Interaction between satellite cells and skeletal muscle fibers

Satellite cells of adult skeletal muscle are mononucleated cells that occupy shallow depressions in the myofiber. The cells have one surface in contact with the basal lamina of the myofiber and the other surface in contact with its plasmalemma with a narrow gap of about 15 nm between the two cells (Ishikawa, 1966). Despite this intimate contact, there is no evidence of electrical coupling between the cells (Bader et al. 1988). Satellite cells of adult muscle have a heterochromatic nucleus and scant cytoplasm with few organelles but both surfaces of the satellite cell show evidence of specialized activity in the form of abundant cell membrane caveolae (Schultz, 1976; Snow, 1977; Bischoff, 1979). The function of these is unknown. Satellite cells are mitotically quiescent in adult muscle but re-enter the cell cycle following muscle injury to generate a population of myoblasts, which may subsequently fuse to form new myofibers (Bischoff, 1975; Konigsberg et al. 1975; Snow, 1977). In developing muscle, satellite cells are mitotically active and contribute virtually all of the nuclei in the secondary generation of myofibers (Kelly and Zacks, 1969; Cardasis and Cooper, 1975; Kelly, 1978; Ross et al. 1987). Identifying signals that govern satellite cell behavior is of central importance in understanding muscle growth and regeneration.

A study of the regulation of satellite cell division was carried out using a tissue culture system consisting of single myofibers with attached satellite cells from adult rats (Bischoff, 1986a, 19866). The satellite cells are retained beneath the basal lamina of these fibers and most remain mitotically quiescent in vitro unless mitogens are added to the culture medium. Effective growth stimulators include fibroblast growth factor and an extract of crushed adult muscle (Bischoff, 19866). Results of these studies show that proliferation of satellite cells is under positive control since death of a cultured myofiber of itself had little effect on proliferation of attached satellite cells, but mitogens extracted from whole crushed muscles were able to stimulate the growth of satellite cells on intact myofibers. Muscle regeneration is a complex phenomenon, however. The regenerative response is proportional to the extent of tissue injury (McGeachie and Grounds, 1987) and the final outcome is influenced by innervation, vascularity, hormonal status and nutrition (Jirmanova and Thesleff, 1972; Hansen-Smith and Carlson, 1979; d’Albis et al. 1987; Phillips et al. 1987; Mulvaney et al. 1988). Thus it is likely that satellite cell growth is regulated by both positive and negative factors.

Among local factors that may influence satellite cell growth are the plasmalemma and the basal lamina of the myofiber since satellite cells have extensive area in contact with both of these surfaces. The basal lamina is thought to facilitate regeneration by providing a framework upon which new myofibers can form and become reinnervated (Vracko and Benditt, 1972; Sanes et al. 1978). The mechanisms by which the basal lamina influences regeneration are unknown, but may include stimulation of satellite cell growth and division.

The present study utilized myofibers with attached satellite cells in culture to evaluate separately the influence of the basal lamina and the plasmalemma of the myofiber on satellite cell proliferation induced by adding muscle extract to the culture medium. Results indicate that satellite cells in contact with the plasmalemma of the myofiber have reduced sensitivity to mitogen stimulation.

Single isolated fibers and attached satellite cells were prepared from the flexor digitorum brevis muscles of 2-to 3-month-old male rats as described previously (Bischoff, 1986a). The fibers were separated from non-muscle cells and cultured on coverslips in 35 mm plastic culture dishes. Several methods were used to attach fibers to the coverslips depending upon the requirements of the experiment. In some experiments, the fibers were embedded in a thin gel made from 50 μl rat tendon collagen solution containing about 3 mg protein ml⁻¹ 0.1% acetic acid. After restoring salt and pH to physiological levels, the collagen was mixed with 50 μl of a suspension of myofibers and incubated at 37 °C for 10 min to induce polymerization (Elsdale and Bard, 1972). In experiments requiring the fibers to be free on the substratum, coverslips were coated with a thin layer of the rat tendon collagen prepared as described and polymerized at 37°C. The coated coverslips were treated with 1 % glutaraldehyde (Electron Microscopy Sciences, Fort Washington, PA) in phosphate-buffered saline (PBS) for 1 h at room temperature, washed 5 times with PBS, treated with 0.5 mg ml’¹ polylysine (Sigma, St. Louis, MO) in PBS for 30 min and washed once with PBS. The treated coverslips were placed in 35 mm culture dishes and myofibers were attached by centrifugal force (100g for 10 min) using a custom-designed dish carrier in an HN-S centrifuge (International Equipment Co., Needham Heights, MA). No satellite cells were detached from the myofibers at this centrifugal force (see below). Culture medium was Eagle’s MEM containing 10% horse serum and 1% antibiotic-antimycotic mixture (GIBCO, Grand Island, NY). Culture medium was changed daily. In some recent experiments, the 10% horse serum was replaced with 20% Controlled Process Serum Replacement No.2 (Sigma) plus 1 % horse serum. Satellite cells remain viable, but few proliferate under these conditions. A saline extract of crushed adult rat muscle prepared as described (Bischoff, 19865) was added to the culture medium to induce proliferation of satellite cells.

