Two Structural Configurations of the Skeletal Muscle Calcium Release Channel

Elena V. Orlova^1,2, Irina I. Serysheva^1,3, Marin van Heel^2,4, Susan L. Hamilton³ and Wah Chiu^1,3

¹Verna and Marrs McLean Department of Biochemistry and The W. M. Keck Center for Computational Biology, Baylor College of Medicine, One Baylor Plaza, Houston, Texas 77030, USA;
²Department of Molecular Physiology and Biophysics, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, USA
³Fritz Haber Institute of the Max Planck Society, Faradayweg 4-6, D-14195, Berlin, Germany
⁴Department of Biochemistry, Imperial College of Science, Medicine and Technology, London SW7 AY, UK

ABSTRACT

Here we present the determination of the three-dimensional structure of the skeletal muscle Ca²⁺-release channel in an open state using electron cryomicroscopy and angular reconstitution. In contrast to our reconstruction of the channel in its closed state, the density map of the channel driven towards its open state by the presence of Ca²⁺ and ryanodine, features a central opening in the transmembrane region - the likely passageway for Ca²⁺ ions from the sarcoplasmic reticulum to the cytosol. The opening of the channel is associated with a 4° rotation of the transmembrane region with respect to the cytoplasmic region of the channel tetramer, and with significant mass translocations within the entire cytoplasmic region of the channel.

Contraction of the skeletal muscle is triggered by the release of Ca²⁺ from the sarcoplasmic reticulum (SR) after transverse tubule depolarization¹. The Ca²⁺ release channel, associated with the junctional face membrane of the terminal cisternae of the SR^2,3 is responsible for the rapid release of Ca²⁺ during excitation-contraction (E-C). The identification and isolation of the Ca²⁺ release channel was greatly facilitated by the finding that ryanodine, a paralyzing neutral plant alkaloid which induces Ca²⁺ release from the muscle SR, binds to the channel protein with high specificity and affinity^4,5. The purified ryanodine receptor exhibits channel activity upon reconstitution into planar lipid bilayers and has similar pharmacological properties as the native Ca²⁺ release channel from the SR^6-8. The channel complex consists of four identical subunits of 565,000 Mr each⁹. A 12,000 Mr FK506-binding protein (FKBP12) was recently reported to be bound to each subunit¹⁰. The entire structure of the Ca²⁺ release channel, including associated proteins, is greater than 2.3 million daltons and is the largest ion channel sequenced⁹ to date. The full tetramer is apparently necessary for ryanodine binding and channel activity¹¹.

Under physiological conditions the Ca²⁺ release channel exists in different functional states. For simplicity we designate the states we study as the "open" and "closed" states. Recent evidence has indicated a preferential interaction of the voltage sensor of the t-tubule with the open Ca²⁺ release channel¹¹. Activators - including Ca²⁺, ATP, caffeine, and polylysine^12-14 - increase the probability that the channel will open. These modulators also enhance [³H]-ryanodine binding to the channel¹⁵. Although the transition of the Ca²⁺-release channel from one functional state to another has been studied extensively by electrophysiological and radioligand binding analyses, the molecular mechanism by which ligands influence Ca²⁺ release remains unknown.

The understanding of the E-C coupling mechanism in muscle will require detailed knowledge of the three-dimensional (3D) architecture of the Ca²⁺-release channel in different functional states. The quaternary structure of the isolated skeletal muscle Ca²⁺-release channel has been investigated by electron cryomicroscopy and 3D reconstruction techniques^16,17.In the closed state of the channel ¹⁷, no detectable opening on the SR lumenal side was found. Here we describe the use of the angular reconstitution approach^18,19 to study the channel in its open state.

Open and closed Ca²⁺-release channels

The particles selected (Fig. 1a) from the digitized micrographs of ice-embedded Ca²⁺-release channels in open state were averaged into characteristic views (Fig. 1b) to reduce noise. The characteristic views, representing the particles in different orientations within the embedding matrix^18,19, were used to calculate the 3D structure of the channel. Reprojections (Fig. 1c) of the final structure presented different orientations (Fig. 1d) that match well with the corresponding characteristic views (Fig. 1b).

