Longitudinal growth of striated muscle fibres

During post-natal growth, the skeletal muscles of most mammals increase in length as a result of the longitudinal growth of their component muscle fibres (Kitiyakara & Angevine, 1963; Goldspink, 1968; Bridge & Allbrook, 1970). There is evidence that, in the mouse, this lengthening of fibres, though due partly to an increase in the length of individual sarcomeres, is mainly due to an increase in the number of sarcomeres along the length of the fibres (Goldspink, 1968). In this investigation further studies have been made of the post-natal increase in fibre length by directly counting the number of sarcomeres along the length of teased single fibres taken from mice of different ages. It has been shown that the growth of muscle fibres is accompanied by an increase in the number of muscle fibre nuclei (Montgomery, 1962; Enesco & Puddy, 1964). Direct counts of nuclei have been made on teased single fibres in order to determine exactly the rate of increase of muscle fibre nuclei during post-natal growth.

Many workers consider that during the lengthening process the incorporation of new sarcomeres into existing myofibrils takes place by serial addition at the ends of the muscle fibres (Holtzer, Marshall & Finck, 1957; Kitiyakara & Angevine, 1963; Ishikawa, 1965; Goldspink, 1968; Mackay, Harrop & Muir, 1969). Some authors, however, suggest that new sarcomeres may be inserted at any point along the length of the fibres (Ruska & Edwards, 1957; Schmalbruch, 1968). In order to clarify this situation an attempt has been made to label the newly formed actin filaments and ribosomes, using the method of Griffin & Goldspink (in preparation), and thus determine the region where new sarcomeres are laid down. Using the electron microscope, an investigation has been made into the way in which newly synthesized contractile proteins are assembled on to the existing myofibrils.

Experiments have shown that surgical modification of the distance the muscle has to contract affects the longitudinal growth of the muscle as a whole (Ctawford, 1954; Alder, Crawford & Edwards, 1959). In order to determine whether it is possible to alter the rate and the extent of the increase in sarcomere number, and to ascertain the importance of movement for normal post-natal growth, some immobilization experiments have been carried out.

The mice used were normal homozygous males of the 129/Re strain. The animals were reared in the Department of Zoology from a colony which was originally obtained from Jackson Memorial Laboratories, Bar Harbor, U.S.A. They were fed on a modified diet formula 41b (Oxoid Ltd.) with food and water available at all times. The biceps brachii and soleus muscles were the muscles chosen for this study on account of their relatively simple, fusiform structure with fibres running from tendon to tendon. Also these muscles, in the mouse, are small enough to be fixed in situ.

Mice of ages ranging from newborn to 2 years were used. Young mice were killed with ether vapour and splints were tied to the limbs so that the biceps brachii and soleus muscles were fully extended. Older mice were killed by dislocation of the cervical vertebrae and the limbs were pinned out on a cork board so that the muscles were again in the fully extended position. The biceps brachii and soleus muscles were exposed by removing the overlying skin and other tissues and they were fixed in situ by pipetting fixative on to the muscles for 20 min. The fixative used was 2·5% glutaraldehyde buffered at pH 7·5 with 0·1 M phosphate buffer and containing 0·5% glucose. The whole limbs were then removed from the animals and placed in fixative. After 1 h, the biceps brachii and soleus muscles were dissected out and placed in fresh fixative for a further 90 min. After fixation the muscles were washed in buffer and the distance between the muscle-tendon junctions was measured for each muscle using micrometer calipers. The muscles were placed in 30 % (w/v) HNO_:I for 2 days to hydrolyse most of the connective tissue. They were then washed and stored in 50 % glycerol to remove the soluble proteins and make the myofibrils more clearly visible. Usingelectrolytically sharpened tungsten needles and with the aid of a dissecting microscope, individual whole fibres were teased out from random positions within the muscles and measured with micrometer calipers. They were then mounted in glycerol jelly with the coverslips supported by strips of glass to avoid compressing the fibre. The fibres were then photographed on 35-mm film, using a Leitz Ortholux microscope with a camera attachment. Photomicrographs were made of 3 fibres from each muscle and the number of sarcomeres along the length of the fibres determined by counting each sarcomere.

Mice of ages ranging from 1 day to 2 years were used. Muscles were fixed in Carnoy’s fixative. After 30 min fixation on the bone, the biceps brachii and soleus muscles were removed and fixed for a further 1 h, then placed in absolute ethanol for 90 min. After hydration, they were immersed for 48 h in gallocyanin stain, adjusted to pH 1·5. The muscles were washed in distilled water and stored in 50 % glycerol. Single fibres were teased out and mounted in glycerol jelly. The fibres were viewed through a binocular microscope and the number of nuclei in each fibre was determined by counting. Counts were made on 3 fibres taken from random positions within each muscle. Care was taken to distinguish between muscle fibre nuclei and the nuclei of the adhering endomysium. The muscle fibre nuclei were found to lie parallel to the axis of the fibre and were rod shaped, whereas the connective tissue nuclei were irregularly arranged, fusiform, and stained more darkly.

