The expression of myosin isoforms was studied during development of calf muscles in foetal and neonatal rats, using monoclonal antibodies against slow, embryonic and neonatal isoforms of myosin heavy chain (MHC). Primary myotubes had appeared in all prospective rat calf muscles by embryonic day 16 (E16). On both E16 and E17, primary myotubes in all muscles with the exception of soleus stained for slow, embryonic and neonatal MHC isoforms; soleus did not express neonatal MHC. In earlier stages of muscle formation staining for the neonatal isoform was absent or faint. Secondary myotubes were present in all muscles by E18, and these stained for both embryonic and neonatal MHCs, but not slow. In mixed muscles, primary myotubes destined to differentiate into fast muscle fibres began to lose expression of slow MHC, and primary myotubes destined to become slow muscle fibres began to lose expression of neonatal MHC. This pattern was further accentuated by E19, when many primary myotubes stained for only one of these two Isoforms. Chronic paralysis or denervation from E15 or earlier did not disrupt the normal sequence of maturation of primary myotubes up until E18, but secondary myotubes did not form. By E19, however, most primary myotubes in aneural or paralyzed tibialis anterior muscles had lost expression of slow MHC and expressed only embryonic and neonatal MHCs. Similar changes occurred in other muscles, except for soleus which never expressed neonatal MHC, as in controls. Paralysis or denervation commencing later than E15 did not have these effects, even though it was initiated well before the period of change in expression of MHC isoforms. In this case, some secondary myotubes appeared in treated muscles. Paralysis initiated on E15, followed by recovery 2 days later so that animals were motile during the period of change in expression of MHC isoforms, was as effective as full paralysis. These experiments define a critical period (E15–17) during which foetuses must be active if slow muscle fibres are to differentiate during E19–20. We suggest that changes in expression of MHC isoforms in primary myotubes depend on different populations of myoblasts fusing with the myotubes, and that the normal sequence of appearance of these myoblasts has a stage-dependent reliance on active innervation of foetal muscles. A critical period of nerve-dependence for these myoblasts occurs several days before their actions can be noted.

Adult muscles contain a mixture of muscle fibre types whose properties vary in several dimensions: from fast to slow contracting; fatiguable to fatigue-resistant; and anaerobic to oxidative in metabolism. There is recent evidence favouring the idea that the existence of different classes of adult muscle fibres reflects their origins from different populations of myoblasts (Miller and Stockdale, 1987; Hoh et al. 1988; reviewed by Stockdale and Miller, 1987) or satellite cells (Maier et al. 1986; Hoh and Hughes, 1988). Other, transitory, fibre type differences arise during development as stage-dependent changes in expression of several families of muscle-specific proteins, for example the sequential expression of embryonic, neonatal and adult myosin heavy chain (MHC) isoforms (Hoh and Yeoh, 1979; Whalen et al. 1981; Fitzsimons and Hoh, 1981). Furthermore, in the adult, muscles in different individuals may contain different fibre types according to their history of use (Brown et al. 1983; reviewed by Jolesz and Sréter, 1981 and Pette and Vrbova, 1985).

Muscle fibres arise from myotubes, which can be categorized into two major groups, primary and secondary (Couteaux, 1937; Kelly and Zacks, 1969) which form sequentially. Myotubes, in turn, appear to arise from the fusion of at least two separate populations of myoblasts; an early population that forms the primary myotubes, and a later-developing nerve-dependent population which gives rise to secondary myotubes (Bonner, 1978, 1980). The onset of secondary myotube formation can usefully be considered to mark the boundary between embryonic and foetal phases of development. Recently, Dhoot (1986) and Narusawa et al. (1987) have shown that all primary myotubes in rat muscles express the slow as well as the embryonic MHC isoform from their earliest stages of formation. Later, some lose expression of the slow isoform while continuing to express embryonic MHC; these myotubes later differentiate into muscle fibres expressing a fast MHC isoform. Secondary myotubes, by contrast, first express the embryonic MHC isoform and generally go on to become fast muscle fibres, but some may later express slow MHC and differentiate into slow muscle fibres in the adult.

The extent to which muscle fibre properties are determined by the particular motoneuron that innervates them is a complex question. In a series of seminal experiments, Buller et al. (1960) cross-reinnervated soleus (SOL) and extensor digitorum longus (EDL) muscles in young kittens, and found that the contractile speeds of these muscles were reversed, EDL becoming slower and SOL faster contracting. These observations led to speculations that nerves might determine muscle fibre properties during embryonic and foetal development. Salmons and Sreter (1976) implanted stimulating electrodes to vary the pattern of activation of muscle fibres while leaving their innervation intact, and demonstrated that the pattern of use of a muscle fibre can modulate its properties without the need to postulate a separate ‘trophic’ mechanism. Their results, however, do not prove that innervation is responsible for the initial determination of muscle fibre type. Indeed, although all the fibres within a single motor unit form a more coherent population than fibres in a whole muscle (Nemeth et al. 1981; Gauthier et al. 1983) they still are not homogeneous (Martin et al. 1988), making it unlikely that innervation alone determines muscle fibre type, even in the adult.

Experiments on the early embryonic stages of muscle development have shown that different primary myotube types appear with normal numbers and distribution even in the absence of innervation (Butler et al. 1982; Laing and Lamb, 1983; Phillips and Bennett, 1984). Later in development, groups of muscle fibres with similar characteristics are collected into individual motor units. All primary myotubes in the rat IVth lumbrical muscle, for example, develop into adult slow muscle fibres, and all constitute a single motor unit (Jones et al. 1987). Motoneurons exclusively contact primary myotubes during the motoneuron cell death period which takes place before the onset of secondary myotube formation (Ross et al. 1987a). In consequence of this, during foetal development there must exist a process of active sorting of nerve–muscle connections, with individual motoneurons recognising their appropriate end-organ rather than determining the properties of each muscle fibre.

Maier et al. (1986) studied the process of transformation of muscle fibre type caused by chronic low-frequency electrical stimulation in adult animals, and found that in a proportion of cases stimulated fibres degenerated, and were rapidly replaced by new fibres with different properties. Hence the effect of stimulus frequency was not only to transform the properties of muscle fibres, but also to induce the formation of new fibres with the appropriate properties. They suggest that adult muscles contain a heterogeneous population of satellite cells, supporting the regeneration of muscle fibres with different properties. This hypothesis has been elegantly confirmed by Hoh and Hughes (1988) who instead of crossing nerves, transplanted muscles from one bed to another. All muscle fibres in the transplanted muscles were made to degenerate, leaving only satellite cells which regenerated new muscle fibres in the transplanted muscles. The jaw-closing muscles of the cat contain an unusual ‘super-fast’ form of myosin. Jaw muscles transplanted into the bed of EDL, a fast muscle, and reinnervated by the EDL nerve, regenerated fibres containing superfast myosin, but not fast myosin. EDL motoneurons do not normally induce the production of this type of myosin. In contrast, jaw muscles transplanted to the soleus muscle bed and reinnervated by slow motoneurons initially expressed both slow and superfast myosin isoforms, but expression of the superfast MHC was not maintained. Taken together, these experiments show that appropriate innervation supports but does not determine the expression of superfast myosin, which requires the presence of particular subpopulations of myoblasts/ satellite cells.

Different muscle-specific gene products are expressed sequentially during development. DNAsel nuclease protection experiments have been used to show that muscle-specific genes are exposed only during the last mitosis prior to fusion into myotubes (Carmon et al. 1982). Once fused, nuclei can no longer undergo mitosis. Hence, changes in expression of genes within nuclei in myotubes must result either from the modulation of expression of genes already exposed in the last prefusion mitosis, or in consequence of fusion of a later, different, population of myoblast or satellite cell nuclei. In principle, these two alternatives could be distinguished, as the first possibility should see some form of cytoplasmic regulation affecting all the nuclei in the myotube, while the second should see expression of ‘late’ genes only in association with recently arrived nuclei.

In this paper, we describe the patterns of maturation of slow and fast myotubes during embryonic and foetal development of rat calf muscles, both in controls and in animals paralyzed or denervated during the period of muscle formation. In this way, we are able to determine the times when myotubes in particular muscles undergo transitions in expression from one MHC isoform to another, and can describe the extent to which these transitions are intrinsic to the sequence of muscle development, and how much they reflect the type and the activity of muscle innervation. Rat calf muscles are particularly suited to the study, as they contain a range of muscles varying from 90 % type I through muscles with a mixture of fibre types to muscles that are almost pure type II (Armstrong and Phelps, 1984). Individual muscles also have characteristic patterns of distribution of fibres with different MHC isoforms, providing a natural form of intrinsic control for experiments where the process of development is perturbed by denervation or paralysis.

The experiments were done with white Wistar rats (brief mention is made of a slight difference in muscle development in Sprague Dawley rats). Pregnancies were dated by the presence of a copulation plug (day 0). Embryos were denervated by injecting 1 μg of β-bungarotoxin (β-BTX) into individual embryos (McCaig et al. 1987), or paralyzed by implanting a slow-release capillary containing tetrodotoxin (TTX) (Mills and Bray, 1979) into the amnion. This advance on our earlier technique of intraembryonic insertion of the capillary (Harris, 1981) made it possible to paralyze embryos from as early as E14 onwards. When paralysis was to be maintained for some time, additional capillaries were inserted every 2 or 3 days.

Immunohistochemistry

Pregnant animals were killed with a blow to the head, and the foetuses removed. Each foetus was pithed, and the hindlegs cut free and immersed in isopentane cooled in liquid nitrogen and then stored at –80°C until required. Before sectioning, the legs were placed in small aluminium foil boats containing Tissue-Tek (at 0°C to prevent tissue thawing) and oriented for sectioning. The Tissue-Tek was solidified at –20°C, and 10 μm frozen sections cut. Transverse sections of the entire calf musculature were made, commencing near the ankle joint and continuing until reaching the proximal insertion of the soleus muscle near the knee. In some experiments, cross-sections of the hindfoot were taken to demonstrate the lumbrical muscles. The sections were collected as ribbons, dried onto subbed slides, fixed for 4 min in 1% paraformaldehyde in phosphate-buffered saline (PBS; 100mM-NaCl, 20mM-phosphate buffer, pH7.4), and washed several times in PBS. Ribbons were collected onto 6–10 slides, depending on age and size of the limb, placing successive ribbons on different slides in rotation, so that each slide had a series of sections taken from the full length of the limb. The primary antibodies were diluted in PBS containing 0.5 % bovine serum albumen (BSA) and 0.5% Triton X-100 detergent. Sections were incubated in primary antibody overnight at 4 °C. After washing in PBS, they were then incubated for l h at room temperature in biotinylated sheep anti-mouse (Amersham) diluted 1:250 in PBS containing 0.5% BSA, and l h at room temperature in biotin–streptavidin–HRP complex (Amersham) diluted 1:250 in PBS containing 0.5 % BSA. The final wash included 10 min in PBS at double the normal salt concentration to eliminate nonspecific binding. The HRP was revealed using diaminobenzidine (DAB) as chromagen. Sections were then cleared and permanently mounted without further staining.

