Satellite and invasive cells in frog sartorius muscle

TISSUE

& CELL

1978 10 (4) 749-772

Puhlishrd by Longman Crortp Ltd. Printed in Great Britain

ESTEL.A CASTILLO de MARUENDA” and CLARA FRANZINI-ARMSTRONG

SATELLITE SARTORIUS

AND INVASIVE MUSCLE

CELLS

IN FROG

ABSTRACT. The occurrence and distribution of two cell types associated with normal and denervated frog skeletal muscle fibers are described. The first is the satellite cell. The general appearance and the number of satellite cells are not affected by long-term denervation. The second type of cell is the invasive cell. Invasive cells penetrate across the basal lamina and up to the core of the muscle fiber, without fusing with it. It is suggested that the origin of invasive cells is extramuscular, probably circulatory. Although invasive cells are more numerous in some denervated muscle, it is established that this is not a direct effect of denervation.

Introduction

satellite cells of skeletal muscle lie between the basal lamina and plasma membrane of the extra- and intrafusal muscle fibers (Mauro, 1961; Katz, 1961; Karlsson et al., 1966; Venable, 1966; Allbrook et al., 1971). A narrow, 20 nm gap separates the opposed membranes of satellite cells and muscle fibers: the two cells are not continuous with each other (Muir et al., 1965). Satellite cells of normal adult muscle were initially described as having scanty cytoplasm and they are thought to be undifferentiated, dormant cells. More recent evidence has suggested that size and shape of satellite cells may vary considerably, particularly during development (Ishikawa, 1970; Schultz, 1976) but also in totally differentiated muscles (Lee, 1965; Flood, 1971; Kryvi, 1975; Kelly, 1978a). It is evident that the initial descriptions of satellite cells structure need to be reconsidered, to take into account these variabilities. The aim of this study was to provide a comprehensive and detailed qualitative and quantitative description of satellite cells in THE

the sartorius muscle of frogs. This is to serve as a baseline for future interpretation of the effects of pathological processes and experimental interventions on satellite cells of this muscle. In addition, we describe cells that are part of a transient population. These are seen in the process of crossing the basal lamina and/or moving into the core of the muscle fiber, without fusing with it. Cells of this type are found rarely in muscles from normal frog and more frequently, but not constantly, in muscles from frogs subjected to unilateral leg denervation. The relationships of these cells to satellite cells and to cells belonging to the interstitium are discussed. Materials and Methods

Adult frogs (Rana pipirns), weighing 30--40 g were kept under several conditions: (I) sacrificed immediately;? (2) maintained up to 30 days at room temperature, fed with crushed liver once a week and forced to swim for a few minutes on alternate days; (3) maintained at room temperature, without feeding or activity for up to 60 days; (4) denervated

Departments of Biology and Anatomy, University of Pennsylvania, Philadelphia, Pa. 19104. * Present address: Institute de Parologia. Apartado 705, Lima 1, Peru.

t Commercial frog suppliers usually provide freshly captured frogs during the summer, and frogs which have been maintained in the refrigerator for some period of time during the winter. Thus frogs in group (1) were ‘fresh’ only during the summer.

Received 22 May 1978. 749

MARUENDA

750

by transecting a short segment of sciatic nerve at the level of the hip and maintained for up to 90 days, as in (2). Failure of neuromuscular transmission in this group was checked by stimulating the distal nerve stump and observing the muscle for twitching. Some of the late muscles showed a faint response, i.e. they were in the process of reinnervation. These are not considered here. Sartorii were fixed in 3 “//,glutaraldehyde in 0.1 cacodylate buffer (pH 7.4) for 1 hr at room temperature. Following fixation, most muscles were cut into three segments of approximately equal length: distal, central and proximal. The muscles were post-fixed in 2 % 0~04 in the same buffer and embedded in Epon. Sections were cut on a Cambridge (A. F. Huxley pattern) microtome. Semi-thin (l-3 pm) sections were examined by phase contrast and Nomarsky differential interference contrast optics. Thin sections were contrasted with uranyl acetate and lead and examined in AEI 1OOBand JEOL 1OOBand S microscopes. For counting of satellite cell nuclei in the electron microscope, sections were mounted on uncoated hexagonal mesh Athena grids which have large openings and small grid bars. Results

In the following description, satellite cells, interstitial cells and cells intermediate between the two (invasive cells) are first described separately. 1. Satellite cells (a) Definition. The term satellite cell is re-

AND

FRANZINI-ARMSTRONG

stricted here to cells located at the periphery of the muscle fiber, and completely separated from the extracellular space by the muscle fiber’s basal lamina (Figs. l-8). (b) Structure. Satellite cells are mononucleated, bipolar cells, i.e. their shape is fusiform, the central region of the cell contains a nucleus and at either end are cytoplasmic tails (Fig. 1). The exact length of the tails cannot be determined from longitudinal sections, because the tails may escape from the plane of the section. Tails up to 10 pm long were observed. Tails were always seen as single profiles. The long axes of satellite cells are quite precisely aligned with the long axes of the muscle fibers. Unless specified, the description below is obtained from crosssections of the muscles and thus from sections at right angles to the long axes of satellite cells. The disposition of basal lamina around satellite cell profiles is quite variable. In fresh frogs the majority of satellite cell profiles (71% of a total of 120 profiles examined) are covered by the muscle fiber’s basal lamina only over the portion of satellite cells plasmalemma that faces towards the intercellular spaces. No basal lamina is visible in the narrow space between the facing plasmalemmas of satellite cells and muscle fibers (Figs. 1, 2). In 21% of the profiles, the basal lamina also penetrates for a variable distance between these plasmalemmas (Fig. 6) and a smaller number of profiles (6%) are totally surrounded by an individual capsule of basal lamina, which is continuous with the basal lamina surrounding the muscle fiber (Fig. 5).

Fig. 1. Longitudinal section of a normal muscle from a recently captured Satellite cell’s nucleus is deeply indented and mostly heterochromatic. forms a thin annulus. Tails are not visible. x 6000.

(fresh) frog. Cytoplasm

Fig. 2. Cross-section through a cell similar to that of Fig. 1, also from a fresh frog. Note presence of coated vesicle (arrow), numerous caveolae, bundles of 10 nm filaments (encircled) and microtubule. No basal lamina is present between satellite cell and muscle fiber. This is the case for most satellite cells. x 31,300. Figs. 3, 4. Two regions of the same satellite cell from a fresh frog having larger amount of cytoplasm. Golgi and centriole mark the cell center. A small lysosome (1) and two multivesicular bodies (m) are present. The tail (Fig. 3) contains RER cisternae, few ribosomes and microtubules. x 16,600.

