Embryonic differentiation of fibre types in normal, paralysed and aneural avian superior oblique muscle

Two main types of muscle fibres, twitch and tonic, can be distinguished in the adult skeletal muscle of lower vertebrates on the basis of their morphology, metabolism, pattern of innervation and contractile speed (Kuffler & Vaughan-Williams, 1953; Couteaux, 1955; Ginsborg & Mackay, 1961; Hess, 1961, 1970; Page, 1969). Twitch and tonic fibres can be distinguished histochemically at the light microscopic level according to their myosin-ATPase staining patterns (Guth & Samaha, 1969). In slow twitch muscle fibres (type I) the ATPase activity is acid-stable and alkali-labile (Barnard, Lyles & Pizzey, 1982). In contrast, the fast twitch fibres have alkali-stable and acid-labile ATPase activity (type II). In the slow tonic (type III) fibres the ATPase activity is resistant to both acid and alkali treatment.

The role of motor nerve fibres in regulating the slow and fast muscle fibres has been investigated in newly hatched chicken muscle (Jirmanova, Hnik & Zelena, 1971; Bennett & Pettigrew, 1974; Hnik, Jirmanova & Syrovy, 1977; Cosmos, Butler, Allard & Mazliah, 1979). These studies indicate that the muscle fibres become slow or fast as a result of the type of nerve fibres innervating them. This point has been illustrated by cross reinnervation of slow tonic and fast twitch muscles. For example, Gordon and colleagues (Gordon, Perry, Srihari & Vrbova, 1971; Gordon & Vrbova, 1975) removed one anterior latissimus dorsi (ALD) and one posterior latissimus dorsi (PLD) muscle, minced each, and switched their positions before placing them back. This arrangement resulted in reinnervation of the minced (regenerating) ALD muscle by the nerve of the PLD muscle and vice versa. After regeneration and reinnervation, the ALD muscle fibres acquired characteristics of those normally seen in the fast fibres (PLD) and PLD became the slow muscle. In other words, the muscle fibre types transformed as a result of the nerve impulse activity, trophic substances or because of their new location. When a slow twitch muscle in its normal location is innervated by a nerve from the fast muscle it acquires properties characteristic of the fast muscle (Buller, Eccles & Eccles, 1960; Mommaerts et al. 1977). Likewise, when the nerve to a fast muscle is artificially stimulated at low frequency, the muscle becomes slow (Salmons & Sreter, 1976). This type of electrical stimulation of nerve induces synthesis of slow myosin in fast fibres (Pette & Schnez, 1977; Rubinstein et al. 1978). These experiments suggest that the expression of slow and fast muscle fibres in the mature muscle depends on the impulse activity of the innervating motor neurones.

The objective of the present study was to investigate whether the initial differentiation of fibre types, as detected by myosin-ATPase staining, during the course of embryonic development is also regulated by the motor neurones. The superior oblique muscle in duck embryo was used as a model to study fibre typing for several reasons. The extraocular muscles contain a mixture of fibre types (Maier, Eldred & Edgerton, 1972) which provides a unique situation to examine the effects of various experimental manipulations on the differentiation of all three fibre types in the same muscle. This muscle receives its innervation solely from a relatively small neurone pool, the trochlear nucleus. The trochlear nucleus is located dorsally in the midbrain which provides easy access for microsurgical destruction of innervation during development. Since eyes are usually visible through the egg shell window, the eye movements can be used as a rough index to monitor the effectiveness of neuromuscular transmission blockade or loss of innervation. Finally, a good deal of information on the development of connections between the trochlear nucleus and the superior oblique muscle already exists for the duck embryo (Sohal et al. 1985). In this report we describe the type of muscle fibres found in the adult superior oblique muscle and compare the differentiation of fibre types between the muscles developing normally with those either paralysed or developing in the absence of innervation.

