Calcium release during skeletal muscle excitation-contraction (EC) coupling occurs at the junctions between the sarcoplasmic reticulum (SR) and either the plasma membrane or the transverse (T) tubules. These Ca2+ release units are characterized by a specific molecular composition and their specific structural organization, both of which are important for the tissue-specific mode of skeletal muscle EC coupling. Upon depolarization Ca2+ is being released from the SR via the skeletal muscle ryanodine receptor (RyR1). The voltage sensor and trigger for this Ca2+ release process is the skeletal isoform of a L-type Ca2+ channel, the dihydropyridiene (DHP) receptor, which activates the RyR1 faster and at lower voltages than its own conductance pore. It is assumed that this voltage-dependent activation of SR Ca2+ release occurs by (direct or indirect) physical interactions between the DHP-receptor and the RyR1. This physical interaction, in turn requires the close association of the two channels within the SR-T tubule or SR-plasma membrane junctions. Indeed, DHPRs in the junctional membranes are regularly arranged in groups of four, called the tetrads, which correspond in size and orientation exactly to the position of every other RyR in the opposing SR membrane. That this highly coordinated molecular arrangement of DHPR and RyR is important for the tissue specific mode of EC-coupling in skeletal muscle is further substantiated by the observation that in cardiac muscle, which requires the influx of trigger Ca2+ for the activation of EC coupling, the DHPRs do not form tetrads opposite the RyR arrays [1].
Figure 1. Molecular organization of RyRs and DHPRs in skeletal muscle triads. Left: Electronmicrographs of transverse section and freeze-fracture showing feet and tetrads, respectively. Center: Corresponding models of RyR and DHPR assemblies in the junction. Right: Double immunofluorescence images showing colocalization of RyR and a heterologously expressed Ca2+ channel construct (a1Aas1607-1661) in the junctions of a dysgenic myotube.

In this project we address the questions as to how this highly organized molecular assembly is formed during muscle development? What are the mechanisms for the targeting and immobilization of the DHPRs and RyRs into their respective membrane systems? And how do alterations of this structural organization affect their function in EC coupling?

A powerful approach to address these questions has been the reconstitution of null-mutant myotubes with recombinant DHPR and RyR isoforms. Myotubes grown from mutant muscle or immortalized cell lines thereof are transiently transfected with expression plasmids encoding wild type, mutant, or chimeric isoforms of the lacking EC coupling protein. The ability of the heterologously expressed constructs to restore normal structure and function of the EC coupling apparatus can be analyzed using fluorescence microscopy, patch-clamp and fluorometric Ca2+ recording techniques. This approach combines all the possibilities of recombinant DNA methods with the virtues of studying the EC coupling proteins in their native environment, skeletal muscle cells. The dysgenic mouse, with a spontaneous mutation in the skeletal muscle DHPR a1S subunit gene, was the first to be used in such studies and aided in the molecular identification of the a1 subunit as the voltage sensor in EC coupling and in the characterization of tissue-specific properties to specific regions of its primary structure [2]. With the advent of gene targeting methods, recombinant reconstitution of null-mutant muscle cells has been extended to the study of the RyR and to other DHPR subunits [3,4]. We used this approach for the analysis of the molecular mechanisms underlying the molecular assembly of the Ca2+ release units in skeletal muscle.
The first important clues about the nature of the targeting mechanisms for the DHPR and RyR came from the analysis of the null-mutants themselves. Since T tubule-SR junctions formed in both the dysgenic (DHPR a1S -/-) [5] and dyspedic (RyR1 -/-) [3] myotubes and these junctions contained clusters of RyR1 and DHPRs, respectively, direct interactions between the two Ca2+ channels could not be important for junction formation nor for the incorporation of either one of the channels into the junctions. Consequently, both DHPRs and RyRs need to possess individual triad targeting mechanisms. With respect to the DHPR, analysis of the dysgenic myotubes indicated that the crucial information for triad targeting resides in the a1S subunit rather than in the accessory a2d or b subunits. Whereas the a2d subunit is expressed in dysgenic myotubes, without a1S it is mistargeted into the plasma membrane [6]. When a b1a -GFP fusion protein was overexpressed in dysgenic myotubes, it was not incorporated into triad junctions but remained diffusely distributed throughout the cytoplasm [7]. Only when it was coexpressed together with the a1S subunit both became correctly incorporated in the junctions. Moreover, disruption of the a1S-b binding by a single residue substitution within the b-binding motif within the I-II loop of a1S, dissociated the b subunit from a1S but did not block the normal incorporation of a1S in the junctions. Interestingly without the b subunit, in myotubes of the b knock-out mouse, a1S is not expressed in the outer membranes [4], and when the b interaction domain in the I-II loop of a1S is fully deleted a1S remains stuck in the ER [7]. Thus, the formation of a1Sˇb complexes may play an important role in early steps of the targeting process. Finally, the fact that a g knock-out is viable and capable of skeletal muscle type EC coupling indicates that this subunit is not essential for targeting or immobilization of the DHPR in the triad junctions [8].
