The myotubal origin of rat muscle fibres affects the extent of tenotomy-induced atrophy

Abstract

1. Immature muscle fibres (myotubes) can be divided into primary and secondary generations, which differ from each other in their time of formation, growth rates and myosin isoform expression. It is unclear whether the intrinsic differences between primary and secondary myotubes are totally extinguished once they mature and extrinsic factors, such as load, become important.

2. Four pregnant rats were injected with 5-bromo-2′-deoxyuridine (BrdU) on the 14th and 15th days of gestation. This selectively and permanently labels primary myotubes. Ten rats from four litters were killed when 8 months old, with three males (365–430 g) from a single litter being used for the quantitative study and the remainder being examined qualitatively. The extensor digitorum longus muscle (EDL) in each rat was tenotomized for 14 days. The sizes of fibres in the EDL were then correlated with their fibre type and whether they contained BrdU-labelled nuclei.

3. We reported that (i) II A and II B fibres derived from primary myotubes atrophied significantly less after tenotomy than II A and II B fibres derived from secondary myotubes and (ii) BrdU-labelled myonuclei were retained in the tenotomized muscle, even though tenotomy resulted in a substantial loss of myonuclei from the EDL.

4. We conclude that the origin of a fibre is a determinant of its response to the external forces which control its size, and hence force generation.

Muscle fibres can be subdivided into classes (fibre types) based on which isoform of myosin ATPase they express. The various classes of muscle fibre also differ in their sizes and profile of metabolic enzymes: II B fibres are large and glycolytic, whereas II A fibres are smaller with intermediate oxidative capacity. However, the size and the relative amounts of metabolic enzymes in a fibre are only loosely connected to the expression of a particular myosin ATPase. Consequently, the sizes of fibres belonging to a particular fibre type fall into a broad spectrum rather than a narrow band (Pette & Staron, 1997; Zhang et al. 1998).

As an animal ages, changes in its behaviour and body size alter the physiological demands being placed on its muscles. Increases in body weight require muscles to be stronger, which is achieved by hypertrophy of fibres (McLennan, 1990). The production of new fibres only occurs under non-physiological conditions, such as extreme weight training (Antonio & Gonyea, 1993). The size of individual fibres is controlled directly by the loading of the fibre, and indirectly through change in fibre type which is influenced by the pattern of neural activation of the fibre. This external regulation of the characteristics of fibres ensures that the strength of a muscle is exquisitely synchronized with the use of the muscle (Pette & Staron, 1997).

The plasticity of muscle noted above led to the view that the diverse characteristics of muscle fibres are entirely created by external factors. However, this view has been challenged by developmental biologists, as immature muscle fibres have intrinsic characteristics. Specifically, in rodents there are two generations of myotubes (primary and secondary) which differ in their time of formation, growth rates and types of myosin ATPase expressed (McLennan, 1994).

Although myogenic diversity clearly exists in developing muscles, the relevance of this diversity to adult muscles is unclear. Of particular importance here is the fact that the forces which shape adult muscles are minimal during development. Immature neuromuscular junctions can only sustain a low frequency of activation (Grinnell, 1995), which is well below the patterns of activation which define adult fibre types (Pette & Staron, 1997). The fetus develops in a fluid environment and maintains no fixed relationship to the earth's axis. This negates the importance of load and the distinction between muscles which act with or against gravity. Furthermore, the intrinsic characteristics of myotubes seem to be regulated by their originating myoblasts (Stockdale, 1997; Zhang et al. 1998). At birth, each immature muscle fibre contains a relatively small number of nuclei, most of which originate from myoblasts (Wigmore et al. 1992). By adulthood, muscle fibres have large numbers of nuclei, which are principally derived from satellite cells. Does this change in the nuclear composition of fibres, coupled with increased importance of external forces, totally extinguish the intrinsic characteristics of muscle fibres? In order to examine this question, we have developed a method of labelling primary myotubes that permits their mature characteristics to be examined and compared with fibres derived from secondary myotubes (Zhang & McLennan, 1995, 1998).

