HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 31, 23745-23750, August 4, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/31/23745    most recent C000186200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Maeshima, Y. Articles by Kalluri, R. Search for Related Content PubMed PubMed Citation Articles by Maeshima, Y. Articles by Kalluri, R. Social Bookmarking                      What's this?

Two RGD-independent _v3 Integrin Binding Sites on Tumstatin Regulate Distinct Anti-tumor Properties^*

Yohei Maeshima, Pablo C. Colorado, and Raghu Kalluri^§

From the Department of Medicine and the Cancer Center, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts 02215

Received for publication, March 21, 2000, and in revised form, May 30, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Vascular basement membrane is an important regulator of angiogenesis and undergoes many alterations during angiogenesis and these changes are speculated to influence neovascularization. Recently, fragments of collagen molecules have been identified to possess anti-angiogenic activity. Tumstatin (3(IV)NC1 domain) is one such novel molecule with distinct anti-tumor properties and possesses an N-terminal (amino acids 54-132) anti-angiogenic and a C-terminal (amino acids 185-203) anti-tumor cell activity (Maeshima, Y., et al. 2000) J. Biol. Chem. 275, 21340-21348). Previous studies have identified the 185-203 amino acid sequence as a ligand for _v3 integrin (Shahan, T. A., et al. (1999) Cancer Res. 59, 4584-4590). In the present study, we found distinct additional RGD-independent _v3 integrin binding site within 54-132 amino acids of tumstatin. This site is not essential for inhibition of tumor cell proliferation but necessary for the anti-angiogenic activity. A fragment of tumstatin containing 54-132 amino acid (tum-2) binds both endothelial cells and melanoma cells but only inhibited proliferation of endothelial cells, with no effect on tumor cell proliferation. A similar experiment with fragment of tumstatin containing the 185-203 amino acid (tum-4) demonstrates that it binds both endothelial cells and melanoma cells but only inhibits the proliferation of melanoma cells. The presence of cyclic RGD peptides did not affect the _v3 integrin-mediated activity of tumstatin, although significant inhibition of endothelial cell binding to vitronectin was observed. The two distinct RGD-independent binding sites on tumstatin suggest unique _v3 integrin-mediated mechanisms governing the two distinct anti-tumor properties of tumstatin.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Tumor growth and metastasis is associated with angiogenesis (1, 2). Integrin _v3 is associated with angiogenic vascular cells and plays a critical role in angiogenesis promoting endothelial cell survival (3, 4). Vascular basement membrane is speculated to play an important role in regulating angiogenesis (5, 6). Type IV collagen is expressed as six distinct -chains, namely 1-6 (7), and assembled into triple helices and further forms a network to provide a scaffold for other macromolecules in basement membranes. These -chains are composed of three domains, N-terminal 7 S domain, the middle triple-helical domain, and C-terminal globular noncollagenous domain (NC1)¹ (8). Type IV collagen is thought to play a crucial role in endothelial cell proliferation and behavior during the angiogenic process (5, 6, 9). Synthetic peptides derived from NC1 domain of 3 chain of type IV collagen (3(IV)NC1) has been shown to bind and inhibit the proliferation of melanoma, and other epithelial tumor cell lines, in vitro (10). Shahan et al. (11) localized the binding site for melanoma cells to amino acids 185-203 of 3(IV)NCI and also found this site bound to _v3 integrin and CD47/IAP. Recently we identified that 3(IV)NC1 (termed "tumstatin" to include it in the family of endogenous inhibitor of angiogenesis derived from larger matrix protein with tumor "stasis" property) possessed a novel anti-angiogenic activity, and this activity was derived from amino acids 54-132, which does not contain the previously reported anti-tumor cell proliferation site (12). Recently, Petitclerc et al. (13) reported similar results further supporting the notion that recombinant 3(IV)NCI inhibited angiogenesis and tumor growth. Although in these studies the issue of 185-203 _v3 binding site or RGD-independent sites within the NC1 domain were not addressed (13). In the present study, we demonstrate that amino acids 54-132 binds to _v3 integrin in an RGD-independent manner, and this binding is essential for the anti-angiogenic property.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Production of Recombinant Tumstatin (₃(IV)NCI), Deletion Mutants and Endostatin, and Synthetic Peptides-- The sequence encoding tumstatin or deletion mutants (tum-1-4) was amplified using polymerase chain reaction from the 3(IV)NCI/pDS vector (14). The resulting cDNA fragment was ligated into pET22b(+) or pET28a(+) (Novagen, Madison, WI). Expression of recombinant protein in Escherichia coli and purification using nickel-nitrilotriacetic acid-agarose column (Qiagen) was performed as described previously (9, 12). Recombinant mouse endostatin was expressed and purified as described previously (15). Only soluble protein was used in these experiments. Synthetic peptide CDCRGDCFC (RGD-4C), peptide CNGRC, and peptide PGLKGKRGDSGSPATWTTRG, which consists of the N-terminal 20 amino acids of tumstatin, were kindly provided by Ilex oncology, Inc. (San Antonio, TX) as gifts. These peptides were synthesized and characterized as described previously (16, 17).

