β1 Integrin Regulates Fibroblast Viability during Collagen Matrix Contraction through a Phosphatidylinositol 3-Kinase/Akt/Protein Kinase B Signaling Pathway

Integrins regulate cell viability through their interaction with the extracellular matrix. Integrins can sense mechanical forces arising from the matrix and convert these stimuli to chemical signals capable of modulating intracellular signal transduction. The phosphatidylinositol 3-kinase (PI3K)/Akt signaling pathway is a major regulator of cell survival. It is not known, however, whether integrins, acting as mechanoreceptors, regulate cell survival via the PI3K/Akt pathway. Here, we show that in response to a matrix-derived mechanical stimulus, β1integrin regulated cell viability by regulating Akt activity in a PI3K-dependent fashion. To accomplish this, we employed fibroblasts cultured in collagen gels. During contraction of collagen matrices, fibroblasts underwent apoptosis. We demonstrate that ligation of β1 integrin with anti-β1 integrin antibodies protected fibroblasts from apoptosis. The nature of the survival signal activated by β1 integrin engagement with antibody was mediated by PI3K acting through Akt/protein kinase B. We show that Akt phosphorylation decreased during collagen contraction and that this decrease correlated precisely with the onset of fibroblast apoptosis. Fibroblasts transfected with constitutively active PI3K displayed increased Akt phosphorylation and were protected from anoikis and collagen gel contraction-induced apoptosis. Our data identify a novel role for β1 integrin in regulating fibroblast viability through a PI3K/Akt/protein kinase B signaling pathway in response to a matrix-derived mechanical stimulus. Integrins regulate cell viability through their interaction with the extracellular matrix. Integrins can sense mechanical forces arising from the matrix and convert these stimuli to chemical signals capable of modulating intracellular signal transduction. The phosphatidylinositol 3-kinase (PI3K)/Akt signaling pathway is a major regulator of cell survival. It is not known, however, whether integrins, acting as mechanoreceptors, regulate cell survival via the PI3K/Akt pathway. Here, we show that in response to a matrix-derived mechanical stimulus, β1integrin regulated cell viability by regulating Akt activity in a PI3K-dependent fashion. To accomplish this, we employed fibroblasts cultured in collagen gels. During contraction of collagen matrices, fibroblasts underwent apoptosis. We demonstrate that ligation of β1 integrin with anti-β1 integrin antibodies protected fibroblasts from apoptosis. The nature of the survival signal activated by β1 integrin engagement with antibody was mediated by PI3K acting through Akt/protein kinase B. We show that Akt phosphorylation decreased during collagen contraction and that this decrease correlated precisely with the onset of fibroblast apoptosis. Fibroblasts transfected with constitutively active PI3K displayed increased Akt phosphorylation and were protected from anoikis and collagen gel contraction-induced apoptosis. Our data identify a novel role for β1 integrin in regulating fibroblast viability through a PI3K/Akt/protein kinase B signaling pathway in response to a matrix-derived mechanical stimulus. extracellular matrix phosphatidylinositol 3-kinase protein kinase B monoclonal antibody terminal deoxynucleotidyltransferase-mediated biotinylated UTP nick end labeling constitutively active p110 subunit of PI3K internal ribosome entry site green fluorescent protein fluorescence-activated cell sorter Integrins are a family of αβ-heterodimeric cell-surface receptors that mediate attachment to the extracellular matrix (ECM)1 (1Ruoslahti E. Tumor Biol. 1996; 17: 117-124Crossref PubMed Scopus (124) Google Scholar). Integrins associate with signaling molecules in the focal adhesion complex, which serves as a signaling device and provides a direct link to the cytoskeleton (1Ruoslahti E. Tumor Biol. 1996; 17: 117-124Crossref PubMed Scopus (124) Google Scholar, 2Giancotti F.G. Ruoslahti E. Science. 