EXPERIMENTAL
CEI,I.
RESEARCH
193,
2’??:11
( 1991)
SHORT NOTE In Vitro Attachment of Skeletal Muscle Fibers to a Collagen Gel Duplicates the Structure of the Myotendinous Junction SOMPORN SWASDISON AND RICHARD
The myotendinous junction (MTJ) and its associated cells and connective tissue are important structures involved in transmission of contractile force from skeletal muscle to tendon. A model culture system was developed to investigate the formation of the MTJ and its attachment to collagen fibers. Skeletal muscle cells were cultured in a well modeled from two layers of a native gel of type I collagen. Muscle cells cultured in this manner formed attachments to the collagen gel and developed into highly contractile multinucleated muscle fibers with the development of extensive terminal invaginations of the sarcolemma. In addition, the subsarcolemma at the ends of muscle fibers showed areas of increased electron density which corresponded well with the termini of myofibrils. The results indicate that the development of sarcolemmal invaginations at the end of a muscle fiber probably occurs intrinsically durin uiuo. The direct association ing muscle development of collagen fibers with the basal lamina at the end of muscle fibers was only occasionally observed in culture, suggesting that other fibrils or proteins may also be involved in the attachment of collagen fibers to the SC lwl Academic basal lamina of muscle fibers at the MTJ. Press.
Inc.
INTRODUCTION
The myotendinous junction (MTJ) has been extensively studied morphologically [l-6] and forms the principal structure involved in transmission of force
I To whom correspondence and reprint requests should be addressed at Department of Cell Biology, 605 Volker Hall, Box 302, IJniversity of Alabama at Birmingham, Birmingham, AL 35294. Fax: (205) 934-7029.
MAYNE’
generated by muscle contraction to the tendon and extracellular matrix (ECM). The most prominent structural features of the MTJ that are observed by electron microscopy include the extensive invaginations of the sarcolemma and the areas of increased electron density in the subsarcolemma to which actin filaments from the terminal sarcomere are attached. Invaginations of the junctional sarcolemma serve to increase the membrane surface area thus increasing the adhesive interface between the muscle sarcolemma and the tendon so that the contractile stress (force per unit area) loaded on the junctional sarcolemma is decreased [4]. However, the mechanism of attachment of the collagen fibrils of the tendon to the basal lamina of the muscle fiber remains poorly understood. It is known that the sarcolemmal area of the MTJ is enriched in the pl subunit of the integrin family [7-g] and subsarcolemmal regions of increased electron densit,y were shown by immunolocalization studies to contain vinculin, talin, and possibly cu-actinin [lo-131. Previously, these proteins were all found at focal adhesion sites of fibroblasts in culture [ 141. However, for the MTJ, the junctional sarcolemma is surrounded by a continuous basal lamina which appears identical to the basal lamina elsewhere in skeletal muscle. Our previous study of adult chicken skeletal muscle using immunofluorescent staining with specific monoclonal antibodies showed that the basal lamina of the MTJ contains an accumulation of laminin and heparan sulfate proteoglycan but not of type IV collagen [9]. Other studies of the MTJ have shown that from the lamina densa small diameter collagen fibrils run between fibroblasts at the end of muscle fibers and gradually blend into the larger diameter collagen fibers of the tendon which mainly are type I collagen [ 151. This area is called the microtendon [15], and it is a major site of injury during exercise of muscle and tendons [7]. The highly organized development of skeletal muscle and its interaction with the tendon and surrounding ECM cannot be easily investigated in Go. To study the develop-
227
001 l-48%7/91 $X0(1 Copyright c 1991 hy Academic Press. Inc. All rights of’ reproduction in any fbrm reserved.
228
SHORT
NOTE
ment of the MTJ and its association with collagen fibrils, we report an in vitro technique which duplicates the structural features of the MTJ.
