Portal de Periódicos da CAPES

An Endostatin-derived Peptide Interacts with Integrins and Regulates Actin Cytoskeleton and Migration of Endothelial Cells

2004; Elsevier BV; Volume: 279; Issue: 19 Linguagem: Inglês

10.1074/jbc.m312921200

1083-351X

Sara A. Wickström, Kari Alitalo, Jorma Keski‐Oja,

Protease and Inhibitor Mechanisms

Endostatin, the C-terminal fragment of collagen XVIII, is a potent inhibitor of angiogenesis and endothelial cell migration. To define its critical cell interaction domains we used endostatin-derived synthetic peptides containing surface-exposed sequences. We observed that, when immobilized, an arginine-rich peptide of 11 amino acids from its N terminus efficiently promoted endothelial cell adhesion through β1 integrin- and heparin-dependent mechanisms. In addition, the peptide induced the formation of membrane ruffles and focal contacts. In the soluble form, the peptide inhibited basic fibroblast growth factor-induced directional migration and tubular morphogenesis of microvascular endothelial cells. Accordingly, the peptide induced the loss of focal adhesions and actin stress fibers in these cells. Substitution of the arginine residues with alanines resulted in the loss of these properties. In the current study we describe a putative integrin-binding sequence with anti-migratory activity within endostatin. Endostatin, the C-terminal fragment of collagen XVIII, is a potent inhibitor of angiogenesis and endothelial cell migration. To define its critical cell interaction domains we used endostatin-derived synthetic peptides containing surface-exposed sequences. We observed that, when immobilized, an arginine-rich peptide of 11 amino acids from its N terminus efficiently promoted endothelial cell adhesion through β1 integrin- and heparin-dependent mechanisms. In addition, the peptide induced the formation of membrane ruffles and focal contacts. In the soluble form, the peptide inhibited basic fibroblast growth factor-induced directional migration and tubular morphogenesis of microvascular endothelial cells. Accordingly, the peptide induced the loss of focal adhesions and actin stress fibers in these cells. Substitution of the arginine residues with alanines resulted in the loss of these properties. In the current study we describe a putative integrin-binding sequence with anti-migratory activity within endostatin. Angiogenesis, the formation of new blood vessels from preexisting ones, is a critical event in numerous physiological and pathological processes such as wound healing, fracture repair, rheumatoid arthritis, and cancer. The growth of solid tumors beyond a few millimeters in diameter requires the generation of tumor microvasculature (1Hanahan D. Folkman J. Cell. 1996; 86: 353-364Abstract Full Text Full Text PDF PubMed Scopus (6091) Google Scholar). The onset of this angiogenic switch is regulated by a shift in the balance of negative and positive regulators of angiogenesis. The positive regulators are most commonly growth factors, such as basic fibroblast growth factor (bFGF), 1The abbreviations used are: bFGF, basic fibroblast growth factor; HDMEC, human dermal microvascular endothelial cell; PBS, phosphate-buffered saline; ES-1 through ES-5, synthetic endostatin peptides (see text); BSA, bovine serum albumin. which stimulate the migration and proliferation of endothelial cells (2Risau W. Nature. 1997; 386: 671-674Crossref PubMed Scopus (4846) Google Scholar). Negative regulators of angiogenesis are more often fragments of the extracellular matrix, which inhibit endothelial cell migration and promote apoptosis (3Hagedorn M. Bikfalvi A. Crit. Rev. Oncol. Hematol. 2000; 34: 89-110Crossref PubMed Scopus (116) Google Scholar, 4Kalluri R. Nat. Rev. Cancer. 2003; 3: 422-433Crossref PubMed Scopus (1336) Google Scholar). One of the most potent negative regulators of angiogenesis is endostatin, a C-terminal proteolytic fragment of type XVIII collagen (5O'Reilly M.S. Boehm T. Shing Y. Fukai N. Vasios G. Lane W.S. Flynn E. Birkhead J.R. Olsen B.R. Folkman J. Cell. 1997; 88: 277-285Abstract Full Text Full Text PDF PubMed Scopus (4253) Google Scholar). Endostatin is able to inhibit tumor growth through the inhibition of angiogenesis in several animal models (5O'Reilly M.S. Boehm T. Shing Y. Fukai N. Vasios G. Lane W.S. Flynn E. Birkhead J.R. Olsen B.R. Folkman J. Cell. 1997; 88: 277-285Abstract Full Text Full Text PDF PubMed Scopus (4253) Google Scholar, 6Blezinger P. Wang J. Gondo M. Quezada A. Mehrens D. French M. Singhal A. Sullivan S. Rolland A. Ralston R. Min W. Nat. Biotechnol. 1999; 17: 343-348Crossref PubMed Scopus (299) Google Scholar, 7Yoon S.S. Eto H. Lin C.M. Nakamura H. Pawlik T.M. Song S.U. Tanabe K.K. Cancer Res. 1999; 59: 6251-6256PubMed Google Scholar, 8Sauter B.V. Martinet O. Zhang W.J. Mandeli J. Woo S.L. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 4802-4807Crossref PubMed Scopus (245) Google Scholar, 9Feldman A.L. Restifo N.P. Alexander H.R. Bartlett D.L. Hwu P. Seth P. Libutti S.K. Cancer Res. 2000; 60: 1503-1506PubMed Google Scholar). The possible cell biological mechanisms underlying the antiangiogenic effects of endostatin include inhibition of endothelial cell migration, induction of cell cycle arrest, and promotion of apoptosis (5O'Reilly M.S. Boehm T. Shing Y. Fukai N. Vasios G. Lane W.S. Flynn E. Birkhead J.R. Olsen B.R. Folkman J. Cell. 1997; 88: 277-285Abstract Full Text Full Text PDF PubMed Scopus (4253) Google Scholar, 10Dhanabal M. Ramchandran R. Waterman M.J.F. Lu H. Knebelmann B. Segal M. Sukhatme V.P. J. Biol. Chem. 1999; 274: 11721-11726Abstract Full Text Full Text PDF PubMed Scopus (588) Google Scholar, 11Dixelius J. Larsson H. Sasaki T. Holmqvist K. Lu L. Engström A. Timpl R. Welsh M. Claesson-Welsh L. Blood. 2000; 95: 3403-3411Crossref PubMed Google Scholar, 12Hanai J. Dhanabal M. Karumanchi S.A. Albanese C. Waterman M. Chan B. Ramchandran R. Pestell R. Sukhatme V.P. J. Biol. Chem. 2002; 277: 16464-16469Abstract Full Text Full Text PDF PubMed Scopus (217) Google Scholar). In addition, endostatin can induce Src-dependent disassembly of the actin cytoskeleton and act as an inhibitor of Wnt signaling (13Wickström S.A. Veikkola T. Rehn M. Pihlajaniemi T. Alitalo K. Keski-Oja J. Cancer Res. 2001; 61: 6511-6516PubMed Google Scholar, 14Wickström S.A. Alitalo K. Keski-Oja J. Cancer Res. 2002; 62: 5580-5589PubMed Google Scholar, 15Dixelius J. Cross M. Matsumoto T. Sasaki T. Timpl R. Claesson-Welsh L. Cancer Res. 2002; 62: 1944-1947PubMed Google Scholar, 16Hanai J. Gloy J. Karumanchi S.A. Kale S. Tang J. Hu G. Chan B. Ramchandran R. Jha V. Sukhatme V.P. Sokol S. J. Cell Biol. 2002; 158: 529-539Crossref PubMed Scopus (142) Google Scholar). Both heparan sulfate proteoglycans and integrins bind endostatin on the endothelial cell surface (17Karumanchi S.A. Jha V. Ramchandran R. Karihaloo A. Tsiokas L. Chan B. Dhanabal M. Hanai J.I. Venkataraman G. Shriver Z. Keiser N. Kalluri R. Zeng H. Mukhopadhyay D. Chen R.L. Lander A.D. Hagihara K. Yamaguchi Y. Sasisekharan R. Cantley L. Sukhatme V.P. Mol. Cell. 2001; 7: 811-822Abstract Full Text Full Text PDF PubMed Scopus (271) Google Scholar, 18Rehn M. Veikkola T. Kukk-Valdre E. Nakamura H. Ilmonen M. Lombardo C. Pihlajaniemi T. Alitalo K. Vuori K. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 1024-1029Crossref PubMed Scopus (421) Google Scholar). Numerous studies have underlined the role of integrin binding in the cell biological effects of endostatin (14Wickström S.A. Alitalo K. Keski-Oja J. Cancer Res. 2002; 62: 5580-5589PubMed Google Scholar, 19Wickström S.A. Alitalo K. Keski-Oja J. J. Biol. Chem. 2003; 278: 37895-37901Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar, 20Sudhakar A. Sugimoto H. Yang C. Lively J. Zeisberg M. Kalluri R. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 4766-4771Crossref PubMed Scopus (436) Google Scholar). Recent findings suggest that interaction of endostatin with both of these putative receptors is needed for the endostatin-induced disassembly of the actin cytoskeleton (19Wickström S.A. Alitalo K. Keski-Oja J. J. Biol. Chem. 2003; 278: 37895-37901Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar). The crystal structure of mouse endostatin revealed a compact globular fold and a basic patch of 11 arginine residues, which have been suggested to act as binding sites for heparin (21Hohenester E. Sasaki T. Olsen B.R. Timpl R. EMBO J. 1998; 17: 1656-1664Crossref PubMed Scopus (199) Google Scholar). However, the roles of the arginine residues have remained unclear so far, because both decreased and unaltered anti-migratory potential has been reported in endostatin mutants, which are unable to bind heparin (22Sasaki T. Larsson H. Kreuger J. Salmivirta M. Claesson-Welsh L. Lindahl U. Hohenester E. Timpl R. EMBO J. 1999; 18: 6240-6248Crossref PubMed Scopus (200) Google Scholar, 23Yamaguchi N. Anand-Apte B. Lee M. Sasaki T. Fukai N. Shapiro R. Que I. Lowik C. Timpl R. Olsen B.R. EMBO J. 1999; 18: 4414-4423Crossref PubMed Scopus (425) Google Scholar). Both mouse and human endostatin lack RGD sequences, and specific integrin-binding sites have not been identified yet. The current work has been carried out to identify specific sequence motifs of human endostatin that would interact with endothelial cell surface integrins or heparan sulfate proteoglycans. We found that an arginine-rich endostatin-derived peptide, when immobilized, efficiently promoted endothelial cell adhesion via β1 integrin and heparin. The peptide was a potent inhibitor of endothelial cell migration and tubular morphogenesis, and induced the disassembly of the actin cytoskeleton when in soluble form. The activity was found to be lost in mutated peptides where arginine residues were substituted with alanines. We describe here a novel RGD-independent sequence motif within endostatin capable of promoting endothelial cell adhesion via integrins. Because this sequence is also an inhibitor of endothelial cell migration and tube formation, it is plausibly an important sequence motif for the anti-angiogenic activity of endostatin. Reagents and Antibodies—Recombinant human endostatin was purchased from Calbiochem (San Diego, CA). bFGF was from R&D Systems