Cytological Differentiation of Human Fetal Skeletal Muscle ROBERT J. TOMANEK AND ANN-SOFI COLLING-SALTIN Lkpartment ofrlnatomy, College of Medicine, University of Iowq Iowa City, Iowa 52242 and Department of Obstetrics and Gynecologv, Karolinska Sjukhuset, Stockholm, Sweden
ABSTRACT The ultrastructural differentiation of several different muscles was investigated in human fetuses ranging in age from 13 weeks to neonatal. A t approximately 16 weeks of gestation cell clusters containing both myotubes and satellite cells lie enclosed by a newly formed basal lamina and show evidence of fusion. The development of organelles is evident in myoblasts, proceeds as the cells transform into myofibers, and continues in the neonate. Filament synthesis occurs primarily in the cell periphery where thin filaments appear to align themselves in relations to parallel arrays of ribosome-studded thick filaments; Z line formation follows the appearance of thin filaments. Intermediate filaments, approximately 10-12 nm thick, were also consistently observed in perinuclear regions and distal to filament assembly. Although sarcoplasmic reticulum (SR) development is closely related to fibril formation, connections between Z lines and SR are not consistent, thus supporting the conclusion that SR does not evoke the formation of the Z line. Bristlecoated vesicles appear to be the precursors of elements of the SR, possibly the lateral sacs. Development of the transverse tubules, as invaginations of the sarcolemma, is closely associated with the formation of lateral sacs since the latter occur along the sarcolemma as soon as transverse tubules appear. Cytological differentiation is similar, though not identical, in several different muscles. During the last trimester muscle fibers show some evidence of diversity, mainly of variation in Z line width. In general the results suggest that the sequence and stages of human myogenesis are similar to those of other species. Nearly 60 years ago the histogenesis of striated muscle of man and other vertebrates was comprehensively described (Tello, '17). Since that time numerous studies have contributed to a basic understanding of myogenesis in vertebrate species (Fishman, '73). Studies dealing with the ultrastructure of human fetal muscle, however, have been less comprehensive; several investigations have focused on the role of the satellite cell (Ishikawa, '66, '70; Conen and Bell, '70).Other studies employing electron microscopy have provided a general description of myogenesis occurring during the first five and one half months in utero (Hudgson et al., '70; Ochoa and Mair, '68), established similarities between chick and human myogenesis (Hudgson et al., '701, elucidated the morphological relationship between polyribosomes and myosin filaments (Larson et al., '691, and suggested a relationship between the sarcoplasmic reticulum, fibril and triad (Walker et al., '75). AM. J. ANAT., 149: 227-246.
Since most of the work on human fetal skeletal muscle has focused on rather specific aspects of differentiation and often considered only specific periods of fetal development, a more comprehensive approach appeared to be necessary for the elucidation of the sequence of cytological differentiation. The observations reported here are based on several muscles of fetuses ranging in age from 13 weeks to neonatal. Our observations have taken into account the relationship of various organelles and cells and the consistency of sequential events. MATERIALS AND METHODS
Muscle tissues from human fetuses and infants were obtained a t necropsy from the vastus lateralis (19 specimens), soleus (10 specimens) and rectus abdominis (8 specimens). Strips of muscle, approximately 3 mm Accepted January 19, '11.
227
228
ROBERT J. TOMANEK AND ANN-SOFI COLLING-SALTIN
in thickness, were tied to wooden splints and fixed in a solution containing: 2.5% glutaraldehyde, 0.1% sodium cacodylate, 0.8%paraformaldehyde and 0.02 M CaClz (ph 7.39). Following an overnight wash in buffer containing 3% sucrose, the tissues were post-fixed in veronal acetate-buffered 1% OsO,, dehydrated in alcohol and embedded in Epon. A Siemens 101 Elmiscope was used to examine thin sections which were stained with uranyl acetate and lead citrate. Sections one micron thick were stained with Richardson’s stain (Richardson et al., ’60) and used for light microscopy. The problem of post-mortem artifact was evident in several specimens and attributed to a lengthy time lapse between death and fixation (the availability of the tissue varied from a few minutes to two hours post-mortem). For this reason it was necessary to limit the use of such inadequately fixed specimens to cursory observations. Z line width was estimated in photomicrographs ( X 30,000) with the aid of a n ocular micrometer (Tomanek, ’76). RESULTS
Nomenclature of myogenesis
The nomenclature regarding the various cell stages comprising myogenesis is defined as follows. “Presumptive myoblasts” are similar in morphology to other mesenchymal cells; they divide mitotically, but do not fuse. Although recent evidence indicates t h a t presumptive myoblasts transcribe myosin mRNAs (Chi e t al., ’751, these cells lack myofibrils. “Myoblasts,” on the other hand, synthesize and organize thin and thick filaments (Holtzer e t al., ’75). The contractile elements are more notable when myoblasts fuse to form a multinucleated syncytium termed a “myotube.” Circumferentially distributed myofibrils or filaments, centrally placed nuclei, and a central core of sarcoplasm are characteristic of this stage. Obviously there is no sharp point of demarcation between this stage of myogenesis and a mature cell or “myofiber” (Fishman, ’73). The latter arises as nuclei are peripherally displaced and myofibrils and other organelles fill the core of the myofiber. With the advent of the basal lamina it is possible to identify a cell which resembles a presumptive myoblast but lies closely apposed to a myofiber or myotube and within the confines of its basal lamina. Such a spindleshaped, mononucleated cell constitutes a “satellite cell” (Ishikawa, ’70).
Stages of. myogenesis ~-
Most of the muscle cells encountered in specimens from 13 to 16-week-old fetuses were myotubes or closely apposed undifferentiated cells. The latter resembled presumptive myoblasts or satellite cells. In 13- to 14-weekold fetuses these cells lacked a well defined basal lamina, which was presumably undergoing development since a faint, fuzzy structure could often be visualized. A closely associated undifferentiated cell appeared to lie enclosed within the same presumptive basal lamina which surrounded one to several cells. This relationship was more evident a t the 16week stage and thereafter. Myotubes representing various levels of maturity were common. Some contained myofibrils which were composed of only a few sarcomeres, while in others myofibrils were well established with numerous sarcomeres and were longitudinally oriented. This stage was typified by clusters of three to ten cells each separated by a narrow (15-50 nm) interspace. Between 16 and 20 weeks the nuclei in some myotubes became peripherally displaced giving rise to early myofibers. Many cells still occurred in groups and lacked an intervening basal lamina. Numerous satellite cells were encountered during this period as well as during later stages of fetal development. Commonly one satellite cell was associated with two myofibers or myotubes, i.e., the three cells were encased by a common basal lamina. In addition, we observed cells which appeared to be connected in part to adjacent cells as indicated either by “fusion” of their cell membranes, by sarcoplasmic bridges, or, as shown in figures 3 and 4, filaments or a myofibril were seen extending between two cells. This phenomenon was limited to cells which were in the myotube stage and involved two myotubes, or a myotube and a myoblast or satellite cell. Myotubes were found throughout fetal development although their frequency of occurrence decreased with time (figs. 5-71. Myofibers predominated after approximately 20 to 22 weeks of gestation and were characterized by relatively wide separation of fibrils which often contained only a few sarcomeres and lacked a longitudinal orientation. A predominance of uniform fibrils with many Z lines in registry with adjacent fibrils was not evident until late fetal development, i.e., approximately 30 weeks of gestation. This later pe-
DIFFERENTIATION OF HUMAN FETAL SKELETAL MUSCLE
riod of fetal life was typified by some increase as well as variation in fiber diameter. Prior to this time fiber diameter did not increase markedly, since the increasing number of contractile filaments replaced the sarcoplasm characteristic of myotubes. The most marked increase in fiber diameter occurred a t the time of birth (fig. 8) and thereafter. The presence of glycogen-containing vacuoles was evident in fibers during late fetal (3036 weeks) and early post-natal (up to 1 month) life. They usually consisted of a double membrane (fig. 24) or membranous fragments (fig. 25). In muscles from post-natal animals the glycogen-containing vacuoles lay in cell regions which were relatively free of glycogen particles.
Contractile filaments In the early myotube, the contractile filaments were first observed in the cell periphery where disorganized or poorly aligned thin filaments were positioned close to the cell membrane, with clusters of thick filaments in close proximity along with their associated ribosomes (fig. 9). As illustrated in figure 10, lattice formation involving thin and thick filaments as well as Z lines was evident as soon as both types of filaments appeared. This hexagonal array was evident in cross sections even when only a few thick filaments were seen. Unoriented thin filaments were peripheral to the lattice structure, suggesting their assembly as new myosin was synthesized. Filament assembly and lattice formation may not be dependent upon the presence of Z bodies. Clusters of thin and thick filaments without an associated Z body were most frequently encountered. A failure to observe such structures could be due to the plane of section. In such situations, however, the Z body would lack connections with the filaments; yet, whenever we encountered Z bodies, fine filaments were seen extending from them. The earliest appearance of dense material suggestive of Z bodies was consistently observed in foci of thin filaments (figs. 13,14). Some possible stages of sarcomere formation may be visualized by moving inward from the cell membrane (fig. 14). This illustration suggests: (1) rather independent sites of thin and thick filament synthesis-the myosin (thick) filaments lying further from the cell membrane, (2) a decrease in the number of myosin-associated ribosomes as lattice formation proceeds, and (3) Z body formation
229
in foci of thin filaments in the absence or presence of thick filaments. Subsequent stages of myogenesis involve patterns similar to those observed in early myotubes. Perinuclear regions were typified by disoriented filaments intermediate in diameter (approximately 10-12 nm) t o thin and thick contractile elements (fig. 12). The intermediate filaments were also found distal to regions of sarcomere formation (fig. 15). They may be distinguished from thin filaments by comparing figures 13-15. Microtubules, approximately 25 nm in diameter, were characteristic of regions of filament synthesis and early assembly (figs. 9,12, 15), but were absent in regions of more advanced myofibril formation. Usually their orientation was longitudinal or oblique and they were situated near the ends of developing sarcomeres. In addition to intermediate filaments and microtubules, numerous vesicles and ribosomes were typical of regions distal to forming sarcomeres (fig. 15). The ribosomes included three varieties: free, membrane-associated and polyribosomes.