Our initial study demonstrated that the origin of a fibre (whether it is primary or secondary) is one of several factors which influences its fibre type, and hence speed of contraction. In the soleus, all primary myotubes develop into type I fibres whereas in the EDL, primary myotubes develop into a variety of fibre types, but with a significant bias towards types I or II B (Zhang & McLennan, 1998). Interestingly, in the soleus the type I fibres of primary myotubal origin are larger than those of secondary myotubal origin (Zhang & McLennan, 1998). This raised the issue of whether primary and secondary myotubal origin fibres have permanent differences in their ability to hypertrophy. In possible contradiction of this, the sizes of II A and II B fibres in the EDL are independent of myotubal origin (Zhang & McLennan, 1998). In this paper, we have examined this question further by determining the characteristics of adult type II fibres after reduction of the external force acting on the muscle, by tenotomy. We report that in tenotomized muscles, primary and secondary myotubal origin fibres differ significantly in their extent of atrophy, thus suggesting that the origin of a fibre is an important determinant of its response to the external forces which control its size, and hence force generation.

METHODS

Experimental animals

All experiments were approved by the University of Otago's Committee on Ethics in the Care and Use of Laboratory Animals and conform to the NIH principles of laboratory animal care. All animals were bred by the Department of Laboratory Animal Science, University of Otago.

Determination of the origin of adult muscle fibres

We have recently developed a method that uses 5-bromo-2′-deoxyuridine (BrdU) to selectively and irreversibly label muscle fibres that are derived from primary myotubes (Zhang & McLennan, 1995, 1998). BrdU is an analogue of thymidine that can be stably incorporated into the DNA of dividing cells. BrdU is rapidly degraded in vivo and cells only synthesize significant levels of brominated DNA if they are in the S-phase of a cell cycle within an hour of the injection (Kriss & Revesz, 1962). Cell proliferation dilutes the BrdU label as the bromine is spread between the daughter cells. Thus, the only cells that are heavily labelled after BrdU injection are those that were in the S-phase of a terminal cell division around the time of injection. Ninety-nine per cent of the myoblasts undergoing their terminal cell division on either embryonic days 14 or 15 subsequently fuse into primary myotubes (Zhang & McLennan, 1995). Injection of BrdU at this time thus leads to BrdU-labelled nuclei being absorbed into primary but not secondary myotubes (Zhang & McLennan, 1995, 1998). As primary myotubes grow, many unlabelled myoblasts fuse with them, so that by the time a labelled primary myotube has fully matured fewer than 1 % of its nuclei contain the BrdU label.

When a BrdU-labelled muscle from a mature adult is sectioned, virtually all of the fibres with a BrdU-labelled nucleus are of primary myotubal origin. On the other hand, the ‘unlabelled’ fibres are either of secondary myotubal origin or are primary myotubal origin fibres in which the labelled nuclei are located above or below the sections being examined. However, over 90 % of fibres in a rat muscle are derived from secondary myotubes (McLennan, 1994). Thus the mean characteristics of the unlabelled fibres in sections of muscles usually closely approximate the characteristics of the secondary myotubes. The type I fibres in a predominantly fast muscle are the only exception to this generalization as all such fibres are solely derived from primary myotubes (Zhang & McLennan, 1998). Thus, in the EDL both the labelled and unlabelled type I fibres are of primary myotubal origin and their characteristics should not be significantly different (see Table 1). For the sake of clarity, we emphasize that this rule does not apply to predominantly slow muscles such as the soleus, as the type I fibres in these muscles are derived from both primary and secondary myotubes (Zhang & McLennan, 1998).

Table 1.

Comparison of the sizes of muscle fibres in control and tenotomized EDL muscles

Fibre type	Control (mm3 per fibre)	Tenotomized (mm3 per fibre)	Extent of decrease (%)
I	0.019 ± 0.003 (185)	0.012 ± 0.002 (151)**	38.9 ± 2.3
IIA	0.025 ± 0.004 (316)	0.014 ± 0.002 (421)**	47.7 ± 2.2
IIB	0.056 ± 0.013 (539)	0.036 ± 0.008 (520)**	43.8 ± 9.2
Labelled I	0.017 ± 0.004 (71)	0.012 ± 0.002 (34)**	31.4 ± 5.6
Labelled IIA	0.032 ± 0.003 (19)	0.027 ± 0.003 (17)*	16.7 ± 2.7 ††
Labelled IIB	0.064 ± 0.012 (61)	0.060 ± 0.012 (63)	14.4 ± 3.2 †

The data are means ± standard error of the mean. The number of fibres examined are indicated in parentheses. The data in the right-hand column are the means of the three muscles and are therefore slightly different to the values obtained by simply dividing the mean size of control and tenotomized fibres. Statistical comparison with the corresponding values from control muscles (Student's t test):

^**P < 0.01

^*P < 0.05. Statistical comparison with the corresponding values from unlabelled fibres (Student's t test)

^††P < 0.01

^†P < 0.05.

