showed a sharp first-shell peak at ∼ 2.1 Å across all three systems, confirming the persistence of hexahydrated coordination during the simulations. In contrast, the Mg²⁺–O­(D4945) RDF revealed a broader peak in the range of ∼ 4–6 Å, characteristic of second-shell interactions mediated by water or flexible side chains. Notably, the coordination number for Mg²⁺–O­(D4945) increased to approximately 4 in the HMg²⁺RyR1 system, whereas only 1–2 D4945 residues contributed to coordination in the opRyR1 and clRyR1 systems. This indicates that in the Mg²⁺-inhibited state, all four D4945 residues simultaneously interact with the hydrated Mg²⁺ ion, albeit through a more loosely organized, water-bridged network. The enhanced stabilization at this site reflects the specific ionic environment observed in the cryo-EM structure of the inhibited state and supports the functional role of D4945 in mediating Mg²⁺-dependent inhibition.

D4945 has been proposed as a second-shell ligand, indirectly stabilizing Mg²⁺ via hydrogen bonding with first-shell water molecules. This indirect coordination mechanism distinguishes Mg²⁺ binding at D4945 from direct chelation, allowing for greater ion mobility depending on the conformational state of RyR1. Our findings indicate that in HMg²⁺RyR1, D4945 engages in strong interactions, effectively retaining Mg²⁺ at the site. This reinforces the role of Mg²⁺ as a physiological inhibitor, as its strong binding likely contributes to pore occlusion and ion conduction blockage. This analysis demonstrates that Mg²⁺ binding strength correlates with pore constriction, with the electrostatic component playing the dominant role in stabilization. These findings provide deeper insights into how localized residue interactions shape conduction energetics within the RyR1 pore, directly influencing ion permeation and gating mechanisms.

4. Discussion

The complex mechanisms governing ion permeation, selectivity, and blockage in the RyR1 pore provide critical insights into its functional regulation across different conformational states. By integrating multiple computational approaches, including pore property analysis, PMF profiles, and binding energy calculations using quantum mechanics and MM-GBSA, this study presents a comprehensive assessment of the energetic and structural factors shaping ion behavior within the RyR1 conduction pathway. Structural insights from cryo-EM studies by Nayak et al. revealed RyR1 in multiple conformational states, including open (opRyR1), closed (clRyR1), and high Mg²⁺ concentration states (HMg²⁺RyR1). In the HMg²⁺RyR1 state, cryo-EM identified multiple Mg²⁺ binding sites, directly correlating with Mg²⁺-mediated inhibition of RyR1 function. Notably, four D4945 residues within the RyR1 pore were found to coordinate a Mg²⁺-D4945₄ complex, effectively blocking ion permeation. This Mg²⁺ binding-induced obstruction highlights a key mechanism of RyR1 inhibition, where the pore is physically occluded, preventing ion flux and stabilizing the nonconductive state.

Our computational approaches provided valuable insights into the structural and functional properties of the RyR1 pore, revealing both distinct and shared features among the opRyR1, clRyR1, and HMg²⁺RyR1 states. Notably, the HMg²⁺RyR1 structure exhibits a significantly narrowed pore radius at the hydrophobic gate I4937, reducing to less than 1 Å (Figure A-B and S3). This observation strongly suggests that, under high Mg²⁺ conditions, the pore structure corresponds to a closed channel, implying that ion conduction is highly restricted or unlikely to occur in this state. Interestingly, the HMg²⁺RyR1 structure features an even narrower pore radius between the gate (Q4933) and the hydrophobic gate (I4937) compared to clRyR1 (Figure A-B). This additional constriction is associated with a pronounced reduction in water solvation, a sharp contrast to the fully open state (Figure ). The decrease in water occupancy in this region correlates with an increase in solvation energy (Figure B and C), further emphasizing the notion that Mg²⁺ inhibits ion conduction by stabilizing a structurally and energetically unfavorable pore conformation.