1999; 285: 1028-1032Crossref PubMed Scopus (3821) Google Scholar). The continuous link between the ECM, integrins, signaling molecules, and the cytoskeleton has led to the discovery that integrins are capable of serving as mechanoreceptors (3Davies P.F. Physiol. Rev. 1995; 75: 519-560Crossref PubMed Scopus (2343) Google Scholar, 4Wilson E. Sudhir K. Ives H.E. J. Clin. Invest. 1995; 96: 2364-2372Crossref PubMed Scopus (268) Google Scholar, 5Ishida T. Peterson T.E. Kovach N.L. Berk B.C. Circ. Res. 1996; 79: 310-316Crossref PubMed Scopus (211) Google Scholar, 6MacKenna D.A. Dolfi F. Vuori K. Ruoslahti E. J. Clin. Invest. 1998; 101: 301-310Crossref PubMed Scopus (273) Google Scholar). As mechanoreceptors, integrins detect matrix-derived mechanical stimuli and convert these to chemical signals capable of modulating signal transduction. Integrin-ECM interaction activates signal transduction pathways that regulate a variety of cellular functions, including cell viability (7Meredith J.E., Jr. Fazeli B. Schwartz M.A. Mol. Biol. Cell. 1993; 4: 953-961Crossref PubMed Scopus (1398) Google Scholar, 8Frisch S.M. Francis H. J. Cell Biol. 1994; 124: 619-626Crossref PubMed Scopus (2775) Google Scholar, 9Brooks P.C. Montgomery A.M.P. Rosenfeld M. Reisfeld R.A., Hu, T. Klier G. Cheresh D.A. Cell. 1994; 79: 1157-1164Abstract Full Text PDF PubMed Scopus (2182) Google Scholar, 10Zhang Z. Vuori K. Reed J. Ruoslahti E. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 6161-6165Crossref PubMed Scopus (574) Google Scholar, 11Frisch S.M. Vuori K. Kelaita D. Sicks S. J. Cell Biol. 1996; 135: 1377-1382Crossref PubMed Scopus (227) Google Scholar, 12Stromblad S. Becker J.C. Yebra M. Brooks P.C. Cheresh D.A. J. Clin. Invest. 1996; 98: 426-433Crossref PubMed Scopus (349) Google Scholar). Not all integrins appear to be capable of regulating cell viability. However, β1 integrin, which mediates attachment to type I collagen, can regulate cell survival (7Meredith J.E., Jr. Fazeli B. Schwartz M.A. Mol. Biol. Cell. 1993; 4: 953-961Crossref PubMed Scopus (1398) Google Scholar). The biological effects of integrin-ECM binding are mediated through the interaction of the cytoplasmic domain of integrins with associated lipid and protein kinases. Specific lipid and protein kinase molecules known thus far to promote cell survival include pp125FAK(13Crouch D.H. Fincham V.J. Frame M.C. Oncogene. 1996; 12: 2689-2696PubMed Google Scholar, 14Frisch S.M. Vuori K. Ruoslahti E. Chan-Hui P.Y. J. Cell Biol. 1996; 134: 793-799Crossref PubMed Scopus (998) Google Scholar, 15Hungerford J.E. Compton M.T. Matter M.L. Hoffstrom B.G. Otey C.A. J. Cell Biol. 1996; 135: 1383-1390Crossref PubMed Scopus (333) Google Scholar, 16Levkau B. Herren B. Koyama H. Ross R. Raines E. J. Exp. Med. 1998; 187: 579-586Crossref PubMed Scopus (227) Google Scholar, 17Chen H.C. Guan J.L. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 10148-10152Crossref PubMed Scopus (477) Google Scholar, 18Ilic D. Almeida E.A.C. Schlaepfer D.D. Dazin P. Aizawa S. Damsky C.H. J. Cell Biol. 1998; 143: 547-560Crossref PubMed Scopus (436) Google Scholar), integrin-linked kinase (19Hannigan G.E. Leung-Hagesteijn C. Fitz-Gibbon L. Coppolini M. Radeava G. Filmus J. Bell J.C. Dedhar S. Nature. 1996; 379: 91-96Crossref PubMed Scopus (967) Google Scholar, 20Wu C. J. Cell Sci. 1999; 112: 4485-4489Crossref PubMed Google Scholar, 21Delcommenne M. Tan C. Gray V. Rue L. Woodgett J. Dedhar S. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 11211-11216Crossref PubMed Scopus (947) Google Scholar), and phosphatidylinositol 3-kinase (PI3K) and Akt/PKB (22Khwaja A. Rodriguez-Viciana P. Wennstrom S. Warne P.H. Downward J. EMBO J. 1997; 16: 2783-2793Crossref PubMed Scopus (939) Google Scholar, 23Lee J.W. Juliano R.J. Mol. Biol. Cell. 2000; 11: 1973-1987Crossref PubMed Scopus (143) Google Scholar, 24Kauffmann-Zeh A. Rodriguez-Viciana P. Ulrich E. Gilbert C. Coffer P. Downward J. Evan G. Nature. 1997; 385: 544-548Crossref PubMed Scopus (1074) Google Scholar, 25Dudek H. Datta S.R. Franke T.F. Birbaum M.J. Yao R. Cooper G.M. Segal R.A. Kaplan D.R. Greenberg M.E. Science. 1997; 275: 661-668Crossref PubMed Scopus (2218) Google Scholar). Evidence has emerged demonstrating the important role of an integrin/PI3K/Akt/PKB signaling pathway in regulating epithelial cell survival (22Khwaja A. Rodriguez-Viciana P. Wennstrom S. Warne P.H. Downward J. EMBO J. 1997; 16: 2783-2793Crossref PubMed Scopus (939) Google Scholar, 23Lee J.W. Juliano R.J. Mol. Biol. Cell. 2000; 11: 1973-1987Crossref PubMed Scopus (143) Google Scholar). Attachment of these cells to the underlying basement membrane via integrins activates the PI3K/Akt/PKB survival signal. However, upon detachment, epithelial cells undergo anoikis. Unlike epithelial cells, which are polarized and attach to the underlying basement membrane at their basal surface, mesenchymal cells are frequently surrounded by the ECM. Because mesenchymal cells are surrounded by the matrix, it is unlikely that unnecessary mesenchymal cells are eliminated by detachment and anoikis. However, recent studies indicate that matrix-derived mechanical signals can regulate cell survival (26Dimmeler S. Assmus B. Hermann C. Haendeler J. Zeiher A.M. Circ. Res. 1998; 83: 334-341Crossref PubMed Scopus (369) Google Scholar, 27Chen K.D., Li, Y.S. Kim M., Li, S. Yuan S. Chien S. Shyy J.Y. J. Biol. Chem. 1999; 274: 18393-18400Abstract Full Text Full Text PDF PubMed Scopus (494) Google Scholar). Currently, it is not known whether integrins, acting as mechanoreceptors, can regulate cell survival via the PI3K pathway in response to a matrix-derived mechanical stimulus. To begin to examine this issue, we employed three-dimensional type I collagen gels. We used three-dimensional collagen gels for our studies because it has been found that three-dimensional matrix adhesions differ in structure, localization, and function from two-dimensional adhesions; and therefore, three-dimensional cell-matrix interactions may be more relevant biologically (28Cukierman E. Pankov R. Stevens D.R. Yamada K.M. Science. 2001; 294: 1708-1712Crossref PubMed Scopus (2449) Google Scholar). This is especially true in the context of examination of mesenchymal cells such as fibroblasts, which typically exist encompassed by the matrix. Recently, it has been discovered that during collagen gel contraction, fibroblasts undergo apoptosis in response to changes in mechanical tension in the collagen matrix (29Grinnell F. Curr. Top. Pathol. 1999; 93: 61-73Crossref PubMed Scopus (23) Google Scholar, 30Grinnell F. Trends Cell Biol. 2000; 10: 362-365Abstract Full Text Full Text PDF PubMed Scopus (338) Google Scholar, 31Fluck J. Querfeld C. Cremer A. Niland S. Krieg T. Sollberg S. J. Invest. Dermatol. 