Tubular system Early fetal development (13-22 weeks) was characterized by various tubular or vesicular structures. Pleomorphic vesicular structures were most common in the early myotubes, while scattered tubular elements resembling sarcoplasmic reticulum (SR)occurred more frequently in late myotubes and early myofibers. Some of the latter were recognized as T-tubules since they could be seen as invaginations of the sarcolemma (figs. 16, 17). It is inferred that the development of the sarcoplasmic reticulum is related to that of the contractile filaments, since the appearance of sarcomeres was accompanied by scattered SR elements (figs. 19-22). The tubules, which varied in width (20-60 nm) were found in the peripheral sarcoplasm along with newly aligning filaments, and in perinuclear regions. They were notably absent from the core of the myotube, which contained numerous glycogen granules. Sarcolemmal and subsarcolemmal vesicles were also common, particularly in myotubes (fig. 23, 26). A coordinated development of the SR and filament components of muscle is suggested by the fact that cells which were in the early myotube stage had a limited amount of SR. This was also true of specimens from older
230
ROBERT J. TOMANEK AND ANN-SOFI COLLING-SALTIN
fetuses (22-30 weeks of age). While components of the sarcoplasmic reticulum were frequently closely apposed to Z lines (fig. 221, this relationship was not evident in the earliest stages of development, i.e., when Z bodies were first seen. On the other hand, SR fragments could be found associated with a few filaments prior to Z line formation (fig. 20). Further development of the SR was evidenced by branching a t the Z line as noted in crosssectional fields. We did not observe extensive branching and anastomoses, typical of adult fibers, even in the neonate. However, some variability in the amount of SR in different fibers was indicated during late fetal life. Triads or diads were evident a t relatively early stages of development. Isolated triads, i.e., without apparent connections with other elements of sarcoplasmic reticulum, were evident near or a t the sarcolemma in the 19week fetus (fig. 16). Examples of sarcolemmal invaginations contributing to a triadic component were found in relatively early myotubes (fig. 17). The absence of triads in deeper regions of the cell a t this stage suggests that the terminal cisternae as well as the T-tubule may originate a t or near the sarcolemma. Between 22 to 25 weeks of fetal life numerous triads, most of which were longitudinally oriented, were found dispersed throughout the myofibers; a few were aligned a t the A-1 junction, while most lay disoriented. Even a t this later developmental stage connections with the longitudinal elements of the sarcoplasmic reticulum were not as common as in adult mammalian muscle. Since terminal cisternae were consistently found in association with Ttubules, the development of the two structures appeared to be closely associated. The longitudinal components of the SR, on the other hand, may be developed independently. Spherical vesicles characterized by the presence of bristles on their surfaces (figs. 2, 14, 18, 19) were frequently encountered in myofibers, satellite cells and myotubes. As suggested by the profile in figure 19, the spherical vesicles appeared to be the terminal components of a tubular system. Such structures were occasionally observed in proximity to sarcomeres (fig. 21), but were more common in the perinuclear and subsarcolemmal regions of fibers or randomly distributed in undifferentiated (satellite) cells.
Cell variation in differentmuscles Most of the observations described apply to
specimens from different muscles. Since variations might suggest differentiation into specific fiber types, Z line width was measured. Between 13 and 19 weeks, Z line widths were similar and not highly variable in both the soleus and vastus lateralis, ranging between 74-118 nm. Later in fetal life (25-31 weeks) the soleus contained fibers which were more homogeneous; Z line width in most fibers fell within the 112-160 nm range. The fibers of the vastus lateralis were more variable in this regard with values ranging from 71-140 nm; in the rectus abdominis the Z line widths were even more varied (75-170 nm). Because all of the muscles studied eventually become diversified, we lacked a model with a homogeneous fiber population. However, the soleus, with predominantly one fiber type, can be contrasted with the other muscles and, therefore most closely represents such a model. In comparing the soleus to other muscles (primarily the vastus lateralis) during early fetal development few differences were apparent. However, in late fetal life most fibers comprising the soleus and many found in the rectus abdominis were characterized by wide Z lines and poorly developed SR. Thus, an apparent differentiation into specific fibers is suggested. Compared to the vastus lateralis, the soleus matured more rapidly. Not uncommonly, cells with relatively densely packed myofilaments and peripherally situated nuclei were evident in specimens from 13-week-old fetuses. During this early period of fetal development a considerable variation in cell diameter was evident in the soleus and could be contrasted with a rather uniform cell size in the vastus lateralis and rectus abdominis. This variability was not evident in the latter muscles until 18 weeks in utero. DISCUSSION
In vitro, time-lapse studies on chick embryos have demonstrated that myogenic cells fuse to form muscle syncytia or myotubes (Capers, '60). Subsequent electron microscopic investigations have supported this finding (Shimada et al., '67; Fishman, '70). Multinucleation during myogenesis includes cytoplasmic fusion of myoblasts and myotubes in the absence of a nerve supply (Konigsberg et al., '60; Stockale and Holtzer, '61; Shimada et al., '67). Our own observations in human specimens are consistent with these reports and suggest that undifferentiated cells lie closely
DIFFERENTIATION OF HUMAN FETAL SKELETAL MUSCLE
apposed to myotubes, are incorporated within a common basal lamina, and may fuse with the latter myotubes. Evidence for this contention is based on observations illustrating: (1) fusion of cell membranes, (2) formation of sarcoplasmic bridges which include filaments or fibrils, and (3) similar cell associations between satellite cells and muscle cells, or between myotubes or early muscle fibers. The fact that the number of satellite cells decreases during late fetal life (Ishikawa, '66) supports evidence that fusion occurs immediately or subsequent to the production of filaments. This hypothesis is based on our findings that small cells which lie under the basal lamina of a muscle cell may have the characteristics of a satellite cell, an early differentiating cell (similar to a satellite cell but containing a few filaments), or a myotube. Since no mode of nuclear replication is known to occur in multinucleated muscle cells (Stockdale and Holtzer, '611, the sarcoplasmic bridges are not likely to be due to cytokinesis. The increase in the total number of nuclei in a given muscle during growth is accounted for by multiplication of satellite cells (Enesco and Puddy, '64; Allbrook, '71; Moss and Leblond, '71). Glycogen-containing vacuoles, which appear during late fetal life, have been described in r a t skeletal muscle (Schiaffino and Hanzlikova, '72). They are considered to be autophagic vacuoles, i.e., components of the lysosomal system which mobilize glycogen a t the time of birth. As in liver, glycogen accumulates in skeletal muscle during the late fetal stages and is rapidly mobilized a t birth (Dawes and Shelley, '68). Our observations in the human appear to be consistent with studies on the rat (Schiaffino and Hanzlikova, '72), with the exception that they were encountered earlier in human muscle.
231
synthesis is not known. In any case, both thick and thin filaments occur in early muscle cells as evidenced in all of our human specimens; thus the process is comparable to chick myogenesis (Fischman, '67). A third type of filament evident in human fetal muscle is of the same dimensions as the intermediate filament demonstrated by Ishikawa et al. ('68). Such filaments appear to differ from actin in that they do not form the characteristic arrowhead complexes with heavy meromyosin (Ishikawa et al., '69). The possibility t h a t earlier reports described intermediate filaments as thin filaments has been suggested (Ishikawa et al., '68). Our observations, however, clearly indicate the presence of both types of filaments in myotubes, with most of the cortical filaments being classified as thin. Polyribosomes are structurally related to the formation of thick filaments in human muscle where they can be found consistently near the end of the filament nearest the Iband (Larson et al., '69). Our observations support this finding and underline the consistency of the association throughout the cell, but a structural association between thin filaments and ribosomes is not indicated since regions containing only thin filaments may or may not contain scattered ribosomes. With the appearance of thick filaments a hexagonal array incorporating the thin filaments is detected, as has been observed in chick muscle (Allen and Pepe, '65; Fischman, '67). The absence of Z line material or Z bodies in regions of early filament associations is consistent with evidence that such structures are not necessary for the formation of the hexagonal filament pattern (Stromer et al., '69). While Z lines are not requisite for this aspect of myogenesis their formation is closely linked with that of other sarcomere components and is essential for longitudinal growth Filaments (Fischman and Zak, '71). Since the appearIn early myotubes, i.e., cells with limited ance of Z bodies is invariably linked with the numbers of cytofilaments, thin filaments (5-6 presence of thin filaments, the synthesis of nm in diameter) predominate. Although corti- the former may be dependent upon the prescal filaments bind heavy meromyosin (Ishi- ence of the latter. kawa et al., '69), there is not yet ample eviTubular system dence demonstrating that this binding is Sarcoplasmic reticulum (SR) is derived limited to actin. Some authors have suggested that thin filaments are synthesized prior to from the endoplasmic reticulum of undifferenthick filaments in chick (Allen and Pepe, '65; tiated cells (Ezerman and Ishikawa, '67). Hudgson et al., '70; Ogawa, '62) as well as Vesicular structures observed in the early human (Hudgson et al., '70) muscle. Whether myotube are probably an early stage of SR. the presence of actin is essential for myosin Irregular tubules show some branching as the
232
ROBERT J. TOMANEK AND ANN-SOFl COLLING-SALTIN
cell matures and, as previously described in the human diaphragm, (Walker et al., '75) are associated with regions of the cell containing filaments and sarcomeres. This is in contrast to an earlier report which indicated little evidence for SR in human myotubes which are maturing into fibers (Ochoa and Mair, '68). A relationship between SR and fibril development in the rat (Walker and Edge, '711, monkey (Walker et al., '75) and man (Walker et al., '75) has been suggested on the basis that SR tubules appear to be connected to and encircle Z lines during myogenesis. Since the SR network has been observed during early fibril development, it has been hypothesized that the SR forms a framework for filament and Z line assembly (Walker et al., '75). Conversely, it has been suggested that the SR develops around assembling filaments (Fischman, '73). Our own observations indicate numerous Z lines in continuity with the SR, but this feature is by no means consistent. If these tubules completely encircle the Z line, as previously reported (Walker et al., '751, tubular segments should be included in any longitudinal section through a Z line. Our observations indicate this is not the rule. This, of course, does not rule out a structural connection between the two components. Since Z bodies which include attached thin filaments are infrequently associated with tubules it appears doubtful that the Z line and SR play a significant role in the assembly of filaments. A more plausible contention may be that the early development and organization of SR coincides with early sarcomere formation. A subsequent positioning of Z lines in register with adjacent fibrils (Walker et al., '75) may, however, be a function of Z line-associated tubules. The T system in cultured chick muscle has been shown to develop by invaginations of the sarcolemma of the myotube (Ezerman and Ishikawa, '67; Ishikawa, '68). Although a previous study on human fetal muscle could not find direct evidence for this pattern of development (Hudgson et al., '701, our observations suggest that this is the case. First, invaginations were observed and seen to form diads or triads in the cell periphery. Second, diads and triads in early myotubes were seen only in the cell periphery and most frequently in association with the sarcolemma. Finally, diads and triads occur in deeper areas of the cell only after myofibrils become centrally displaced.