Embryonic myoblasts were prepared from 19 day rat embryos by digestion of muscle tissue using 0.2 % crude trypsin (Sigma, 1:250) in Earle’s balanced salt solution for 30 min at 37°C. Cells were dissociated by trituration and separated from clumps and debris by filtration through 10 μm Nitex cloth (TETCO, NY). Suspensions were enriched for myogenic cells by preplating (Konigsberg, 1979) and the purified cells were cultured in Eagle’s MEM with 10% horse serum, 5% embryo extract and antibiotics. The embryo extract was prepared from 11 day chick embryos (Bischoff and Holtzer, 1969). All other chemicals were from Sigma.

To remove satellite cells from intact myofibers, a suspension of myofibers in PBS was placed in 35 mm culture dishes bearing collagen-polylysine-treated coverslips. The dishes were centrifuged in an HN-S centrifuge at full speed (about 1500g) for 10 min to allow attachment of the fibers to the substratum then the PBS was replaced with culture medium and centrifugation was continued for another 20 min.

Satellite cells attached to the basal lamina in the absence of the myofiber were obtained by killing the myofibers with a myotoxic anaesthetic. Cultures were incubated for 10 min in PBS containing 0.05% Marcaine (Winthrop-Breon, NY), washed twice with PBS and returned to normal culture medium. Myonuclei in Marcaine-killed myofibers completely lyse after overnight incubation. In some experiments, however, it was necessary to remove the myonuclei more rapidly and for this 33 μg ml⁻¹ deoxyribonuclease I (Sigma, type IV) was added to the culture medium after Marcaine treatment.

Satellite cell proliferation was measured by radioautography after incubating coverslip cultures in medium containing 0.5 μCiml⁻¹ [rne/W-³H]thymidine ([³H]TdR, 6.7CimM⁻¹, New England Nuclear, Boston, MA) from 44 to 48h in vitro. Coverslips were fixed in alcohol, Formalin, acetic acid (20:2:1), attached to slides and dipped in NTB2 bulk emulsion (Eastman Kodak, Rochester, NY) diluted 1,1 with water. After photographic processing according to the manufacturer’s recommendations, the cells were stained with Gill’s hematoxylin (Polysciences, Warrington, PA). Labeled satellite cells were scored in at least 50 fibers from each treatment.

Cultures were fixed with 4% formaldehyde in PBS, washed and incubated for 1 h in rabbit antibody against laminin diluted 1:200 with PBS (Bethesda Research Lab, Bethesda, MD). After washing, the cultures were incubated for 1 h in fluorescein-labeled goat anti-rabbit (GAR) IgG diluted 1:50 (Miles, Elkhart, IN). Control cultures were incubated in normal rabbit serum and fluorescein-GAR IgG or fluorescein-GAR IgG alone.

A Zeiss microscope fitted with an epifluorescent condenser and halogen lamp was used for fluorescent microscopy. Specimens were mounted in glycerol: PBS (9:1), pH8.7. For scanning electron microscopy, myofibers were centrifuged onto polylysine-coated coverslips and cultured for various periods. The cultures were fixed in 3 % glutaraldehyde in PBS, dehydrated in methanol, critical point dried, and sputter-coated with 12 – 15 nm gold. Cultures were examined with a Philips 501 microscope.

