Fibroblast-mediated collagen gel contraction does not require fibronectin-╬▒5╬▓1 integrin interaction.код для вставкиСкачать
THE ANATOMICAL RECORD 234153-160 (1992) Fibroblast-Mediated Collagen Gel Contraction Does Not Require Fibronectin-a$, lntegrin Interaction - JAMES J. TOMASEK AND STEVEN K. AKIYAMA Department of Anatomical Sciences, University of Oklahoma Health Sciences Center, Oklahomu City, Oklahoma (J.J.T.); Laboratory of Developmental Biology, National Institute of Dental Research, National Institutes of Health, Bethesda, Maryland 20892 (S.K.A.) ABSTRACT Fibroblasts cultured within free-floating collagen gels can bind to and reorganize the surrounding collagen fibrils into a more dense and compact arrangement. Collagen gel contraction provides an in vitro model for studying fibroblast-collagen interactions important in wound healing, fibrosis, scar contraction, and connective tissue morphogenesis. We have assessed the role of fibronectin and its interaction with the aspl “high affinity” fibronectin-specific integrin receptor in collagen gel contraction. A variety of agents, which specifically inhibit fibronectin-a,P, interactions, were tested for their abilities to inhibit fibroblast-mediatedcollagen gel contraction. These included anti-a,p, monoclonal antibodies, the synthetic peptide GRGDSP, the cell adhesive fragment of fibronectin, and an antibody against the cell adhesive region of fibronectin. None of these agents inhibited collagen gel contraction. Therefore, it is concluded that fibronectin-a,P, interactions are not necessary for collagen gel contraction. However, collagen gel contraction is dependent on a member or members of the PI subfamily of integrin matrix receptors. A polyclonal antiserum and a monoclonal antibody, both directed against the P1 subunit of integrin matrix receptors, inhibited the spreading of fibroblasts in the collagen gel and inhibited collagen gel contraction. This study demonstrates that fibroblast-mediated collagen gel contraction is independent of fibronectin-a,& interactions but dependent on an interaction of p1 integrin matrix receptors with collagen fibers. o 1992 Wiley-Liss, Inc. Key words: Extracellular matrix, Wound healing, Morphogenesis Fibroblasts cultured within a collagen gel attach and to be the result of tractional forces exerted by fibrospread on the surrounding fibrils, assuming an elon- blasts during their spreading and migration in the colgate bipolar morphology (Elsdale and Bard, 1972; lagen gel (Stopak and Harris, 1982). These actin-genTomasek et al., 1982). Over time, the fibroblasts reor- erated forces must be transmitted to the collagen ganize the collagen fibrils and contract the collagen gel fibrils, either directly or indirectly, to result in collagen (Bell et al., 1979). Collagen gel contraction is believed gel contraction. The integrins, a family of extracellular matrix recepto resemble processes that are important in wound healing, fibrosis, scar contraction, and connective tis- tors, are present on the surface of fibroblasts (Albelda sue morphogenesis (Bell et al., 1979; Stopak and Har- and Buck, 1991; Hynes, 1987). These receptors appear ris, 1982). The contraction of collagen gels is an active to function by linking extracellular macromolecules cellular process. The rate of contraction is dependent with the cytoskeleton and promoting cell attachment, on cell number, the type and concentration of the col- migration and shape changes (Wayner et al., 1988; lagen, and serum concentration (Bell et al., 1979; But- Burridge et al., 1988; Akiyama et al., 1989; Clyman et tle and Ehrlich, 1983; Guidry and Grinnell, 19851, but al., 1990). The integrins are heterodimers consisting of not on collagen synthesis or degradation (Guidry and noncovalently associated a and p subunits (Albelda Grinnell, 1985). The interaction of fibroblasts with the and Buck, 1991; Hynes, 1987). They can be divided into surrounding collagen fibrils results in a more dense and compact organization of the matrix. This reorganization is dependent on an organized actin cytoskeleton, as treatment with cytochalasin D or B will inhibit Received October 30, 1991; accepted February 28, 1992. the spreading of fibroblasts in collagen gels and collaAddress reprint requests to Dr. James J. Tomasek, Department of gen gel contraction (Bell et al., 1979; Guidry and Grin- Anatomical Sciences, Biomedical Sciences Bldg., Rm. 553, The Uninell, 1985; Tomasek and Hay, 1984). The mechanical versity of Oklahoma Health Sciences Center, P.O. Box 26901, Oklareorganization of the collagen fibrils has been proposed homa City, OK 73190. 