Since my happy days as a graduate student at the University of Wisconsin under the supervision of Bob Cassens, I have been convinced that the manipulation of prenatal development and innervation of muscle offers some tremendous opportunities for the long-term improvement of meat yield and quality (Ph.D. Thesis, 1971; Development and Innervation of Muscle Subject to Selective Breeding). After a decade or so of descriptive histology and histochemistry, however, it became equally obvious that the tools required for manipulation primarily would be biochemical and genetic. With a poor grasp of these subjects, I decided to leave the task to others better qualified than myself and to redirect my efforts elsewhere (developing methods for the on-line evaluation of meat quality . However, I have enjoyed keeping pace as a general reader with the advances made in prenatal muscle development and innervation, which I will attempt to introduce below (details and references are available).


In normal development, the prenatal origin of striated skeletal muscles is from myoblasts. The cells that give rise to myoblasts may be called premyoblasts or presumptive myoblasts. The origin of premyoblasts is rather difficult to determine since premyoblasts are difficult to distinguish morphologically from other types of stem cells destined to give rise to other types of tissues.

This diagram shows a greatly simplified plan of a transverse section through an embryo. The bottom of the figure, in particular, bears little or no resemblence to either mammalian or avian embryos but is included to indicate the general relationships of the somites to the endoderm of the future gut (11), to the coelom or body cavity (12) and to the parietal lateral plate mesoderm or future ventral body wall (9) that develops later. Other key landmarks are the hollow dorsal nerve cord (5), notochord (1), and neural crest (4). Each somite is composed of three zones around the myocoele cavity:

In bovine embryos, the cells of the dermatome and myotome differ in their arrangement so that the myocoele cavity may be merely an artifact produced by the histological processing of embryos.


Limb bud formation is regulated by a dialogue between ectoderm and mesoderm:
  The initial amount of premuscular mesoderm may be controlled by ectodermal factors.

The mesoderm of the limb bud separates into muscle-forming and cartilage-forming regions. The chondrogenic regions that give rise to cartilage can be detected because they synthesize extracellular proteoglycan. Molecular differentiation is preceded by differential vascularization so that vascular growth also might be involved in establishing metabolic gradients in the developing limb.

Very little is known about the factors that regulate the quantitative distribution of premuscle mesoderm, and nothing is known about the role of such factors in the regulation of the postnatal potential for muscle growth in meat animals. For example, the maximum postnatal number of muscle fibers might be predetermined by the number of stem cells and by the number of times that their offspring divide. At present it appears likely that the numbers of mitotic divisions in a cell lineage might be genetically programmed, but with environmental factors controlling the rate of division.

Cell lines can also be traced back to the early development of the somites. In poultry, for example, each somite makes specific contributions to the development of particular muscles. The point at which multipotential precursor cells differentiate to become myogenic stem cells is genetically regulated by a single gene called myd. In other words, is myd a primary control point for genetically-regulated muscle mass in meat animals?

The cleavage of premuscle mesoderm in the limb buds of embryonic chicks has been examined to establish muscle homologies as a basis for comparative anatomy. The evolutionary origin of tetrapod limb muscles from the dorsal and ventral muscle masses of ancestral fish fins is reflected in a primary cleavage of premuscle limb bud mesoderm into dorsal and ventral masses. Dorsal and ventral masses undergo a further sequence of cleavages (2 in the lizard and 5 in mice) to form the individual muscles found postnatally. These events are completed quite early in development; by 13.5 days in mice, by 8 days in chicks, and by 17 mm crown-rump length (approximately 27 days gestation) in pigs. Cleavages to form future muscles may depend on the presence of the skeleton, although osteofascial compartments such as that of the avian supracoracoideus muscle may form in the absence of any muscle tissue.