SR cisternae are presumed to expand and form their characteristic structure after contacts between SR and T-tubules have formed (Edge, '70). The observation that during early fetal development cisternae seldom show connections to other SR components suggests either that they develop independently or that they form on short fragments of SR which eventually unite with other SR tubules. Spherical vesicles characterized by bristles on their outer surface were previously seen in primate muscle as evaginations of T-tubules (Walker et al., '75). That such vesicles are components of a tubular system is also evident from our observations. Their location in the cell periphery (particularly in the early myotube) fits with the suggestion that they are associated with the T-tubules, although their density and dilated form is more akin to SR terminal cisternae than to a T-tubule. Their presence in satellite cells may lend support to the suggestion that they are components of the SR, since the T-system is not evident until the myotube stage. Characteristically the SR terminal cisternae contain a dense material which may consist of mucopolysaccharides or glycoproteins and accordingly provide sites for calcium binding (Philpott and Goldstein, '67). Diversification of fibers From our observations, it is evident that cytological differentiation in the human fetus is similar, though not identical, in several different muscles. The development of various organelles may thus constitute a primary phase, common to all muscle fibers (Eccles, '631, which is under genetic regulation. Fiber diversification, on the other hand, is controlled by neural factors (Hanzlikova and Schiaffino, '73). Such diversification is not clearly evident during the first six months of fetal life, after which time the soleus becomes more homogeneous while the fibers in the vastus lateralis become more diversified with respect to size, Z line width and extensiveness of sarcoplasmic reticulum. A more complete diversification undoubtedly occurs after birth as evidenced by studies on the cat (Tomanek, '76). In this respect as well as the sequence of organelle development and stages of myogenesis, human fetal muscle development and differentiation are similar to those of other species.
DIFFERENTIATION OF HUMAN FETAL SKELETAL MUSCLE LITERATURE CITED Allen, E. R., and F. A. P e p 1965 Ultrastructure of developing muscle cells in the chick embryo. Am. J. Anat., 116: 115-147. Bergman, R. A. 1962 Observations on the morphogenesis of r a t skeletal tissue. Bull. Johns Hopkins Hosp., 110: 187-201. Capers, C. R. 1960 Multinucleation of skeletal muscle in uitro. J. Biophys. Biochem. Cytol., 7: 559-566. Chi, J. C., S. A. Fellini and H. Holtzer 1975 Differences among myosins synthesized in non-myogenic cells, presumptive myoblasts and myoblasts. Proc. Nat. Acad. Sci., (U. S. AJ, 72: 4999-5003. Conen, P. E., and C. D. Bell 1970 Study of satellite cells in mature and fetal muscle and rhabdomyosarcoma. In: Regeneration of Striated Muscle and Myogenesis. H. Mauro, S . A. Shafiq and A. T. Milhorat, eds. Excerpta medica, Amsterdam, pp. 194-211. Cullen, M. J., and D. Weightman 1975 The ultrastructure of normal human muscle in relation to fibre type. J. Neurol Sci., 25:(1): 43-56. Dawes, G. S.,and J. H. Shelley 1968 Physiological aspects of carbohydrate metabolism in the foetus and newborn. In: Carbohydrate Metabolism and Its Disorders. F. Dickens, P. J. Randle and W. T. Whelan, eds. Academic Press Inc., New York, 2: 87-121. Edge, M. B. 1970 Development of apposed sarcoplasmic reticulum a t th e T system and sarcolemma and t he change in orientation of triads in r a t skeletal muscle. Develop. Biol., 23: 634-650. Elliott, B. J., and D. G. F. Harriman 1975 Growth of human muscle spindles in uitro. Nature, 251: 622-624. Enesco, M., and D. Puddy 1964 Increase in the number of nuclei and weight in skeletal muscles of rats of various ages. Am. J. Anat., 114: 235-244. Ezerman, E. B., and H. Ishikawa 1967 Differentiation of the sarcoplasmic reticulum and T system in developing chick skeletal muscle in vitro. J. Cell Biol., 35: 405-420. Fischman. D. A. 1967 An electron microscoDic studv of myofibril formation in embryonic chick skeletal muscle. J. Cell. Biol., 32: 557-575. 1972 Development of striated muscle. In: The Structure and Function of Muscle. Vol. 1. Second ed. G. H. Bourne, ed. Academic Press, New York. Fischman, D. A,, and R. Zak 1971 Assembly of myofibrils in the absence of protein synthesis. J. Gen. Physiol., 57: 245. Holtzer, H. 1970 Myogenesis In: Cell Differentiation. 0. Schjeide and J. deVellis, eds. Van Nostrand-Reinhold, Princeton, pp. 476-503. Holtzer, H., K. Strahs and J. Biehl 1975 Thick and thin filaments in postmitotic, mononucleated myoblasts. Science, 188: 943-945. Hudgson, P., M. Jenkison and P. F. Larson 1970 An ultrastructural study of chick embryo and human foetal muscle. In: Proceedings International Congress on Muscle Diseases, Milan 1969.J. N. Walton, N. Canal andG. Scarlato, eds. Excerpta medica, Amsterdam, pp. 90-97. Ishikawa, H. 1966 Electron microscopic observations of satellite cells with special reference to the development of mammalian skeletal muscles. Z. Anat. Entw. Gesch., 125: 43-63. 1968 Formation of elaborate networks of T system tubules in cultured skeletal muscle with special reference to the T. system formation. J. Cell Biol., 38: 51. 1970 Satellite cells in developing muscle and tissue culture. In: Regeneration of Striated Muscle and
233
Myogenesis. A. Mauro, S. A. Shafiq and A. T. Milhorat, eds. Excerpta medica, Amsterdam, pp. 167-179. Ishikawa, H., R. Bischoff and H. Holtzer 1968 Mitosis and intermediate-sized filaments in developing skeletal muscle. J. Cell Biol., 38: 538-555. 1969 Formation of arrowhead complexes with heavy meromyosin in a variety of cell types. J. Cell Biol., 43: 312-328. Konigsberg, I. R. 1965 Aspects of cytodifferentiation of skeletal muscle. In: Organogenesis. R. L. de Haan and H. Ursprung, eds. Holt, Rinehart and Winston, Inc., New York. Konigsberg, I. R. N., N. Mcelvain, M. Tootle and H. Herrmann 1960 The dissociability of ribonucleic acid synthesis from the development of multinuclearity of muscle cells in culture. J. Biophys. Biochem. Cytol., 8: 333-343. Larson, P. F., P. Hudgson and J. N. Walton 1969 Morphological relationship of polyribosomes and myosin filaments in developing and regenerating skeletal muscle. Nature, 222: 1168-1169. Lipton, B. H., and I. R. Koningsberg 1972 A fine-structural analysis of the fusion of myogenic cells. J. Cell Biol., 5 3 W : 348-364. Mauro, A. 1961 Satellite cell of skeletal muscle fibers. J. Biophys. Biochem. Cytol., 9: 493-495. Moss, F. P., and C. P. Leblond 1970 Nature of dividing nuclei in skeletal muscle of growing rats. J. Cell Biol., 44: 459-462. Ochoa, J., and W. G. P. Mair 1968 The ultrastructure of normal foetal muscle and foetal muscle from known dystrophic-carriers. In: Proceedings of 4th Symposium on Current Research in Muscular Dystrophy. Pitman Medical, London, pp. 223-239. Ogawa, Y. 1962 Synthesis of skeletal muscle proteins in early embryo and regenerating tissue of chick and triturus. Exp. Cell Res., 26: 269-274. Philpott, C. W., and M. A. Goldstein 1967 Sarcoplasmic reticulum of striated muscle: localization of potential calcium binding sites. Science, 155: 1019-1020. Richardson, K. C., L. J a re tt and E. H. Finke 1960 Embedding in epoxy resins for ultrathin sectioning in electron microscopy. Stain Tech., 35: 313-323. Schiaffino, S., and V. Hanzlikova 1972 Autophagic degradation of glycogen in skeletal muscles of the newborn rat. J. Cell Biol., 52: 41-51. Shimada, Y., D. A. Fischman and A. A. Moscona 1967 The fine structure of embryonic chick skeletal muscle cells differentiated in vitro. J. Cell Biol., 35: 445-453. Stockdale, F. E., and H. Holtzer 1961 DNA synthesis and myogenesis. Exp. Cell Res., 24: 508-520. Stromer, M. H., D. J. Hartshorne, H. Mueller and R. V. Rice 1969 The effect of various protein fractions on Z- and Mline reconstitution. J. Cell Biol., 40: 167-178. Tello, J. F. 1917 Genesis de las terminaciones nerviosas motrices y sensitivas. I: En el sisterma lacomotor de las vertebrados superiores. Histogenesis muscular. Trabajos Lab. Invest. Biol. Univ. Madrid, 15: 101.199. Tomanek, R. J. 1976 Ultrastructural Differentiation of skeletal muscle fibers and their diversity. J. Ultrastr. Res., 55: 212-227. Walker, S.M., and M. B. Edge 1971 The sarcoplasmic reticulum and development of Z lines in skeletal muscle fibers of fetal and postnatal rats. Anat. Rec., 169: 661-678. Walker, S. M., G. R. Schrodt and G. J. Currier 1975 Evidence for a structural relationship between successive parallel tubules in the SR network and supernumerary striations of 2 line material in purkinje fibers of the chicken, sheep, dog and Rhesus monkey heart. J. Morph., 147: 459-474.