Injection and detection of BrdU

Four pregnant rats were injected intraperitoneally with 5 mg of BrdU on both the 14th and 15th days of pregnancy. When the resulting pups had matured into 8-month-old adults, one of their extensor digitorum longus muscles was tenotomized. Fourteen days later, the rats were killed by cervical dislocation with death verified by the absence of a pulse. The resting length of each muscle was then measured with a dial calliper and the muscles serially sectioned (10 μm) in the transverse plan, using a cryostat. Three groups of sections were collected from each muscle (Fig. 1).

Figure 1. Schematic illustration of how the muscles were sampled.

Three groups of sections were collected from each muscle. Group 1 (thin lines) was used to estimate the number of labelled myonuclei in the muscle (Table 2). A minimum of 30 sections equally spaced along the length of the muscle (random systematic sampling; Gundersen et al. 1988) were double stained with anti-BrdU to detect labelled nuclei and anti-collagen IV (CoIV) to delineate the boundary of each fibre. Group 2 (thick lines) was used to: (i) estimate the total number of myonuclei in the muscles (Table 2), (ii) identify the fibre type of fibres, and (iii) measure the cross-sectional area of fibres (Table 1). An additional 4 sections were collected at every 4th level of group 1. One of each of these sections were stained with Haematoxylin and Eosin (H.E) to identify nuclei whereas the other 3 sections were stained for myosin ATPase, with various preincubation steps - pH 4.3, pH 4.6 or no preincubation to identify fibre types (Fig. 2). Group 3 (dashed lines) was used to measure the cross-sectional areas of labelled fibres (Table 1). A montage of a cross-section of the muscle was made and the location of labelled fibres from all twelve sections were marked on the montage. The fibre types of the labelled fibres were then identified using adjacent group 2 sections.

Labelled nuclei in the sections were detected using an antibody to brominated DNA. The anti-BrdU was a murine monoclonal antibody (Becton-Dickinson, San Jose, CA, USA) that was used in conjunction with an horseradish peroxidase-conjugated anti-mouse Ig antibody, with diaminobenzidine as the chromogen. The anti-collagen IV antibody was a rabbit polyclonal (Weislab, Sweden), with the immunoreactivity being developed with a Texas Red-conjugated secondary antibody (Molecular Probes, Eugene, OR, USA). The immunohistochemical procedures were as previously described (Zhang & McLennan, 1995) with the following minor modifications: (1) the sections were fixed in 1 % paraformaldehyde at 4°C for 30 min; (2) after fixation, each section was treated with 60 × 10⁻³ units ml⁻¹ of proteinase K in proteinase K buffer for 5 min.

Tenotomy

Ten rats were anaesthetized with an intraperitoneal injection of 90 mg kg⁻¹ ketamine and 3 mg kg⁻¹ xylazine. Two small incisions were made through the skin lateral to the knee joint and at the front of the ankle, and the proximal and distal tendons of the EDL were identified and severed. The incisions were closed with sutures and the animals allowed to recover in warm cages. The rats were monitored during the recovery period and daily thereafter to ensure that they suffered no pain or distress. Fourteen days later they were killed as described above and their EDL muscles were dissected from the operated and contralateral limbs. The muscles were frozen at their resting positions in melting isopentane and stored in a −80°C freezer until required. Three of the male rats (365–430 g) from a single litter were used for quantitative investigations, with the other rats being used for qualitative studies of the effect of tenotomy.

Fibre typing by ATPase histochemistry

Sections were preincubated at room temperature for 10 min in solutions consisting of 50 mM potassium acetate and 18 mM CaCl₂, adjusted to pH 4.3, 4.35, 4.4, 4.5, 4.6 or 4.7. One further section was processed without preincubation. After washing with 100 mM Tris-maleate buffer, pH 7.4, the sections were incubated at 37°C for 10 min in a solution containing 4.9 mM ATP, 71.4 mM CaCl₂, 7.1 mM citric acid and 100 mM Tris-maleate, adjusted to pH 9.4 with NaOH. After two successive 2 min incubations in 68 mM CaCl₂, the sections were incubated in 2 % (w/v) CoCl₂ for 3 min and then washed twice for 2 min in distilled water. After 1 min of incubation in 1 % (v/v) (NH₄)₂S, the sections were washed in tap water, dehydrated in ethanol, cleared in xylene, and mounted. Types I, II A and II B fibres were identified according to the criteria of Brooke & Kaiser (1970) (Fig. 2).

