Exercise-induced alterations and loss of sarcomeric M-line organization in the diaphragm muscle of obscurin knockout mice.

We recently reported that skeletal muscle fibers of obscurin knockout (KO) mice present altered distribution of ankyrin B (ankB), disorganization of the subsarcolemmal microtubules, and reduced localization of dystrophin at costameres. In addition, these mice have impaired running endurance and increased exercise-induced sarcolemmal damage compared with wild-type animals. Here, we report results from a combined approach of physiological, morphological, and structural studies in which we further characterize the skeletal muscles of obscurin KO mice. A detailed examination of exercise performance, using different running protocols, revealed that the reduced endurance of obscurin KO animals on the treadmill depends on exercise intensity and age. Indeed, a mild running protocol did not evidence significant differences between control and obscurin KO mice, whereas comparison of running abilities of 2-, 6-, and 11-mo-old mice exercised at exhaustion revealed a progressive age-dependent reduction of the exercise tolerance in KO mice. Histological analysis indicated that heavy exercise induced leukocyte infiltration, fibrotic connective tissue deposition, and hypercontractures in the diaphragm of KO mice. On the same line, electron microscopy revealed that, in the diaphragm of exercised obscurin KO mice, but not in the hindlimb muscles, both M-line and H-zone of sarcomeres appeared wavy and less defined. Altogether, these results suggest that obscurin is required for the maintenance of morphological and ultrastructural integrity of skeletal muscle fibers against damage induced by intense mechanical stress and point to the diaphragm as the skeletal muscle most severely affected in obscurin-deficient mice.

Lifelong physical exercise delays age-associated skeletal muscle decline.

Aging is usually accompanied by a significant reduction in muscle mass and force. To determine the relative contribution of inactivity and aging per se to this decay, we compared muscle function and structure in (a) male participants belonging to a group of well-trained seniors (average of 70 years) who exercised regularly in their previous 30 years and (b) age-matched healthy sedentary seniors with (c) active young men (average of 27 years). The results collected show that relative to their sedentary cohorts, muscle from senior sportsmen have: (a) greater maximal isometric force and function, (b) better preserved fiber morphology and ultrastructure of intracellular organelles involved in Ca(2+) handling and ATP production, (c) preserved muscle fibers size resulting from fiber rescue by reinnervation, and (d) lowered expression of genes related to autophagy and reactive oxygen species detoxification. All together, our results indicate that: (a) skeletal muscle of senior sportsmen is actually more similar to that of adults than to that of age-matched sedentaries and (b) signaling pathways controlling muscle mass and metabolism are differently modulated in senior sportsmen to guarantee maintenance of skeletal muscle structure, function, bioenergetic characteristics, and phenotype. Thus, regular physical activity is a good strategy to attenuate age-related general decay of muscle structure and function (ClinicalTrials.gov: NCT01679977).

Deletion of small ankyrin 1 (sAnk1) isoforms results in structural and functional alterations in aging skeletal muscle fibers.

Muscle-specific ankyrins 1 (sAnk1) are a group of small ankyrin 1 isoforms, of which sAnk1.5 is the most abundant. sAnk1 are localized in the sarcoplasmic reticulum (SR) membrane from where they interact with obscurin, a myofibrillar protein. This interaction appears to contribute to stabilize the SR close to the myofibrils. Here we report the structural and functional characterization of skeletal muscles from sAnk1 knockout mice (KO). Deletion of sAnk1 did not change the expression and localization of SR proteins in 4- to 6-mo-old sAnk1 KO mice. Structurally, the main modification observed in skeletal muscles of adult sAnk1 KO mice (4-6 mo of age) was the reduction of SR volume at the sarcomere A band level. With increasing age (at 12-15 mo of age) extensor digitorum longus (EDL) skeletal muscles of sAnk1 KO mice develop prematurely large tubular aggregates, whereas diaphragm undergoes significant structural damage. Parallel functional studies revealed specific changes in the contractile performance of muscles from sAnk1 KO mice and a reduced exercise tolerance in an endurance test on treadmill compared with control mice. Moreover, reduced Qγ charge and L-type Ca(2+) current, which are indexes of affected excitation-contraction coupling, were observed in diaphragm fibers from 12- to 15-mo-old mice, but not in other skeletal muscles from sAnk1 KO mice. Altogether, these findings show that the ablation of sAnk1, by altering the organization of the SR, renders skeletal muscles susceptible to undergo structural and functional alterations more evident with age, and point to an important contribution of sAnk1 to the maintenance of the longitudinal SR architecture.