PLATE 1 EXPLANATION OF FIGURES
1 A myotube and its satellite cell (SC) (rectus abdominis, 28-week-old fetus). The latter contains free ribosomes and polyribosomes (arrowheads). A basal lamina (bl) encloses the two cells but is absent between their apposing plasma membranes (arrows). x 20,000.
2 Structural association of two myotubes (vastus lateralis, 18-week-oldfetus). The two cells are enclosed by a common basal lamina (bl) which is (as in fig. 1). absent between their apposing plasma membranes. One cell (A) has a peripheral rim of contractile elements not yet organized into defined myofibrils and also contains numerous polyribosomes (arrowheads). A more “mature” myotube (B) contains some defined myofibrils. Note the peripherally situated bristle-coated vesicle (arrow). X 20,000.
3 Apparent fusion of two cells (vastus lateralis, 22-week-old fetus). One cell (A) is either an early myotube or a myoblast containing only a few contractile filaments. Filaments from a more advanced myotube (B) a re seen extending into the neighboring cell. ES, extracellular space; sa, sarcolemma. X 50,000.
234
DIFFERENTIATION OF HUMAN FETAL SKELETAL MUSCLE Robert J . Tomanek and Ann-Sofi Colling-Saltin
PLATE 1
PLATE 2 EXPLANATION OF FIGURES 4
Apparent fusion of two myotubes (more advanced fusion from the soleus of a 16 to 17week-old fetus. A myofibril (my) extends between the two cells, which are joined by a narrow channel of sarcoplasm. Polyribosomes (arrowheads) and unassembled filaments (fi) a r e present in both cells. ES, extracellular space. X 19,500.
Figs. 5-8 Cross-sectional fields of 1-pm Epon sections stained with Richardson’s solution. 5 Muscle cells from a 13-week-old fetus (vastus lateralis). Virtually all cells are in the myotube stage as evidenced by centrally situated nuclei and, in non-nuclear regions, a central core of cytoplasm (arrow-heads). X 830. 6 Muscle cells from a 13-week-old fetus (soleus). Compared to the vastus lateralis, cell diameter varies considerably and some cells are early myofibers as evidenced by peripherally situated nuclei (arrows). X 830. 7 Muscle cells from an 18-week-old fetus (vastus lateralis). Myofibrils in most cells are
relatively densely packed. In some cells nuclei are situated peripherally while in others they a r e centrally positioned. Unlike the earlier stage (fig. 5) cells vary considerably in diameter. X 830. 8
236
Muscle cells from a one-month-old infant (rectus abdominis). Cell diameter is considerably greater than in any fetal stage and myotubes are virtually absent. Some variation in cell diameter is evident. X 830.
DIFFERENTIATION OF HUMAN FETAL SKELETAL MUSCLE Robert J. Tomanek and Ann-Sofi Colling-Saltin
PLATE 2
237
PLATE 3 EXPLANATION OF FIGURES
9 Early myotube (vastus lateralis, 18-week-old fetus). The region above the arrowheads lies in proximity to the plasma membrane and contains thin filaments and extremely limited numbers of ribosomes. Below and to the left of this region thick filaments lie in parallel and are associated with ribosomes (rb). Microtubules (mt) are common near the ends of the assembled thick filaments. X 30,000. 10 Early myotube (soleus, 18-week-oldfetus). The early formation of sarcomeres is illustrated; both thick and thin filaments lie in close proximity and appear to undergo assembly along with Z bodies (Z). Ribosomes (rb) lie adjacent to thick filaments. The presence of occasional tubules (tb) suggests the early development of the sarcoplasmic reticulum. X 30,000. 11 Focus of thin filaments associated with a 2 body (Z) (vastus lateralis, 17-week-old fetus). The majority of the filaments in the field are thin. This illustration suggests t h a t filament assembly in conjunction with a Z body is independent of thick filaments. Note the virtual absence of ribosomes. X 30,000. 12 Perinuclear region (vastus lateralis, 25-week-old fetus). Such regions appear to be the prime focus for
“intermediate” filament (fi) synthesis. Such filaments, approximately 10-12 nm in diameter, are longer than contractile filaments, and are unaligned; scattered polyribosomes, free ribosomes and microtubules (mt) also typify such regions. Go, Golgi apparatus; nu, nucleus. X 30,000.
238
DIFFERENTIATION OF HUMAN FETAL SKELETAL MUSCLE Robert J . Tomanek and Ann-Sofi Colling-Saltin
PLATE 3
239
PLATE 4 EXPLANATION OF FIGURES
13 Early formation of Z bodies (*I. Such structures appear to be developed in conjunction with unaligned thin filaments. Relatively parallel thick filaments (arrowheads) are seen adjacent to these areas and appear to forecast sarcomere formation. X
30,000.
14 Filament assembly and sarcomere formation (vastus lateralis, 18-week-old fetus). The most peripheral region (1)of this myotube contains chiefly thin filaments, occasional Z bodies (*) and a few parallel thick filaments. Adjacent to this area (2) lies a focus of numerous thick filaments and their ribosomes. Isolated filaments are less evident but can be seen in the vicinity of a small Z body (arrow). Some thin filaments appear assembled in parallel to thick filaments. Alignment of thin and thick filaments and the appearance of a Z line (Z) are illustrated in a deeper cell region (3). Note the I (I) and A (A) bands. X 30,000. 15 Longitudinal growth (rectus abdominis, neonate). This micrograph shows a region distal to sarcomeres and illustrates the presence of disoriented "intermediate" filaments (fi) and microtubules (mt) typical of the peripheral and perinuclear regions of fetal specimens. Numerous sarcolemmal and subsarcolemmal vesicles (ve) are evident. mi, mitochondria. X 30,000.
240
DIFFERENTIATION OF HUMAN FETAL SKELETAL MUSCLE Robert J . Tomanek and Ann-Sofi Colling-Saltin
PLATE 4
241
PLATE 5 EXPLANATION OF FIGURES
16 Development of T-system (vastus lateralis, 18-week-old fetus). Two T-tubules (TI are seen a t the plasma membrane; associated with each tubule are dense structures which appear to be terminal cisternae of the sarcoplasmic reticulum. Such tubules are limited to the peripheral cell regions containing contractile filaments. Vesicular or tubular structures (*I lacking associated terminal cisternae are also seen. X 30,000. 17 Extension of a T-tubule (TI from the plasma membrane (arrowheads) of a soleus myotube (18-weekold fetus). Terminal cisternae (tc) are seen forming triads or diads. Branching of T-tubules is common. X 30,000. 18 Bristle-coated vesicles (arrowheads) from the vastus lateralis (28-week-old fetus). This micrograph illustrates an apparent fusion of an undifferentiated (satellite) cell (lower half of field) with a myotube. The apposed plasma membranes of the two cells are seen (between arrows). This type of vesicle is found throughout fetal devevlopment; they are most common in myotubes and early myofibers but also occur in satellite cells. X 30,000. 19 Bristle-coated vesicles (*) as components of a tubular (tb) system (soleus, 18-week-old fetus). The vesicles, as illustrated here, assume the position of terminal cisternae. X 96,000. 20 Sarcoplasmic reticulum (arrows) in a myotube (vastus lateralis, 7-month-fetus).A focus of assembling filaments with ribosomes often includes tubular structures even before the estalishment of sarcomeres. X 30,000. 21 Tubule (tb) with bristle-coated vesicle (arrowhead) from the same specimen a s figure 20. In the later stages of fetal life these structures are seen in deeper portions of the cell and may lie in the proximity of a sarcomere as shown here. Triads (*) are increasingly common a t this stage of development. x 30,000. 22 Sarcoplasmic reticulum (rectus abdominis, 7-month-old fetus). In conjunction with sarcomere development, elements of the sarcoplasmic reticulum (arrowheads) frequently are closely associated with Z lines. This association is not evident in early fetal life (13-18). X 30,000.
242
DIFFERENTIATION OF HUMAN FETAL SKELETAL MUSCLE Robert J. Tomanek and Ann-Sofi CollingSaltin
PLATE 5
243
PLATE 6 EXPLANATION OF FIGURES
23 Organelle development within a myotube (soleus, 18-week-old fetus). Filament alignment is evident in the cell a t the top of the illustration, but polarization of forming myofibrils is not uniform. Numerous polyribosomes as well as a profile of rough endoplasmic reticulum (er) are seen. Most membranous structures lie near the plasma membrane; some resemble sarcoplasmic reticulum tubules (tb). Mitochondria (mi) a r e sparse and are limited in structural detail. X 20,000. 24
Double-membraned vesicular structure (rectus abdominis, 1-month-old infant). Such a structure (arrow) lies in a perinuclear region and contains glycogen; i t may be a n early autophagic vacuole. X 12,000.
25 Glycogen-containing vacuole (soleus, 30- to 32-week-old fetus). Note t h a t the glycogen is limited primarily to the vacuole, suggesting t h a t the vacuole is autophagic. X 30,000. 26 A muscle cell in the later myotube stage (vastus lateralis, 19-week-old fetus). As myofibrils become demarcated and attain a sufficient diameter, ribosomes become sparse, but other organelles are seen more frequently: small mitochondria (mi) with ill-defined cristae, subsarcolemmal vesicles (ve), sarcoplasmic reticulum (sr), rough endoplasmic reticulum (er) and bristle-coated vesicles (arrowhead). A structure resembling a multivesicular body is seen (*). X 39,000.
