We have used antibodies to three of the isomorphic forms of collagen, types I, III and V, in an immunofluorescence microscopy study of myogenesis in the embryonic chick wing, concentrating on the period between stages 27 and 30 (5 to 7·5 days incubation) which is when the dorsal and ventral muscle masses separate into discrete muscles.

We have demonstrated the presence of all three collagen types at the ectoderm-mesenchyme junction from stage 27 onwards. Type I collagen and then type III collagen are found in progressively deeper layers of the dermis at the later stages. Both types I and V collagen are initially present in the cartilage elements, but type 1 collagen becomes restricted to the periphery of these structures at later stages. The developing muscle areas show a lack of staining at all stages and it is only at the latest stages that types I and III collagen first appear in the surrounding epimysium.

We discuss possible mechanisms for the division of the muscle masses in the light of this information on the distribution of collagen types.

The first morphological indication of muscle development in the tetrapod limb is the appearance of distinct muscle masses dorsal and ventral to the chondrogenic core. Repeated divisions and subdivisions of these muscle masses produce the pattern of discrete limb muscles characteristic of each particular species. Although there is variation in the pattern of adult musculature from species to species, Sullivan (1962) states that the overall sequence of divisions is similar in all those tetrapods that have been studied, such as the newt, lizard, turtle, chick and opossum.

These similarities in the process of muscle mass division in widely different species argue strongly for a single underlying principle. In view of the central role of this process in the formation of the often complex pattern of limb muscles it is surprising that relatively few recent studies have examined it in detail. Shellswell (1977 a) showed that these divisions can take place and result in a normal pattern of muscle in the absence of nerves. Bradley (1970) confirmed the findings of Hunt (1932) that tension within the longitudinally growing limb was not the stimulus for the divisions.

Recent findings of Christ, Jacob & Jacob (1977), Chevallier, Kieny & Mauger (1977) and Chevallier (1979) have shown that cells of the muscle connective tissue (epi-, peri-, and endomysium) in the wing are derived from the somatopleural mesoderm while only the muscle cells are of somitic origin. Chevallier et al. (1977) have also shown that grafts of somites from any cephalocaudal level (e.g. lumbar, cervical) to the brachial level always result in a normal pattern of wing muscles. One interpretation of these results is that the non-muscle cells in some way control the formation of the pattern of muscles and that muscle cells are essentially passive partners in this process, solely concerned with muscle cytodifferentiation.

Although the precise mechanism by which non-muscle cells control this process is unknown, it might be expected from their purely physical proximity that cells of the muscle connective tissue play a major role. The present study follows the development of connective tissue collagen as the muscle masses divide. We make use of our recent finding that there is a specific distribution of the various collagen types within the different components of muscle connective tissue in the chick - type I collagen is found mainly in the epimysium, type III in the perimysium and type V in the endomysium (Bailey, Shellswell & Duance, 1979). It is well known that the presence of collagen greatly enhances cytodifferentiation in vitro (Hauschka & Konigsberg, 1966). In this paper we have used immunofluorescence microscopy to follow the appearance of the different collagen types in vivo to assess their importance in the division of the muscle masses.

Embryos of a Light Sussex × Rhode Island Red strain (Lansdown Poultry Farm, Bath) were removed from the eggs after periods of incubation varying between 4 and 10 days. Embryos were staged according to Hamilton (1952), fixed in cold 95 % alcohol and processed for wax embedding as described in Sainte-Marie (1962).

Serial, transverse 5 μm-thick sections were cut from the limbs and those corresponding to the mid-lower arm region were rehydrated as described by Sainte-Marie (1962). Sections were incubated for 1 h with 1 % testicular hyaluronidase-(Sigma, 1:100 in 0·1 M phosphate buffer pH 5·6). After washing in phosphate-buffered saline (PBS), sections were treated for 30 min with one of the following rabbit sera: (i) anti-chick type I collagen, (ii) anti-chick type III collagen, (iii) anti-human type V collagen or (v) non-immune rabbit serum as control. Collagen-type-specific antibodies were prepared and tested for crossreactivity according to Duance et al. (1977). After further washing for 1 h in PBS, sections were treated with fluorescein isothiocyanate conjugated antirabbit IgG (Wellcome Reagents Ltd) for 30 min before final washing for 1 h in PBS and mounting in Glycerine/PBS.