Proliferation of satellite cells in response to muscle extract was measured by [³H]TdR labeling in radioautographs under three conditions with respect to cell contact (Fig. 1). First, satellite cells were left in their normal location between the basal lamina and plasmalemma of the myofiber. Second, satellite cells were detached from the myofiber and basal lamina by centrifugation and deposited on the culture substratum directly beneath the myofiber. Third, satellite cells were removed from the influence of the plasmalemma, but left in contact with the basal lamina by killing the myofiber. Because of technical limitations, experiments comparing the three conditions were carried out separately using the normal myofibers as control in each case. In evaluating the results, however, the three conditions should be considered together.

Attempts to provide cultures consisting of satellite cells on the myofiber plasmalemma in the absence of the basal lamina were unsuccessful. The basal lamina could be removed with trypsin or other enzymes (Bischoff, 1975; Bischoff, 1986a) but satellite cells did not remain on the myofibers under these conditions.

A suspension of myofibers with attached satellite cells was centrifuged in culture dishes bearing coverslips with an adhesive surface (see Material and Methods). As centrifugal force was applied, the myofibers attached to the coverslip within a few seconds, then, during the following half hour of centrifugation, satellite cells detached from the myofibers and were deposited on the coverslip beneath the myofibers. This technique avoids further enzymatic treatment and leaves the satellite cells close to their associated myofiber. Preliminary experiments showed that most cells were detached from the myofibers after 10 min at 1000g, but a higher force/time was used routinely to ensure complete removal of all cells. The myofiber itself appeared unchanged after centrifugation (Fig. 2) and remained viable in culture. The satellite cells were often deposited to one side of the myofiber. The uneven distribution may have resulted from the release of cells before the dish holder reached a horizontal position in the centrifuge. In addition to satellite cells, fibrillar debris and a few small vesicles were deposited on the substratum following centrifugation. This material was not identified further, but scanning electron microscopy reveals a few collagen fibrils and debris from the dissociation procedure adherent to the surface of the basal lamina of isolated myofibers before centrifugation (Bischoff, 1986a).

The basal lamina was not removed from the myofiber by centrifugation as evidenced by the persistence of staining for laminin, a major component of the basal lamina, following immunocytochemistry. Most myofibers had several small laminin-negative patches on the surface (Fig. 3). These are not present in uncentrifuged myofibers (Bischoff, 1986a) and were interpreted as sites at which centrifugal force had removed satellite cells from the surface of the myofiber leaving tears in the overlying basal lamina.

Satellite cells deposited on the substratum remained rounded for several hours then spread out as plump, spindle-shaped cells. No satellite cells remained on the myofibers if the centrifugal force was 1500g, although at lower forces not all satellite cells were removed. The cells did not grow in the absence of mitogen but, when muscle extract was added, a group of satellite cell progeny surrounded the myofiber after 2 days in culture (Fig. 4). Although most satellite cells remained in the vicinity of the associated myofiber, some migrated up to 100 μm away. Only an occasional satellite cell migrated away from its associated myofiber in control cultures that were not centrifuged. When grown in 500 μgml ⁻¹ muscle extract for 2 days and labeled with [³H]TdR, 48 % (437/905) of satellite cells on the substratum were labeled while only 38% (410/1078) of myofiber-associated satellite cells were labeled in uncentrifuged companion cultures (Fig. 5). The number of labeled cells per myofiber was not counted because the dispersion of satellite cells on the substratum during the 2 day culture period made it difficult to reliably assign each satellite cell to a myofiber.

The satellite cell is in contact with both the plasmalemma and the basal lamina of the myofiber. To separate the effects of these two surfaces on satellite cell replication, the myofiber was killed with the myotoxic anaesthetic Marcaine (Hall-Craggs, 1974) leaving the satellite cells in contact with the basal lamina tube but removing them from the influence of the myofiber plasmalemma (Fig. 6). When provided with mitogen, the satellite cells proliferated and most remained within the basal lamina tube (Fig. 7). The live and Marcainekilled myofibers were compared for response to muscle extract (Fig. 8). Satellite cells associated with the killed myofibers were about 40 % more responsive to a given dose of muscle extract than satellite cells associated with live myofibers. Marcaine alone had no effect on satellite cell proliferation, confirming earlier studies (Schultz and Lipton, 1978). The protease inhibitors had no effect on satellite cell proliferation, but myonuclei and debris from the Marcaine-killed myofibers persisted for at least several days within the basal lamina tube. In the absence of protease inhibitors, nuclei within the killed myofiber became pyknotic and dissolved after a few hours.