I.

n2

752

MARUENDA

In the case of three cells we established that the capsule covered the entire length of the satellite cells. In other cells the capsule is complete at some levels but not at others along the same cell. Finally, very few cells (2 %) are completely surrounded by a double capsule of basal lamina (Fig. 7). The frequency of cells surrounded by a partial, complete and double capsule of basal lamina is higher in frog maintained for l-2 months in the laboratory. Of 170 satellite cell profiles from these animals, 18, 20 and 6%, respectively, were surrounded by a partial, complete or double layer of basal lamina in the space adjacent to the muscle fiber. The position of the satellite cell relative to the muscle fiber is also variable. Some cells lie above the surface in such a way that only a minor portion of their plasmalemma faces the muscle fiber. Other cells lie in a shallow groove (Fig. 2) and some are partially surrounded by extensions of the muscle fiber (Figs. 5, 7). At the myotendon junction, satellite cells penetrate between the fingerlike projections of the muscle fiber. The cytoplasm of satellite cells occupies the tails and a thin annulus in the perinuclear region, which has a width variable between 0.03 and 1.6 pm. The cytoplasm contains, in small numbers, cytoplasmic organelles (ribosomes and polyribosomes, elements of smooth and rough endoplasmic reticulum, mitochondria, Golgi, glycogen). In cells with a larger amount of cytoplasm, cell organelles are more abundant (Figs. 3, 4). In the cell illustrated in Figs. 3 and 4, the Golgi apparatus is well developed and there are multivesicular bodies and lysosomes. We should like to emphasize that satellite cells

AND

FRANZINI-ARMSTRONG

also contain other organelles. Occasionally, a pair of centrioles is seen. Most satellite cell cross-sections at the level of the nucleus contain several microtubules and approximately one-third of the cross sections contain ‘10 nm’ filaments, arranged in bundles (Figs. 5, 6) and a fuzzy layer, probably containing actin filaments, under the plasma membrane. The same organelles are prominent in the tails, together with rough endoplasmic reticulum cisternae, often having a visible content (Fig. 6). Coated vesicles are present at the cell’s periphery (Figs. 2, 5). We notice that the content and distribution of cellular organelles is quite variable in the profiles of satellite cells as seen in cross-sections. In longitudinal sections, on the other hand, a single satellite cell has a uniform appearance, i.e. it does not show a larger accumulation of organelles in some areas than in others (see Figs. 1, 3, 4). It can thus be concluded that the variation in organelle content is mostly due to differences among satellite cells and not to differences along an individual cell. The surface forms numerous, small invaginations (Fig. 2). We have not determined if these should be considered pinocytotic vesicles or caveolae, like *those decorating the surface of the muscle fibers. The latter is probably true, because the invaginations are fairly numerous and there are no obvious signs of pinocytotic activity. Satellite cell nuclei are single, elongated in the longitudinal direction, and have an approximately oval or round shape in crosssections, with undulations of the surface and occasional indentations. These may be deep (Fig. 1) and it is probable that they are responsible for the rare appearance of a

Fig. 5. Cross-section through a muscle from a frog maintained in the laboratory for a period of time. A complete capsule of basal lamina surrounds the satellite cell. In the cytoplasm are prominent bundles of 10 nm filaments. x 25,700. Fig. 6. Section across the tail of a satellite cell which is encapsulated by basal lamina. RER, of which one cisterna is dilated, and bundles of 10 nm filaments are the most prominent cytoplasmic features. x 20,200. Fig. 7. The basal lamina capsule around this cell is complete Laboratory maintained frog, as for Fig. 5. x 25,000.

and, in places, double.

754

MARUENDA

double nuclear profile. The chromatin forms a dense ring at the periphery of the nucleus and nucleoli are present in some profiles. No cells undergoing mitosis have been found in this investigation. Length, width (in the direction of the muscle fiber’s circumference) and depth (perpendicular to the circumference) of nuclei were measured in electron micrographs of longitudinal sections along 15 satellite cells and they are 20.4 & 4.9 pm, 4.7 + 1.7 pm and 1.5 + 0.7 pm, respectively. Muscle fiber nuclei (myonuclei) markedly differ from satellite cells’ nuclei in several structural details. The shape is angular with sharp corners. The chromatin is more euchromatic and the nucleolus is large and obviously separated from the heterochromatic regions. With few exceptions, myonuclei are not located at the periphery. The differences in density of the two types of nuclei are sufficient to allow identification at the level of the light microscope (Fig. S), a fortunate circumstance which has permitted us, as well as numerous other investigators (see Table 4), to obtain good counts of satellite cell distribution (see below). In the frog, identification is facilitated by the different locations of nuclei. The average dimensions of myonuclei are 22k 7.3 pm, 4.652.5 pm, 1.2 f 0.5 pm (from 15 myonuclei measured from electron micrographs). The above description applies also to satellite cells in denervated muscles, where we did not find any particularly large satellite cells. Counts of satellite cells and myonuclei. Nuclei belonging to myofibers and satellite cells were counted in semi-thin cross-sections (1.5-3 pm thick) using phase contrast optics (Table l), and in thin cross sections (about 70 nm thick) using the electron microscope (Table 2). Sections were cut across the entire muscle. By marking over the coverslip with a felt-tip pen, random areas of several cross-sections were obscured. Satellite cell profiles were counted in two or three visible areas for each muscle. For those sections that contained only 20-30 fibers overall, two or three consecutive sections were used for the counts. Sections taken near the distal end of two only were in this category. An oil immersion lens (final magnification of x 1000) was used for examining sections from

AND

FRANZINI-ARMSTRONG

animals l-5 and a x 40 dry objective (final magnification x 400) was used for animals 6-10. Identification of satellite cell nuclei was based on the peripheral position, the uniformly dense appearance and the round or oval shape. In the counts of Table 1 (animals l-5), 169 peripherally placed nuclei were identified as belonging to satellite cells (column 4); 10 similarly placed nuclei were identified as myonuclei on the basis of their irregular shape and thin rim of heterochromatin at the edges. The uncertainty in this identification produces a maximum error of only 6 %. Electron microscope analysis was performed on material from animals l-5. The face of the block was trimmed to a smaller size and all the fibers included in the final section were counted (Table 3). In Tables 1 and 2, columns 2 to 4 give the crude counts of muscle fibers, myonuclei and satellite cell nuclei profiles. The ratio of myonuclei to fiber counts and of satellite cell nuclei to fiber counts are calculated in columns 5 and 6 and these are used below to estimate the average number of myonuclei and of satellite cells per unit length of fiber. The ratio of satellite cell nuclei to total nuclei is also given (column 7), to be used for comparison with data in the literature. In Table 2, cross-sectional profiles of satellite cell tails are given (column 8). These, of course, are not visible in the light microscope. For animals l-5, counts made on sections belonging to three different segments of approximately equal length into which the muscle was divided at the moment of embedding (see Methods) are tabulated separately. For muscles 6-10, sections were cut where the muscle was at its widest, and no special note was taken of the location of the sections. The distal region of the muscle is tapered as fibers gradually terminate into the distal tendon and sections cut across the entire muscle contain up to 200 fibers. In crosssections very close to the distal end of the muscle (e.g. animal 5) all fibers are cut close to the myotendon junction. The central region of the muscle contains fibers which are cut at a distance from their termination. Few fibers terminate in this region. In the central region, the fiber content of a cross-section varies between 200 and 1000 fibers. At the proximal tendon, fibers terminate within a short distance from each other. The fiber

SATELLITE

AND INVASIVE

75s

CELLS

Table 1. Counts of nuclei of muscle fibers and satellite cells in cross-section of normal frog sartorilu. Light microscope

I Segment

Animal

I

Distal

2 3 4 5 Subtotal

2

3

4

No.

No.

No.

fibers

mn

scn

59 80 73 17 48

153 183 201 176 125

10 16 17 IO 13

337

838

66

Mean+s.d. (partial)

I

Central

2 3 4 5 Subtotal

80 76 92 89 77 414

___106 136 160 116 135

11 10 14 7 8

653

50

Mean + s.d. (partial) 1 2 3 4 5

Proximal

Subtotal

72 76 74 89 102

129 171 136 126 153

10 11 11 13 8

413

715

53

Mean f s.d. (partial) 6 7 8 9 10 Subtotal

Proximal and central

148 149 89 114 322

213 221 157 222 579

20 22 21 13 33

822

1452

109

Mean f s.d. (partial) Total Mean + s.d.