The major divisions of the dorsal and ventral muscle masses at the midforearm level occur in a binary sequence between stages 26 and 32 (Shellswell & Wolpert, 1977).

Stage 27 : (5 days incubation)

Stage 27 is a convenient time to study the early events in this process because the ventral mass is just making its first division while the dorsal mass is yet to divide (Fig. 1 d).

Fig. 1.

(a, b, c, d and f). Transverse sections of mid-forearm level, stage-27 limbs, stained with: (a) Haematoxylin-Eosin (H/E); (b) and (c) anti-type I collagen serum; (d) anti-type III collagen serum; (f) anti-type V collagen serum, (e) Aorta, stained as in (d). Treatment with non-immune serum gave negligible staining at all stages. A = aorta, BW = body wall, C = cartilage element, D = dorsal muscle mass, E/M = ectoderm-mesenchyme junction, V = ventral muscle mass. Scale bars represent : 100 μm in a, b, d and f; 50 μm in c and e.

Fig. 1.

(a, b, c, d and f). Transverse sections of mid-forearm level, stage-27 limbs, stained with: (a) Haematoxylin-Eosin (H/E); (b) and (c) anti-type I collagen serum; (d) anti-type III collagen serum; (f) anti-type V collagen serum, (e) Aorta, stained as in (d). Treatment with non-immune serum gave negligible staining at all stages. A = aorta, BW = body wall, C = cartilage element, D = dorsal muscle mass, E/M = ectoderm-mesenchyme junction, V = ventral muscle mass. Scale bars represent : 100 μm in a, b, d and f; 50 μm in c and e.

Immunofluorescent staining with anti-type I collagen antibodies demonstrated a clear presence of type I collagen at the ectoderm-mesoderm interface (Fig. 1b). As noted by Von der Mark, Von der Mark & Gay (1976) this staining was mainly restricted to a sharp line under the ectoderm although there was some staining of the outermost cell layers in the presumptive dermis (Fig. 1c). It should be noted that the dermis of the body wall was also stained with antitype I collagen antibodies but to a greater depth than in the limb (Fig. 1c). In addition, type I collagen was found in the developing cartilage elements. However, the individual dorsal muscle masses displayed a definite lack of staining although there was an indication of anti-type I collagen staining along the dorsal surface of the muscle mass. A similar lack of staining was found in the ventral muscle mass which has just divided at this stage.

Staining with anti-type III collagen antibodies was solely restricted to the ectoderm-mesoderm junction in the limb (Fig. 1 d). Any suggestion that the antitype III collagen serum was inactive could be ruled out because there was clear staining around some of the major blood vessels in the body of the embryo as shown in Fig. 1 (e).

Type V collagen presented a similar distribution to type III collagen at the ectoderm-mesoderm junction but was also detected in the cartilage elements (Fig. 1f). In summary, at stage 27, there was no indication of a band of any type of collagen appearing at either the site of the future division of the dorsal muscle mass or between the two parts of the ventral muscle mass.

Stage 29 : (6·5 days incubation)

Fig. 2a shows that the dorsal mass has completed its first division and is in the process of subdividing again. The ventral mass remains in two parts. The positions of these various areas of muscle tissue are revealed as areas lacking distinct staining with anti-type I collagen antibodies (Fig. 2b and c ). There was still a clear staining of the ectoderm-mesoderm junction and more layers of prospective dermis were stained at this stage compared to Stage 27. There remains, however, a clear difference in the staining of the dermis of the body wall which was both intense and present to a greater and more sharply defined depth.

Fig. 2.

Transverse sections of mid-forearm level, stage-29 limb, stained with: (a) H/E; (b) and (c) anti-type 1 collagen serum; (d) anti-type III collagen serum. Abbreviations as in Fig. 1 and N = nerve. Scale bars represent 100 μm.

Fig. 2.

Transverse sections of mid-forearm level, stage-29 limb, stained with: (a) H/E; (b) and (c) anti-type 1 collagen serum; (d) anti-type III collagen serum. Abbreviations as in Fig. 1 and N = nerve. Scale bars represent 100 μm.

Type I collagen was still present in the cartilage elements although it was more localized at the periphery of the elements, in agreement with Von der Mark et al. (1976).

Anti-type III collagen staining was again restricted to the ectoderm-mesoderm junction (Fig. 2d). Anti-type V staining was also found at this location but there was additional staining in the dermis and other areas of loose mesenchyme surrounding the muscle areas.

