Structural basis for rosuvastatin-induced activation of skeletal muscle ryanodine receptor 1 Authors: Gunnar Weninger, Steven Reiken, Qi Yuan, Marco C Miotto, Alexander Chang, Yang Liu, Anetta Wronska, Haikel Dridi, Andrew R Marks Published: 2026 ABSTRACT AbstractStatin therapies, used to lower high blood cholesterol, have been proposed to cause skeletal muscle side effects including myalgia and rhabdomyolysis. The ryanodine receptor 1 (RyR1)/Ca²⁺ release channel in skeletal muscle has recently been identified as an off-target binder of statins. Here, we report the cryo-electron microscopy structure of RyR1 in complex with rosuvastatin. Rosuvastatin occupies a single binding site in the pore region of RyR1 stabilizing the open conformation. In the closed conformation, the rosuvastatin binding site is occupied by the RyR1 auxiliary transmembrane helices (TMx1-2). Structural comparison indicates steric overlap of rosuvastatin with simvastatin (at Sim-1) but not with atorvastatin on RyR1, while all these statins may interfere with the conformational stability of the TMx1-2 helices. In contrast to simvastatin and atorvastatin, rosuvastatin does not exhibit multiple binding sites on RyR1 while similarly promoting RyR1 activation, indicating its potential as a preferable lead structure for developing statins that minimizes adverse side effects on RyR1.Subject terms: Calcium channels, Cryoelectron microscopy, Cryoelectron microscopy Introduction Statins, used to treat high cholesterol, have been associated with adverse drug reactions in skeletal muscle. These Statin Associated Muscle Symptoms (SAMS) range from myalgia and muscle weakness to rhabdomyolysis1–3. However, conflicting reports state the incidence of SAMS as ranging from 0 to ~10% of patients on statins4–7. Given that statins are among the most widely prescribed medications8,9, even a marginal SAMS incidence rate would affect a substantial number of individuals, constituting a meaningful public health burden; at the higher end of estimates, the absolute numbers could be quite large. The dispute appears to be centered around whether the SAMS are measured in clinical trials, which tend to show little or no increase in SAMS, or in observational studies, which tend to report ~10% incidence of SAMS6,7. One prominent off-target effector of statins is the ryanodine receptor 1 (RyR1)/Ca2+ release channel crucial for excitation-contraction coupling in skeletal muscle10–12. Pathogenic mutations of RyR1 (such as T4709M) may promote simvastatin intolerance13,14. Using cryo-electron microscopy (cryo-EM), we recently identified two simvastatin binding sites in the pore region of RyR1: a higher-affinity site situated between transmembrane (TM) helices S3 and S4 (Sim-1) and a lower-affinity site near the N-terminal S6 region (Sim-2)14. Because of their structural similarities with simvastatin, lovastatin and pravastatin (type-1 statins) are predicted to affect the structure-function relationship of RyR1 in a comparable manner10,14. Atorvastatin and rosuvastatin (type-2 statins) differ structurally from simvastatin in having more diverse hydrophobic core moieties (pyrrole in atorvastatin, pyrimidine in rosuvastatin, hexahydro-naphthalene in simvastatin), and a 4-fluorophenyl group as a characteristic substituent of type-2 statins10,14. Atorvastatin and rosuvastatin have been shown to activate RyR1 similarly to simvastatin10. A recent report revealed that atorvastatin occupies three clustered binding sites (forming triplets) within the pore region of RyR115 that do not overlap with Sim-1 or Sim-214. The 4‑fluorophenyl moieties of all three bound atorvastatin molecules engage in extensive hydrophobic interactions with RyR1, suggesting that other type-2 statins could interact with RyR1 in a similar manner15. However, to the best of our knowledge, structural data about other type-2 statins have been missing until now. Here, we provide the structural basis for the adverse off-target effects of rosuvastatin on skeletal muscle RyR1. Results Structure of RyR1 in complex with rosuvastatin To analyze the RyR1 structure in complex with rosuvastatin, endogenous RyR1 was purified from rabbit skeletal muscle and exposed to 2 mM rosuvastatin lactone before vitrification under RyR1 channel-activating conditions (30 µM free Ca2+, 5 mM caffeine, and 10 mM ATP). The samples were assessed by cryo-EM single-particle analysis. Rosuvastatin binding yielded ~49% open particles, separated by 3D classification (cryoSPARC) (Supplementary Fig. 1, Supplementary Table 1). In the absence of statin, we previously reported ~10% open particles for RyR1 purified from mouse skeletal muscle using similar sample preparation conditions and cryo-EM workflows14. The structure of RyR1 in the open state was reconstructed at 2.9 Å global and 2.6 Å local resolution in the pore region (Supplementary Fig. 2) exhibiting one rosuvastatin molecule per protomer bound to the Sim-1 site (Fig. 1a). Local resolution estimation (cryoSPARC) of the Sim-1 site indicated that rosuvastatin itself was resolved at ~3.0 Å, while the surrounding binding pocket reached ~2.6 Å resolution (Supplementary Fig. 3). The Sim-2 and atorvastatin binding sites remained unoccupied. At the Sim-1 site, rosuvastatin displaced the W4794 side chain (Supplementary Fig. 4), exposing a hydrophobic pocket between TM helices S3 and S4, into which the 4-fluorophenyl group of rosuvastatin binds tightly (Fig. 1a). Structural comparison indicates that the binding position of the 4-fluorophenyl group overlaps with the 2,2-dimethylbutyrate ester group of simvastatin within the hydrophobic pocket of the Sim-1 site (Fig. 1b). Within the pocket, hydrophobic interactions with the 4-fluorophenyl group of