Ca2+ inactivation of the mammalian ryanodine receptor type 1 in a lipidic environment revealed by cryo-EM

Figure 1.. RyR1 at high Ca2+ concentration exhibits a mixture of open and inactivated conformations. (A) Ryanodine binding of rabbit skeletal sarcoplasmic reticulum (SR) microsomes showing the Ca2+-induced inactivation of RyR1; activators ATP or ACP (2 mM) increased channel open probability by 3- and 4.5-fold, respectively. Ca2+ concentrations of 100 μM to 2 mM progressively decrease the probability of opening (Po). Mean specific [3H]-ryanodine binding ± SEM from four independent experiments. (B) Pore profile of RyR1-ACP/EGTA, and of open and inactivated conformations in the RyR1-ACP/Ca2+A dataset calculated with the program HOLE (Smart et al., 1993). The position of relevant landmarks is indicated with violet shadowing. Radius corresponding to a dehydrated Ca2+ ion shown with a dashed line. (C) Dotted surfaces of RyR1 ion permeation pathway in closed, open, and inactivated conformations, color-coded according to pore radius. Overlaid coordinates of the S5, S6, pore helices, and selectivity filter are shown for two diagonal protomers. Insets: cytoplasmic views of the Ile4937 constriction and corresponding pore diameter measured at the Cα backbone. Similar results were obtained for…

Figure 1—figure supplement 1.. Image processing scheme for the RyR1-ACP/Ca2+A dataset. (A) Representative micrograph collected on a Titan Krios at ×81,000 magnification with a K3 camera in counting mode. (B) Representative 2D class averages obtained by reference-free 2D classification in RELION. (C) Image processing workflow with tetrameric and single subunit-masked RyR1 particles in RELION. Overall resolution values obtained for the inactivated conformations after density modification in PHENIX.Resolve. (D) Euler angle distribution of the particles contributing to the RyR1-ACP/Ca2+ inactivated (3.8 Å resolution) and RyR1-ACP/Ca2+ open (4.6 Å resolution) maps. (E) Left: gold standard Fourier shell correlation prior to density modification. Right: local resolution of the symmetry-expanded maps calculated with ResMap. (F) Main 3D classes obtained from the focused 3D classification after symmetry expansion with NTD (N-terminal domain) and CD (Central domain) domains highlighted in yellow and green, respectively, with the pore axis on the right. Classification yielded one open conformation and three closed conformations that differ in the flexion angle of the cytoplasmic domain.

Figure 1—figure supplement 2.. Correlation of RyR1-ACP/Ca2+A inactivated model with the cryogenic electron microscopy (cryo-EM) density. Model quality for RyR1-ACP/Ca2+A inactivated in the consensus cryo-EM map; domains in the central and transmembrane regions are shown.

Figure 1—figure supplement 3.. Image processing scheme for the RyR1-ACP/Ca2+B dataset. (A) Representative micrograph collected on a Titan Krios at ×130,000 magnification with a K2 camera in super-resolution mode. (B) Representative 2D class averages obtained by reference-free 2D classification in RELION. Overall resolution values obtained for inactivated 3D class after a density modification in PHENIX.Resolve. (C) Workflow of image processing with tetramer and single subunit-masked particles carried out in RELION. (D) Euler angle distribution of the particles contributing to the RyR1-ACP/Ca2+B-inactivated and RyR1-ACP/Ca2+B open maps at 4.1 Å and 5.8 Å resolution, respectively. (E) Left: gold standard Fourier shell correlation prior to density modification. Right: local resolution of the symmetry-expanded map calculated with ResMap with flexion angle of the cytoplasmic domain indicated.

Figure 1—figure supplement 4.. Image processing scheme for the RyR1-ACP/EGTA dataset. (A) Representative micrograph collected on a Titan Krios at ×130,000 magnification with a K2 camera in super-resolution mode. (B) Representative 2D class averages obtained by reference-free 2D classification in RELION. (C) Image processing workflow including 3D classification focused on RyR1’s quarter carried out in RELION. (D) Euler angle distribution of the particles contributing to the 4.3 Å resolution 3D map. (E) Left: gold standard Fourier shell correlation. Right: local resolution of the symmetry-expanded map calculated with ResMap.

Figure 1—figure supplement 5.. Pore conformation in the RyR1-ACP/Ca2+B dataset. (A) Dotted surfaces of RyR1 ion-permeation pathway in open and inactivated conformations. Inset: fourfold cytoplasmic views of Ile4937 constriction and corresponding pore diameter (measured at Cα backbone) with respective cryogenic electron microscopy (cryo-EM) maps (mesh). (B) Left: pore profiles of inactivated and open conformations calculated with HOLE. Radius corresponding to a dehydrated Ca2+ ion is shown as a dotted line. Right: fourfold cytoplasmic views of the Gln4933 constriction in RyR1-ACP/Ca2+A and RyR1-ACP/Ca2+B pores with their respective cryo-EM maps.