Further insights into Mg²⁺-induced RyR1 inhibition were obtained through free energy analysis of ion translocation, particularly via potential of mean force calculations (Figure A). These simulations revealed distinctive PMF patterns along with the RyR1 conduction pathway, which varied depending on the functional states of the channel. Key free energy peaks were observed at critical regions of the pore, including the selectivity filter (SF), the SF-gate junction, the gate (G), the hydrophobic gate (HG), and the cytoplasmic vestibule. The corresponding key residues including D4899, G4894, Q4933, I4937, D4938, and D4945, play essential roles in shaping the energy landscape of ion permeation. A key finding from these analyses in that in nonconductive states (clRyR1 and HMg²⁺RyR1), the PMF profile exhibits a sharp energy increase at I4937, highlighting a significant energy barrier associated with pore closure. Furthermore, Mg²⁺ translocation within the HMg²⁺RyR1 pore was markedly hindered, as evidenced by a deep free energy well at the D4945 site (Figure A). This suggests that Mg²⁺ becomes trapped at D4945, emphasizing its role in pore blockage and RyR1 inhibition. These findings provide a detailed mechanistic understanding of how Mg²⁺ stabilizes a closed RyR1 conformation, ultimately preventing ion permeation. The evaluation of ΔE _bind across key regions of the RyR1 pore provides valuable insight into ion-protein interactions (Figure ). Together with the electrostatic potential map analysis (Figure B), these findings highlight the energetic costs associated with ion interactions in distinct pore regions. Our results demonstrate that the SF and G regions serve as favorable sites for attractions, primarily due to the presence of polar residues D4899 and Q4933, which contribute to significantly lower ΔE _bind values. This trend is consistently observed across all three functional states of RyR1, emphasizing the role of these regions in facilitating ion permeation. In contrast, the HG region (I4937) serves as an energetic barrier, a consequence of its hydrophobicity nature, which impedes ion transport. This effect is particularly pronounced in nonconductive states (clRyR1 and HMg²⁺RyR1), where the binding energy differences between states further reinforce the inhibitory role of the hydrophobic gate.

The pulling path used in SMD was aligned with the pore axis from cryo-EM structures, representing the most probable central conduction pathway in RyR1. As forced ion displacement can disrupt solvation, we employed a slow pulling speed of 1 Å/ns and a soft spring constant of 5 kcal·mol^–1·Å^–2 to minimize such artifacts. These setting allowed ions to retain their hydration shells throughout the trajectory. The smooth, linear ion displacement over time (Figure S4C) confirms the stability of the pulling process and the absence of abrupt force-induced disruptions. Importantly, the SMD simulations were not analyzed in isolation; instead, they provided a series of intermediate configurations used to initiate PMF calculations (Figure ). Thus, the resulting PMF profiles reflect equilibrium free energy landscapes, enabling a more accurate and thermodynamically meaningful assessment of ion permeation through RyR1.

At the cytoplasmic vestibule, the negatively charged residues D4938 and D4945 play a crucial role in attracting ions and facilitating their initial binding. These residues lower the energy barrier after ion cross the hydrophobic gate (Figure A). This electrostatic attraction likely acts as a driving force, guiding ions toward the conduction pathway and enhancing their permeation efficiency. As described above, the negative charge distribution in this region is fundamental to regulating ion flux, and our findings further support the critical role of these residues in RyR1 ion permeation. In addition, mutations at these residues can disrupt RyR1 function, potentially leading to malignant hyperthermia (MH) and central core disease (CCD). ,,

In interpreting the energetic and structural roles of D4938 and D4945, it is important to distinguish between transient ion association and stable ion retention. Our DFT calculations (Figure ) indicate that D4945 exhibits a stronger binding affinity for divalent cations, particularly Mg²⁺, than D4938, which aligns with its location in the wider and more solvent-accessible cytoplasmic vestibule. This strong electrostatic interaction underlies D4945’s key role in Mg²⁺-mediated inhibition, effectively stabilizing the closed conformation by forming a high-affinity barrier to ion passage. In contrast, D4938 resides closer to the hydrophobic gate and contributes less prominently to ion retention, likely due to steric constraints and a more restrictive local environment.