rosuvastatin or the 2,2-dimethylbutyrate ester group of simvastatin are mediated by residues Y4791, Y4795, and M4798 of S3, as well as A4811-H4812 of S4 (Fig. 1b and Supplementary Fig. 5). These hydrophobic interactions likely drive statin binding to the Sim-1 site. Structure of RyR1 with rosuvastatin. a Side view of RyR1 in the open conformation. Rosuvastatin (orange) binds to the pore region between transmembrane helices S3 and S4. In the magnification, the atomic model indicates that the 4-fluorophenyl group (4-FPh) of rosuvastatin occupies the hydrophobic pocket formed by residues Y4791, Y4795, and M4798 of S3, and H8412 of S4. The lactone group interacts with S3 at L4790-Y4791 near the cytosolic TMD surface. The methyl sulfonamide group (Me-SA) interacts with S4 at residues L4814, A4817, and M4818, as well as the S4S5 linker at L4827. The pyrimidine core moiety putatively makes contacts with acyl chains of PC lipids. The density map (gray) of the rosuvastatin binding site is shown at a contour level of 5.3σ. b Structural comparison of the rosuvastatin (orange) and simvastatin (green, Sim-1) binding sites revealing steric overlap. Both statins share similar binding poses for the lactone group. 4-FPh sterically overlaps with the 2,2-dimethylbutyrate ester group of simvastatin in the hydrophobic pocket. c Side view of TMD of RyR1 in the closed conformation exhibiting the auxiliary transmembrane helices TMx1 and TMx2 (blue) but no rosuvastatin. d Atomic model showing TMx1 and TMx2. e Rosuvastatin binding site occupied by TMx1 in the closed conformation (circle). Moreover, the lactone groups of both statins adopt comparable binding modes at L4790-Y4791 of S3 near the cytosolic interface of the transmembrane domain (TMD; Fig. 1b). Given that the lactone group is not tightly encased by the Sim-1 site, the acid form of rosuvastatin could adopt a binding pose similar to the lactone form without steric clashes. The more extended hydrophilic dihydroxyheptanoic acid group (opened lactone ring) would putatively protrude toward N4787 of S3 at the cytosolic TMD surface. This suggests that rosuvastatin binds to RyR1 independently of lactone ring hydrolysis. In contrast to simvastatin, rosuvastatin binding involves additional interactions mediated by the methyl sulfonamide group with the S4 helix at L4814 and A4817 (Fig. 1b and Supplementary Fig. 5). Of note, the density for the methyl sulfonamide group is less well resolved, implying a higher local mobility (Supplementary Fig. 3b). The hydrophobic pyrimidine core moiety of rosuvastatin putatively makes contacts with the acyl chains of phosphatidylcholine (PC) lipids (Fig. 1b). Comparing RyR1 with and without rosuvastatin, we detected no significant alteration in the S3 and S4 helical arrangement forming the Sim-1 site (Supplementary Fig. 6). However, when aligned to helices S1 to S4, the S4S5 linker exhibited a trend of lateral movement (~0.7 Å) toward S5 upon rosuvastatin binding (Supplementary Fig. 7) that might support pore opening. In line with this, RMSD comparison of the transmembrane helices in the presence versus absence of rosuvastatin showed a slightly increasing trend from S4 to S6 (Cα-RMSD: ~0.7 Å for S6; Supplementary Fig. 8a). The cytosolic shell remained structurally unaltered (Supplementary Fig. 8a). Quantitative analysis of the pore radii along the conduction pathway of the RyR1 channel (calculated using HOLE) confirmed the closed and open conformations (minimum pore radius: ~0.8 versus ~3.0 Å, respectively) separated by 3D classification but showed no substantial differences between RyR1 in the presence or absence of rosuvastatin (Supplementary Fig. 8c). Taken together, rosuvastatin binding enhanced occupancy of the open state (~49% open particles; Supplementary Fig. 1), yet did not produce substantial structural alterations in the open conformation of RyR1 (Supplementary Fig. 8a). In the closed conformation of RyR1 reconstructed at 2.7 Å resolution, no density corresponding to rosuvastatin was detected (Fig. 1c–e). Notably, in the pore region, the W4794 side chain of S3 remained in the altered conformation, thereby exposing the hydrophobic pocket of the Sim-1 site. The Sim-1 site was occupied by the auxiliary transmembrane helix TMx1 (Fig. 1e and Supplementary Fig. 9), an otherwise unstable structural element that, to the best of our knowledge, has not previously been reported. TMx1 is situated between S3 and TMx2 (Fig. 1d and Supplementary Fig. 8b), a second auxiliary transmembrane helix previously designated as TMx14,16 or S017. The TMx1 density is not contiguous with other densities in the pore region, while the C-terminal region of TMx1 is largely exposed at the protein-membrane interface, suggesting that TMx1 could be (part of) an unidentified RyR1-associated protein that would bind to the Sim-1 site; however, aside from Calstabin, no additional proteins were detected in this or previous RyR1 preparations using similar purification protocols14,18–21. Notably, rosuvastatin was added post‑purification and therefore could not have affected the purification. TMx1 is thus more likely intrinsic to RyR1, putatively corresponding to the previously predicted transmembrane hairpin at residues ~4280–430022, an alanine-rich sequence conserved across mammalian species but not across RyR isoforms (Supplementary Fig. 10). Stabilization of TMx1 induced a conformational shift of TMx2 next to it (Supplementary Fig. 11), while the S4 helix shifted slightly (Cα-RMSD: ~0.8 Å) toward the S4S5 linker compared to the statin-free condition (Supplementary Fig. 8b). Otherwise, no substantial structural alterations were detected in the pore or cytosolic shell attributable to TMx1-2 (Supplementary Fig. 