Eggs of White Peking duck were incubated in a force-draft incubator at 37·5 °C and in a humidity of approximately 80 %. On the third day of incubation an opening in the shell of all eggs was made with a sander to acquire access to the embryo for application of curare, for microsurgery and for direct visual observations. Embryos were paralysed with daily application of 3 mg d-tubocurarine (d-TC, Sigma) in 0-1 ml saline from day 9 onwards. This is two days before the neuromuscular transmission begins in the trochlear nucleus-superior oblique muscle as determined electrophysiologically (Stoney & Sohal, 1978). d-TC solution was directly dropped onto the vascularized chorioallantoic membrane through the egg-shell window. Previous studies on embryonic motility have shown that this dosage of d-TC is sufficient to abolish all visible movements of the limb and eye (Sohal & Wrenn, 1984). A total of 38 embryos was paralysed and the results are based on 26 embryos that survived the treatment.

The trochlear nucleus becomes identifiable in histological sections on embryonic day 7 (Sohal, 1976). The trochlear nerve fibres first enter the superior oblique muscle on day 10 as determined by retrograde flow of HRP injected into the muscle on various days of embryonic development (Sohal & Holt, 1978). The trochlear nucleus, the sole source of innervation of the superior oblique muscle, was destroyed on embryonic day 7 as described previously (Sohal & Wrenn, 1984). Briefly, the microsurgical procedure involved cutting the vitelline membrane in a relatively avascular region, raising the head with a glass rod, and destruction of the dorsal midbrain with Malis Bipolar coagulator (Codlman & Shurtleff, Inc.). The embryo was replaced and the egg was sealed. The entire procedure took less than 1 min per embryo. The muscle was made aneural in 85 embryos and the results are based on 22 embryos that survived the operation. Because of the radical nature of the operation this treatment results in permanent, bilateral destruction of the trochlear and oculomotor nuclear complexes innervating the extraocular muscles (Sohal & Wrenn, 1984). The destruction of motor neurones was confirmed from histological sections of the brain and the lack of muscle innervation was verified with silver staining (see below).

In addition to the normal, paralysed, and aneural preparations described above a fourth group of embryos was also utilized. In this group the use of d-TC and motor neurone destruction was combined. In other words embryos with aneural muscle also received d-TC in the same amounts and for the same duration as described above for the paralysed group. A total of 24 embryos was subjected to this dual treatment and the results are based on nine surviving cases.

The superior oblique muscle was dissected, mounted on a piece of cork with OCT compound and rapidly frozen in liquid nitrogen. Serial transverse sections at 10 pm thickness were cut with a cryostat and alternate sections were processed for histochemical demonstration of (i) alkaline-stable actomyosin-ATPase activity, (ii) acid-stable actomyosin-ATPase activity or (iii) motor endplate cholinesterase activity and silver staining of nerve fibres. ATPase staining was performed according to the method of Guth & Samaha (1970) as modified by Butler & Cosmos (1981). To be more specific on critical details, alkaline preincubations were conducted for 10 min at pH 10·0 at room temperature. Acid preincubations were initially performed for 1·5 min at pH 3·7 to 4·7 at room temperature. pH lability patterns of all three muscle fibres from all ages permitted the use of only two pHs (10·0 and 4·3) in reliable identification of various fibre types and these pHs were used throughout this study. Adjacent sections were processed for combined cholinesterase and silver impregnation staining according to the method outlined by Toop (1976) to verify absence of innervation in the aneural muscle. Muscles on embryonic days 9–12,15–17, 22 and 27 (hatching); and of 3-week post-hatching and adult ducks (minimum age 1 year) were used for histochemical staining. The number of animals utilized varied from one to eight per age.

In order to characterize fully the fibre types in the superior oblique muscle, additional histochemical methods for demonstration of metabolic enzymes were performed on 3-week post-hatching and adult animals, α-glycerophosphate dehydrogenase (α-GPDH) activity was demonstrated by the method of Wattenburg & Leong (1960) with a 30-min incubation. Nicotinamide adenine dinucleotide tetrazolium reductase (NADH-TR) activity was demonstrated by a 20-min incubation using the method of Scarpelli, Hess & Pearse (1958). Phosphorylase (Phos) activity was shown by the protocol outlined in Pearse (1972) with an incubation time of 90 min including 20 % ethanol to inhibit branching enzyme activity.