Massive alterations of sarcoplasmic reticulum free calcium in skeletal muscle fibers lacking calsequestrin revealed by a genetically encoded probe.

The cytosolic free Ca(2+) transients elicited by muscle fiber excitation are well characterized, but little is known about the free [Ca(2+)] dynamics within the sarcoplasmic reticulum (SR). A targetable ratiometric FRET-based calcium indicator (D1ER Cameleon) allowed us to investigate SR Ca(2+) dynamics and analyze the impact of calsequestrin (CSQ) on SR [Ca(2+)] in enzymatically dissociated flexor digitorum brevis muscle fibers from WT and CSQ-KO mice lacking isoform 1 (CSQ-KO) or both isoforms [CSQ-double KO (DKO)]. At rest, free SR [Ca(2+)] did not differ between WT, CSQ-KO, and CSQ-DKO fibers. During sustained contractions, changes were rather small in WT, reflecting powerful buffering of CSQ, whereas in CSQ-KO fibers, significant drops in SR [Ca(2+)] occurred. Their amplitude increased with stimulation frequency between 1 and 60 Hz. At 60 Hz, the SR became virtually depleted of Ca(2+), both in CSQ-KO and CSQ-DKO fibers. In CSQ-KO fibers, cytosolic free calcium detected with Fura-2 declined during repetitive stimulation, indicating that SR calcium content was insufficient for sustained contractile activity. SR Ca(2+) reuptake during and after stimulation trains appeared to be governed by three temporally distinct processes with rate constants of 50, 1-5, and 0.3 s(-1) (at 26 °C), reflecting activity of the SR Ca(2+) pump and interplay of luminal and cytosolic Ca(2+) buffers and pointing to store-operated calcium entry (SOCE). SOCE might play an essential role during muscle contractures responsible for the malignant hyperthermia-like syndrome in mice lacking CSQ.

Stable muscle atrophy in long-term paraplegics with complete upper motor neuron lesion from 3- to 20-year SCI.

STUDY DESIGN: Unrandomized trial. OBJECTIVES: To investigate the structural and functional relationships and the progression of muscle atrophy up to 20 years of spastic paraplegia. SETTING: Clinical follow-up in Vienna, Austria; muscle biopsies analyzed by light microscopy in Padova and by electron microscopy (EM) in Chieti, Italy. METHODS: Force was measured as knee extension torque; trophism by computer tomography scan; tissue composition and fiber morphology by histopathology and EM. RESULTS: In the long-term group of patients (17.0+/-2.6 years), force and size of thigh muscles were only slightly different from those of mid-term subjects (2.2+/-0.5 years). Histology and ultrastructure confirm that the difference in average size of muscle fibers between long-term and mid-term paralyzed leg muscles is actually very small. In addition, muscle fibers maintain the striated appearance characteristic of normal skeletal fibers even after 14-20 years of paralysis. Ultrastructural alterations of the activating and metabolic machineries, and the presence of fibers with lower motor neuron denervation features, may explain the low-force output and the reduced endurance of paretic muscles. CONCLUSION: The stable muscle atrophy that characterizes long-lasting spastic paraplegia suggests that there are no upper-time limits to begin a training program based on functional electrical stimulation.

Atrophy, but not necrosis, in rabbit skeletal muscle denervated for periods up to one year.