8b, c). Interestingly, both the occupied Sim-1 and Sim-2 sites would sterically clash with TMx1 and TMx2, respectively (Supplementary Fig. 12). This suggests that statins may interfere with TMx1-2 conformations and stability. Given their location at the TMD-lipid interface inside the membrane, the TMx1-2 helices might be involved in interactions with other SR membrane proteins forming larger RyR heterocomplexes, such as junctin, triadin, or junctophilin23,24. Dose-dependent activation of RyR1 by rosuvastatin The rosuvastatin dose-response of RyR1 was assessed using single-channel recordings at 0.15 µM free Ca2+ (Fig. 2 and Supplementary Fig. 13), in which recombinant WT RyR1 channels were reconstituted in planar lipid bilayers and then subjected to increasing rosuvastatin lactone or acid concentrations sequentially in three 10-fold steps from 0.1 to 100 μM. Rosuvastatin lactone significantly increased the open probability (PO) from 0.3% at baseline to 7.6% at 1 µM rosuvastatin, 13.1% at 10 µM rosuvastatin, and 31.5% at 100 µM rosuvastatin (Fig. 2, left). Compared to the lactone form, the acid form of rosuvastatin yielded similar dose-dependent increases in PO (Fig. 2, right). Thus, at the low free Ca2+ condition, substantial RyR1 activation required low micromolar concentrations of rosuvastatin in both its lactone and acid forms. Interestingly, simvastatin has been previously reported to activate RyR1 in single-channel recordings in a similar dose-dependent fashion14. Single-channel recordings of RyR1 treated with rosuvastatin. Recombinant RyR1 channels were prepared from HEK293 cells after transient transfection and reconstituted in planar lipid bilayers. Current traces (top) were recorded at 150 nM free Ca2+ without (control) or with rosuvastatin in its lactone (left) or acid (right) form, and the corresponding open probabilities (bottom) were quantified from three independent channel recordings (N = 3 biological replicates). Single-channel traces shown are from the same RyR1 channel treated with increasing rosuvastatin concentrations (from 0.1 µM to 100 µM). Channel openings in current traces of RyR1 are represented as upward deflections, while baseline currents correspond to the closed state (c). Upon addition of rosuvastatin at micromolar concentrations, the open probability (PO) of RyR1 significantly increased (reaching a ~23-fold increase at 1 µM rosuvastatin lactone, ~40-fold at 10 µM, and ~95-fold at 100 µM). The acid form of rosuvastatin yielded a comparable dose-response curve to that of the lactone form. Data are expressed as mean ± SEM. Statistical significance was determined by one-way repeated measures ANOVA with a Greenhouse-Geisser correction (rosuvastatin lactone: F(1.25, 2.52) = 56.00, P = 0.009, ε = 0.32; rosuvastatin acid: F(1.09, 2.18) = 77.59, P = 0.010, ε = 0.27), followed by multiple comparisons against the control using two-sided paired t-tests with Holm-Bonferroni adjusted P values. Significance is indicated as *P < 0.05 (for rosuvastatin lactone: t(2) = 12.22, P = 0.027 at 0.1 µM; t(2) = 5.93, P = 0.036 at 1 µM; t(2) = 9.04, P = 0.036 at 10 µM; and t(2) = 8.26, P = 0.036 at 100 µM; Hedges’ g ranged from 3.79–5.72; for rosuvastatin acid: t(2) = 3.00, P = 0.096 at 0.1 µM; t(2) = 7.99, P = 0.031 at 1 µM; t(2) = 14.09, P = 0.020 at 10 µM; and t(2) = 9.79, P = 0.031 at 100 µM; Hedges’ g ranged from 2.50 to 9.66). The rosuvastatin-induced RyR1 activation was further assessed in [3H]ryanodine binding assays at 10 µM free Ca2+ (Fig. 3), in which ryanodine binding positively correlates with RyR1 activity because ryanodine preferentially binds to the open conformation of the channel10,14. The corresponding [3H]ryanodine binding curves exhibited a sigmoidal shape, with a half maximal effective concentration (EC50) of 0.49 ± 0.07 μM and a maximum fold change (Emax) of 1.98 ± 0.02 for recombinant WT RyR1 with rosuvastatin lactone (Fig. 3a). No significant difference was detected for the acid form of rosuvastatin (EC50: 0.50 ± 0.12 µM, Emax: 2.02 ± 0.04, at 10 µM free Ca2+; Fig. 3b). Interestingly, WT RyR1 exhibited similar a dose-response behavior to simvastatin (acid form: EC50: 0.73 ± 0.09 μM, Emax: 2.08 ± 0.02, at 10 µM free Ca2+) in previously reported [3H]ryanodine binding data14. In parallel, we examined the rosuvastatin dose-responses of the binding site mutants BSM2 (Y4791A + W4794A in rabbit RyR1) and BSM3 (Y4912A) that have previously been shown to impair simvastatin binding to the Sim-1 and Sim-2 sites, respectively14. BSM2 but not BSM3 almost completely suppressed the dose-response to rosuvastatin, confirming that rosuvastatin binds to the Sim-1 site but not to the Sim-2 site (Fig. 3a, b). Next, we analyzed atorvastatin in the [3H]ryanodine binding assays. Compared to rosuvastatin, atorvastatin acid exhibited a similar dose-response of WT RyR1 (EC50: 0.98 ± 0.19 μM) with a slightly higher maximum fold change (Emax: 2.32 ± 0.05) (Fig. 3c). The BSM2 and BSM3 mutations did not affect the atorvastatin dose-responses, confirming that atorvastatin does not occupy the Sim-1 and Sim-2 sites14,15. Dose-responses of RyR1 to rosuvastatin and atorvastatin in [3H]ryanodine binding assays. a–c [3H]Ryanodine binding data are shown for recombinant WT RyR1 (green) and binding site mutants of the Sim-1 site (BSM2: Y4791A + W4794A; yellow) and Sim-2 site (BSM3: Y4912A; red) at 10 µM free Ca2+. Dose-responses to rosuvastatin lactone (a), rosuvastatin acid (b) and atorvastatin acid (c) were assessed at concentrations ranging from 0.01 to 100 µM. Biological quadruplicates (N = 4) are reported for each sample, with individual values shown as single data points. Data are expressed as mean ± SEM (left). Emax and EC50 values for rosuvastatin and atorvastatin