If the accessory DHPR subunits depend on the a1S subunit for their own targeting into the triad, the signal for triad targeting must reside in the a1S subunit itself. Of the a1 subunit isoforms hitherto expressed in dysgenic myotubes, the L-type channels (class S, C, and D) but not the none-L-type channel (class A) were correctly targeted into the triads. Thus, chimeras of the targeted a1S and the non-targeted a1A were created and expressed in dysgenic myotubes to probe for the location of the targeting signal in the primary sequence of a1S [9]. Of all the large intracellular segments tested, only the C-terminus of a1S seemed to be essential for triad targeting. The targeting property of a1S could be conferred to a1A by replacing its C-terminus with that of a1S. But even though a1S becomes incorporated into triads as the full-length form, the distal portion of the C-terminus is not critical for normal targeting or restoration of EC coupling [10]. Successive shortening of the a1S C-terminal sequence able to confer the targeting property to a1A revealed that an essential triad targeting signal is contained within a 55 amino acid sequence between position 1607 and 1661 of a1S. Further shortening of the sequence or replacement of individual residues conserved between targeted a1 isoforms reduced the efficiency of targeting indicating that the signal is not encoded in a simple sequence motif of conserved residues but also requires the integrity of the overall structure of this C-terminal domain.
Reconstitution of dysgenic myotubes with a1S not only restores triad targeting but also the organization of DHPRs into tetrads, the depolarization-induced release of Ca2+ from the SR and the RyR1-dependent enhancement of L-type Ca2+ currents [9,11]. Expression of the cardiac a1C subunit restores triad targeting without the formation of DHPR tetrads, and cardiac EC coupling via Ca2+-induced Ca2+ release. In contrast, expression of the neuronal a1A, which is not targeted into triads, yields large current densities, which were however incapable of restoring EC coupling except in few isolated instances. But when a1A was rerouted into the triad by transferring onto it the C-terminal targeting signal, EC coupling properties improved in parallel with its triad targeting properties. These experiments suggest that the localization of the DHPR opposite the RyR is not only important for the Ca2+-independent, skeletal mode of EC coupling, but also for cardiac-type coupling by Ca2+-induced Ca2+ release. This interpretation is further substantiated by recent observations showing that the correctly targeted a1D, which has current densities as small as those of a1S but lacks the ability of direct skeletal coupling, very efficiently restores EC coupling via Ca2+-induced Ca2+ release [12]. Thus, for the activation of cardiac-type EC coupling (at least in reconstituted dysgenic myotubes) the location of the Ca2+ influx is much more important than the absolute amounts of Ca2+ entering the cell in response to depolarization.
The precise superposition of four DHPRs, the tetrads, on every other RyR in the junctions is believed to be the structural basis of Ca2+-independent EC coupling in skeletal muscle and for the retrograde current enhancement. The domain responsible for these tissue specific functional interactions between the DHPR and RyR1 has been carefully traced to a 45 residue sequence within the II-III loop of a1S using the chimera approach [11,13]. The question as to whether the same sequence determines also the assembly of DHPR tetrads is currently subject of investigations. Preliminary results of studies in the DHPR as well as in the RyR1 suggest that the domains responsible for tetrad formation overlap but do not fully correspond to those responsible for tissue-specific functional interactions.
Figure 2. Three consecutive steps (right) and possible molecular interactions (left) in the organization of the DHPR in the triad. (1) Interactions of b1a with the I-II loop of a1S facilitate export from the ER; (2) a C-terminal signal sequence is responsible for the targeting of the a1S into the junctional T-tubule membrane; (3) interactions between a1S and the RyR1 (direct or indirect) are responsible for the tissue-specific arrangement of the DHPR opposite the RyRs.