1998; 110: 153-157Abstract Full Text Full Text PDF PubMed Scopus (121) Google Scholar, 32Grinnell F. Zhu M. Carlson M.A. Abrams J.M. Exp. Cell Res. 1999; 248: 608-619Crossref PubMed Scopus (239) Google Scholar). Furthermore, it is known that α2β1 integrin mediates collagen gel contraction and that β1 integrin can function as a mechanoreceptor and can also regulate cell viability. Therefore, we examined whether β1 integrin in response to a mechanical stimulus generated by collagen contraction can regulate fibroblast apoptosis. Using anti-β1 integrin antibodies, we show that ligation of β1 integrin protected fibroblasts from undergoing apoptosis in response to collagen gel contraction. The nature of the survival signal activated by β1 integrin engagement with anti-β1 integrin antibody was mediated by PI3K acting through Akt/PKB. We demonstrate that Akt activity precipitously declined during collagen matrix contraction and that this decrease correlated precisely with the onset of collagen matrix contraction-induced fibroblast apoptosis. Up-regulation of PI3K activity using fibroblasts transfected with the catalytically active p110 subunit of PI3K protected fibroblasts from undergoing anoikis in suspension cultures and apoptosis in response to collagen gel contraction. These studies identify a novel role for β1integrin in regulating fibroblast viability via modulation of a PI3K/Akt signaling pathway in response to contraction of type I collagen matrices. Human lung fibroblasts (CCL-210, American Type Culture Collection, Manassas, VA) were cultured in Dulbecco's modified Eagle's medium (Sigma) containing 10% heat-inactivated fetal calf serum and used between passages 5 and 15. Collagen gels were prepared as described by Fluck et al. (31Fluck J. Querfeld C. Cremer A. Niland S. Krieg T. Sollberg S. J. Invest. Dermatol. 1998; 110: 153-157Abstract Full Text Full Text PDF PubMed Scopus (121) Google Scholar) with minor modifications. Human lung fibroblasts (1.2 × 105) were resuspended in 2.0 ml of 1.5× Dulbecco's modified Eagle's medium containing 15% fetal calf serum. Vitrogen (3 mg/ml; Collagen Esthetics, Palo Alto, CA) was added to the cell suspension to achieve a final concentration of 0.5 mg/ml. The cell/collagen solution was incubated in a water bath for 10 min at 37 °C; poured into 3.5-cm uncoated plastic dishes; and placed into a cell culture incubator at 10% CO2 and 37 °C, where the gels polymerized in ∼30–60 min. Mouse monoclonal antibody P5D2 (raised against the human β1 integrin subunit) was provided by Dr. Leo Furcht (University of Minnesota). Monoclonal antibody M-13 (directed against the β1 integrin chain) was a gift from Dr. Kenneth Yamada (National Institutes of Health). Monoclonal antibodies (mAbs) 1981 and 2253 (raised against the β1 integrin chain), mAb 1950 (directed against the α2 integrin chain), mAb 1998 (raised against α2β1integrin), mAb 1973 (directed against the α1 chain of integrin), and antibody LM609 (directed against αvβ3 integrin) were purchased from Chemicon International, Inc. (Temecula, CA). Mouse anti-human CD49b blocking antibody MCA743, which recognizes the α2 integrin subunit, was purchased from Serotec Inc. (Raleigh, NC). Anti-polyclonal Akt antibody was obtained from Santa Cruz Biotechnologies (Santa Cruz, CA), and anti-phosphorylated Akt polyclonal antibody was purchased fromCell Signaling (Beverly, MA). TUNEL assay was performed according to the methods of Gavrieli et al. (33Gavrieli Y. Sherman Y. Ben-Sasson S.A. J. Cell Biol. 1992; 119: 493-501Crossref PubMed Scopus (9156) Google Scholar) with minor modifications. After deparaffinization, nuclei containing degraded DNA within apoptotic cells in paraffin gel sections were identified using an in situ cell death terminal deoxynucleotidyltransferase dideoxy-UTP-fluorescein isothiocyanate detection kit (Roche Molecular Biochemicals) according to the manufacturer's instructions. Alternatively, fibroblasts recovered from anoikis assays and type I collagen gels were analyzed for apoptosis using an in situcell death detection kit (fluorescence TUNEL assay; Roche Molecular Biochemicals) according to the manufacturer's instructions. Briefly, recovered cells were fixed with 