An Analysis of Nuclear Numbers in Individual Muscle Fibers during Differentiation and Growth: A Satellite Cell-Muscle Fiber Growth Unit CONSTANCE A . CARDASIS273 A N D GEORGE W. COOPER Department of A n a t o m y , College of Physicians and Surgeons, Columbia University, New Y o r k , N e w Y o r k 10032
ABSTRACT A numerical analysis of changes in the populations of nuclei in individual, intact muscle fibers was made to study how multinucleation arises during normal differentiation and growth. Gastrocnemius muscle fibers from pre- and post-natal mice were isolated with guanidine (Cardasis and Cooper, '75) and examined. Satellite cells associated with muscle fibers were first observed at 19 days of gestation. The number of nuclei per muscle fiber (muscle satellite cell nuclei) averages 83 at this age, 157 at birth and continues to increase to 354 by 63 days of age. However, the rate of increase during growth is not constant. Estimates of satellite cell and muscle nuclei in histological cross sections indicate that there is a decrease in the percentage of satellite cells from 32% at birth to 6% in the adult. However, the numbers of satellite cells associated with individual muscle fibers, calculated from these percentages and the nuclear counts on whole fibers, decreases only between 2 and 4 weeks of age. Cytosine arabinoside was injected subcutaneously during the first two weeks of age. Pairs of satellite cells, abnormal nuclei and elevated percentages of satellite cells were observed. This evidence as well as the numerical analysis of nuclear populations in whole fibers lends further support to the hypothesis that satellite cells account for the increase in muscle nuclei from birth to maturity.
+
How multinucleation occurs in muscle cells during normal differentiation and growth has not been fully explained. During the early stages of differentiation, mononucleated myoblasts fuse to form multinucleated myotubes. Myotubes contain rounded, centrally located nuclei aligned end to end, which are surrounded by myofilaments. The fusion is well documented by transplantation experiments (Loeffler, '70), the use of allophenic mice (Mintz and Baker, '67) and photography of fusion in living cultures (Bassleer, '62). It seems likely that, in tissue culture, DNA synthesis is limited to mononuclear myoblasts. Okazaki and Holtzer ('66), employing both 3H-thymidine autoradiography and fluorescent antibodies demonstrated that nuclei in myotubes in which myosin is detected do not incorporate "-thymidine for DNA synthesis and that myoblasts which incorporate "-thymidine do not contain myosin. J . EXP.ZOOL., 191: 347-358.
Myotubes initiate the process of differentiation into myofibers prior to birth. Following birth there are a number of changes still occurring in the morphology of the muscle fibers as well as an increase in size of muscle fibers. Whether muscle nuclei also increase in number during postnatal growth has been controversial. Some early investigators found no increase in the number of nuclei with age (Godlewski, '02) while others reported an increase (Morpungo, 1898; Schiefirdecker, '09). More recent investigations indicate that the muscle cells continue to grow and the muscle nuclei continue to increase in number following birth (Kitiyakara and Angevine, '63). Nuclear counts based on 1 Supported in part by a grant of the Institute of General Medical Sciences, GM 15289. 2 This paper is based on a dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Faculty of Pure Science, Columbia University. 8 Present address: Department of Anatomy, Downstate Medical Center, Brooklyn, New York 11203.
34 7
348
CONSTANCE A . CARDASIS A N D GEORGE W . COOPER
histological sections (Montgomery, '62), two mice at each of the following periods chemical determinations of DNA content of development: the 19th day of gestation of whole muscle (Enesco and Puddy, '64), (day - 2 ) , birth (day 0) and 1, 2, 4, 5, 6, cinephotomicrographic analysis of serial 7 and 9 weeks of age. cross-sections (Bridge and Allbrook, '70) The day of gestation of fetal stages was and the analysis of electron microscopic determined by examination of age dependmontages of muscle cross-sections (Schultz, ent external characteristics of the fetus '72) all demonstrate an increase in the (Grunberg, '42). The lower right limbs number of muscle nuclei during post-natal were dissected from unanesthetized fetuses growth. This raises the question of how at 19 days of gestation and placed intact this increase is occurring, when the mus- into 1% glutaraldehyde (Polysciences) in cle nuclei are thought to have lost their a 0.1 M phosphate buffer, pH 7.4 (Karlsability to undergo DNA synthesis and son and Schultz, '65). Animals killed at divide. birth to one week of age were sacrificed by Satellite cells are mononucleated cells decapitation, older animals by etherization. which lie between the muscle basal lamina The lower right legs from all animals were and plasma membrane (Mauro, '61). They skinned and initially placed in the above may be myoblasts which were incorporated fixative and the gastrocnemius muscles within the basement membrane of the were removed from the bones. The total muscle without fusion with the muscle time of fixation for the intact muscle was plasma membrane or they may be derived one hour. Fixed muscles were stored in from cells penetrating the basement mem- 0.1 M phosphate buffer (pH 7.4). Groups brane later in life. They have the ability of fibers from the margins of these musto undergo DNA synthesis (Moss and Le- cles were dissected free from the muscle blond, '70) and mitosis (Shafq et al., '68) and individual fibers were isolated in a during growth and during regeneration solution of 0.02 M guanidine-HC1 (Car(Reznik, '69). These cells may serve as a dasis and Cooper, '75). The remainder of source of reserve myoblasts in adult life. the muscle was prepared for light and The present investigation presents a electron microscopy. The semi-thin light description of the increase in nuclei during microscopic sections were used to deterpre- and post-natal development until sex- mine the satellite cell percentages (see ual maturity. It is based on nuclear counts RESULTS for criteria). made on individual fibers which have been The results of the counts of nuclear isolated by the guanidine technique (Car- number per fiber are based on the numerdasis and Cooper, '75). This procedure al- ical average of the counts of eight individlows a more direct determination of nu- ual fibers from two animals (4 fibers from clear numbers as it provides intact muscle each animal) for each time period. An fibers with their satellite cells but free of analysis of the standard deviations from their connective tissue. the means was performed on the numerSpecifically, this investigation answers ical count data. the following questions: (1) Is there a n inCytosine arabinoside treatment of growcrease in the total number of nuclei per ing animals. Eight growing mice, of which fiber during post-natal growth? (2) At any four served as controls, were used in this given age, what percentage of the isolated study. The times chosen for treatment fiber's nuclei are satellite cells? (3) When were based on the critical periods of nudo increases in nuclear number occur and clear increase obtained from the study of what are the rates of increase during spe- normal growth. Cytosine arabinoside (Sigcific time intervals? ma) was employed at a dose of 0.002 mg/ gm body weight. One animal received a MATERIALS AND M E T H O b S subcutaneous injection of cytosine arabinoMaterials for the study of normal growth. side at three days of age and was killed The growth of muscle was studied in white by ether at four days of age. Another animice of the Princeton-Rockefeller strain. mal received cytosine arabinoside at ten This strain is sexually mature at approxi- days of age and was killed at 11 days. The mately six weeks of age. The right gas- remaining animals received multiple introcnemius muscles were obtained from jections of cytosine arabinoside (each in-
NUCLEAR NUMBERS AND GROWTH OF MUSCLE FIBERS