~___~ 1986

3658

5

mnjfiber

6

scn/iiber

7

scn/scn + mn

2.59 2.29 2.75 2.29 2.60

0.17 0.20 0.23 0.13 0.27

0.05 0.14 0.12 0.29 0.10

2.50 f 0.21

0.20 + 0.05

0. I4 $0.9

1.33 1.79 1.74 1.30 1.75

0.14 0.13 0.15 0.08 0.10

0.0 0.13 0.12 0.17 0.13

1.58kO.24

@12+0.03

0.1 1 + 0.06

1.79 2.25 1.84 1.42 1.50

0.14 0.14 0.15 0.15 0.08

0.07 0.06 0.17 0.14 0.03

1.76kO.33

0~13rf:O.O3

0.09 + 0.06

1.84 1.48 1.76 1.95 1.80

0.14 0.15 0.24 @ll O,lO

o-01 0.09 0.12 0.06 0.05

1.76kO.17

0.15+0.06

0.07 + 0.04

1.70+0.25*

0.13 * 0.04*

0.09+ 0.05*

278

mn=myonuclei. scn= satellite cell nuclei. * Mean is for central and proximal regions of animals l-5 and for animals 6-10.

MARUENDA

756

Table 2. Counts of myonuclei

and satellite

AND FRANZINI-ARMSTRONG

cells in cross-sections

of normal frog

sartorius.

Electron

microscope

1

Animal 1 2 3 4 5

Region Distal

Subtotal

2

3

4

5

6

I

8

No. fibers

No. mn

No. scn

mn/fiber

scn/fiber

scn/scn + mn

No. of tails

35 7 27 24 7

38 24 53 10 19

2 4 I 4 2

100

144

19

Mean+s.d. (partial) 1 2 3 4 5

Central

Subtotal

72 47 67 31 40

79 40 68 43 40

0 6 9 9 6

257

270

30

Mean + s.d. (partial) 1 2 3 4 5

Proximal

23 28 33 58 26

28 30 30 36 65

2 2 6 6 2

168

189

18

1.09 3.43 1.96 0.42 2.71

0.06 0.57 0.26 0,17 0.29

0.05 0.14 0.12 0.29 0.10

1.91* 1.21

0.27kO.19

0.14kO.09

1.10 0.85 1.01 1.39 1.00

0.00 0.13 0.13 0.29 0.15

0.00 0.13 0.12 0.17 0.13

1.07+0.20

0.14kO.10

0.11+0.06

1.22 1.07 0.91 0.62 2.50

0.09 0.07 0.18 0.10 0.08

0.07 0.06 0.17 0.14 0.03

1 1 3 1 2

1 2 0 4 1

0 0 0

1 1 _

Subtotal Mean+s.d. (partial) Total

525

603

2 1.26kO.73

O.lO+OG4

0.09kO.06

1.17*0.51* .._

0.12-1@08*

0.10&0.06*

67

Mean+s.d.

18

mn = myonuclei. scn = satellite cell nuclei. * Mean is for central and proximal regions only content of cross-sections within the proximal segments is 1000-1200. Since in sartorii end plates are scattered over two broad bands in the muscle, we did not try to establish a relationship between satellite cell counts and neuromuscular junctions. The following information is obtained from analysis of the data in Tables 1 and 2. 1. The mean number of myonuclei and satellite cell nuclei per fiber is higher in the distal than in the central and proximal segments. In Table 1, the counts from animals 6-10 are pooled with those from

proximal and central segments of animals l-5 and the mean of these counts is compared with that of the distal region. The difference is statistically significant at a level of 0.001 (Student’s t-test). By analogy, when data from light microscope and electron microscope counts are pooled, the mean number of myonuclei and of satellite cell nuclei is significantly lower in the central and proximal than in the distal segments (significance level 0.01 for myonuclei and 0.05 for satellite cell nuclei). 2. The mean number of myonuclei per

SATkLLITE

AND

INVASIVE

751

CELLS

from the silver interference color of the fiber cross-section is 1-95 in light microscope counts (mean of entire column 5, Table 1) sections. The number of nuclei N per unit length of and 1.42 in electron microscope counts (mean of entire column 5, Table 2). The difference is fiber is: statistically significant at a level of 0.01. The mean counts of satellite cell nuclei per fiber (column 6, Tables 1 and 2) on the other hand, are not statistically significant between Combining (1) and (2), the two methods of counting. Section thickness (see below) can thus be excluded as a cause of this difference in nuclear complement, since it would equally affect counts of For proximal and central regions of the both types of nuclei. A more likely cause for muscle, the number of myonuclei iV, and the discrepancy in counts of myonuclei satellite cells nuclei N, per mm of fiber length obtained by light and electron microscope is are : grouping of the myonuclei (each cross,$I, = _. ~ !.I?’~~ = 69,39/mm section contains between 0 and 5 profiles of (22 + 2.5) 10m3mm myonuclei). This, if present, would more strongly affect counts at the EM level, where 0.13 sampling is more limited. ‘lis=(20+2.5j~~O~3~m = 5,78/mm Agreement between counts of satellite cells at the light and electron microscope levels from light microscope counts, and indicate that the criteria used for identifica1.17 N,,$= _. ~-_ --=53.01/mm tion in the light microscope were adequate (22 + 0.07) 10m3mm and that they do not suffer from the sort of systematic error that leads to the discrepancy N, = __ ~ 0-12 =5,98/mm in counts of myonuclei. (20 f 0.07) 10-3mt?tm The average content of myonuclei and satel!ite cell nuclei per unit length of muscle from electron microscope counts. fiber is calculated from the data of Tables 1 For the distal regions, the corresponding and 2, using the equation developed by numbers are: Abercrombie (1946) to correct for the effect 2.5 N,,L= -~ ~--~= 102.O/mm of nuclear length and section thickness on (22+2.5) 10m3mm crude nuclear counts. P=A

M L,+M

(1)

where, in our case, P is the mean number of nuclei contained in a slice of the fiber of the same thickness as the section thickness, if each nucleus appeared only once. A is the mean number of nuclear profiles per fiber profile, from the counts (column 5 for myonuclei and column 6 for satellite cell nuclei). Ln is the length of the nucleus. This is 20 pm for satellite cell nuclei and 22 pm for myonuclei (see page 754). M is the thickness of the section. The section thickness for calculations from light microscope data was taken to be 2.5 pm. The sections used were within +_I pm of that thickness. This introduces, at most, an error of 5% (since the nuclear lengths, see below, are more than 20 pm). For the electron microscope counts, the section thickness was assumed to be 70 nm

Ns=

O? ~~= 8.89/mm (20+2.5) 10m3mm

from light microscope counts and I.91 = 86,54/mm NnL= (if+ 0.07) 10~3~mm Ns=

0.27 = 13.45/mm (2O-tO.07) lO-3 mm

from electron microscope counts. By taking into account the section thickness, this calculation reduces, in part, the difference in crude counts obtained by light microscopy and electron microscopy, but the differences in myonuclei content obtained by the two methods remains significant. 2. Invasive cells (a) Dejinition. Cells that are caught in the process of crossing the muscle fiber’s basal

758

MARUENDA

lamina and/or moving towards the interior of the fiber, without fusing with it (Figs. 9-23). (b) General shape and internal structure. The shape of the cell is quite varied, depending on the stage of invasion. In general, it is elongated in the longitudinal direction and it is characterized by the presence of long fingerlike projections, which penetrate between the fibrils of the muscle fiber. The cytoplasm contains numerous ribosomes, free as well as attached to elements of the endoplasmic recticulum (including the nuclear envelope). Other prominent, but not constant, features of the cytoplasm are a