244
DIFFERENTIATION OF HUMAN FETAL SKELETAL MUSCLE Robert J. Tomanek and Ann-Sofi Colling-Saltin
PLATE 6
ABSTRACT The ultrastructural differentiation of several different muscles was investigated in human fetuses ranging in age from 13 weeks to neonatal. A t approximately 16 weeks of gestation cell clusters containing both myotubes and satellite cells lie enclosed by a newly formed basal lamina and show evidence of fusion. The development of organelles is evident in myoblasts, proceeds as the cells transform into myofibers, and continues in the neonate. Filament synthesis occurs primarily in the cell periphery where thin filaments appear to align themselves in relations to parallel arrays of ribosome-studded thick filaments; Z line formation follows the appearance of thin filaments. Intermediate filaments, approximately 10-12 nm thick, were also consistently observed in perinuclear regions and distal to filament assembly. Although sarcoplasmic reticulum (SR) development is closely related to fibril formation, connections between Z lines and SR are not consistent, thus supporting the conclusion that SR does not evoke the formation of the Z line. Bristlecoated vesicles appear to be the precursors of elements of the SR, possibly the lateral sacs. Development of the transverse tubules, as invaginations of the sarcolemma, is closely associated with the formation of lateral sacs since the latter occur along the sarcolemma as soon as transverse tubules appear. Cytological differentiation is similar, though not identical, in several different muscles. During the last trimester muscle fibers show some evidence of diversity, mainly of variation in Z line width. In general the results suggest that the sequence and stages of human myogenesis are similar to those of other species. Nearly 60 years ago the histogenesis of striated muscle of man and other vertebrates was comprehensively described (Tello, '17). Since that time numerous studies have contributed to a basic understanding of myogenesis in vertebrate species (Fishman, '73). Studies dealing with the ultrastructure of human fetal muscle, however, have been less comprehensive; several investigations have focused on the role of the satellite cell (Ishikawa, '66, '70; Conen and Bell, '70).Other studies employing electron microscopy have provided a general description of myogenesis occurring during the first five and one half months in utero (Hudgson et al., '70; Ochoa and Mair, '68), established similarities between chick and human myogenesis (Hudgson et al., '701, elucidated the morphological relationship between polyribosomes and myosin filaments (Larson et al., '691, and suggested a relationship between the sarcoplasmic reticulum, fibril and triad (Walker et al., '75). AM. J. ANAT., 149: 227-246.
Since most of the work on human fetal skeletal muscle has focused on rather specific aspects of differentiation and often considered only specific periods of fetal development, a more comprehensive approach appeared to be necessary for the elucidation of the sequence of cytological differentiation. The observations reported here are based on several muscles of fetuses ranging in age from 13 weeks to neonatal. Our observations have taken into account the relationship of various organelles and cells and the consistency of sequential events. MATERIALS AND METHODS
Muscle tissues from human fetuses and infants were obtained a t necropsy from the vastus lateralis (19 specimens), soleus (10 specimens) and rectus abdominis (8 specimens). Strips of muscle, approximately 3 mm Accepted January 19, '11.
227
228
ROBERT J. TOMANEK AND ANN-SOFI COLLING-SALTIN
in thickness, were tied to wooden splints and fixed in a solution containing: 2.5% glutaraldehyde, 0.1% sodium cacodylate, 0.8%paraformaldehyde and 0.02 M CaClz (ph 7.39). Following an overnight wash in buffer containing 3% sucrose, the tissues were post-fixed in veronal acetate-buffered 1% OsO,, dehydrated in alcohol and embedded in Epon. A Siemens 101 Elmiscope was used to examine thin sections which were stained with uranyl acetate and lead citrate. Sections one micron thick were stained with Richardson’s stain (Richardson et al., ’60) and used for light microscopy. The problem of post-mortem artifact was evident in several specimens and attributed to a lengthy time lapse between death and fixation (the availability of the tissue varied from a few minutes to two hours post-mortem). For this reason it was necessary to limit the use of such inadequately fixed specimens to cursory observations. Z line width was estimated in photomicrographs ( X 30,000) with the aid of a n ocular micrometer (Tomanek, ’76). RESULTS
Nomenclature of myogenesis
The nomenclature regarding the various cell stages comprising myogenesis is defined as follows. “Presumptive myoblasts” are similar in morphology to other mesenchymal cells; they divide mitotically, but do not fuse. Although recent evidence indicates t h a t presumptive myoblasts transcribe myosin mRNAs (Chi e t al., ’751, these cells lack myofibrils. “Myoblasts,” on the other hand, synthesize and organize thin and thick filaments (Holtzer e t al., ’75). The contractile elements are more notable when myoblasts fuse to form a multinucleated syncytium termed a “myotube.” Circumferentially distributed myofibrils or filaments, centrally placed nuclei, and a central core of sarcoplasm are characteristic of this stage. Obviously there is no sharp point of demarcation between this stage of myogenesis and a mature cell or “myofiber” (Fishman, ’73). The latter arises as nuclei are peripherally displaced and myofibrils and other organelles fill the core of the myofiber. With the advent of the basal lamina it is possible to identify a cell which resembles a presumptive myoblast but lies closely apposed to a myofiber or myotube and within the confines of its basal lamina. Such a spindleshaped, mononucleated cell constitutes a “satellite cell” (Ishikawa, ’70).
Stages of. myogenesis ~-
Most of the muscle cells encountered in specimens from 13 to 16-week-old fetuses were myotubes or closely apposed undifferentiated cells. The latter resembled presumptive myoblasts or satellite cells. In 13- to 14-weekold fetuses these cells lacked a well defined basal lamina, which was presumably undergoing development since a faint, fuzzy structure could often be visualized. A closely associated undifferentiated cell appeared to lie enclosed within the same presumptive basal lamina which surrounded one to several cells. This relationship was more evident a t the 16week stage and thereafter. Myotubes representing various levels of maturity were common. Some contained myofibrils which were composed of only a few sarcomeres, while in others myofibrils were well established with numerous sarcomeres and were longitudinally oriented. This stage was typified by clusters of three to ten cells each separated by a narrow (15-50 nm) interspace. Between 16 and 20 weeks the nuclei in some myotubes became peripherally displaced giving rise to early myofibers. Many cells still occurred in groups and lacked an intervening basal lamina. Numerous satellite cells were encountered during this period as well as during later stages of fetal development. Commonly one satellite cell was associated with two myofibers or myotubes, i.e., the three cells were encased by a common basal lamina. In addition, we observed cells which appeared to be connected in part to adjacent cells as indicated either by “fusion” of their cell membranes, by sarcoplasmic bridges, or, as shown in figures 3 and 4, filaments or a myofibril were seen extending between two cells. This phenomenon was limited to cells which were in the myotube stage and involved two myotubes, or a myotube and a myoblast or satellite cell. Myotubes were found throughout fetal development although their frequency of occurrence decreased with time (figs. 5-71. Myofibers predominated after approximately 20 to 22 weeks of gestation and were characterized by relatively wide separation of fibrils which often contained only a few sarcomeres and lacked a longitudinal orientation. A predominance of uniform fibrils with many Z lines in registry with adjacent fibrils was not evident until late fetal development, i.e., approximately 30 weeks of gestation. This later pe-
DIFFERENTIATION OF HUMAN FETAL SKELETAL MUSCLE
riod of fetal life was typified by some increase as well as variation in fiber diameter. Prior to this time fiber diameter did not increase markedly, since the increasing number of contractile filaments replaced the sarcoplasm characteristic of myotubes. The most marked increase in fiber diameter occurred a t the time of birth (fig. 8) and thereafter. The presence of glycogen-containing vacuoles was evident in fibers during late fetal (3036 weeks) and early post-natal (up to 1 month) life. They usually consisted of a double membrane (fig. 24) or membranous fragments (fig. 25). In muscles from post-natal animals the glycogen-containing vacuoles lay in cell regions which were relatively free of glycogen particles.
Contractile filaments In the early myotube, the contractile filaments were first observed in the cell periphery where disorganized or poorly aligned thin filaments were positioned close to the cell membrane, with clusters of thick filaments in close proximity along with their associated ribosomes (fig. 9). As illustrated in figure 10, lattice formation involving thin and thick filaments as well as Z lines was evident as soon as both types of filaments appeared. This hexagonal array was evident in cross sections even when only a few thick filaments were seen. Unoriented thin filaments were peripheral to the lattice structure, suggesting their assembly as new myosin was synthesized. Filament assembly and lattice formation may not be dependent upon the presence of Z bodies. Clusters of thin and thick filaments without an associated Z body were most frequently encountered. A failure to observe such structures could be due to the plane of section. In such situations, however, the Z body would lack connections with the filaments; yet, whenever we encountered Z bodies, fine filaments were seen extending from them. The earliest appearance of dense material suggestive of Z bodies was consistently observed in foci of thin filaments (figs. 13,14). Some possible stages of sarcomere formation may be visualized by moving inward from the cell membrane (fig. 14). This illustration suggests: (1) rather independent sites of thin and thick filament synthesis-the myosin (thick) filaments lying further from the cell membrane, (2) a decrease in the number of myosin-associated ribosomes as lattice formation proceeds, and (3) Z body formation
229
in foci of thin filaments in the absence or presence of thick filaments. Subsequent stages of myogenesis involve patterns similar to those observed in early myotubes. Perinuclear regions were typified by disoriented filaments intermediate in diameter (approximately 10-12 nm) t o thin and thick contractile elements (fig. 12). The intermediate filaments were also found distal to regions of sarcomere formation (fig. 15). They may be distinguished from thin filaments by comparing figures 13-15. Microtubules, approximately 25 nm in diameter, were characteristic of regions of filament synthesis and early assembly (figs. 9,12, 15), but were absent in regions of more advanced myofibril formation. Usually their orientation was longitudinal or oblique and they were situated near the ends of developing sarcomeres. In addition to intermediate filaments and microtubules, numerous vesicles and ribosomes were typical of regions distal to forming sarcomeres (fig. 15). The ribosomes included three varieties: free, membrane-associated and polyribosomes.