were determined from the [3H]ryanodine binding curves of WT RyR1, BSM2, and BSM3 (right). The dose-responses to rosuvastatin lactone and acid were similar and were almost completely suppressed by BSM2 (Sim-1 site), but not by BSM3 (Sim-2 site). These binding site mutants did not substantially alter the atorvastatin dose-response. Source data are provided as a Source data file. To assess interactions of the lactone group or dihydroxyheptanoic acid group (opened lactone ring) in the rosuvastatin binding site, the RyR1 mutants BSM4 (L4790A) and BSM5 (N4787A + L4790A), targeting residues predicted to interact with the rosuvastatin lactone group, were subjected to [3H]ryanodine binding assays (Supplementary Fig. 14). The dose-responses of BSM4 and BSM5 to rosuvastatin in its lactone or acid forms resembled those of WT RyR1. For rosuvastatin lactone, BSM5 showed a trend toward decreased affinity, as indicated by a slightly increased EC50 value (1.09 ± 0.16 µM for BSM5 versus 0.49 ± 0.07 µM for WT). Thus, BSM4 and BSM5 did not suppress rosuvastatin binding as efficiently as BSM2, suggesting that the residues Y4791 and W4794 of S3 (both mutated in BSM2) are of higher relevance for the interaction than N4787A and L4790A (both mutated in BSM5). To assess rosuvastatin binding directly, RyR1 was incubated with radiolabeled rosuvastatin acid at concentrations ranging from 0.01 to 50 μM in radioligand binding assays (Supplementary Fig. 15). The corresponding dose-response curve showed a typical sigmoidal shape with an estimated dissociation constant (KD) of 0.49 ± 0.02 µM, reaching a maximum binding capacity (Bmax) of 3.75 ± 0.03 rosuvastatin molecules per RyR1 channel. The KD value is in line with the IC50 value from [3H]ryanodine binding assays for WT RyR1 with rosuvastatin (Fig. 3a, b). The Bmax value indicates one rosuvastatin binding site per RyR1 protomer, consistent with our structural data (Fig. 1a). Interestingly, previously reported simvastatin binding data indicated two simvastatin binding sites per RyR1 protomer: a higher-affinity site (Sim-1) and a lower-affinity site (Sim-2)14. Rosuvastatin exhibits a submicromolar affinity similar to that of simvastatin at the Sim-1 site (KD ≈ 0.7 µM)14, which is shared by both statins, as indicated by our structural data (Fig. 1b). Discussion Statins primarily act by inhibiting HMG‑CoA reductase to limit cholesterol biosynthesis3,25. Recent evidence regarding adverse statin‑induced responses in skeletal muscle suggests that their impact can extend beyond lipid metabolism and directly influence intracellular Ca2+ homeostasis1,2,12. In particular, RyR1 has emerged as a potential off-target effector of statins10,14,15. The structural and functional data presented here reveal that rosuvastatin directly binds to and activates RyR1 channels to release Ca2+ in a manner similar to simvastatin14, but not to atorvastatin15, even though rosuvastatin shares fewer structural similarities with simvastatin than with atorvastatin. Rosuvastatin binding is observed exclusively in the open conformation of RyR1, implying conformational stabilization, whereas the closed conformation features the TMx1-2 helices occupying the rosuvastatin binding site (Fig. 1). In the previously reported cryo-EM dataset of RyR1 with simvastatin, we did not detect the TMx1-2 helices, but simvastatin bound to the Sim-1 and Sim-2 sites in both closed and open conformations14. Simvastatin binding at the Sim-2 site likely impaired stabilization of the TMx1-2 helices due to steric overlap of Sim-2 with TMx2. Moreover, in the previous dataset, the higher concentration of simvastatin (10 mM) compared to rosuvastatin (2 mM) might have facilitated occupation of the Sim-1 site by simvastatin in the closed RyR1 conformation14. Since rosuvastatin and simvastatin occupy the hydrophobic pocket of the Sim-1 site with distinct functional groups (4-fluorophenyl versus 2,2-dimethylbutyrate ester, respectively), it seems plausible that other drug-like molecules or endogenous regulatory ligands may also bind to it, modulating RyR1 activity. The binding mode of rosuvastatin, with its 4-fluorophenyl moiety inserted between S3 and S4, might imply a molecular “wedge” mechanism that could induce an altered S4S5 linker conformation, thereby supporting pore opening, similar to what has been described for atorvastatin15. Using cryo-EM, atorvastatin was observed to form triplets within the pore region of RyR1 in the open conformation15. Notably, atorvastatin’s large hydrophobic aromatic ring system and the high concentrations (10 mM atorvastatin) used in cryo-EM preparations likely contributed to the formation of these triplets. In the RyR1 structure, the atorvastatin triplet (Ator-1, Ator-2, Ator-3) engages S1, S3, S4, the S4S5 linker, and S5′ of the adjacent subunit near the cytosolic TMD surface (Supplementary Fig. 16); furthermore, Ator-3 contacts the thumb-and-forefinger domain in the RyR1 core region15. As with rosuvastatin, the 4-fluorophenyl moiety of Ator-3 binds between S3 and S4; however, it approaches from the opposite side, positioned closer to the cytosolic TMD surface compared with the Sim-1 site (Supplementary Fig. 6). The binding of the atorvastatin triplet expands the S3 N-terminus to S4 C-terminus distance and shifts the S4S5 linker and S5′, potentially triggering pore opening (Supplementary Fig. 6). Interestingly, a single atorvastatin binding site was observed in the closed state15. This atorvastatin molecule is situated similarly to Ator-1 between S4, S5′, and the S4S5 linker, slightly shifting S5′ and the S4S5 linker without opening the pore (Supplementary Fig. 17a, b)15. In contrast to atorvastatin, we detected no substantial alteration in the helical