The primary antibodies were applied in a range of dilutions. Antibody NOQ7.1.1A was a hybridoma supernatant kept frozen at –80 °C, and was effective in dilutions of 1:500–1:5000, while antibody MY32 (purchased from Sigma) was effective at 1:2000–1:20000. Anti-embryonic MHC monoclonal antibody 2B6 (Gambke and Rubinstein, 1984) (gift from Dr N. Rubinstein) was effective in dilutions up to 1:20000, and was used at 1:2000–1:4000. When testing for an absence of response, antibodies were used at higher concentrations, usually 1:10 for NOQ7.1.1A and 1:400 for MY32. Controls included serial dilution of the primary antibody, and absence of the primary antibody. Trials were made with products from several manufacturers to ensure that the second antibody and biotin–streptavidin–HRP system gave no background staining. Sections were examined with brightfield, phase-contrast and dark-field microscopy, using a Zeiss microscope.

An anti-desmin monoclonal antibody purchased from Amersham, used as a control to identify muscle tissue, was applied in a dilution of 1:50.

Tissue culture

L6 myoblasts were stored at –80°C and cultured when required. Shortly after fusion, L6 myotubes express only embryonic MHC (Whalen et al. 1979). An aliquot of cells was thawed and plated in 60 mm Falcon tissue culture dishes in Dulbecco’s modified Eagle’s medium (DMM) with 20% foetal calf serum. The medium was changed after 24h, and cells reached confluency 3 days later. The medium was then changed to DMM with 10% horse serum. Cells began fusing within 24h and were maintained for various times. For immunohistochemistry, cells were fixed for 10 min in 4% paraformaldehyde in 0.1 M-phosphate buffer, pH7.4; washed three times in PBS including 20 min with 0.1 M-glycine added, and incubated with 2 ml of antibody solutions using the concentrations and times as for frozen section immunohistochemistry.

Gel electrophoresis

Myosin from bulk hindlimb muscle from embryonic and neonatal animals, from individual muscles from adult rats, and from cultures of L6 myotubes, was partially purified by the method of Hoh et al. (1976) and stored at –20°C in pyrophosphate buffer (40mM-Na4P2O7, lmM-MgC12, 0.01% 2-mercaptoethanol, pH8.8), diluted 1:1 with glycerol, until required. Protein concentrations were estimated using a Pierce BCA kit. Two methods of gel electrophoresis were employed. The first was an SDS–polyacrylamide system (Danieli-Betto et al. 1986; Biral et al. 1988) modified by using midget gels (8 cm high and 0.7 mm thick) with a stacking gel of 3 % and separating gel of 5 % polyacrylamide. These gels were run in the cold room at 17 Vcm-1. The stored samples were diluted 1:1 with 3% SDS, 62.5mM-Tris, 48% glycerol, 0.5% dithiothreitol, pH6.8 and heated to 100°C for 2min. Gel loadings were 1–2 μg per lane, and gels were silver stained as required by the method of Morrissey (1981).

The second system employed pyrophosphate gels (Hoh et al. 1976; d’Albis et al. 1979), using 0.7 mm thick midget gels cooled with glycol circulating at 0°C, with the buffer recirculated. Gels were run for 250–300Vhcm-1 at 9–17Vcm-1. The gels contained 4 % polyacrylamide and were loaded with 1–2μg of protein per lane. Pyrophosphate gels were stained with silver, as for the SDS gels.

Western blots

Proteins separated on SDS–polyacrylamide gels were transferred to a nitrocellulose membrane using an LKB 2117–250 semidry blotter and a transfer buffer containing 39 mM-glycine, 48mM-Tris–HCl, 0.0375% SDS and 20% methanol, pH8.5. Transfer time was 1h at 0.8mAcm-2. A similar procedure was used for pyrophosphate gels, except that they were soaked for 1h prior to transfer in 62.5mM-Tris–HCl, 1% SDS, 1% 2-mercaptoethanol and 10% glycerol, pH 8.5 (Gambke and Rubinstein, 1984), and the transfer time was 1.5 h. The membranes were then stained in 0.3 % Ponceau S, 0.1% acetic acid, for 4min followed by washing in distilled water, to visualise the protein bands. After blocking the membrane in 5 % blotto (5 % nonfat skim milk in PBS with 0.1% Tween-20, pH7.4) the myosin isoforms were detected by incubation with an anti-MHC antibody (MY32 1:2000 to 1:20000; NOQ7.1.1A 1:20 to 1:100; 2B6 1:2000 dilution in 1% blotto) 2h at room temperature or overnight at 4 °C, followed by biotinylated sheep anti-mouse (1:2000) and then streptavidin–biotin–HRP complex (1:2000). Finally, the HRP was visualized, using DAB as a substrate. All incubations were separated by 3×15 min washes in 0.1% Tween PBS, except for the last two washes and incubation with DAB, where 150mM-NaCl, 50mM-Tris–HCl, pH7.6, was used.

Elisa

Native myosin in pyrophosphate buffer was allowed to bind overnight at 4°C to Linbro Titertek Elisa plates (1 μg per well). After blocking in 2% BSA, 0.05% Tween-20 in PBS for 2 h at 37 °C, plates were incubated for 1 h at 37 °C in serial dilutions of the primary antibody. When constructing antibody dilution curves, the anti-sarcomeric MHC antibody MF20 (Bader et al. 1982) (gift of D. Fishman) was used to confirm equality of MHC concentrations in preparations from different tissues. Binding was visualized with the Amersham biotin-streptavidin system, used according to the manufacturer’s instructions, with OPD as the chromogen.

Characterization of antibodies

Samples of myosin purified from bulk muscle from E17, E20 and 5 days postbirth (PN5) rats, from adult SOL and EDL muscles, and from recently fused cultures of L6 myotubes were run on SDS–polyacrylamide and pyrophosphate gels, and transferred to a nitrocellulose filter by Western blotting (Fig. 1). Silver staining demonstrated that SDS electrophoresis resolved the myosin heavy chains from adult muscles into at least 4 groups. Myosin heavy chains from SOL consisted mainly of the fastest moving isoform, although a faint band of the slowest migrating band was also present. In EDL, by contrast, approximately equal quantities of 2 slower migrating forms were present, with only a faint band visible of the fastest migrating form, and none of the slowest. In foetal and neonatal tissues, a strong band(s) of slower mobility was present, together with a fainter band that migrated in the same position as the fastest migrating band in adult muscle.

Fig. 1.

Separation of rat MHC isoforms by gel electrophoresis. (A) 4% pyrophosphate gels; (B, C) 5% SDS–PAGE. Gels were stained with silver (1, 4) or blotted onto nitrocellulose and marked with HRP after incubation with MAb NOQ7.1.1A (anti-slow MHC) (2), with MAb MY32 (anti-neonatal/IIa/IIb/lld MHC) (3, 6), or MAb 2B6 (anti-embryonic MHC) (5). Lanes in the gels are: (a) E17 bulk hindlimb muscle; (b) E20 bulk hindlimb muscle; (c) PN5 bulk hindlimb muscle; (d) adult EDL muscle; (e) adult SOL muscle; (f) L6 myotubes; (g) PN5 bulk hindlimb muscle.

Fig. 1.

Separation of rat MHC isoforms by gel electrophoresis. (A) 4% pyrophosphate gels; (B, C) 5% SDS–PAGE. Gels were stained with silver (1, 4) or blotted onto nitrocellulose and marked with HRP after incubation with MAb NOQ7.1.1A (anti-slow MHC) (2), with MAb MY32 (anti-neonatal/IIa/IIb/lld MHC) (3, 6), or MAb 2B6 (anti-embryonic MHC) (5). Lanes in the gels are: (a) E17 bulk hindlimb muscle; (b) E20 bulk hindlimb muscle; (c) PN5 bulk hindlimb muscle; (d) adult EDL muscle; (e) adult SOL muscle; (f) L6 myotubes; (g) PN5 bulk hindlimb muscle.

Antibody NOQ7.1.1A (similar to NOQ7.5.4D: Kelly, 1987; Draeger et al. 1987) bound to the fastest migrating band on SDS gels. This myosin isoform (type I MHC) was most strongly expressed in SOL, but faint bands were visible on the other samples, reflecting small proportions of type I muscle fibres in these muscles. There was no cross-reaction with the other myosin isoforms. This antibody did not stain myosin extracted from cultures of newly fused L6 myotubes (results not shown).

Antibody MY32 bound to the 2 MHC isoforms from adult EDL (types Ila and lIb MHC) and to a faint slowly migrating band in SOL (lId MHC; Bär and Pette, 1988). MY32 bound to the intermediate migrating band (neonatal MHC) in the foetal and neonatal muscles. This antibody did not stain myosin extracted from cultures of newly fused L6 myotubes (Fig. 1C). By comparing the silver-stained gels with the Western blots in Fig. 1B it can be seen that neither MY32 nor NOQ7.1.1A bound to the slow-migrating band representing the majority species of MHC isoform (embryonic MHC, see below) in foetal muscles.

Antibody 2B6 (anti-embryonic MHC, Gambke and Rubinstein, 1984) bound the single band of MHC seen on SDS gels from myosin extracted from newly fused L6 myotubes (Fig. 1C). On gels of myosin extracted from foetal bulk muscle (E17 or E20), it bound the majority band which ran slightly slower than the band stained by MY32 (results not shown). This isoform was still present in myosin extracted from neonatal (PN5) bulk muscle, which also contained a substantial proportion of neonatal MHC, stained by MY32 (Fig. 1C).