Tubular system Early fetal development (13-22 weeks) was characterized by various tubular or vesicular structures. Pleomorphic vesicular structures were most common in the early myotubes, while scattered tubular elements resembling sarcoplasmic reticulum (SR)occurred more frequently in late myotubes and early myofibers. Some of the latter were recognized as T-tubules since they could be seen as invaginations of the sarcolemma (figs. 16, 17). It is inferred that the development of the sarcoplasmic reticulum is related to that of the contractile filaments, since the appearance of sarcomeres was accompanied by scattered SR elements (figs. 19-22). The tubules, which varied in width (20-60 nm) were found in the peripheral sarcoplasm along with newly aligning filaments, and in perinuclear regions. They were notably absent from the core of the myotube, which contained numerous glycogen granules. Sarcolemmal and subsarcolemmal vesicles were also common, particularly in myotubes (fig. 23, 26). A coordinated development of the SR and filament components of muscle is suggested by the fact that cells which were in the early myotube stage had a limited amount of SR. This was also true of specimens from older
230
ROBERT J. TOMANEK AND ANN-SOFI COLLING-SALTIN
fetuses (22-30 weeks of age). While components of the sarcoplasmic reticulum were frequently closely apposed to Z lines (fig. 221, this relationship was not evident in the earliest stages of development, i.e., when Z bodies were first seen. On the other hand, SR fragments could be found associated with a few filaments prior to Z line formation (fig. 20). Further development of the SR was evidenced by branching a t the Z line as noted in crosssectional fields. We did not observe extensive branching and anastomoses, typical of adult fibers, even in the neonate. However, some variability in the amount of SR in different fibers was indicated during late fetal life. Triads or diads were evident a t relatively early stages of development. Isolated triads, i.e., without apparent connections with other elements of sarcoplasmic reticulum, were evident near or a t the sarcolemma in the 19week fetus (fig. 16). Examples of sarcolemmal invaginations contributing to a triadic component were found in relatively early myotubes (fig. 17). The absence of triads in deeper regions of the cell a t this stage suggests that the terminal cisternae as well as the T-tubule may originate a t or near the sarcolemma. Between 22 to 25 weeks of fetal life numerous triads, most of which were longitudinally oriented, were found dispersed throughout the myofibers; a few were aligned a t the A-1 junction, while most lay disoriented. Even a t this later developmental stage connections with the longitudinal elements of the sarcoplasmic reticulum were not as common as in adult mammalian muscle. Since terminal cisternae were consistently found in association with Ttubules, the development of the two structures appeared to be closely associated. The longitudinal components of the SR, on the other hand, may be developed independently. Spherical vesicles characterized by the presence of bristles on their surfaces (figs. 2, 14, 18, 19) were frequently encountered in myofibers, satellite cells and myotubes. As suggested by the profile in figure 19, the spherical vesicles appeared to be the terminal components of a tubular system. Such structures were occasionally observed in proximity to sarcomeres (fig. 21), but were more common in the perinuclear and subsarcolemmal regions of fibers or randomly distributed in undifferentiated (satellite) cells.
Cell variation in differentmuscles Most of the observations described apply to
specimens from different muscles. Since variations might suggest differentiation into specific fiber types, Z line width was measured. Between 13 and 19 weeks, Z line widths were similar and not highly variable in both the soleus and vastus lateralis, ranging between 74-118 nm. Later in fetal life (25-31 weeks) the soleus contained fibers which were more homogeneous; Z line width in most fibers fell within the 112-160 nm range. The fibers of the vastus lateralis were more variable in this regard with values ranging from 71-140 nm; in the rectus abdominis the Z line widths were even more varied (75-170 nm). Because all of the muscles studied eventually become diversified, we lacked a model with a homogeneous fiber population. However, the soleus, with predominantly one fiber type, can be contrasted with the other muscles and, therefore most closely represents such a model. In comparing the soleus to other muscles (primarily the vastus lateralis) during early fetal development few differences were apparent. However, in late fetal life most fibers comprising the soleus and many found in the rectus abdominis were characterized by wide Z lines and poorly developed SR. Thus, an apparent differentiation into specific fibers is suggested. Compared to the vastus lateralis, the soleus matured more rapidly. Not uncommonly, cells with relatively densely packed myofilaments and peripherally situated nuclei were evident in specimens from 13-week-old fetuses. During this early period of fetal development a considerable variation in cell diameter was evident in the soleus and could be contrasted with a rather uniform cell size in the vastus lateralis and rectus abdominis. This variability was not evident in the latter muscles until 18 weeks in utero. DISCUSSION
In vitro, time-lapse studies on chick embryos have demonstrated that myogenic cells fuse to form muscle syncytia or myotubes (Capers, '60). Subsequent electron microscopic investigations have supported this finding (Shimada et al., '67; Fishman, '70). Multinucleation during myogenesis includes cytoplasmic fusion of myoblasts and myotubes in the absence of a nerve supply (Konigsberg et al., '60; Stockale and Holtzer, '61; Shimada et al., '67). Our own observations in human specimens are consistent with these reports and suggest that undifferentiated cells lie closely
DIFFERENTIATION OF HUMAN FETAL SKELETAL MUSCLE
apposed to myotubes, are incorporated within a common basal lamina, and may fuse with the latter myotubes. Evidence for this contention is based on observations illustrating: (1) fusion of cell membranes, (2) formation of sarcoplasmic bridges which include filaments or fibrils, and (3) similar cell associations between satellite cells and muscle cells, or between myotubes or early muscle fibers. The fact that the number of satellite cells decreases during late fetal life (Ishikawa, '66) supports evidence that fusion occurs immediately or subsequent to the production of filaments. This hypothesis is based on our findings that small cells which lie under the basal lamina of a muscle cell may have the characteristics of a satellite cell, an early differentiating cell (similar to a satellite cell but containing a few filaments), or a myotube. Since no mode of nuclear replication is known to occur in multinucleated muscle cells (Stockdale and Holtzer, '611, the sarcoplasmic bridges are not likely to be due to cytokinesis. The increase in the total number of nuclei in a given muscle during growth is accounted for by multiplication of satellite cells (Enesco and Puddy, '64; Allbrook, '71; Moss and Leblond, '71). Glycogen-containing vacuoles, which appear during late fetal life, have been described in r a t skeletal muscle (Schiaffino and Hanzlikova, '72). They are considered to be autophagic vacuoles, i.e., components of the lysosomal system which mobilize glycogen a t the time of birth. As in liver, glycogen accumulates in skeletal muscle during the late fetal stages and is rapidly mobilized a t birth (Dawes and Shelley, '68). Our observations in the human appear to be consistent with studies on the rat (Schiaffino and Hanzlikova, '72), with the exception that they were encountered earlier in human muscle.
231
synthesis is not known. In any case, both thick and thin filaments occur in early muscle cells as evidenced in all of our human specimens; thus the process is comparable to chick myogenesis (Fischman, '67). A third type of filament evident in human fetal muscle is of the same dimensions as the intermediate filament demonstrated by Ishikawa et al. ('68). Such filaments appear to differ from actin in that they do not form the characteristic arrowhead complexes with heavy meromyosin (Ishikawa et al., '69). The possibility t h a t earlier reports described intermediate filaments as thin filaments has been suggested (Ishikawa et al., '68). Our observations, however, clearly indicate the presence of both types of filaments in myotubes, with most of the cortical filaments being classified as thin. Polyribosomes are structurally related to the formation of thick filaments in human muscle where they can be found consistently near the end of the filament nearest the Iband (Larson et al., '69). Our observations support this finding and underline the consistency of the association throughout the cell, but a structural association between thin filaments and ribosomes is not indicated since regions containing only thin filaments may or may not contain scattered ribosomes. With the appearance of thick filaments a hexagonal array incorporating the thin filaments is detected, as has been observed in chick muscle (Allen and Pepe, '65; Fischman, '67). The absence of Z line material or Z bodies in regions of early filament associations is consistent with evidence that such structures are not necessary for the formation of the hexagonal filament pattern (Stromer et al., '69). While Z lines are not requisite for this aspect of myogenesis their formation is closely linked with that of other sarcomere components and is essential for longitudinal growth Filaments (Fischman and Zak, '71). Since the appearIn early myotubes, i.e., cells with limited ance of Z bodies is invariably linked with the numbers of cytofilaments, thin filaments (5-6 presence of thin filaments, the synthesis of nm in diameter) predominate. Although corti- the former may be dependent upon the prescal filaments bind heavy meromyosin (Ishi- ence of the latter. kawa et al., '69), there is not yet ample eviTubular system dence demonstrating that this binding is Sarcoplasmic reticulum (SR) is derived limited to actin. Some authors have suggested that thin filaments are synthesized prior to from the endoplasmic reticulum of undifferenthick filaments in chick (Allen and Pepe, '65; tiated cells (Ezerman and Ishikawa, '67). Hudgson et al., '70; Ogawa, '62) as well as Vesicular structures observed in the early human (Hudgson et al., '70) muscle. Whether myotube are probably an early stage of SR. the presence of actin is essential for myosin Irregular tubules show some branching as the
232
ROBERT J. TOMANEK AND ANN-SOFl COLLING-SALTIN
cell matures and, as previously described in the human diaphragm, (Walker et al., '75) are associated with regions of the cell containing filaments and sarcomeres. This is in contrast to an earlier report which indicated little evidence for SR in human myotubes which are maturing into fibers (Ochoa and Mair, '68). A relationship between SR and fibril development in the rat (Walker and Edge, '711, monkey (Walker et al., '75) and man (Walker et al., '75) has been suggested on the basis that SR tubules appear to be connected to and encircle Z lines during myogenesis. Since the SR network has been observed during early fibril development, it has been hypothesized that the SR forms a framework for filament and Z line assembly (Walker et al., '75). Conversely, it has been suggested that the SR develops around assembling filaments (Fischman, '73). Our own observations indicate numerous Z lines in continuity with the SR, but this feature is by no means consistent. If these tubules completely encircle the Z line, as previously reported (Walker et al., '751, tubular segments should be included in any longitudinal section through a Z line. Our observations indicate this is not the rule. This, of course, does not rule out a structural connection between the two components. Since Z bodies which include attached thin filaments are infrequently associated with tubules it appears doubtful that the Z line and SR play a significant role in the assembly of filaments. A more plausible contention may be that the early development and organization of SR coincides with early sarcomere formation. A subsequent positioning of Z lines in register with adjacent fibrils (Walker et al., '75) may, however, be a function of Z line-associated tubules. The T system in cultured chick muscle has been shown to develop by invaginations of the sarcolemma of the myotube (Ezerman and Ishikawa, '67; Ishikawa, '68). Although a previous study on human fetal muscle could not find direct evidence for this pattern of development (Hudgson et al., '701, our observations suggest that this is the case. First, invaginations were observed and seen to form diads or triads in the cell periphery. Second, diads and triads in early myotubes were seen only in the cell periphery and most frequently in association with the sarcolemma. Finally, diads and triads occur in deeper areas of the cell only after myofibrils become centrally displaced.