Quantification of fibre types for the normally developing muscle was performed on embryonic days 9, 12, 16, 22, 27 and 3 weeks post-hatching. Quantification of fibre types in paralysed and aneural muscles was done on embryonic days 16 and 22. Fibre type populations were calculated from photographs (final magnification of x250 or 500) of serial sections of the muscle processed for alkaline- and acid-stable ATPase activity. Due to the uneven distribution of muscle fibre types across the muscle, all muscle fibres within the muscle sections were counted. All unstained fibres in the acid preincubation sections were type II fibres; all unstained fibres in the alkaline preincubation sections were type I. The number of type III fibres was determined by subtraction (total number of fibres minus type I and type II = number of type III). All counts were made blind and all fibres fell into one of the three categories. Quantification of fibre types of the adult muscle was not attempted.

The fibre type counts of 16- and 22-day paralysed and aneural muscles were compared to normals with a two-way analysis of variance. A one-way ANOVA and Tukey’s HSD post hoc test was used to determine significant differences in total number and number of each fibre type in normal, paralysed and aneural muscles. A Student’s t-test was used to determine differences between 16- and 22-day muscles.

Qualitative analysis of actomyosin-ATPase staining patterns revealed three types of muscle fibres in the adult superior oblique muscle. Type I fibres contain an ATPase activity that is stable to preincubation media of an acidic pH and labile under alkaline conditions (Figs 1,2). Following exposure to acidic media of pH 3·7 for 1·5 min, the ATPase activity of type I fibres is completely inactivated. A small amount of activity is retained after exposure to pH 3·9, a moderate amount at pH 4·1, and all activity is retained at pH 4·3 and higher. Following a 10·min preincubation at pH 10·0, all activity is lost. Type II fibres lose their activity when exposed to preincubation media of pH3·7 to 4·1 (Fig. 1). At pH4·3 a punctate staining pattern is observed in some of these fibres and is likely to be due to mitochondrial Ca²⁺ activated ATPase (Samaha & Yunis, 1973). At pH4·5 and above, the ATPase activity of these fibres is moderate to high. These fibres are intensely stained following exposure to pH 10·0 preincubation (Figs 1, 2). Type III fibres possess ATPase activity that is resistant to both acidic and alkaline conditions (Figs 1, 3). These fibres are intensely stained following pH 10·0 and pH4·3 to 4·7 preincubations. At pHs of 3·9 to 4·1 the ATPase activity is retained in most of the fibres; however, some fibres show a partial lability of this activity as indicated by their moderate staining.

Histochemical staining to demonstrate the relative activities of glycolysis (α-GPDH), glycogenolysis (Phos) and oxidative (NADH-TR) metabolism showed a considerable variability of activity within each fibre type (Figs 2–4). α-GPDH activity paralleled the alkaline-stable ATPase activity; type II and III fibres possessed moderate to high activity while type I fibres had little activity of this enzyme. In general, Phos activity was lowest in type I fibres. However, as may be observed in Figs 2,3, this activity was extremely variable among the different types of fibres. Some type I fibres contained Phos activity that was greater than some type II and III fibres. The range of NADH-TR activity in all three fibre types was also widespread as fibres with low, moderate, and high activity could be observed in all three types of fibres (Figs 2, 3, 4). In addition, a small number of type III fibres showed undetectable amounts of NADH-TR activity. In general, type I and II fibres are located at the peripheral zone adjacent to the orbit whereas the type III fibres are concentrated at the global (central) portion of the muscle (Figs 2, 3, 5). Correlation between ATPase pH labilities and the activity of metabolic enzymes in the adult muscle fibre types is summarized in Fig. 4.

Although the percentage of type I, II and III fibres in the superior oblique muscle varied with age (see below), the same pH lability patterns as in the adult muscle were observed in the embryonic muscle. Reliable identification of type I, II and III embryonic fibres could be made using only pH 4·3 and 10·0 preincubations. We therefore define type I fibres as those that are stained following pH 4·3 preincubation and unstained at pH 10· 0; type II fibres are stained at pH 10· 0 and unstained at pH 4· 3 and type III fibres are intensely stained at both pHs. The word ‘fibre’ is used throughout this study purely for the sake of consistency even though some stained cells represent myotubes, especially during earlier stages of development.