0 1992 WILEY-LISS, INC 154 J.J. TOMASEK AND S.K. AKIYAMA at least six different subfamilies, each with a different p subunit. Integrins from the p, subfamily are important in fibroblast adhesion to fibronectin and to collagen (Wayner and Carter, 1987; Takada et al., 1988; Wayner et al., 1988; Akiyama et al., 1989; Gullberg e t al., 1989). Fibroblast adhesion to fibronectin involves the integrin receptor a5p1 (Wayner and Carter, 1987; Akiyama et al., 1989), while fibroblast adhesion to collagen may be mediated by the integrin receptors alp1, a2p1and a3p1(Wayner and Carter, 1987; Takada et al., 1988; Wayner et al., 1988; Gullberg et al., 1989; Clyman et al., 1990). Fibronectin has been implicated to play a role in collagen gel contraction (Guillery et al., 1986). Fibronectin could potentially act as a ligand linking the cell surface to surrounding collagen fibers (Kleinman et al., 1981). However, other studies have suggested that fibronectin may not be necessary for collagen gel contraction (Guidry and Grinnell, 1985; Gullberg et al., 1990; Asaga et al., 1991). In addition, recent studies have demonstrated that adhesion to collagen can be independent of fibronectin (Wayner and Carter, 1987; Wayner et al., 1988; Takada et al., 1988; Gullberg et al., 1989; Clyman et al., 1990). To elucidate fibronectin’s role in this process, we investigated whether the “high affinity” interaction of fibronectin with the fibronectin-specific receptor is necessary for collagen gel contraction. Antibodies, synthetic peptides or fibronectin fragments, which can either compete with or block the fibronectin-a& integrin interaction, were used in collagen gel contraction assays. In addition, the role of p, integrins in collagen gel contraction was investigated using a n antiserum and a monoclonal antibody, which can inhibit the function of p1 integrins. Antibodies and Fibronectin Fragment Monoclonal anti-p, antibody (mAb 13) and monoclonal anti-a, antibody (mAb 16) were produced and characterized as previously described (Akiyama et al., 1989). Fab fragments of mAb 13 were prepared a s previously described (Akiyama et al., 1989). Monoclonal antibody against the fibronectin cell adhesive fragment was produced and characterized as previously described (McDonald et al., 1987). The 75 kD cell adhesive fragment of fibronectin was produced and characterized as previously described (Hayashi and Yamada, 1983; Yamada and Kennedy, 1984). Polyclonal anti-p, antiserum was obtained from Dr. Martin Hemler (Dana-Farber Cancer Inst., Boston, MA) and has been previously characterized (Takada et al., 1987). Monoclonal anti-a, antibody (PlD6) was obtained from Telios Pharmaceuticals and has been previously characterized (Wayner et al., 1988). Collagen Gels Collagen gels were manufactured by rapidly mixing together 0.125 ml of M-199 media with or without antibodies, synthetic peptides or fibronectin fragment with 0.125 ml of cells (4 x lo5 cells/ml) and 0.25 ml of collagen solution (1.5 mg/ml) in the well of a 24 well bacteriologic plate (Corning, Corning, NY). Plates were immediately incubated at 37°C in 5% C 0 2 and 95% air to promote collagen fibrillogenesis. Cells were harvested as described above except that they were washed and suspended in M-199 supplemented with 10% fibronectin-depleted FBS. Collagen solution was prepared as described previously (Tomasek et al., 1982; Tomasek and Hay, 19841, except that fibronectin-depleted FBS was used. In some experiments, complete MATERIALS AND METHODS FBS was used in place of the fibronectin-depleted FBS. The final cell concentration was 1 x lo5 cells/ml and Reagents Type I collagen (rat tail tendon, acetic acid extracted) final collagen concentration was 0.75 mg/ml. Gel contraction was determined by measuring the was obtained commercially (Collaborative Research, Bedford, MA). GRGDSP- and GRGESP-peptides were diameter of the gel to the nearest 0.25 mm using a obtained from Telios Pharmaceuticals (San Diego, CA). Nikon SMZ-1 stereoscope. Fibroblast morphology in M-199 medium was supplemented with 2 mM glu- the gel was documented with a n Olympus IMT2 intamine and 1%antibiotic-antimycotic (GIBCO, Grand verted microscope with Hoffman modulation contrast Island, NY). Fetal bovine serum (FBS) was purchased optics. from Irvine Scientific (Santa Ana, CA). FibronectinAssay