SR cisternae are presumed to expand and form their characteristic structure after contacts between SR and T-tubules have formed (Edge, '70). The observation that during early fetal development cisternae seldom show connections to other SR components suggests either that they develop independently or that they form on short fragments of SR which eventually unite with other SR tubules. Spherical vesicles characterized by bristles on their outer surface were previously seen in primate muscle as evaginations of T-tubules (Walker et al., '75). That such vesicles are components of a tubular system is also evident from our observations. Their location in the cell periphery (particularly in the early myotube) fits with the suggestion that they are associated with the T-tubules, although their density and dilated form is more akin to SR terminal cisternae than to a T-tubule. Their presence in satellite cells may lend support to the suggestion that they are components of the SR, since the T-system is not evident until the myotube stage. Characteristically the SR terminal cisternae contain a dense material which may consist of mucopolysaccharides or glycoproteins and accordingly provide sites for calcium binding (Philpott and Goldstein, '67). Diversification of fibers From our observations, it is evident that cytological differentiation in the human fetus is similar, though not identical, in several different muscles. The development of various organelles may thus constitute a primary phase, common to all muscle fibers (Eccles, '631, which is under genetic regulation. Fiber diversification, on the other hand, is controlled by neural factors (Hanzlikova and Schiaffino, '73). Such diversification is not clearly evident during the first six months of fetal life, after which time the soleus becomes more homogeneous while the fibers in the vastus lateralis become more diversified with respect to size, Z line width and extensiveness of sarcoplasmic reticulum. A more complete diversification undoubtedly occurs after birth as evidenced by studies on the cat (Tomanek, '76). In this respect as well as the sequence of organelle development and stages of myogenesis, human fetal muscle development and differentiation are similar to those of other species.
DIFFERENTIATION OF HUMAN FETAL SKELETAL MUSCLE LITERATURE CITED Allen, E. R., and F. A. P e p 1965 Ultrastructure of developing muscle cells in the chick embryo. Am. J. Anat., 116: 115-147. Bergman, R. A. 1962 Observations on the morphogenesis of r a t skeletal tissue. Bull. Johns Hopkins Hosp., 110: 187-201. Capers, C. R. 1960 Multinucleation of skeletal muscle in uitro. J. Biophys. Biochem. Cytol., 7: 559-566. Chi, J. C., S. A. Fellini and H. Holtzer 1975 Differences among myosins synthesized in non-myogenic cells, presumptive myoblasts and myoblasts. Proc. Nat. Acad. Sci., (U. S. AJ, 72: 4999-5003. Conen, P. E., and C. D. Bell 1970 Study of satellite cells in mature and fetal muscle and rhabdomyosarcoma. In: Regeneration of Striated Muscle and Myogenesis. H. Mauro, S . A. Shafiq and A. T. Milhorat, eds. Excerpta medica, Amsterdam, pp. 194-211. Cullen, M. J., and D. Weightman 1975 The ultrastructure of normal human muscle in relation to fibre type. J. Neurol Sci., 25:(1): 43-56. Dawes, G. S.,and J. H. Shelley 1968 Physiological aspects of carbohydrate metabolism in the foetus and newborn. In: Carbohydrate Metabolism and Its Disorders. F. Dickens, P. J. Randle and W. T. Whelan, eds. Academic Press Inc., New York, 2: 87-121. Edge, M. B. 1970 Development of apposed sarcoplasmic reticulum a t th e T system and sarcolemma and t he change in orientation of triads in r a t skeletal muscle. Develop. Biol., 23: 634-650. Elliott, B. J., and D. G. F. Harriman 1975 Growth of human muscle spindles in uitro. Nature, 251: 622-624. Enesco, M., and D. Puddy 1964 Increase in the number of nuclei and weight in skeletal muscles of rats of various ages. Am. J. Anat., 114: 235-244. Ezerman, E. B., and H. Ishikawa 1967 Differentiation of the sarcoplasmic reticulum and T system in developing chick skeletal muscle in vitro. J. Cell Biol., 35: 405-420. Fischman. D. A. 1967 An electron microscoDic studv of myofibril formation in embryonic chick skeletal muscle. J. Cell. Biol., 32: 557-575. 1972 Development of striated muscle. In: The Structure and Function of Muscle. Vol. 1. Second ed. G. H. Bourne, ed. Academic Press, New York. Fischman, D. A,, and R. Zak 1971 Assembly of myofibrils in the absence of protein synthesis. J. Gen. Physiol., 57: 245. Holtzer, H. 1970 Myogenesis In: Cell Differentiation. 0. Schjeide and J. deVellis, eds. Van Nostrand-Reinhold, Princeton, pp. 476-503. Holtzer, H., K. Strahs and J. Biehl 1975 Thick and thin filaments in postmitotic, mononucleated myoblasts. Science, 188: 943-945. Hudgson, P., M. Jenkison and P. F. Larson 1970 An ultrastructural study of chick embryo and human foetal muscle. In: Proceedings International Congress on Muscle Diseases, Milan 1969.J. N. Walton, N. Canal andG. Scarlato, eds. Excerpta medica, Amsterdam, pp. 90-97. Ishikawa, H. 1966 Electron microscopic observations of satellite cells with special reference to the development of mammalian skeletal muscles. Z. Anat. Entw. Gesch., 125: 43-63. 1968 Formation of elaborate networks of T system tubules in cultured skeletal muscle with special reference to the T. system formation. J. Cell Biol., 38: 51. 1970 Satellite cells in developing muscle and tissue culture. In: Regeneration of Striated Muscle and
233
Myogenesis. A. Mauro, S. A. Shafiq and A. T. Milhorat, eds. Excerpta medica, Amsterdam, pp. 167-179. Ishikawa, H., R. Bischoff and H. Holtzer 1968 Mitosis and intermediate-sized filaments in developing skeletal muscle. J. Cell Biol., 38: 538-555. 1969 Formation of arrowhead complexes with heavy meromyosin in a variety of cell types. J. Cell Biol., 43: 312-328. Konigsberg, I. R. 1965 Aspects of cytodifferentiation of skeletal muscle. In: Organogenesis. R. L. de Haan and H. Ursprung, eds. Holt, Rinehart and Winston, Inc., New York. Konigsberg, I. R. N., N. Mcelvain, M. Tootle and H. Herrmann 1960 The dissociability of ribonucleic acid synthesis from the development of multinuclearity of muscle cells in culture. J. Biophys. Biochem. Cytol., 8: 333-343. Larson, P. F., P. Hudgson and J. N. Walton 1969 Morphological relationship of polyribosomes and myosin filaments in developing and regenerating skeletal muscle. Nature, 222: 1168-1169. Lipton, B. H., and I. R. Koningsberg 1972 A fine-structural analysis of the fusion of myogenic cells. J. Cell Biol., 5 3 W : 348-364. Mauro, A. 1961 Satellite cell of skeletal muscle fibers. J. Biophys. Biochem. Cytol., 9: 493-495. Moss, F. P., and C. P. Leblond 1970 Nature of dividing nuclei in skeletal muscle of growing rats. J. Cell Biol., 44: 459-462. Ochoa, J., and W. G. P. Mair 1968 The ultrastructure of normal foetal muscle and foetal muscle from known dystrophic-carriers. In: Proceedings of 4th Symposium on Current Research in Muscular Dystrophy. Pitman Medical, London, pp. 223-239. Ogawa, Y. 1962 Synthesis of skeletal muscle proteins in early embryo and regenerating tissue of chick and triturus. Exp. Cell Res., 26: 269-274. Philpott, C. W., and M. A. Goldstein 1967 Sarcoplasmic reticulum of striated muscle: localization of potential calcium binding sites. Science, 155: 1019-1020. Richardson, K. C., L. J a re tt and E. H. Finke 1960 Embedding in epoxy resins for ultrathin sectioning in electron microscopy. Stain Tech., 35: 313-323. Schiaffino, S., and V. Hanzlikova 1972 Autophagic degradation of glycogen in skeletal muscles of the newborn rat. J. Cell Biol., 52: 41-51. Shimada, Y., D. A. Fischman and A. A. Moscona 1967 The fine structure of embryonic chick skeletal muscle cells differentiated in vitro. J. Cell Biol., 35: 445-453. Stockdale, F. E., and H. Holtzer 1961 DNA synthesis and myogenesis. Exp. Cell Res., 24: 508-520. Stromer, M. H., D. J. Hartshorne, H. Mueller and R. V. Rice 1969 The effect of various protein fractions on Z- and Mline reconstitution. J. Cell Biol., 40: 167-178. Tello, J. F. 1917 Genesis de las terminaciones nerviosas motrices y sensitivas. I: En el sisterma lacomotor de las vertebrados superiores. Histogenesis muscular. Trabajos Lab. Invest. Biol. Univ. Madrid, 15: 101.199. Tomanek, R. J. 1976 Ultrastructural Differentiation of skeletal muscle fibers and their diversity. J. Ultrastr. Res., 55: 212-227. Walker, S.M., and M. B. Edge 1971 The sarcoplasmic reticulum and development of Z lines in skeletal muscle fibers of fetal and postnatal rats. Anat. Rec., 169: 661-678. Walker, S. M., G. R. Schrodt and G. J. Currier 1975 Evidence for a structural relationship between successive parallel tubules in the SR network and supernumerary striations of 2 line material in purkinje fibers of the chicken, sheep, dog and Rhesus monkey heart. J. Morph., 147: 459-474.