Inc. (Minneapolis, MN). Mouse monoclonal anti-vinculin antibodies were from Sigma Chemical Co. (St. Louis, MO). Monoclonal antibodies against integrins α5β1 (cloneJBS5), α5 (clone P1D6), β1 (clone P5D2), αv (clone M9), and polyclonal antibodies against integrin β3 (catalog no. 1932) were from Chemicon Inc. (Temecula, CA). Cell Culture—Human dermal microvascular endothelial cells (HDMECs) were purchased from Promocell (Heidelberg, Germany) and were cultured in endothelial cell growth medium (Promocell), at 37 °C in a humidified 5% CO2 atmosphere. The cells used for the experiments were from passages 3–6. Unless indicated otherwise, the cells were washed twice and incubated in serum-free Medium 199 for at least 12 h prior to treatment with the various proteins or chemicals. All experiments were carried out under serum-free conditions. Immunofluorescence—Cells cultured on glass coverslips were washed with phosphate-buffered saline (PBS, 170 mm NaCl, 10 mm sodium phosphate buffer, pH 7.4) and fixed with 3% paraformaldehyde at 4 °C for 10 min. Nonspecific protein binding sites were then saturated with 5% bovine serum albumin (BSA) in PBS for 30 min. The cells were then washed with PBS and incubated with polyclonal antibodies against fibronectin or monoclonal antibodies against vinculin and paxillin. Rhodamine-conjugated phalloidin (Sigma) was used for the staining of the actin cytoskeleton. Unbound proteins were removed by washing, followed by incubation with Texas Red- or fluorescein isothiocyanate-labeled secondary antibodies (Jackson ImmunoResearch Laboratories, West Grove, PA) for 1 h. The coverslips were then washed and mounted on glass slides using Vectashield (Vector Laboratories, Burlingame, CA). The fluorescent images were obtained using an epifluorescence microscope. Cell Adhesion Assays—Plastic 96-well cell culture plates were coated with endostatin or endostatin-derived peptides at concentrations indicated (50 μl/well) at 4 °C for 16 h. Nonspecific adhesion sites were saturated by incubating the wells with 1% heat-inactivated BSA (70 °C for 1 h) at 22 °C for 1 h. After washing the wells with PBS, 0.5 × 105 cells in 100 μl of prewarmed serum-free medium were seeded into each well. Antibodies against integrins α5, β1, αv, and β3 or heparin (100 IU/ml) were added to the cells at a concentration of 30 μg/ml where indicated. After 90 min of incubation at 37 °C, nonattached cells were removed by washing with PBS. The cells were then fixed and stained in 30% methanol, 10% acetic acid, containing 0.1% Coomassie Blue. Subsequently, the cells were washed extensively with 30% methanol, 10% acetic acid, and once with PBS. The cells were then lysed in 1% SDS, after which the amount of lysed cells were quantified by measuring their absorbance at 620 nm using a Multiscan plate reader (Labsystems, Finland). Each experiment contained three independent wells per coated protein. Values presented represent at least three independent experiments. Cell Migration Assays—Cell migration assays were performed in modified Boyden chambers containing polyethylene membranes (8-μm pore size; BD Falcon, Franklin Lakes, NJ) coated with 50 μg/ml fibronectin at 4 °C for 16 h. After washing the membranes with PBS, nonspecific adhesion sites were saturated with 1% heat-inactivated BSA at 22 °C for 1 h. Cells (1 × 104) suspended in serum-free medium were placed in the upper compartment of the chambers. Medium containing 50 ng/ml bFGF was added to the lower compartment. The chambers were then incubated for 4 h at 37 °C. Subsequently, the membranes were fixed and stained in 30% methanol, 10% acetic acid containing 0.1% Coomassie Blue, and the cells migrated to the lower surface were counted at ×200 magnification. Three random fields from each two wells were counted. Values represent three independent experiments. Endothelial Cell Tube Formation Assay—24-well cell culture plates were coated with 200 μl of growth factor-reduced Matrigel basement membrane matrix (BD Biosciences, Bedford, MA) and allowed to polymerize at +37 °C for 45 min. The cells (4 × 105) were suspended in 600 μl of pre-warmed serum-free Medium 199 containing endostatin or endostatin-derived peptides at concentrations indicated. The cells were seeded on Matrigel-coated plates, treated with 10 ng/ml bFGF, and incubated at 37 °C for 16 h. The cells were then fixed, and tube formation was assessed using an inverted phase-contrast microscope. Tubular branches were counted from three random fields. Values represent three independent experiments. Construction of Synthetic Endostatin-derived Peptides Containing Surface-exposed Basic Amino Acid Residues—Using crystal structure analysis, an extensive basic patch of arginine and lysine residues has been identified on the surface of mouse and human endostatins (21Hohenester E. Sasaki T. Olsen B.R. Timpl R. EMBO J. 1998; 17: 1656-1664Crossref PubMed Scopus (199) Google Scholar). To identify specific sequence motifs that interact with endothelial cell surface integrins, and to investigate the role of the basic amino acids in endostatin function, we constructed five