However, the relevance of D4938N, which has been experimentally linked to altered Ca²⁺ selectivity and RyR1 dysfunction, can be rationalized by considering its functional role in the initial stages of ion entry. Rather than acting as a deep binding pocket, D4938 may facilitate transient coordination or steering of Ca²⁺ ions into the conduction pathway. Its strategic location near the gate likely influences local electrostatics and the energetic landscape of ion selection, thereby playing a regulatory role distinct from that of D4945. Thus, while D4945 is a stronger binding site in static energy terms, D4938 may exert a subtler but significant influence on Ca²⁺ discrimination,.

Additionally, the functional consequences of D4945 mutations, such as D4945N, merit closer consideration. Prior electrophysiological studies have demonstrated that neutralizing this residue increases K⁺ conductance and diminishes Ca²⁺ selectivity. Our findings provide a mechanistic rationale for this observation: D4945 serves as a critical electrostatic anchor for divalent ions, and its mutation would weaken this attraction, lowering the energy barrier for monovalent cation passage and disrupting selective ion retention. Furthermore, our PMF and MM-GBSA analyses reveal that Mg²⁺ binding at D4945 is essential for pore occlusion in the inhibited state. Loss of this interaction would likely impair Mg²⁺-mediated inhibition, potentially increasing basal RyR1 activity and contributing to dysregulated Ca²⁺ homeostasis. These insights underscore the dual functional role of D4945 in both ion selectivity and inhibitory gating and suggest that mutations at this site could have profound physiological consequences.

This conformation, as captured in cryo-EM structures and supported by our simulations, reveals a distinct electrostatically stabilized closure of the pore mediated by Mg²⁺ binding at D4945. The enhanced ion binding observed in HMg²⁺RyR1, specifically the sharp energy well for Mg²⁺ at D4945 and the pronounced pore narrowing, suggests that Mg²⁺ actively induces and stabilizes a functionally inhibited state rather than passively occupying an already closed conformation. This inhibitory mechanism plays a protective role by preventing inadvertent Ca²⁺ release under basal conditions.

Clinically, disruption of this Mg²⁺-mediated inhibition, either through mutations that diminish D4945 coordination or through reduced cytosolic Mg²⁺ availability, can weaken the structural integrity of the inhibited state. This may lead to pathological channel leakage, contributing to Ca²⁺ dysregulation implicated in MH and CCD. Thus, our results highlight the importance of the Mg²⁺-bound inhibited conformation as a distinct regulatory state, complementing the open and closed configurations in the broader functional landscape of RyR1.

The ΔE _bind analysis for Mg²⁺ at D4945 in HMg²⁺RyR1 revealed the strongest binding affinity, exhibiting the lowest binding energy among to other ions (Ca²⁺, K⁺ and Na⁺) and RyR1 states, highlighting the stability of the Mg²⁺-D4945 complex. Consistently, the MM-GBSA-derived energy landscape at D4945, shows that Mg²⁺ binding in HMg²⁺RyR1 is the most stable, as indicated by its tightly clustered spatial distribution and the most favorable ΔG _bind values (Figure ).

The strongly negative binding energy at this site indicates a high-affinity interaction that effectively immobilizes Mg²⁺ within the pore, supporting its role in pore blockage and channel inhibition. These findings align with cryo-EM structures, computational results and electrophysiology studies, ,,,, further emphasizing the potent inhibitory capacity of Mg²⁺. Consistent with previous studies, our results confirm that Mg²⁺ acts as a physiological inhibitor by stabilizing a nonconductive pore conformation. To summarize these findings, Figure provides a schematic representation of the HMg²⁺RyR1 pore domain, illustrating how Mg²⁺ interacts with various regions of the pore, leading to trapping and inhibition. This figure visualizes key regions where Mg²⁺ bindings alter the structural landscape, demonstrating how pore rearrangements facilitate Mg²⁺ retention and ultimately block ion conduction.

7.

Schematic of the HMg²⁺RyR1 permeation pathway and Mg²⁺ inhibition mechanism. Hydrated Mg²⁺ becomes trapped at D4945, where strong binding stabilizes the S6 helices, inducing narrowing of the closed gate (Q4933) and hydrophobic gate (I4937). This further restricts ion permeation and limits water accessibility. The diagram highlights key structural features, ion interactions, and pore constriction in the inhibited state.