Immunoblotting-- Recombinant tumstatin and deletion mutants were analyzed by SDS-PAGE and immunoblotting as described previously (18). Rabbit antibody raised against C-terminal 36 amino acids of tumstatin (tum-4 antibody) was prepared as described previously (19). Goat anti-rabbit IgG antibody conjugated with horseradish peroxidase was purchased from Sigma.

Cell Lines and Culture-- Bovine pulmonary arterial endothelial cells (C-PAE) and human umbilical vein endothelial cells (HUVECs) were all obtained from American Type Culture Collection. These cell lines were maintained in DMEM (C-PAE; Life Technologies, Inc.) supplemented with 10% fetal calf serum (FCS), 100 units/ml of penicillin, and 100 mg/ml streptomycin, or in EGM-2 (HUVEC; Clonetics, San Diego, CA). The melanoma cell line WM-164 was obtained from Dr. Meenhard Herlyn at the Wistar Institute (Philadelphia, PA) and maintained as described previously (20).

Proliferation Assay-- A suspension of C-PAE cells (7,000 cells/well, passage 2-4) in DMEM containing 1% FCS was added onto 96-well plates precoated with fibronectin. After 24 h, medium was replaced with DMEM containing 20% FCS and recombinant protein. Then, after 48 h methylene blue staining was performed as described previously (9). Polymyxin B (Sigma, 5 µg/ml) was used to inactivate endotoxin (21). In separate experiments, T1 synthetic peptides were used for treatment.

Competition Proliferation Assay-- C-PAE cells were plated onto 96-well plates and serum-depleted, as described above, and tumstatin (0.1 µg/ml, final concentration), tum-2 (1 µg/ml, final concentration), or tum-4 (50 µg/ml, final concentration) was incubated with varying concentration of human _v3 or ₅₁ protein (Chemicon) for 30 min at room temperature. Then the mixture was added onto wells and incubated for 48 h. Proliferation assays were performed using methylene blue staining method. In separate experiments, WM-164 cells were examined.

Cell Attachment Assay-- This assay was performed as described previously (22). 96-Well plates were coated with 10 µg/ml recombinant protein, mouse laminin-1, or human type IV collagen (Collaborative Biomedical Products) overnight. Vitronectin (Collaborative Biomedical Products) was coated at a concentration of 0.5-2.5 µg/ml. Plates were blocked with 100 mg/ml of bovine serum albumin (BSA, Sigma) for 2 h. Synthetic peptides RGD-4C or CNGRC were coated at a concentration of 25 µg/ml, and plates were blocked with 30 mg/ml of BSA (Sigma) for an hour. T1 peptides were coated at a concentration of 40 µg/ml, and plates were blocked as described above. HUVECs or C-PAEs (1.5 × 10⁵ cells/ml, in M-199) were incubated with 10 µg/ml of antibody for 15 min. In some assays, cells were incubated with synthetic peptides RGD-4C or CNGRC (1-5 µg/ml, final concentration). Then, 100 µl of the cell suspension/well was added onto plates and incubated for 45 min at 37 °C. After washing, the number of attached cells was determined with methylene blue staining. Control mouse IgG1 and mouse monoclonal antibody to the human ₁ integrin (clone P4C10) were purchased from Life Technologies, Inc. Monoclonal antibody to the human ₁-₆ (Alpha Integrin Blocking and IHC kit), _v5 integrin (clone P1F6) were purchased from Chemicon. WM-164 cells were examined similarly in separate experiments.