The overall structural features revealed in the open state 3D reconstruction of the channel (Fig. 2, 3) resemble those described in the closed channel¹⁷. The large square-shaped cytoplasmic region is rather hollow and consists of "clamp"-shaped domains (C) at its periphery, interconnected by "handle" domains (H) in the outer part and by ring connections surrounding the open inner part of the cytoplasmic region (CY) of the channel (Fig. 3). The cytoplasmic region is connected to the smaller transmembrane domain (TM) by four columns. The columns and the transmembrane domain together form the "stem" of the mushroom-shaped protein (Fig. 3, middle).

In the open state reconstruction a central cavity is revealed in the putative transmembrane portion of the channel (Fig. 3 top and bottom), whereas there is no apparent hole in the absence of ryanodine and Ca^{2+ 17}. The diameter of the narrowest part of the channel has been estimated to be ~ 7 Å, based on the permeability of the channel to organic cations ²⁰ whereas the length of the narrowest part of the channel has been estimated, by streaming potentials, to be ~10 Å^21,22. The opening in the transmembrane domain in our open-state reconstruction has a diameter of ~18 Å+7Å in its narrowest region, which has a length of ~20 Å and is located at ~20 Å from the lumen face of the channel. The selectivity portion of the channel could be formed by just one or a few amino-acid residues protruding into the channel from each of the four symmetry-related polypeptides surrounding the pore. Such small details would not be detectable at the current level of resolution (30 Å). Our reconstructions of the channel in two functional states, however, do elucidate a substantial structural rearrangement in this region of the molecule.

On the SR lumenal side, the transmembrane domain in the open state is twisted counter clockwise by ~4deg. with respect to its position in the closed state. On the cytoplasmic side of the channel protein, the opening across the putative membrane region widens to a diameter of 30 Å at a distance of ~45 Å from the lumen surface. The clamp-shaped domains at the four corners of the cytoplasmic region appear in a more open conformation in the presence of ryanodine and Ca²⁺ than they do in the closed state (Fig. 3, right column). In the side view (Fig. 3, middle), the channel protein in the open state has a height of ~200 Å which is slightly higher than that observed for the closed state ( ~190 Å)¹⁷. The stem portions of the protein channel appear similar in both states. The Euler angle distributions of the particle orientations in the closed and open state samples were quite similar and these distributions thus do not artificially cause the small height difference between two states. The structural differences between the states are observed at contour levels corresponding to 2.2-2.7x10⁶ Mr.

Interpretation of the two configurations

We found significant structural changes in the protein in the presence of Ca²⁺ and ryanodine (Figs. 3). These changes are global and are detected in both the putative membrane spanning portion and in the cytoplasmic region. Several lines of evidence have supported global rather than local conformational changes in the Ca²⁺-release channel. All the mutations in the channel associated with malignant hyperthermia and with central core disease are in the putative cytoplasmic regions of the protein, but alter the activity of the channel forming parts of the protein ^{23, 24}. Fluorescent probe studies by Ikemoto et al.²⁵ also support global conformational changes in response to activators. Global conformational changes upon binding ligands have been observed with the gap junction protein²⁶ and with the nicotinic acetylcholine receptor²⁷. In the case of acetylcholine receptor where the structure was solved to a resolution ~9 Å, the narrowest region of the channel is 9-10 Å, a value close to the value predicted from permability measurement for various organic molecules. For the structure of the gap junction, however, which was determined to 20 Å resolution, no obvious hole was seen in the reconstruction, although permeability measurements on mammalian cells predicted a hole size of 16 -20 Å^{28, 29}.