synthetic peptides of 11–13 amino acids. These peptides, termed ES-1 to ES-5, contained 12 basic amino acids out of the total 18 as determined on the basis of the crystal structure. ES-1 consisted of the amino acids 23–34 and two surface exposed basic amino acids, Arg24 and Arg27. ES-2 contained the amino acids 60–70 and three surface exposed arginine residues, Arg62, Arg63, and Arg66. ES-3 contained the amino acids 99–111, and three surface-exposed basic amino acids, Arg99, Lys106, and Arg110. ES-4 contained the amino acids 127–139 and three surface-exposed arginines, Arg128, Arg129, and Arg139. ES-5 covered the C-terminal amino acids 171–183, containing C-terminal Lys183 (Fig. 1, upper panel). To characterize the role of the arginines for endostatin function, three mutant ES-2-peptides were constructed. In ES-2mut1, the first two arginine residues, Arg62 and Arg63, from ES-2 were substituted with alanines. In ES-2mut2 Arg66 was replaced by an alanine. Accordingly, in ES-2mut3 all three arginines, Arg62, Arg63, and Arg66, were replaced by alanines (Fig. 1, lower panel). Immobilized ES-2 Mediates Endothelial Cell Adhesion—Integrins have been implicated to serve as receptors for endostatin, and immobilized endostatin mediates cell adhesion in an α5β1 integrin-dependent manner (18Rehn M. Veikkola T. Kukk-Valdre E. Nakamura H. Ilmonen M. Lombardo C. Pihlajaniemi T. Alitalo K. Vuori K. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 1024-1029Crossref PubMed Scopus (421) Google Scholar, 20Sudhakar A. Sugimoto H. Yang C. Lively J. Zeisberg M. Kalluri R. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 4766-4771Crossref PubMed Scopus (436) Google Scholar). To investigate whether the endostatin-derived peptides would serve as substrates for cell adhesion, the cells were plated under serum-free conditions on culture dishes coated with 5 μm endostatin or 50 μm endostatin-derived peptides. Cells were allowed to attach for 90 min after which cell adhesion was quantified. Cell adhesion was significantly increased when cells were plated on immobilized endostatin compared with the BSA-coated control wells. ES-2 was even more efficient than full-length endostatin in inducing cell attachment. ES-5 and ES-3 induced a less pronounced effect on cell adhesion but were as efficient as full-length endostatin. ES-4 was slightly less efficient, whereas ES-1 had no effect (Fig. 2A). To investigate which of the cell surface molecules might mediate endothelial cell adhesion to ES-2, antibodies inhibiting integrin ligation as well as heparin were applied to the cell suspension before allowing the cells to adhere. Both heparin and antibodies against β1 integrin efficiently inhibited HDMEC adhesion to ES-2. Antibodies against α1, α2, α5, αv, and β3 integrins displayed no dramatic effects on cell adhesion (Fig. 2B). To investigate the roles of the arginine residues in the adhesive effect of ES-2, cell adhesion assays were performed using mutated peptides (Fig. 1). Cells plated on ES-2 mutants ES-2mut1, ES-2mut2, or ES-2mut3 lacking one, two, or three arginines, respectively, displayed reduced levels of adhesion (Fig. 2C). Because peptides ES-3 and ES-5 also promoted cell adhesion, it was of interest to investigate whether α5β1 integrin or heparin would be involved in these interactions. Antibodies inhibiting α5 and β1 integrin ligation as well as heparin were applied to the cell suspension before allowing the cells to adhere. Antibodies against α5, and β3 integrins had no significant effects on cell adhesion, but heparin efficiently inhibited HDMEC adhesion to both ES-3 and ES-5 (Fig. 2D). Immobilized ES-2 Promotes a Limited Degree of Cell Spreading—Integrin-dependent cell adhesion to the extracellular matrix results in cell spreading, actin polymerization, and formation of focal adhesions (24Clark E.A. King W.G. Brugge J.S. Symons M. Hynes R.O. J. Cell Biol. 1998; 142: 573-586Crossref PubMed Scopus (533) Google Scholar). To examine whether immobilized ES-2 promotes cell spreading, we performed immunofluorescence analysis on cells attached on glass coverslips coated with ES-2 under serum-free conditions. Control coverslips were coated with BSA. Cells were allowed to attach and spread for 2 h prior to fixation. Actin staining of cells attached onto ES-2 revealed the formation of membrane ruffles at the cell periphery. No formation of actin stress fibers was observed. The actin-rich membrane protrusions were also positive for vinculin staining. Integrin α5β1 was accordingly observed to localize to the membrane ruffles in the cell periphery. Cells plated on ES-2 displayed limited cell spreading when compared with cells plated on fibronectin, the major substrate for α5β1 integrin. Cells adhering on fibronectin formed actin stress fibers and displayed cytoplasmic focal adhesions visualized by vinculin staining. Cells attached on BSA remained rounded with no evidence of cell spreading (Fig. 3). Soluble ES-2 Inhibits Cell Adhesion on Endostatin—To investigate whether endostatin and the endostatin-derived peptides occupy the same cell surface receptors, soluble