Native pyrophosphate gels resolved the myosin isoforms into a multitude of bands, although expression in individual samples was more limited. E17 muscle was characterized by one major and one faint band; E20 was characterized by 2 strong bands of slightly faster mobility; while in PN5 muscle 3 bands were apparent, which migrated faster than all except the fastest band present on E20. In adult EDL, at least 3 major bands were present, although separation and resolution of these bands was consistently poorer than in other species (rabbit, sheep, chicken; results not presented). These bands were of intermediate mobility, and slower migrating than the E20 and PN5 samples. In SOL, a markedly slower migrating band was apparent and was the major species, with another faint band migrating slightly above or equivalent to the slowest major band in EDL.

Antibody NOQ7.1.1A bound specifically to the slowly migrating band present in large quantities only in SOL. In contrast, MY32 bound to the faster migrating bands in all samples, but not to the majority band (embryonic MHC) in E17 and E20 samples (Fig. 1A).

Elisa antibody dilution curves for MY32, NOQ7.1.1A and 2B6 (Fig. 2) were made using myosin samples characterized by the SDS gels, pyrophosphate gels and Western blots (Fig. 1). To confirm that the protein content of the samples accurately reflected their content of MHC, dilution curves were constructed using the anti-sarcomeric MHC antibody MF20 (Fig. 2A). Scanning densitometry of serially diluted PN5 and E20 myosins run on SDS–PAGE gels also was used to demonstrate that concentrations of MHC in the samples were not significantly different (results not shown).

Fig. 2.

ELISA analysis of specificity of the monoclonal antibodies, using antibody dilution curves. (A) MF20; (B) MY32; (C) 2B6; (D) MY32; (E) N0Q.7.1A. Rat myosin samples were prepared from: (•) E20 bulk hindlimb muscle (embryonic and neonatal MHC); (▴) PN5 bulk hindlimb muscle (embryonic and neonatal MHC); (◼) adult EDL (fast MHC, types Ila and IIb); (○) adult soleus muscle (slow MHC, type I); (□) L6 myotubes (embryonic MHC). The curves in A show equality of loading of total MHC in each sample (MF20 is a pan anti-sarcomeric MHC). The curves in B show that MY32 does not bind myosin extracted from L6 myotubes (embryonic MHC), and show an increase in the ratio of neonatal/ embryonic MHC between E20 and PN5. The curves in C identify embryonic MHC as the principal component of L6 myosin. In D the minor MY32 response to SOL myosin is due to a small proportion of fast MHC (type lid) in this muscle, and in E the response of NOQ.1.1A to myosin from adult EDL and from PN5 and E20 hindlimb muscle depends on the variable content of type I MHC (see Fig. 1B).

Fig. 2.

ELISA analysis of specificity of the monoclonal antibodies, using antibody dilution curves. (A) MF20; (B) MY32; (C) 2B6; (D) MY32; (E) N0Q.7.1A. Rat myosin samples were prepared from: (•) E20 bulk hindlimb muscle (embryonic and neonatal MHC); (▴) PN5 bulk hindlimb muscle (embryonic and neonatal MHC); (◼) adult EDL (fast MHC, types Ila and IIb); (○) adult soleus muscle (slow MHC, type I); (□) L6 myotubes (embryonic MHC). The curves in A show equality of loading of total MHC in each sample (MF20 is a pan anti-sarcomeric MHC). The curves in B show that MY32 does not bind myosin extracted from L6 myotubes (embryonic MHC), and show an increase in the ratio of neonatal/ embryonic MHC between E20 and PN5. The curves in C identify embryonic MHC as the principal component of L6 myosin. In D the minor MY32 response to SOL myosin is due to a small proportion of fast MHC (type lid) in this muscle, and in E the response of NOQ.1.1A to myosin from adult EDL and from PN5 and E20 hindlimb muscle depends on the variable content of type I MHC (see Fig. 1B).

MY32 bound with a high affinity to the samples of PN5 and adult fast MHCs (Figs 2B and 2D). In contrast, binding to E20 bulk muscle myosin was about one order of magnitude weaker, reflecting the preponderance of embryonic rather than neonatal MHC present at this time. MY32 did not bind to myosin extracted from L6 myotubes (Fig. 2B), which contained principally embryonic MHC (Fig. 2C). Binding to adult soleus muscle MHC was revealed only with high concentrations of antibody (Fig. 1D), reflecting the small quantity of type lid myosin in this sample. Similar antigen dilution curves were constructed for NOQ1.1.1A (Fig. 1E), which showed the highest binding affinity for adult soleus myosin. When tested on other samples, the responses varied with their content of slow myosin, which ranged from approximately 9% of total myosin in neonatal bulk hindlimbs to less than 3% in adult EDL.

Dilution curves of antibody 2B6 were made using myosin extracted from L6 myotube cultures, and from bulk muscle from E20 and PN5 rats (Fig. 2B). This antibody reacted strongly with all three myosin preparations (Fig. 2C). The difference in binding to L6 MHC by antibodies MF20 and 2B6 (Fig. 1A, 1C) suggests that another, as yet unidentified, MHC isoform may also be expressed in these cells.

Antibody specificities were also tested with immuno-histochemistry on cultures of L6 myoblasts and myotubes (results not shown). ‘Early’ fused myotubes in cultures of L6 myoblasts (1–3 days after changing to fusion medium) bound the anti-embryonic MHC MAb 2B6, but were not stained by MY32. L6 myotubes maintained for 5–7 days after fusion also bound MY32, but more faintly than the anti-embryonic MHC MAb.

We conclude that MAb NOQ7.1.1A is specific for slow MHC (type I), while MY32 stains neonatal MHC and adult fast Ila, lib and lid MHC isoforms. We henceforth refer to MY32 as ‘anti-neonatal/f MHC’ as it is the neonatal isoform that it most likely revealed during the time-span of our experiments.

Immunohistochemistry: MHC isoforms in new-formed muscles

All myotubes at all stages examined in this project were stained by the anti-embryonic MHC monoclonal antibody 2B6, so these results are not individually presented. MAb N0Q7.1.1A (anti-slow MHC) strongly stained myotubes from young (E16) muscles. Antibody MY32 (anti-neonatal/f MHC) also stained these fibres, although it was less effective than the anti-slow MHC MAb. Muscles of this age contain only primary myotubes. As many authors have described a sequence of expression of myosin isoforms, with neonatal MHC being expressed later than embryonic MHC, we examined E15 diaphragm and E16 lumbrical muscles, where in both cases about one-half the final number of primary myotubes has formed. Myotubes in whole-mounts of E15 diaphragm reacted strongly with anti-slow MHC, and also with anti-desmin (used as a control for access of a muscle-specific antibody), but not with anti-neonatal/f MHC (Fig. 3). Similarly, cross-sections of hindlimb lumbrical muscles from E16 foetuses were stained by anti-slow but not by anti-neonatal/f MHC (Fig. 3D, E). This contrasts with the high sensitivity to staining with MY32 (anti-neonatal/f MHC) in older muscles. We conclude that, in this strain of rats, neonatal MHC is first expressed slightly later in primary myotube development than slow MHC. In another strain of rats (Sprague-Dawley), primary myotubes were stained by MY32 as soon as they formed, and we could not demonstrate any delay in the expression of neonatal MHC (results not shown).

Fig. 3.

New-formed primary myotubes react with anti-slow but not with anti-neonatal/f MHC MAbs in Wistar rats. (A) Whole-mount of E15 diaphragm muscle stained with anti-slow MHC; all fibres show a strong reaction. (B) A control experiment; primary myotubes in an E15 diaphragm show a strong positive reaction with muscle-specific anti-desmin MAb; nonmuscle tissue is not stained. (C) An E15 diaphragm shows no reactivity with anti-neonatal/f MHC MAb. (D) intrinsic muscles of the foot from an E16 hind-limb react strongly for anti-slow MHC, but not at all (E) for anti-neonatal/f MHC (dark-field views, DAB reaction product appears white). Magnification bars, 250 μm.

Fig. 3.

New-formed primary myotubes react with anti-slow but not with anti-neonatal/f MHC MAbs in Wistar rats. (A) Whole-mount of E15 diaphragm muscle stained with anti-slow MHC; all fibres show a strong reaction. (B) A control experiment; primary myotubes in an E15 diaphragm show a strong positive reaction with muscle-specific anti-desmin MAb; nonmuscle tissue is not stained. (C) An E15 diaphragm shows no reactivity with anti-neonatal/f MHC MAb. (D) intrinsic muscles of the foot from an E16 hind-limb react strongly for anti-slow MHC, but not at all (E) for anti-neonatal/f MHC (dark-field views, DAB reaction product appears white). Magnification bars, 250 μm.

Development of muscles in the lower hind-limb

Sectioned limbs from foetuses sampled at daily intervals from E16 until E20 were incubated with each of the two primary antibodies. Muscles examined included tibialis anterior, extensor digitorum longus, peroneal group, soleus and the red portion of gastrocnemius. The percentages of adult fibre types in these muscles are listed in Table 1. In presenting the results we pay particular attention to TA and SOL. TA develops into a mixed muscle with a well-defined inner red (slow) portion, and outer white (fast) portion. SOL is an almost pure red (slow) muscle.

Table 1.

Muscle fibre types in rat hindlimb muscles (from Armstrong and Phelps, 1984)

Muscle fibre types in rat hindlimb muscles (from Armstrong and Phelps, 1984)
Muscle fibre types in rat hindlimb muscles (from Armstrong and Phelps, 1984)

El6 limb muscles

All muscles stained strongly and uniformly with antislow MHC (Fig. 4A, left). In contrast, staining with anti-neonatal/f MHC was weak (Fig. 4A, right), although muscle-specific. SOL was not stained with anti-neonatal/f MHC (Fig. 4A, right), and EDL stained more faintly than TA. Neither antibody showed differences in staining between inner and outer portions of TA. Only primary myotubes are present at this time.

Fig. 4.