arrangement of S3, S4, or S5′ upon rosuvastatin binding (Supplementary Fig. 6). However, when aligned at S1-S4, the S4S5 linker exhibited a trend of lateral shifting toward S5 (by ~0.7 Å) in the presence of rosuvastatin, potentially facilitating pore opening (Supplementary Fig. 7). For all the reported statin binding modes on RyR1 structures, the hydrophobic core moieties bind within the TMD of RyR1, while the lactone group (as shown for the rosuvastatin or simvastatin lactone forms) or the dihydroxyheptanoic acid moiety (open-ring form seen in atorvastatin) extends toward the polar interface of the TMD at the cytosolic side. A notable exception is the Sim-2 site, where the lactone group projects toward the luminal side (Supplementary Fig. 17a). For rosuvastatin and simvastatin, the hydrophobic interactions within the pocket between S3 and S4 formed by displacing W4794 likely dominate the binding affinity at the Sim-1 site, mediated by the 4-fluorophenyl and 2,2-dimethylbutyrate ester moieties, respectively. The hydrophobic core moieties (pyrimidine in rosuvastatin, hexahydro-naphthalene in simvastatin) show no overlap in their binding poses; both may contribute to interactions with the hydrophobic acyl chains of phospholipids at the Sim-1 site (Supplementary Fig. 17b). The atorvastatin triplet binding site does not overlap with Sim-1 or Sim-2 but likewise exhibits predominantly hydrophobic interactions. At the Ator-1 site, initially occupied by a single atorvastatin molecule in the closed conformation of RyR1, these involve the 4-fluorophenyl group (typical of type-2 statins) interacting with V4820 of S4S5, and the phenylcarbamoyl and phenyl groups (specific to atorvastatin) interacting with V4838′, G4842′ of S5′ and I4816 of S4, respectively (Supplementary Fig. 17b, d). The triplet formation is enabled by atorvastatin’s large hydrophobic aromatic ring system mediating hydrophobic and π-π stacking interactions both intermolecularly and with the surrounding triplet binding site of RyR1 in the open conformation15. In particular, the phenylcarbamoyl and phenyl groups, unique to atorvastatin, participate in interactions with the clustered Ator-1, Ator-2, and Ator-3 binding sites, and the phenyl group of Ator-2 engages Ator-115. However, given that other statins, such as simvastatin and rosuvastatin, do not possess a comparable extended hydrophobic aromatic ring system, they might be unable to form triplets similar to those observed for atorvastatin, even though functional assays indicate that they activate RyR1 in a comparable manner10,14. Interestingly, similar to Sim-2, atorvastatin binding would sterically clash with the previously described TMx2 conformation (Supplementary Fig. 6)14. Thus, it seems plausible that the interference of these statin drugs with the TMx1-2 helices could alter the stability of the closed pore and facilitate pore opening. Future studies are warranted to elucidate the effect of the various type-1 and -2 statins on the TMx1-2 conformational stability and structure-function relationship of RyR1 in detail, as understanding how external ligands like statins trigger or exacerbate dysfunctional RyR1 might be critical for drug safety and optimization. Rosuvastatin adopts a binding pose at the Sim-1 site that substantially differs from its pose within the active site of the HMG-CoA reductase, its pharmacological target. HMG-CoA reductase inhibition is mediated by the hydrolyzed lactone group of statins, dihydroxyheptanoic acid, which mimics the HMG moiety and sterically hinders substrate binding25. In contrast to HMG-CoA reductase, the rosuvastatin binding site on RyR1 does not form a tight binding pocket for the hydrophilic dihydroxyheptanoic acid group (Supplementary Fig. 17c). Instead, hydrophobic interactions of the pyrimidine core moiety and 4-fluorophenyl group occupying the hydrophobic pocket of the Sim-1 site are crucial for rosuvastatin binding to RyR1 (Supplementary Fig. 17b, d). Thus, the differences between the rosuvastatin binding modes on RyR1 and HMG-CoA reductase may open mechanistic avenues for future therapeutic studies aimed at developing statin drugs with a reduced incidence of SAMS. As a study limitation, we acknowledge that local statin concentrations reached near RyR1 in skeletal muscle during statin therapies remain uncertain. While therapeutic systemic concentrations of statins are typically in the nanomolar range26,27, our in vitro assays required micromolar concentrations (KD ≈ 0.49 µM for rosuvastatin) to characterize binding and functional effects. However, previous studies suggest that statins can reach substantially higher levels in skeletal muscle than those measured in plasma due to membrane partitioning, active transport, and chronic drug accumulation over time28–30. Moreover, given that statins are lipophilic, they may accumulate within the SR membrane or at protein-lipid interfaces, resulting in higher local concentrations that facilitate interactions with RyR1 and putatively trigger SAMS30,31. Interestingly, rosuvastatin is less lipophilic than most other statins30,31. Its lower lipophilicity may limit non-specific penetration into muscle tissue via passive diffusion, putatively contributing to a lower clinical incidence of SAMS, although this remains controversial32–35. A recent meta-analysis study reported similar intolerance rates between lipophilic (atorvastatin, simvastatin, lovastatin, fluvastatin, and pitavastatin) and hydrophilic (pravastatin and rosuvastatin) statins, with no significant difference (4.0 versus 5.0%, p = 0.33)35. In randomized controlled clinical trials (in which participants are randomly assigned to control and statin treatment groups under strictly monitored and