for Cell Spreading depleted FBS was obtained by passing the FBS over a gelatin-sepharose affinity column (Sigma, Chemical Quantitation of fibronectin-mediated cell spreading Co., St. Louis, MO) as described (Engvall and Ruo- of human palmar fibroblasts was performed by a modification of previously described methods (Yamada and slahti, 1977). Kennedy, 1984; Akiyama et al., 1986, 1989). Multiwell Cells tissue culture dishes (96 wells, Falcon) were incubated Monolayer cultures of adult human palmar fibro- with 100 ~1 of 20 Fg/ml human plasma fibronectin blasts were established from explant cultures of pal- (Collaborative Research, Inc.) in Dulbecco’s phosphate mar aponeurosis. Normal-appearing palmar aponeu- buffered saline (D-PBS) overnight at 4°C and blocked rosis was obtained a s surgical discard tissue from for 30 min with 100 ~1 of 10 mg/ml heat-denatured patients undergoing carpal tunnel release. Pieces of (SOOC for 3 min) bovine serum albumin in D-PBS withtissue were placed onto 60 mm tissue culture dishes out C a + + or M g + + . The wells were washed seven (Falcon, Oxnard, CA), allowed to attach, and cultured times with D-PBS and the prepared substrates were in supplemented M-199 containing 10% FBS. Cells covered with serum-free M-199 which was removed were subcultured by treatment with 0.05% trypsin- just prior to use. All subsequent steps of the assay were 0.02% EDTA in Ca/Mg-free Hanks BSS (GIBCO) for 2 performed at 37°C. Subconfluent palmar fibroblasts min. Cells were washed three times with supplemented were washed with D-PBS and incubated for 2 min in media containing 10% FBS and cultured in 75 cm2 tis- 100 Fg/ml TPCK trypsin (Worthington Biochemical sue culture flasks (Falcon). Fibroblasts used in these Corp.) in D-PBS. The cells were then dislodged by experiments were between cell passages 4 and 10. shaking, suspended in a n equal volume of regular cul- FIBRONECTIN IN COLLAGEN GEL CONTRACTION 155 ture media, centrifuged, resuspended in regular culture media, allowed to recover from the trypsinization for 20 min a t 37"C, and counted. The cells were then centrifuged and resuspended in serum-free-M-199 at a concentration of 2 x lo5 celldml. Aliquots (50 p1) of cells were mixed with 50 pl aliquots of various agents to be tested in serum-free M-199, added to fibronectin prepared substrates, and incubated for 1hr at 37°C in a 95% air, 5% CO,, humidified atmosphere. Cells were then fixed by adding 100 ~1of glutaraldehyde in D-PBS directly to the wells. Cell spreading was quantitated by phase contrast microscopy as previously described (Grinnell et al., 1977; Yamada and Kennedy, 1984; Akiyama et al., 1986). The percentage of cells that spread was determined by counting four random microscopic fields of about 100 cells per field and calculating the average percent of cells spread per well. Duplicate wells were then averaged together. RESULTS Removal of Serum Fibronectin From FBS Has no Effect on Collagen Gel Contraction Normal palmar fibroblasts, when incorporated into a free-floating collagen gel, will reorganize the collagen fibrils resulting in contraction of the gel (Fig. 1). Contraction was quantitated by measuring the diameter of the gel at different time points. The majority of the contraction of the collagen gel occurred within the first 10 hr under the conditions used in this study (Fig. 2). Contraction occurred in the presence of fibronectin-depleted fetal bovine serum (Fig. 2). The plasma fibronectin concentration was reduced by at least 95% as determined by immunoblot analysis (data not illustrated). No differences were observed in collagen gel contraction in complete FBS or in fibronectin-depleted FBS, suggesting that exogenous plasma fibronectin is not necessary for collagen lattice contraction (Fig. 2). Fibronectin-a,p, interaction is not Necessary for Collagen Gel Contraction To determine whether fibronectin-a5Pl integrin matrix receptor interactions play a role in collagen gel contraction, a variety of agents which compete with or block this interaction were added to collagen gels. No inhibition of collagen gel contraction was observed with either the GRGDSP peptide (Fig. 3) or the 75 kd fibronectin cell adhesive fragment (Fig. 41, both of which compete with fibronectin-a5P1 interactions (Pytela et al., 1985a; Hayashi and Yamada, 1983; Yamada and Kennedy, 1984). In addition, no inhibition was observed with either a monoclonal antibody that binds to the cell adhesive