Tritiated adenosine was used to locate the region of longitudinal growth in young biceps and soleus muscles. The adenosine is incorporated into newly formed actin and ribosomes. Free adenosine and other adenosine-containing compounds are removed by glycerol extraction (Griffin & Goldspink, in preparation).

Mice weighing approximately 5 g were given daily intraperitoneal injections of 25 μCi [³H]adenosine for periods ranging from 2 days to 3 weeks. The animals were then sacrificed and the hind and forelimbs removed and pinned to pieces of cork board. They were then immersed for 2 weeks in a mixture of 50 % glycerol and 50 % 0·2 M phosphate buffer, pH 7·6, and maintained at − 5°C. During this period the glycerol mixture was changed frequently. Following glycerol extraction, the limbs were washed twice in phosphate buffer only. The biceps brachii and soleus muscles were dissected out, fixed in glutaraldehyde and embedded in ester wax. The muscles were sectioned longitudinally at a thickness of 5 μm and ribbons of sections were mounted on subbed slides and coated with Ilford K2 nuclear emulsion. The dried slides were placed in light-tight boxes and exposed at 4 °C. Test slides were developed at intervals of 1, 2, 3 and 4 weeks. The slides were stained with haematoxylin and eosin and examined under reflected-light illumination using Leitz Ultropak optics (the silver grains appear as white spots).

Young mice weighing approximately 5 g were injected daily with 25 μCi [³H]adenosine for 3 days. The forelimbs were removed and immersed for 2 weeks in 50 % glycerol/buffer mixture. Following this they were washed for 2 days in buffer only. The biceps brachii muscles were removed and supported on the microtome chuck in brain tissue taken from a non-injected animal. The microtome chuck was then immersed in Freon 12 which had previously been cooled to − 160 °C using liquid nitrogen. Each muscle plus brain tissue was sectioned transversly on a Pearse Slee cryostat at a thickness of 20 um. The first section containing muscle was placed on a slide and the following 40 sections were collected in a scintillation vial. The next section was put on a slide, the following 40 into a second vial. This procedure was repeated until the entire muscle had been sectioned. The sections on slides were fixed and stained and the area of muscle in each was determined using a slide projector and a planimeter. From these measurements, the volume of muscle tissue in each vial could be estimated.

To each scintillation vial, 0-7 ml of hyamine IOX hydroxide was added and the vials were sealed and rotated for 36 h in a water bath at 37 °C to dissolve the tissue sections. After cleaning the outside of the vials, 10 ml of PPG in toluene were added to each. The vials were placed in a Packard automatic scintillation counter and counted for too min. To determine the counting efficiency of each sample, 50 μl of tritiated hexadecane was added to each vial and the resulting mixture recounted. The radioactivity of each sample was expressed as dpm/mm³ tissue.

In a second experiment, adult mice were injected daily with 50 μ Ci [³H]adenosine for 3 days. The limbs were removed and the muscles treated in the same way as the young muscles.

Biceps brachii and soleus muscles from foetal and newborn mice were fixed in glutaraldehyde, washed overnight in phosphate buffer at 4 °C, and postfixed for 2 h in OsO₄. The muscle fibres were teased apart in 70% ethanol and dehydrated in 100% ethanol and 100% acetone. Short lengths of single fibres or small bundles were taken from the middle and end regions of the muscles and embedded in Araldite (Ciba). Longitudinal sections were cut using a Reichert ultramicrotome. Ribbons of sections showing gold or silver interference colours were collected on Celloidin-coated grids and double stained with uranyl acetate and lead citrate solutions. The sections were examined with a JEOL JEM 7 A electron microscope.

Soleus muscles of young mice were immobilized in different positions during the growth period.

Plaster casts were used to hold the hind limbs of mice with the whole limb in the extended position, that is to say with the soleus in its shortened position. Animals weighing approximately 5 g had plaster casts put on one hind limb; the contralateral leg served as the control. The casts were changed twice weekly throughout the experiment. The presence of the cast did not usually cause any decrease in the growth of the bones and in those few cases where it did, the animals were discarded from the experiment.

Mice from one group were sacrificed at intervals ranging from 2 to 6 weeks after immobilization. The number of sarcomeres per fibre in the soleus muscles from experimental and contralateral limbs was determined. Mice from a second group had their casts removed when the mice weighed between 18 and 20 g. They were left for 3-4 weeks before being sacrificed and the number of sarcomeres per fibre was determined in the experimental and contralateral muscles.