To obtain additional information on the effect of the intact myofiber, 1 examined the dose-response of mitogen-stimulated satellite cell proliferation on killed and living myofibers. In these experiments, a mixture of killed and live myofibers was obtained in the same culture dish by killing a fraction (10 – 50%) of the myofibers just prior to plating. The myofibers are very sensitive to mechanical damage and variable numbers can be killed by vigorous trituration or by triturating with a pipet having a tip diameter less than the myofiber length (1mm). These cultures were grown in 0 – 1.8 mg ml⁻¹ muscle extract for 2 days and [^rH]TdR was added during the last 4 h. Labeled satellite cells were counted separately in killed and live myofibers. The killed myofibers were easily identified by hypercontraction of their myofibrils and the absence of myonuclei. There were substantially more labeled satellite cells associated with the killed myofibers over the entire range of muscle extract, even at concentrations in the plateau region of the dose-response curve (Fig. 9).

Inhibition of proliferation may induce myogenic cells to differentiate (Lathrop et al. 1985; Linkhart et al. 1982). Since the mature myofibers suppress proliferation of attached myogenic cells, I tested whether the cells would fuse with the myofibers they contacted. Previous studies have shown that progeny derived from satellite cells on single myofibers are fusion competent and form myotubes by fusion among themselves when the associated mature myofiber is killed (Bischoff, 1979; Bischoff, 1980).

This experiment was designed to challenge labeled satellite cells to fuse with their associated myofiber and then to detect the resulting labeled myonuclei by their disappearance after Marcaine treatment. Although satellite cells and myonuclei can be distinguished microscopically in whole mounts of myofibers (Bischoff, 1986a), the overlying radioautographic silver grains makes identification uncertain in some cases. Single myofibers were cultured in medium containing a low concentration of muscle extract (100μgml⁻¹) plus 0.1 μ Ciml⁻¹ [³H]TdR for 68h. Half the cultures were then treated with 0.05 % Marcaine to kill myofibers while the other half was left as control. Both groups were incubated with medium containing 0.32 mM unlabeled thymidine and 33 μ gml⁻¹ deoxyribonuclease and sacrificed after 4h, by which time the myonuclei were lysed in the killed myofibers. If any of the satellite cells had fused during the 3 day period, the resulting labeled myonuclei would be eliminated by Marcaine and the control myofibers would therefore contain more labeled nuclei than the Marcaine-treated myofibers. Counts made of total labeled nuclei per myofiber in radioautographs revealed no significant difference between the killed and living myofibers (killed: 16.7±0.8, live: 17.5±1). Therefore, although contact with the myofiber inhibits proliferation of associated satellite cells, it does not induce them to fuse with the mature myofiber during a 3 day period.

Since single myofibers undergo morphological and biochemical changes after several days in culture (Bischoff, 1980; Glavinovic et al. 1983; Bischoff, 1986a), I wished to determine whether the fusion competence of the myofibers for myogenic cells changes with time. The proliferating satellite cells in single myofiber cultures, however, begin to fuse among themselves before substantial changes in the myofibers occur (Bischoff, 1986a). Since this activity would complicate the assay, embryonic myoblasts were challenged to fuse with myofibers that had been stripped of satellite cells and cultured for several days. Single myofibers were treated with trypsin as described (Bischoff, 1986a) to remove the basal lamina and satellite cells. The fibers were attached to polylysine-treated coverslips by centrifugation and cultured with daily feeding of MEM with 10% horse serum and 5% chick embryo extract. Myogenic cells were prepared from 19 day rat embryos as described and labeled overnight with 0.1 μCiml⁻¹ [³H]TdR. Radioautographs showed that this procedure results in the labeling of 95±1.6% of the cells. The cells were dissociated with trypsin, counted with a hemocytometer and suspended in the same medium containing 0.32mM unlabeled thymidine. Myogenic cells (3 × 10⁵ per 35 mm dish) were added to the single myofiber cultures at 6 or 7 days in vitro. By this time, most of the myofibers had pseudopodial outgrowths from the myotendon ends (Fig. 10). Some myogenic cells were also grown in fresh polylysine-treated dishes as a control for fusion competence of the cells. The cultures were fixed at 24, 48 and 72 h after adding labeled cells and radioautographs were prepared. The labeled myogenic cells fused with each other and with the sprouting ends of the mature myofibers beginning at 24 h, but did not fuse with the original portion of the mature myofibers at any time period tested (Fig. 11).