5. Conclusions

This study provides detailed mechanistic insights into Mg²⁺-mediated inhibition of the RyR1 channel through a multiscale computational approach that integrates molecular dynamics (MD), potential of mean force (PMF) calculations, Poisson–Boltzmann solvation energy analysis, MM-GBSA binding free energy estimation, and quantum mechanical (DFT) modeling. By analyzing three functional states of RyR1, Ca²⁺-activated (opRyR1), closed (clRyR1), and Mg²⁺-inhibited (HMg²⁺RyR1), we elucidated how pore geometry, hydration behavior, and ion-residue interactions differ across functional conformations. A key contribution of this work lies in the identification of Mg²⁺-specific energetic features that stabilize the nonconductive conformation. PMF analysis revealed that Mg²⁺ encounters the highest permeation barrier among all studied ions and exhibits a pronounced energy well at the D4945 site, indicating strong retention within the cytoplasmic vestibule. DFT and MM-GBSA results further demonstrated that this interaction is driven by a hexahydrated Mg²⁺ core stabilized by second-shell coordination with D4945 residueshighlighting an inhibitory mechanism based on electrostatic and hydration-driven trapping rather than direct chelation. These structural and energetic features distinguish the Mg²⁺-bound inhibited state from the closed apo conformation, revealing a mechanistically distinct mode of channel regulation. Importantly, our findings showcase how computational techniques can resolve dynamic and energetic properties that are difficult to capture through cryo-EM or mutagenesis alone. The hydrated ion behavior, coordination environment, and ion-specific gating energetics revealed here deepen our understanding of RyR1’s regulatory landscape and provide a foundation for future work on pathogenic variants and therapeutic interventions. Rather than focusing on specific disease states, this work contributes broadly to the mechanistic modeling of large ion channels and highlights the utility of advanced computational strategies in uncovering regulatory principles that are otherwise experimentally inaccessible.

The molecular modeling and simulation tools used in this study include PROPKA, NAMD, VMD, HOLE, and APBS, all available for noncommercial use under distribution-specific licenses. Quantum calculations were performed using the Gaussian 09 software package under an academic license. Custom analysis scripts were developed in TCL and executed within VMD; these scripts were specifically created in our laboratory for this research. All software tools are referenced in the Methods and Results sections. Simulation trajectories, structural files, and analysis scripts generated in this study will be made available on request.

The Supporting Information is available free of charge on the Web site at DOI: The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.5c03018.

• RMSD analyses show structural stability and no cross-state convergence (Figure S1); comparison of pore radius profiles and structural alignments (Figure S2); pore shape and diameter inside the cryo-EM structures of RyR1 (Figure S3); protein stability during SMD simulations (Figure S4); representative configurations of Ca²⁺, Mg²⁺, Na⁺ and K⁺ at the D4945 site (Figure S5); representative configurations of ion–residue complexes used in DFT calculations (Figure S6); RDF analysis of Mg²⁺ coordination with water and D4945 (Figure S7); full ΔG _solv and ΔG _elec profiles along the RyR1 pore axis (Figure S8); predicted pK _a values of ionizable residues (Table S1); simulated systems for MD of the RyR1 pore domain (Table S2); simulated systems for SMD of the RyR1 pore domain, followed by ABF calculations (Table S3); summary of residues lining the interior of the RyR1 pore used in solvation energy analysis (Table S4); summary of average binding energy (ΔE _bind ) of the different ion-residue complexes (Table S5); number of distinct metal–residue complex models (Table S6) (PDF)

P.B. performed computational experiments and interpreted the data, P.T. and P.L. contributed to experimental design and assisted in discussion. P.B. and P.S. wrote the main manuscript text and prepared all figures, R.B.P. and M.S. provided critical discussions and valuable intellectual input. All authors reviewed the manuscript.

Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the Rights Link permission system: http://.

The authors declare no competing financial interest.