Hind limbs of young mice were again held in plaster casts but this time with the limb in a flexed position, that is to say with the soleus muscle in the fully extended position. The mice were sacrificed at intervals ranging from 2 to 6 weeks after immobilization. The number of sarcomeres per fibre was again determined in soleus muscles from experimental and contralateral limbs.

The number of sarcomeres in single fibres teased from biceps brachii and soleus muscles is shown in Fig. 1. It will be seen that there was very little variation in the number of sarcomeres along the length of different fibres in a muscle at any particular age but that there was a very marked increase in sarcomere number with age (body weight). The greatest increase was found to occur during the first 3 weeks after birth. In the soleus muscle the rate of increase in sarcomere number was greater than in the biceps brachii. When the animal reached a weight of approximately 22 g (6 weeks) no further increase in sarcomere number was detected. From the data in Figs. 1 and 2, it will be seen that in both the biceps brachii and the soleus muscles, the rate and extent of increase in fibre length are approximately the same as the rate and extent of the increase in sarcomere number.

The number of nuclei in single fibres teased from biceps brachii and soleus muscles is shown in Fig. 3. The number of nuclei per fibre increased with body weight, and this increase continued beyond the point at which the sarcomere increase had ceased. The rate of increase in the number of nuclei per fibre was greater in the soleus than in the biceps brachii. The nuclei were found to be scattered randomly along the fibres of both muscles. Again the variation between the measurements on different fibres in a muscle at any particular age was quite small.

Fig. 5 shows an autoradiograph of a longitudinal section of the biceps brachii muscle of a 5-g mouse which had been injected with [³H]adenosine. It can be seen that most of the label was located at the ends of the muscle fibres. These results indicate that the newly formed actin and newly formed ribosomes are found mainly in the end regions of growing muscle fibres.

It can be seen from Table 1 that in the biceps brachii muscles taken from young mice the tritiated adenosine was incorporated more into the ends of the muscles than into the middle regions. By grouping vials 1 and 5 and vials 2, 3 and 4, and applying a t test, it was shown that the difference between the counts in the ends and the middle regions of the fibres was significant at the 5% probability level. Very little [³H]-adenosine was incorporated into the adult muscles and there was no difference between the ends and the middle regions.

Newborn and foetal muscle fibres were seen to contain narrow myofibrils which were located around the periphery of the fibres (Fig. 6). Near the ends of the myofibrils ribosomes were often found surrounding the myosin filaments (Figs. 7, 8). At the muscle-tendon junction the myofibrils appeared to be connected to the plasma membrane in the region of the sarcolemmal clefts (Fig. 9). Myofilaments which were not organized into myofibrils were frequently seen in the sarcoplasm at the ends of the muscle fibres near to the plasma membrane (Fig. 10).

The results of these experiments are shown in Fig. 4. As will be seen from this figure, the effect of immobilization during the growth period was to reduce the rate and extent of sarcomere addition. The number of sarcomeres in fibres from muscles which had been immobilized in the shortened position was approximately half that in fibres from the normal, contralateral muscles. When plaster casts were removed from the immobilized limbs the number of sarcomeres rapidly increased, so that after a period of 4 weeks the sarcomere number in the experimental and the contralateral muscle fibres did not differ significantly.

In fibres from muscles which had been immobilized in the extended position, the number of sarcomeres was reduced to almost the same extent as in those which had been immobilized in the shortened position (Fig. 4). Again, removal of plaster casts after a period of immobilization was followed by an increase in sarcomere number and a return to the normal level in a period of 4 weeks.

Whilst previous work (Goldspink, 1968) has indicated that the main factor in the increase in fibre length during post-natal growth is an increase in the number of sarcomeres along the length of the muscle fibres, the number of sarcomeres was estimated only from measurements of sarcomere length and muscle belly length. The direct determination of sarcomere number as reported here proves conclusively that increase in fibre length during normal growth is accompanied by a large increase in the number of sarcomeres in series. This increase in sarcomere number was sufficient to account for the increase in fibre length, but it is possible that the increase in fibre length is also accompanied by changes in sarcomere length (Goldspink, 1968). Changes in the sarcomere length would not be detected in this study since the nitric acid treatment causes uneven shrinkage of the sarcomeres.

The apparent discrepancy between the extent of increase in fibre length and muscle length is interesting. After the point at which the fibres cease to grow in length, further increase in muscle belly length must presumably be due to a rearrangement of the fibres. Observations made while teasing the muscles indicated that in the adult mice the points of insertion of the fibres into the tendon were more staggered than in the young mice.

It is known that different muscles mature at different rates and in mammals the musculature of the front end differentiates first. Since it has been observed that the extent of increase in the number of sarcomeres and the number of nuclei per fibre is greater in the soleus muscle than in the biceps brachii, it may be assumed that the biceps brachii is more mature at birth.