jection was 0.002 mg/gm body weight). One received the drug at 2 and 5 days of age and was killed on day 7. Another received the drug at 2, 5 and 11 days of age and was killed at 14 days of age. Control animals were uninjected littermates killed by ether at the same ages as the experimental animals. The gastrocnemius muscles from all animals were fixed in 1% glutaraldehyde as described previously and prepared for light and electron microscopy. The light microscopic sections were used to determine the percentages of normal and abnormal appearing muscle and satellite cell nuclei. Electron microscopy was used to verify the location of abnormal nuclei. Preparation of light and electron microscopic sections Blocks of tissue were obtained and fixed in 1% glutaraldehyde by the identical method employed for the whole fibers. Following a wash in 0.1 M phosphate buffer all tissue to be embedded were post-fixed one hour in 1 % osmium tetroxide in 0.1 M phosphate buffer (pH 7.4). They were then dehydrated in a graded series of ethyl alcohols to absolute ethanol and in propylene oxide. The tissues were infiltrated with Epon 812 and embedded. Both the semithin (0.5-1 .O micron) and thin (50-70 nm) sections were cut on an LKB Ultratome. The semi-thin sections were stained with the Azure 11-methylene blue method of Richardson ('60) modified as follows: 1% periodic acid was used for ten minutes (instead of 5 minutes) and the staining solution was diluted to half of the original concentration. Heating of the sections during staining was not performed. These sections were used for determining the percentage of satellite cells, which contained the more heterochromatic nuclei as compared with muscle nuclei. The percentage of satellite cells was calculated from 200 nuclei (muscle and satellite cell nuclei) counted in two blocks of muscle tissue from each animal (total 400 nuclei/ animal). The number of satellite cell nuclei within the basement membrane of individual muscle cells was calculated by multiplying these percentages of satellite cells by the average of the total nuclei counted in isolated fibers. Phase photomicrographs of light micro-
349
scopic sections were taken on a Reicher microscope with Kodak Panatomic X film and developed in Microdol X, diluted 1 to 3. The thin sections for electron microscopy were placed on copper grids and stained with a filtered solution of saturated aqueous uranyl acetate and lead citrate (Reynolds, '63). Sections were examined and photographed on an RCA EMU3-G electron microscope at 50 KV. RESULTS
Light and electron microscopic observations. There are differences in the structure of young and mature mouse gastrocnemius muscle. Some of the structural changes which can be correlated with the changes in nuclear numbers during growth will be described. At a gestational age of 19 days (fig. 1A) and at birth (figs. l B , 2), the cross striations are not as distinct as in adult muscle due to the fact that all the myofibrils are not aligned. Many lipid droplets are present and glycogen is abundant and diffusely scattered throughout the sacroplasm. Muscle fibers have satellite cells associated with them by the 19th day of gestation. The cytoplasm of satellite cells associated with muscle fibers contain many free ribosomes and mitochondria. Their nuclei contain a large amount of heterochromatin near the nuclear envelope. In contrast, the muscle nuclei contain mostly euchromatin and have very large, distinct nucleoli (fig. 2). These two nuclear characteristics are the basis for distinguishing satellite from muscle nuclei with the light microscope. The muscle and satellite cell nuclei seen in whole mounts of young isolated fibers appear oval (figs. lA,B). By four weeks (fig. 1C) the nuclei have the elongated or fusiform appearance of adult muscle nuclei. Electron microscopy demon strates that by four weeks of age the internal structure of the muscle fiber looks very much like that of the adult. The cross striations are distinct, the lipid droplets are rare and glycogen particles are limited to the interfibrillar spaces near the sarcoplasmic reticulum. Changes in nuclear numbers in isolated fibers. Nuclei counted in fibers isolated by the guanidine-HC1 technique include both muscle nuclei and the satellite cell
350
CONSTANCE A. CARDASIS AND GEORGE W. COOPER
35 1
NUCLEAR NUMBERS AND GROWTH OF MUSCLE FIBERS TABLE 1
N o r m a l growth of mouse gastrocnemius muscle fibers ~
Length of fiber in mm? S.D.
No. of nuclei (muscle and satellite cell)/ fiber? S . D .
0.90 2 0.05 1.23 t 0.23 1.53 t 0.27 2.98 2 0.22 4.00 2 0.32 4.29 2 0.37 4.52 f 0.34 4.76 f 0.34 4.42 f 0.43 5.20 C 0.41
83 t 11.0 1 5 7 t 42.0 199 t 32.4 241 t 23.6 276 f 26.4 302 f 36.1 312 i25.1 341 i29.5 355 f 49.7 349 i 32.3
Age in days
relative to birth
-2
~~~~
~~
No. of nuclei (muscle and satellite cell)/ 170 @ ? S.D.
No. of sateilite cells per fiber
Yo 2
02
72 14 28 35 42 49 63 6 mos.-1
yr.3
15.8 f 2 . 1 21.6 f 2.4 22.7 f 5.3 13.8f 1.2 11.820.7 12.0 t 0.8 11.8f0.7 12.2 1.0 13.6 t 0.8 11.5 2 0.5
50 (32) 55 (27) 48 (20) 28 (10) 15 (5) 22 (7) 25 (7) 21 (6)
*
' Each average i s based on an analysis of eight fibers from two individual animals.
Averages are based o n an analysis of less than eight fibers from two individual animals, Averages are based on an analysis of 21 animals (4 fibers from each animal). ( 1 o/o satellite cell nuclei of all nuclei within muscle basal lamina as observed in cross sections.
3
nuclei associated with individual fibers. The number of nuclei per fiber increases from 84 nuclei per fiber at the gestational age of 19 days to 354 nuclei per fiber at 63 days of age (fig. 3, table 1). The rate of increase can be divided into three distinct intervals. The greatest rate of increase occurs in the two day interval between the gestational age of 19 days (84 nucIei) and birth (158 nuclei), when the number of nuclei per fiber almost doubles. The next largest increase in total number of nuclei occurs between birth, seven days (208 nuclei) and 14 days (241 nuclei) of age. After 14 days the rate of increase in nuclei per fiber becomes progressively smaller. Changes in the length of isolated fibers. While the nuclei are increasing in number during growth, the size of the fiber is also increasing. The increase in the latFig. 1 A. Photomicrograph of gastrocnemius muscle fiber isolated in guanidine from a mouse embryo of a gestational age of 19 days. Oval nuclei (arrow) and some striations are present. x 1,600. B. Photomicrograph of gastrocnemius muscle fiber isolated in guanidine from mouse a t birth. Nuclei (large arrow) and lipid droplets (small arrow) can be seen. X 1,600. C. Photomicrograph of mouse gastrocnemius fiber isolated in guanidine at four weeks of age. The area of tendon attachment (t) is demonstrated. The muscle nuclei are elongated (arrow), rather than round to oval as in younger muscle. X 1,600. Fig. 2 Electron micrograph of mouse gastrocnemius muscle a t birth. Note heterochromatic satellite cell nucleus (sc), euchromatic muscle nuclei (m) and lipid droplets in muscle cytoplasm (arrow). x 6,100.
"1
I
I
0
Blrtn
10
,
20
30
,
40
,
50
I
60
DnYS of CGE
Fig. 3 Graph showing an increase in the number of nuclei per muscle fiber with age. Brackets indicate standard deviation.
era1dimension is apparent in whole mounts (figs. lA,B,C) but was not measured. The lengths of each of the fibers was measured (fig. 4, table 1). The greatest increase in length occurs between days 7 and 14 (1.53 mm vs. 2.98 mm). This is approximately 40% of the total increase in length. The increases in nuclear number and in lengths of the fibers during growth do not always occur at the same rate. The numbers of nuclei per fiber for a given age as seen in figure 3 was divided by the number of 170 micron segments in the fiber
CONSTANCE A. CARDASIS AND GEORGE W . COOPER
of nuclei and the rate of increase in length are closely correlated events. Satellite cell numbers. Percentages of satellite cell nuclei were determined on histological cross-sections of muscle (fig. 6, table 1). The percentage of satellite cells could not be determined with any degree of confidence in fetal muscles. The results demonstrate that there is an increase in the percentage of satellite cell nuclei from 32% at birth to 10% at 4 weeks of age. From this time till 63 days of age the percentage of satellite cells only decreased to -2 :O I0 i0 30 40 50 $0 6%. BIRTH AGE IN DAYS The percentage of satellite cells estimated by counts made on histological secFig. 4 Changes in the length of muscle fibers as related to age. Brackets indicate standard detions were related to the numbers of nuviation. clei found in individual fibers, isolated by the guanidine method. The total number (fig. 5, table 1). This calculation was done of satellite cell nuclei per individual musto determine the relationship between the cle fiber was calculated by multiplying the increase in the number of nuclei and the estimated percentages by the average of increase in length of the fiber. The in- the total number of nuclei per fiber (fig. crease in nuclei per unit length increases 6, table 1). Although the percentages of from the gestational age of 19 days (day satellite cells are decreasing, these calcu- 2) ( 15 nuclei/ 170 micron segment) to lated absolute values do not differ greatly birth (day 0) (21.6 nuc€ei/l70 segment) from birth through 14 days. There is a due to an increase in the number of nu- decrease between 14 and 28 days of age. clei with little increase in fiber length. However, following this age, the number This high number per standard of length of satellite cell nuclei per fiber again reis maintained during the first seven post- mains constant. natal days. There is a sharp reduction beCytosine arabinoside treatment of growtween 7 and 14 days due to the great ini n g animals. Injection of cytosine aracrease in fiber length. From 14 to 63 days of age, there is only slight fluctuation in binoside on day 3 resulted in an elevated this value, with overlapping standard de400 viations. This indicates that from 2 weeks of age the rate of increase in the number 3SO ,
I
,
w K
m T
300
w
2
10
250
I
zoo u 3 150 (r w
g
100
z 3
50
2
,
,
I0
I
$0
33
40
2 Q
.