domain of fibronectin (Fig. 5) or two different monoclonal anti-a, integrin antibodies, mAb 16 (Fig. 6) or P1D6 (data not shown), all of which block fibronectin-a,P, interactions (McDonald et al., 1987; Akiyama et al., 1989; Wayner et al., 1988). In addition, none of the agents tested affected the spreading of cells in collagen gels (data not shown). These agents were tested for their ability to specifically block the spreading of human Palmar fibroblasts on fibronectin. Palmar fibroblasts spread on substrates prepared with fibronectin (Fig. 7). shown in Figure 7, all these agents significantly inhibited cell spreading on fibronectin ( p < 0.01) at the Same concentrations as those used in the collagen gels. Fig. I . Darkfield photomicrographs of free-floating collagen gels after 0 (a),2 (b), and 24 hr ( c ) of incubation. Human palmar fibroblasts and type I collagen were mixed to yield 0.75 mg/ml and 1 X lo6 cells/ml. The collagen gels were cultured in M-199 supplemented with 10%fibronectin-depletedFBS as described in Materials and Methods. Contraction resulted in a symmetrical reduction in the diameter of the lattice. Bar = 3.5 mm. 156 J.J. TOMASEK AND S.K. AKIYAMA n N 150 E E W 0 100 2 6 50 " 0 0 5 10 15 20 25 0 5 10 Fig. 2.Effect of exogenous plasma fibronectin on contraction of collagen gels. The collagen gels were prepared as described in Materials and Methods and cultured in M-199 supplemented with either 10% normal FBS (-1 or 10% fibronectin-depleted FBS (&A). Similar contraction occurred under both conditions. Averages of duplicate collagen gels are shown and the total ranges are indicated by the vertical bars. 0 ' ' 5 . ' 10 ' ' 15 . 20 25 Time (hrs) Time (hrs) 01 15 ' 20 . ' 25 ' Fig. 4. Effect of the 75 kD fibronectin cell adhesive fragment on contraction of collagen gels. The collagen gels were prepared as described in Materials and Methods and cultured in M-199 supplemented with fibronectin-depleted FBS. Contraction occurred in the presence (&A) or absence (0-0) of 1mglml of the fragment. Averages of duplicate collagen gels are shown and the total ranges are indicated by the vertical bars. 0 5 10 15 20 25 Time (hrs) Time (hrs) Fig. 3.Effect of a soluble RGD-containing peptide on contraction of collagen gels. The collagen gels were prepared as described in Materials and Methods and cultured in M-199 supplemented with 10% fibronectin-depleted FBS. Contraction occurred in the presence of 1 mg/ml GRGDSP (&A), 1mglml GRGESP (U), and no peptide (-1. Averages of duplicate collagen gels are shown and the total ranges are indicated by the vertical bars. Fig. 5. Effect of a monoclonal antibody that binds to fibronectin's cell adhesive domain (mAb 333) on contraction of collagen gels. The collagen gels were prepared as described in Materials and Methods and cultured in M-199 10% fibronectin-depleted FBS.Contraction occurred in the presence (&A) or absence (c-0) of 100 pgiml of the antibody. Averages of duplicate collagen gels are shown and the total ranges are indicated by the vertical bars. Necessity of the p, Subfamily of lntegrin Matrix Receptors in Collagen Gel Contraction plete FBS (data not shown). This inhibition of contraction was dose-dependent, with the highest concentration resulting in almost total inhibition of contraction. Collagen gel contraction was not inhibited by either control nonimmune rabbit antiserum added a t the same dilution as the anti+, antiserum (Fig. 8A) or by 1 mg/ml of a control rat IgG,, monoclonal antibody, which is the same antibody subclass as mAb 13 (Fig. 8B). Also, monoclonal antibodies against the a,p? integrin matrix receptor, at much higher concentrations, did not affect collagen gel contraction (see above). In To determine the role of PI integrins, in general, in collagen gel contraction, antibodies were used which functionally block the interaction of all PI integrins with their extracellular matrix ligands. The addition of either a polyclonal anti-p, antiserum or a monoclonal anti+, antibody (mAb 13) inhibited collagen gel contraction (Fig. 8A,B). Inhibition occurred in the presence of fibronectin-depleted FBS (Fig. 8A,B) or com- + FIBRONECTIN IN COLLAGEN GEL CONTRACTION 0 5 10 15 20 157 25 Time (hrs) Fig. 6. Effect of anti-a, integrin antibody (mAb 16) on contraction of collagen gels. The collagen gels were prepared as described in Materials and Methods and cultured in M-199 + 10% fibronectin-depleted , pg/ml FBS. Contraction occurred in the presence of 1m g / m l ( ~ )200 (V-01,and 100 pg/ml (A-A) of anti-a, integrin antibody (mAb 16). Also shown is contraction in the presence of 1 mg/ml of a control rat IgG,, monoclonal antibody (0-0).Averages of duplicate collagen gels are shown and the total ranges are indicated by the vertical bars. I . 