Using biochemical methods based on the determination of DNA content and morphological methods involving tissue sections, some workers (Montgomery, 1962; Enesco & Puddy, 1964) have reported that the number of nuclei increases in muscle fibres growing in vivo. Direct determination of the absolute number of muscle fibre nuclei proves conclusively, first, that there is a large increase in the number of nuclei during the post-natal development of muscle fibres and, secondly, that the number of nuclei per fibre continues to increase beyond the point at which the fibres have stopped increasing in length. This continued increase is probably associated with increase in girth of the muscle fibres. Moss (1968) found that in growing chick muscle the crosssectional area of the fibres and the total number of nuclei (estimated from DNA determination) maintained a constant ratio. It is known (Rowe & Goldspink, 1968) that mouse muscle fibres continue to increase in girth beyond the 20-g stage at which sarcomere increase ceases. If a certain number of nuclei are required for a given volume of cytoplasm, then it is obviously necessary for the nuclei to continue to increase in number as long as cytoplasm is being added to the fibre, whether it be due to an increase in length of the fibre or to an increase in girth.

During the differentiation of muscle tissue, nuclear replication ceases once the myoblasts have fused to form myotubes (Stockdale & Holtzer, 1961). It seems probable that additional nuclei are provided by the fusion of satellite cells with the muscle fibre, as suggested by Shafiq, Gorychi & Mauro (1968). In the electron-microscope studies reported here some cells were observed in the process of fusion with muscle fibres. These cells were located beneath the basal membrane and may, therefore, have originated as satellite cells. It is perhaps significant that such cells were found near the ends of the fibres. Kitiyakara & Angevine (1963) noted that during the first few days after injection of pHjthymidine, a high proportion of the labelled muscle fibre nuclei was located near the ends of the muscle, which indicates that the ends of the muscle fibres are areas of active development.

The use of tritiated adenosine proved to be very valuable in locating the growth region of the muscle fibres. (The advantage of using this isotope rather than a labelled amino acid lies in the fact that it is a fairly specific label for actin. Labelled amino acids are incorporated into all the proteins which are being synthesized by the muscle cells, and, due to their slower rate of synthesis, the myofibrillar proteins do not incorporate much of the label. Experiments have shown (Griffin & Goldspink, in preparation) that in myofibrils prepared from young mice which had been injected with [TI]-adenosine, approximately two thirds of the label was in the ADP of the actin and approximately one third was in the RNA of the ribosomes which are bound to the myofibrils.) In young muscle it was found that the isotope was located more in the ends than in the middle portions of the fibres, thus strongly suggesting that the new actin filaments and ribosomes, and hence the new sarcomeres, were added to the ends of the existing myofibrils during the process of longitudinal growth. The fact that sometimes only one end of the muscle was heavily labelled may be attributable to intermittent growth.

In the electron-microscope studies, a variety of evidence was found indicating that new sarcomeres are laid down at the ends of the muscle fibres. In the first place, the myofibrils in this region were not well organized; free filaments could be seen, particularly round the periphery of the fibre. Secondly, numerous ribosomes were found, particularly in foetal muscle. These were often associated with the terminal myosin filaments in the way described by Larson, Hudson & Walton (1969). Sometimes the appearance of these ribosomes suggested that they were helically arranged round the myosin filaments.

Crawford (1961) has shown that the immobilization of limb joints results in the muscles failing to attain their normal length. The present investigation demonstrates that this is due to a smaller number of sarcomeres along the length of the constituent fibres. When the lower joints of the hind limbs were immobilized with the soleus in either the extended or the shortened position, the sarcomere number fell far short of that of the controls. Coupled with this, the fact that the bone length was not affected implies that the tendon was correspondingly longer. When the plaster casts were removed, the total number of sarcomeres in series increased almost to the normal level during a period of a few weeks. Moreover, the sarcomere number returned to the normal level during a stage in development at which the addition of sarcomeres would normally be almost completed. According to the work of Tabary, Goldspink, Tardieu, Tabary & Tardieu (in preparation), immobilization of muscles by means of plaster casts in adult animals resulted in a loss of sarcomeres in series, followed by a return to normal levels once the restriction was removed.

The changes found in immobilized muscles may be due mainly to a decrease in the activity of the muscles. Since, however, immobilized muscles are able to develop some isometric tension (Fischback & Robbins, 1969) it is perhaps necessary that they contract isotonically for normal fibre growth to occur and be maintained. There is apparently a strict relationship between the length of the limb bone, the degree of movement and the number of sarcomeres in series. The nature of this relationship between the function of the muscle and its longitudinal growth is being investigated further.

This work was supported by a research grant from the Agricultural Research Council. The authors also wish to acknowledge the advice and help received from Mr G. E. Griffin, Mr S. Waterson and Mrs Lesley Clapison.