50
60
Birth
D A Y S of AGE
Fig. 5 The frequency of nuclei per 170 micron segment of muscle fiber as related to age. Brackets indicate standard deviation.
Birth
7
14
21
56
DAYS of AGE
Fig. 6 The number of nuclei per muscle fiber as related to age. Blackened areas indicate the % of nuclei which are satellite cell nuclei within the basement membrane of a musclefiber.
NUCLEAR NUMBERS A N D GROWTH OF MUSCLE FIBERS
percentage of satellite cells (43%) as compared to control animals (25 % ) on day 4. Similar elevated values for the percentage of satellite cells were found regardless of whether the animals received multiple or single injections during the first 11 days post-natal (fig. 7, table 2). Satellite cells in close proximity with one another are frequently seen in animals treated with cytosine arabinoside (figs. 8, 9). In addition, certain satellite cells appear binucleate or possibly contain nuclei which have constricted or are dumbbell shape (fig. 9). These abnormalities are not seen in muscle cell nuclei. The percentage of satellite cells which exhibit these nuclear abnormalities comprise approximately 20 % of the satellite cell population or 8% of the total nuclear population. Whole mounts of individual fibers were
-
$ 1 Y
5
loo
CONTROL
CYTOSINE ARABINOSIOE
-X
t3
80
I
2
3
4
MULTIPLE INJECTIONS
5 6 7 8 9 1 0 1 l 1 2 1 3 1 4 AGE IN DAYS
Fig. 7 Effect of 0.2 mg/kg of cytosine arabinoside on the percentage of nuclei within the basal lamina of muscle fibers which are satellite cell nuclei compared with control values. TABLE 2
Cytosine arabinoside t r e a t m e n t during jirst t w o w e e k s of growth
Age in days relative to birth
Cytosine arabinoside (0.2 mglkg) Satellite cells
Satellite cellsabnormal
Control
Satellite cells
Satellite cellsmitotic figures
70
70
70
%
19
7 11 1
43 39 40
26 20
14
28
14
25 27 26 17
4 4 4 0
4
*
Animals killed at 4 and 11 days were injected on day 3 and day 10, respectively. 1 Injected at days 3 and 5 and sacrificed on day 7. 3 Injected at days 3, 5 and 11, and sacrificed on day 14.
353
prepared from each animal which had received cytosine arabinoside. Although the lengths of these fibers were similar to those of control fibers, the numbers of nuclei per fiber were considerably lower than control values. This effect of cytosine arabinoside on the total number of nuclei per fiber is consistent with the interference with normal DNA synthesis in satellite cells. The extent of this reduction can only be stated qualitatively as the number of fibers counted were too few in number per animal to make a direct comparison with previous data (table 1). DISCUSSION
Individual striated muscle cells can contain several hundreds of nuclei. The guanidine technique for isolating individual muscle fibers has been employed to facilitate the study of how multinucleation arises during normal differentiation and growth. The earliest stage at which intact muscle fibers can be isolated in the mouse by this method is the gestational age of 19 days. The fibers at this time contain both muscle and satellite cell nuclei but their ultrastructure differs from adult muscle. It is likely that prior to this time the gastrocnemius muscle consisted mainly of mononucleated myoblasts and early myotubes. In the rat, many of the myoblasts begin fusing to form multinucleated myotubes at 16 days of gestation (Zelena, '62). Events in the development of the mouse are generally considered to occur at least a day in advance of those in the rat (Snell, '41). Thus, the first time fibers can be isolated by the guanidine technique in the mouse is consistent with the findings that at this time most of the myotubes have been formed and are beginning their early differentiation into muscle fibers. During differentiation the nuclei come to lie adjacent to the plasma membrane as more myofibrils are formed. The sharp decrease in nuclear number per unit segment (fig. 4) observed between 1 and 2 weeks of age in this study can be correlated with findings based on cinephotomicrography of marsupial muscle (Bridge and Allbrook, '70), which showed a decrease in the nuclear frequency along the fiber during development. A certain proportion of the nuclei counted in the isolated fibers were satellite
354
CONSTANCE A. CARDASIS A N D GEORGE W. COOPER
cell nuclei. Their percentages were determined in order to understand how the increase in numbers of fiber nuclei are related to satellite cells. The percentages of satellite cells are considered estimates, as they were based on morphological differences between satellite cell and muscle nuclei observable at the light microscopic level. However, the results agree closely with those of other investigators employing electron microscopy. At birth, mouse gastrocnemius fibers were estimated to contain 32% satellite cell nuclei, whereas the peroneus longus muscle contained 30% to 35% satellite cell nuclei (Allbrook et al., '71). In the present investigation the gastrocnemius fibers were estimated to contain 27% satellite cell nuclei at 1 week of age, whereas the mouse lumbrical muscle contained 30% (Schultz, '72). These percentages decrease to 6 % satellite cell nuclei in the gastrocnemius muscle of adult mice and a similar decrease was found in the peroneus longus ( 5 % ) (Allbrook, '71) and lumbrical muscle (7% ) (Schultz, '72). A theoretical model accounting for the continuing increases observed in the nuclear numbers of individual gastrocnemius muscle fibers during growth can be constructed on the basis of the results. An individual muscle fiber with its associated satellite cells constitutes a morphological unit at 19 days of gestation. During the two-day interval between this fetal age and birth, the number of nuclei per fiber almost doubles. This doubling may be explained by assuming fusion of the myoblasts and/or satellite cells with the muscle fiber or by assuming fusion of whole fibers with each other. Following birth, it is unlikely that fusion of whole fibers accounts for changes in nuclear numbers of length, since the rates of increase are too small to be accounted for by fusion of whole muscle cells. Furthermore, the number of fibers in a cross section of mouse muscle have not been observed to change between 1 day and 15 days of age (Kitiyakara and Angevine, '63). From birth to 2 weeks of age, one division of each satellite cell per week, with fusion of one of the daughter cells can account for the increase in nuclear number. Between 1 and 2 weeks, the greatest increase
in length of the fiber occurs, resulting in a decrease in the frequency of nuclei. Between 2 and 4 weeks, the rate of increase in nuclear number per fiber decreases as well as the number of satellite cells associated with individual muscle fibers. This indicates that there may be a high frequency of fusion of both satellite daughter cells with the muscle fiber. This fusion depletes the source of nuclei, thereby accounting for the observed decrease in the rate of nuclear increase. From 4 weeks to 9 weeks of age the total nuclear numbers and the fiber lengths increase slowly and the satellite cell population per fiber remains stable. Although the ultrastructure was not studied at 3 weeks, by 4 weeks the typical adult muscle ultrastructure was observed. This is consistent with the findings in the pectoralis major muscle of the rat that the T system does not form at the A-I junction and glycogen does not become limited to the interfibrillar space close to the sarcoplasmic reticulum until 20 days following birth (Yokota, '69). Based on this sequence, various procedures which affect mitotic division, cell fusion or cytoplasmic growth can be employed. Cytosine arabinoside, a drug which affects DNA synthesis, was employed during the first 2 weeks of age, a time of rapid increase in nuclear number in myofibers. The drug was used at this time to further test the hypothesis that the observed increase in nuclear number is a result of division and fusion of satellite cells. The occurrence of nuclear abnormalities in satellite cell nuclei (figs. 8, 9) indicate that DNA synthesis in satellite cells is altered by cytosine arabinoside treatment. Nuclear abnormalities following cytosine arabinoside treatment have also been described in other actively dividing cells such a s bone marrow (Telley and Vaitkevicus, '63) and crypt epithelial cells of the intestinal mucosa (Leach et al., '69). The nuclear abnormalities seen in the present investigation suggest that satellite cells are actively dividing during the first 2 weeks of life. Satellite cells often occur in pairs following cytosine arabinoside treatment, a condition not observed in normal develop-