Stage 30 : (7-5 days incubation)

At this stage a further condensation of the muscle areas has occurred such that they are now clearly defined (Fig. 3 d).

Fig. 3.

Transverse sections of mid-forearm level, stage-30 limb, stained with: (a) H/E; (b) and (c) anti-type I collagen serum; (d) anti-type III collagen serum; (e) anti-type V collagen serum. Abbreviations as in Fig. 1. Scale bars represent: 100μm in a, b, d and e; 50μm in c.

Fig. 3.

Transverse sections of mid-forearm level, stage-30 limb, stained with: (a) H/E; (b) and (c) anti-type I collagen serum; (d) anti-type III collagen serum; (e) anti-type V collagen serum. Abbreviations as in Fig. 1. Scale bars represent: 100μm in a, b, d and e; 50μm in c.

The distribution of anti-type I collagen staining was similar to Stage 29 with a clearer indication of the presence of type I collagen in the dermis and loose mesenchyme connective tissue around the muscle areas (Fig. 3b and c ). Compared with Stage 29 there was a further restriction of the anti-type I collagen staining to the periphery of the cartilage elements.

Staining with antibodies to type III collagen revealed that this collagen type was beginning to appear in the outermost layers of the dermis at Stage 30 (Fig. 3d) and there was some indication of staining in the perichondrium. By contrast, anti-type V collagen serum produced an overall staining of all nonmuscle mesenchymal tissue including the cartilage elements (Fig. 3e). A thin band of brighter fluorescence was also observed corresponding to the basal lamina of the central blood vessel.

Stage 35: (9 days incubation)

The final subdivisions of the muscle masses have taken place by this stage (Fig. 4a), so that the process of muscle pattern formation has ended. This stage is therefore more concerned with the further cytodifferentiation and growth of the separate muscles. It is at this stage that the compartmentalization of muscle fibres into bundles becomes evident, with the appearance of the first clear signs of the epi- and perimysium components of the muscle connective tissue (Shellswell, 1977b).

Fig. 4.

Transverse sections of mid-forearm level, stage-35 limb, stained with: (a) H/E; (b) anti-type 1 collagen serum; (c) anti-type III collagen serum. E = é pimysium, M = individual muscles, T = tendon. Scale bars represent: 200μm in a; 100 μm in b ande.

Fig. 4.

Transverse sections of mid-forearm level, stage-35 limb, stained with: (a) H/E; (b) anti-type 1 collagen serum; (c) anti-type III collagen serum. E = é pimysium, M = individual muscles, T = tendon. Scale bars represent: 200μm in a; 100 μm in b ande.

Antibody staining reveals that both types I and III collagen are found in the developing epimysium (Fig. 4b and c ). Thus it is interesting that type III collagen has yet to achieve its preponderance in the perimysium as described in our earlier work on adult chicken muscle (Bailey et al. 1979). Treatment with anti-type V collagen serum revealed an overall staining throughout non-muscle mesenchymal tissues which suggests that the restricted localization of this collagen type in the endomysium of adult muscle also occurs at a later stage.

There is now clear evidence that the cytodifferentiation of limb muscle cells is specified in somitic cells well before limb outgrowth. Shellswell & Wolpert (1977) have described how the formation of the pattern of muscles in the wing can be viewed in the same terms as have been suggested for the formation of the pattern of wing cartilage elements (Summerbell, Lewis & Wolpert, 1975). These latter authors suggested that pattern formation in the wing depended on the assignment of positional values to cells within the early undifferentiated limb bud. These positional values are then interpreted by the cells to give the appropriate cytodifferentiation, for example muscle, cartilage, tendon or connective tissue. Shellswell & Wolpert (1977) and Shellswell (1977b) have demonstrated that the wing muscle pattern can respond to changes in positional value within the limb produced by various experimental procedures. Thus it is clear that although the formation of the muscle pattern may depend upon a system of positional information within the wing, this process must occur independently of the assignment of muscle cytodifferentiation which occurs at a much earlier stage.