AND

FRANZINI-ARMSTRONG

fairlv large number of mitochondria (Fig. 23), mulfivesicular bodies and small lysosomes, with a dense content (Figs. 21-23). No phagocytic vacuoles and/or large lysosomes enclosing recognizable cell debris are present. Microtubules are more numerous than in satellite cells. Some, but not all invasive cells contain several cisternae of the rough endoplasmic reticulum (Figs. 22, 23). Overall, the cytoplasm may be either darker than the myofibrils, both in phase and EM images (Figs. 9, 10, 14), or lighter (Figs. 11, 12, 18, 21). Invasion can occur either singly or more frequently, in groups of two or three at each level of the fiber. The nucleus is fairly dense and similar, but

Fig. 8. Cross-section through a normal muscle. Phase contrast image to show differences between satellite cell nucleus (arrow) and myonuclei. x 1140. Fig. 9. Cross-section of an invaded unaffected muscle fibers in a normal round the invaded fibers. One of these the fiber. Dense nuclei within muscle are myonuclei (m). x 1140.

muscle fiber (center of the image) surrounded by muscle. Numerous cells with dense nuclei sur(arrow) sends a slender cytoplasmic process into fiber’s outline belong to invasive cells (i), others

Fig. 10. An invaded fiber from a control muscle (No. 8c in Table 3). An invasive cell (arrows) comes from pool of interstitial cells surrounding muscle fiber. Nuclei of invasive cells (arrows) and muscle fiber (m) are easily distinguishable. x 1140. Figs. 11, 12. An invaded fiber from a control muscle (No. 8c in Table 3) in phase contrast and Nomarsky interference contrast respectively. In the latter, slender cytoplasmic processes of invasive cells are better visible. x 1140. Figs. 13, 14. Invasive cells presumably in an early stage of invasion. Most of the cell body is outside the basal lamina, but a cytoplasmic finger penetrates across it. In Fig. 14, the invasive cell contains microtubules and small lysosomes (1); the muscle fiber contains a network of tubules deriving from the T-system. Both fibers are from frogs maintained in the laboratory. x 10,300 and x 41,800. Figs. 15, 16. Sections at two different levels across the same invasive cell. In Fig. 16, the cell is totally surrounded by basal lamina and it could be mistakenly identified as a satellite cell, if it was not for the presence of obvious evagination of its surface. In Fig. 15, the muscle fiber has numerous peripheral couplings in the area facing the satellite cell. From a denervated muscle (muscle 9d, Table 3). x 14,500 and x 12,100. Fig. 17. An invasive cell within the outlines of a muscle fiber. Long processes of the invasive cell penetrate between bundles of fibrils. The nucleus takes on elongated shape. At top, the muscle fiber shows signs of degeneration. Over most of the image, however, the structure is normal. Small T-tubules networks are forming in proximity of the invasive cell (arrows). Denervated muscle (No. 6d, Table 3). x 14,800. Fig. 18. An invasive cell penetrates into satellite cell. The satellite cell’s nucleus and rounded by the invasive cell. The invasive cytoplasm is less dense than that of the Table 3). x 14,800.

the muscle fiber in the area occupied by a thin layer of cytoplasm are completely surcell’s nucleus is more euchromatic and its satellite cell. Denervated muscle (No. 7d,

SATELLITE

AND

INVASIVE

CELLS

759

162

MARUENDA

not identical, in its shape and chromatin distribution to that of satellite cells, particularly when seen in the light microscope (Figs. 9-12). No mitotic figures are present. (c) The process of invasion. Cells are found in different stages of invasion, and these are described here in a sequence. From the fixed material it is not possible to tell if the actual sequence of invasion is as described and, particularly, if the cells at the moment of fixation are moving in the direction indicated here, i.e. from extracellular space to the center of the fiber (invading), or vice versa (retreating). Overall, the next movement is obviously from extracellular space to fiber’s interior. (9 Cells located totally outside of the basal lamina have finger-like projections which penetrate across a small gap in the basal lamina and come to close proximity to the muscle fiber’s plasmalemma (Figs. 13, 14). (ii) Cells whose body is partly on one side and partly on another of the basal lamina. (iii) Cells that are at the periphery of the muscle fiber and almost entirely covered by the basal lamina. These cells are classified as invasive if they present evaginations either crossing the basal lamina or penetrating towards the interior of the muscle fiber (Figs. 15, 18). Cells in this group are most easily mistaken for satellite cells, when the section cuts at a level

AND

FRANZINI-ARMSTRONG

where no special features are visible. For example, the cell in Fig. 15 is obviously ‘invasive’, but at a different level of sectioning (Fig. 16) it is very similar to a satellite cell, because it is totally surrounded by basal lamina. (iv) Cells whose cytoplasm and nucleus are contained within the boundary of a muscle fiber, up to the center core (Figs. 17, 21-23). These cells are characterized by very long, slender processes which penetrate between bundles of fibrils, so that the muscle fiber is, incompletely, divided into several domains. In no case has fusion of the plasmalemma of invading cells and muscle fibers been observed. Curiously, invasive cells often are seen to penetrate at a site occupied by a satellite cell and to surround it (Fig. 18). It seems that the natural niche on the fiber’s surface, in which the satellite cell is located, provides an easy access to invasive cells. In such a case it is possible to see that satellite and invasive cells are not identical. (d) Frequency of invasive cells. factors affecting them. In normal

Possible

muscles invasive cells are present, but rare. Few muscle fibers occupied by invasive cells have been observed in muscles from fresh frogs, in which several hundreds of muscle fibers were screened for nuclear counts. In muscles from frogs maintained in the laboratory, the frequency of invaded fibers is higher. It is interesting that when invaded, a fiber is

Fig. 19. A pericyte (P) and a fibroblast (F) occupy the interstitial is surrounded by a complete layer of basal lamina. The fibroblast and it contains few cisternae of RER. x 5900.

spaces. The pericyte has no basal lamina

Fig. 20. One invasive cell penetrates in the muscle fiber and several others with similar chromatin disposition in the nucleus occupy the interstitial spaces in this denervated muscle (No. 8d, Table 3). Cells with these characteristics are not usually present in frog sartorius. x 6500. Fig. 21. Two invasive cells within the outline of a muscle fiber from a normal muscle (same fiber is shown in Fig. 9). The cytoplasm contains free ribosomes and small lysosomes. x 15,100:~ Figs. 22, 23. Details of the cytoplasm of an invasive cell. There are numerous somes and some cisternae of rough ER. x 15,100 and x 22,200.

ribo-

MARUENDA

764

occupied by several foreign nuclei at each level of sectioning. The fiber in Fig. 9, for example, is a single invaded fiber in the entire cross-section of an otherwise normallooking sartorius. This same fiber was followed for at least 30 pm in non-serial semithin and thin sections. At all levels, it was surrounded and penetrated by numerous invasive cells. In the process of examining sartorii muscles from the legs of frogs subjected to unilateral denervation (see Methods), we found some muscles in which numerous fibers showed penetration of invading cells towards their center (Figs. 10-12). Counts of nuclei belonging to the three categories (muscle fibers, satellite cells and invasive cells) were performed at the light microscope level from sections of a group of denervated muscles and their controls (i.e. the muscles from the contralateral, normally innervated leg). These are presented in Table 3. Nuclei of satellite

AND

FRANZINI-ARMSTRONG

cells and invasive cells are tabulated together (columns 6 and 7) because at the light microscope level it is not always possible to distinguish between the two. Thin sections adjacent to those used for counts showed that nuclei having a dense chromatin and located within the outline but not at the periphery of the muscle fiber belonged to invasive cells as defined above. From the data of Table 3, the following observations are derived. (1) In most denervated muscles, the number of nuclei and satellite cell nuclei are not statistically different from those in normal muscles and in the control muscles from the same animal. (2) In muscles marked by asterisks (column l), many (up to half) of the muscle fibers are invaded, i.e. they show dense nuclei up to their center. In these muscles the number of satellite cell plus invasive cell nuclei is larger than in normal muscles (from Table 1). Thus in these muscles there is an increase in the

Table 3. Counts of nuclei of muscle fibers, satellite cells and invasive cells in a group of denervated muscles and their controls, Light microscope

Days denerv. 11 11

1

2

3

(2) Animal

No. fibers

No.