NUCLEAR NUMBERS AND GROWTH OF MUSCLE FIBERS
ment. This suggests that the increase in percentage of satellite cell nuclei above control values does not result from an influx of cells from outside the basal lamina but from division of the satellite cells, without fusion of the daughter cells. It is conceivable that the interference with fusion may be due to the inability of the abnormal DNA molecule to transcribe the proper mRNA for the synthesis of molecules necessary for fusion. The failure to summate nuclear abnormalities with multiple drug doses in the animals examined on day 14 may be due to the fact that the animals had time to recover from the drug, as occurs with the crypt epithelial cells of the intestine which recover fully if the injection is not repeated more often than 72 hours (Leach, '69). Preliminary evidence indicates that changes in the lengths of the fibers were not affected by the drug, indicating that during the first week of age the addition of nuclei to the fiber is not necessary for cytoplasmic growth. However, the relationship between the addition of nuclei to the fiber and cytoplasmic growth is not clear beyond the first week due to the possibility that the cells recovered from the drug. In summary, the following points have been made: (1) The population of satellite cells associated with each fiber is large enough to account for the observed increases in the numbers of muscle nuclei from birth till adulthood. (2) Changes in nuclear number per fiber have been correlated with certain developmental events (i.e. birth, increases in fiber length). ( 3 ) Cytosine arabinoside, causes alterations in the morphology, distribution and percentage of nuclei in the satellite cell population, indicating that division and fusion of satellite cells is a factor in normal postnatal development of myofibers. (4) A hypothetical dynamic model of muscle fiber growth was constructed on the basis of a numerical analysis of nuclear numbers. LITERATURE CITED Allbrook, D. B., M. F. Han and A. E. Hellmuth 1971 Population of muscle satellite cells in r e lation to age and mitotic activity. Pathology, 3: 233-243. Bassleer, R. 1962 Etude de l'augmentation due nombre de noyaux dans dens bourgeons mus-
355
culaired cultives in vitro. Observations sur le vivant, dosages cytophotometriques et histoautoradiographies. 2. Anat. Entwick., 123: 184-205. Bridge, D. T., and D. Allbrook 1970 Growth of striated muscle in a n Australian marsupial (Se tonix brachyiris). J. Anat., 106: 283-293. Cardasis, C. A , , and G. W. Cooper 1975 A method for the chemical isolation of individual muscle fibers and its application to a study of the effect of denervation on the number of nuclei per muscle fiber. J. Exp. Zool., 191: 333346. Enesco, M., and D. Puddy 1964 Increase in number and weight in skeletal muscle of rats of various ages. Am. J. Anat., 114: 235-244. Godlewski 1902 Die Entwicklung des Skelettund Herzmuskel gewebes der Saugetiere. Arch. mickroskop. Anat. Entwicklungsmech., 60: 111. Grunberg, H. 1943 The development of some external features in mouse embryos. J. Hered., 34: 8 S 9 2 . Karlsson, U., and R. L. Schultz 1965 Fixation of the central nervous system for electron microscopy by aldehyde perfusion. I. Preservation with aldehyde perfusates versus direct perfusion with osmium tetroxide with special reference to membranes and the extracellular space. J . Ultrastruct. Res., 12: 160-186. Kitiyakara, A,, and D. M. Angevine 1963 A study of the pattern of postembryonic growth of M. gracilis in mice. Devel. Biol., 8: 322-340. Leach, W. B . , W . R. Laster, Jr., J. G . Mayo, Jr., D. P. Griswold, Jr. and F. M. Schabel, Jr. 1969 Toxicity studies in mice treated with 1-B-D Arabinofuranosyl cytosine (ara-C). Cancer Research, 29: 529-535. Loeffler, C. A. 1970 Evidence for the fusion of myoblasts in amphibian embryos. 11. Xenoplastic transplantation of somatic cells from anuran to urodele embryos. J. Morph., 130: 491-500. Mauro, A. 1961 Satellite cell of skeletal muscle fibers. J. Biophy. Biochem. Cyt., 9: 493-495. Mintz, B., and W. W. Baker 1967 Normal mammalian muscle differentiation and gene control of isocitrate dehydrogenase synthesis. Proc. Nat'l Acad. Sci., 58: 592-598. Montgomery, R. D. 1962 Growth of human striated muscle. Nature, 195: 194-195. Morpungo, B. 1898 Uber die postembryonale Entwicklung der quergestrieften Muskeln von weissen Ratten. Anat. Anz., 15: 300-306. Moss, F. P., and C. P. Leblond 1970 Nature of dividing nuclei in skeletal muscle of growing rats. J. Cell Biol., 44: 4 5 9 4 6 2 . Okazaki, K., and H. Holtzer 1966 Myogenesis: Fusion, myosin synthesis and the mitotic cycle. Proc. Nat'l Acad. Sci., 56: 1484-1490. Reynolds, E. S. 1963 The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. J. Cell Biol., 17: 210-213. Reznik, M. 1969 Thymidine3H uptake by satellite cells of regenerating skeletal muscle. J. Cell Biol., 40: 568-571. Richardson, K. C., L. Jarett and E. H. Finke 1960 Embedding in epoxy resins for ultrathin sectioning in electron microscopy. Stain Tech., 35: 3 1 3-323. Schiefirdecker, P. 1909 Muskeln und Muskelkern. Leipzig.
356
CONSTANCE A. CARDASIS AND GEORGE W. COOPER
Schultz, E. 1972 Changing satellite cell populations in skeletal muscle of the neonatal mouse. Anat. Rec., 172: 401. S h d q , S. A,, M. A. Gorycki a n d A . Mauro 1968 Mitosis during postnatal growth in skeletal and cardiac muscle of the rat. J. Anat., 103: 135-141. Snell, G. D. 1941 T h e early embryology of the mouse. In: Biology of the Laboratory Mouse. G . D. Snell, ed. Dover Publ., New York, Chap. I , 1-54. Telley, R. W., and V. K . Vaitkevicus 1963 Mega-
loglastosis produced by a cytosine antagonist I-B-D arabinofuranosyl-cytosine. Blood, 21 : 352362. Yokota, S. 1969 Electron microscopic studies on the development of the skeletal muscle (pectoralis major muscle) of the albino rats. Okajima Folia Anat. Jap., 4 5 : 309-333. Zelena, J. 1962 The effect of denervation on muscle development. I n : The Denervated Muscle. E. Gutman, ed. Publ. House of Czeck. Acad. Sci., Prague, Chap. 111, 103-126.
PLATE 1 EXPLANATION
OF FIGURES
8
Phase contrast photomicrograph of gastrocnemius muscle from mouse which received 0.2 mglkg cytosine arabinoside o n third day of age and was killed o n day 4. Note occurrence of two pairs of satellite cells (large arrows) and abnormal satellite cell nucleus (small arrow). X 4,000.
9
Phase contrast photomicrographs of gastrocnemius muscle from mouse which received 0.2 mg/kg cytosine arabinoside on third day of age and was killed on day. 4. Note dumbbell shaped nucleus in satellite cell (arrow). x 4,000.