Development of patterns of reactivity to anti-neonatal/f and anti-slow MHC MAbs. Left column: anti-slow; right column: anti-neonatal/f. (A) Cross-section of lower leg at E16, viewed under dark-field to show DAB reaction product as white. Myotubes in all muscles react strongly with anti-slow MHC MAb, but weakly with anti-neonatal/f MHC, with EDL notably less intensely stained than TA, and SOL not stained. (B) Bright-field view of TA muscles at E18, showing differentiation into future white and red zones. Only primary myotubes are stained with anti-slow MHC MAb (left illustration), and some in the future white zone of the muscle (TAW) have lost their reactivity to this MAb, and are seen here as pale grey in colour. Both primary and secondary myotubes react with anti-neonatal/f MHC (right illustration), but some primary myotubes on the inner (future red) zone of the muscle (TAR) have lost their reactivity to this MAb and are seen here as pale grey in colour, surrounded by dark-staining secondary myotubes. Portions of EDL muscles appear in these sections. Here, all primary myotubes stain with anti-slow MHC, but some on the inner surface of these muscles have begun to lose expression of fast MHC. (C) TA and EDL muscles at E20 (left, dark-field; right, bright-field illumination). In the TA muscle, the future red (TAR) and white (TAW) zones are clearly delineated with both MAbs. In EDL, by contrast, all primary myotubes still react with anti-slow MHC and the anti-neonatal/f MAb stains only secondary myotubes, as in the inner, red zone of TA. Magnification bars, 100μm. Abbreviations, in this and subsequent figures: f, fibula; t, tibia; TAR, red portion of tibialis anterior; TAW, white portion of TA; EDL, extensor digitorum longus; PER, peroneal group of muscles; SOL, soleus.

Fig. 4.

Development of patterns of reactivity to anti-neonatal/f and anti-slow MHC MAbs. Left column: anti-slow; right column: anti-neonatal/f. (A) Cross-section of lower leg at E16, viewed under dark-field to show DAB reaction product as white. Myotubes in all muscles react strongly with anti-slow MHC MAb, but weakly with anti-neonatal/f MHC, with EDL notably less intensely stained than TA, and SOL not stained. (B) Bright-field view of TA muscles at E18, showing differentiation into future white and red zones. Only primary myotubes are stained with anti-slow MHC MAb (left illustration), and some in the future white zone of the muscle (TAW) have lost their reactivity to this MAb, and are seen here as pale grey in colour. Both primary and secondary myotubes react with anti-neonatal/f MHC (right illustration), but some primary myotubes on the inner (future red) zone of the muscle (TAR) have lost their reactivity to this MAb and are seen here as pale grey in colour, surrounded by dark-staining secondary myotubes. Portions of EDL muscles appear in these sections. Here, all primary myotubes stain with anti-slow MHC, but some on the inner surface of these muscles have begun to lose expression of fast MHC. (C) TA and EDL muscles at E20 (left, dark-field; right, bright-field illumination). In the TA muscle, the future red (TAR) and white (TAW) zones are clearly delineated with both MAbs. In EDL, by contrast, all primary myotubes still react with anti-slow MHC and the anti-neonatal/f MAb stains only secondary myotubes, as in the inner, red zone of TA. Magnification bars, 100μm. Abbreviations, in this and subsequent figures: f, fibula; t, tibia; TAR, red portion of tibialis anterior; TAW, white portion of TA; EDL, extensor digitorum longus; PER, peroneal group of muscles; SOL, soleus.

E17 limb muscles

TA and EDL stained strongly and uniformly with both anti-slow and anti-neonatal/f antibodies. SOL and the red (medial) portion of gastrocnemius (GR) stained strongly with anti-slow, but not with anti-neonatal/f antibody. The lumbrical muscles reacted strongly with both antibodies.

E18 limb muscles

The TA muscle had begun the process of differentiation into future red and white zones. Primary myotubes in the inner, axial region of the muscle stained with antislow MHC, but not with anti-neonatal/f MHC MAbs. Primary myotubes in the outer region were stained strongly by anti-neonatal/f MHC antibody, but about half had lost the strong reactivity for slow MHC they had demonstrated a day earlier (Fig. 4B). Electron microscopy (results not shown) revealed that secondary myotubes first appear in TA on E17, and on E18 they were clearly visible with light microscope immunohistochemistry (Fig. 4B). Secondary myotubes did not stain with anti-slow MHC, even in SOL, but in all muscles they stained strongly with anti-neonatal/f MHC. Within the EDL, a future fast muscle (Table 1), all primary myotubes continued to stain with both antislow and anti-neonatal/f MAbs, and all secondary myotubes stained with anti-neonatal/f MHC alone.

E19 limb muscles

TA muscles were clearly divided into inner and outer zones, and primary myotubes in these zones were stained only by anti-slow or by anti-neonatal/f MHC MAbs, respectively. Each primary myotube was surrounded by a rosette of secondary myotubes, reacting strongly with anti-neonatal/f and not at all with antislow MHC Ab.

E20 limb muscles

The days E18 through E20 were marked by a rapid increase in size of each limb muscle as the secondary myotubes, destined to give rise to more than 80% of the adult muscle fibres, appeared in each muscle. By E20, the future fast and slow zones of TA were clearly established (Figs4C, 5A, 5B). The number of slow MHC-positive fibres in this muscle (323±15, n=3) was significantly less than at E17 or E18 (585±4, n=3), showing that slow MHC-negative primary myotubes had lost expression of this isoform, rather than being late-developing pure ‘fast’ primary myotubes. These regions differed only in their primary myotube MHC isoforms; we could not discriminate types of secondary myotube with these antibodies. Some secondary myotubes in SOL, identified by the criteria of size and of relation to a primary myotube, were becoming reactive to anti-slow MHC Ab (Fig. 5C), although they still reacted strongly with anti-neonatal/f MHC (Fig. 5D). In EDL, a future fast muscle, primary myotubes continued to express both neonatal and slow MHC isoforms.

Fig. 5.

E20 hindlimb muscles stained with anti-neonatal/f or anti-slow MHC MAbs. (A) External (white) TA muscle stained with anti-neonatal/f MHC. Both primary (large diameter) and secondary (small diameter) myotubes are stained; reaction product in secondary myotubes is more dense than in primaries. (B) Internal (red) TA muscle stained with anti-neonatal/f MHC. Only secondary myotubes are stained; primary myotubes react only with anti-slow MHC (see Fig. 4C). (C) Soleus muscle stained with anti-slow MHC. Primary myotubes show a strong reaction, and most secondary myotubes in this muscle (but in no others) react weakly with this MAb. (D) Soleus muscle (right) stained with anti-neonatal/f MHC. Secondary myotubes show a strong reaction, while primary myotubes show little or no reaction. By contrast, both primary and secondary myotubes in plantaris muscle (bottom left) react strongly. Magnification bar, 50 μm.

Fig. 5.

E20 hindlimb muscles stained with anti-neonatal/f or anti-slow MHC MAbs. (A) External (white) TA muscle stained with anti-neonatal/f MHC. Both primary (large diameter) and secondary (small diameter) myotubes are stained; reaction product in secondary myotubes is more dense than in primaries. (B) Internal (red) TA muscle stained with anti-neonatal/f MHC. Only secondary myotubes are stained; primary myotubes react only with anti-slow MHC (see Fig. 4C). (C) Soleus muscle stained with anti-slow MHC. Primary myotubes show a strong reaction, and most secondary myotubes in this muscle (but in no others) react weakly with this MAb. (D) Soleus muscle (right) stained with anti-neonatal/f MHC. Secondary myotubes show a strong reaction, while primary myotubes show little or no reaction. By contrast, both primary and secondary myotubes in plantaris muscle (bottom left) react strongly. Magnification bar, 50 μm.

Role of innervation in supporting muscle development

In a series of experiments, individual embryos had all peripheral motor and sensory innervation destroyed by an injection of β-BTX (McCaig et al. 1987). The aneural foetuses were removed on E18, 19 or 20, and their muscle development compared with controls from the same litter. Denervation on E14 or E15, but not later, caused major changes in muscle development; these changes were evident by E19.

The effects of early (E14 or E15) denervation are illustrated in Figs 6 and 7. Little change in the patterns of anti-MHC reactivity was seen on E18 (Fig. 6A, B). Muscles were smaller than normal due to absence of secondary myotube formation (Harris, 1981; Ross etal. 1987b), and all primary myotubes, except in SOL and GR, stained with both anti-neonatal/f and anti-slow MHC MAbs; there was no sign of differentiation into fast or slow zones within muscles. By E19, however, the pattern had changed dramatically (Fig. 6C). Few myotubes, with the constant exception of SOL myotubes, expressed slow MHC; all others stained with anti-neonatal/f MHC alone (as in controls, all myotubes in experimental animals expressed embryonic MHC throughout this time period; results not shown). These effects were even more clearly seen by E20 (Fig. 6E, F; Fig. 7). Within TA, no more than a dozen or so primary myotubes expressed slow MHC, but the rest all stained strongly for neonatal MHC (Fig. 6E, F; Fig. 7B, C). The total number of myotubes in denervated TA muscles on E20 (561±4, n=3) was similar to the number of primary myotubes in E17 and E18 control TA muscles (585 ±4, n=3), suggesting that a change in MHC expression had occurred, rather than selective myotube death. MHC expression in the SOL muscle was less affected by denervation: primary myotubes in SOL muscle stained with anti-slow MHC, as in controls, and only a small proportion also stained with anti-neonatal/f MHC (Fig. 7A).

Fig. 6.

Effects on muscle development of denervation with β-BTX. Cross-sections of hindleg examined with dark-field microscopy; DAB reaction product (and some unstained tissues) appears white. (A) E14–18 β-BTX (i.e. denervated at E14, examined on E18), stained for slow MHC. Only primary myotubes are present, and all react strongly with this MAb. (B)E14–18 β-BTX, stained for neonatal/f MHC. Primary myotubes in all muscles except SOL stain strongly with this MAb.(C)E14–19 β-BTX, stained for slow MHC. In TA, almost no primary myotubes react with anti-slow MHC, and reactivity is reduced below normal in other muscles. Only SOL is strongly stained with this MAb. By contrast, all primary myotubes (except in SOL) react strongly with anti-neonatal/f MHC (not shown). (D) E20 control muscles, stained for anti-slow MHC. TA shows its normal delineation into red and white zones, with primary myotubes in the inner (red) zone reacting strongly with anti-slow MAb. SOL is very strongly stained because, in contrast to other muscles, secondary as well as primary myotubes are stained. (E, F) E15–20 β-BTX, stained with anti-slow (E) and anti-neonatal/f (F) MHC MAbs. Even fewer myotubes stain with anti-slow MHC than in the E14-19 limb, showing that denervation one day later has made no difference to the pattern of development of aneural muscles. TA contains only about a dozen slow MHC-positive myotubes, and only SOL is strongly stained. All primary myotubes, except in SOL, react strongly with anti-neonatal/f MHC MAb. Magnification bars 250 μm. Note the great reduction in size of aneural as compared to control muscles at E20, due to the absence of secondary myotubes. A, B, D, left hindlimb; C, E, F, right hindlimb.

Fig. 6.