Proliferation of satellite cells following muscle injury is initiated by the release of diffusible growth factors from the damaged muscle (Bischoff, 19866). In addition to this positive stimulation, the present results indicate that satellite cell growth is regulated by negative factors, including contact with the surface of a viable myofiber. Satellite cells on the culture substratum or on the myofiber basal lamina both proliferated to a greater extent than cells on intact myofibers. Although it is possible that the substratum stimulated proliferation, this seems unlikely since satellite cells on the substratum were not stimulated to proliferate without added mitogen. Furthermore, cells on the intact myofiber were suppressed relative to those on the basal lamina alone and the only difference in this case was the presence of the myofiber plasmalemma.

A growth strategy based upon both positive and negative elements affords a greater degree of fine tuning of muscle regeneration since satellite cell response depends upon the balance between positive and negative regulators. For example, death of a single myofiber, or a small group, would release its associated satellite cells from contact inhibition and allow the cells to proliferate in response to the small amount of growth factor liberated from the degenerating myofiber. At the same time, satellite cells on neighboring viable myofibers would be prevented from responding to the growth factor by contact inhibition and remain quiescent.

The present results show that with a low dose of muscle extract it is possible to stimulate a moderate level of satellite cell proliferation on killed myofibers but virtually none on live myofibers (Fig. 9). There is abundant evidence in the literature for local control of regeneration. Small foci of regenerating muscle, involving one or a few myofibers, are often observed following strenuous exercise (Giddings et al. 1985; Darr and Schultz, 1987; Irintchev and Wernig, 1987) and in muscular dystrophy (Engel, 1977). Contact-dependent communication between satellite cells and myofiber may also explain mobilization of satellite cells upstream or downstream from the actual site of injury (Mazanet et al. 1982; Klein-Ogus and Harris, 1983). Here, the response might include migration of satellite cells as well as proliferation, but the response would be much greater in the axial than the transverse dimension of the muscle. Unfortunately, the reaction of satellite cells distant from a wound has not been investigated quantitatively. Finally, contact inhibition may be important in muscle development as a mechanism for coordinating mitotic quiescence of satellite cells. As the myofiber reaches a certain level of maturity, changes in the plasmalemma might signal attached satellite cells to withdraw from the cell cycle. It would be of interest to measure the degree of contact inhibition as a function of myofiber age or maturity.

Contact inhibition is not absolute and can be overcome by increasing the concentration of muscle extract in vitro. Similarly, in certain situations in vivo it may be advantageous for satellite cells to proliferate on viable myofibers. There are examples of this during compensatory hypertrophy (Hanzlfkovà et al. 1975; Schiaffino et al. 1976), ischemia (Jennische, 1986) and toxinstimulated regeneration (Maltin et al. 1983). Some of the satellite cell progeny may fuse with the myofibers (Schiaffino et al. 1976). It is not known whether these cases represent increased release of growth factors or reduced contact inhibition by the myofiber, or both.

The muscle basal lamina survives various types of tissue injury and is believed to have a beneficial effect on regeneration (Le Gros Clark, 1946; Vracko and Benditt, 1972; Carlson and Faulkner, 1983), but its role has not been systematically investigated. Unfortunately, in the present experiments it was not possible to directly test the effect of the basal lamina on satellite cell proliferation. The basal lamina can be removed from single isolated myofibers with a low concentration of trypsin or Pronase (Bischoff, 1986a), but the satellite cells invariably become detached from the myofibers and thus their response to mitogen cannot be tested on fibers with and without a basal lamina covering. In the Marcaine experiments, the satellite cells were in contact with basal lamina in both the live and killed myofibers, yet were more responsive to mitogen in the absence of a live myofiber. Thus it seems that if basal lamina components, such as laminin, stimulate proliferation of myogenic cells, as has been shown in previous studies using monolayer cultures (Podleski et al. 1979; Foster et al. 1987; Öcalan et al. 1988), the effect can be suppressed by contact with the viable myofiber.