Our understanding of the effects of long-term denervation on skeletal muscle is heavily influenced by an extensive literature based on the rat. We have studied physiological and morphological changes in an alternative model, the rabbit. In adult rabbits, tibialis anterior muscles were denervated unilaterally by selective section of motor branches of the common peroneal nerve and examined after 10, 36, or 51 wk. Denervation reduced muscle mass and cross-sectional area by 50-60% and tetanic force by 75%, with no apparent reduction in specific force (force per cross-sectional area of muscle fibers). The loss of mass was associated with atrophy of fast fibers and an increase in fibrous and adipose connective tissue; the diameter of slow fibers was preserved. Within fibers, electron microscopy revealed signs of ultrastructural disorganization of sarcomeres and tubular systems. This, rather than the observed transformation of fiber type from IIx to IIa, was probably responsible for the slow contractile speed of the muscles. The muscle groups denervated for 10, 36, or 51 wk showed no significant differences. At no stage was there any evidence of necrosis or regeneration, and the total number of fibers remained constant. These changes are in marked contrast to the necrotic degeneration and progressive decline in mass and force that have previously been found in long-term denervated rat muscles. The rabbit may be a better choice for a model of the effects of denervation in humans, at least up to 1 yr after lesion.

Type 1 and type 3 ryanodine receptors generate different Ca(2+) release event activity in both intact and permeabilized myotubes.

In this investigation we use a "dyspedic" myogenic cell line, which does not express any ryanodine receptor (RyR) isoform, to examine the local Ca(2+) release behavior of RyR3 and RyR1 in a homologous cellular system. Expression of RyR3 restored caffeine-sensitive, global Ca(2+) release and causes the appearance of relatively frequent, spontaneous, spatially localized elevations of [Ca(2+)], as well as occasional spontaneous, propagating Ca(2+) release, in both intact and saponin-permeabilized myotubes. Intact myotubes expressing RyR3 did not, however, respond to K(+) depolarization. Expression of RyR1 restored depolarization-induced global Ca(2+) release in intact myotubes and caffeine-induced global release in both intact and permeabilized myotubes. Both intact and permeabilized RyR1-expressing myotubes exhibited relatively infrequent spontaneous Ca(2+) release events. In intact myotubes, the frequency of occurrence and properties of these RyR1-induced events were not altered by partial K(+) depolarization or by application of nifedipine, suggesting that these RyR1 events are independent of the voltage sensor. The events seen in RyR1-expressing myotubes were spatially more extensive than those seen in RyR3-expressing myotubes; however, when analysis was limited to spatially restricted "Ca(2+) spark"-like events, events in RyR3-expressing myotubes were larger in amplitude and duration compared with those in RyR1. Thus, in this skeletal muscle context, differences exist in the spatiotemporal properties and frequency of occurrence of spontaneous release events generated by RyR1 and RyR3. These differences underscore functional differences between the Ca(2+) release behavior of RyR1 and RyR3 in this homologous expression system.

RYR1 and RYR3 have different roles in the assembly of calcium release units of skeletal muscle.