controlled clinical conditions), the prevalence of statin intolerance is typically low [~4.9%; 95% confidence interval (CI): 4.0–6.0%]35, and often shows no statistically significant difference compared with placebo groups. In contrast, observational studies (which track pre-existing statin treatment groups over time) report substantially higher prevalences of statin intolerance (~17%; CI: 14–19%) and of SAMS specifically (~8.4%; CI: 5.7–11%)35, suggesting that real‑world statin intolerance rates are considerably higher than those captured in clinical trials (Supplementary Fig. 18a). A key distinction from observational studies is that clinical trials typically apply strict exclusion criteria, often omitting individuals with comorbidities, advanced age, or a prior history of statin intolerance, which may result in an artificially lower observed SAMS incidence. Data obtained from 36 clinical trials (ClinicalTrials.gov) involving individuals not pre-screened for prior statin intolerance or muscle weakness showed a trend toward higher SAMS incidence for atorvastatin [214/2701 (7.9%)] and simvastatin [160/1762 (9.1%)] compared to placebo [795/10874 (6.8%)] but not for rosuvastatin [854/11584 (6.9%); Supplementary Fig. 18b]. Hence, rosuvastatin’s pharmacophore properties could make it a superior lead compound for the development of next-generation statins with improved myotoxicity profiles. Importantly, comorbidities have been associated with increased SAMS incidence rates35. In particular, emerging evidence indicates RyR1 as a critical off‑target effector in adverse statin responses, with pathogenic RyR1‑related conditions predisposing individuals to heightened statin susceptibility14,15. The dominant pathogenic RyR1-Y522S mutation was previously reported to trigger hypermetabolic responses to simvastatin in ex vivo mouse muscle36. Our recent data demonstrate a similar susceptibility in a mouse model expressing the patient‑derived RyR1‑T4709M mutation associated with profound statin intolerance, where simvastatin caused muscle weakness14. Under the same experimental conditions, WT mice remained unaffected by adverse side effects14. This might mirror clinical trial settings where healthy individuals without underlying medical conditions or risk factors presumably exhibit a lower likelihood of developing statin intolerance. The discrepancy underscores that individual statin susceptibilities driven by either genetically determined or acquired comorbidities affecting RyR1 potentially play a crucial role in triggering severe SAMS. Taken together, our findings confirm that simvastatin and rosuvastatin share the same binding site on RyR1, designated as Sim-1, which could serve as a structural basis for designing statin drugs that minimize adverse side effects on RyR1. In contrast to simvastatin and atorvastatin, rosuvastatin does not exhibit multiple binding sites on RyR1 while similarly promoting RyR1 activation (Figs. 2 and 3)10,14,15, indicating its potential as a preferable lead structure. Methods Ethical statement Rabbit muscle tissue was purchased from BioIVT [https://bioivt.com/]. Purification of endogenous RyR1 from rabbit skeletal muscle Endogenous skeletal muscle RyR1 was purified in a tag-free approach by successive anion exchange and size-exclusion chromatography as previously described14,18,19. Buffers were kept ice-cold unless otherwise stated. Frozen rabbit back and thigh muscle tissue (BioIVT; ~20 g) was homogenized in buffer H (~200 ml; 10 mM Tris maleate pH 6.8, 1 mM EGTA, 1 mM benzamidine hydrochloride, 0.5 mM AEBSF) using a Waring blender. Debris was pelleted by centrifugation at 11,000 × g for 10 min at 4 °C, and the supernatant was filtered through cheesecloth. Membrane fractions were pelleted by centrifugation at 36,000 × g for 30 min and solubilized in buffer S (10 mM HEPES pH 7.4, 0.8 M NaCl, 1% CHAPS, 0.1% phosphatidylcholine, 1 mM EGTA, 2 mM DTT, 0.5 mM AEBSF, 1 mM benzamidine hydrochloride, one protease inhibitor tablet (Pierce)) using a glass tissue grinder (Kontes). The resulting membrane suspension was diluted with buffer D (same as buffer S but without NaCl) at a 1:1 ratio, again using the glass tissue grinder. CHAPS-insoluble material was pelleted by centrifugation at 100,000 × g for 30 min, after which the supernatant was filtered and loaded onto a HiTrap Q HP column (5 mL, Cytiva) pre‑equilibrated with buffer A [10 mM HEPES pH 7.4, 400 mM NaCl, 0.25% CHAPS, 1 mM EGTA, 0.5 mM TCEP, 0.01% 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC, Avanti)]. After washing with six column volumes of buffer A, immobilized protein was eluted using a linear gradient from 480 to 550 mM NaCl generated with buffers A and B (same as buffer A but with 600 mM NaCl). RyR1 fractions were pooled and concentrated to ~0.4 mL using 100,000 kDa cut-off centrifugation filters (MilliporeSigma). Size-exclusion chromatography was performed on a TSKgel G4SWXL column (TOSOH Biosciences) using buffer C (same as buffer A but with 150 mM NaCl). RyR1 fractions were pooled and concentrated to >10 g/L using 100,000 kDa cut-off centrifugation filters (MilliporeSigma). For cryo-EM, RyR1 samples (final: ~15 µM RyR1 protomer) were prepared with 10 mM Na-ATP, 5 mM caffeine, and 30 μM free Ca2+, and were incubated with 2 mM rosuvastatin lactone (ChemScene) dissolved in DMSO (final: ~1%) for at least 20 min on ice. Free Ca2+ concentrations were calculated using MaxChelator (https://somapp.ucdmc.ucdavis.edu/pharmacology/bers/maxchelator/webmaxc/webmaxcE.htm). Cryo-EM data collection, processing, and model building Cryo-EM grids were prepared as described previously14,18–21. A Vitrobot Mark IV (Thermo Fisher Scientific) was used for vitrification. Purified RyR1 (3 µL) was