ES-2, ES-3, and ES-5 were used to compete with endostatin for endothelial cell adhesion. Cell culture dishes were coated with 5 μm endostatin (100 μg/ml). HDMECs in suspension containing the peptides (6 μm) were then added, and the cells were allowed to attach for 90 min. ES-2 effectively inhibited endothelial cell adhesion to immobilized endostatin. ES-5 had a moderate inhibitory effect on cell adhesion to endostatin, whereas ES-3 and ES-2mut3 displayed negligible effects (Fig. 4A). To further analyze the effect of ES-2 on endothelial cell adhesion to endostatin, HDMECs in suspension containing increasing concentrations of ES-2 were allowed to attach on cell culture dishes coated with 5 μm endostatin (100 μg/ml). Quantification of cell attachment after 90 min indicated that ES-2 inhibited HDMEC adhesion to endostatin in a concentration-dependent manner (Fig. 4B). Soluble Endostatin-derived Peptides Induce Disassembly of Focal Adhesions and Actin Stress Fibers—We have previously observed that human endostatin induces the disassembly of focal adhesions and actin stress fibers in cultured endothelial cells (13Wickström S.A. Veikkola T. Rehn M. Pihlajaniemi T. Alitalo K. Keski-Oja J. Cancer Res. 2001; 61: 6511-6516PubMed Google Scholar). To analyze whether endostatin-derived peptides would induce similar changes in cytoskeletal morphology, we carried out immunofluorescence analysis of peptide-treated HDMECs using antibodies against the focal adhesion protein vinculin and the actin fiber staining agent rhodamine-phalloidin. Cells were treated with 500 nm concentrations of ES-1 through ES-5 peptides, and after 1 h of incubation the cells were fixed and subjected to immunofluorescence analysis. Untreated cells and cells treated with ES-1, ES-3, or ES-5 displayed distinct focal adhesion structures as visualized by vinculin staining, and multiple actin stress fibers. In cells treated with ES-2, the vinculin staining was restricted to the cell periphery with no cytoplasmic plaques visible. In addition, rhodamine-phalloidin staining of the actin cytoskeleton revealed the absence of a stress fiber network. In cells treated with ES-4, a fraction of the cells displayed intact focal adhesions and actin stress fibers, whereas in some cells these structures were absent (Fig. 5A, arrows). As the substitution of the arginine residues with alanines resulted in decreased adhesive activity of ES-2, we sought next to investigate the role of these basic amino acids in the ES-2-induced disruption of cytoskeletal structures. Cells were treated with 50 nm endostatin or 500 nm ES-2, ES-2mut1, ES-2mut2, or ES-2mut3, and immunofluorescence analysis of their cytoskeletons was carried out. We found that untreated control cells and cells treated with ES-2mut3 displayed multiple cytoplasmic focal adhesions (vinculin staining) and a distinct network of actin stress fibers. In contrast, endostatin-treated cells and cells treated with ES-2 completely lacked these structures, and vinculin staining was visible only in the cell periphery together with cortical actin bundles. In cells treated with ES-2mut1 or ES-2mut2 the effect was partial, as evidenced by the presence of cells with intact focal adhesions and clear stress fibers, as well as cells void of these structures (Fig. 5B). Effect of Soluble Endostatin-derived Peptides on Endothelial Cell Migration—Endostatin inhibits the migration of cultured endothelial cells (5O'Reilly M.S. Boehm T. Shing Y. Fukai N. Vasios G. Lane W.S. Flynn E. Birkhead J.R. Olsen B.R. Folkman J. Cell. 1997; 88: 277-285Abstract Full Text Full Text PDF PubMed Scopus (4253) Google Scholar). The strength of cell adhesion to the ECM regulates the choice between adhesion and migration. If the adhesion is too strong, the cell remains stationary, and if the adhesion is too weak, sufficient traction cannot be generated for movement. Cell migration occurs thus at intermediate levels of adhesion (25Schwarzbauer J.E. Curr. Biol. 1997; 7: R292-R294Abstract Full Text Full Text PDF PubMed Google Scholar). Because endostatin and ES-2 induced changes in cell-matrix interactions, we explored next their effects on directional cell motility using a modified Boyden chamber assay. Endostatin or endostatin-derived peptides were added to both upper and lower compartments of the chamber. Cell migration through a fibronectin matrix toward bFGF was inhibited ∼50%, when the endothelial cells were treated with 50 nm recombinant endostatin. A comparable effect was observed in cells treated with 500 nm ES-2. ES-4 had a moderate effect cell migration, whereas ES-1, ES-3, and ES-5 had only negligible effects on endothelial cell migration (Fig. 6A). To investigate the role of the arginine residues of ES-2 for its anti-migratory effect, Boyden chamber migration assays were performed using the mutant peptides. Cells treated with 50 nm endostatin or 500 nm ES-2 displayed reduced ability to migrate toward the bFGF stimulus. ES-mut1 and ES-mut2 possessed moderate anti-migratory potential but were less effective than full-length endostatin or ES-2. When the cells were treated with