Calcium release units (CRUs) are junctions between the sarcoplasmic reticulum (SR) and exterior membranes that mediates excitation contraction (e-c) coupling in muscle cells. In skeletal muscle CRUs contain two isoforms of the sarcoplasmic reticulum Ca(2+)release channel: ryanodine receptors type 1 and type 3 (RyR1 and RyR3). 1B5s are a mouse skeletal muscle cell line that carries a null mutation for RyR1 and does not express either RyR1 or RyR3. These cells develop dyspedic SR/exterior membrane junctions (i.e., dyspedic calcium release units, dCRUs) that contain dihydropyridine receptors (DHPRs) and triadin, two essential components of CRUs, but no RyRs (or feet). Lack of RyRs in turn affects the disposition of DHPRs, which is normally dictated by a linkage to RyR subunits. In the dCRUs of 1B5 cells, DHPRs are neither grouped into tetrads nor aligned in two orthogonal directions. We have explored the structural role of RyR3 in the assembly of CRUs in 1B5 cells independently expressing either RyR1 or RyR3. Either isoform colocalizes with DHPRs and triadin at the cell periphery. Electron microscopy shows that expression of either isoform results in CRUs containing arrays of feet, indicating the ability of both isoforms to be targeted to dCRUs and to assemble in ordered arrays in the absence of the other. However, a significant difference between RyR1- and RyR3-rescued junctions is revealed by freeze fracture. While cells transfected with RyR1 show restoration of DHPR tetrads and DHPR orthogonal alignment indicative of a link to RyRs, those transfected with RyR3 do not. This indicates that RyR3 fails to link to DHPRs in a specific manner. This morphological evidence supports the hypothesis that activation of RyR3 in skeletal muscle cells must be indirect and provides the basis for failure of e-c coupling in muscle cells containing RyR3 but lacking RyR1 (see the accompanying report, ).

Expression of ryanodine receptor RyR3 produces Ca2+ sparks in dyspedic myotubes.

Discrete, localized elevations of myoplasmic [Ca2+], Ca2+ 'sparks', were readily detected using the fluorescent Ca2+ indicator fluo-3 and laser scanning confocal microscopy in 'dyspedic' 1B5 myotubes, i.e. myotubes which do not express ryanodine receptors (RyRs), transduced with virions containing cDNA for RyR type 3 that were saponin permeabilized to allow dye entry. Ca2+ sparks were never observed in non-transduced RyR null myotubes. The spatial locations of sparks observed in permeabilized myotubes roughly corresponded to regions of RyR protein expression in the same myotube as detected after subsequent fixation and antibody staining. Permeabilized RyR3-transduced myotubes exhibited similar punctate peripheral RyR3 protein immunohistochemical patterns as myotubes fixed before permeabilization indicating that permeabilization did not affect the structural organization of the triad. Ca2+ sparks, recorded in line scan mode, in permeabilized myotubes expressing RyR3 exhibited mean amplitudes (change in fluorescence/mean fluorescence, DeltaF/F: 1.20 +/- 0.04) and temporal rise times (10-90%; 6.31 +/- 0.12 ms) similar to those of sparks recorded in permeabilized frog skeletal muscle fibres (0.98 +/- 0.01; 6.11 +/- 0.07, respectively) using the same confocal system. Spatial extent and temporal duration of the Ca2+ sparks were approximately 40% larger in the RyR3-expressing myotube cultures than in frog fibres. Ca2+ sparks recorded in line scan mode often occurred repetitively at the same spatial location in RyR3-expressing myotubes. Such repetitive events were highly reproducible in amplitude and spatio-temporal properties, as previously observed for repetitive mode sparks in frog skeletal muscle. Ca2+ sparks recorded in xy mode were frequently compressed in the y (slower scan) direction compared to the x direction. This asymmetry was reproduced assuming spatially symmetric events having the time course of Ca2+ sparks recorded in line scan (xt) mode. These expression studies demonstrate that the presence of RyR3 is sufficient for the production of Ca2+ sparks in a skeletal muscle system lacking the expression of any other RyR isoform.

HSV-1 amplicon vectors are a highly efficient gene delivery system for skeletal muscle myoblasts and myotubes.