Effects on muscle development of denervation with β-BTX. Cross-sections of hindleg examined with dark-field microscopy; DAB reaction product (and some unstained tissues) appears white. (A) E14–18 β-BTX (i.e. denervated at E14, examined on E18), stained for slow MHC. Only primary myotubes are present, and all react strongly with this MAb. (B)E14–18 β-BTX, stained for neonatal/f MHC. Primary myotubes in all muscles except SOL stain strongly with this MAb.(C)E14–19 β-BTX, stained for slow MHC. In TA, almost no primary myotubes react with anti-slow MHC, and reactivity is reduced below normal in other muscles. Only SOL is strongly stained with this MAb. By contrast, all primary myotubes (except in SOL) react strongly with anti-neonatal/f MHC (not shown). (D) E20 control muscles, stained for anti-slow MHC. TA shows its normal delineation into red and white zones, with primary myotubes in the inner (red) zone reacting strongly with anti-slow MAb. SOL is very strongly stained because, in contrast to other muscles, secondary as well as primary myotubes are stained. (E, F) E15–20 β-BTX, stained with anti-slow (E) and anti-neonatal/f (F) MHC MAbs. Even fewer myotubes stain with anti-slow MHC than in the E14-19 limb, showing that denervation one day later has made no difference to the pattern of development of aneural muscles. TA contains only about a dozen slow MHC-positive myotubes, and only SOL is strongly stained. All primary myotubes, except in SOL, react strongly with anti-neonatal/f MHC MAb. Magnification bars 250 μm. Note the great reduction in size of aneural as compared to control muscles at E20, due to the absence of secondary myotubes. A, B, D, left hindlimb; C, E, F, right hindlimb.

Fig. 7.

Effects on muscle development of denervation with β-BTX on E15, examined at E20. Left, stained with anti-slow MHC MAb; right, stained with anti-neonatal/f MHC MAb. (A) Soleus muscle. All myotubes react strongly with anti-slow but not with anti-neonatal/f MHC MAb. (B, C) tibialis anterior muscle with low and high magnification. Almost no myotubes stain with anti-slow, but all stain with anti-neonatal/f MHC MAb. Secondary myotube formation is much reduced in these muscles. Calibrations, 100 μm (A and C at same magnifications).

Fig. 7.

Effects on muscle development of denervation with β-BTX on E15, examined at E20. Left, stained with anti-slow MHC MAb; right, stained with anti-neonatal/f MHC MAb. (A) Soleus muscle. All myotubes react strongly with anti-slow but not with anti-neonatal/f MHC MAb. (B, C) tibialis anterior muscle with low and high magnification. Almost no myotubes stain with anti-slow, but all stain with anti-neonatal/f MHC MAb. Secondary myotube formation is much reduced in these muscles. Calibrations, 100 μm (A and C at same magnifications).

To see whether the neural support of normal muscle development is due simply to activation of muscle contraction or of muscle electrical activity, or to some other form of neurotrophic communication, perhaps chemically mediated, the effects of denervation were compared with those of chronic paralysis with TTX. This agent blocks action potentials in both nerves and muscles, but does not block the spontaneous release of acetylcholine from nerve terminals, which remain physically intact. Paralyzed muscles were indistinguishable, insofar as patterning and differentiation of muscle fibre types is concerned, from those denervated over the same period by treatment with β-BTX (Fig. 8).

Fig 8.

Effects on muscle development of chronic paralysis with TTX, in animals paralysed from E15 onwards, examined on E20. Left column: stained with anti-slow MHC MAb; right column: stained with anti-neonatal/f MHC MAb. (A) Crosssection of hind leg at midcalf level, examined under dark-field illumination. Slow MHC reactivity persists in SOL and GR muscles; almost no myotubes are stained in TA. Anti-neonatal/f MHC MAb stained all myotubes, except in SOL. (B) Soleus muscle is strongly stained with anti-slow (left) but not with anti-neonatal/f (right) MHC MAb (bright-field illumination). (C) Phase-contrast views of TA to show lack of staining with anti-slow (left) and strong staining with anti-neonatal/f (right) MHC MAbs. The section stained with anti-slow MHC includes all positive myotubes present in this muscle. Secondary myotube formation is much reduced following paralysis at this time. Calibrations: A, 250 μm; B, C, 50 μm.

Fig 8.

Effects on muscle development of chronic paralysis with TTX, in animals paralysed from E15 onwards, examined on E20. Left column: stained with anti-slow MHC MAb; right column: stained with anti-neonatal/f MHC MAb. (A) Crosssection of hind leg at midcalf level, examined under dark-field illumination. Slow MHC reactivity persists in SOL and GR muscles; almost no myotubes are stained in TA. Anti-neonatal/f MHC MAb stained all myotubes, except in SOL. (B) Soleus muscle is strongly stained with anti-slow (left) but not with anti-neonatal/f (right) MHC MAb (bright-field illumination). (C) Phase-contrast views of TA to show lack of staining with anti-slow (left) and strong staining with anti-neonatal/f (right) MHC MAbs. The section stained with anti-slow MHC includes all positive myotubes present in this muscle. Secondary myotube formation is much reduced following paralysis at this time. Calibrations: A, 250 μm; B, C, 50 μm.

Critical period for neural determination of muscle development

Denervation or paralysis were effective in changing the pattern of muscle differentiation only if they began on E15 or earlier. Denervation or paralysis commencing on E16 did not markedly change the pattern of distribution of muscle fibre types on E20, except for a reduction in the number of secondary myotubes formed. Fig. 9 illustrates myotubes in an E20 TA muscle, following denervation with β-BTX on E16. Here, primary myotubes express either slow or neonatal MHC, and secondary myotubes, stained with anti-neonatal/f MHC, are associated with some primary myotubes. The reduction in secondary myotube numbers can be judged by comparison with the E20 control TA muscle illustrated in Fig. 5A, B.

Fig. 9.

Denervation from E16 onwards does not affect the patterning of distribution of fast and slow MHC isoforms at E20. TA muscle from an E16–20 β-BTX-treated foetus, stained with anti-slow (left; phase contrast view) or anti-neonatal/f (right; bright-field optics) MHC MAbs. The sections show the transition zone between red (upper part of photos) and white (lower) zones of TA; primary myotubes in the red zone are losing reactivity to anti-neonatal/f and retaining activity to anti-slow MHC MAb, as in controls. The anti-slow MHC stained section is viewed with phase contrast to show unstained myotubes. Some secondary myotubes (smaller diameter, dark-stained with anti-neonatal/f MHC MAb) are present, in numbers reduced below normal. Magnification bar, 50 μm.

Fig. 9.

Denervation from E16 onwards does not affect the patterning of distribution of fast and slow MHC isoforms at E20. TA muscle from an E16–20 β-BTX-treated foetus, stained with anti-slow (left; phase contrast view) or anti-neonatal/f (right; bright-field optics) MHC MAbs. The sections show the transition zone between red (upper part of photos) and white (lower) zones of TA; primary myotubes in the red zone are losing reactivity to anti-neonatal/f and retaining activity to anti-slow MHC MAb, as in controls. The anti-slow MHC stained section is viewed with phase contrast to show unstained myotubes. Some secondary myotubes (smaller diameter, dark-stained with anti-neonatal/f MHC MAb) are present, in numbers reduced below normal. Magnification bar, 50 μm.

The duration of the critical period during which muscle activation is necessary if slow primary myotubes are to differentiate during E19–20 was analyzed further by paralyzing embryos by a single application of a small (2–3 mm length) TTX capillary on E15 or E16. In control experiments, capillaries of this size did not maintain paralysis for more than 2 days. The results of these experiments are illustrated in Fig. 10 and summarized in Fig. 11, where the effects of denervation, chronic paralysis and temporary paralysis are compared.

Fig. 10.

Effects on development of slow muscle fibres in TA muscle of temporary paralysis with TTX. Fibres reacting with anti-slow MHC appear dark in these phase-contrast photographs. Short TTX-capillaries paralysed embryos for about 2 days, after which they recovered their motility. All animals were examined on E20. (A) Paralysis during E15–17; (B) paralysis during E16–18; (C) paralysis during El7–20; (D) E20 control. Scale bar, 100 μm.

Fig. 10.

Effects on development of slow muscle fibres in TA muscle of temporary paralysis with TTX. Fibres reacting with anti-slow MHC appear dark in these phase-contrast photographs. Short TTX-capillaries paralysed embryos for about 2 days, after which they recovered their motility. All animals were examined on E20. (A) Paralysis during E15–17; (B) paralysis during E16–18; (C) paralysis during El7–20; (D) E20 control. Scale bar, 100 μm.

Fig. 11.

Neural regulation of differentiation of slow primary myotubes in tibialis anterior muscle. All muscles were examined at E20, and slow primary myotubes recognized by their reaction with antibody NOQ7.1.1A. Aneural embryos had all motoneurons destroyed by injection of β-BTX on E14, E15, E16 or E17. Paralyzed embryos were immobilized by chronic application of T1X beginning on E15 or E17. The effects of paralysis for about 2 days, followed by recovery, were studied by inserting small TTX capillaries on E15 or E16. (A) Numbers of slow primary myotubes on E20 after denervation, chronic paralysis or temporary paralysis, initiated at the times on the abscissa, compared with E20 controls. (B) The effectiveness of denervation, or chronic or temporary paralysis, in suppressing the differentiation of slow primary myotubes. Each point is the mean of data from A, presented as their inverse so that ‘0%’ would equal control numbers. In both graphs the abscissa gives the time of initiation of the experimental treatment. (○) β-BTX injected; (□) chronic paralysis with TTX; (▴) ≈2 day paralysis with TTX, followed by recovery; (•) control muscles.

Fig. 11.

Neural regulation of differentiation of slow primary myotubes in tibialis anterior muscle. All muscles were examined at E20, and slow primary myotubes recognized by their reaction with antibody NOQ7.1.1A. Aneural embryos had all motoneurons destroyed by injection of β-BTX on E14, E15, E16 or E17. Paralyzed embryos were immobilized by chronic application of T1X beginning on E15 or E17. The effects of paralysis for about 2 days, followed by recovery, were studied by inserting small TTX capillaries on E15 or E16. (A) Numbers of slow primary myotubes on E20 after denervation, chronic paralysis or temporary paralysis, initiated at the times on the abscissa, compared with E20 controls. (B) The effectiveness of denervation, or chronic or temporary paralysis, in suppressing the differentiation of slow primary myotubes. Each point is the mean of data from A, presented as their inverse so that ‘0%’ would equal control numbers. In both graphs the abscissa gives the time of initiation of the experimental treatment. (○) β-BTX injected; (□) chronic paralysis with TTX; (▴) ≈2 day paralysis with TTX, followed by recovery; (•) control muscles.