MATERIALS
AND METHODS
C’ell culture. Hreast muscle from l&day quail embryos was cleaned of connective tissue, minced, and mechanically dissociated by pipetting up and down in culture medium [16]. The cell suspension was plated at a concentration of 500,000 cells/ml into wells made from two layers of a native type I collagen gel. Cultures were incubated at 37°C in a humidified atmosphere with 5% CO,, 95% air. After 3 days of culture, the muscle cells were covered with an additional collagen gel. Culture medium was changed daily. The medium consisted of 10% Eagle’s minimum essential medium (MEM) supplemented with I? antibioticcantimycotic solution, 1% L-glutamine, lo!‘; horse serum. and 5% embryo extract (all tissue culture reagents were obtained from GIBCO, Grand Island, NY). Culture u!rll preparation. The wells were prepared from native collagen. One milliliter of neutralized collagen solution (2 mg/ml Vitrogen, Collagen Corp., Palo Alto, CA) was poured into a 35.mm culture dish and allowed to polymerize at 37°C for 30 min. A cylindrical insert made by gluing a plastic mesh to the bottom of a small plastic ring was centrally placed on the first layer of gel and a second gel layer poured outside the insert. After polymerization, the insert was removed to create a circular well. The muscle cell suspension (0.1 ml) was plated into the well and, after 3 days, the cells were covered with an additional layer of gel. Contraction of individual muscle fibers was recorded using an Ikegami Model ITC-510 television camera and the micrographs were taken on Kodak Tmax-100 tape. After 3 weeks of culture within the collagen gel, cultures were prepared for electron microscopy using standard procedures. Small cubes of the gel con taining the ends of muscle fibers were fixed with 3% glutaraldehyde in 0.1 M phosphate-buffered saline (PBS), pH 7.3, at 4°C for 2 h and postfixed in 2’:; 0~0, in PBS for 1.5 h. After extensive rinsing with PBS, the samples were dehydrated in graded alcohol and propylene oxide before being embedded in Spurr’s resin (Electron Microscopy Sciences, Ft. Washington, PA). Clltrathin sections were prepared using an I,KB NOVA llltrotome and stained with uranyl acetate (1%) and lead tit rate (0.2”b ).
RESULTS Myotube formation was observed 48-72 h after plating muscle cells in a collagen well. After 6 days of culture within the gel, the culture was rich in muscle fibers which became spontaneously contractile. The developing fibers tended to grow together laterally and began to enter the collagen gel outside the well. The muscle fibers formed attachments to the collagen gel and began to pull on the gel during spontaneous contractions (Fig. 1).
micrographs taken from a video FIG. 1. Series of phase-contrast camera recording of 6-day muscle culture in the collagen gel showing contraction of a single muscle fiber. The distance of the contraction can be observed by comparing the position of the end of the muscle fiber relative to the fixed arrow in each photograph. The time in the contraction cycle in seconds is labeled for each frame at the lower right corner (bar = 50 pm).
SHORT
NOTE
229
FIG. 2. Electron micrograph of a longitudinal section of a single muscle fiber grown in a collagen gel for 3 weeks. The end of the muscle fiber shows extensive invaginations with electron-dense subsarcolemmal attachment sites. Note the alignment of the collagen fibers to the end of the muscle fiber (arrows) (magnification, 5940X).
Eventually, after several seconds of relaxation, each muscle fiber returned to its original location (Fig. 1). Electron micrographs of longitudinal sections of muscle fibers after 3 weeks in culture showed well-developed muscle fibers with highly organized sarcomeres. The ends of muscle fibers showed extensive invaginations of the sarcolemma (Fig. 2) with electron-dense attachment sites immediately below the sarcolemma (Fig. 3, arrowheads). These dense areas were present where actin filaments from the last sarcomere terminated at the end of muscle fibers (Fig. 3, large arrows). External to the sarcolemma, muscle fibers were surrounded by a continuous basal lamina which always extended into the invaginations (Fig. 3, small arrows). Close association of collagen fibers with the basal lamina of the muscle were sometimes observed (Fig. 3, open arrows), but direct insertions of collagen fibers into the basal lamina at the end of the muscle fiber were not observed.