Analysis of RyR1 structure function in muscle cells is made difficult by the low (<5%) transfection efficiencies of myoblasts or myotubes using calcium phosphate or cationic lipid techniques. We inserted the full-length 15.3-kb RyR1 cDNA into a herpes simplex virus type 1 (HSV-1) amplicon vector, pHSVPrPUC between the ori/IE 4/5 promoter sequence and the HSV-1 DNA cleavage/packaging signal (pac). pHSVGN and pHSVGRyR1, two amplicons that expressed green fluorescent protein, were used for fluorescence-activated cell sorter analysis of transduction efficiency. All amplicons were packaged into HSV-1 virus particles using a helper virus-free packaging system and yielded 10(6) transducing vector units/ml. HSVRyR1, HSVGRyR1, and HSVGN virions efficiently transduced mouse myoblasts and myotubes, expressing the desired product in 70-90% of the cells at multiplicity of infection 5. The transduced cells appeared healthy and RyR1 produced by this method was targeted properly and restored skeletal excitation-contraction coupling in dyspedic myotubes. The myotubes produced sufficient protein to allow single-channel analyses from as few as 10 100-mm dishes. In most cases this method could preclude the need for permanent transfectants for the study of RyR1 structure function.

Shape, size, and distribution of Ca(2+) release units and couplons in skeletal and cardiac muscles.

Excitation contraction (e-c) coupling in skeletal and cardiac muscles involves an interaction between specialized junctional domains of the sarcoplasmic reticulum (SR) and of exterior membranes (either surface membrane or transverse (T) tubules). This interaction occurs at special structures named calcium release units (CRUs). CRUs contain two proteins essential to e-c coupling: dihydropyridine receptors (DHPRs), L-type Ca(2+) channels of exterior membranes; and ryanodine receptors (RyRs), the Ca(2+) release channels of the SR. Special CRUs in cardiac muscle are constituted by SR domains bearing RyRs that are not associated with exterior membranes (the corbular and extended junctional SR or EjSR). Functional groupings of RyRs and DHPRs within calcium release units have been named couplons, and the term is also loosely applied to the EjSR of cardiac muscle. Knowledge of the structure, geometry, and disposition of couplons is essential to understand the mechanism of Ca(2+) release during muscle activation. This paper presents a compilation of quantitative data on couplons in a variety of skeletal and cardiac muscles, which is useful in modeling calcium release events, both macroscopic and microscopic ("sparks").

Role of ryanodine receptors in the assembly of calcium release units in skeletal muscle.

In muscle cells, excitation-contraction (e-c) coupling is mediated by "calcium release units," junctions between the sarcoplasmic reticulum (SR) and exterior membranes. Two proteins, which face each other, are known to functionally interact in those structures: the ryanodine receptors (RyRs), or SR calcium release channels, and the dihydropyridine receptors (DHPRs), or L-type calcium channels of exterior membranes. In skeletal muscle, DHPRs form tetrads, groups of four receptors, and tetrads are organized in arrays that face arrays of feet (or RyRs). Triadin is a protein of the SR located at the SR-exterior membrane junctions, whose role is not known. We have structurally characterized calcium release units in a skeletal muscle cell line (1B5) lacking Ry1R. Using immunohistochemistry and freeze-fracture electron microscopy, we find that DHPR and triadin are clustered in foci in differentiating 1B5 cells. Thin section electron microscopy reveals numerous SR-exterior membrane junctions lacking foot structures (dyspedic). These results suggest that components other than Ry1Rs are responsible for targeting DHPRs and triadin to junctional regions. However, DHPRs in 1B5 cells are not grouped into tetrads as in normal skeletal muscle cells suggesting that anchoring to Ry1Rs is necessary for positioning DHPRs into ordered arrays of tetrads. This hypothesis is confirmed by finding a "restoration of tetrads" in junctional domains of surface membranes after transfection of 1B5 cells with cDNA encoding for Ry1R.

Contractile impairment and structural alterations of skeletal muscles from knockout mice lacking type 1 and type 3 ryanodine receptors.