The formation of the muscle pattern is achieved through the division of the muscle masses, and Wolpert (1978) has suggested that this depends on some activity of the muscle connective tissue cells. There is no information on the distribution of these somatopleurally derived cells at early stages but it is likely that they are present within the muscle masses from the beginning because there is evidence that they do not invade at later stages (Shellswell, 1977b). If invasion does not occur, how might muscle connective tissue cells influence where the divisions are made ? We have shown here that muscle mass division does not depend on the gross production of those particular types of collagen which are synthesized by myogenic cultures in vitro (Bailey et al. 1979). Thus the possibility that the differentiating epimysium separates the muscles by wrapping them up with layers of collagen which characterize adult epimysium appears unlikely. Indeed, we have found that the spatial organization of the collagen types which characterizes the various components of the muscle connective tissue (Duance et al. 1977; Bailey et al. 1979) does not appear until after the muscle masses have ceased dividing. This is particularly clear in the case of type III collagen which is only beginning to appear in the perimysium at Stage 36 (10 days). Von der Mark, Von der Mark, Herrmann & Timpl (1978) have reported in a brief abstract that they find type III collagen as a reticular meshwork in muscle some time after day 5 but give no further details as to the precise timing of appearance of this collagen type. Similarly, type V collagen has also been shown to be present in myogenic areas at an early stage, stage 25 (Sasse, Von der Mark & Von der Mark, 1978); however, our own studies could not confirm this finding. It may be that these authors were studying proximal limb muscle which differentiates earlier than at more distal levels. Shellswell (1977b) found that patchy areas of basement membrane around the developing myotubes did not become evident in the mid-forearm level until about stage 30. Our present study did not detect type V collagen until well after this stage, possibly reflecting the different resolution obtainable with electron versus immunofluorescence microscopy.

In a more recent publication, Von der Mark & Von der Mark (1979) have described how type V collagen originates in the chick embryo mainly from skeletal, cardiac and smooth muscle but state that they could not detect it in cartilage or bone. However, this finding is at variance with that of Rhodes & Miller (1978), who found the B chain of type-V collagen in infant human cartilage. Our own results show type V collagen at early stages of cartilage formation and support these findings of Rhodes and Miller. These conflicting results could be explained by the specificity of the different antibody preparations, with our antibodies recognizing antigenic determinants on the B chain whilst those of Von der Mark recognizing determinants only on the A chain, which according to Rhodes and Miller is absent from cartilage.

In conclusion, our studies have shown the production of collagen is not a gross mechanical basis for the separation of the embryonic limb muscle masses. However, collagen fibres make up only one component of the extracellular matrix and it is possible that connective tissue may control muscle mass division through the production and spatial arrangement of other components of extracellular matrix. In fact, there have been clear indications of the involvement of glycosaminoglycans (Morriss & Solursh, 1978) and fibronectin (Vaheri, Ruoslahti & Mosher, 1978) in early morphogenetic events. However, it may well be that only small amounts of these molecules are required, such that collagen may still be able to provide the necessary information to guide a rearrangement of myogenic cells, on a ‘sorting-out’ basis as proposed by Wolpert (1978). Indeed, Reddi (1976) has suggested that subtle changes in the charge of the collagen molecules may provide significant developmental information, and the amounts of collagen necessary for this may be too small for resolution by immunofluorescence microscopy. To determine whether collagen or any other connective tissue component is involved in the morphogenesis of muscle at this more subtle level, we are currently extending our immunohistochemical studies to the electron microscope level.

This work was supported by the Medical Research Council and the Arthritis and Rheumatism Council. We are grateful to A. Cousins and R. Mounsdon for their expert technical assistance.