DISCUSSION
It appears that the present culture system represents a simple model with which to study the attachment of a muscle fiber to fibrils of native collagen. In previous studies in which a monolayer of muscle cells was grown on a thin film of collagen, the myotubes on becoming contractile often detach from the culture dish and degenerate. This does not occur if the muscle cells are
grown within and attach to a collagen gel. We have not observed degenerating cells in our cultures on electron microscopy and the cultures can be maintained for several weeks. At present, we are unable to recognize many attachment sites of collagen to the muscle basal lamina but it would appear that lateral associations of collagen fibrils with the surface of the basal lamina (Fig. 3) may be important as attachment sites. We also do not know the potential role of type VI collagen at the myotendinous junction. This collagen has numerous potential collagen binding domains and is often observed as a network between the larger collagen fibrils [17-181. The results suggest that the formation of the sarcolemma1 invaginations probably arises intrinsically from the muscle fiber during myogenesis. It may arise from the contractile force of myofilaments since it has been shown that sarcolemma folding can occur at the areas where myofilaments are associated with subsarcolemma1 densities during chick MTJ development [6]. Immunofluorescent staining of the present cultures after 3 weeks showed that the basal lamina surrounding the developing muscle fibers contains type IV collagen, laminin, and heparan sulfate proteoglycan and that, intracellularly, the muscle fibers contain both neonatal and adult type myosin heavy chain (unpublished results). Recently, two other groups of investigators have reported culture systems for the long-term development of skeletal muscle in vitro [19, 201. However, in neither study was the end of the muscle fibers investigated. In one study, chicken skeletal muscle organogenesis was
230
SHORT
NO’I’JC
FIG. 3. Higher magnification electron micrograph of the same muscle fiber in Fig. 2. Note electron-dense attachment sites (arrowheads) to which myofilament,s terminate (large arrows). Also, note the extension of’ basal lamina into the invaginations (small arrows). The collagen fibers do not appear to insert directly into the muscle hasal lamina but are often observed parallel to the basal lamina (open arrows) (magnification, 23,430X).
investigated by using computerized mechanical cell stimulator [19]. In this culture system, muscle cells are grown on a collagen-coated flexible membrane which is slowly stretched so that the developing myotubes become aligned. In another study, the morphogenesis and histogenesis of long-term muscle cultures were studied in a chamber made from a membrane (Saran Wrap) pinned to Sylgard resin [20]. In both culture systems, muscle fibers formed a bundle-like appearance with connective tissue compartments which resembled the endomysium, perimysium, and endomysium of muscle in uiuo. Biochemical and morphological studies showed that the cultured muscle fibers were well differentiated. However, the culture system which we have developed potentially provides a simple way to study the mechanism of transmission of force from muscle to tendon.
This investigation was supported hy research Grant AR S7984 f’rom the National Institutes of Health and by the MDA.
REFERENCES 1. 2. :i. 4. 5. 6. 7.
Ishikawa. H. (196:i) Arch Histo/. Jnpon. 25(X), 275-296. Mackay, B., Harrop, ‘I’. CJ.,and Muir, A. R. (1969) Acta Amt. '73, 588-604. Trotter, .J. A., Corhett, K., and Avner, H. P. (1981) Anc11.h!w. 20 1, 29%:w2. Trotter, ,J. A., Hsi, K., Samora, A., and Wof’sy, C. (198.5) Amt. Rec. 213, 26-32. Ovalle, W. K. (1987) And. Enhyvol. 176, 281-294. Tidball, ?J.G., and Lin, C. (1989) Cell ‘I’issw Krs. 257, 77-84. Garrett, W., ,Jr., and Tidball, J. (1988) in Injury and Repair of the Musculoskeletal Soft Tissues (Woo, S. L.-Y., and Buck-
SHORT Walter, J. A., Eds.). pp. lil-207, paedic Surgeons.