NUCLEAR NUMBERS A N D GROWTH OF MUSCLE FIBERS Constance A. Cardasis and George W . Cooper
PLATE 1
357
ABSTRACT A numerical analysis of changes in the populations of nuclei in individual, intact muscle fibers was made to study how multinucleation arises during normal differentiation and growth. Gastrocnemius muscle fibers from pre- and post-natal mice were isolated with guanidine (Cardasis and Cooper, '75) and examined. Satellite cells associated with muscle fibers were first observed at 19 days of gestation. The number of nuclei per muscle fiber (muscle satellite cell nuclei) averages 83 at this age, 157 at birth and continues to increase to 354 by 63 days of age. However, the rate of increase during growth is not constant. Estimates of satellite cell and muscle nuclei in histological cross sections indicate that there is a decrease in the percentage of satellite cells from 32% at birth to 6% in the adult. However, the numbers of satellite cells associated with individual muscle fibers, calculated from these percentages and the nuclear counts on whole fibers, decreases only between 2 and 4 weeks of age. Cytosine arabinoside was injected subcutaneously during the first two weeks of age. Pairs of satellite cells, abnormal nuclei and elevated percentages of satellite cells were observed. This evidence as well as the numerical analysis of nuclear populations in whole fibers lends further support to the hypothesis that satellite cells account for the increase in muscle nuclei from birth to maturity.
+
How multinucleation occurs in muscle cells during normal differentiation and growth has not been fully explained. During the early stages of differentiation, mononucleated myoblasts fuse to form multinucleated myotubes. Myotubes contain rounded, centrally located nuclei aligned end to end, which are surrounded by myofilaments. The fusion is well documented by transplantation experiments (Loeffler, '70), the use of allophenic mice (Mintz and Baker, '67) and photography of fusion in living cultures (Bassleer, '62). It seems likely that, in tissue culture, DNA synthesis is limited to mononuclear myoblasts. Okazaki and Holtzer ('66), employing both 3H-thymidine autoradiography and fluorescent antibodies demonstrated that nuclei in myotubes in which myosin is detected do not incorporate "-thymidine for DNA synthesis and that myoblasts which incorporate "-thymidine do not contain myosin. J . EXP.ZOOL., 191: 347-358.
Myotubes initiate the process of differentiation into myofibers prior to birth. Following birth there are a number of changes still occurring in the morphology of the muscle fibers as well as an increase in size of muscle fibers. Whether muscle nuclei also increase in number during postnatal growth has been controversial. Some early investigators found no increase in the number of nuclei with age (Godlewski, '02) while others reported an increase (Morpungo, 1898; Schiefirdecker, '09). More recent investigations indicate that the muscle cells continue to grow and the muscle nuclei continue to increase in number following birth (Kitiyakara and Angevine, '63). Nuclear counts based on 1 Supported in part by a grant of the Institute of General Medical Sciences, GM 15289. 2 This paper is based on a dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Faculty of Pure Science, Columbia University. 8 Present address: Department of Anatomy, Downstate Medical Center, Brooklyn, New York 11203.
34 7
348
CONSTANCE A . CARDASIS A N D GEORGE W . COOPER
histological sections (Montgomery, '62), two mice at each of the following periods chemical determinations of DNA content of development: the 19th day of gestation of whole muscle (Enesco and Puddy, '64), (day - 2 ) , birth (day 0) and 1, 2, 4, 5, 6, cinephotomicrographic analysis of serial 7 and 9 weeks of age. cross-sections (Bridge and Allbrook, '70) The day of gestation of fetal stages was and the analysis of electron microscopic determined by examination of age dependmontages of muscle cross-sections (Schultz, ent external characteristics of the fetus '72) all demonstrate an increase in the (Grunberg, '42). The lower right limbs number of muscle nuclei during post-natal were dissected from unanesthetized fetuses growth. This raises the question of how at 19 days of gestation and placed intact this increase is occurring, when the mus- into 1% glutaraldehyde (Polysciences) in cle nuclei are thought to have lost their a 0.1 M phosphate buffer, pH 7.4 (Karlsability to undergo DNA synthesis and son and Schultz, '65). Animals killed at divide. birth to one week of age were sacrificed by Satellite cells are mononucleated cells decapitation, older animals by etherization. which lie between the muscle basal lamina The lower right legs from all animals were and plasma membrane (Mauro, '61). They skinned and initially placed in the above may be myoblasts which were incorporated fixative and the gastrocnemius muscles within the basement membrane of the were removed from the bones. The total muscle without fusion with the muscle time of fixation for the intact muscle was plasma membrane or they may be derived one hour. Fixed muscles were stored in from cells penetrating the basement mem- 0.1 M phosphate buffer (pH 7.4). Groups brane later in life. They have the ability of fibers from the margins of these musto undergo DNA synthesis (Moss and Le- cles were dissected free from the muscle blond, '70) and mitosis (Shafq et al., '68) and individual fibers were isolated in a during growth and during regeneration solution of 0.02 M guanidine-HC1 (Car(Reznik, '69). These cells may serve as a dasis and Cooper, '75). The remainder of source of reserve myoblasts in adult life. the muscle was prepared for light and The present investigation presents a electron microscopy. The semi-thin light description of the increase in nuclei during microscopic sections were used to deterpre- and post-natal development until sex- mine the satellite cell percentages (see ual maturity. It is based on nuclear counts RESULTS for criteria). made on individual fibers which have been The results of the counts of nuclear isolated by the guanidine technique (Car- number per fiber are based on the numerdasis and Cooper, '75). This procedure al- ical average of the counts of eight individlows a more direct determination of nu- ual fibers from two animals (4 fibers from clear numbers as it provides intact muscle each animal) for each time period. An fibers with their satellite cells but free of analysis of the standard deviations from their connective tissue. the means was performed on the numerSpecifically, this investigation answers ical count data. the following questions: (1) Is there a n inCytosine arabinoside treatment of growcrease in the total number of nuclei per ing animals. Eight growing mice, of which fiber during post-natal growth? (2) At any four served as controls, were used in this given age, what percentage of the isolated study. The times chosen for treatment fiber's nuclei are satellite cells? (3) When were based on the critical periods of nudo increases in nuclear number occur and clear increase obtained from the study of what are the rates of increase during spe- normal growth. Cytosine arabinoside (Sigcific time intervals? ma) was employed at a dose of 0.002 mg/ gm body weight. One animal received a MATERIALS AND M E T H O b S subcutaneous injection of cytosine arabinoMaterials for the study of normal growth. side at three days of age and was killed The growth of muscle was studied in white by ether at four days of age. Another animice of the Princeton-Rockefeller strain. mal received cytosine arabinoside at ten This strain is sexually mature at approxi- days of age and was killed at 11 days. The mately six weeks of age. The right gas- remaining animals received multiple introcnemius muscles were obtained from jections of cytosine arabinoside (each in-
NUCLEAR NUMBERS AND GROWTH OF MUSCLE FIBERS
jection was 0.002 mg/gm body weight). One received the drug at 2 and 5 days of age and was killed on day 7. Another received the drug at 2, 5 and 11 days of age and was killed at 14 days of age. Control animals were uninjected littermates killed by ether at the same ages as the experimental animals. The gastrocnemius muscles from all animals were fixed in 1% glutaraldehyde as described previously and prepared for light and electron microscopy. The light microscopic sections were used to determine the percentages of normal and abnormal appearing muscle and satellite cell nuclei. Electron microscopy was used to verify the location of abnormal nuclei. Preparation of light and electron microscopic sections Blocks of tissue were obtained and fixed in 1% glutaraldehyde by the identical method employed for the whole fibers. Following a wash in 0.1 M phosphate buffer all tissue to be embedded were post-fixed one hour in 1 % osmium tetroxide in 0.1 M phosphate buffer (pH 7.4). They were then dehydrated in a graded series of ethyl alcohols to absolute ethanol and in propylene oxide. The tissues were infiltrated with Epon 812 and embedded. Both the semithin (0.5-1 .O micron) and thin (50-70 nm) sections were cut on an LKB Ultratome. The semi-thin sections were stained with the Azure 11-methylene blue method of Richardson ('60) modified as follows: 1% periodic acid was used for ten minutes (instead of 5 minutes) and the staining solution was diluted to half of the original concentration. Heating of the sections during staining was not performed. These sections were used for determining the percentage of satellite cells, which contained the more heterochromatic nuclei as compared with muscle nuclei. The percentage of satellite cells was calculated from 200 nuclei (muscle and satellite cell nuclei) counted in two blocks of muscle tissue from each animal (total 400 nuclei/ animal). The number of satellite cell nuclei within the basement membrane of individual muscle cells was calculated by multiplying these percentages of satellite cells by the average of the total nuclei counted in isolated fibers. Phase photomicrographs of light micro-