No.

mn

Id

88

IC

76

2d 2c

28

3d 3c

4

5

6

7

scn + icn

mn/fiber

scn + icn/fiber

scn + icn/mn + scn + icn

230 187

12 8

2.61 (2) 2.46 (2)

0.14 0.11

0.05 0.04

130 74

212 115

8 7

1.63 1.55

0.06 0.09

0.04 0.06

119 62

198 96

4 4

1.66 1.55

0.03 (2) 0.06

0.02 0.04

28

4d

81

154

8

1.90

0.10

0.05

28

5C

104

200

12

1.92

0.12

0.06

41

6d*

55 65

114 104

27 11

2.07 1.60

0.49 0.11

0.19 0.10

83 125

123 183 144 151 221 147

38 14 20 63 6 4

1.48 1.46 1.67 1.64 1.69 1.55

0.31 0.08 0.23 0.68 0.05 0.04 (2)

0.24 0.07 0.12 0.29 0.03 0.03

6c 41

Id* 7c

41

8d* SC*

61

9d 9c

86 92 131 95

d = denervated muscle. c = contralateral, control muscle. (2) See text. mn= myonuclei. scn= satellite cell nuclei. icn = invasive cell nuclei. * = muscles containing numerous, up to 50 %, invaded fibers.

SATELLITE

AND

INVASIVE

CELLS

number of nuclei other than myonuclei which are contained within the outline of the muscle fiber and the increase is due to the presence of invasive cells. In the same muscles the number of myonuclei is not significantly different than in the control muscles. Large numbers of invasive cells are not limited to denervated muscle, but they are present also in some of the control muscles from the same animals (i.e. animal 8). In a muscle denervated for 11 days and its control, the number of myonuclei per fiber is higher than the mean for normal muscles. In two muscles (one denervated for 28 days and the control of a 6-day denervation), the number of satellite cell nuclei per fiber is lower than the mean for normal muscles. (e) Response qf’ muscle fiber to invasion. The

invaded muscle fiber usually presents interesting changes. In proximity to the branches of invading fibers there are large honeycomblike arrangements of tubules, which can be traced to the transverse tubular system (Fig. 14). Smaller, less precisely arranged networks of T-tubules are also seen (Fig. 17). In addition, cisternae belonging to the sarcoplasmic reticulum form numerous peripheral couplings facing the areas of plasma membrane located in proximity of invading cells. These are identified by the presence of a dense content within the cisternae and by the junctional feet in the space between the cisternae and the plasmalemma (Fig. 15). Peripheral couplings are present, but rare, in normal adult frog muscle (Spray et al., 1974). Rarely, areas of damaged cytoplasm are present (Fig. 17) and one invaded fiber was in hyaline necrosis. (f) Fiber necrosis and macrophages. We have also seen fibers in necrosis without apparent invasion, in a muscle from a frog kept in the laboratory for about 2 months. In this muscle, macrophages are caught in the process of surrounding and presumably digesting large pieces of degenerated muscle fiber. 3. Interstitial cells Dqfinition. The endomysium contains cells belonging to the connective, vascular and nerve tissue. The following description is limited to cells that have some morphological similarity with satellite and invasive cells and that are mobile. Identification is based on (a)

765

previous descriptions in the literature, but it is not unequivocal, since most descriptions are derived from mammalian tissues. (b) Fibroblasts and jibrocytes (Fig. 19). NO attempt is made here to distinguish between the two, although the latter has a larger amount of rough endoplasmic reticulum. The overall shape of the cell is elongated with slender processes. The plasma membrane is not surrounded by a basal lamina and it small invaginations forms numerous, (caveolae and/or pynocytotic vesicles). The cytoplasm contains prominent Golgi and rough endoplasmic reticulum elements. Lysosomes, 10 nm filaments and microtubules are also present. The heterochromatin band of the edge of the nucleus is narrow. The nucleolus is prominent. Even though the fibroblast may be located very close to the muscle fiber, the space between the two is usually resolvable in the light microscope and it is unlikely that nuclei from fibroblasts are included in the counts for satellite cells. (c) Pericytes (Fig. 19) are adventitial

cells located at the periphery of capillaries, under the lamina propria. Their major distinctive feature is the complete capsule of basal lamina that surrounds them. Their cytoplasm is filled by numerous organelles, including centrioles. The nucleus is mostly euchromatic. Rarely, pericytes are seen to move away from the periphery of capillar:es. Lt is unlikely that pericytes are mistakenly identified as satellite cells in counts at the light microscope level, because they are located at some distance from the periphery of the muscle fibers. (d) Schwann cells cover one side of the nerve terminal and their nuclei are quite dense. They lie very close to the muscle fiber. particularly where the nerve branch that they cover is a terminal branch, close to the end plate. Thus they can, under those conditions, be mistaken for satellite cells at the light microscope level. However, there are, at most, two end plate regions on each muscle fiber and thus very few Schwann cells per fiber. The error introduced by their mistaken identification with satellite cells is minimal. (e> Additiorlal cells in the interstitium, o/ probable blooa’origin. In normal muscles the

766

MARUENDA

interstitium is poorly populated by the cell types mentioned above. Other cells, sometimes in large numbers, are present in proximity of muscle fibers that are invaded (Figs. 9-12, 20). Invasive cells come from this population. The number of cells occupying the interestitium in heavily invaded muscles is quite large (Figs. 9-12) and since no mitosis are seen, it is likely that they are the result of migration from the blood stream. The organelle content of these interstitial cells is similar to that of the invasive cells. The cytoplasm is variable in amount and mostly light, and it contains lysosomes, mostly small, filaments and microtubules, and few elements of the rough endoplasmic reticulum. The shape may be either elongated (Figs. 13, 14) or round (Fig. 20), and we have not determined whether the two shapes represent different stages of the same cell. The most characteristic feature is the presence of long, slender processes, arising at several sites along the periphery of the cell. Discussion

This paper presents two major results. One is a description of satellite cells in a frog muscle, which includes an overall view of variabilities in shape, organelle content, basal lamina and counts of satellite cells frequency. The second is a new finding: invasive cells which may penetrate to the interior of a muscle fiber, without piercing or fusing with its plasmalemma. We suggest that these cells are of blood origin and we provide a description of features that allow us to distinguish them from satellite cells. Descriptions of satellite cells are usually derived from sections at the level of the nucleus and in these the cytoplasm forms a thin perinuclear annulus containing few organelles. Most of the satellite cell profiles that we have seen conform to this general description but with additional details. First of all, the amount of cytoplasm and the content of organelles can be quite variable from one satellite cell to another, even within the same muscle. Some of the variabilities that we describe here have been already recognized in satellite cells from other animals particularly during development (see lshikawa, 1966; Schultz, 1976; Conen and Bell, 1970; Wakajama, 1976; Kahn and