2% paraformaldehyde in phosphate-buffered saline (pH 7.4) for 60 min at room temperature and permeabilized with 0.1% Triton X-100 for 2 min at 4 °C. The cells were resuspended in TUNEL reaction mixture, incubated for 60 min at 37 °C, washed, and analyzed by fluorescence microscopy. Fibroblasts embedded in collagen gels were recovered by gel solubilization for ∼30 min at 37 °C using 0.5 ml of bacterial collagenase (2 mg/ml; Sigma). The reaction was stopped with Dulbecco's modified Eagle's medium and 10% fetal calf serum; the cell solution was centrifuged at 600 × g for 10 min; the resultant pellet of cells was resuspended in trypan blue in phosphate-buffered saline; and cells were counted. Fibroblasts were incorporated into contractile and non-contractile collagen gels for various periods of time. Gels were then transferred to microcentrifuge tubes, and lysis buffer containing protease inhibitors was applied directly onto the gels for 30 min on ice. The samples were centrifuged at 20,000 ×g for 15 min at 4 °C, and lysates were precleared for 1 h at room temperature with protein G-beads and immunoprecipitated for 16 h at 4 °C with anti-Akt polyclonal antibody. Samples were processed for Western analysis. Western analysis of human lung fibroblasts was performed as previously described (34Henke C.A. Roongta U. Mickelson D.J. Knutson J.R. McCarthy J.B. J. Clin. Invest. 1996; 97: 2541-2552Crossref PubMed Scopus (127) Google Scholar). Equal amounts of protein from the cell lysate were subjected to 8% SDS-PAGE and transferred (45 min, 24 V) to nitrocellulose membrane. Membranes were blocked with 20 mm Tris-HCl (pH 7.6), 137 mm NaCl, and 0.05% Tween 20 containing 6% nonfat dry milk; incubated with the primary antibody; washed; and incubated with horseradish peroxidase-conjugated secondary antibody. The membranes were developed using the ECL method (Amersham Biosciences). The constitutively active p110 subunit of PI3K (CA-p110) was provided by Dr. Julian Downward. CA-p110 was subcloned into the multiple cloning site of the retroviral plasmid MIGR1.str containing IRES-GFP, creating the bicistronic construct CA-p110/IRES-GFP. The packaging cell line HEK293 was cotransfected with this plasmid and the packaging vector pCL2-DNA using the CaCl2 precipitate method, and the resulting viral stock was used to infect human lung fibroblasts (CCL-210). A population of CA-p110/GFP-positive cells was obtained by cell sorting using FACS analysis. Anoikis assays were performed as described by Frisch and Francis (8Frisch S.M. Francis H. J. Cell Biol. 1994; 124: 619-626Crossref PubMed Scopus (2775) Google Scholar). Tissue culture plates were coated twice with Poly-HEME (4 mg/ml in ethanol; Sigma) and rinsed extensively with phosphate-buffered saline. Fibroblasts resuspended in Dulbecco's modified Eagle's medium were plated on the Poly-HEME plates. At the indicated times, the cells in suspension were recovered and analyzed by fluorescence TUNEL assay. All data are expressed as means ± S.E. Experiments were performed three times. For assessment of the percentage of apoptotic cells within collagen gels using the TUNEL assay, microscopic analysis of at least 200 cells/slide was performed. Paired evaluations were made for experimental and control conditions within each experiment, and significance was determined by Student'st test. Significance level was set at p < 0.05. It has previously been shown that primary human skin fibroblasts undergo apoptosis during contraction of collagen gels. We examined whether this also was the case for primary human lung fibroblasts. Cells were seeded into 0.5 mg/ml collagen gels in uncoated Petri dishes (contractile gels) or in tissue culture dishes (non-contractile gels). Fibroblasts incorporated into 0.5 mg/ml