ES-2mut3, the level of migration was not significantly reduced from that of untreated cells (Fig. 6B). Tubular Morphogenesis of HDMECs in Inhibited by Soluble ES-2—To assess the effect of the ES-2 peptide on the formation of vascular structures, HDMECs were seeded on growth factor-reduced Matrigel basement membrane matrix in the presence of bFGF. After 16 h of incubation the cultures were fixed and assayed for tube formation using an inverted phase-contrast microscope. In untreated control cells, bFGF induced the formation of organized tubular structures with branching sprouts and anastomoses (Fig. 7A). In cells treated with 50 nm endostatin or 500 nm ES-2, tube formation was significantly decreased, and the cells remained as an adherent monolayer with only a few tubular structures. ES-2mut3 had no inhibitory effect on the tubular morphogenesis of HDMECs (Fig. 7, A and B). Current knowledge of the effects of endostatin on endothelial cell cultures suggests that the anti-migratory and anti-proliferative properties of this molecule are the major mechanisms underlying its anti-angiogenic potential. Recent studies provide evidence that the binding of endostatin to α5β1 integrin on the endothelial cell surface in part modulates these effects (14Wickström S.A. Alitalo K. Keski-Oja J. Cancer Res. 2002; 62: 5580-5589PubMed Google Scholar, 19Wickström S.A. Alitalo K. Keski-Oja J. J. Biol. Chem. 2003; 278: 37895-37901Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar, 20Sudhakar A. Sugimoto H. Yang C. Lively J. Zeisberg M. Kalluri R. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 4766-4771Crossref PubMed Scopus (436) Google Scholar). It has also been suggested that the basic patch formed on the surface of endostatin by clusters of arginine residues could play a role in the activity of this molecule (21Hohenester E. Sasaki T. Olsen B.R. Timpl R. EMBO J. 1998; 17: 1656-1664Crossref PubMed Scopus (199) Google Scholar, 26Sasaki T. Fukai N. Mann K. Gohring W. Olsen B.R. Timpl R. EMBO J. 1998; 17: 4249-4256Crossref PubMed Scopus (327) Google Scholar). However, no detailed structure-function analyses of human endostatin have been carried out. In the current study we designed peptides derived from areas of human endostatin containing multiple surface-exposed basic amino acid residues. We analyzed the effects of these peptides on endothelial cell adhesion and characterized the cytoskeletal alterations of cells treated with the peptides. We also analyzed the effects of these peptides on endothelial cell migration and tubular morphogenesis. The results indicated that an arginine-rich peptide was able to promote endothelial cell adhesion and spreading via heparin- and integrin β1-dependent interactions. In addition, the soluble peptide induced the disassembly of the actin cytoskeleton and inhibited cell migration and tubular morphogenesis as described earlier for endostatin (5O'Reilly M.S. Boehm T. Shing Y. Fukai N. Vasios G. Lane W.S. Flynn E. Birkhead J.R. Olsen B.R. Folkman J. Cell. 1997; 88: 277-285Abstract Full Text Full Text PDF PubMed Scopus (4253) Google Scholar, 11Dixelius J. Larsson H. Sasaki T. Holmqvist K. Lu L. Engström A. Timpl R. Welsh M. Claesson-Welsh L. Blood. 2000; 95: 3403-3411Crossref PubMed Google Scholar). The arginine residues of the peptide were found to be critical for its activity. Immobilized endostatin has previously been observed to promote cell adhesion via integrins (18Rehn M. Veikkola T. Kukk-Valdre E. Nakamura H. Ilmonen M. Lombardo C. Pihlajaniemi T. Alitalo K. Vuori K. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 1024-1029Crossref PubMed Scopus (421) Google Scholar). The current studies on HDMEC adhesion onto endostatin-derived peptides identified a putative integrin-binding sequence motif of human endostatin. An arginine-rich, endostatin-derived peptide termed ES-2 most efficiently promoted endothelial cell adhesion. Our results imply that the adhesion is mediated by β1 integrin, which is the major fibronectin receptor in mammalian cells as a complex with the α5 subunit. This integrin has previously been observed to modulate endostatin-induced signal transduction and disassembly of the actin cytoskeleton (19Wickström S.A. Alitalo K. Keski-Oja J. J. Biol. Chem. 2003; 278: 37895-37901Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar, 20Sudhakar A. Sugimoto H. Yang C. Lively J. Zeisberg M. Kalluri R. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 4766-4771Crossref PubMed Scopus (436) Google Scholar). Function-neutralizing antibodies against integrins α1, α2, α5, and αv did not significantly decrease cell adhesion. In contrast, the function-neutralizing β3 integrin antibodies moderately enhanced cell adhesion to ES-2. The reason for this enhancement is unclear at present but might result from trans-regulation, a phenomenon by which specific integrins bound by agonists or antagonists regulate the function and activation status of other integrins (27Schwartz M.A. Ginsberg M.H. Nat. Cell Biol. 2002; 4: E65-E68Crossref PubMed Scopus (681) Google Scholar). In addition to integrins, cell adhesion is modulated by cell surface heparan sulfate proteoglycans. These two families of adhesion receptors act often in coordination to promote cell spreading and rearrangement of the cytoskeleton (28Woods A. Couchman J.R. Trends Cell Biol. 1998; 8: 189-192Abstract Full Text Full Text PDF PubMed Scopus (169) Google Scholar). Adhesion of endothelial cells to ES-2 was strongly inhibited by heparin. This suggests that heparan sulfate proteoglycans could be involved in ES-2-mediated cell attachment. Fulllength endostatin binds with low affinity to glypican, a cell surface heparan sulfate proteoglycan expressed on endothelial cells (17Karumanchi S.A. Jha V. Ramchandran R. Karihaloo A. Tsiokas L. Chan B. Dhanabal M. Hanai J.I. Venkataraman G. Shriver Z. Keiser N. Kalluri R. Zeng H. Mukhopadhyay D. Chen R.L. Lander A.D. Hagihara K. Yamaguchi Y. Sasisekharan R. Cantley L. Sukhatme V.P. Mol. Cell. 2001; 7: 811-822Abstract Full Text Full Text PDF PubMed Scopus (271) Google Scholar). In addition, endostatin-induced disassembly of the actin cytoskeleton has been observed to require an interaction with an unidentified heparan sulfate proteoglycan (19Wickström S.A. Alitalo K. Keski-Oja J. J. Biol. Chem. 2003; 278: 37895-37901Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar). Heparan sulfate proteoglycans act as co-receptors also for bFGF, but endostatin does not seem to compete for the same binding sites on the endothelial cell surface (29Chang Z. Choon A. Friedl A. Am. J. Pathol. 1999; 155: 71-76Abstract Full Text Full Text PDF PubMed Scopus (61) Google Scholar). The possibility that endothelial cell adhesion to ES-2 would be a nonspecific effect resulting from the basic residues of the peptide was excluded by the observations that peptide ES-4, which also contains three basic amino acid residues, did not efficiently promote cell adhesion. Peptides ES-5 and ES-3 were almost as efficient as full-length endostatin in promoting cell attachment, but they did not seem to act via α5 or β1 integrins. Interestingly, cell adhesion to both of the peptides was inhibited by heparin. ES-5 also inhibited endothelial cell adhesion to immobilized endostatin, but was less effective than ES-2. However, it did not induce disruption of the actin cytoskeleton or inhibit endothelial cell migration. Therefore, the effects of this peptide appear to be very different from those of ES-2. The corresponding sequence motif in endostatin might, however, be involved in the cell surface association of the protein. A mutant form of mouse endostatin lacking nine C-terminal amino acids and thus most of the amino acids included in the ES-5 peptide has been found to be as effective as full-length endostatin in inhibiting tumor growth in mice (30Dhanabal M. Ramchandran R. Volk R. Stillman I.E. Lombardo M. Iruela-Arispe M.L. Simons M. Sukhatme V.P. Cancer Res. 1999; 59: 189-197PubMed Google Scholar). This implies that this sequence motif might not be essential for endostatin function. The ligands of α5β1 integrins consist of ECM or cell surface proteins containing the Arg-Gly-Asp (RGD) sequence (31Ruoslahti E. Annu. Rev. Cell Dev. Biol. 1996; 12: 697-715Crossref PubMed Scopus (2570) Google Scholar). This sequence motif is not present in endostatin or ES-2, which suggests that the binding of these two peptides to the integrin could occur at a site distinct from the ligand binding pocket. In support of this, ES-2 did not significantly inhibit endothelial cell adhesion on fibronectin. 2S. A. Wickström and J. Keshi-Oja, unpublished observations. In addition, in contrast to the integrin ligands containing the RGD-sequence, ES-2 promoted only a limited degree of cell spreading and did not induce the formation of actin stress fibers or abundant focal adhesions. Endothelial cell contact with an increased concentration of immobilized α5β1 integrin ligand is known to enhance cell migration in a haptotactic manner (32Bauer J.S. Schreiner C.L. Giancotti F.G. Ruoslahti E. Juliano R.L. J. Cell Biol. 1992; 116: 477-487Crossref PubMed Scopus (116) Google Scholar). We observed that immobilized full-length endostatin or ES-2 were inefficient in promoting haptotactic cell migration.2 However, endostatin seems to exist as a monomer, and monovalent ligand occupancy of integrins is considered not to be sufficient for efficient accumulation of cytoskeletal proteins (33Miyamoto S. Akiyama S.K. Yamada K.M. Science. 