Bailey
,
A. J.
,
Shellswell
,
G. B.
&
Duance
,
V. C.
(
1979
).
Identification and change of collagen types in differentiating myoblasts and developing chick muscle
.
Nature, Lond
.
278
,
67
69
.
Bradley
,
S. J.
(
1970
).
An analysis of self differentiation of chick limb buds in chorio-allantoic grafts
.
J. Anat
.
107
,
479
490
.
Chevallier
,
A.
(
1979
).
Role of the somitic mesoderm in the development of the thorax in bird embryos. II. Origin of the thoracic and appendicular musculature
.
J. Embryol. exp. Morph
.
49
,
73
88
.
Chevallier
,
A.
,
Kieny
,
M.
&
Mauger
,
A.
(
1977
).
Limb-somite relationship: origin of the limb musculature
.
J. Embryol. exp. Morph
.
43
,
263
278
.
Christ
,
B.
,
Jacob
,
H. J.
&
Jacob
,
M.
(
1977
).
Experimental analysis of the origin of the wing musculature in avian embryos
.
Anat. Embryol
.
150
,
171
186
.
Duance
,
V. C.
,
Restall
,
D. J.
,
Beard
,
H.
,
Bourne
,
F. J.
&
Bailey
,
A. J.
(
1977
).
The location of three collagen types in skeletal muscle
.
FEBSLett
.
79
,
248
252
.
Hamilton
,
H. L.
(
1952
).
Lillie’s Development of the Chick
.
New York
:
Holt, Rinehart & Winston
.
Hauschka
,
S. D.
&
Konigsberg
,
I. R.
(
1966
).
The influence of collagen on the development of muscle clones
.
Proc. natn. Acad. Sci., U.S.A
.
55
,
119
126
.
Hunt
,
E. A.
(
1932
).
The differentiation of chick limb buds in chorioallantoic grafts with special reference to the muscles
.
J. exp. Zool
.
62
,
57
91
.
Morriss
,
G. M.
&
Solursh
,
M.
(
1978
).
Regional differences in mesenchymal cell morphology and glycosaminoglycans in early neural-fold stage rat embryos
.
J. Embryol. exp. Morph
.
46
,
37
52
.
Reddi
,
A. H.
(
1976
).
Collagen and cell differentiation
.
In ’Biochemistry of Collagen’1
(ed.
G. N.
Ramachandran
&
A. H.
Reddi
.
New York
:
Plenum Press
.
Rhodes
,
R. K.
&
Miller
,
E. J.
(
1978
).
Physicochemical characterisation and molecular organisation of collagen A and B chains
.
Biochem
.
17
,
3442
3448
.
Sainte-Marie
,
G.
(
1962
).
A paraffin embedding technique for studies employing immunofluorescence
.
J. Histoch. Cytoch
.
10
,
250
256
.
Sasse
,
J.
,
Von Der Mark
,
K.
&
Von Der Mark
,
H.
(
1978
).
AB collagen: a new marker for studies of muscle differentiation
.
J. Cell Biol
.
79
,
323a
.
Shellswell
,
G. B.
(
1977a
).
The formation of discrete muscles from the chick wing dorsal and ventral muscles in the absence of nerves
.
J. Embryol. exp. Morph
.
41
,
269
277
.
Shellswell
,
G. B.
(
1977b
).
Studies on the development of the pattern of muscles and tendons in the chick wing
.
Ph.D. thesis
,
University of London
.
Shellswell
,
G. B.
&
Wolpert
,
L.
(
1977
).
The pattern of muscle and tendon developments in the chick wing
.
In ‘Vertebrate Limb and Somite Morphogenesis’
, (ed.
D. A.
Ede
,
J. R.
Hinchliffe
&
M.
Balls
).
Cambridge
:
Cambridge University Press
.
Sullivan
,
G. E.
(
1962
).
Anatomy and embryology of the wing musculature of the domestic fowl (Gallus)
.
Aust. J. Zool
.
10
,
458
518
.
Summerbell
,
D.
,
Lewis
,
J. H.
&
Wolpert
,
L.
(
1975
).
Positional information in chick limb morphogenesis
.
Nature, Lond
.
244
,
492
496
.
Vaheri
,
A.
,
Ruoslahti
,
E.
&
Mosher
,
D. F.
(
1978
).
Fibroblast surface protein
.
Annis N.Y. Acad. Sci
.
312
,
1
456
.
Von Der Mark
,
H.
&
Von Der Mark
,
K.
(
1979
).
Isolation and characterisation of collagen A and B chains from chick embryos
.
FEBS Lett
.
99
,
101
105
.
Von Der Mark
,
K.
,
Von Der Mark
,
H.
&
Gay
,
S.
(
1976
).
Study of differential collagen synthesis during development of the chick embryo by immunofluorescence. II. Localisation of Type I and Type II collagen during long bone development
.
Devi Biol
.
53
,
153
170
.
Von Der Mark
,
K.
,
Von Der Mark
,
H.
,
Herrmann
,
H.
&
Timpl
,
R.
(
1978
).
Immuno-fluorescent studies of type I, II, III and IV collagen development in the chick embryo
.
Upsala, J. Med. Sci
.
82
,
110
.
Wolpert
,
L.
(
1978
).
Pattern formation in biological development
.
Sci. Amer
.
239
,
124
137
.