The structures of the Ca²⁺-release channel shown here provide new insights into understanding the possible path followed by calcium ions upon opening of the Ca²⁺ release in the SR. The newly identified "hole" in the putative membrane spanning region of the protein may represent the exit route for Ca²⁺ from the SR lumen when the channel opens. Freeze-fractured images of the t-tubule show the putative voltage-sensors to be arranged as square groups of four, with their centers of 260-320 Å apart³⁰. Such arrangement of these receptors appears to match the spatial organization of the four clamp-shaped domains of the channel. One could thus speculate that the cytoplasmic clamp domain is the site of interaction with the voltage sensor. A change in the conformation of the voltage sensor could lead to the observed large conformational changes of the Ca²⁺-release channel, resulting in its opening. In summary, our analysis reveals global conformational changes when the Ca²⁺-release channel opens. This is the first time that angular reconstitution^18,19,31 approach has been used to demonstrate structural differences between two functional states of a protein.

Methods:

Electron cryomicroscopy. The Ca²⁺-release channel protein was purified to homogeneity from rabbit fast twitch skeletal muscle³². To bring the majority of the channels into their open state, the purified sample was incubated in 100 uM Ca²⁺ and 100 nM ryanodine for 3 hours at 20deg.C. The channel molecules were frozen in a thin layer of vitrous ice on a thin carbon film¹⁷.

Images of the frozen, hydrated Ca²⁺-release channel protein (Fig. 4) were recorded at 100 keV under minimal dose conditions (5-7 e/Ų), and at a magnification of 30,000x in a JEOL 1200 electron cryomicroscope. The first zero of the contrast transfer function (CTF) at the defocus values used were at ~26 Å as determined by methods previously described ³³, and no CTF correction was thus applied at the current level of resolution. The seven best micrographs were digitized with a Perkin Elmer microdensitometer using a step size of 20 um per pixels (6.7 Å in the object). After the micrographs were confirmed to have similar defocus values, ~ 5,700 channel particles were selected interactively, extracted into individual images of 80x80 pixels, and processed as a single data set. The data processing was performed in the context of the IMAGIC-5³⁴ software system on either Silicon Graphics Indigo or DEC alpha workstations.

Angular reconstitution. Reference-free alignment by classification ³⁵ was applied to find a set of first references for multi-reference alignment of the data set¹⁷. Multivariate statistical data compression and automatic classification ³⁶ were used to sum similar images into class averages or "characteristic views" to reduce noise. Each of the characteristic views was assigned a set of Euler angles^{17,18, 31} for the subsequent 3D reconstruction procedures calculated, using the exact filter back-projection algorithm³⁷. Starting with this preliminary 3D reconstruction, several rounds of iterative refinement were carried out ^17,31. Different quality criteria were used to follow the improvements attained during the iterative procedures and to discard poor class averages³⁶ or class averages that were assigned a poor set of Euler angles³¹ from the 3D reconstructions.

In our final reconstruction 165 classes (out of 320) were used, which together contained ~ 3700 of the 5700 original molecular images. The Euler angle distribution of the classes used in the final reconstruction is shown in Fig. 5a. The asymmetric 'triangle' for a four-fold symmetric protein, spans a quarter of the unit sphere with ß ranging from 0°-180° and [[gamma]] from -45° to 45°. Since the Euler directions north of the equator are equivalent to their mirror images south of the equator, the overall coverage of the asymmetric triangle by characteristic views is complete (Fig. 5b). The resolution in the reconstruction of the channel in the open state was found to be 30 Å, using the Fourier Shell Correlation³⁷ between two independent reconstructions, each based on half the number of class averages used for the final 3D reconstruction. The relative distribution of the energy in the power spectrum of the 3D reconstructions of the Ca²⁺-release channel in its open and closed states were matched so that the surface views of the reconstructions are comparable to each other.

1. Ebashi, S. & Endo, M. Calcium ion and muscle contraction. Biophys. Mol. Biol. 18, 123-183 (1968).

2. Franzini-Armstrong, C. Studies of the triad. I. Structure of the junction in frog twitch fibers. J. Cell Biol. 47, 488-499 (1970).