collagen gels in uncoated Petri dishes contracted the collagen gels from 8.0 cm2 at time 0 (time 0 = cell/collagen mixture placed into dish) to 0.8 cm2 at 24 h (Fig.1 A). Little additional collagen gel contraction occurred over the next 48 h. The degree of collagen gel contraction was inversely proportional to the concentration of the collagen gel. At 1 mg/ml, collagen gels contracted from 8.0 cm2 at time 0 to 2.3 cm2 at 24 h; and at 2 mg/ml, collagen gels contracted from 8.0 to 5.1 cm2 at 24 h (Fig. 1 A). The decrease in cell number was directly related to the degree of collagen gel contraction (Fig. 1 B). A greater proportion of cells underwent apoptosis in the highly contractile 0.5 mg/ml collagen gels as opposed to the less contractile 2 mg/ml collagen gels. The cell number in the 0.5 mg/ml collagen gels declined from 1.2 × 105 cells seeded into the gels at time 0 to (4.8 ± 0.3) × 104 cells/gel at 24 h (Fig. 1 B). By 48 h, the cell number in the 0.5 mg/ml collagen gels had dropped to (2.8 ± 0.4) × 104 cells/gel; and by 72 h, the cell number was (2.4 ± 0.4) × 104cells/gel. In contrast, at 24 h, the cell number in 2 mg/ml contractile collagen gels declined from 1.2 × 105 to (8.4 ± 0.3) × 104 cells/gel. However, at 48 h, the cell number had increased slightly to (8.8 ± 0.4) × 104 cells/gel. The numbers of fibroblasts incorporated into non-contractile collagen gels composed of varying concentrations of collagen (0.5, 1, and 2 mg/ml) did not change much. For example, the cell number in non-contractile gels (0.5 mg/ml collagen) increased slightly from 1.2 × 105 cells/gel at time 0 to (1.4 ± 0.4) × 105 cells/gel at 48 h. In accord with several prior reports (31Fluck J. Querfeld C. Cremer A. Niland S. Krieg T. Sollberg S. J. Invest. Dermatol. 1998; 110: 153-157Abstract Full Text Full Text PDF PubMed Scopus (121) Google Scholar, 32Grinnell F. Zhu M. Carlson M.A. Abrams J.M. Exp. Cell Res. 1999; 248: 608-619Crossref PubMed Scopus (239) Google Scholar), we found that the decline in human lung fibroblast numbers during collagen gel contraction in our system was due to apoptosis. In situlabeling of DNA fragmentation by TUNEL assay revealed that at 48 h, most cells within the contracted 0.5 mg/ml collagen gels were rounded up and contained nuclei staining bright reddish brown (Fig.2 A). In contrast, fibroblasts within the 2 mg/ml collagen gels were well spread, and the nuclei were negative by TUNEL assay (Fig. 2 B). The changes in cell shape noted by TUNEL assay of fibroblasts in contractile collagen gels were also noted by phase-contrast microscopy. Prior to collagen gel contraction, fibroblasts within the collagen matrix were well spread and assumed a stellate appearance with extension of filopodia into the collagen matrix (Fig.3 A). During contraction of the collagen matrix, many fibroblasts became progressively round in appearance and retracted their filopodia that were attached to the matrix. As they retracted their filopodia, the fibroblasts pulled in collagen fibrils such that a meshwork of collagen fibrils could be seen surrounding the cells (Fig. 3 B). Furthermore, membrane blebbing was apparent in some rounded cells after completion of gel contraction, indicating that these cells were undergoing apoptosis (Fig. 3, C and D). α2β1 integrin mediates contraction of type I collagen gels, and the β1 integrin subunit can function as a mechanoreceptor and regulates cell viability. It has been shown that disruption of cell adhesion to the underlying matrix by interference with integrin function causes anchorage-dependent cells to undergo a form of apoptotic death known as anoikis. However, whether β1 integrin can also regulate cell viability in response to a mechanical stimulus generated by contraction of a collagen