The mesodermal cells of somites and limb buds undergo frequent mitosis, with a variety of factors such as IGF-I and PDGF being mitogenic. The peak of mitotic activity in the limb buds of the chick embryo is at about 5 days incubation. Dividing cells are rounded in shape and are locked into a mitotic cycle. The escape from this cycle, when a stem cell becomes a postmitotic myoblast, appears to be irreversible. The cycle preceding a cell's escape has been termed the quantal divisionThe number of times that a clone of cells remains locked into the mitotic cycle might have a profound importance on myoblast numbers: just one extra cycle by all cells might double the number of myoblasts and give rise to extra muscle fibers (hyperplasia). The population of premyoblasts capable of mitosis may not be completely homogeneous since it might contain true stem cells and committed precursors. A committed precursor is a cell that may give rise to a cohort of 16 terminally differentiated muscle cells. Obviously, factors that may regulate myoblast proliferation, such as triiodothyronine are extremely important to the meat industry.

Another way of looking at this system of cell proliferation is to consider cells at the escape point in their mitotic cycle. Both the daughter cells produced by mitosis may stay in the cycle, both may escape to become myoblasts, or one may stay in and one may escape. With a population of cells, the percentage of escaping cells starts at 0% in very young embryos, before the appearance of any myoblasts, and then increases towards, but never reaches 100% (some stem cells remain as satellite cells, a source of muscle nuclei during growth and regeneration). Cell populations containing mixtures of premyoblast stem cells, mononucleate myoblasts and fused myoblasts can be sorted with arabinocytidine. This prevents the formation of new myoblasts but does allow cell fusion. In cultures from 11-day chick embryos, about 20% of cells are myoblasts, but the percentage is lower in younger embryos. Another way of sorting cells is to determine what percentage may be cloned to give rise to myoblasts capable of fusion. Chick leg bud mesoderm at 72 hours incubation contains 0%, at 80 hours it contains 10%, and at 6 days it reaches 60%. In human limb buds, comparable values are 14% at 36 days, with a 90% plateau from 100 to 172 days.

Another factor controlling cell proliferation might be the duration of the mitotic cycle, possibly by a variation of the duration of G1 . Cells that have escaped from the mitotic cycle to become myoblasts eventually fuse together, but the fusion of cells eventually becomes less frequent, as if inhibited. Alternatively, escape from the mitotic cycles may be in late G1. Cells in G1 may respond to PROSTAGLANDIN E1 with a transient increase in intracellular cyclic AMP. This may activate protein kinase and the onset of myoblast fusion. As discussed later, the nervous system exerts some regulation over muscle development, and its control over myoblast proliferation is probably achieved by varying the duration of G1 rather than G2. Because of the importance of G1 in the regulation of cell numbers, it is interesting to note that the G1 -S boundary is the point at which the cell synthesizes calmodulin. Calmodulin is a protein that binds calcium ions, and which is thought to be involved together with cyclic AMP in the regulation of many aspects of cell metabolism, growth and division.


The morphological features of postmitotic cells and of myoblasts prior to fusion are not unlike those of other types of precursor cells in the embryo. RNA synthesis dominates cell activity and results in a large ovoid nucleus, prominent nucleoli (which vary in number between species), diffuse chromatin and many ribosomes. Myoblasts are bipolar spindle-shaped cells, whereas fibroblasts tend to be triangular in shape. Myoblasts may form tight junctions where they are in contact with each other, usually at the tips of their elongated cytoplasmic extensions. Myoblasts may be categorized into three types or cellular isoforms.

In the chick embryo, there are three different cell lineages of myoblasts early in myogenesis (with fast-contracting myosin, with mixed fast and slow, and with slow) that are independent of any innervation. The non-neural cues that initiate myogenesis appear to originate externally to the future muscle and are, in some way, related to position within the embryo. Later in development there are myoblasts whose developmental fate is determined by the nervous system.

The overall sequence of events in myogenesis may be separated into commitment, differentiation and maturation.

Myogenin and MyoD are genes in a family that is activated at commitment to a myogenic lineage and could be very useful in exploring the factors that determine muscle size in meat animals. Myogenin and MyoD are sensitive to thyroid hormones, as well as being regulated by muscle electrical activity, possibly via a mechanism dependent on cyclic-AMP. Innervation controls the abundance of myogenic factors such as MyoD1 and myogenin, and denervated muscle reverts to a neonatal phenotype. Subject to neural regulation, MyoD is prevalent in fast muscles, and myogenin in slow muscles.