Statistical Analysis-- All values are expressed as mean ± S.E. Analysis of variance with a one-tailed Student's t test was used to identify significant differences in multiple comparisons. A level of p < 0.05 was considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Expression and Purification of Human Tumstatin and Deletion Mutants-- Recombinant human tumstatin was produced in E. coli using an expression plasmid, pET22b or pET28a, as a fusion protein with a C-terminal six-histidine tag (12). Four different deletion mutants of tumstatin were expressed in E. coli using pET28a system (Table I). Tumstatin consists of 244 amino acids, including 12 amino acids from triple-helical portion located in the N-terminal portion and 232 amino acids derived from NC1 domain. tum-1 consists of 191 amino acids and is lacking N-terminal 53 amino acids. tum-2 consists of 132 amino acids in the N-terminal half portion of tumstatin, and tum-3 is the C-terminal half portion (112 amino acids). tum-4 consists of 64 amino acids in the C terminus of tumstatin (amino acids 181-244). The E. coli expressed deletion mutants were isolated predominantly as a soluble protein and SDS-PAGE analysis revealed a monomeric band corresponding to the size of each protein as described previously (12). Tumstatin, tum-1, tum-3, and tum-4 were immunodetectable using anti-tum-4 antibody in Western blotting (Fig. 1 panel A). tum-2 protein, which does not contain the tum-4 sequence and is anti-angiogenic, was not detectable by anti-tum-4 antibody (Fig. 1, panel A).

View larger version (43K): [in this window] [in a new window]   Fig. 1.   Immunoblot of tumstatin and deletion mutants and cell attachment assay using tumstatin. Panel A, immunoblot analysis using anti-tum-4 antibody (polyclonal rabbit IgG) with the dilution of 1:250. Molecular weight marker (MW): lane 1, tumstatin; lane 2, tum-1; lane 3, tum-2; lane 4, tum-3; lane 5, tum-4. 500 ng of protein was loaded in each lane under nonreducing condition. The arrow indicates the band composed of tum-3 monomer. The arrowhead shows the position of the tum-4 monomer. Panel B, attachment assay was performed with HUVECs using tumstatin, type IV collagen, vitronectin, and laminin-1 for coating culture plates. Cells were incubated with monoclonal anti-human integrin antibodies or control mouse IgG (10 µg/ml), plated onto precoated 96-well plates, and then the number of cells attaching to the plates was determined. Matrix used for precoating plates is indicated on the top of each graph (B-F). Antibodies used for incubation with cells are described at the bottom (B-F). The number of cells on BSA-coated plates was shown as a negative control. Cell attachment onto tumstatin-coated plates was significantly inhibited by anti-6, -1, or -v3 antibody as compared with control IgG-treated. When v3 and 1 antibody were used together, cell attachment was further inhibited. Panels C-E, type IV collagen-, vitronectin-, and laminin-1-coated plates were used to demonstrate the activity of antibodies in blocking cell binding. Panel F, attachment assay was performed with C-PAE cells as described above. Cell attachment onto tumstatin-coated plates were significantly inhibited by anti-1 or v3 antibody as compared with control IgG-treated. When v3 and 1 antibodies were used together, cell attachment was further inhibited. When v5 antibody was used, cell attachment was not inhibited. Each column represents the mean ± S.E. of triplicate wells. These experiments were repeated three times. *, p < 0.05 by one-tailed Student's t test.