Denervation or maintained paralysis commencing on E14 or E15 resulted in the appearance of only about 10 slow primary myotubes in the TA muscle on E20, compared to more than 300 present in control muscles (Figs 8C, 11 A). Muscles paralyzed during E15–E17 and then allowed to recover showed a comparable suppression (Fig. 10A), with 48±6 (n=7) slow primary myotubes present on E20. Temporary or chronic paralysis from E16 was less effective, and paralysis or denervation from E17 had little effect (Fig. 10B, C). These data are replotted in Fig. 11B to show the effectiveness of the various treatments in suppressing differentiation into slow primary myotubes. Treatment on E14 or E15 results in virtually complete suppression; on E16, in suppression in half the primary myotubes; and, on E17, there is little effect. At all times, paralysis for 2 days is as effective as paralysis or denervation throughout the whole period, showing that the critical period is during E15–16, and that it precedes the time of differentiation by at least 2 days.

Muscle cleavage

The 4 peroneal muscles, P. longus, P. digiti quinti, P. digiti quarti and P. brevis form late in development from a single muscle mass which cleaves to form two muscles (P. longus plus another premuscle mass) by E17, followed by a trinary division of the last premuscle mass (Kieny et al. 1986) so that all four muscles are present by E18. These cleavages did not occur in paralyzed or denervated muscles. A summary graph of the effects of paralysis or denervation, initiated at various times, on cleavage of the peroneal muscle mass is presented in Fig. 12. The peroneal muscle mass remained in the same state as when paralysis was initiated; there was no regression within this time period. Temporary paralysis was as effective in stopping cleavage as chronic paralysis or denervation.

Fig. 12.

Failure of peroneal muscle mass to cleave into individual peroneal muscles in denervated and paralysed embryos. (○) β-BTX injected; (□) chronic paralysis with TTX; (▴) ≈2 day paralysis with TTX, followed by recovery; (•) control muscles.

Fig. 12.

Failure of peroneal muscle mass to cleave into individual peroneal muscles in denervated and paralysed embryos. (○) β-BTX injected; (□) chronic paralysis with TTX; (▴) ≈2 day paralysis with TTX, followed by recovery; (•) control muscles.

This study had two major goals: to study the role of innervation in the determination and subsequent differentiation of muscle fibre types, and to define the utility of two particular anti-MHC MAbs in identifying primary and secondary myotubes, so that individual isolated myotubes could be categorized in the study presented in the following paper (Harris et al. this journal). In addition, it extends previous observations on the normal sequence of development of muscle fibre types.

Specificity of antibodies

The skeletal muscle myosin multigene family includes at least 6 MHC genes and 6 light chain genes (Buckingham, 1985; Mahdavi et al. 1986). Individual myosin molecules are made up of 2 heavy chains and 4 light chains, so there is a large number of possible combinations, many of which actually occur in the adult or as stage-specific isoforms expressed during development. The sequence homology between members of the MHC gene family makes it difficult to obtain antibodies specific to only one form (Whalen et al. 1979; Mahdavi et al. 1986). These difficulties, and the problem of remaining contamination in polyclonal antibodies, have led to some ambiguities in the literature describing the expression of isomyosins during embryonic, foetal and postnatal development of particular muscle fibre types.

The migration patterns of rat MHCs on SDS gels have been examined by Danieli-Betto et al. (1986) who showed that the fastest migrating MHC is type I. This is consistent with its preponderance in SOL muscle, and our results from Western blotting show that antibody NOQ7.1.1A is specific for this isoform. These authors concluded also that the slowest migrating adult rat MHC isoform on SDS gels is type Ila, while the intermediate migrating form is type IIb. This pattern has been confirmed by Bar and Pette (1988), who also identified a fourth band, type IId, which migrates more slowly even than type IIa. These authors show that embryonic MHC runs slower than neonatal, with neonatal MHC running between types IIa and IIb. The pattern of silver staining of samples of adult rat muscle run on SDS gels in the present study closely follows the relative abundance of fibre types as measured by Armstrong and Phelps (1984) using traditional ATPase histochemistry (Table 1). The major MHC isoform from PN5 bulk hindlimb muscle ran in a position corresponding to Bär and Pette’s description of neonatal MHC, and was stained by MY32 (Fig. 1B). A band running at the same position in gels from E17 and E20 muscles was stained by MY32. In silver-stained gels, this band was obscured by another, heavier band which ran a little slower, corresponding to embryonic MHC in Bar and Pette’s (1988) description of mobilities on SDS gels; this band was not stained by MY32. Further confirmation of the binding specificity of MY32 came from studying myosin extracted from L6 myotubes, which have been shown to express only embryonic MHC (Whalen et al. 1979). Anti-embryonic MHC antibody 2B6 (Gambke and Rubinstein, 1984) bound L6 myosin, while MY32 did not (Figs 1, 2).

Native pyrophosphate gel electrophoresis separates myosin isoforms on the basis of both charge and molecular weight, and is useful in that at least some of the individual myosin subunit combinations can be distinguished. Furthermore, separation of isoforms including embryonic and neonatal MHC is possible, given appropriate acrylamide concentrations (Hoh and Yeoh, 1979; Whalen et al. 1981; Fitzsimons and Hoh, 1981; Lyons et al. 1983). The present study, using Western blots of SDS–PAGE and pyrophosphate–PAGE, indicates that MY32, as well as reacting with adult type Ila, lib and lid MHC’s, also reacts strongly with neonatal MHC. This was independently confirmed by Elisa analysis of antibody serial dilution curves (Fig. 2). These conclusions are supported by the results of Mahdavi et al. (1986) who showed with SI nuclease protection hybridization that adult fast MHC Ila and lib mRNAs are not synthesized until after birth in the rat.

Results from the pyrophosphate gel experiments confirm the SDS gel results that MAb NOQ7.1.1A is specific for type I MHC. In conclusion, the Western blots show that NOQ7.1.1A is highly specific for MHC type I in rats, and MY32 is specific for the so-called ‘fast’ class of MHC’s which includes Ila, lib, lid and neonatal MHC isoforms. (Neonatal muscle fibres containing only embryonic and neonatal MHC are actually slow-contracting; the designation ‘fast’ presumably refers to the final state of the muscle fibres in which they are found.)

Sequence of differentiation of myotubes in rat hindlimb muscles

Results from gel electrophoresis and immunohistochemistry showed that embryonic MHC is the major isoform present during embryonic and foetal development in the rat, and probably accounts for the physiological properties of foetal muscles (slow-contracting). In addition, particular myotubes show stage- and nervedependent expression of relatively small proportions of the slow and neonatal MHC isoforms.

All new-formed primary myotubes are slow and embryonic MHC-positive, neonatal/fast MHC-negative. Soon (⩽1 day) after their formation all primary myotubes, except in SOL muscle, also express neonatal MHC. In another strain of rats (Sprague-Dawley), no delay was evident in the expression of neonatal MHC. Simultaneous expression of embryonic and neonatal MHC mRNAs was seen in E15 mouse muscles by Weydert et al. (1987), but we are not aware of any other studies at such an early time from rat. The SDS gels show that embryonic MHC was the major isoform in bulk muscle up to and including E20, with neonatal MHC accounting for only a few per cent of total myosin content.

A day or two after all primary myotubes have formed, secondary myotubes begin to appear. These are all embryonic and neonatal MHC-positive, slow MHC-negative. At this time, primary myo tubes in mixed muscles begin to convert to fast (embryonic and neonatal MHC-positive, slow MHC-negative) or slow (neonatal MHC-negative, embryonic and slow MHC-positive) types, according to their position in the muscle. In general, axial primary myotubes become slow, and peripheral ones become fast (Dhoot, 1986; Narusawa et al. 1987). Primary myotubes in SOL never express neonatal MHC but only slow MHC, and secondary myotubes in this muscle begin to express slow MHC by E20, although at this time they also still express neonatal MHC.

As mentioned above, the use of different antibodies and difficulties in defining their specificities has led to some confusion in the literature describing the developmental sequence of appearance of myosin isoforms. With the work of Dhoot (1986), Narusawa et al. (1987) and the current study, in rats, and the results of Hoh et al. (1988) in cats, some general agreement has been reached.

We found, in future mixed muscles, that all primary myotubes of age E16–17 reacted with both anti-neonatal/f and anti-slow MHC MAbs, but subsequently they switched to express either neonatal or slow MHC, not both. Those expressing slow MHC are destined to become slow adult muscle fibres, specifically located in the axial part of the muscle. Soleus muscle primary myotubes never expressed neonatal or fast MHC, only slow. Secondary myotubes in soleus at first expressed only embryonic and neonatal MHCs, but by E20 were also expressing low levels of slow MHC. ITius, the generalization that primary myotubes develop into slow muscle fibres and secondary myotubes into fast muscle fibres (Kelly and Rubinstein, 1980; Rubinstein and Kelly, 1981) is useful but must be applied with care. In the rat IVth lumbrical muscle, it is strictly true (Jones et al. 1987); in TA, slow fibres have developed from primary myotubes, but primary myotubes in the superficial region of the muscle become fast; in SOL, secondary myotubes switch from fast to slow.

Determination of different muscle fibre types

The pattern of expression of MHC isoforms in chick muscle has been shown, on the basis of tissue culture experiments, to depend on the population of myoblasts which fused together to form the myotube (reviewed by Stockdale and Miller, 1987). Clones of early myoblasts form homogenous populations of myotubes (analogous to primary myotubes in vivo) expressing only fast, or mixed fast/slow, or only slow MHC. The fact that the same classes of myotube arise in mixed cultures suggests that, at least in this circumstance (absence of nerves and nerve-induced contraction), there is hemophilic fusion within each class of myoblast. While there are few similar experiments on mammalian skeletal muscle it is likely that similar mechanisms are employed to determine myotube properties, even if the phenomenology may differ.

The early stages of muscle formation are independent of innervation or contractile activity (Harris, 1981), but later development of muscle shows increasing reliance on these factors. Secondary myotube formation, for example, depends absolutely on innervation, due to nerve-dependence of mitosis in the population of myoblasts which should form secondary myotubes (Bonner, 1978, 1980; Ross et al. 1987b).