PLATE 1 EXPLANATION OF FIGURES
1 A myotube and its satellite cell (SC) (rectus abdominis, 28-week-old fetus). The latter contains free ribosomes and polyribosomes (arrowheads). A basal lamina (bl) encloses the two cells but is absent between their apposing plasma membranes (arrows). x 20,000.
2 Structural association of two myotubes (vastus lateralis, 18-week-oldfetus). The two cells are enclosed by a common basal lamina (bl) which is (as in fig. 1). absent between their apposing plasma membranes. One cell (A) has a peripheral rim of contractile elements not yet organized into defined myofibrils and also contains numerous polyribosomes (arrowheads). A more “mature” myotube (B) contains some defined myofibrils. Note the peripherally situated bristle-coated vesicle (arrow). X 20,000.
3 Apparent fusion of two cells (vastus lateralis, 22-week-old fetus). One cell (A) is either an early myotube or a myoblast containing only a few contractile filaments. Filaments from a more advanced myotube (B) a re seen extending into the neighboring cell. ES, extracellular space; sa, sarcolemma. X 50,000.
234
DIFFERENTIATION OF HUMAN FETAL SKELETAL MUSCLE Robert J . Tomanek and Ann-Sofi Colling-Saltin
PLATE 1
PLATE 2 EXPLANATION OF FIGURES 4
Apparent fusion of two myotubes (more advanced fusion from the soleus of a 16 to 17week-old fetus. A myofibril (my) extends between the two cells, which are joined by a narrow channel of sarcoplasm. Polyribosomes (arrowheads) and unassembled filaments (fi) a r e present in both cells. ES, extracellular space. X 19,500.
Figs. 5-8 Cross-sectional fields of 1-pm Epon sections stained with Richardson’s solution. 5 Muscle cells from a 13-week-old fetus (vastus lateralis). Virtually all cells are in the myotube stage as evidenced by centrally situated nuclei and, in non-nuclear regions, a central core of cytoplasm (arrow-heads). X 830. 6 Muscle cells from a 13-week-old fetus (soleus). Compared to the vastus lateralis, cell diameter varies considerably and some cells are early myofibers as evidenced by peripherally situated nuclei (arrows). X 830. 7 Muscle cells from an 18-week-old fetus (vastus lateralis). Myofibrils in most cells are
relatively densely packed. In some cells nuclei are situated peripherally while in others they a r e centrally positioned. Unlike the earlier stage (fig. 5) cells vary considerably in diameter. X 830. 8
236
Muscle cells from a one-month-old infant (rectus abdominis). Cell diameter is considerably greater than in any fetal stage and myotubes are virtually absent. Some variation in cell diameter is evident. X 830.
DIFFERENTIATION OF HUMAN FETAL SKELETAL MUSCLE Robert J. Tomanek and Ann-Sofi Colling-Saltin
PLATE 2
237
PLATE 3 EXPLANATION OF FIGURES
9 Early myotube (vastus lateralis, 18-week-old fetus). The region above the arrowheads lies in proximity to the plasma membrane and contains thin filaments and extremely limited numbers of ribosomes. Below and to the left of this region thick filaments lie in parallel and are associated with ribosomes (rb). Microtubules (mt) are common near the ends of the assembled thick filaments. X 30,000. 10 Early myotube (soleus, 18-week-oldfetus). The early formation of sarcomeres is illustrated; both thick and thin filaments lie in close proximity and appear to undergo assembly along with Z bodies (Z). Ribosomes (rb) lie adjacent to thick filaments. The presence of occasional tubules (tb) suggests the early development of the sarcoplasmic reticulum. X 30,000. 11 Focus of thin filaments associated with a 2 body (Z) (vastus lateralis, 17-week-old fetus). The majority of the filaments in the field are thin. This illustration suggests t h a t filament assembly in conjunction with a Z body is independent of thick filaments. Note the virtual absence of ribosomes. X 30,000. 12 Perinuclear region (vastus lateralis, 25-week-old fetus). Such regions appear to be the prime focus for
“intermediate” filament (fi) synthesis. Such filaments, approximately 10-12 nm in diameter, are longer than contractile filaments, and are unaligned; scattered polyribosomes, free ribosomes and microtubules (mt) also typify such regions. Go, Golgi apparatus; nu, nucleus. X 30,000.
238
DIFFERENTIATION OF HUMAN FETAL SKELETAL MUSCLE Robert J . Tomanek and Ann-Sofi Colling-Saltin
PLATE 3
239
PLATE 4 EXPLANATION OF FIGURES
13 Early formation of Z bodies (*I. Such structures appear to be developed in conjunction with unaligned thin filaments. Relatively parallel thick filaments (arrowheads) are seen adjacent to these areas and appear to forecast sarcomere formation. X
30,000.
14 Filament assembly and sarcomere formation (vastus lateralis, 18-week-old fetus). The most peripheral region (1)of this myotube contains chiefly thin filaments, occasional Z bodies (*) and a few parallel thick filaments. Adjacent to this area (2) lies a focus of numerous thick filaments and their ribosomes. Isolated filaments are less evident but can be seen in the vicinity of a small Z body (arrow). Some thin filaments appear assembled in parallel to thick filaments. Alignment of thin and thick filaments and the appearance of a Z line (Z) are illustrated in a deeper cell region (3). Note the I (I) and A (A) bands. X 30,000. 15 Longitudinal growth (rectus abdominis, neonate). This micrograph shows a region distal to sarcomeres and illustrates the presence of disoriented "intermediate" filaments (fi) and microtubules (mt) typical of the peripheral and perinuclear regions of fetal specimens. Numerous sarcolemmal and subsarcolemmal vesicles (ve) are evident. mi, mitochondria. X 30,000.
240
DIFFERENTIATION OF HUMAN FETAL SKELETAL MUSCLE Robert J . Tomanek and Ann-Sofi Colling-Saltin
PLATE 4
241
PLATE 5 EXPLANATION OF FIGURES
16 Development of T-system (vastus lateralis, 18-week-old fetus). Two T-tubules (TI are seen a t the plasma membrane; associated with each tubule are dense structures which appear to be terminal cisternae of the sarcoplasmic reticulum. Such tubules are limited to the peripheral cell regions containing contractile filaments. Vesicular or tubular structures (*I lacking associated terminal cisternae are also seen. X 30,000. 17 Extension of a T-tubule (TI from the plasma membrane (arrowheads) of a soleus myotube (18-weekold fetus). Terminal cisternae (tc) are seen forming triads or diads. Branching of T-tubules is common. X 30,000. 18 Bristle-coated vesicles (arrowheads) from the vastus lateralis (28-week-old fetus). This micrograph illustrates an apparent fusion of an undifferentiated (satellite) cell (lower half of field) with a myotube. The apposed plasma membranes of the two cells are seen (between arrows). This type of vesicle is found throughout fetal devevlopment; they are most common in myotubes and early myofibers but also occur in satellite cells. X 30,000. 19 Bristle-coated vesicles (*) as components of a tubular (tb) system (soleus, 18-week-old fetus). The vesicles, as illustrated here, assume the position of terminal cisternae. X 96,000. 20 Sarcoplasmic reticulum (arrows) in a myotube (vastus lateralis, 7-month-fetus).A focus of assembling filaments with ribosomes often includes tubular structures even before the estalishment of sarcomeres. X 30,000. 21 Tubule (tb) with bristle-coated vesicle (arrowhead) from the same specimen a s figure 20. In the later stages of fetal life these structures are seen in deeper portions of the cell and may lie in the proximity of a sarcomere as shown here. Triads (*) are increasingly common a t this stage of development. x 30,000. 22 Sarcoplasmic reticulum (rectus abdominis, 7-month-old fetus). In conjunction with sarcomere development, elements of the sarcoplasmic reticulum (arrowheads) frequently are closely associated with Z lines. This association is not evident in early fetal life (13-18). X 30,000.
242
DIFFERENTIATION OF HUMAN FETAL SKELETAL MUSCLE Robert J. Tomanek and Ann-Sofi CollingSaltin
PLATE 5
243
PLATE 6 EXPLANATION OF FIGURES
23 Organelle development within a myotube (soleus, 18-week-old fetus). Filament alignment is evident in the cell a t the top of the illustration, but polarization of forming myofibrils is not uniform. Numerous polyribosomes as well as a profile of rough endoplasmic reticulum (er) are seen. Most membranous structures lie near the plasma membrane; some resemble sarcoplasmic reticulum tubules (tb). Mitochondria (mi) a r e sparse and are limited in structural detail. X 20,000. 24
Double-membraned vesicular structure (rectus abdominis, 1-month-old infant). Such a structure (arrow) lies in a perinuclear region and contains glycogen; i t may be a n early autophagic vacuole. X 12,000.
25 Glycogen-containing vacuole (soleus, 30- to 32-week-old fetus). Note t h a t the glycogen is limited primarily to the vacuole, suggesting t h a t the vacuole is autophagic. X 30,000. 26 A muscle cell in the later myotube stage (vastus lateralis, 19-week-old fetus). As myofibrils become demarcated and attain a sufficient diameter, ribosomes become sparse, but other organelles are seen more frequently: small mitochondria (mi) with ill-defined cristae, subsarcolemmal vesicles (ve), sarcoplasmic reticulum (sr), rough endoplasmic reticulum (er) and bristle-coated vesicles (arrowhead). A structure resembling a multivesicular body is seen (*). X 39,000.
244
DIFFERENTIATION OF HUMAN FETAL SKELETAL MUSCLE Robert J. Tomanek and Ann-Sofi Colling-Saltin
PLATE 6