Effect of Deletion Mutants of Tumstatin on Proliferation of C-PAE and WM-164 Cells-- tum-1 and tum-2 inhibited the proliferation of C-PAEs in a dose-dependent manner. At 15 µg/ml, tum-1 or tum-2 inhibited C-PAE proliferation by 65.6 and 73.3%, respectively (Table I). Tumstatin at 15 µg/ml inhibited the proliferation of C-PAE cells by 78.5%. In contrast, tum-3 and tum-4 did not inhibit the proliferation of C-PAE cells. tum-4 inhibited the proliferation of WM-164, and tum-1 or tum-2 did not (12). At 50 µg/ml, tum-4 inhibited the proliferation of WM-164 by 46.1% (Table I). In contrast, 50 µg/ml of tum-4 did not inhibit the proliferation of C-PAE cells (data not shown).

Tumstatin Binds to _v3 and ₆₁ Integrin on Endothelial Cells-- We examined the attachment of HUVECs and C-PAEs to tumstatin-coated plates in the presence of ₁-₆, ₁, and _v3 integrin blocking antibody. As shown in Fig. 1, panel B, _v3 antibody inhibited the attachment of HUVECs by 80%, and ₆ or ₁ antibody blocked by 54% as compared with control IgG treatment. Although ₅ antibody exhibited minor inhibition (20%), ₁-₄ antibody did not block cell attachment. When _v3 antibody and ₁ antibody were used together, cell binding was blocked by more than 91%. Comparable inhibition was also observed using C-PAE cells instead of HUVECs (Fig. 1, panel F). The binding of C-PAEs to tumstatin was not inhibited by _v5 antibody (Fig. 1, panel F). Collectively, these results suggest that _v3 and ₆₁ integrins may play roles in the anti-angiogenic property of tumstatin.

As controls, culture plates coated with type IV collagen, vitronectin, and laminin-1 were used to demonstrate the blocking activity of various antibodies used in this study (Fig. 1, panels C-E). The ₁₁ and ₂₁ integrins bind collagens (23, 24). Cell binding onto type IV collagen coated plates was partially inhibited by ₁ (20%), ₂ (27%) antibody, and ₁ antibody (53%), as compared with cells incubated with control IgG. _v3 integrin is a major receptor for vitronectin (25). Cell binding onto vitronectin-coated plates was inhibited by _v3 antibody by 61%. Previous studies have shown that ₅₁ and ₆₁ integrins bind laminin-1 (26, 27). Anti-₅ or ₆ antibody blocked the binding of endothelial cells onto laminin-1-coated plates by 50 and 89%, respectively. These control experiments further support the functional integrin-blocking activities of these antibodies.

Tumstatin Binds to Endothelial Cells via _v3 Integrin, Independent of RGD Sequence-- Within the sequence of tumstatin used in our study, an RGD sequence at amino acids 7-9 is present, and it is possible this RGD sequence binds to _v3 integrin on endothelial cells. To address this issue, synthetic peptide RGD-4C, peptides CNGRC, which were previously reported to bind to vascular endothelial cells (16), and T1 peptide, which contains the N-terminal RGD containing sequence of tumstatin, were used in cell attachment assays. When plates were coated with RGD peptide, cell attachment of C-PAEs was increased by 144.25% as compared with noncoated (PBS) plates (Fig. 2, panel A). Cell attachment to plates coated with peptide CNGRC was also increased by 106.00%, as was expected from the previous studies (16). Next, plates were coated with vitronectin, and C-PAE cells were preincubated with RGD-4C or CNGRC peptides and plated on vitronectin. RGD peptide (5 µg/ml) significantly inhibited cell attachment onto vitronectin coated plates by 76.5%, whereas CNGRC peptides did not inhibit the binding of C-PAEs to vitronectin (Fig. 2, panel B). In contrast, attachment of C-PAEs to tumstatin-coated plates was not inhibited by preincubation with RGD-4C peptide (Fig. 2, panel C). When cells were plated on tumstatin in the presence of soluble _v3 integrin protein to compete with endothelial cell surface binding to tumstatin, cell attachment was inhibited by 46.7%. Incubation of cells with RGD peptides in addition to _v3 integrin protein did not alter the inhibitory effect of _v3 integrin protein on cell attachment to tumstatin (Fig. 2, panel C). When HUVECs were plated on T1 peptide (tumstatin RGD peptide)-coated plates, the cell attachment was similar to noncoated (PBS) plates (Fig. 2, panel D), indicating that endothelial cells do not exhibit specific attachment to this tumstatin derived RGD peptide. We observed similar results using C-PAEs (data not shown). These results suggest that _v3 integrin may bind to tumstatin via RGD-independent mechanism.