349
scopic sections were taken on a Reicher microscope with Kodak Panatomic X film and developed in Microdol X, diluted 1 to 3. The thin sections for electron microscopy were placed on copper grids and stained with a filtered solution of saturated aqueous uranyl acetate and lead citrate (Reynolds, '63). Sections were examined and photographed on an RCA EMU3-G electron microscope at 50 KV. RESULTS
Light and electron microscopic observations. There are differences in the structure of young and mature mouse gastrocnemius muscle. Some of the structural changes which can be correlated with the changes in nuclear numbers during growth will be described. At a gestational age of 19 days (fig. 1A) and at birth (figs. l B , 2), the cross striations are not as distinct as in adult muscle due to the fact that all the myofibrils are not aligned. Many lipid droplets are present and glycogen is abundant and diffusely scattered throughout the sacroplasm. Muscle fibers have satellite cells associated with them by the 19th day of gestation. The cytoplasm of satellite cells associated with muscle fibers contain many free ribosomes and mitochondria. Their nuclei contain a large amount of heterochromatin near the nuclear envelope. In contrast, the muscle nuclei contain mostly euchromatin and have very large, distinct nucleoli (fig. 2). These two nuclear characteristics are the basis for distinguishing satellite from muscle nuclei with the light microscope. The muscle and satellite cell nuclei seen in whole mounts of young isolated fibers appear oval (figs. lA,B). By four weeks (fig. 1C) the nuclei have the elongated or fusiform appearance of adult muscle nuclei. Electron microscopy demon strates that by four weeks of age the internal structure of the muscle fiber looks very much like that of the adult. The cross striations are distinct, the lipid droplets are rare and glycogen particles are limited to the interfibrillar spaces near the sarcoplasmic reticulum. Changes in nuclear numbers in isolated fibers. Nuclei counted in fibers isolated by the guanidine-HC1 technique include both muscle nuclei and the satellite cell
350
CONSTANCE A. CARDASIS AND GEORGE W. COOPER
35 1
NUCLEAR NUMBERS AND GROWTH OF MUSCLE FIBERS TABLE 1
N o r m a l growth of mouse gastrocnemius muscle fibers ~
Length of fiber in mm? S.D.
No. of nuclei (muscle and satellite cell)/ fiber? S . D .
0.90 2 0.05 1.23 t 0.23 1.53 t 0.27 2.98 2 0.22 4.00 2 0.32 4.29 2 0.37 4.52 f 0.34 4.76 f 0.34 4.42 f 0.43 5.20 C 0.41
83 t 11.0 1 5 7 t 42.0 199 t 32.4 241 t 23.6 276 f 26.4 302 f 36.1 312 i25.1 341 i29.5 355 f 49.7 349 i 32.3
Age in days
relative to birth
-2
~~~~
~~
No. of nuclei (muscle and satellite cell)/ 170 @ ? S.D.
No. of sateilite cells per fiber
Yo 2
02
72 14 28 35 42 49 63 6 mos.-1
yr.3
15.8 f 2 . 1 21.6 f 2.4 22.7 f 5.3 13.8f 1.2 11.820.7 12.0 t 0.8 11.8f0.7 12.2 1.0 13.6 t 0.8 11.5 2 0.5
50 (32) 55 (27) 48 (20) 28 (10) 15 (5) 22 (7) 25 (7) 21 (6)
*
' Each average i s based on an analysis of eight fibers from two individual animals.
Averages are based o n an analysis of less than eight fibers from two individual animals, Averages are based on an analysis of 21 animals (4 fibers from each animal). ( 1 o/o satellite cell nuclei of all nuclei within muscle basal lamina as observed in cross sections.
3
nuclei associated with individual fibers. The number of nuclei per fiber increases from 84 nuclei per fiber at the gestational age of 19 days to 354 nuclei per fiber at 63 days of age (fig. 3, table 1). The rate of increase can be divided into three distinct intervals. The greatest rate of increase occurs in the two day interval between the gestational age of 19 days (84 nucIei) and birth (158 nuclei), when the number of nuclei per fiber almost doubles. The next largest increase in total number of nuclei occurs between birth, seven days (208 nuclei) and 14 days (241 nuclei) of age. After 14 days the rate of increase in nuclei per fiber becomes progressively smaller. Changes in the length of isolated fibers. While the nuclei are increasing in number during growth, the size of the fiber is also increasing. The increase in the latFig. 1 A. Photomicrograph of gastrocnemius muscle fiber isolated in guanidine from a mouse embryo of a gestational age of 19 days. Oval nuclei (arrow) and some striations are present. x 1,600. B. Photomicrograph of gastrocnemius muscle fiber isolated in guanidine from mouse a t birth. Nuclei (large arrow) and lipid droplets (small arrow) can be seen. X 1,600. C. Photomicrograph of mouse gastrocnemius fiber isolated in guanidine at four weeks of age. The area of tendon attachment (t) is demonstrated. The muscle nuclei are elongated (arrow), rather than round to oval as in younger muscle. X 1,600. Fig. 2 Electron micrograph of mouse gastrocnemius muscle a t birth. Note heterochromatic satellite cell nucleus (sc), euchromatic muscle nuclei (m) and lipid droplets in muscle cytoplasm (arrow). x 6,100.
"1
I
I
0
Blrtn
10
,
20
30
,
40
,
50
I
60
DnYS of CGE
Fig. 3 Graph showing an increase in the number of nuclei per muscle fiber with age. Brackets indicate standard deviation.
era1dimension is apparent in whole mounts (figs. lA,B,C) but was not measured. The lengths of each of the fibers was measured (fig. 4, table 1). The greatest increase in length occurs between days 7 and 14 (1.53 mm vs. 2.98 mm). This is approximately 40% of the total increase in length. The increases in nuclear number and in lengths of the fibers during growth do not always occur at the same rate. The numbers of nuclei per fiber for a given age as seen in figure 3 was divided by the number of 170 micron segments in the fiber
CONSTANCE A. CARDASIS AND GEORGE W . COOPER
of nuclei and the rate of increase in length are closely correlated events. Satellite cell numbers. Percentages of satellite cell nuclei were determined on histological cross-sections of muscle (fig. 6, table 1). The percentage of satellite cells could not be determined with any degree of confidence in fetal muscles. The results demonstrate that there is an increase in the percentage of satellite cell nuclei from 32% at birth to 10% at 4 weeks of age. From this time till 63 days of age the percentage of satellite cells only decreased to -2 :O I0 i0 30 40 50 $0 6%. BIRTH AGE IN DAYS The percentage of satellite cells estimated by counts made on histological secFig. 4 Changes in the length of muscle fibers as related to age. Brackets indicate standard detions were related to the numbers of nuviation. clei found in individual fibers, isolated by the guanidine method. The total number (fig. 5, table 1). This calculation was done of satellite cell nuclei per individual musto determine the relationship between the cle fiber was calculated by multiplying the increase in the number of nuclei and the estimated percentages by the average of increase in length of the fiber. The in- the total number of nuclei per fiber (fig. crease in nuclei per unit length increases 6, table 1). Although the percentages of from the gestational age of 19 days (day satellite cells are decreasing, these calcu- 2) ( 15 nuclei/ 170 micron segment) to lated absolute values do not differ greatly birth (day 0) (21.6 nuc€ei/l70 segment) from birth through 14 days. There is a due to an increase in the number of nu- decrease between 14 and 28 days of age. clei with little increase in fiber length. However, following this age, the number This high number per standard of length of satellite cell nuclei per fiber again reis maintained during the first seven post- mains constant. natal days. There is a sharp reduction beCytosine arabinoside treatment of growtween 7 and 14 days due to the great ini n g animals. Injection of cytosine aracrease in fiber length. From 14 to 63 days of age, there is only slight fluctuation in binoside on day 3 resulted in an elevated this value, with overlapping standard de400 viations. This indicates that from 2 weeks of age the rate of increase in the number 3SO ,
I
,
w K
m T
300
w
2
10
250
I
zoo u 3 150 (r w
g
100
z 3
50
2
,
,
I0
I
$0
33
40
2 Q
.