Inhibiting the proliferation of attached satellite cells by contact with a myofiber does not increase the likelihood that the cells will fuse with the myofiber. In fact, there was no instance of fusion of either satellite cells or embryonic myoblasts with the surface of a mature myofiber. The enzymatic treatment used to disaggregate tissue into single myofibers may have damaged surface receptors for fusion. In the case of embryonic myoblasts, however, the single fibers had been in culture for up to 7 days, allowing ample time for recovery. These results, together with earlier work, suggests that the plasma membrane becomes refractory to fusion as part of the terminal differentiation of myofibers. Bischoff and Holtzer (1969) found that fusion-competent myogenic cells failed to fuse with well-differentiated myotubes in monolayer cultures and speculated that the block to fusion might reside in the basal lamina or extracellular matrix. This possibility can be ruled out in the present study. The proliferating satellite cells on single myofiber are beneath the basal lamina, yet do not fuse. In the experiments with embryonic myoblasts, all extracellular coats were removed prior to challenging the myofibers with fusioncompetent cells. Trypsin-treated myofibers lack a basal lamina and the openings of the transverse tubules are clearly visible with the scanning electron microscope (Bischoff, 1986a).

A fusion block might function to regulate the nuclear density of myofibers during myogenesis. Most mononucleated myogenic cells adhere to the surface of myotubes during development in vivo (Tello, 1922) and in vitro (Bischoff and Lowe, 1974). Unfused myogenic cells on the surface of a myotube that became closed to further fusion would be encouraged to fuse among themselves and thus initiate the formation of a new generation of myotubes. The fusion block may also be involved in the formation of Go satellite cells by inducing myogenic cells to withdraw from the cell cycle, perhaps as the concentration of growth factors becomes limiting during the late stages of development. According to this scheme, quiescent satellite cells could exist on mature myofibers at the same time as dividing satellite cells on young myofibers nearby. Satellite cell proliferation would.be regulated both by concentration of growth factors and by the surface properties of the myofiber. Further studies are needed to investigate the fusion properties of myofibers. Lipton and Schultz (1979) reported that myogenic cells injected into adult muscle can fuse with mature myofibers. Fusion of the injected cells among themselves was noted by 4 days and many of the injected cells had migrated to a position beneath the basal lamina of the myofibers by this time, but the cells did not fuse with mature myofibers until 7 days after transplantation. There is no explanation for the delay in fusion. Since there was myofiber degeneration and regeneration at the injection site, it is possible that the implanted cells first fused here with the ends of injured myofibers and the myonuclei required several days to migrate into undamaged regions of the muscle.

Although myogenic cells did not fuse with mature myofibers in the present experiments, they readily fused with nucleated sprouts growing from the ends of the mature myofibers. The sprouts consist of pseudopodia! extensions that begin to appear after 4 or 5 days in culture. Eventually the entire myofiber is transformed into an elongate, sometimes branched, myotube-like structure (Bischoff, 1980; Bischoff, 1986a). Although myonuclei migrate from the original myofiber into the sprouts, it seems unlikely that labeled cells first fused with the original myofiber then their nuclei migrated centrifugally, for labeled nuclei were never found in the original myofiber.

Sprouting has also been observed in injured muscle in vivo, where it is believed to contribute to a process termed continuous regeneration. Immature myotubes with clustered nuclei and sparse myofibrils are often found near the wound edge several days after injury, in apparent continuity with surviving mature myofibers (Le Gros Clark, 1946; Hall-Craggs, 1974). The nature of these myotubes is difficult to determine in light microscope studies, but electron microscopic examination clearly shows cytoplasmic continuity with mature myofibers (Shafiq and Gorycki, 1965). The sprouts may arise from myofibers that were transected by injury and subsequently healed over (Echeverria et al. 1987). This process offers several advantages to muscle regeneration. By elongating across the wound area, the sprouts bridge the gap to join myofibers from the opposite edge (Gay and Hunt, 1954; Hudgson and Field, 1973; Webb, 1973) The growth rate of the sprouts is rapid enough to establish electrical continuity before satellite-cellderived myotubes have had much opportunity to form (Stuart et al. 1981). Sprouting involves dedifferentiation of the myofibers, a process that includes the reappearance of embryonic isozymes of creatine kinase (Bischoff, 1986a). Another aspect of dedifferentiation, as demonstrated in the present study, is the reacquisition of a surface membrane compatible with fusion. This would allow transected myofibers to restore their original nuclear complement by fusion with satellite-cell-derived myoblasts, or alternatively to fuse with satel-lite-cell-derived myotubes or with sprouting myofibers from the opposite wound edge.

I thank Kym Hallsworth for competent assistance with these experiments.

This research was supported by a grant from the National Institutes of Health (AM36853).