American
Academy
of’ Ortho-
8. Bozyczko. D., Decker, C., Muschler. ,J., and Horwitz, A. F. (1589) Ewp ('Pll RPS.183, 52-91. 9. Swasdison. S., and Mayne, H. (1989) (‘ell ?‘issuc Iics. 257, 59’i543.
10. Shear, C. K.. and Bloch, K. -1. (19%) J (‘r/l Rio/. 101, 240-256. Il. I”. 13.
Tidball, J. C;., O’Halloran, ‘I‘.. and Burridge, K. (1986) J. (‘P/I Rid 103, 1465 147%. Trotter, ,J. A.. Eberhard. S.. and Samora, S. (1983) (‘PII Mdil. 3, 431 43X. Tidball, .J. (>. ( 198’i) fLr[~. C’(zli Kcs. 170, 469-482.
Received September ]:I, lY9tl Revised version received October :{o. 1990
231
NOTE 14. If,. 16.
Burridge, K. (1986) Cancer Rev. 4, 18-78. Moore, M. CJ.(1983) Muscles Ncrue 6, 416-422. Caplan, A. I. (1976) J. Embryol. E:rp. Morphol.
17. Bonaldo,
P.. Russo, V., Bucciotti, F., Doliana, Patti, A. (1990) Hiochemistr~~ 29, 1246-1254. E., Timpl,
36, 175.-181. R., and Colom-
18.
Bruns. H. K., Press, W., Engvall, (1986) J. (‘dl Aid. 103, XX-404.
K., and Gross, J.
19.
Vandenbnrgh, H. H., and Karlisch, Hid 25, 607 616.
20.
Strohman, K. C., Bayne, E., Spector, D., Ohinata, T., MicouEastwood, ,J., and Maniotis, A. (1990) In Vitro (‘d. Zku. Hiol. 26. 2OlL208.
P. (1989) In Vitro (‘cdl. Iku.
CEI,I.
RESEARCH
193,
2’??:11
( 1991)
SHORT NOTE In Vitro Attachment of Skeletal Muscle Fibers to a Collagen Gel Duplicates the Structure of the Myotendinous Junction SOMPORN SWASDISON AND RICHARD
The myotendinous junction (MTJ) and its associated cells and connective tissue are important structures involved in transmission of contractile force from skeletal muscle to tendon. A model culture system was developed to investigate the formation of the MTJ and its attachment to collagen fibers. Skeletal muscle cells were cultured in a well modeled from two layers of a native gel of type I collagen. Muscle cells cultured in this manner formed attachments to the collagen gel and developed into highly contractile multinucleated muscle fibers with the development of extensive terminal invaginations of the sarcolemma. In addition, the subsarcolemma at the ends of muscle fibers showed areas of increased electron density which corresponded well with the termini of myofibrils. The results indicate that the development of sarcolemmal invaginations at the end of a muscle fiber probably occurs intrinsically durin uiuo. The direct association ing muscle development of collagen fibers with the basal lamina at the end of muscle fibers was only occasionally observed in culture, suggesting that other fibrils or proteins may also be involved in the attachment of collagen fibers to the SC lwl Academic basal lamina of muscle fibers at the MTJ. Press.
Inc.
INTRODUCTION
The myotendinous junction (MTJ) has been extensively studied morphologically [l-6] and forms the principal structure involved in transmission of force
I To whom correspondence and reprint requests should be addressed at Department of Cell Biology, 605 Volker Hall, Box 302, IJniversity of Alabama at Birmingham, Birmingham, AL 35294. Fax: (205) 934-7029.