1 . I . I . I . 1 I 0 loo 0 5 10 15 20 25 Time (hrs) Fig. 8. Effects of different anti-p, antibodies on the contraction of collagen gels. The collagen gels were prepared as described in Materials and Methods and cultured in M-199 + 10% fibronectin-depleted FBS containing different concentrations of anti-p, antibody or control antibody. A Contraction in the presence of 1/10 (C.) and 1/50 (MI polyclonal anti+, antiserum. Also shown is contraction in the presence of 1/10 (u and )1/50 (A-A) pre-immune rabbit serum. B Contraction in the presence of 250 pg/ml(-), 100 pg/ml(V-V), 50 pgiml (A-A), and 25 pg/ml (M) anti-p, antibody (mAb 13). Also shown is contraction in the presence of 1 mg/ml of a control rat IgG,, monoclonal antibody (-). Averages of duplicate collagen gels are shown and the total ranges are indicated by the vertical bars. C G F 16 13 P 333 B Fig. 7. Effect of agents on palmar fibroblast spreading on fibronectin substrates. Palmar fibroblasts were added alone (C) or in the presence of GRGDSP (1mg/ml) (GI, 75 kD fibronectin cell adhesive fragment (1 mg/ml) (F), mAb 16 anti-a, integrin antibody (1mg/ml) (16), mAb 13 anti-p, integrin antibody (250 pg/ml) (131, P1D6 anti-a, integrin antibody (1:1500 dilution) (P),and mAb 333 anti-cell adhesive domain of fibronectin (100 pg/ml) (333). There was no spreading if the substrate was not coated with fibronectin (B). Each bar is the average of eight determinations from random fields of cells ( 2SEM). the control cells assumed an elongate bipolar configuration in the collagen gel (Fig. 9b). The inhibition of cell spreading and collagen gel contraction by anti-p, antibodies was reversible. Replacement of the antibody solution with media lacking antibody, after 24 hr in culture, resulted in cell elongation and collagen gel contraction (data not shown). DISCUSSION another control experiment, Fab fragments of mAb 13 also inhibited collagen gel contraction (data not shown). Cell spreading within the collagen gel was also inhibited by the anti+, antibodies a t concentrations which inhibited collagen gel contraction (Fig. 9). After 24 hours in the presence of the antibody, most of the cell bodies were still round with only short pseudopodia projecting into the collagen gel (Fig. 9a). In contrast, It has been previously demonstrated that fibroblasts will spread on collagen fibrils in hydrated collagen lattices and assume an elongate bipolar configuration (Elsdale and Bard, 1972; Tomasek et al., 1982). In addition, fibroblasts can exert force upon the collagen fibrils, resulting in reorganization and contraction of the collagen gel (Bell et al., 1979). Collagen gels provide a simple and very sensitive model for examining the functional interaction of cell surface receptors with their extracellular ligands. In this study, we demon- 158 J.J. TOMASEK AND S.K. AKIYAMA Fig. 9. Effect of anti-p, antibody (mAb 13) on cell spreading in collagen gels. The collagen gels were prepared as described in Materials and Methods and cultured for 24 hr in M-199 + 10% fibronectindepleted FBS containing 250 kg/ml of anti-p, antibody (a)or 1mg/ml of a control rat IgG,, monoclonal antibody (b). Cells were photographed with Hoffman modulation contrast optics. Bar = 75 pm. strate that fibroblasts can contract collagen gels independent of an interaction between fibronectin and the a& “high affinity” fibronectin-specific receptor. The ability of fibroblasts to spread on collagen fibrils and to contract the collagen gel is dependent upon the p1 subfamily of integrin matrix receptors. These results suggest that collagen gel contraction can occur by a direct interaction of integrin matrix receptors for collagen with surrounding collagen fibers. The role of P,-integrin matrix receptors in the contraction of collagen gels was examined with the use of antibodies against p1 integrins. Two different anti+, antibodies were used: (1)a polyclonal antiserum previously demonstrated to inhibit fibroblast attachment to and spreading on a collagen substratum (Takada et al., 1988); and (2) a monoclonal antibody that will inhibit the spreading of human palmar fibroblasts on a fibronectin substratum. The inhibition of