AND

FRANZINI-ARMSTRONG

Simpson, 1974; Ishikawa, 1970; Kelly, 1978a), but in some instances they have been attributed to an activating effect of experimental intervention (Studitsky, 1974). We find prominent bundles of 10 nm filaments and numerous microtubules in most crosssections of satellite cells and these are consistent with the elongated shape of the cell. Centrioles are found fairly frequently (see Kahn and Simpson, 1974) and this is consistent with the fact that even though dormant in normal conditions, the cells can be easily activated to divide under a variety of experimental stimuli. It is interesting that satellite cells contain, even if rarely, lysosomes. It is only recently that the presence of elongated tails has been recognized in satellite cells (Kryvi, 1975). We occasionally find long tails, but have never observed them to branch. The tails are particularly rich in filaments and microtubules. One mostly unrecognized variable in the structure of satellite cells is the disposition and abundance of basal lamina. We describe infiltration of the basal lamina into the space between satellite cells and muscle fiber, a feature also observed in aging muscles (Snow, 1978). There are two possible explanations for this extra coat of basal lamina. (1) The initial situation is as usually described, with the basal lamina covering only the outer rim of the satellite cell. A variable amount of basal lamina is secondarily secreted by either muscle or satellite cell and this results in an extra coat. (2) The second possibility is that some cells start with a complete and even double layer of basal lamina around them and then this is gradually reabsorbed. If this were the case, then satellite cells having a more complete coat would more recently have been added. During differentiation and development of mammalian skeletal muscle, most satellite cells are enveloped by the basal lamina at the same time as the muscle fiber and occasionally an extra layer of lamina penetrates between muscle fiber and satellite cell (Schultz, 1976). Less frequently, in later stages of development, new satellite cells penetrate across the basal lamina to take their final position (Kelly and Zacks, 1969a). Thus the situation that we describe here would be unusual. The only cells in the interstitial spaces of adult frog muscle that possess a basal lamina coat of their own are the pericytes. These are the only cells that

SATELLITE

AND

INVASIVE

CELLS

167

would be appropriate candidates for our hypothesis number two. The higher incidence of satellite cells with extra basal lamina coat in frogs maintained in the laboratory indicates that the amount of basal lamina can be induced to change. Our counts of satellite cells, even though more extensive than many in the literature, have brought no surprises. The values that we obtain (Tables 1 and 2) are not very different from those reported in the literature for a variety of muscles (Table 4). Frogs range among the animals that have the highest content of satellite cells, particularly if one considers that the sartorius is composed of fast twitch fibers and thus it should be compared with e.g. the EDL in rat. We find that following denervation the range of structural variability in the satellite cells is

well within the limits found in normal muscle and thus that there are no morphological indications of changes specifically induced by denervation in satellite cells of frog sartorius. Limited counts of satellite cells in muscles denervated for 11-61 days also show that there are no significant changes in the content of satellite cells following this intervention. These results are in agreement with the counts by Cardasis and Cooper (1975a) and with the observed decline in Hs thymidine incorporation (Kelly, 1978b) in satellite cells of muscles denervated during development, but are at variance with the results by Lee (1965), Hess and Rosner (1970), Aloisi et al. (1973), Ontell (1974), Schultz (1974) and Hanzlikova et al. (1975), who observe an increase in the content of satellite cells following denervation. In comparing these

Table 4. Counts of’satellite cell nuclei in N vavietv of rmtscles from ndult animals Muscle

Animal

scnjscn + mn

Method

Reference

Shark Axial red (Galerts nrelastmms) Axial white

0.032 0~016

EM EM

Kryvi (1975)

Axolate Trunk (Siredon mexicanus)

0.040

EM

Flood (1971)

0.016 0.016 0.090 0.100

EM EM LM EM

Trupin (1976) Trupin (I 976)

Tail

0.075 0.048

LM

Kahn and Simpson (1974)

Mouse

Lumbricalis Peroneus longus Gastrocnemius Levator ani

0.040 0.042 0.060 0.058

EM EM EM LM

Schultz (1974) Allbrook et a/. (1971) Cardasis and Cooper (1975b) Venable (1966)

Rat

EDL, tibialis

0.020 OX@7 0.027 0.048 0.058

LM LM EM EM EM

Ontell (1974)

0~038-0~073 0.040

EM EM

0.100

EM

Muir et ml. (1965)

0.04&O. 1to

EM

0.040

EM

Wakayama (1976) Schmalbruch and Hellhammer (I 976)

Frog (Rmza clamitam) (Rtma pipiens)

Lizard (Lposoma sp.) (An~lq’ssp.)

Gastrocnemius Gastrocnemius Sartorius

EDL,

soleus

Soleus Soleus, tibialis ant diaphragm. Subclavius Bat (Eidolon helvum)

Web and others

Human

scn = satellite cell nuclei. mn = myonuclei.

This study

Kelly (1978b) Hanzlikova et al. (1975) Schmalbruch and Hellhammer (1977) Allbrook et al. (1971)

768

MARUENDA

data one should keep in mind the effects of difference in the reactivity of different muscles (frog muscles develop atrophy very slowly relative to those from warm-blooded animals) and, perhaps more importantly, the effect of passive mechanical stresses on the muscle, as indicated by Schiaffino and Hanzlikova (1970). Thus the effect on a single muscle of denervating the entire leg may be quite different from the effect of denervation at the site of nerve entry into the muscle. In addition, our data were obtained from muscles which were denervated for a relatively long period of time. We describe invasive cells independently of satellite cells because the following evidence indicates that invasive cells come from a different population. (1) ‘Invaded’ fibers are surrounded by cells, sometimes in large numbers, lying outside the basal lamina. Invasive cells are seen to come from this pool. (2) In muscles where invaded fibers are numerous (see Table 3) there is also a statistically significant increase in the number of nuclei (other than myonuclei) contained within the basal lamina. Since mitosis is not observed, these extranumerary nuclei must come from the interstitium. In addition, as the interstitium is normally scarcely populated, we postulate that these cells come from the circulation. (3) In injured muscles satellite cells tend to remain within the basal lamina scaffolding, where they regenerate muscle fibers (Price et al., 1964; Church et al., 1966; Church, 1970; Vrako and Benditt, 1972; Jirmanova and Thesleff, 1970; Snow, 1977; Schmalbruch, 1977). The work of Bischoff (1974, 1975) indicates that satellite cells, even when fully active, will not cross the intact basal lamina. Other in vitro investigations confirm that colonies derived from satellite cells grow into the culture medium from the cut ends of muscle fiber segments, where the basal lamina is broken (Konigsberg et al., 1975). Invasive cells, on the other hand, can cross the basal lamina, presumably by producing a small hole in it. (4) There are small but obvious structural differences between invasive and satellite cells in the overall shape and in the appearance of nuclei and cell organelle. These differences, however, are similar to those that would result from an ‘activation’ of satellite cells. (5) In developing muscles, satellite cells occasionally send slender processes towards the muscle

AND

FRANZINI-ARMSTRONG

fiber’s interior (Kelly and Zachs, 1969a, b; Schultz, 1976; Ontell, 1977). These cells presumably are in stages just preliminary to fusion with either a myotube or a muscle fiber. These satellite cells differ from invasive cells in two respects: no migration of satellite cells towards the center of the fiber, prior to fusion, has been described; satellite cells in young muscles contain many polyribosomes, whereas invasive cells mostly do not. (Occasionally, an invasive cell may also contain polyribosomes.) In the course of development, satellite cells are also seen to cross the basal lamina from interstitium towards the muscle fiber (Kelly and O’Donoghue, 1974). Again, there is a difference between the behavior of these cells and that of invasive cells. Future satellite cells remain at the periphery of the muscle fiber, whereas invasive cells typically do not. (6) Invaded fibers do not have a large number of myonuclei, i.e. invasive cells do not fuse with muscle fibers. The fate of invaded fibers is not clear. Even though possessing apparently unaltered cross striation and disposition of internal organelles, invaded fibers, both in normal and denervated muscles, show minor changes. Most prominent is the presence of T-tubules networks. These form during differentiation, as an intermediate stage in T-tubule development (Ishikawa, 1968; Schiaffino and Margreth, 1969), but they also are present in many pathological conditions (Pellegrino and Franzini, 1963; Schotland, 1970). The presence of numerous peripheral couplings near invaded areas of the fiber is also unusual, since the couplings are rare in adult frog muscle (Spray et al., 1974). We have found evidence of necrosis only in one ‘invaded’ fiber and vacuolation and swelling only rarely (see Fig. 17). At the moment, it is not possible to say whether final dissolution of the fiber structure is the outcome of invasion. One conclusion is warranted from our findings: the invaded fiber acts as a target, thus even though this is not always morphologically apparent, it must be altered before invasion. Invasive cells are not limited to frog’s sartorius muscle. They have, so far, been observed in the pectoralis muscles from the same animal (Gorio, 1978) and in rat muscles in the process of compensatory hypertrophy (Schiaffino, 1978). Cells with processes