applied to UltrAuFoil holey gold grids (Quantifoil R 0.6/1.0, Au 300) that had been cleaned with an EasiGlow device (PELCO). Grids were blotted with ashless filter paper (Whatman) using a blot force of 10 for 8 s at 4 °C and 100% relative humidity to generate thin aqueous films. The grids were then vitrified by plunge-freezing into liquid ethane cooled by liquid nitrogen. Cryo-EM data were acquired on a Titan Krios 300 kV microscope (Thermo Fisher Scientific) equipped with a BioQuantum energy filter (20 eV slit width) and a K3 direct electron detector (Gatan). Automated data acquisition was performed using Leginon software. Data were recorded at a nominal magnification of 105,000×, corresponding to a calibrated pixel size of 0.83 Å. The electron dose rate was set to 16e−/pixel/s with a total exposure time of 2.5 s, yielding a cumulative dose of approximately 50–60e−/Ų. Cryo-EM data were processed in cryoSPARC37. Patch Motion Correction and Patch CTF Estimation were used to align image stacks and estimate defocus values, respectively. RyR1 particles were picked using Topaz, which was trained on pre-existing RyR1 templates. After 2D classification, particles from the highest-resolution classes were pooled for further processing. Ab initio 3D reconstructions of RyR1 particles were generated and subjected to C4-symmetric homogeneous refinement, followed by three-class heterogeneous refinement to further isolate high-quality particles. A local refinement followed by 3D classification, using a mask focused on the pore region, was performed to separate the closed and open conformations of RyR1 induced under activating conditions. For subsequent local refinements, particle stacks were symmetry expanded (C4). Refinement masks were generated for the following regions: (i) the N-terminal domain, the three SPRY domains, the RY1&2 tandem repeat domain, and calstabin; (ii) the CSol and JSol domains; (iii) the BSol domain; (iv) the pore region including the thumb-and-forefinger (TaF), transmembrane (TMD), and C-terminal (CTD) domains. C4 symmetry was applied during the local refinement of the pore region. Additional smaller masks targeted more variable regions, including: (v) the TMx region, and the tandem repeat domains (vi) RY1&2-only and (vii) RY3&4-only. Particles in the closed conformation exhibiting TMx1-2 were further selected by 3D classification using the small mask focused on the TMx region, followed by local refinements. Following homogeneous and local refinements, composite maps were independently generated for particles in closed and open conformations (Supplementary Fig. 1 and Supplementary Fig. 2). To assemble composite maps, focused maps from local refinements were aligned to the global consensus map from the corresponding homogeneous refinement and combined using UCSF ChimeraX38. The pixel size was calibrated by calculating correlation coefficients against a map generated from the crystal structure of the N-terminal domain39. Atomic models were manually built in Coot40, using a high-resolution cryo-EM structure of RyR1 (PDB: 7TZC) as a starting template. Atomic models were subjected to real-space refinement in Phenix41. Pore dimensions along the conduction pathway of RyR1 were analyzed with HOLE42. Figures of RyR1 structures were prepared with UCSF ChimeraX38. ER microsome preparation ER microsomes were prepared from HEK293 cells recombinantly expressing WT or mutant RyR1, as previously described14,20. In brief, cell pellets were homogenized in homogenization buffer [20 mM Tris maleate pH 7.4, 1 mM EDTA, 1 mM DTT, and protease inhibitors (Roche)] on ice using a Teflon-glass homogenizer. Debris was removed by centrifugation at 4000 × g for 15 min at 4 °C. Supernatants were centrifuged at 40,000 × g for 30 min at 4 °C. Pellets were resuspended in storage buffer (10 mM MOPS pH 7.4, 1 mM DTT, 1 mM EDTA, 250 mM sucrose, and protease inhibitors). Aliquots were frozen in liquid nitrogen and stored at 80 °C. Single-channel data acquisition and analysis of RyR1 RyR1 channels in ER microsomal preparations from transfected HEK293 cells were reconstituted into planar lipid bilayers for single‑channel recordings, as previously described14,18,20. In brief, phosphatidylethanolamine and phosphatidylcholine (Avanti Polar Lipids) were mixed in a 3:1 ratio and dissolved in decane (30 mg/ml). The lipid mixture was spread across a 200-µm aperture in a polysulfonate cup (Warner Instruments). The planar lipid bilayer formed across the aperture separated two 1‑ml chambers representing the cytoplasmic and SR/ER luminal compartments on the cis and trans sides, respectively. The cis chamber was maintained at virtual ground, whereas the trans chamber was connected to the headstage input of a BC‑525D bilayer voltage‑clamp amplifier (Warner Instruments). The chamber solutions were as follows: the cis side contained 250/125 mM HEPES/Tris pH 7.35, 50 mM KCl, 1 mM EGTA, and 0.64 mM CaCl₂, while the trans side contained 250 mM HEPES pH 7.35, 50 mM KCl, and 53 mM Ca(OH)₂. Free Ca²⁺ concentrations in the cis chamber were determined using MaxChelator. Microsomal RyR1 channels were then introduced to the cis side of the planar lipid bilayer. To promote channel incorporation into the bilayer lipid membrane, the cis side was rendered hyperosmotic by raising the KCl concentration to 0.4–0.5 M. Following channel incorporation, the cis chamber was perfused with fresh cis buffer. RyR1 single-channel currents were recorded at room temperature before and after adding rosuvastatin in its lactone (rosuvastatin lactone, ChemScene) or acid form (rosuvastatin sodium salt, Thermo Fisher Scientific) to the cis chamber. Approximately 1 min after each addition, PO was determined over 3 min of