Figure 2.. Inactivation of RyR1 involves out-of-plane rotation of the central block and rearrangement around the Ca2+ activation site. (A) The high-affinity Ca2+-binding site in the CD/CTD interface with density around the Ca2+ site contoured at 4σ. Contacts within 2.8 Å from Ca2+, as well as additional contact Gln3970-Ser4029 within 3.6 Å, are represented by dashed lines. Channel axis is on the left. During the transition from open to inactivated conformations, the CD/CTD block tilts around the Ca2+-binding site such that the protruding fourth helix of the CD (h) and connected CTD tilt inward, while the CD-C′ tilts upward and away from the sarcoplasmic reticulum (SR) membrane – with Gln3970 separating from Ca2+ by ~6 Å. Arrows and stationary reference lines illustrate the conformational changes undergone with respect to the panel on the left. The region represented relative to the channel is highlighted with a square in panel (C). See Figure 2—figure supplement 1 for the corresponding space-filling representation. (B) Heat map showing Cα-backbone root mean square deviation (RMSD) (in Å) between domain pairs from respective conformations after aligning them (pre-aligned), in their native…

Figure 2—figure supplement 1.. Reorganization of the high-affinity Ca2+-binding site at the CD/CTD interface under different conditions. Ca2+-binding site in the CD/CTD interface in different conformations with involved residues shown with spheres. Residues Glu3983, Glu3967, Gln3970 (CD), and Thr5001 (CTD), within 2.8 Å from Ca2+ in the open state, are shown. In going from RyR1-ACP/Ca2+ open to RyR1-ACP/Ca2+ inactivated, Gln3970 switches its interaction from Ca2+ to Ser4029 (interaction within 3.6 Å). Arrows and stationary reference lines illustrate the conformational changes undergone with respect to the panel on the left. The region represented relative to the channel is highlighted with a square in Figure 2C; fourfold axis is on the left.

Figure 2—figure supplement 2.. Rotation axes of the CD of RyR1 among the different conformational transitions. (A) Reproducibility of the conformation of the ‘central block’ of RyR1-ACP/Ca2+ inactivated from two independent datasets (A and B, superimposed), with a root mean square deviation (RMSD) of 0.95 Å over the Cα atoms. The central block comprises the CD, EF hand, U-motif, S6C′, and CTD. (B) Transition from RyR1-ACP/Ca2+ open (gray) to RyR1-ACP/Ca2+ inactivated (colored) involves an out-of-plane rotation of the CD around the yellow-colored axis. This results in inward tilt in CD-N′ (brown arrows) and upward shift/tilt of CD-C′ and connected EF hand domain (blue arrows). Two orthogonal views are shown, and the fourfold axis on the left panels, in front of the structure, is indicated. (C) Transition from RyR1-ACP/Ca2+ inactivated (colored) to RyR1-ACP/EGTA closed (gray) involves out-of-plane rotation of the CD of RyR1 around the magenta-colored axis.

Figure 3.. Two salt bridges between the EF hands and the S2–S3 loop are determinant for the inactivated state. (A) Cytoplasmic sarcoplasmic reticulum (SR) view (top row) and side view (bottom row) highlighting the EF hand domain (blue) and S2–S3 loop (magenta) at the subunit interface. Counterclockwise in-plane and ascending out-of-plane rotations of the central region in the transition from RyR1-ACP/EGTA (closed) to RyR1-ACP/Ca2+ open brings the two domains closer together. Further rotation of the CD (tan color in top row) and its protruding EF hand domain in RyR1-ACP/Ca2+ inactivated brings the two domains in contact. The inter Cα-Cα contact distances illustrate the progressive approximation of the two domains. Two salt bridges, Lys4101-Asp4730 and Glu4075-Arg4736, form in the RyR1-ACP/Ca2+-inactivated conformation. The location of the domains in the context of RyR1 is shown on the right panels. (B) MH/CCD mutation sites in the interface between the EF hand domain and S2–S3 loop that abolish Ca2+-dependent inactivation (black) versus MH/CCD mutations without effect on inactivation (green). The two EF loops are indicated (EF1, EF2). Residues forming the salt bridges are labeled according…

Figure 4.. Changes in the interaction network of the nucleotide and complexed Ca2+. The ATP-binding pocket is formed by the U-motif, CTD, and S6C′. (A) cryo-EM densities (gray mesh) of ACP in the different states represented at map significance values (root mean square deviation [RMSD] σ) above 12, 7, and 7, respectively. Residues within 4 Å of either ACP, Ca2+, or Zn2+ (from the CTD zinc finger) are color-coded according to domain. Notably, an interaction between His4983 (CTD) and ACP is only observed in the inactivated state. Fitted water densities (w) are represented in red. (B) Left: ACP goes deeper into the ATP-binding pocket in RyR1-ACP/Ca2+ inactivated (colored structure, blue ACP) as compared to the open state (gray structure, red ACP) due to a 6° rotation (see arrows). Right: absence Ca2+ in the closed state (green) causes release of the CTD from the CD, resulting in a 4° tilt of the CTD and U-motif (see arrows). Additional reorganization allows closer interaction between ACP and U-motif in the closed state.

Figure 4—figure supplement 1.. Map significance of the ACP and putative Ca2+ densities under different conditions. Root mean square deviation (RMSD σ) values of the densities attributed to Ca2+ and ACP in the cryogenic electron microscopy (cryo-EM) maps of RyR1-ACP/EGTA closed, RyR1-ACP/Ca2+ open, and RyR1-ACP/Ca2+ inactivated.