Figure 2. A BrdU-labelled type I fibre in an adult EDL muscle.

A, C and E are photographs of the same section: A is a bright-field image showing a nucleus stained with the anti-BrdU antibody; C is an epifluorescence image showing collagen IV immunoreactivity; E is an electronic image made by overlaying image A with image C. B, D and F are a series of sections adjacent to A which were stained with myosin ATPase after preincubation at pH 4.3 (B), pH 4.6 (D) or no preincubation (F). The arrows point to a BrdU-positive nucleus in A, C and E. Representative fibre types I, II A and II B have been labelled in B, D, E and F. Scale bar, 20 μm.

Calculations

The total volume of each muscle was calculated by multiplying the mean cross-sectional area of sections examined by the length of the muscle. The total number of labelled myonuclei within a muscle was calculated by multiplying the total volume of the muscle by the density of the labelled myonuclei. The density of labelled myonuclei was estimated by dividing the number of counted nuclei by the volume of the sections examined.

The mean volume of single fibres was estimated by multiplying the mean area per fibre by the length of the muscle. Ideally, the mean length of individual fibres should be used for this estimation. Practically, this is almost impossible. However, the fibres in the rat EDL muscles are of similar length, are nearly parallel with each others and lie at the same angle to the long axis of the muscle (Close, 1964). Therefore, all the fibres within the same muscle have a similar degree of error when their volumes are estimated by the length of the muscle.

The mean volume of fibres was used in this study. It is more common to use cross-sectional area as a measure of size. However, the cross-sectional area of a fibre is affected by its length and cross-sectional area is only a valid measure of fibre size when all muscles in the study have been stretched to a fixed length. This is not possible after tenotomy (see Results).

RESULTS

The sizes of fibres in the control muscles varied significantly with fibre type (Table 1). When examining the affect of myotubal origin on fibre size, it was therefore necessary to compare the size of labelled (primary) fibres of a particular fibres type with the unlabelled (secondary) fibres of the same fibre type. In the control muscles, the sizes of the labelled types II A and II B were not significantly different from the corresponding unlabelled II A and II B fibres (Table 1).

Tenotomized EDL muscles

The EDL muscles were examined 14 days after cutting both of their tendons. The gap between the two stumps of a tendon was filled with connective tissue. The tenotomized muscles were atrophic, very stiff and could not be stretched to their pre-tenotomized lengths. This shortening of the tenotomized muscles, coupled with decreased cross-sectional area, combined to produce an overall reduction in volume of 29 % (Table 2). This figure for the overall reduction in volume is slightly less than for the mean reduction in the volume of muscle fibres (Table 1) as the connective tissue elements of the muscles atrophied less than the muscle fibres.

Table 2.

The sizes of control and tenotomized EDL muscles

	Control	Tenotomized
Number of myonuclei
Total	2.99 ± 0.14 × 106	1.88 ± 0.05 × 106*
Labelled	8.60 ± 0.26 × 103	8.75 ± 0.45 × 103
Muscle volume (mm3)	66.0 ± 4.4	46.8 ± 1.7*
Muscle length (mm)	27.0 ± 0.6	25.0 ± 1.3

The data are means ± standard error of the mean. Significantly different from control muscles:

^*P < 0.01, Student's t test.

Type II fibres of primary myotubal origin were less affected by tenotomy

The myotubal origin of the type II fibres significantly affected the extent of atrophy. In the tenotomized muscles, the mean volume of the labelled (primary) II A and II B fibres were significantly larger than the mean volume of unlabelled II A and II B fibres, which are predominantly of secondary myotubal origin (Table 1). A similar analysis cannot be done with type I fibres, as all type I fibres in the EDL are of primary myotubal origin (see Methods).

BrdU-positive myonuclei are selectively preserved following tenotomy

The total numbers of myonuclei and BrdU-positive myonuclei were counted in both tenotomized and contralateral control muscles (Table 2). In the 14 days following tenotomy, the EDL muscle lost about one-third of its myonuclei (37 %). However, the total number of BrdU-positive myonuclei did not change (Table 2).