In the normally developing muscle, type II and III fibres can be identified as early as embryonic day 9 (Fig. 7). A majority of the fibres (82 %) belong to type III category. From days 9 to 16 the fibre types remain the same but the proportion of type II fibres increases to 39% (366 fibres) while the proportion of type III decreases to 61 % (571 fibres) on day 16 (Table 1). Type I fibres are first seen on embryonic day 17 (data not shown). On day 22 the percentages of type I, II and III fibres are 29, 53, and 18, respectively (Figs 5–7). On day 22 there is a significant increase in the number of type I and II fibres as well as in the total number of muscle fibres as compared to day 16 (Table 1). The small marginal and the large central portion of the muscle becomes identifiable also on embryonic day 22 (Fig. 5). The transition between the two zones is abrupt. After day 22 the percentage of type I and II fibres decreases whereas the type III increases. In a 3-week post-hatching muscle approximately 76 % of the fibres are type III, 17 % type II and only 7 % type I (Fig. 7).

The paralysed muscles were considerably smaller than normal at the corresponding ages. Histologically the muscle cells were much smaller in diameter (Fig. 6). A total muscle fibre count indicated a significant decrease as an average of 434 cells on embryonic day 16 and 367 cells on day 22 was noticed (Table 1). The normal muscle at the same ages had an average of 937 and 3455 cells, respectively (Table 1). The total fibre count in the paralysed muscle on day 16 is not significantly different from day 22. Type II and III fibres were present in the paralysed muscle from the onset of paralysis. On day 16, 58 % (255 fibres) were type II and the rest belonged to the type III (179 fibres) category (Fig. 7, Table 1). On day 22 the paralysed muscle contained an average of 112 (33 %) type II and 256 (67 %) type III fibres (Table 1). Both of these counts are significantly lower as compared to the 22-day normal. A comparison of fibre types between day 16 and 22 paralysed muscle indicated a significant decrease in the type II fibres on day 22 (Table 1). Type I fibres were never observed in the paralysed muscles (Figs 6, 7). Thus, paralysis of the embryo prevents the differentiation of type I but not type II and III fibres.

The aneural muscles were also much smaller than normal muscles. Histologically the aneural muscle cells resembled those in the paralysed muscle (Fig. 6). The total number of cells was not significantly different from the paralysed muscle as an average of 392 fibres on day 16 and 489 on embryonic day 22 was found in the aneural muscle (Table 1). In both cases there was a severe atrophy of the muscle as compared to the normals. In spite of this the marginal and the central zone of the muscle could be distinguished on day 22 (Fig. 6).

Initially the aneural muscle showed type II and III fibres (data not shown). On embryonic day 16 type I fibres were also present. The muscle on this day contained 27% (107) type I, 45% (177) type II, and 28% (109) type III fibres (Fig. 7, Table 1). It should be pointed out that the appearance of type I fibres in the aneural muscle is one day earlier than that seen in the normal muscle. As compared to the normal muscle on day 16, the aneural muscle contained significantly fewer type II and III fibres but significantly more type I fibres. A comparison between the aneural and paralysed muscle showed a significant difference in type I but not type II and III fibres on day 16. On day 22 a majority of the fibres belongs to the type II class, i.e. 28 % (139) type 1,57 % (277) type II and 15 % (73) type III (Figs 6, 7, Table 1). Although the percentage of each fibre type in the aneural muscle on day 22 resembles that of the normal, the fibre counts in each category are significantly lower in the aneural muscle. A comparison between the paralysed and aneural muscle on day 22 indicated a significant increase in the type I and II fibres and a significant decrease in the type III fibres in the aneural muscle. No significant differences exist between the day-16 and day-22 aneural muscle (Table 1). These results indicate that the initial differentiation of three fibre types in the superior oblique muscle occurs in the absence of innervation.