Skeletal muscle contraction is triggered by the release of Ca2+ from the sarcoplasmic reticulum through the type 1 ryanodine receptor (RyR1). Recently it has been shown that also the type 3 isoform of ryanodine receptor (RyR3), which is expressed in some mammalian skeletal muscles, may participate in the regulation of skeletal muscle contraction. Here we report the generation and the characterization of double mutant mice carrying a targeted disruption of both the RyR1 and the RyR3 genes (RyR1-/-;RyR3-/-). Skeletal muscles from mice homozygous for both mutations are unable to contract in response to caffeine and to ryanodine. In addition, they show a very poor capability to develop tension when directly activated with micromolar [Ca2+]i after membrane permeabilization which indicates either poor development or degeneration of the myofibrils. This was confirmed by biochemical analysis of contractile proteins. Electron microscopy confirms small size of myofibrils and shows complete absence of feet (RyRs) in the junctional SR.

Comparative ultrastructure of Ca2+ release units in skeletal and cardiac muscle.

The sarcoplasmic reticulum (SR) of striated muscle fibers interacts with exterior membranes (surface membrane and transverse tubules) to form junctions that are involved in the internal release of calcium during excitation-contraction coupling. Release of calcium through the ryanodine receptors (RyRs) or calcium release channels of the SR is under the control of the L type calcium channels or dihydropyridine receptors (DHPRs) of exterior membranes. Interacting clusters of the two proteins constitute calcium release units. The cytoplasmic domains of RyRs are visible as large electron-dense structures (the feet) with four identical subunits in the junctional gap separating SR from exterior membranes. In freeze-fracture replicas of skeletal muscle, large intramembrane particles are grouped into clusters of tetrads in the exterior membranes, and the tetrads are located in correspondence of the four subunits of the feet. Lack of tetrads in dysgenic muscle fibers with a null mutation for DHPRs and appearance of the tetrads after transfection with cDNA for DHPR indicate identity of tetrads with four DHPRs. In cardiac muscle, DHPRs are located at the sites of SR-surface junctions, but they are not grouped into tetrads. This is consistent with a possible direct DHPR-RyR interaction in skeletal but not in cardiac muscle. The size and distribution of SR-surface junctions in skeletal and cardiac muscles provide further clues to their function.

Ryanodine receptors of striated muscles: a complex channel capable of multiple interactions.

The ryanodine receptor (RyR) is a high-conductance Ca2+ channel of the sarcoplasmic reticulum in muscle and of the endoplasmic reticulum in other cells. In striated muscle fibers, RyRs are responsible for the rapid release of Ca2+ that activates contraction. Ryanodine receptors are complex molecules, with unusually large cytoplasmic domains containing numerous binding sites for agents that control the state of activity of the channel-forming domain of the molecule. Structural considerations indicate that long-range interactions between cytoplasmic and intramembrane domains control channel function. Ryanodine receptors are located in specialized regions of the SR, where they are structurally and functionally associated with other intrinsic proteins and, indirectly, also with the luminal Ca2(+)-binding protein calsequestrin. Activation of RyRs during the early part of the excitation-contraction coupling cascade is initiated by the activity of surface-membrane Ca2+ channels, the dihydropyridine receptors (DHPRs). Skeletal and cardiac muscles contain different RyR and DHPR isoforms and both contribute to the diversity in cardiac and skeletal excitation-contraction coupling mechanisms. The architecture of the sarcoplasmic reticulum-surface junctions determines the types of RyR-DHPR interactions in the two muscle types.

Coordinated incorporation of skeletal muscle dihydropyridine receptors and ryanodine receptors in peripheral couplings of BC3H1 cells.