collagen gel contraction with these antibodies demonstrates that this process is dependent upon P,-integrin matrix receptors. The effects of these antibodies on spreading and contraction were reversible, demonstrating that their inhibition was not due to toxicity. In addition, the inhibition was not due to effects of cross-linking the P,-integrin receptors on the cell surface with the antibodies, since Fab fragments of mAb 13were effective in inhibiting contraction. Previously, Gullberg et al. (1990)have demonstrated that a monospecific polyclonal anti-rat p1 antibody could inhibit collagen gel contraction by rat heart fibroblasts but could only partially inhibit contraction by human fibroblasts. In this study, we have demonstrated that a monoclonal antibody raised against the human p1 integrin subunit can inhibit collagen gel contraction by human fibroblasts. These results demonstrate that collagen gel contraction by different cell types from different species is dependent upon the P1 subfamily of integrin matrix receptors. In the present study, we determined that fibroblastmediated collagen gel contraction is not affected by a variety of inhibitors of cell-fibronectin interactions. The GRGDSP peptide has been previously shown to inhibit a variety of fibronectin-dependent processes (Hemler et al., 1987; Hynes, 1987; Albelda and Buck, 1991).In this paper, we demonstrate that it can inhibit fibronectin-mediated spreading of human palmar fibroblasts. This peptide does not inhibit fibroblast-mediated collagen gel contraction at similar concentrations, in agreement with a previous report by Gullberg et al. (1990). Although this peptide can inhibit numerous fibronectin-dependent processes, its affinity for the integrin receptor asplis about 100-fold lower than the intact fibronectin molecule (Akiyama and Yamada, 1985). The 75 kD cell-binding fragment of fibronectin interacts with cells with a similar affinity as the intact molecule (Akiyama et al., 1985). The addition of the 75 kD fragment to collagen gels does not inhibit contraction, similar to that observed for GRGDSP. This result demonstrates that even agents that can compete with intact fibronectin for binding to its receptor do not inhibit collagen gel contraction. consistent with these results, we have also observed that a monoclonal antibody which binds to the cell adhesive region of fibronectin and sterically hinders interaction with the aqpl integrin receptor (McDonald et al., 1987) does not inhibit collagen gel contraction. This report also demonstrates that directly blocking the interaction of the a5p1integrin receptor with fibronectin does not inhibit collagen gel contraction. mAb 16 binds specifically to the a5p1 integrin receptor and blocks its binding to fibronectin (Akiyama et al., 1989). As demonstrated here, mAb 16 will bind to human palmar fibroblasts and inhibit fibronectin-mediated spreading; however, mAbl6 does not inhibit collagen gel contraction. These results demonstrate that fibronectin-a5pl integrin interaction is not necessary for collagen gel contraction. Recently, it has been proposed that the “promiscuous” integrin receptor a3p1 also binds to the RGD cell-binding region of fibronectin (Elices et al., 1991). The binding of this receptor to fibronectin can be blocked by RGD peptides (Elices et al., 1991). The lack FIBRONECTIN IN COLLAGEN GEL CONTRACTION of inhibition of collagen gel contraction with GRGDSP would suggest that the interaction of a3p1 integrin with fibronectin is not necessary for this process. It should be emphasized that this study does not rule out the possibility that fibronectin may function as a ligand linking the cell surface with collagen fibrils under certain circumstances. In fact, other studies have suggested that fibronectin can play a role in collagen gel contraction (Guillery et al., 1986; Asaga et al., 1991). However, in those studies, the contraction occurred over a greater period of time compared to our study and that of Gullberg et al. (1990). We purposely set conditions which would allow for collagen gel contraction to occur rapidly for two reasons: (1) the less time in culture, the less cellular fibronectin synthesized, and (2) the forces required t o reorganize collagen fibrils should be the least in a lattice that undergoes the most rapid contraction (Nishiyama et al., 1988). In lattices undergoing slower contraction, fibronectin may augment the binding to collagen to increase the force