matrix is not known. We examined whether β1 integrin regulates fibroblast viability during collagen gel contraction. It has previously been shown that anti-β1 integrin monoclonal antibody M-13 inhibits collagen gel contraction, whereas anti-β1 integrin monoclonal antibody P5D2 has little effect on collagen gel contraction (35Seltzer J.L. Lee A.Y. Akers K.T. Sudbeck B. Southon E.A. Wayner E.A. Eisen A.Z. Exp. Cell Res. 1994; 213: 365-374Crossref PubMed Scopus (122) Google Scholar). Both antibodies have been used to study β1integrin-mediated signal transduction in collagen gels, and both recognize the same ligand-binding domain on the β1integrin chain (amino acids 207–218) (35Seltzer J.L. Lee A.Y. Akers K.T. Sudbeck B. Southon E.A. Wayner E.A. Eisen A.Z. Exp. Cell Res. 1994; 213: 365-374Crossref PubMed Scopus (122) Google Scholar, 36Takada Y. Puzon W. J. Biol. Chem. 1993; 268: 17597-17601Abstract Full Text PDF PubMed Google Scholar). It is important to note that although these antibodies have blocking properties in that they can inhibit cell adhesion, this function does not preclude them from being capable of modulating cell signaling. The binding of these antibodies to fibroblasts has also been shown to regulate specific intracellular signaling pathways (35Seltzer J.L. Lee A.Y. Akers K.T. Sudbeck B. Southon E.A. Wayner E.A. Eisen A.Z. Exp. Cell Res. 1994; 213: 365-374Crossref PubMed Scopus (122) Google Scholar). These reagents put us in a position to separate the effect of β1 integrin signaling from gel contraction per se. Human lung fibroblasts were preincubated with various concentrations of each of the monoclonal antibodies and incorporated into collagen gels (0.5 mg/ml, 1.2 × 105 cells/gel). Consistent with previous reports (35Seltzer J.L. Lee A.Y. Akers K.T. Sudbeck B. Southon E.A. Wayner E.A. Eisen A.Z. Exp. Cell Res. 1994; 213: 365-374Crossref PubMed Scopus (122) Google Scholar), we confirmed that the M-13 antibody to the β1 integrin chain inhibited collagen gel contraction in a dose-dependent fashion (Fig.4 A), whereas the P5D2 antibody to β1 integrin had little effect on gel contraction (Fig.4 B). Interestingly, both the M-13 and P5D2 monoclonal antibodies protected fibroblasts from apoptosis. We expected that the M-13 monoclonal antibody might confer protection from apoptosis by virtue of its ability to inhibit lattice contraction. The M-13 antibody protected fibroblasts from apoptosis in a dose-dependent fashion (Fig. 4 C). Cell counts at 24 h were (3.7 ± 0.3) × 104 cells/gel (0 μg/ml antibody), (4.4 ± 0.4) × 104 cells/gel (1 μg/ml antibody), (5.5 ± 0.4) × 104 cells/gel (10 μg/ml antibody), and (6.3 ± 0.4) × 104cells/gel (50 μg/ml antibody). Even though the P5D2 monoclonal antibody had little effect on collagen gel contraction, like the M-13 antibody, it also inhibited induction of fibroblast apoptosis in a dose-dependent fashion (Fig. 4 D). Fibroblast cell numbers at 24 h were (4.1 ± 0.3) × 104 cells/gel (0 μg/ml antibody), (4.2 ± 0.4) × 104 cells/gel (0.1 μg/ml antibody), (7.2 ± 0.4) × 104 cells/gel (1 μg/ml antibody), and (7.9 ± 0.4) × 104 cells/gel (5 μg/ml antibody). These data indicate that direct engagement of β1 integrin with anti-β1 integrin antibody is capable of regulating fibroblast viability during collagen gel contraction. We also examined whether blocking antibodies to the α2integrin subunit would affect collagen gel contraction or the induction of apoptosis. The adhesion-blocking monoclonal antibody MCA743 (directed against the α2 integrin chain) as well as the blocking mAb 1950 (also directed against the α2 integrin subunit) did not inhibit collagen gel contraction and did not protect against apoptosis (data not shown). Furthermore, blocking monoclonal antibodies against the α1 