3. Saito, A., Inui, M., Radermacher, M., Frank, J. & Fleischer, S. Ultrastructure of the calcium release channel of sarcoplasmic reticulum. J. Cell Biol. 107, 211-219 (1988).

4. Campbell, K., Knudson, C. & Imagawa, T. Identification and characterization of the high affinity [3H]-ryanodine receptor of the junctional sarcoplasmic reticulum Ca²⁺ release channel. J. Biol. Chem. 262, 6460-6463 (1987).

5. Inui, M., Saito, A. & Fleischer, S. Purification of the ryanodine receptor and identity with feet structures of junctional terminal cisternae of sarcoplasmic reticulum from fast skeletal muscle. J. Biol. Chem. 262, 1740-1747 (1987).

6. Imagawa, T., Smith, J. & Coronado, R. Purified ryanodine receptor from skeletal muscle sarcoplasmic reticulum is the Ca⁺² -permeable pore of the Ca⁺² -release channel. J. Biol. Chem. 262, 16636-16643 (1987).

7. Lai, F.A., Erickson, H.P., Rousseau, E., Liu, Q.-Y. & Meissner, G. Purifiation and reconstitution of the calcium release channel from skeletal muscle. Nature 331, 315-319 (1988).

8. Lai, F.A., Misra, M. & Xu, L. The ryanodine receptor Ca²⁺ release channel complex of skeletal muscle sarcoplasmic reticulum: evidence for a cooperatively charged homotetramer. J. Biol. Chem. 264, 16776-16785 (1989).

9. Takeshima, H., et al. Primary structure and expression from complementary DNA of skeletal muscle ryanodine receptor. Nature 339, 439-445 (1989).

10. Jayaraman, T., et al. FK506 binding protein associated with the calcium release channel (ryanodine receptor). J. Biol. Chem. 267, 9474-9477 (1992).

11. Lu, X., Xu, L. & Meissner, G. Activation of the skeletal muscle calcium release channel by a cytoplasmic loop of the dihydropyridine receptor. J. Biol. Chem. 269, 6511-6516 (1994).

12. Smith, J.S., Coronado, R. & Meissner, G. Single channel measurements of the calcium release channel from skeletal muscle sarcoplasmic reticulum. J. Gen. Physiol. 88, 573-588 (1986).

13. Cifuentes, M.E., Ronjat, M. & N.Ikemoto. Polylysine induces a rapid Ca²⁺ release from sarcoplasmic reticulum vesicles by mediation of its binding to the foot protein. Arch. Biochem. Biophys. 273, 554-561 (1989).

14. Pessah, I.N., Stambuk, R.A. & Casida., J.E. Ca²⁺ -activated ryanodine binding: mechanisms of sensitivity and intensity modulation by Mg²⁺, caffeine, and adenine nucleotides. Mol. Pharmacol. 31, 232-238 (1987).

15. Chu, A., Diaz-Muñoz, M., Hawkes, M.J., Brush, K. & Hamilton, S.L. Ryanodine as a probe for the functional state of the skeletal muscle sarcoplasmic calcium release channel. Mol. Pharmacol. 37, 735-741 (1990).

16. Radermacher, M., et al. Cryo-electron microscopy and three-dimensional reconstruction of the calcium channel/ryanodine receptor from skeletal muscle. J. Cell Biol. 127, 411-423 (1994).

17. Serysheva, I.I., et al. Electron cryomicroscopy and angular reconstitution to visualize the skeletal muscle calcium-release channel. Nature Struct. Biol. 2, 18-24 (1995).

18. van Heel, M. Angular reconstitution: A posteriori assignment of projection directions for 3D reconstruction. Ultramicroscopy 21, 111-124 (1987).

19. van Heel, M., Winkler, H., Orlova, E. & Schatz, M. Structural analysis of ice-embedded single particles. Scanning Microscopy Suppl.6, 23-42 (1992).

20. Tinker, A. & Williams, A.J. Probing the structure of the conduction pathway of the sheep cardiac sarcoplasmic reticulum calcium-release channel with permanent and impermeant organic cations. J. Gen. Physiol. 102, 1107-1129 (1993).