As attractive as direct genetic regulation of muscle fiber numbers may be to meat scientists, it is important not to overlook other possibilities. Transforming growth factor beta 1 (TGF-þ1) is a small peptide involved the joint develop of muscle fibers and connective tissues. Following local induction of TGF-þ1, it may produce local gradients that enhance the development of connective tissues by fibroblasts, but inhibit myogenesis. Thus, a reduction of TGF-þ1 gradients might produce a condition similar to that found in double-muscled cattle.

Myoblast fusion

Myoblasts fuse with each other to form multinucleate cells that give rise to multinucleate skeletal muscle fibers. Fusion is initiated at a single site between two myoblasts. The pore formed to link adjacent cells enlarges and leaves no trace of the intervening membranes. The cytoskeleton within myoblasts forms a dense meshwork under the cell membrane and undergoes extensive remodelling at the time that myoblasts fuse to form myotubes. Myofibrils are formed rapidly once fusion has occurred and they accumulate under the cell membrane. The nuclei become restricted to an axial core of sarcoplasm surrounded by myofibrils arranged to form a hollow cylinder or tube. At this stage, the whole cellular structure may be called a myotube because its structure is dominated by the hollow cylinder of myofibrils. Myoblasts do not normally fuse with other types of cells but, experimentally, myoblasts of one species can be induced to fuse with myoblasts of another species.

Myoblast fusion has been observed by time-lapse cinematography, where some myoblasts may be seen to move to suitable positions prior to fusion while, in other cases, this may be unnecessary because aggregates of dividing cells have kept in contact with fused cells. Fusion is preceeded by a period of cell to cell recognition in which the cells may still be dispersed chemically with EDTA. Recognition is followed by a period of adhesion in which trypsin must be added experimentally in order to disperse the cells. Finally, after membrane fusion, fused cells cannot be dispersed. Cultured myoblasts fuse when their numbers reach a certain density, perhaps in response to a chemical signal. Within the myoblast, an increase in the level of cyclic AMP initiates the events that lead to fusion. Myoblasts have surface antigens that are probably involved in cell-cell recognition. Myoblast fusion is triggered by calcium ions but is inhibited by magnesium and potassium ions.

Formation of myofibrils

The synthesis of all the major proteins of the myofibril is simultaneous but a number of different sized filaments that may occur in myogenic cells. The myosin and actin of developing myofibrils appear as 15 to 16 nm diameter and 5 to 6 nm diameter filaments, respectively. The diameter of unincorporated myosin filaments is similar to those that have already joined a myofibril. The filaments of the Z line are 5 to 6 nm in diameter. Filaments with diameters of 5 to 6 nm also occur below the cell membrane, but are probably non-muscle actin. Microtubules, with diameters from 22 to 25 nm, may be found in the axial core of myotubes. Microtubule subunits with a diameter of 10 microns may also be found. Filaments with diameters from 5 to 10 nm may be found at the myotendon junction.

There are numerous possibilities for the method of assembly of myofibrils, and I keep an open mind on which is the most likely.

Many of the earlier studies on the early formation of sarcomeres must be reconsidered to take into account the respective contributions of alpha actinin, desmin and vimentin. At present, it appears that the regular structure and arrangement of sarcomeres proceeds in two stages. The initial formation of Z lines containing alpha actinin is dependent on a process of trial and error to establish alignment and is followed by the appearance of desmin and vimentin at the time that the lateral alignment of Z lines is established.


The first myotubes formed in each embryonic muscle are involved in establishing the future arrangement of muscle fibers, as well as in establishing the approximate size and anatomical location of the muscle. Little is known about any of these three factors in meat animals.