One problem that still remains unresolved concerns the origin and the significance of the alternating association of tetrads and RyRs. Analysis of the recruitment of DHPR particles into the junctions revealed that tetrads are not preformed structures but assemble one-by-one during triad formation [14]. The unexpected finding was that individual DHPR particles in immature junction are found only in the positions of future tetrads over alternating RyRs and seem to avoid the in-between RyRs. But what determines this curious preference. A possible explanation would be RyR arrays composed of two molecular isoforms - one that allows DHPR association and one that does not. Indeed many muscles express two RyR isoforms, the skeletal RyR1 and the ubiquitous RyR3. Moreover RyR3 is localized in the junctions in mammalian muscle fibers [15]. However, alternating dispositions of tetrads have also been observed in muscles not expressing RyR3. If RyR3 do not sit in the alternating positions of the RyR arrays, where in the junctions are they? A recent finding of an electron microscopic study may give the answer to this puzzle. Additional rows of large membrane proteins were observed adjacent to the double rows of RyR "feet" just outside the junctions [16]. These so called "parajunctional feet" were so far only observed in preparations of muscles known to express RyR3 and thus represent good candidates for the position of this second isoform of Ca2+ release channels in the triads.
Recent investigations of the EC coupling apparatus combining molecular, structural, and physiological techniques have shed new light on the extraordinary molecular machine that generates the precisely tuned Ca2+ release in response to skeletal muscle excitation. Increasingly it becoming clearer that the spatial arrangements of the channels within the Ca2+ release units is of utmost significance for their specific function in EC coupling. The picture of how DHPRs are incorporated in the junctions is beginning to take shape (Figure 2): The first hurdle along the way is the export of a1S from the ER that most likely requires the interactions with the b subunit. Next a1S needs to find its way to the triad junction, a process which is guided by a triad targeting signal in the C-terminus of a1S. However, the receptor for this signal in the target membrane is still elusive. Finally, a1S is arranged in groups of four opposite alternating RyR receptors. This process seems to depend on specific DHPR-RyR1 interactions, possibly via some of the same molecular domains that determine the tissue-specific functional cross-talk. Unfortunately, much less is known about the mechanisms that guide the targeting and assembly of the RyR arrays in the junctions. New candidate triad proteins together with the approach of reconstituting null-mutants can be expected to soon fill in the gaps in our understanding of the EC coupling process and help us resolve the mysteries of the molecular organization and assembly of the Ca2+ release units in skeletal muscle.
1. Franzini-Armstrong, C., Protasi, F., and Ramesh, V. 1998. Ann N Y Acad Sci. 853:20-30.
2. Tanabe, T., Beam, K.G., Powell, J.A., and Numa, S. 1988. Nature. 336:134-139.
3. Takekura, H., Nishi, M., Noda, T., Takeshima, H., and Franzini-Armstrong, C. 1995. Proc. Natl. Acad. Sci. USA. 92:3381-3385.
4. Strube, C., Beurg, M., Powers, P.A., Gregg, R.G., and Coronado, R. 1996. Biophys. J. 71:2531-2543.
5. Powell, J.A., Petherbridge, L., and Flucher, B.E. 1996. J. Cell Biol. 134, 375-387.
6. Flucher, B.E., Phillips, J.L.,and Powell, J.A. 1991. J. Cell Biol. 115, 1345-1356.
7. Neuhuber, B., Gerster, U., Döring, F., Glossmann, H., Tanabe, T., and Flucher, B.E. 1998. Proc. Natl. Acad. Sci. USA 95, 515-520.
8. Freise, D., Held, B., Wissenbach, U., Pfeifer, A., Trost, C., Himmerkus, N., Schweig, U., Freichel, M., Biel, M., Hofmann, F., Hoth, M., and Flockerzi, V. 2000. J. Biol. Chem. 275:14476-14481.
9. Flucher, B.E., Kasielke, N., and Grabner, M. 2000. J. Cell Biol, 151, 467-478.
10. Flucher, B.E., Kasielke, N., Gerster, U., Neuhuber, B., and Grabner, M. 2000. FEBS Letters, 474, 93-98.
11. Grabner, M., Dirksen, R.T., Suda, N., and Beam, K.G. 1999. J Biol Chem. 274:21913-21919.
12. Kasielke, N., Reimer, D., Obermair, G., Grabner, M., and Flucher, B.E. 2001. Biophys. J. 80:592a.
13. Wilkens, C.M., Kasielke, N., Flucher, B.E., Beam, K.G., and Grabner, M. 2001. Proc. Natl. Acad. Sci. USA, 98, 5892-7.
14. Protasi, F., Franzini-Armstrong, C., and Flucher, B.E. 1997. J. Cell Biol. 137, 859-870.
15. Flucher, B.E., Conti, A., Takeshima, H., and Sorrentino, V. 1999. J. Cell Biol. 146, 621-630.
16. Felder, E., and Franzini-Armstrong, C. 2002. Proc. Natl. Acad. Sci. USA. 99, 1695-700.

Excitation-Contraction Coupling