50
60
Birth
D A Y S of AGE
Fig. 5 The frequency of nuclei per 170 micron segment of muscle fiber as related to age. Brackets indicate standard deviation.
Birth
7
14
21
56
DAYS of AGE
Fig. 6 The number of nuclei per muscle fiber as related to age. Blackened areas indicate the % of nuclei which are satellite cell nuclei within the basement membrane of a musclefiber.
NUCLEAR NUMBERS A N D GROWTH OF MUSCLE FIBERS
percentage of satellite cells (43%) as compared to control animals (25 % ) on day 4. Similar elevated values for the percentage of satellite cells were found regardless of whether the animals received multiple or single injections during the first 11 days post-natal (fig. 7, table 2). Satellite cells in close proximity with one another are frequently seen in animals treated with cytosine arabinoside (figs. 8, 9). In addition, certain satellite cells appear binucleate or possibly contain nuclei which have constricted or are dumbbell shape (fig. 9). These abnormalities are not seen in muscle cell nuclei. The percentage of satellite cells which exhibit these nuclear abnormalities comprise approximately 20 % of the satellite cell population or 8% of the total nuclear population. Whole mounts of individual fibers were
-
$ 1 Y
5
loo
CONTROL
CYTOSINE ARABINOSIOE
-X
t3
80
I
2
3
4
MULTIPLE INJECTIONS
5 6 7 8 9 1 0 1 l 1 2 1 3 1 4 AGE IN DAYS
Fig. 7 Effect of 0.2 mg/kg of cytosine arabinoside on the percentage of nuclei within the basal lamina of muscle fibers which are satellite cell nuclei compared with control values. TABLE 2
Cytosine arabinoside t r e a t m e n t during jirst t w o w e e k s of growth
Age in days relative to birth
Cytosine arabinoside (0.2 mglkg) Satellite cells
Satellite cellsabnormal
Control
Satellite cells
Satellite cellsmitotic figures
70
70
70
%
19
7 11 1
43 39 40
26 20
14
28
14
25 27 26 17
4 4 4 0
4
*
Animals killed at 4 and 11 days were injected on day 3 and day 10, respectively. 1 Injected at days 3 and 5 and sacrificed on day 7. 3 Injected at days 3, 5 and 11, and sacrificed on day 14.
353
prepared from each animal which had received cytosine arabinoside. Although the lengths of these fibers were similar to those of control fibers, the numbers of nuclei per fiber were considerably lower than control values. This effect of cytosine arabinoside on the total number of nuclei per fiber is consistent with the interference with normal DNA synthesis in satellite cells. The extent of this reduction can only be stated qualitatively as the number of fibers counted were too few in number per animal to make a direct comparison with previous data (table 1). DISCUSSION
Individual striated muscle cells can contain several hundreds of nuclei. The guanidine technique for isolating individual muscle fibers has been employed to facilitate the study of how multinucleation arises during normal differentiation and growth. The earliest stage at which intact muscle fibers can be isolated in the mouse by this method is the gestational age of 19 days. The fibers at this time contain both muscle and satellite cell nuclei but their ultrastructure differs from adult muscle. It is likely that prior to this time the gastrocnemius muscle consisted mainly of mononucleated myoblasts and early myotubes. In the rat, many of the myoblasts begin fusing to form multinucleated myotubes at 16 days of gestation (Zelena, '62). Events in the development of the mouse are generally considered to occur at least a day in advance of those in the rat (Snell, '41). Thus, the first time fibers can be isolated by the guanidine technique in the mouse is consistent with the findings that at this time most of the myotubes have been formed and are beginning their early differentiation into muscle fibers. During differentiation the nuclei come to lie adjacent to the plasma membrane as more myofibrils are formed. The sharp decrease in nuclear number per unit segment (fig. 4) observed between 1 and 2 weeks of age in this study can be correlated with findings based on cinephotomicrography of marsupial muscle (Bridge and Allbrook, '70), which showed a decrease in the nuclear frequency along the fiber during development. A certain proportion of the nuclei counted in the isolated fibers were satellite
354
CONSTANCE A. CARDASIS A N D GEORGE W. COOPER
cell nuclei. Their percentages were determined in order to understand how the increase in numbers of fiber nuclei are related to satellite cells. The percentages of satellite cells are considered estimates, as they were based on morphological differences between satellite cell and muscle nuclei observable at the light microscopic level. However, the results agree closely with those of other investigators employing electron microscopy. At birth, mouse gastrocnemius fibers were estimated to contain 32% satellite cell nuclei, whereas the peroneus longus muscle contained 30% to 35% satellite cell nuclei (Allbrook et al., '71). In the present investigation the gastrocnemius fibers were estimated to contain 27% satellite cell nuclei at 1 week of age, whereas the mouse lumbrical muscle contained 30% (Schultz, '72). These percentages decrease to 6 % satellite cell nuclei in the gastrocnemius muscle of adult mice and a similar decrease was found in the peroneus longus ( 5 % ) (Allbrook, '71) and lumbrical muscle (7% ) (Schultz, '72). A theoretical model accounting for the continuing increases observed in the nuclear numbers of individual gastrocnemius muscle fibers during growth can be constructed on the basis of the results. An individual muscle fiber with its associated satellite cells constitutes a morphological unit at 19 days of gestation. During the two-day interval between this fetal age and birth, the number of nuclei per fiber almost doubles. This doubling may be explained by assuming fusion of the myoblasts and/or satellite cells with the muscle fiber or by assuming fusion of whole fibers with each other. Following birth, it is unlikely that fusion of whole fibers accounts for changes in nuclear numbers of length, since the rates of increase are too small to be accounted for by fusion of whole muscle cells. Furthermore, the number of fibers in a cross section of mouse muscle have not been observed to change between 1 day and 15 days of age (Kitiyakara and Angevine, '63). From birth to 2 weeks of age, one division of each satellite cell per week, with fusion of one of the daughter cells can account for the increase in nuclear number. Between 1 and 2 weeks, the greatest increase
in length of the fiber occurs, resulting in a decrease in the frequency of nuclei. Between 2 and 4 weeks, the rate of increase in nuclear number per fiber decreases as well as the number of satellite cells associated with individual muscle fibers. This indicates that there may be a high frequency of fusion of both satellite daughter cells with the muscle fiber. This fusion depletes the source of nuclei, thereby accounting for the observed decrease in the rate of nuclear increase. From 4 weeks to 9 weeks of age the total nuclear numbers and the fiber lengths increase slowly and the satellite cell population per fiber remains stable. Although the ultrastructure was not studied at 3 weeks, by 4 weeks the typical adult muscle ultrastructure was observed. This is consistent with the findings in the pectoralis major muscle of the rat that the T system does not form at the A-I junction and glycogen does not become limited to the interfibrillar space close to the sarcoplasmic reticulum until 20 days following birth (Yokota, '69). Based on this sequence, various procedures which affect mitotic division, cell fusion or cytoplasmic growth can be employed. Cytosine arabinoside, a drug which affects DNA synthesis, was employed during the first 2 weeks of age, a time of rapid increase in nuclear number in myofibers. The drug was used at this time to further test the hypothesis that the observed increase in nuclear number is a result of division and fusion of satellite cells. The occurrence of nuclear abnormalities in satellite cell nuclei (figs. 8, 9) indicate that DNA synthesis in satellite cells is altered by cytosine arabinoside treatment. Nuclear abnormalities following cytosine arabinoside treatment have also been described in other actively dividing cells such a s bone marrow (Telley and Vaitkevicus, '63) and crypt epithelial cells of the intestinal mucosa (Leach et al., '69). The nuclear abnormalities seen in the present investigation suggest that satellite cells are actively dividing during the first 2 weeks of life. Satellite cells often occur in pairs following cytosine arabinoside treatment, a condition not observed in normal develop-