View larger version (46K): [in this window] [in a new window]   Fig. 2.   Cell attachment assay using synthetic peptides and deletion mutants of tumstatin. Panel A, attachment assay was performed with C-PAEs using synthetic peptides RGD-4C (RGD) or CNGRC (CNGRC; 25 µg/ml) for coating plates. Cell attachment on RGD- or CNGRC-coated plates was increased as compared with noncoated (PBS) plates. Panel B, cell (C-PAE) attachment onto vitronectin-coated plates was significantly inhibited by RGD peptides (RGD, 5 µg/ml). Incubation of cells with CNGRC control peptides did not inhibit cell attachment. The matrix used for precoating plates is indicated on the top of each graph (B, C, E-G). Panel C, cell attachment onto tumstatin-coated plates was significantly inhibited by v3 integrin protein. RGD or CNGRC peptides had no effect in inhibiting cell attachment. Cell attachment after addition of both v3 integrin protein and RGD peptides was not different from incubation with v3 integrin protein alone. Panel D, T1 peptides from N-terminal 20 amino acids of tumstatin containing RGD sequence were coated on plates. Attachment of HUVECs onto T1 peptide-coated plates was similar to the noncoated (PBS) plates. Panels E-G, attachment assay was performed with C-PAEs using tum-1, tum-2, and tum-4 for coating plates. Antibodies used for incubation with cells are described at the bottom (E-G). Panel E, cell attachment onto tum-1-coated plates was significantly inhibited by anti-v3 integrin antibody. Panels F and G, cell attachment onto tum-2- or tum-4-coated plates were significantly inhibited by anti-v3 antibody. Anti-v5 integrin antibody did not inhibit cell attachment onto tum-1-, tum-2-, or tum-4-coated plates. Panel H, cell attachment of WM-164 melanoma cells were significantly increased when cells were plated on tumstatin-, tum-1-, tum-2-, or tum-4 (10 µg/ml)-coated plates as compared with PBS control. Each column represents the mean ± S.E. of triplicate wells. These experiments were repeated three times. *, p < 0.05 by one-tailed Student's t test.

tum-1, tum-2, and tum-4 Bind to _v3 Integrin on Endothelial Cells-- We examined the attachment of C-PAEs to tum-1-, tum-2-, or tum-4-coated plates in the presence of _v3 or _v5 integrin blocking antibody. As shown in Fig. 2, panels E-G, _v3 antibody inhibited the attachment of CPAEs to tum-1, tum-2, or tum-4 by 55.9, 69.8, and 62.6, respectively. Binding of CPAEs to tum-1-, tum-2-, and tum-4-coated plates was not inhibited by _v5 antibody.

Tumstatin, tum-1, tum-2, and tum-4 Binds to WM-164 Melanoma Cells-- We next examined the attachment of WM-164 cells to tumstatin, tum-1-, tum-2-, or tum-4-coated plates. As shown in Fig. 2, panel H, melanoma cells bind specifically to tumstatin, tum-1-, tum-2-, or tum-4-coated plates (8.1-, 8.4-, 9.6-, and 8.2-fold increase as compared with PBS, respectively).

Reversal of Anti-endothelial Cell Proliferative Effect of Tumstatin and tum-2 by Soluble _v3 Integrin Protein-- Tumstatin was incubated with _v3 integrin protein for 30 min and then added to C-PAEs, which were then plated on culture plates and serum-depleted overnight. Anti-proliferative effect of tumstatin was reversed dose-dependently with increasing doses of _v3-soluble protein (Fig. 3, panel A). The _V3 protein at 2.4 µg/ml (3-fold molar excess of tumstatin) significantly reversed tumstatin-induced anti-proliferative effect by 43.1%. Treatment with _v3 protein alone without tumstatin did not inhibit endothelial cell proliferation. When ₅₁ integrin protein was incubated with tumstatin instead of _v3 protein, anti-proliferative effect of tumstatin was not reversed (Fig. 3, panel B). When tum-2 was incubated with _v3 protein for 30 min, and then added to C-PAEs as described above, the anti-proliferative effect of tum-2 was also reversed dose-dependently with increasing doses of _v3-soluble protein (Fig. 3, panel C). The _v3 protein at 2 µg/ml significantly recovered tum-2-induced anti-proliferative effect by nearly 74.1%. This increased recovery of activity, as compared with the recovery of tumstatin-induced anti-proliferative effect, may be partially mediated by the enhanced binding of _v3 integrin protein to tum-2 compared with tumstatin, due to it's significantly smaller size and thus potentially due to decreased steric hindrance for binding to _v3 integrin.