Early myotubes

 The major nerve trunks grow into a limb bud by following the connective tissue framework of the bud, but developing muscles may be necessary to invoke the formation of side branches to the muscle. Muscle fibers themselves may not be absolutely essential since the formation of nerves can be induced by the general growth pattern of somatopleural derivatives in the absence of muscle. Muscles with more than about ten myotubes might be able to initiate the formation of a nerve branch to the muscle.

Muscles may be attached to either the shaft (diaphysis) or the knob (epiphysis) of a bone. But the longitudinal growth of bones occurs at cartilagenous epiphyseal plates, and one of these plates is located between each epiphysis and its diaphysis. Thus, to retain their positions relative to each other during epiphyseal plate growth, some muscle attachments must migrate over the bone surface. Muscle migrations are regulated by the bone and traction by the periosteum is responsible for the migration of tendon insertions. Muscle development in the limbs of fetal pigs may be shaped by a dynamic interaction between linear skeletal growth and the resistance of muscles to stretching. The nervous system appears to have no direct part in the determination of myotube alignment.

If muscle stretching really does shape muscle growth, the determination of muscle fiber arrangment might be explained by the contact guidance theory that attempts to explain how nerve cells invade developing tissues. Myotubes and myoblasts might be guided by a matrix of very fine connective tissue fibers. Migrating myogenic stem cells in chick embryos branch into filopodia at their leading edges, and stem cells follow the alignment of fine connective tissue fibers. The ends of myotubes actively grow through the tissue of the future muscle and have a well developed cytoskeleton dominated by microtubules.

Molecules of fibronectin have binding sites for a number of the components that surround cells (such as for collagen and glycosaminoglycans) but also they can bind to the surfaces of cells. Thus, matrices of fibronectin may be involved in the guiding of cell migrations and the determination of muscle architecture. The initial arrangement of connective tissue fibers probably is organized by fibroblasts that are able to use their intracellular microfilaments to exert a force on the surrounding extracellular matrix. It is likely that a number of other substances also are involved in the determination of cellular arrangement in developing muscles, such as glycoprotein complexes liberated by fibroblasts.

In vitro, myoblasts only develop a parallel alignment if they are cultured on a type of collagen that forms distinct collagen fibers. Myotubes also may be pulled into alignment by their already anchored ends to follow the dominant directions of a stretched matrix, and the chances of myoblast fusion are increased when myoblasts become aligned on parallel collagen fibers.

The angular arrangement of fibers is more difficult to explain. Perhaps the tensile forces that shape the connective tissue matrix of a pennate muscle are transmitted by intramuscular tendons. Another possibility is that myoblast arrangement is influenced by the orientation of electrical fields. Cultured myoblasts become arranged with their long axes perpendicular to electric fields of 36 to 170 mV/cm.

Intracellularly, the parallel arrangement of myofibrils is dependent on the proper attachment of the whole cell. New aggregates of thick and thin filaments appear first at the periphery of cells so that the longitudinal orientation of filaments may follow the direction of membrane stretching.

Many of the early histologists who studied myogenesis were impressed by the widespread evidence of cellular degeneration (retrograde metamorphosis) that they found in developing muscles. Lysosomes capable of causing degeneration are well developed even in myoblasts. More recently, there has been a trend to dismiss degenerative phenomena during myogensis as being a mere consequence of localized myotube contracture (a sustained and destructive contraction). An alternative viewpoint is to regard contracture followed by degeneration as a means of eliminating fibers that have failed to align themselves correctly within a muscle. If cultured intercostal muscles are maintained between pieces of rib, when the muscles are stretched by slow separation of the ribs, muscle fibers continue to develop but, when they are not stetched, the fibers degenerate. The passive stretching of myotubes activates the sodium ion pump of their membranes, and this is followed by increases in amino acid uptake and protein synthesis. In vitro, muscle fibers maintained in a relaxed state by tetrodotoxin exhibit a normal accumulation of vimentin and desmin, but do not accumulate contractile myofibrillar proteins. The stimulation of amino acid transport and protein synthesis induced by the stretching of myotubes acts through some mechanism that is intrinsic to the myotube and which does not rely on circulating hormonal factors. And mechanical stimulation may act on protein synthesis and muscle growth by the release of second messengers such as arachidonic acid, diacylglycerol and prostaglandins. Long-term cultures of muscle that become connected to their substrate and are able to contract then are able to develop structures that resemble myotendinous junctions, as well as endo-, peri- and epimysial layers of connective tissue.