MAYNE’
generated by muscle contraction to the tendon and extracellular matrix (ECM). The most prominent structural features of the MTJ that are observed by electron microscopy include the extensive invaginations of the sarcolemma and the areas of increased electron density in the subsarcolemma to which actin filaments from the terminal sarcomere are attached. Invaginations of the junctional sarcolemma serve to increase the membrane surface area thus increasing the adhesive interface between the muscle sarcolemma and the tendon so that the contractile stress (force per unit area) loaded on the junctional sarcolemma is decreased [4]. However, the mechanism of attachment of the collagen fibrils of the tendon to the basal lamina of the muscle fiber remains poorly understood. It is known that the sarcolemmal area of the MTJ is enriched in the pl subunit of the integrin family [7-g] and subsarcolemmal regions of increased electron densit,y were shown by immunolocalization studies to contain vinculin, talin, and possibly cu-actinin [lo-131. Previously, these proteins were all found at focal adhesion sites of fibroblasts in culture [ 141. However, for the MTJ, the junctional sarcolemma is surrounded by a continuous basal lamina which appears identical to the basal lamina elsewhere in skeletal muscle. Our previous study of adult chicken skeletal muscle using immunofluorescent staining with specific monoclonal antibodies showed that the basal lamina of the MTJ contains an accumulation of laminin and heparan sulfate proteoglycan but not of type IV collagen [9]. Other studies of the MTJ have shown that from the lamina densa small diameter collagen fibrils run between fibroblasts at the end of muscle fibers and gradually blend into the larger diameter collagen fibers of the tendon which mainly are type I collagen [ 151. This area is called the microtendon [15], and it is a major site of injury during exercise of muscle and tendons [7]. The highly organized development of skeletal muscle and its interaction with the tendon and surrounding ECM cannot be easily investigated in Go. To study the develop-
227
001 l-48%7/91 $X0(1 Copyright c 1991 hy Academic Press. Inc. All rights of’ reproduction in any fbrm reserved.
228
SHORT
NOTE
ment of the MTJ and its association with collagen fibrils, we report an in vitro technique which duplicates the structural features of the MTJ.
MATERIALS
AND METHODS
C’ell culture. Hreast muscle from l&day quail embryos was cleaned of connective tissue, minced, and mechanically dissociated by pipetting up and down in culture medium [16]. The cell suspension was plated at a concentration of 500,000 cells/ml into wells made from two layers of a native type I collagen gel. Cultures were incubated at 37°C in a humidified atmosphere with 5% CO,, 95% air. After 3 days of culture, the muscle cells were covered with an additional collagen gel. Culture medium was changed daily. The medium consisted of 10% Eagle’s minimum essential medium (MEM) supplemented with I? antibioticcantimycotic solution, 1% L-glutamine, lo!‘; horse serum. and 5% embryo extract (all tissue culture reagents were obtained from GIBCO, Grand Island, NY). Culture u!rll preparation. The wells were prepared from native collagen. One milliliter of neutralized collagen solution (2 mg/ml Vitrogen, Collagen Corp., Palo Alto, CA) was poured into a 35.mm culture dish and allowed to polymerize at 37°C for 30 min. A cylindrical insert made by gluing a plastic mesh to the bottom of a small plastic ring was centrally placed on the first layer of gel and a second gel layer poured outside the insert. After polymerization, the insert was removed to create a circular well. The muscle cell suspension (0.1 ml) was plated into the well and, after 3 days, the cells were covered with an additional layer of gel. Contraction of individual muscle fibers was recorded using an Ikegami Model ITC-510 television camera and the micrographs were taken on Kodak Tmax-100 tape. After 3 weeks of culture within the collagen gel, cultures were prepared for electron microscopy using standard procedures. Small cubes of the gel con taining the ends of muscle fibers were fixed with 3% glutaraldehyde in 0.1 M phosphate-buffered saline (PBS), pH 7.3, at 4°C for 2 h and postfixed in 2’:; 0~0, in PBS for 1.5 h. After extensive rinsing with PBS, the samples were dehydrated in graded alcohol and propylene oxide before being embedded in Spurr’s resin (Electron Microscopy Sciences, Ft. Washington, PA). Clltrathin sections were prepared using an I,KB NOVA llltrotome and stained with uranyl acetate (1%) and lead tit rate (0.2”b ).