exerted by cells on collagen. Studies are currently underway to test this possibility. The p1 integrins that have been implicated in cell attachment to type I collagen include the alp1,a2p1, and a3p1 integrin matrix receptors (Santoro, 1986; Wayner and Carter, 1987; Wayner et al., 1988; Takada et al., 1988; Gullberg et al., 1989; Clyman et al., 1990). Consistent with our study, fibroblast and platelet binding to collagen is RGD-independent (Gullberg et al., 1989; Santoro 1986). Multiple integrin receptors on the same cell may participate in attachment to collagen. Antibodies against either the azpl or the a3p1integrin matrix receptor can inhibit fibroblast adhesion to collagen (Wayner and Carter, 1987). The integrin matrix receptors alp1and a2p1 both function on rat aortic smooth muscle cells in attachment to type I collagen (Clyman et al., 1990). Whether the direct binding of fibroblasts to collagen by one, two, or all three of these collagen-binding integrins is sufficient for collagen gel contracting is presently unknown. ACKNOWLEDGMENTS The authors thank Mr. Melville Vaughan for his technical assistance with this project. We also thank Dr. Robert Buchanan of the Department of Plastic Surgery and Dr. Gaza Rayan of the Department of Orthopedic Surgery, University of Oklahoma-Health Sciences Center, for their contributions of tissue, and Dr. Martin Hemler for generously providing the anti-p, antiserum used in this study. This research was supported by grants from the Presbyterian Health Foundation (PHF 101 and PHF 85) to Dr. James Tomasek. NOTE ADDED IN PROOF After the submission of this manuscript, two papers have appeared examining the role of the integrin receptor azpl in collagen lattice contraction: Schiro et al. (1991) and Klein et al. (1991). LITERATURE CITED Akiyama, S.K., E. Hasegawa, T. Hasegawa, and K.M. Yamada 1985 The interaction of fibronectin fragments with fibroblastic cells. J. Biol. Chem., 260:13256-13260. Akiyama, S.K., and K.M. Yamada 1985 Synthetic peptides competitively inhibit both direct binding to fibroblasts and functional 159 biological assays for the purified cell-binding domain of fibronectin. J . Biol. Chem., 260:10402-10406. Akiyama, S.K., S.S. Yamada, W.-T. Chen, and K.M. Yamada 1989 Analysis of fibronectin receptor function with monoclonal antibodies: Roles in cell adhesion, migration, matrix assembly, and cytoskeletal organization. J . Cell Biol., 109:863-875. Akiyama, S.K., S.S. Yamada, and K.M. Yamada 1986 Characterization of a 140-kD avian cell surface antigen as a fibronectinbinding molecule. J. Cell Biol., 102:442-448. Asaga, H., S. Kikuchi, and K. Yoshizato 1991 Collagen gel contraction by fibroblasts requires cellular fibronectin but not plasma fibronectin. Exp. Cell Res., 193t167-174. Albelda, S.M., and C.A. Buck 1991 Integrins and other cell adhesion molecules. FASEB J., 4:2868-2880. Bell, E., B. Ivarsson, and C. Merril 1979 Production of a tissue-like structure by contraction of collagen lattices by human fibroblasts of different proliferative potential in uitro. Proc. Natl. Acad. Sci. U.S.A., 76:1274-1278. Burridge, K., K. Fath, T. Kelly, G. Nuckolls, and C. Turner 1988 Focal adhesions: Transmembrane junctions between the extracellular matrix and the cytoskeleton. Annu. Rev. Cell Biol., 4:487-525. Buttle, D.J., and H.P. Ehrlich 1983 Comparative studies of collagen lattice contraction utilizing a normal and transformed cell line, J . Cell Physiol., 116:159-166. Clyman, R.I., D.C. Turner, and R.H. Kramer 1990 An a,/P,-like integrin receptor on rat aortic smooth muscle cells mediates adhesion to laminin and collagen types I and IV. Arteriosclerosis, 10:402-409. Elices, M.J., L.A. Urry, and M.E. Hemler 1991 Receptor functions for the integrins VLA-3: Fibronectin, collagen, and laminin binding are differentially influenced by Arg-Gly-Asp peptide and by divalent cations. J. Cell Biol., 112:169-181. Elsdale, T.R., and J.B.L. Bard 1972 Collagen substrata for studies of cell behavior. J . Cell Biol., 41:298-311. Engvall, E., and E. Ruoslahti 1977 Binding of soluble form of fibroblast surface protein, fibronectin, to collagen, Int. J . Cancer, 20: 1-5. Grinnell, F., D.G. Hays, and D. Minter 1977 Cell adhesion and spreading factor: Partial purification and properties. Exp. Cell Res., 110:175-190. Guidry, C., and F. Grinnell 1985 Studies on the mechanism of hydrated collagen gel reorganization by human skin fibroblasts. J. Cell Sci., 79:67-81. Guillery, P., F . 