Nerves regulate muscle development not only by regulating mitosis in populations of myoblasts with different properties, but by changing the patterns of protein synthesis within multinucleate muscle cells. Adult muscle fibres, for example, respond to changes in their pattern of use with alterations in isoforms of both metabolic and contractile proteins. Thus, while the initial properties of a primary myotube depend only on the population of myoblasts from which it arose and are not affected by innervation, later nerve-determined changes in its properties might arise either from changes in expression of genes in nuclei contributed by these founding myoblasts, or by the appearance of nerve-dependent ‘late’ populations of myoblasts which, by fusing with the myotube, could introduce nuclei with new programs for gene expression. This latter possibility is supported by the results of the experiments on the effects of timing of denervation or paralysis, reported here.

Primary myotubes first appear in the TA muscle of the rat hindlimb on E15, and all have formed by E16. At this time, they all express both slow and neonatal MHC isoforms, as well as the majority class of embryonic MHC. Secondary myotubes begin to form on E17, and are easily visualized with light microscopy on E18, when they stain strongly for embryonic and neonatal MHCs and not at all for slow MHC. The adult muscle has two distinct zones, white (fast) on the outside, and red (slow) in the axial portion. The first signs of this regionalization become evident on E18, with inner primary myotubes staining more strongly for slow MHC and less strongly for neonatal MHC than the outer primary myotubes. By E20, two concentric bands of fibres, the inner presumptive slow and the outer presumptive fast, are clearly evident.

If the muscle is denervated or paralyzed from El5 onwards, commencing before all primary myotubes have been generated, then by E19 almost no myotubes express slow MHC, and all stain strongly for neonatal MHC, regardless of their position in the muscle. This represents a late-occurring change in development of primary myotubes, as on E18 their distribution still is comparable with controls, all fibres staining for both MHC isoforms. The muscle is smaller than normal, due to the absence of secondary myotubes, but there is no loss of primary myotubes. This developmental pattern is not seen if paralysis or denervation commences one day later, at E16. In this case, some secondary myotubes form, although the muscle still is smaller than normal. On E20, a normal pattern of distribution of slow and fast primary myotubes is present, even though paralysis/denervation was initiated at least two days before the change in primary myotube phenotype towards fast or slow could be identified. The fact that some secondary myotubes have formed is proof that a part of the nerve-dependent population of ‘late’ myoblasts was generated.

We have to explain a neural mechanism controlling the sequence of differentiation of a primary myotube that depends on its position in the muscle. It is unlikely that this mechanism involves position-dependent differences in patterns of nervous activation of primary myotubes. Until secondary myotubes start to form, primary myotubes are extensively electrically coupled to one another and the muscle contracts as a unit (Dennis et al. 1981), and all primary myotubes will have similar histories of contraction until a time well after denervation/paralysis ceased to be effective in changing primary myotube differentiation in our experiment. Position-dependent determination is a commonplace event in development, so that it would not be unusual to find induction of different populations of myoblasts in different regions of a muscle (Butler et al. 1988). Our experiments demonstrate that the default condition in muscle development is for fast/slow primary myotubes to stop expressing slow MHC shortly before birth and to become exclusively fast. This observation confirms the results of Weydert et al. (1987) who showed that the developmental sequence of appearance of embryonic, neonatal and adult fast type IIb MHC mRNAs still occurred in foetal mouse muscles denervated by injection of β-BTX. Future slow muscle fibres, in muscles that have been active during E14–16, do the opposite; neonatal MHC is degraded, and there is renewed expression of slow MHC. Our evidence supports the idea that the differentiation of primary myotubes from fast/slow to slow requires the addition of new myonuclei by fusion of a nerve-dependent population of ‘late’ myoblasts. The particular properties of these ‘late’ myoblasts are determined by a position-sensitive process of induction of myogenic precursor cells within the muscle.

Recognition of primary and secondary myotubes

In the following paper, (Harris et al. 1989) we study the birthdates of nuclei in disaggregated myotubes. One of the aims of the study reported here was to distinguish primary and secondary myotubes, using anti-MHC antibodies. Our results show that antibody NOQ7.1.1A unequivocally stains all primary myotubes at early times. Reactivity with this antibody is diminished in some primary myotubes beginning a day or so after secondary myotube formation has commenced. Antibody MY32 reacts strongly with all secondary myotubes but it also reacts, less intensely, with primary myotubes.

Muscle cleavage during development

Limb muscles form by a series of stereotyped binary or trinary cleavages from dorsal and ventral premuscle masses (Kieny et al. 1986). This sequence of patterned cleavage is disturbed in naturally occurring mutations crooked necked dwarf (cnd) in chicks and muscular dysgenesis (mdg) in mice, where developing muscles are chronically paralysed due to absence of the mechanism for excitation–contraction coupling. It was of particular interest, in the present study, to find that restoration of muscle contraction shortly after cleavage would normally have been complete did not result in restoration of a normal pattern of muscle formation (Fig. 11). Cleavage is due to synthesis of extracellular matrix materials in the plane of cleavage, followed by formation of the epimysium. In the chick end mutant some muscles do separate, but the continued paralysis results in these muscles fusing back together again (Kieny et al. 1986). The maintained effect of temporary paralysis indicates that muscle contraction is necessary for the appearance and/or organization of the connective tissue cells that will form the epimysium. Observations on development in the mdg and cnd mutations indicate that later, nerve-induced contraction may be necessary for maintenance of their differentiated state.

Neural determination of muscle fibre type

Several studies of fibre type differentiation in aneural or paralyzed muscles have been made in chicks, but little work has previously been reported from mammals, where most studies begin with denervation at birth. Crow and Stockdale (1986) paralyzed chick embryos with curare, beginning at day 4 in ovo. No differences in distribution of fast and slow myotubes could be seen at day 8, when only primary myotubes are present. By day 16, after secondary myotubes normally would have formed, primary myotubes staining for slow MHC were greatly reduced in number. They did not note whether this was due to specific loss of these myotubes, or to their expression of another MHC isoform. In a similar experiment on curarized chick embryos, Gauthier et al. (1984) noted a partial loss of myotubes staining for slow MHC from the anterior latissimus dorsi muscle; the transformed myotubes stained for a fast myosin light chain isoform not normally present in this muscle.

Our results, in rats, show that denervation or paralysis initiated no later than the first day all primary myotubes are present, prevents the later conversion of primary myotubes to their neonatal pattern of expressing either slow or neonatal MHC (in addition to the majority isoform, embryonic MHC). Instead, they all tend to express only neonatal MHC, and not slow. Treatment at this time also prevents the formation of secondary myotubes. Denervation or paralysis commencing slightly later, early enough to substantially diminish the number of secondary myotubes that form, does not prevent the normal development of adult patterns of distribution of primary myotubes with different MHC isoforms. We suggest that denervation or paralysis at this early time prevents the induction of a population of nerve-dependent myoblasts, and that transformation of primary myotubes to adult-slow fibres may depend on new myoblasts fusing with the muscles. The critical period of nerve-dependence for these myoblasts occurs several days before their actions can be noted.

This work was supported by the New Zealand Medical Research Council, the Vernon Willey Trust and the Muscular Dystrophy Association of America (MDA). We thank N. Rubinstein and D. Fishman for the gift of antibodies.