RESULTS Myotube formation was observed 48-72 h after plating muscle cells in a collagen well. After 6 days of culture within the gel, the culture was rich in muscle fibers which became spontaneously contractile. The developing fibers tended to grow together laterally and began to enter the collagen gel outside the well. The muscle fibers formed attachments to the collagen gel and began to pull on the gel during spontaneous contractions (Fig. 1).
micrographs taken from a video FIG. 1. Series of phase-contrast camera recording of 6-day muscle culture in the collagen gel showing contraction of a single muscle fiber. The distance of the contraction can be observed by comparing the position of the end of the muscle fiber relative to the fixed arrow in each photograph. The time in the contraction cycle in seconds is labeled for each frame at the lower right corner (bar = 50 pm).
SHORT
NOTE
229
FIG. 2. Electron micrograph of a longitudinal section of a single muscle fiber grown in a collagen gel for 3 weeks. The end of the muscle fiber shows extensive invaginations with electron-dense subsarcolemmal attachment sites. Note the alignment of the collagen fibers to the end of the muscle fiber (arrows) (magnification, 5940X).
Eventually, after several seconds of relaxation, each muscle fiber returned to its original location (Fig. 1). Electron micrographs of longitudinal sections of muscle fibers after 3 weeks in culture showed well-developed muscle fibers with highly organized sarcomeres. The ends of muscle fibers showed extensive invaginations of the sarcolemma (Fig. 2) with electron-dense attachment sites immediately below the sarcolemma (Fig. 3, arrowheads). These dense areas were present where actin filaments from the last sarcomere terminated at the end of muscle fibers (Fig. 3, large arrows). External to the sarcolemma, muscle fibers were surrounded by a continuous basal lamina which always extended into the invaginations (Fig. 3, small arrows). Close association of collagen fibers with the basal lamina of the muscle were sometimes observed (Fig. 3, open arrows), but direct insertions of collagen fibers into the basal lamina at the end of the muscle fiber were not observed.
DISCUSSION
It appears that the present culture system represents a simple model with which to study the attachment of a muscle fiber to fibrils of native collagen. In previous studies in which a monolayer of muscle cells was grown on a thin film of collagen, the myotubes on becoming contractile often detach from the culture dish and degenerate. This does not occur if the muscle cells are
grown within and attach to a collagen gel. We have not observed degenerating cells in our cultures on electron microscopy and the cultures can be maintained for several weeks. At present, we are unable to recognize many attachment sites of collagen to the muscle basal lamina but it would appear that lateral associations of collagen fibrils with the surface of the basal lamina (Fig. 3) may be important as attachment sites. We also do not know the potential role of type VI collagen at the myotendinous junction. This collagen has numerous potential collagen binding domains and is often observed as a network between the larger collagen fibrils [17-181. The results suggest that the formation of the sarcolemma1 invaginations probably arises intrinsically from the muscle fiber during myogenesis. It may arise from the contractile force of myofilaments since it has been shown that sarcolemma folding can occur at the areas where myofilaments are associated with subsarcolemma1 densities during chick MTJ development [6]. Immunofluorescent staining of the present cultures after 3 weeks showed that the basal lamina surrounding the developing muscle fibers contains type IV collagen, laminin, and heparan sulfate proteoglycan and that, intracellularly, the muscle fibers contain both neonatal and adult type myosin heavy chain (unpublished results). Recently, two other groups of investigators have reported culture systems for the long-term development of skeletal muscle in vitro [19, 201. However, in neither study was the end of the muscle fibers investigated. In one study, chicken skeletal muscle organogenesis was
230
SHORT
NO’I’JC
FIG. 3. Higher magnification electron micrograph of the same muscle fiber in Fig. 2. Note electron-dense attachment sites (arrowheads) to which myofilament,s terminate (large arrows). Also, note the extension of’ basal lamina into the invaginations (small arrows). The collagen fibers do not appear to insert directly into the muscle hasal lamina but are often observed parallel to the basal lamina (open arrows) (magnification, 23,430X).