During the determination of fiber arrangement, a loss of cells by retrograde metamorphosis may affect muscle size, as in the development of the extrinsic ocular muscles that rotate the eyeball. Transplantation of developing eyes from a large-eyed amphibian species to a small-eyed species results in an enlargement of the host's extrinsic ocular muscles. The cause of muscle enlargement is an increase in fiber numbers, hyperplasia.


In the descriptions of myogenesis given by many histology textbooks, we may read that all skeletal muscle fibers of newborn animals are derived from myotubes by the radial migration of nuclei and a disruption of the tubular arrangement of myofibrils. In pigs, however, the great histologist Theodore Schwann (way back in 1839!) discovered that two types of muscle fiber precursors exist in the fetus: one type has a tubular arrangement of its myofibrils and may be called a myotube, while the other type lacks a tubular appearance and cannot reasonably be called a myotube. As yet, there is no generally agreed name for this second type of muscle fiber precursor. Here it is called a secondary fetal muscle fiber, shortened to secondary fiber.

Secondary fibers.

The classical myotube may then be called a primary fiber or primary myotube. Unfortunately, this is the reverse of Schwann's original terminology since he worked backwards from late to early embryos. The duality of muscle fiber precursors was almost completely ignored until recently, now there are gangs of people rushing around who claim the discovery.

The myoblasts that contribute to the formation of secondary fibers are derived from a special cell lineage distinguished by a developmental dependency on their innervation.

There is no general agreement yet on the histological significance of secondary fibers. The viewpoint taken here is that secondary fiber formation is a process for the rapid mass production of relatively large numbers of new muscle fibers, taking advantage of two factors; firstly, that the general features of muscle architecture have already been established by the arrangement of primary myotubes and, secondly, that the surfaces of gently contracting myotubes provide an ideal site for myoblast contact and fusion. Myoblasts are commonly found clinging to myotubes. As myotubes contract, these surface myoblasts probably bump against each other. The dimensional changes that occur on the myotube surface, as it decreases in length and increases in radius, favor myoblast contact along the length of the myotube. Thus, the long axis of fused myoblasts will follow that of their supporting myotube. With an in vitro model created by stretching the substrate of cultured avian myoblasts, the optimal rate of stretching for maximum myoblast alignment is 0.2 mm/hour.

Strings of fused myoblasts that have been assembled on a primary myotube are soon reinforced continuously along their length by new myofibrils. New myofibrils are grouped as a solid core, mostly located away from the supporting myotube. Most secondary fibers formed in this way retain a more or less axial core of myofibrils rather than having a tubular arrangement of their myofibrils. Secondary fibers adhere to their primary myotubes by means of pseudopodial processes that project into invaginations on the surface of the primary myotube. Further contractions of the supporting myotube (as indicated by their short sarcomeres), do not spread to secondary fibers (as indicated by their long sarcomeres). Once a secondary fiber has acquired a substantial core of fibrils, differences in length due to contraction may create a shear force between the secondary fiber and its supporting myotube that leads to the separation of secondary fibers from their supporting myotubes, and accounts for the fact that secondary fibers are often sinuously folded when their myotubes are contracted. Sinuous secondary splitting off its primary.