Armstrong
,
R. H.
&
Phelps
,
R. O.
(
1984
).
Muscle fibre type composition of the rat hindlimb
.
Am. J. Anat
.
171
,
259
272
.
Bader
,
D.
,
Masaki
,
T.
&
Fishman
,
D. A.
(
1982
).
Immunochemical analysis of myosin heavy chain during avian myogenesis in vivo and in vitro in chicken skeletal muscle
.
Devi Biol
.
95
,
763
770
.
Bär
,
A.
&
Pette
,
D.
(
1988
).
Three fast myosin heavy chains in adult rat skeletal muscle
.
FEBS Letts
235
,
153
155
.
Biral
,
D.
,
Betto
,
R.
,
Danieli-Betto
,
D.
&
Salviati
,
G.
(
1988
).
Myosin heavy chain composition of single fibres from normal human muscle
.
Biochem. J
.
250
,
307
308
.
Bonner
,
P. H.
(
1978
).
Nerve dependent changes in clonable myoblast populations
.
Devi Biol.
66
,
207
219
.
Bonner
,
P. H.
(
1980
).
Differentiation of chick embryo myoblasts is transiently sensitive to functional denervation
.
Devi Biol
.
76
,
79
86
.
Brown
,
W. E.
,
Salmons
,
S.
&
Whalen
,
R. G.
(
1983
).
The sequential replacement of myosin subunit isoforms during muscle type transformation induced by long term electrical stimulation
.
J.biol. Chem
.
258
,
14 686
14 692
.
Buckingham
,
M. E.
(
1985
).
Actin and myosin multigene families: their expression during the formation of skeletal muscle
.
Essays in Biochem
.
20
,
77
109
.
Buller
,
A. J.
,
Eccles
,
J. C.
&
Eccles
,
R. M.
(
1960
).
Interactions between motoneurons and muscles in respect of the characteristic speeds of their responses
.
J. Physiol
.
150
,
417
439
.
Butler
,
J.
,
Cosmos
,
E.
&
Brierley
,
J.
(
1982
).
Differentiation of muscle fiber types in aneurogenic brachial muscles of the chick embryo
.
J. exp. Zool
.
224
,
65
80
.
Butler
,
J.
,
Cosmos
,
E.
&
Cauwenbergs
,
P.
(
1988
).
Positional signals: evidence for a possible role in muscle fibre-type patterning of the embryonic avian limb
.
Development
102
,
763
772
.
Carmon
,
Y.
,
Czosnek
,
H.
,
Nudel
,
U.
,
Shani
,
M.
&
Yaffó
,
D.
(
1982
).
DNAse 1 sensitivity of genes expressed during myogenesis
.
Nucleic Acids Res
.
10
,
3085
3097
.
Couteaux
,
R.
(
1937
).
L’origine des ‘myotubes secondaries’ chez les embryons de mammiferes
.
C. r. hebd. Seanc. Acad. Set
.
205
,
1179
1181
.
Crow
,
M. T.
&
Stockdale
,
F. E.
(
1986
).
Myosin expression and specialization among the earliest muscle fibres of the developing avian limb
.
Devi Biol
.
113
,
238
254
.
D’Albis
,
A.
,
Pantaloni
,
C.
&
Bechet
,
J.-J.
(
1979
).
An electrophoretic study of native myosin isozymes and of their subunit content
.
Eur. J. Biochem
.
99
,
261
272
.
Danieli-Betto
,
D.
,
Zerbato
,
E.
&
Betto
,
R.
(
1986
).
Type 1, 2A and 2B myosin heavy chain electrophoretic analysis of rat muscle fibres
.
Biochim. biophys. Res. Commttn
.
138
,
981
987
.
Dennis
,
M. J.
,
Ziskind-Conhaim
,
L.
&
Harris
,
A. J.
(
1981
).
Development of neuromuscular junctions in rat embryos
.
Devi Biol
.
81
,
266
279
.
Dhoot
,
G. K.
(
1986
).
Selective synthesis and degradation of slow skeletal myosin heavy chains in developing muscle fibres
.
Muscle & Nerve
9
,
155
164
.
Draeger
,
A.
,
Weeds
,
A. G.
&
Fitzsimons
,
R. B.
(
1987
).
Primary, secondary and tertiary myotubes in developing skeletal muscle: a new approach to the analysis of human myogenesis
.
J. Neurol. Sci
.
81
,
19
43
.
Fitzsimons
,
R. B.
&
Hoh
,
J. F. Y.
(
1981
).
Embryonic and foetal myosins in human skeletal muscle. The presence of foetal myosins in Duchenne muscular dystrophy and infantile spinal muscular atrophy
.
J. Neurol. Sci
.
52
,
367
384
.
Gambke
,
B.
&
Rubinstein
,
N. A.
(
1984
).
A monoclonal antibody to the embryonic myosin heavy chain of rat skeletal muscle
.
J. biol. Chem
.
259
,
12091
12100
.
Gauthier
,
G. F.
,
Burke
,
R. E.
,
Lowey
,
S.
&
Hobbs
,
A. W.
(
1983
).
Myosin isozymes in normal and cross-reinnervated cat skeletal muscle fibres
.
J. Cell Biol
.
97
,
756
771
.
Gauthier
,
G. F.
,
Ono
,
R. D.
&
Hobbs
,
A. W.
(
1984
).
Curare-induced transformation of myosin pattern in developing skeletal muscle fibers
.
Devi Biol
.
105
,
144
154
.
Harris
,
A. J.
(
1981
).
Embryonic growth and innervation of rat skeletal muscles. 1. Neural regulation of muscle fibre numbers
.
Phil. Trans. Roy. Soc. Lond. B
293
,
257
277
.
Harris
,
A. J.
,
Duxson
,
M. J.
,
Fitzsimons
,
R. B.
&
Rieger
,
F.
(
1989
).
Myonuclear birthdates distinguish the origins of primary and secondary myotubes in embryonic mammalian skeletal muscles
.
Development
107
,
771
784
.
Hoh
,
J. F. Y.
&
Hughes
,
S.
(
1988
).
Myogenic and neurogenic regulation of myosin gene expression in cat jaw-closing muscle regenerating in fast and slow limb muscle beds
.
J. Muse. Res. Cell Motility
9
,
59
72
.
Hoh
,
J. F. Y.
,
Hughes
,
S.
,
Hale
,
P. T.
&
Fitzsimons
,
R. B.
(
1988
).
Immunocytochemical and electrophoretic analyses of changes in myosin gene expression in cat limb fast and slow muscles during postnatal development
.
J. Muse. Res. Cell Motility
9
,
30
47
.
Hoh
,
J. F. Y.
,
McGrath
,
P. A.
&
White
,
R. I.
(
1976
).
Electrophoretic analysis of multiple forms of myosin in fasttwitch and slow-twitch muscles of the chick
.
Biochem. J
.
157
,
87
95
.
Hoh
,
J. F. Y.
&
Yeoh
,
G. P. S.
(
1979
).
Rabbit skeletal myosin isozymes from foetal, fast-twitch and slow-twitch muscles
.
Nature, Lond
.
280
,
321
323
.
Jolesz
,
F.
&
Srgter
,
F. A.
(
1981
).
Development, innervation and activity-pattern induced changes in skeletal muscle
.
A. Rev. Physiol
.
43
,
531
552
.
Jones
,
S. P.
,
Ridge
,
R. M. A. P.
&
Rowlerson
,
A.
(
1987
).
Rat muscle during post-natal development: evidence in favour of no interconversion between fast- and slow-twitch fibres
.
J. Physiol., Lond
.
386
,
395
406
.
Kelly
,
A. M.
(
1987
).
Slow myosin in developing rat skeletal muscle
.
J. Cell Biol
.
104
,
447
459
.
Kelly
,
A. M.
&
Rubinstein
,
N. A.
(
1980
).
Why are fetal muscles slow?
Nature, Lond
.
288
,
266
269
.
Kelly
,
A. M.
&
Zacks
,
S. I.
(
1969
).
The histogenesis of rat intercostal muscle
.
J. Cell Biol
.
42
,
135
153
.
Kieny
,
M.
,
Pautou
,
M.-P.
,
Chevallier
,
A.
&
Mauger
,
A.
(
1986
).
Spatial organization of the developing limb musculature in birds and mammals
.
Bibl. anat
.
29
,
65
90
.
Laing
,
N. G.
&
Lamb
,
A. H.
(
1983
).
Development and motor innervation of a distal pair of fast and slow wing muscles in the chick embryo
.
J. Embryol. exp. Morph
.
78
,
53
66
.
Lyons
,
G. E.
,
Haselgrove
,
J.
,
Kelly
,
A. M.
&
Rubinstein
,
N.A.
(
1983
).
Myosin transitions in developing fast and slow muscles of the rat hindlimb
.
Differentiation
25
,
168
175
.
Mahdavi
,
V.
,
Strehler
,
E. E.
,
Periasamy
,
M.
,
Wieczorek
,
D. F.
,
Izumok
,
S.
&
Nadal-Ginard
,
B.
(
1986
).
Sarcomeric myosin heavy chain gene family: organization and pattern of expression
.
Med. Sci. Sports Exerc
.
18
,
299
308
.
Maier
,
A.
,
Gambke
,
B.
&
Pette
,
D.
(
1986
).
Degenerationregeneration as a mechanism contributing to the fast to slow conversion of chronically stimulated fast-twitch rabbit muscle
.
Cell Tissue Res
.
244
,
635
643
.
Martin
,
T. P.
,
Bonine-Fowler
,
S.
,
Roy
,
R. R.
,
Eldred
,
E.
&
Edgerton
,
V. R.
(
1988
).
Metabolic and fiber size properties of cat tibialis anterior motor units
.
Am. J. Physiol
.
255
(Cell Physiol. 24),
C43
C50
.
McCaig
,
C. D.
,
Ross
,
J. J.
&
Harris
,
A. J.
(
1987
).
Embryonic somatic nerve destruction with /3-bungarotoxin
.
Cell Tissue Res
.
247
,
41
50
.
Miller
,
J. B.
&
Stockdale
,
F. E.
(
1987
).
Developmental regulation of the multiple myogenic cell lineages of the avian embryo
.
J. Cell Biol
.
103
,
2197
2208
.
Mills
,
R. G.
&
Bray
,
J. J.
(
1979
).
A slow-release technique for inducing prolonged paralysis by tetrodotoxin
.
Pfliigers Arch
.
383
,
67
70
.
Morrisey
,
J. H.
(
1981
).
Silver stain for proteins in polyacrylamide gels: a modified procedure with enhanced uniform sensitivity
.
Anal. Biochem
.
117
,
307
310
.
Narusawa
,
M.
,
Fitzsimons
,
R. B.
,
Izumo
,
S.
,
Nadal-Ginard
,
B.
,
Rubinstein
,
N. A.
&
Kelly
,
A. M.
(
1987
).
Slow myosin in developing rat skeletal muscle
.
J. Cell Biol
.
104
,
447
459
.
Nemeth
,
P. M.
,
Pette
,
D.
&
Vrbova
,
G.
(
1981
).
Comparison of enzyme activities among single muscle fibres within defined motor units
.
J. Physiol., Lond
.
311
,
489
495
.
Pette
,
D.
&
Vrbova
,
G.
(
1985
).
Neural control of phenotypic expression in mammalian muscle fibers
.
Muscle & Nerve
8
,
676
689
.
Phillips
,
W. D.
&
Bennett
,
M. R.
(
1984
).
Differentiation of fiber types in wing muscles during embryonic development: effect of neural tube removal
.
Devi Biol
.
106
,
457
468
.
Ross
,
J. J.
,
Duxson
,
M. J.
&
Harris
,
A. J.
(
1987a
).
Formation of primary and secondary myotubes in rat lumbrical muscles
.
Development
100
,
383
394
.
Ross
,
J. J.
,
Duxson
,
M. J.
&
Harris
,
A. J.
(
1987b
).
Neural determination of muscle fibre numbers in embryonic rat lumbrical muscles
.
Development
100
,
395
409
.
Rubinstein
,
N. A.
&
Kelly
,
A. M.
(
1981
).
Development of muscle fiber specialization in the rat hindlimb
.
J. Cell Biol
.
90
,
128
144
.
Salmons
,
S.
&
Sreter
,
F. A.
(
1976
).
Significance of impulse activity in the transformation of skeletal muscle type
.
Nature, Lond
.
263
,
30
34
.
Stockdale
,
F. E.
&
Miller
,
J. B.
(
1987
).
The cellular basis of myosin heavy chain isoform expression during development of avian skeletal muscles
.
Devi Biol
.
123
,
1
9
.
Weydert
,
A.
,
Barton
,
P.
,
Harris
,
A. J.
,
Pinset
,
C.
&
Buckingham
,
M.
(
1987
).
Developmental pattern of mouse skeletal myosin heavy chain gene transcripts in vivo and in vitro
.
Cell
49
,
121
129
.
Whalen
,
R. G.
,
Schwartz
,
K.
,
Bouveret
,
P.
,
Sell
,
S. M.
&
Gros
,
F.
(
1979
).
Contractile protein isozymes in muscle development: identification of an embryonic form of myosin heavy chain
.
Proc. natn. Acad. Sci. U.S.A
.
76
,
5197
5201
.
Whalen
,
R. G.
,
Sell
,
S. M.
,
Butler-Browne
,
G. S.
,
Schwartz
,
K.
,
Bouveret
,
P.
&
Pinset-Harstrom
,
I.
(
1981
).
Three myosin heavy chain isozymes appear sequentially in developing rat muscle
.
Nature, Lond
.
292
,
805
809
.