4 . Maquart, and J.-P. Bore1 1986 Fibronectin dependence of the contraction of collagen lattices by human skin fibroblasts. Exp. Cell Res., 167r29-37. Gullberg, D., A. Ingstrom, A,-C. Thuresson, L. Olsson, L. Terracio, T.K. Borg, and K. Rubin 1990 PI integrin-mediated collagen gel contraction is stimulated by PDGF. Exp. Cell Res., 186t264-272. Gullberg, D., L. Terracio, T.K. Borg, and K. Rubin 1989 Identification of integrin-like matrix receptors with affinity for interstitial collagens. J . Biol. Chem., 264:12686-12694. Hayashi, M., and K.M. Yamada 1983 The domain structure of the carboxyl-terminal half of human plasma fibronectin. J . Biol. Chem., 258:3332-3340. Hemler, M.E., C. Huang, and L. Schwartz 1987 The VLA protein family. J . Biol. Chem., 262:3300-3309. Hynes, R.O. 1987 Integrins: A family of cell surface receptors. Cell, 48549-544. Klein, C.E., D. Dressel, T. Steinmayer, C. Mauch, B. Eckes, T. Kreig, R.B. Bankert, and L. Weber 1991 Integrin a,#,is upregulated in fibroblasts and highly aggressive melanoma cells in three-dimensional collagen lattices and mediates the reorganization of collagen I fibrils. J . Cell Biol., 115:1427-1436. Kleinman, H., R.J. Klebe, and G.R. Martin 1981 Role of collagenous matrices in the adhesion and growth of cells. J. Cell Biol., 88: 473-485. McDonald, J.A., B.J. Quade, T.J. Broekelman, R. LaChance, K. Forsman, E. Hasegawa, and S. Akiyama 1987 Fibronectins cell-adhesive domain and a n amino terminal matrix assembly domain participate in its assembly into fibroblast pericellular matrix. J. Biol. Chem., 272:2957-2967. Nishiyama, T., N. Tominaga, K. Nakajima, and T. Hayashi 1988 Quantitative evaluation of the factors affecting the process of fibroblast-mediated collagen gel contraction by separating the process into three phases. Collagen Rel. Res., 8:259-273. Pytela, R., M.D. Pierschbacher, and E. Ruoslahti 1985a Identification and isolation of a 140 kd cell surface glycoprotein with properties expected of a fibronectin receptor. Cell, 40:191-198. Pytela, R., M.D. Pierschbacher, and E. Rouslahti 198513 A 1251115kDa cell surface receptor specific for vitronectin interacts with 160 J.J. TOMASEK AND S.K. AKIYAMA the arginine-glycine-aspartic acid adhesion sequence derived from fibronectin. Proc. Natl. Acad. Sci. U.S.A., 825766-5770. Santoro, S.A. 1986 Identification of a 160,000 dalton platelet membrane protein that mediates the initial divalent cation-dependent adhesion of platelets to collagen. Cell, 46~913-920. Schiro, J.A., B.M.C. Chan, W.T. Roswit, P.D. Kassner, A.P. Pentland, M.E. Hemler, A.Z. Eisen, and T.S. Kupper 1991 Integrin a'& (VLA-2) mediates reorganization and contraction of collagen matrices by human cells. Cell, 67:403-410. Stopak, D., and A.K. Harris 1982 Connective tissue morphogenesis by fibroblast traction. Dev. Biol., 90:383-398. Takada, Y., J.L. Strominger, and M.E. Hemler 1987 The very late antigen family is part of a superfamily of molecules involved in adhesion and embryogenesis. Proc. Natl. Acad. Sci. U.S.A., 84: 3239-3243. Takada, Y., E.A. Wayner, W.G. Carter, and M.E. Carter 1988 Extracellular matrix receptors, ECMRII and ECMRI, for collagen and fibronectin correspond to VLA-2 and VLA-3 and in the VLA family of heterodimers. J . Cell Biochem., 37:385-393. Tomasek, J.J., and E.D. Hay 1984 Analysis of the role of microfilaments and microtubules in acquisition of bipolarity and elonga- tion of fibroblasts in hydrated collagen gels. J . Cell Biol., 99: 536-549. Tomasek, J.J., E.D. Hay, and K. Fujiwara 1982 Collagen modulates cell shape and cytoskeleton of embryonic corneal and fibroma fibroblasts: Distribution of actin, alpha-actinin, and myosin. Dev. Biol., 92:107-122. Wayner, E.A., and W.G. Carter 1987 Identification of multiple cell adhesion receptors for collagen and fibronectin in human fibrosarcoma cells possessing unique (L and common p subunits. J . Cell Biol., 105:1873-1884. Wayner, E.A., W.G. Carter, R.S. Piotrowicz, and T.J. Kunicki 1988 T h e function of multiple extracellular matrix receptors in mediating cell adhesion to extracellular matrix: Preparation of monoclonal antibodies to the fibronectin receptor that specifically inhibit cell adhesion to fibronectin and react with platelet glycoproteins Ic-IIa. J. Cell Biol., 107:1881-1891. Yamada, K.M., and D.W. Kennedy 1984 Dualistic nature of adhesive protein function: Fibronectin and its biologically active peptide fragments can autoinhibit fibronectin function. J . Cell Biol., 99: 29-36.