continuous recording using 50% threshold analysis. The trans-bilayer potential was maintained at 0 mV using the Bilayer Clamp amplifier. Single-channel currents were low-pass filtered at 1 kHz and digitized at a sampling rate of 4 kHz. Software used for data acquisition: Digidata 1440A and Axoscope 10.2. Software used for data analysis: Clampfit 10.2 (Molecular Devices), GraphPad Prism, and Microsoft Excel 2010. Radiometric rosuvastatin and ryanodine binding assays Recombinant RyR1 [3H]ryanodine binding or [3H]rosuvastatin binding was radiometrically measured in microsomal ER preparations as previously described14,20,21. In brief, microsomal ER preparations (~0.1 mg) from HEK293 cells transiently transfected with recombinant WT or mutant RyR1 were incubated with [3H]rosuvastatin acid (American Radiolabeled Chemicals) or [3H]ryanodine (American Radiolabeled Chemicals) in binding buffer (50 mM Tris-HCl pH 7.5, 25 mM MgCl2, 150 mM NaCl) at room temperature for 30 min in presence or absence of rosuvastatin (rosuvastatin sodium salt, Thermo Fisher Scientific) or atorvastatin (atorvastatin sodium salt, TargetMol). Incubations were halted by dilution with ice-cold binding buffer, followed by rapid filtration through Whatman GF/B filters. Filters were rinsed three times with 5 mL of wash buffer (10 mM MOPS pH 7.4, 200 mM NaCl). Radioactivity retained on dried filters was quantified by scintillation counting to determine [3H]rosuvastatin or [3H]ryanodine binding. Non-specific binding was determined in the presence of a 20-fold excess of unlabeled ryanodine. Data were normalized to saturating [3H]ryanodine binding levels. Reporting summary Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article. Supplementary information Supplementary Information Transparent Peer Review file Reporting Summary Source data Source data Acknowledgements Cryo-EM data were collected at the Columbia University Cryo-Electron Microscopy Center. The authors thank Robert Grassucci and Zhening Zhang for assistance with the cryo-EM data collection. Author contributions G.W. and A.R.M. designed research. G.W., S.R., Q.Y., and A.W. performed research. G.W., S.R., Q.Y., M.C.M., A.C., and A.R.M. analyzed data. G.W. wrote the manuscript. G.W., M.C.M., Y.L., H.D., and A.R.M. edited and revised the manuscript. Peer review Peer review information Nature Communications thanks Zhiguang Yuchi, Rouslan Efremov, and Angela F Dulhunty for their contribution to the peer review of this work. A peer review file is available. Funding This work was supported by NIH grants R01 DK118240, R01 HL140934, R01 HL178501, R25 HL156002R25, and P01 HL164319 (to ARM); NIH grant R01 AR085078 (to H.D.). Data availability Atomic coordinates have been deposited in the Protein Data Bank (PDB) and cryo-EM density maps in the Electron Microscopy Data Bank (EMDB). The PDB accession codes for RyR1 in the presence of rosuvastatin are 9ZO8 (closed) and 9ZOE (open). The corresponding EMDB accession codes are EMD-74471 [https://www.ebi.ac.uk/pdbe/entry/emdb/EMD-74471] and EMD-74486 [https://www.ebi.ac.uk/pdbe/entry/emdb/EMD-74486], respectively. The following PDB accession code was used to guide model building: 7TZC The following PDB and corresponding cryo-EM map accession codes were used for comparison regarding the statin binding sites on RyR1 and HMG-CoA reductase: 9NMO/EMD-49535 [https://www.ebi.ac.uk/emdb/EMD-49535] (RyR1, closed), 9NMN/EMD-49534 [https://www.ebi.ac.uk/emdb/EMD-49534] (RyR1, open), 9NMP/EMD-49536 [https://www.ebi.ac.uk/emdb/EMD-49536] (RyR1-simvastatin, open), 9OL4/EMD-70589 [https://www.ebi.ac.uk/emdb/EMD-70589] (RyR1-atorvastatin, open), 9OL5/EMD-70590 [https://www.ebi.ac.uk/emdb/EMD-70590] (RyR1-atorvastatin, closed), 1HWL (HMG-CoA reductase with rosuvastatin), 1HW9 (HMG-CoA reductase with simvastatin), and 1HWK (HMG-CoA reductase with atorvastatin). Source data are provided with this paper. Competing interests A.R.M. owns stock in RyCarma Therapeutics Inc., a company developing compounds targeting RyR. A.R.M. is coinventor on US patent nos. US8022058 and US8710045. G.W., H.D., M.C.M., and A.R.M. are inventors on the patent application entitled “Statin Innovation for Muscle-Friendly Cholesterol Management” (Invention Report CU24350) to be filed by Columbia University. The remaining authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s41467-026-77372-9. References Associated Data Supplementary Materials Supplementary Information Transparent Peer Review file Reporting Summary Source data Data Availability Statement Atomic coordinates have been deposited in the Protein Data Bank (PDB) and cryo-EM density maps in the Electron Microscopy Data Bank (EMDB). The PDB accession codes for RyR1 in the presence of rosuvastatin are 9ZO8 (closed) and 9ZOE (open). The corresponding EMDB accession codes are EMD-74471 [https://www.ebi.ac.uk/pdbe/entry/emdb/EMD-74471] and EMD-74486 [https://www.ebi.ac.uk/pdbe/entry/emdb/EMD-74486], respectively. The following PDB accession code was used to guide model building: 7TZC The following PDB and corresponding cryo-EM map accession codes were used for comparison regarding the statin binding sites on RyR1 and HMG-CoA reductase: 9NMO/EMD-49535 [https://www.ebi.ac.uk/emdb/EMD-49535] (RyR1, closed), 9NMN/EMD-49534 [https://www.ebi.ac.uk/emdb/EMD-49534] (RyR1, open), 9NMP/EMD-49536 [https://www.ebi.ac.uk/emdb/EMD-49536] (RyR1-simvastatin, open), 9OL4/EMD-70589 [https://www.ebi.ac.uk/emdb/EMD-70589] (RyR1-atorvastatin, open), 9OL5/EMD-70590 [https://www.ebi.ac.uk/emdb/EMD-70590] (RyR1-atorvastatin, closed), 1HWL (HMG-CoA reductase with rosuvastatin), 1HW9 (HMG-CoA reductase with simvastatin), and 1HWK (HMG-CoA reductase with atorvastatin). Source data are provided with this paper.