SATELLITE

AND

INVASIVE

extending towards the extracellular spaces and the interior of the muscle fibers have been described in a fish muscle (Kryvi, 1975) and the possibility exists that these also are invasive rather than satellite cells. Identification of invasive cells’ origin is uncertain, but some hypotheses can be formulated on the basis of their morphology and from comparisons with cell types which invade muscles under a variety of pathological conditions. Identity with neutrophils can be excluded outright on the basis of strong differences (Yoramy et a/., 1976). Histiocytes (macrophages), such as those involved in the process of phagocytizing damaged end plate regions in experimentally induced myasthenia (Engel et al., 1976) and necrotic fibers in damaged muscles (Church et al., 1966; Carlson and Gutman, 1975; Jirm&nova and Thesleff, 1972; Trupin, 1978), are different from invasive cells in some structural detail (more ribosomes, large phagocytic vacuoles) and in apparent behavior (they tend to try to engulf damaged regions, rather than to penetrate into them). Invasive cells resemble monocytes, which tend to assume an elongated shape when inside the outlines of the basal lamina (see Trupin, 1978) and contain small lysosomes. If this identification is correct, then it would indicate that in the particular condition that induces invasion, monocytes delay their differentiation into active macrophages for a prolonged period of time. This is a behavior which is quite unusual. On the other hand, there is a strong pos-

49

769

CELLS

sibility that invasive cells are lymphocytes. This alternate identification is favored by the finding that invaded fibers seem to act as targets for the invasion and by the overall rounded appearance of invasive cells when free in the interstitial spaces. Invasion of other cell types (emperipolesis) is not uncommon behavior for a lymphocyte. In lymph nodes, for example, lymphocytes gain access to the capillaries by penetrating across the body of the endothelial cells from the basal side and exiting on the luminal side, without ever fusing with it (Marchesi and Gowans, 1964). A similar process is apparently used by lymphocytes to penetrate into the lumen of the intestine (Andrew, 1965; Schmedtje, 1965). Finally, intact lymphocytes are harbored within large macrophages (Pulverte and Humble, 1962). More significantly, it was shown that sensitized lymphoid cells infiltrate muscle fibers (Kakulas, 1966) in a manner very similar to that described here (Mostaglie (of al., 1975). If the final fate of invaded fibers is indeed necrosis, it may be assumed that this is due to toxins injected by the invasive cells into the muscle fiber (see Peter and Dawkins, 1971, for the existence of target cell lysis). Acknowledgements

We thank Mrs S. Johnson for her high quality assistance. Dr Castillo de Maruenda is a recipient of an MDAA fellowship. This research was supported by MDAA (Henry M. Watts Neuromuscular Disease Center).

MARUENDA

770

AND

FRANZINI-ARMSTRONG

References ABERCROMBIE,M. 1946. Estimation of nuclear population from microtome sections. Anat. Rec., Lond., 94, 239-247. ALLBROOK, D. B., HAN, M. F. and HELLMUTH, A. E. 1971. Population of muscle satellite cells in relation to age and mitotic activity. Pathology, 3, 233-243. ALOISI, M., MUSSINI, I. and SCHIAFFINO, S. 1973. Activation of satellite cells nuclei in denervation and hypertrophy. In Basic Research in Myology (ed. B. A. Kokulas), Part I, ICS 294, 338-342. Excerpta Medica, Amsterdam. ANDREW, W. 1965. Cytoplasmic and nuclear alterations in migrating lymphocytes of intestinal mucosa of the mouse: a light and electron microscopic study. Anat. Rec., 151, 319a. BISCHOFF, R. 1974. Enzymatic liberation of myogenic cells from adult rat muscle. Anat. Rec., 180, 645-662. BISCHOFF, R. 1975. Regeneration of single skeletal muscle fibers in vitro. Anat. Rec., 182, 215-236. CARDASIS, C. A. and COOPER, G. W. 1975a. A method for the chemical isolation of individual fibers and its application to a study of the effect of denervation on the number of nuclei per fiber. J. exp. Zool., 191, 333-346.

CARDASIS,C. A. and COOPER, G. W. 1975b. An analysis of nuclear numbers in individual muscle fibers during differentiation and growth: a satellite cell muscle fiber growth unit. J. exg. Zool., 191, 347-358. CARLSON, B. M. and GUTMAN, E. 1975. Regeneration in free grafts of normal and denervated muscles in the rat: morphology and histochemistry. Anat. Rec., 183, 47-62. CHURCH, J. C. T. 1970. A model for myogenesis using the concept of the satellite cell segment. In Regeneration of Striated Muscle and Myogenesis (eds. A. Mauro, S. A. Shafig and A. T. Milhorat), pp. 118-121. Excerpta Medica, Amsterdam. CHURCH, J. C. T., NORONHA, R. F. X. and ALLBROOK, D. B. 1966. Satellite cells and skeletal muscle regeneration. Br. J. Surg., 53, 638642. CONEN, P. E. and BELL C. D. 1970. Study of satellite cells in mature and fetal human muscle and rhabdomyosarcoma. In Regeneration of Striated Muscle and Myogenesis (eds. A. Mauro, S. A. Shafig and A. T. Milhorat), pp. 194-211. Excerpta Medica, Amsterdam. ENGEL, A. G., TSUJIHATA, M., LAMBERT, E. H., LINDSTROM, J. and LENNON, V. A. 1976. Experimental autoimmune myasthenia gravis: a sequential and quantitative study of the neuromuscular junction ultrastructure and electrophysiologic correlations. J. Neuropath. exp. Neurol., 35, 569-587. FLOOD, L. 1971. The three-dimensional structure and frequency of myosatellite cells in trunk muscle of the Axolotl (Siredon mexicanus). J. Ultrastruct. Res., 36, 523-524. Gomo, A. 1978. Personal communication. HANZL~KOVA, V., MACKOV.&,E. C. and HN~K, P. 1975. Satellite cells of rat soleus muscle in the process of compensatory hypertrophy combined with denervation. CeN Tiss. Res., 160, 411421. HESS, A. and ROSNER, S. 1970. The satellite cell bud and myoblast in denervated mammalian muscle fibers. Am. J. Anat., 129, 21-40. ISHIKAWA, H. 1966. Electron microscopic observations of satellite cells with special reference to the development of mammalian skeletal muscles. Z. Anat. EntwGesch., 125, 43-63. ISHIKAWA, H. 1968. Formation of elaborate networks of T-system tubules in cultured skeletal muscle with special reference to the T-system formation. J. CeN Biol., 38, 51-66. ISHIKAWA, H. 1970. Satellite cells in developing muscle and tissue culture. In Regeneration of Striated Muscle and Myogenesis (eds. A. Mauro, S. A. Shafiq and A. T. Milhorat), pp. 167-179. Excerpta Medica, Amsterdam. JIRMANOVA, Y. and THESLEFF, S. 1970. Ultrastructural study of experimental muscle degeneration and regeneration in the adult rat. Z. Zellforsch. Mikrosk. Anat., 131, 71-97. KAHN, E. B. and SIMPSON, S. B., JR 1974. Satellite cells in mature uninjured skeletal muscles of the lizard tail. Devl Biol., 37, 219-233. KAKULAS, B. A. 1966. Destruction of differentiated muscle cultures by sensitized lymphoid cells. J. Path. Bact., 91, 495-503. KARLSON, U., ANDERSSON-CEDERGRAN,E. and OTTOSON, D. 1966. Cellular organization of the frog muscle spindle as revealed by serial sections for electron microscopy. J. Ultrastruct. Res., 14, l-35. KATZ, F. R. S. 1961. The termination of the afferent nerve fiber in the muscle spindle of the frog. Phil. Trans. R. Sot., Lond., Series B, 243, 221-240.