21. Tu, Q., Velez, P., Browick, M. & Fill, M. Streaming potentials reveal a short ryanodine-sensitive selectivity filter in cardiac Ca2+ release channel. Biophys. J. 67, 2280-2285 (1994).

22. Tinker, A. & Williams, A.J. Measuring the length of the pore of the sheep cardiac sarcoplasmic reticulum calcium-release channel using related trimethylammonium ions as molecular calipers. Biophys. J. 68, 111-120 (1995).

23. Zhang, Y., Chen, H.S. & Khanna, V.K. A mutation in the human ryanodine receptor gene associated with central core disease. Nature Genetics 5, 46-50 (1993).

24. Fujii, J., Otsu, K. & Zorzoto, F. Identification of a mutation in porcine ryanodine receptor associated with malignant hyperthermia. Science 253, 448-451 (1991).

25. Ikemoto, N., Antoniuu, B. & Meszaros, L.G. Rapid flow chemical quench studies of calcium release from isolated sarcoplasmic reticulum. J. Biol. Chem. 260, 14096-14100 (1985).

26. Unwin, P.N.T. & Ennis, P.D. Calcium-mediated Changes in Gap Junction Structure: Evidence from the Low Angle X-ray Pattern. J. Cell Biol. 97, 1459-1466 (1983).

27. Unwin, P.N.T. Acetylcholine receptor channel imaged inthe open state. Nature 373, 37-43 (1995).

28. Flagg-Newton, J., Simpson, I. & Loewenstein, W.R. Permeability of the cell-to-cell membrane channels in mammalian cell junction. Science 205, 404-407 (1979).

29. Schwarzmann, G., Wiegandt, H. & Rose, B. Diameter of the cell-to-cell junctional membrane channels as probed with neutral molecules. Science 213, 551-553 (1981).

30. Loesser, K.E., Catellani, L. & Franzini-Armstrong, C. Disposition of junctional feet in muscles of invertebrates. J. Muscle Res. Cell Motil. 13, 161-173 (1992).

31. Schatz, M., Orlova, E.V., Dube, P., Jäger, J. & van Heel, M. Structure of Lumbricus terrestris hemoglobin at 30 Å resolution determined using angular reconstitution. J. Struct. Biol. 114, 28-40 (1995).

32. Hawkes, M.J., Diaz-Muñoz, M. & Hamilton, S.L. A procedure for purification of the ryanodine receptor from skeletal muscle. Memb. Biochem. 8, 133-145 (1989).

33. Zhou, Z.H., et al. CTF determination of images of ice-embedded single particles using a graphics interface. J. Struct. Biol. 116, 216-222 (1996).

34. van Heel, M., Harauz, G., Orlova, E.V., Schmidt, R. & Schatz, M. A new generation of the IMAGIC image processing system. J. Struct. Biol. 116, 17-24 (1996).

35. Dube, P., Tavares, P., Lurz, R. & van Heel, M. The portal protein of bacteriophage SPP1: a DNA pump with 13-fold symmetry. EMBO J. 12, 1303-1309 (1993).

36. van Heel, M. Classification of very large electron microscopical image data sets. Optik 82, 114-126 (1989).

37. Harauz, G. & van Heel, M. Exact filters for general geometry three dimensional reconstruction. Optik 73, 146-156 (1986).

38. van Heel, M. & Harauz, G. Resolution criteria for three dimensional reconstructions. Optik 73, 119-122 (1986).

Acknowledgment: We thank Dr. Michael Schatz and Mr. Ralf Schmidt of Image Science Software GmbH for software support. This work is supported by grants from the Deutsche Forschungsgemeinschaft to MvH ; from MDA and the National Institutes of Health to SLH; from the NCRR of National Institutes of Health, National Science Foundation, the W. M. Keck Foundation, the R. Welch Foundation and Office of Research at Baylor College of Medicine to WC.

to NCMI home page.