The morphological categorization of prenatal muscle fibers into primary myotubes and secondary fibers is a general principle which, at best, can only account for a majority of cases. At worst, it does not take long to find a few fibers with features that are intermediate between those of primary and secondary fetal fibers. Schwann in 1839 was the first to notice these transitional cases. Why should early fibers have a tubular structure and later fibers be different? Early in muscle development, strings of myoblasts start to produce filaments below their plasma membrane. If sustained by tension from successful terminal attachments, the continuous peripheral formation of new filaments gives rise to a complete tube of myofibrils below the plasma membrane. This radial symmetry provides the maximum surface area for mass production of secondary fibers. As secondary fibers start to acquire fibrils that do not contract synchronously with those of the supporting myotube, shear forces may develop between supporting myotubes and secondary fibers and further proliferation of fibrils in the secondary fibers may lead to their separation.

Now we can understand the smaller fibers in this image we saw before. 

Once a secondary fiber has separated, its core of fibrils may become crescent-shaped in transverse section as the secondary fiber starts to utilize the free space below the membrane originally apposed to the supporting myotube. The oldest secondary fibers, which are now pushed away from their parent myotubes by younger secondary fibers, start to assume a tubular structure. Perhaps now they may support production of further secondary fibers themselves, but secondary fiber production slows down as fetal development nears completion. The axial nuclei of myotubes move, or are pushed to a peripheral position below the cell membrane, and the morphological distinction between primary myotubes and secondary fibers is obscured.

The radial dimensions of primary and secondary fibers are difficult to measure histologically. Primary myotubes may appear to become smaller during postnatal development, while secondary fibers may remain constant in size until just before birth. However, the decrease in mean size of primary myotubes may be related to their less frequent contraction or to the detachment of secondary fibers. Similarly, the mean size of secondary fibers may be biased by the constant addition of newly formed fibers with a small diameter. However, just before before, the radial growth is secondary fibers is unmistakable.

It is difficult to estimate the numbers of secondary fetal fibers formed by mass production. The fibers that can be counted in a muscle cross section (apparent number) comprise only a fraction of those present in the whole muscle (real number) so that the apparent numbers of primary myotubes and secondary fibers only indicate the ratio of primary myotubes to secondary fibers. Even the interpretation of this ratio rests on the unproved assumption that both types of fibers maintain equal or constantly proportional lengths. Even the apparent ratio of primary myotubes to secondary fibers is difficult to determine towards the end of prenatal development because a few of the oldest secondary fibers may develop a tubular structure at the same time that primary myotubes start to lose their tubular structure.

The ratio of tubular to nontubular fibers describes a curve that starts with all tubular fibers and ends with all nontubular fibers. From this curve, the start of secondary fiber production may be estimated at approximately 55 days gestation in pigs. The end of secondary fiber production is more difficult to pinpoint: subjective estimates range from 70 days, through 85 to 95 days, to 100 days, depending on whether secondary fibers are counted while they are still on their supporting myotubes or when they are first detached. These estimates are based on studies using three different methods of embedding (thin epon sections, frozen sections and paraffin sections, respectively) - another factor that may have influenced the estimates.

One way to interpret the curve of the ratio of tubular to nontubular fibers is to regard the rapid change in ratio from 55 to 70 days as a consequence of secondary fiber mass production and the slow change in ratio from 70 to 110 days as a consequence of nuclear migration. Thus, at 70 days in porcine muscle, it appears that each primary myotube has supported the production of approximately five secondary fibers.

Differences between genetically obese and genetically lean pigs become apparent at 80 days gestation for muscle DNA in the semitendinosus and at 90 days onwards for muscle weight. In other words, differences are detected soon after the mass production of secondary fibers is complete. Secondary fiber formation is reduced in runt piglets. In rats, where the effects of a restricted maternal nutrition during gestation and lactation are more easily induced than in farm animals, restriction causes a reduction in the numbers of secondary fibers but not of primary myotubes.

In fetal calves, secondary fiber formation is completed by 20 cm crown-rump length, at about 205 days gestation. In chick embryos, primary myotubes are all formed by approximately 11 days and the mass production of secondary fibers occurs up to 16 days of incubation.

The logical place to go from here is to muscle fiber histochemsitry. The link is:

primary myotubes ---> red fibers

secondary myotubes ---> white fibers

although there is plenty of plasticity so muscles can adapt functionally.