NUCLEAR NUMBERS AND GROWTH OF MUSCLE FIBERS
ment. This suggests that the increase in percentage of satellite cell nuclei above control values does not result from an influx of cells from outside the basal lamina but from division of the satellite cells, without fusion of the daughter cells. It is conceivable that the interference with fusion may be due to the inability of the abnormal DNA molecule to transcribe the proper mRNA for the synthesis of molecules necessary for fusion. The failure to summate nuclear abnormalities with multiple drug doses in the animals examined on day 14 may be due to the fact that the animals had time to recover from the drug, as occurs with the crypt epithelial cells of the intestine which recover fully if the injection is not repeated more often than 72 hours (Leach, '69). Preliminary evidence indicates that changes in the lengths of the fibers were not affected by the drug, indicating that during the first week of age the addition of nuclei to the fiber is not necessary for cytoplasmic growth. However, the relationship between the addition of nuclei to the fiber and cytoplasmic growth is not clear beyond the first week due to the possibility that the cells recovered from the drug. In summary, the following points have been made: (1) The population of satellite cells associated with each fiber is large enough to account for the observed increases in the numbers of muscle nuclei from birth till adulthood. (2) Changes in nuclear number per fiber have been correlated with certain developmental events (i.e. birth, increases in fiber length). ( 3 ) Cytosine arabinoside, causes alterations in the morphology, distribution and percentage of nuclei in the satellite cell population, indicating that division and fusion of satellite cells is a factor in normal postnatal development of myofibers. (4) A hypothetical dynamic model of muscle fiber growth was constructed on the basis of a numerical analysis of nuclear numbers. LITERATURE CITED Allbrook, D. B., M. F. Han and A. E. Hellmuth 1971 Population of muscle satellite cells in r e lation to age and mitotic activity. Pathology, 3: 233-243. Bassleer, R. 1962 Etude de l'augmentation due nombre de noyaux dans dens bourgeons mus-
355
culaired cultives in vitro. Observations sur le vivant, dosages cytophotometriques et histoautoradiographies. 2. Anat. Entwick., 123: 184-205. Bridge, D. T., and D. Allbrook 1970 Growth of striated muscle in a n Australian marsupial (Se tonix brachyiris). J. Anat., 106: 283-293. Cardasis, C. A , , and G. W. Cooper 1975 A method for the chemical isolation of individual muscle fibers and its application to a study of the effect of denervation on the number of nuclei per muscle fiber. J. Exp. Zool., 191: 333346. Enesco, M., and D. Puddy 1964 Increase in number and weight in skeletal muscle of rats of various ages. Am. J. Anat., 114: 235-244. Godlewski 1902 Die Entwicklung des Skelettund Herzmuskel gewebes der Saugetiere. Arch. mickroskop. Anat. Entwicklungsmech., 60: 111. Grunberg, H. 1943 The development of some external features in mouse embryos. J. Hered., 34: 8 S 9 2 . Karlsson, U., and R. L. Schultz 1965 Fixation of the central nervous system for electron microscopy by aldehyde perfusion. I. Preservation with aldehyde perfusates versus direct perfusion with osmium tetroxide with special reference to membranes and the extracellular space. J . Ultrastruct. Res., 12: 160-186. Kitiyakara, A,, and D. M. Angevine 1963 A study of the pattern of postembryonic growth of M. gracilis in mice. Devel. Biol., 8: 322-340. Leach, W. B . , W . R. Laster, Jr., J. G . Mayo, Jr., D. P. Griswold, Jr. and F. M. Schabel, Jr. 1969 Toxicity studies in mice treated with 1-B-D Arabinofuranosyl cytosine (ara-C). Cancer Research, 29: 529-535. Loeffler, C. A. 1970 Evidence for the fusion of myoblasts in amphibian embryos. 11. Xenoplastic transplantation of somatic cells from anuran to urodele embryos. J. Morph., 130: 491-500. Mauro, A. 1961 Satellite cell of skeletal muscle fibers. J. Biophy. Biochem. Cyt., 9: 493-495. Mintz, B., and W. W. Baker 1967 Normal mammalian muscle differentiation and gene control of isocitrate dehydrogenase synthesis. Proc. Nat'l Acad. Sci., 58: 592-598. Montgomery, R. D. 1962 Growth of human striated muscle. Nature, 195: 194-195. Morpungo, B. 1898 Uber die postembryonale Entwicklung der quergestrieften Muskeln von weissen Ratten. Anat. Anz., 15: 300-306. Moss, F. P., and C. P. Leblond 1970 Nature of dividing nuclei in skeletal muscle of growing rats. J. Cell Biol., 44: 4 5 9 4 6 2 . Okazaki, K., and H. Holtzer 1966 Myogenesis: Fusion, myosin synthesis and the mitotic cycle. Proc. Nat'l Acad. Sci., 56: 1484-1490. Reynolds, E. S. 1963 The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. J. Cell Biol., 17: 210-213. Reznik, M. 1969 Thymidine3H uptake by satellite cells of regenerating skeletal muscle. J. Cell Biol., 40: 568-571. Richardson, K. C., L. Jarett and E. H. Finke 1960 Embedding in epoxy resins for ultrathin sectioning in electron microscopy. Stain Tech., 35: 3 1 3-323. Schiefirdecker, P. 1909 Muskeln und Muskelkern. Leipzig.
356
CONSTANCE A. CARDASIS AND GEORGE W. COOPER
Schultz, E. 1972 Changing satellite cell populations in skeletal muscle of the neonatal mouse. Anat. Rec., 172: 401. S h d q , S. A,, M. A. Gorycki a n d A . Mauro 1968 Mitosis during postnatal growth in skeletal and cardiac muscle of the rat. J. Anat., 103: 135-141. Snell, G. D. 1941 T h e early embryology of the mouse. In: Biology of the Laboratory Mouse. G . D. Snell, ed. Dover Publ., New York, Chap. I , 1-54. Telley, R. W., and V. K . Vaitkevicus 1963 Mega-
loglastosis produced by a cytosine antagonist I-B-D arabinofuranosyl-cytosine. Blood, 21 : 352362. Yokota, S. 1969 Electron microscopic studies on the development of the skeletal muscle (pectoralis major muscle) of the albino rats. Okajima Folia Anat. Jap., 4 5 : 309-333. Zelena, J. 1962 The effect of denervation on muscle development. I n : The Denervated Muscle. E. Gutman, ed. Publ. House of Czeck. Acad. Sci., Prague, Chap. 111, 103-126.
PLATE 1 EXPLANATION
OF FIGURES
8
Phase contrast photomicrograph of gastrocnemius muscle from mouse which received 0.2 mglkg cytosine arabinoside o n third day of age and was killed o n day 4. Note occurrence of two pairs of satellite cells (large arrows) and abnormal satellite cell nucleus (small arrow). X 4,000.
9
Phase contrast photomicrographs of gastrocnemius muscle from mouse which received 0.2 mg/kg cytosine arabinoside on third day of age and was killed on day. 4. Note dumbbell shaped nucleus in satellite cell (arrow). x 4,000.
NUCLEAR NUMBERS A N D GROWTH OF MUSCLE FIBERS Constance A. Cardasis and George W . Cooper
PLATE 1
357