integrin subunit, the α5 integrin subunit, and αvβ3integrin did not affect lattice contraction or cell viability. In addition, we examined the effect of RGD peptides on collagen gel contraction and induction of apoptosis. Preincubation of fibroblasts with RGD peptides did not block collagen gel contraction and did not inhibit apoptosis (data not shown). Our data indicate that the β1 integrin subunit (but not the α2integrin subunit) regulates fibroblast viability during collagen gel contraction. Several integrin-mediated viability signal transduction pathways have begun to be characterized. In epithelial and endothelial cells, the PI3K/Akt/PKB survival pathway can be activated by growth factor and integrin binding (17Chen H.C. Guan J.L. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 10148-10152Crossref PubMed Scopus (477) Google Scholar, 18Ilic D. Almeida E.A.C. Schlaepfer D.D. Dazin P. Aizawa S. Damsky C.H. J. Cell Biol. 1998; 143: 547-560Crossref PubMed Scopus (436) Google Scholar, 22Khwaja A. Rodriguez-Viciana P. Wennstrom S. Warne P.H. Downward J. EMBO J. 1997; 16: 2783-2793Crossref PubMed Scopus (939) Google Scholar,23Lee J.W. Juliano R.J. Mol. Biol. Cell. 2000; 11: 1973-1987Crossref PubMed Scopus (143) Google Scholar). We examined whether engagement of β1 integrin on fibroblasts with antibody was capable of activating Akt/PKB. We examined the phosphorylation of Ser473 by Western blotting using a specific antibody. Phosphorylation of Ser473 of Akt indicates activation of the kinase (37Chan T.O. Rittenhouse S.E. Tsichlis P.N. Annu. Rev. Biochem. 1999; 68: 965-1014Crossref PubMed Scopus (876) Google Scholar). To perform these experiments, fibroblasts were plated on tissue culture dishes and serum-starved overnight prior to addition of anti-β1 integrin antibody. It is important to note that during these experiments, the addition of the P5D2 or M-13 antibody to fibroblasts already adhered to tissue culture plastic did not affect their attachment. However, as a control to confirm P5D2 and M-13 antibody function, preincubation of fibroblasts with either antibody inhibited their ability to adhere to tissue culture plastic (data not shown). Interestingly, we found that ligation of β1 integrin with the P5D2 antibody increased the phosphorylation of Ser473 within 30 min and began to decrease by 2 h (Fig.5 A). However, anti-β1 integrin antibody M-13 and the isotype control antibody had no effect on Akt phosphorylation (Fig. 5 B). We also sought to determine whether ligation of β1 integrin by the P5D2 antibody independent of cell adhesion would alter Akt/PKB activation. Serum-starved fibroblasts in suspension culture were treated with anti-β1 integrin antibody. Similar to adherent fibroblasts, Akt was phosphorylated at Ser473 in non-adherent fibroblasts by ligation of β1 integrin in a time-dependent manner (Fig. 5 C). To determine whether activation of Akt/PKB was PI3K-dependent, the fibroblasts were pretreated with wortmannin and exposed to anti-β1 integrin antibody. Wortmannin abrogated the increase in Akt/PKB phosphorylation mediated by ligation of β1 integrin with antibody (Fig. 5 D). These data indicate that the P5D2 antibody, when preincubated with cells prior to adherence, has blocking properties in that it inhibits attachment. However, our data also show that when the P5D2 antibody binds to β1 integrin on cells that are already attached or are in suspension, it functions in an agonist fashion by activating the Akt survival signaling pathway via a PI3K-dependent mechanism. Our data suggest that fibroblast protection conferred by ligation of β1 integrin with the P5D2 antibody during collagen gel contraction may be mediated via a β1integrin/PI3K/Akt/PKB survival signaling pathway. Our data suggest that