Rapid release of calcium from the sarcoplasmic reticulum (SR) of skeletal muscle fibers during excitation-contraction (e-c) coupling is initiated by the interaction of surface membrane calcium channels (dihydropyridine receptors; DHPRs) with the calcium release channels of the SR (ryanodine receptors; RyRs, or feet). We studied the early differentiation of calcium release units, which mediate this interaction, in BC3H1 cells. Immunofluorescence labelings of differentiating myocytes with antibodies against alpha1 and alpha2 subunits of DHPRs, RyRs, and triadin show that the skeletal isoforms of all four proteins are abundantly expressed upon differentiation, they appear concomitantly, and they are colocalized. The transverse tubular system is poorly organized, and thus clusters of e-c coupling proteins are predominantly located at the cell periphery. Freeze fracture analysis of the surface membrane reveals tetrads of large intramembrane particles, arranged in orderly arrays. These appear concomitantly with arrays of feet (RyRs) and with the appearance of DHPR/RyS clusters, confirming that the four components of the tetrads correspond to skeletal muscle DHPRs. The arrangement of tetrads and feet in developing junctions indicates that incorporation of DHPRs in junctional domains of the surface membrane proceeds gradually and is highly coordinated with the formation of RyR arrays. Within the arrays, tetrads are positioned at a spacing of twice the distance between the feet. The incorporation of individual DHPRs into tetrads occurs exclusively at positions corresponding to alternate feet, suggesting that the assembly of RyR arrays not only guides the assembly of tetrads but also determines their characteristic spacing in the junction.

Functional nonequality of the cardiac and skeletal ryanodine receptors.

Dihydropyridine receptors (DHPRs), which are voltage-gated Ca2+ channels, and ryanodine receptors (RyRs), which are intracellular Ca2+ release channels, are expressed in diverse cell types, including skeletal and cardiac muscle. In skeletal muscle, there appears to be reciprocal signaling between the skeletal isoforms of both the DHPR and the RyR (RyR-1), such that Ca2+ release activity of RyR-1 is controlled by the DHPR and Ca2+ channel activity of the DHPR is controlled by RyR-1. Dyspedic skeletal muscle cells, which do not express RyR-1, lack excitation-contraction coupling and have an approximately 30-fold reduction in L-type Ca2+ current density. Here we have examined the ability of the predominant cardiac and brain RyR isoform, RyR-2, to substitute for RyR-1 in interacting with the skeletal DHPR. When RyR-2 is expressed in dyspedic muscle cells, it gives rise to spontaneous intracellular Ca2+ oscillations and supports Ca2+ entry-induced Ca2+ release. However, unlike RyR-1, the expressed RyR-2 does not increase the Ca2+ channel activity of the DHPR, nor is the gating of RyR-2 controlled by the skeletal DHPR. Thus, the ability to participate in skeletal-type reciprocal signaling appears to be a unique feature of RyR-1.

Formation and maturation of the calcium release apparatus in developing and adult avian myocardium.

Muscle fibers release large amounts of calcium from an internal compartment, the sarcoplasmic reticulum (SR), during activation. Two proteins are involved in this process and its control: plasma membrane calcium channels, or dihydropyridine receptors (DHPRs), and SR calcium release channels, or ryanodine receptors (RyRs). The two proteins form part of a structural complex, perhaps unique to muscle cells, which allows an interaction between plasma membrane and SR, resulting in calcium release from the latter. The surface-SR interaction is a step in the coupling between electrical events in the plasma membrane and contraction (excitation-contraction coupling). The structural complexes have been called calcium release units. One key to further understanding the control of calcium homeostasis in muscle is knowledge of how DHPRs and RyRs assemble into calcium release units. We have studied the development of avian myocardium, using immunocytochemistry to locate DHPRs and RyRs and electron microscopy to follow the formation of calcium release units containing feet (RyRs) and large membrane particles (presumably DHPRs). We find that the initial step is a docking of SR vesicles to the plasma membrane, followed by the appearance of feet in the junctional gap between SR and plasma membrane. Feet aggregate in ordered arrays, and the arrays increase in size until they fill the entire junctional gap. Clustering of membrane particles, presumably DHPRs, is apparently coupled to clustering of feet, since the two junction components assemble within patches of membrane of approximately equal size and containing an approximately constant ratio of particles to feet. Thus, despite the fact that no evidence exists for a direct interaction between DHPRs and RyRs in cardiac muscle, some mechanism exists to ensure that the two molecules are clustered in proximity to each other and in the appropriate proportion.