Under the combined experimental conditions we were interested in determining whether or not type I muscle fibres developed. Embryos in this group showed the most severe atrophy of muscles (Fig. 8). Muscles in all cases contained fewer than 200 muscle fibres, which is roughly half the number observed in paralysed and aneural muscles. Over the time period examined (16– 22 days), type I fibres were not observed (Fig. 8). The other two fibre types were present in proportions similar to those reported for the 22-day paralysed muscle in Table 1.

The results of this investigation indicate that at least three types of fibres are present in the superior oblique muscle of the adult duck. Type I fibre ATPase activity is acid-stable, alkali-labile; type II fibres possess alkali-stable and acidlabile ATPase; and type III fibre ATPase is both acid- and alkali-stable. No consistent pattern of staining was observed with metabolic enzyme activities. These observations on fibre types are in general agreement with other studies on the extraocular muscles of birds and mammals (Miller, 1967; Yellin, 1969; Maier et al. 1972).

The distribution of type I, II and III fibres within the superior oblique muscle is not homogeneous. Detailed quantitative studies of Maier et al. (1972) on the cross-sectional area of superior oblique muscle fibres in the marginal and central portions of adult quail, pigeon and canary have revealed significant differences in the diameter of fibres between the two portions of the muscle. Segregation of fibres into two distinct zones has been observed in the extraocular muscles of fish, amphibians, birds and mammals (Kilarski & Bigaj, 1969; Yellin, 1969; Maier et al. 1972). However, not all extraocular muscles exhibit different regions of fibre segregation. Segregation of zones is a feature unique to the superior and inferior oblique and rectus muscles. The mechanism by which segregation of fibre types is brought about is currently unknown. It has been suggested by Maier et al. (1972) that the two zones may result from differential interactions between the motor neurone axons and the developing myotubes. Since we have shown segregation of fibre zones in aneural muscle it is concluded that motor neurone influences are not directly involved in this process.

In the present study type II and III fibres could be identified with ATPase staining as early as day 9 in the normally developing muscle. This observation indicates that differentiation of type II and III fibre occurs very early in development; even prior to the arrival of motor nerve fibres in the muscle. Our previous studies utilizing retrograde uptake of horseradish peroxidase have indicated that the trochlear nerve fibres first appear in the superior oblique muscle on embryonic day 10 (Sohal & Holt, 1978). We have also shown that at this time the muscle is primarily composed of myoblasts and myotubes (Sohal & Holt, 1980). Together, these observations suggest that the initial differentiation of muscle cells into type II and III is independent of direct neural influences and that young myotubes can be stained for ATPase activity characteristic of type II and III muscle fibres.

In the present study type I, II and III fibres, in proportions similar to the normal muscle, especially on day 22, were observed in muscle developing in the absence of innervation. The simplest explanation for these findings is that a direct contact by the motor neurones is not essential for the initial expression of muscle fibre types during the course of embryonic development. This interpretation is consistent with the recent observations of others. For example, switching of motor neurone pools during development has no effect on fibre type differentiation (Khaskiya et al. 1980; Butler, Cosmos & Brierley, 1982b; Laing & Lamb, 1983). Butler et al. (1982a) reported that in the absence of motor neurones the differentiation of myotubes into slow and fast fibres occurs in the brachial muscles of the chick embryo. Phillips & Bennett (1984) recently reported differentiation of type I, II and III fibres in forelimb muscles of chick after early removal of neural tube. Although this aspect of our study confirms the conclusions drawn by the above authors it is difficult to compare our data with these studies since quantification of fibre types was not reported in the above studies and owing to the fact that in our study all three different types of muscle fibres were examined in a single muscle under several conditions.