View larger version (45K): [in this window] [in a new window]   Fig. 3.   Competitive proliferation assay and the effect of anti-v3 integrin antibody on endothelial cell proliferation. Panel A, tumstatin was incubated with v3 integrin protein (Chemicon) for 30 min at room temperature and added onto C-PAE cells grown on 96-well plates. After incubating 48 h, the cell number was determined using methylene blue staining. Anti-proliferative effect of tumstatin (0.1 µg/ml, final concentration) was significantly decreased by v3 protein (0.8-2.4 µg/ml, final concentration). v3 protein itself did not inhibit proliferation. Panel B, tumstatin was incubated with 51 integrin protein (Chemicon) for 30 min at room temperature and added onto C-PAE cells. Anti-proliferative effect of tumstatin (0.1 µg/ml, final concentration) was not decreased by 51 protein (0.08-2.6 µg/ml, final concentration). 51 protein itself did not inhibit proliferation. Panel C, tum-2 was incubated with v3 integrin protein (Chemicon) for 30 min at room temperature and added onto C-PAE cells. Anti-proliferative effect of tum-2 (1 µg/ml, final concentration) was significantly decreased by v3 protein (1-2 µg/ml, final concentration). Panel D, tum-4 was incubated with v3 integrin protein (Chemicon) for 30 min at room temperature and added onto WM-164 cells grown on 96-well plates. After incubating 48 h, the cell number was determined using methylene blue staining. Anti-proliferative effect of tum-4 (50 µg/ml, final concentration) was significantly decreased by v3 protein (1-2 µg/ml, final concentration). Panel E, T1 peptides were used for cell proliferation assay using C-PAEs. Relative cell number was determined by methylene blue staining method. T1 peptides did not inhibit proliferation of C-PAE cells even at 10 or 50 µg/ml. Panel F, equimolar amount of human tumstatin and anti-v3 integrin antibody was added to C-PAE cells to compare the effect on endothelial cell proliferation. Increasing amount of anti-v3 integrin antibody did not inhibit proliferation, in contrast human tumstatin exhibited dose-dependent inhibition of proliferation. Mouse endostatin that was used as a positive control also inhibited endothelial cell proliferation. Control IgG did not inhibit proliferation of C-PAEs. Each column or points represents the mean ± S.E. of triplicate wells. These experiments were repeated three times. *, p < 0.05 by one-tailed Student's t test.

Reversal of Anti-proliferative Effect of tum-4 on Melanoma Cells by Soluble _v3 Integrin Protein-- In our previous study we reported that tumstatin, tum-1, and tum-2 contained sequences that caused inhibition of endothelial cell proliferation (12). These proteins did not cause inhibition of melanoma cell proliferation. In contrast, tum-3 and tum-4 protein did not cause inhibition of endothelial cell proliferation, but tum-4 caused inhibition of melanoma cell proliferation. In this study, tum-4 was incubated with _v3 protein for 30 min and then added to WM-164 cells. Anti-proliferative effect of tum-4 was reversed dose-dependently with increasing doses of _v3-soluble protein (Fig. 3, panel D). The _v3 protein at 2 µg/ml significantly recovered tum-4-induced anti-proliferative effect by 76.7%. Treatment with _v3 protein alone without tum-4 did not inhibit melanoma cell proliferation.