investigated by using computerized mechanical cell stimulator [19]. In this culture system, muscle cells are grown on a collagen-coated flexible membrane which is slowly stretched so that the developing myotubes become aligned. In another study, the morphogenesis and histogenesis of long-term muscle cultures were studied in a chamber made from a membrane (Saran Wrap) pinned to Sylgard resin [20]. In both culture systems, muscle fibers formed a bundle-like appearance with connective tissue compartments which resembled the endomysium, perimysium, and endomysium of muscle in uiuo. Biochemical and morphological studies showed that the cultured muscle fibers were well differentiated. However, the culture system which we have developed potentially provides a simple way to study the mechanism of transmission of force from muscle to tendon.
This investigation was supported hy research Grant AR S7984 f’rom the National Institutes of Health and by the MDA.
REFERENCES 1. 2. :i. 4. 5. 6. 7.
Ishikawa. H. (196:i) Arch Histo/. Jnpon. 25(X), 275-296. Mackay, B., Harrop, ‘I’. CJ.,and Muir, A. R. (1969) Acta Amt. '73, 588-604. Trotter, .J. A., Corhett, K., and Avner, H. P. (1981) Anc11.h!w. 20 1, 29%:w2. Trotter, ,J. A., Hsi, K., Samora, A., and Wof’sy, C. (198.5) Amt. Rec. 213, 26-32. Ovalle, W. K. (1987) And. Enhyvol. 176, 281-294. Tidball, ?J.G., and Lin, C. (1989) Cell ‘I’issw Krs. 257, 77-84. Garrett, W., ,Jr., and Tidball, J. (1988) in Injury and Repair of the Musculoskeletal Soft Tissues (Woo, S. L.-Y., and Buck-
SHORT Walter, J. A., Eds.). pp. lil-207, paedic Surgeons.
American
Academy
of’ Ortho-
8. Bozyczko. D., Decker, C., Muschler. ,J., and Horwitz, A. F. (1589) Ewp ('Pll RPS.183, 52-91. 9. Swasdison. S., and Mayne, H. (1989) (‘ell ?‘issuc Iics. 257, 59’i543.
10. Shear, C. K.. and Bloch, K. -1. (19%) J (‘r/l Rio/. 101, 240-256. Il. I”. 13.
Tidball, J. C;., O’Halloran, ‘I‘.. and Burridge, K. (1986) J. (‘P/I Rid 103, 1465 147%. Trotter, ,J. A.. Eberhard. S.. and Samora, S. (1983) (‘PII Mdil. 3, 431 43X. Tidball, .J. (>. ( 198’i) fLr[~. C’(zli Kcs. 170, 469-482.
Received September ]:I, lY9tl Revised version received October :{o. 1990
231
NOTE 14. If,. 16.
Burridge, K. (1986) Cancer Rev. 4, 18-78. Moore, M. CJ.(1983) Muscles Ncrue 6, 416-422. Caplan, A. I. (1976) J. Embryol. E:rp. Morphol.
17. Bonaldo,
P.. Russo, V., Bucciotti, F., Doliana, Patti, A. (1990) Hiochemistr~~ 29, 1246-1254. E., Timpl,
36, 175.-181. R., and Colom-
18.
Bruns. H. K., Press, W., Engvall, (1986) J. (‘dl Aid. 103, XX-404.
K., and Gross, J.
19.
Vandenbnrgh, H. H., and Karlisch, Hid 25, 607 616.
20.
Strohman, K. C., Bayne, E., Spector, D., Ohinata, T., MicouEastwood, ,J., and Maniotis, A. (1990) In Vitro (‘d. Zku. Hiol. 26. 2OlL208.
P. (1989) In Vitro (‘cdl. Iku.