The results of the present study also indicate that the type I fibres do not appear in muscles innervated but paralysed with d-TC. This observation confirms the finding of McLennan (1983a) in the leg muscles of the chick embryo. Why type I fibres do not appear following d-TC treatment while their differentiation is not affected in the absence of innervation is unknown. It is not due to a delay in differentiation and maturation as these developmental processes are more severely affected in the aneural than in the paralysed muscle (Sohal & Holt, 1980). It has been suggested by McLennan (1983a) that loss of twitch activity prevents the differentiation of type I fibres. Although our results are consistent with his suggestion they also suggest another possibility. It is possible that some unique action of d-TC on muscle, independent of innervation, is responsible for lack of type I differentiation since type I fibres were also absent in our curarized aneural muscles. In addition to the well-known effect of d-TC in decreasing twitch response of muscle it has been reported that d-TC has a direct effect on the muscle in that it increases the baseline tension in innervated and denervated adult mammalian muscles (McIntyre, King & Dunn, 1945; Bean & Elwell, 1951; Katz & Eakins, 1967). Although the effects of d-TC on tension of embryonic avian muscles are currently unknown it is possible that such muscles may also respond to d-TC with an increase in baseline tension. This assumption would imply that increased tension somehow prevents the appearance of type I fibres. This could explain the absence of type I fibres in our curarized innervated and curarized aneural muscles. An alternative explanation has been provided by Gauthier, Ono & Hobbs (1984). These authors observed that following treatment with d-TC, a type of myosin that is not normally present was observed in the ALD (slow) but not in the PLD muscle of chick embryo as detected immunocytochemically. They suggested that changes in myosin synthesis, rather than selective loss of slow myosin, may be related to the different composition of motor neurone pools as paralysis augments motor neurone survival by reducing the magnitude of normally occurring cell death (Pittman & Oppenheim, 1978). This would imply that the new myosin synthesis may be directed by the additional motor neurones. It should be pointed out that d-TC also prevents the death of the trochlear motor neurones in the duck embryo (Sohal, Leshner & Swift, 1983).

The results of the present study and of others indicate that motor neurones do not specify differentiation of fibre types during the course of embryonic development. This is contrary to the results of the cross-reinnervation experiments in the newly hatched chicken which have shown that the motor neurones control muscle fibre types. Apparently, some time during the course of embryonic development the motor neurones begin to exert control over the muscle. It appears that the muscles lose their independence some time after the initial nerve-muscle connections have been established. Neural control of muscle soon after innervation during in vivo development has recently been demonstrated (Toutant, Toutant, Renaud & Le Douarin, 1979; Renaud, Gardahaut, Rouaud & Le Douarin, 1983). These authors reported the persistence of type I fibres, which normally disappear during development, following artificial electrical stimulation of the PLD muscle in the chick. The precise mechanisms by which muscle cells lose their independence or the motor neurones exert their control over the muscle are unknown.

Finally, a comment should be made regarding the influence of innervation on the production of myotubes. The myotubes present at the beginning of myogenesis are termed primary myotubes (Kelly & Zacks, 1969). The myoblasts at the periphery of the primary myotube fuse to form secondary myotubes. As myogenesis proceeds the secondary myotubes are separated from the primary myotubes to form distinct muscle fibres (Ontell, 1977). In the present study the number of myotubes in the curarized muscle was similar to that seen in the aneural muscle and in both cases contained significantly fewer myotubes than in the normal muscle. This decrease is interpreted as lack of generation of secondary myotubes since d-TC treatment (McLennan, 1983b) and a lack of innervation (Harris, 1981) are known to inhibit the production of secondary but not primary myotubes. Since all three fibre types were seen in the aneural muscle our results suggest that primary myotubes are capable of differentiating into type I, II and III fibres (but see McLennan, 1983b). Our results also suggest that some fibres must switch types as the muscle matures. This is based on the fact that from day 16 to 22 there is a significant decrease in the number of type II fibres in the curarized muscles. It is unlikely that reduction in type II fibres is due to selective degeneration as the total number of primary myotubes remains stable during this period. In the curarized muscle type I fibres were absent and yet all fibres were stained. The number of type II and type III fibres differs significantly between the paralysed and the aneural muscle whereas there is no significant difference in the total number of myotubes between these two groups.

We thank Teena Knox and Greg Oblak for technical assistance, Sharlene Booker for typing the manuscript, and Harry Davis for statistics. This work was supported by grants from the National Institutes of Health (HD 17800, HD 18280, OH02020).