1995; 267: 883-885Crossref PubMed Scopus (790) Google Scholar). Further studies are required to define the nature of the interaction and possible co-receptors between α5β1 integrin and endostatin. Modified Boyden chamber migration assays using bFGF as a chemotactic stimulus revealed that the peptide ES-2, containing the surface-exposed arginine residues Arg62, Arg63, and Arg66 displayed anti-migratory potential comparable to full-length endostatin when used in a 10-fold molar excess. The functional role of these specific arginine residues has been assessed in full-length mouse endostatin. The arginine residues of ES-2 are essential for the heparin-binding properties of endostatin. In addition, endostatin mutants in which these arginines have been substituted with alanines lose their ability to inhibit endothelial cell migration in vitro and angiogenesis in vivo (22Sasaki T. Larsson H. Kreuger J. Salmivirta M. Claesson-Welsh L. Lindahl U. Hohenester E. Timpl R. EMBO J. 1999; 18: 6240-6248Crossref PubMed Scopus (200) Google Scholar). The subsequent observation of partial loss of function in ES-2 mutants lacking one or two out of the three arginines, and complete loss of function in the mutant lacking all three arginines, further emphasized the importance of the arginine residues in the anti-migratory potential of these peptides. ES-4 containing three surface-exposed arginines, Arg128, Arg129, and Arg139, was significantly less potent than ES-2 in inhibiting endothelial cell migration. Interestingly, whereas directed cell migration was inhibited by endostatin and ES-2, random cell migration was essentially unaffected or even modestly increased.2 This could be a consequence of alterations in cell-matrix interactions induced by these treatments. Both contractile and protrusive forces are required to move the cell body forward. These forces are generated by actin stress fiber contraction and traction from the focal adhesions (34Lauffenburger D.A. Horwitz A.F. Cell. 1996; 84: 359-369Abstract Full Text Full Text PDF PubMed Scopus (3278) Google Scholar). The strength of cell adhesion to the ECM determines the speed of migration. If the adhesion is too strong, the cell remains stationary; if it is too weak, enough traction cannot be generated for movement. The direction of cell movement is controlled by growth factors such as bFGF, which engage in extensive cross-talk with integrins to enhance cell polarization and directed cell motility (35Eliceiri B.P. Circ. Res. 2001; 89: 1104-1110Crossref PubMed Scopus (314) Google Scholar). Disruption of actin stress fibers and focal adhesions by endostatin or ES-2 would thus result in decreased adhesion and increased random migration. However, the loss of these structures at the leading edge could inhibit directional migration toward a chemotactic stimulus. Another possibility is that the disassembly of focal adhesions could lead to alterations in bFGF signaling, which occurs in a focal adhesion-dependent manner. Previous studies have, however, not provided evidence for endostatin-induced modulation of major growth factor signaling pathways (36Eriksson K. Magnusson P. Dixelius J. Claesson-Welsh L. Cross M.J. FEBS Lett. 2003; 536: 19-24Crossref PubMed Scopus (119) Google Scholar). Finally, the observed inhibition of tubular morphogenesis in endostatin- and ES-2-treated cells could reflect the loss of directed cell migration. In a recent study four endostatin-derived peptides spanning amino acids 6–178 were used to identify regions of the protein with anti-angiogenic potential (37Chillemi F. Francescato P. Ragg E. Cattaneo M.G. Pola S. Vicentini L. J. Med. Chem. 2003; 46: 4165-4172Crossref PubMed Scopus (35) Google Scholar). The study revealed a peptide containing amino acids 6–49 to be capable of inhibiting human umbilical vein-derived endothelial cell proliferation and migration and effective in inhibiting angiogenesis in vivo. No activity was observed for the 42-amino acid peptide covering amino acids 50–92, which contains the sequence for the ES-2 peptide (11 amino acids) used in the current study. Importantly, the study did not assess the ability of these peptides to bind to the endothelial cell surface. The reason for the differences between those results and the current study is unclear at present. However, the report (37Chillemi F. Francescato P. Ragg E. Cattaneo M.G. Pola S. Vicentini L. J. Med. Chem. 2003; 46: 4165-4172Crossref PubMed Scopus (35) Google Scholar) describes aggregation of the relatively long endostatin-derived peptides and suggests that the differences in the secondary structures may result in differences in the activities of the peptides. The observed aggregation could significantly alter the biological properties of the peptides and mask putative intra-peptide cell-binding sites such as in peptide ES-2. In addition, studies using peptide fragments from another basement membrane-derived angiogenesis inhibitor, tumstatin, suggest that the length of the peptide is a critical determinant of biological activity. This is most likely due to differences in secondary structure of peptides of different lengths (38Floquet N. Pasco S. Ramont L. Derreumaux P. Laronze J.Y. Nuzillard J.M. Maquart F.X. Alix A.J. Monboisse J.C. J. Biol. Chem. 2003; 279: 2091-2100Abstract Full Text Full Text PDF PubMed Scopus (73) Google Scholar). Furthermore, the peptides with anti-angiogenic activity (37Chillemi F. Francescato P. Ragg E. Cattaneo M.G. Pola S. Vicentini L. J. Med. Chem. 2003; 46: 4165-4172Crossref PubMed Scopus (35) Google Scholar) do not contain the specific arginine residues found to be important for the biological activity of endostatin, suggesting that additional novel domains could be involved in the activity of endostatin. We thank Sami Starast and Anne Remes for technical assistance.