SATELLITE

AND

INVASIVE

CELLS

771

KELLY, A. M. 1978a. Perisynaptic satellite cells in the developing and mature rat soleus muscle. Anar. Rec. (in press). KELLY, A. M. 1978b. Satellite cells and myofiber growth in the rat soleus and extensor digitorum longus muscles. Devl Biol. (in press). KELLY, A. M. and ZACHS, S. I. 1969a. The histogenesis of rat intercostal muscle. J. Cell Biol., 42, 135-153. KELLY, A. M. and ZACHS, S. I. 1969b. The fine structure of motor end plate morphogenesis. J. CeN Biol., 42, 154-I 69. KELLY, A. M. and O’DONOGHUE, .I. L. 1974. Wandering satellite cells in skeletal muscle. J. CeNBiol., 63, 164a. KONIGSBERG, U. R., LIPTON, B. H. and KONIGSBERG, I. R. 1975. The regenerative response of single mature muscle fibers isolated in vitro. Devl Biol., 45. 260. KRYVI, H. 1975. The structure of the myosatellite cells in axial muscles of the shark Galeus melrrsfomus. Anat. Embryo/., 143, 3544. LEE, C. 1965. Electron microscope observations on myogenic cells of denervated skeletal muscle. Expl Neural., 12, 123-135. MARCHESI, V. T. and COWAN, J. L. 1964. The migration of lymphocytes through the endothelium of venules in lymph nodes: an electron microscope study. Proc. R. SW., Lond., B, 159, 283-290. MASTAGLIA, F. L., PAPAOLOMIVRIOU,J. M. and DAWKINS, R. L. 1975. Mechanisms of cell-mediated myotoxicity. Morphological observations in muscle grafts and in muscle exposed to sensitized spleen cells in viva. J. neural. Sri., 25, 269-282. MAURO, A. 1961. Satellite cell of skeletal muscle fibers. J. biophys. biochem. Cytol., 9, 493-495. MUIR, A. R., KANJI, A. H. M. and ALLBROOK, D. 1965. The structure of the satellite cells in skeletal muscle. J. Anat., Lond., 99, 435-444. ONTELL, M. 1974. Muscle satellite cells: a validated technique for light microscopic identification and a quantitative study of changes in their population following denervation. Anat. Rec., 178, 21 l-228. ONTELL, M. 1977. Neonatal muscle: an electron microscopic study. Anat. Rec., 189, 669-690. PELLEGRINO, C. and FRANZINI, C. 1963. An electron microscope study of denervation atrophy in red and white skeletal muscle fibers. J. Cell Biol., 17, 327-349. PETER, J. B. and DAWKINS, R. L. 1971. Target cell lysis mediated by soluble cytotoxin released from stimulated lymphocytes. Nature (New Biol.), 232, 79-80. PRICE, H., HOWES, E. and BLUMBERG,J. 1964. Ultrastructural alterations in skeletal muscle fibers injured by cold. II. Cells of the sarcolemmal tube. Observations on discontinuous regeneration and myofibril formation. Lab. Invest., 13, 1279-l 302. PULVERTAFT,R. J. V. and HUMBLE, J. G. 1962. Intracellular phase of existence of lymphocytes during remission of acute lymphatic leukemia. Nature, 194, 194-195. REZNICK, M. 1969. Thymidine-H uptake by satellite cells of regenerating skeletal muscle. J. Cell Biol., 40, 586-590. SCHIAFFINO, S. 1978. In Proc. II Con$ on Regeneration of Striated Muscle (ed. A. Mauro), (in press). SCHIAFFINO, S. and MARGRETH, A. 1969. Coordinated development of the sarcoplasmic reticulum and T-system during postnatal differentiation of rat skeletal muscle. J. Cell Biol., 41, 855-875. SCHIAFFINO, S. and HANzLiKovA, V. 1970. On the mechanisms of compensatory hypertrophy in skeletal muscles. Experientia, 26, 152-153. SCIIMALBRLICH,H. 1977. Regeneration of soleus muscles of rat autographed in foto as studied by electron microscopy. Cell Tiss. Res., 177, 159-180. SCHMALBRUCH,H. and HELLHAMMER,U. 1976. The number of satellite cells in normal human muscle. Anar. Rec., 185, 279-288. SCHMALBRUCH, H. and HELLHAMMER, U. 1977. The number of nuclei in adult rat muscles with special reference to satellite cells. Anat. Rec., 189, 169-176. SCHMEDTJE,J. F. 1965. Some histochemical characteristics of lymphoepithelial cells of the rabbit appendix. Anat. Rec., 151, 412a. SCHOTLAND, D. L. 1970. An electron microscopic investigation of myotonic dystrophy. J. Neuroputh. exp. Neural., 29, 241-253. SCHUL-~Z, E. 1974. A quantitative study of the satellite cell population in postnatal mouse lumbrical muscle. Anat. Rec., 180, 589-596. SCHULTZ, E. 1976. Fine structure of satellite cells in growing skeletal muscles. Am. J. Anat., 147, 49-70. SNOW, M. H. 1977. Myogenic cell formation in regenerating rat skeletal muscle injured by mincing. I. A fine structural study. Anat. Rec., 188, 181-199. SPRAY, T. L.. WAUGH, R. A. and SOMMER, J. R. 1974. Peripheral couplings in adult vertebrate skeletal muscle. Anatomical observations and functional implications. J. Cell Biol., 62, 223-227. STLIDITSKY,A. N. 1974. The neural factor in the development of transplanted muscles. In Exp1orator.v Concepts in Muscular Dystrophy (ed. A. T. Milhorat), pp. 351-366. Excerpta Medica, Amsterdam. TRIJPIN, G. L. 1976. The satellite cells of normal anuran skeletal muscle. Deal. Biol. 50, 517--524.

112

MARUENDA

AND

FRANZINI-ARMSTRONG

TRUPIN, G. L. 1978. Proc. II Conf. on Muscle Regeneration (ed. A. Mauro), (in press). VENABLE, J. H. 1966. Morphology of the cells of normal, testosterone-deprived and testosterone-stimulated

levator ani muscles. J. Anar., 119, 271-302. VRACHO, R. and BENDITT, E. P. 1972. Basal lamina: the scaffold for orderly cell replacement. Observations on regeneration of injured skeletal muscle fibers and capillaries. J. Cell Biol., 55, 406-419. WAKAYAMA, Y. 1976. Electron microscopic study of the satellite cell in the muscle of Duchenne muscular

dystrophy.

J. Neuropath. exp. New&

35, 532-540.

YOKAM, R., BEHAR, A. J., YANKO, L., HALL, T. N. and PETERS, P. D. 1976. Gold tracer

regeneration.

J. Neuropath. exp. Neurol., 35, 444.

studies of muscle