DISCUSSION

Force generation by fibres is related to their size. The size of a fibre is regulated by multiple influences. Loading or unloading directly affects size, whereas the pattern of neural activation has an indirect effect through the regulation of fibre type (Pette & Staron, 1997). We have recently shown that the size of type I fibres in the soleus varies depending on whether the fibre is derived from a primary or secondary myotube (Zhang & McLennan, 1998). In this paper, we have demonstrated that the mean sizes of the II A and II B fibres derived from primary myotubes are significantly greater than the corresponding sizes of II A and II B fibres of secondary myotubal origin, but only when the effect of load is removed by tenotomy.

In a mature fast muscle, the contribution of myotubal origin to the size of a fibre is relatively small compared with the influence of fibre type and load. This can be seen by examining the data in Table 1: II B fibres of primary myotubal origin are substantially bigger than type I or II A fibres of primary myotubal origin (cf. labelled types I, II A and II B fibres); tenotomized II B secondary myotubal origin fibres are substantially smaller than normal II B secondary myotubal origin fibres (cf. unlabelled normal and tenotomized II B fibres). The influence of myotubal origin only becomes apparent when the effects of fibre type and load are eliminated by comparing fibres with the same fibre type, and the same loading but with different myotubal origins. This, however, does not imply that the effect of myotubal origin is irrelevant: the difference in the sizes of primary and secondary myotubal origin type II fibres in a tenotomized muscle is large enough to be biologically important (cf. labelled with unlabelled II B fibres in a tenotomized muscle, Table 1).

We hypothesize that the primary myotubal origin fibres may act as a small conservative base for a muscle, which only slowly and/or minimally adapts to external forces. The more abundant secondary myotubal origin fibres may constitute the highly adaptive bulk of a muscle. Consistent with this, the production of primary myotubes during development is relatively invariant compared with secondary myotubes (McLennan, 1994) and primary myotubal origin fibres atrophy minimally after tenotomy (Table 1). If this hypothesis is correct, then primary myotubal origin fibres should also atrophy less after denervation or starvation and be relatively resistant to fibre-type transformation.

Preferential retention of labelled nuclei

Muscle fibres retain a relatively constant ratio of nuclei to cytoplasm (Enesco & Puddy, 1964; McCall et al. 1998). Consistent with this, the number of myonuclei in the tenotomized EDL decreased in parallel with the overall decrease in muscle volume (Table 2). However, the number of labelled nuclei in the EDL was not affected by tenotomy (Table 2). This preferential retention of labelled nuclei was an unexpected finding that our study was not designed to examine. Nevertheless some preliminary points can be raised. First, this observation could be a trivial consequence of where the labelled nuclei are located within the fibres. During elongation, muscle fibres grow by the fusion of myoblast or satellite cells to their ends (Goldspink, 1980; Ontell & Kozeka, 1984). This results in the labelled nuclei being concentrated in the middle portion of a muscle (Zhang & McLennan, 1995). Thus, if nuclei are preferentially lost from the ends of a muscle after tenotomy, then the labelled nuclei will be largely retained.

Conversely, the preservation of labelled nuclei may be biologically important. Primary myotubes originate from a class of myoblasts known as early myoblasts, whereas secondary myotubes originate from late myoblasts. This difference in originating myoblasts may underlie the differences between primary and secondary myotubal origin fibres (for a discussion of this see McLennan, 1994; Stockdale, 1997). The BrdU label used in this study is introduced into the primary myotubes via their originating myoblasts. The BrdU-labelled nuclei will thus be predominantly or exclusively derived from early myoblasts. Primary and secondary myotubes elongate and hypertrophy by the addition of nuclei from satellite cells. A priori it is possible that when a fibre atrophies, it selectively eliminates nuclei of satellite cell origin and retains nuclei from the originating myoblasts, thus helping to preserve the intrinsic characteristics of the fibre.

Conclusion

Muscle fibres are heterogeneous, even within a given fibre type. Part of this heterogeneity is due to differences in load and the pattern of activation. This, and our previous paper (Zhang & McLennan, 1998), establish for the first time that the embryological origin of a fibre also affects its mature characteristics. Further studies are needed to determine the range of mature characteristics which are affected by embryological origin and the relative importance of embryological origin versus extrinsic forces.