T1 Peptide (RGD Containing Peptide from Tumstatin) Does Not Inhibit Proliferation of C-PAE Cells-- T1 peptide, which contains N-terminal RGD sequence of tumstatin, was examined for it's effect on endothelial cell proliferation. T1 peptides at 10 and 50 µg/ml did not inhibit proliferation of C-PAE cells (Fig. 3, panel E). This result further indicates that tumstatin's activity to inhibit endothelial cell proliferation is RGD-independent.

Tumstatin Inhibits Proliferation of Endothelial Cells in Contrast to Anti-_v3 Integrin Antibody (LM609)-- To compare the anti-endothelial cell property of tumstatin and anti-_v3 integrin antibody, an equimolar amount of tumstatin and anti-_v3 integrin antibody was added to C-PAE cells to assess their impact on endothelial cell proliferation. Anti-_v3 integrin antibody (LM609) did not inhibit proliferation at any of the doses used in this study. In contrast, both tumstatin and endostatin exhibited dose-dependent inhibition of endothelial cell proliferation (Fig. 3, panel F). Control IgG did not inhibit proliferation of these cells.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Recent reviews have suggested that a switch to an angiogenic phenotype requires both up-regulation of angiogenic stimulators and down-regulation of angiogenesis inhibitors (1, 28). In this regard, vascular endothelial growth factor (VEGF) and basic fibroblast growth factor (bFGF) are major inducers of angiogenesis. A number of angiogenesis inhibitors have been recently identified, and certain factors, such as angiostatin (29), endostatin (30), canstatin (9), and arresten (6), are tumor-associated angiogenesis inhibitors that are generated in vivo. Recently, tumstatin, recombinant NC1 domain of the 3 chain of human type IV collagen (8, 31), was identified as a protein with the dual anti-tumor activities. Tumstatin was identified as an inhibitor of vascular endothelial cell proliferation and formation of new blood vessels using in vitro and in vivo assays of angiogenesis and tumor growth (10, 12). Petitclerc et al. (13) recently suggested that NC1 domain of ₂(IV), 3(IV), and 6(IV) may exert their anti-angiogenic activity via binding to _v3 or ₁ integrins. Although in these studies, functional experiments were not performed (13). According to their study utilizing cell adhesion assays, 3(IV)NC1 (tumstatin) was a strong RGD-dependent ligand for _v3, and not for _v5, integrin (13). Our results using cell adhesion method with two different endothelial cell types suggest that _v3 and ₆₁ integrin bind to tumstatin and that the _v3 binding is RGD-independent. Recently, synthetic peptides (19 amino acids) corresponding to the C-terminal portion of tumstatin (amino acids 185-203) was reported to bind to _v3 integrin and inhibit tumor cell proliferation including melanoma cell lines (11). In our previous study, we identified that anti-angiogenic activity of tumstatin is located within amino acids 54-124 by deletion mutagenesis and showed that the C-terminal fragment of tumstatin, including amino acids 185-203, did not possess anti-angiogenic activity (12).

In this study, we used tumstatin and deletion mutants of tumstatin in cell adhesion assays to analyze for integrin binding sites. Within the N-terminal portion of tumstatin used in this study, there is a RGD sequence (amino acids 7-9) derived from triple-helical noncollagenous portion. The RGD sequence is generally considered as an important binding site for _v3 integrin receptor. In this study, we show that tum-1, lacking N-terminal 53 amino acids (including this RGD site), still binds to _v3 integrin. Also, a synthetic peptide T1 from 1-20 amino acids of tumstatin containing the RGD sequence neither bound to endothelial cells nor inhibited endothelial cell proliferation. These results suggest that RGD sequence in the N-terminal of tumstatin is not potentially a functional _v3 integrin binding site. Therefore, the binding property of tumstatin to endothelial cells is likely RGD-independent as was previously shown for the 185-203 amino acid sequence of tumstatin (11). Furthermore, tum-2 (amino acids 1-132), which does not contain the C-terminal _v3 binding site (amino acids 185-203), was shown to bind to _v3 in cell adhesion assay and inhibit endothelial cell proliferation. When tumstatin or tum-2 were preincubated with _v3 integrin protein to block their binding sites, anti-proliferative effect of tumstatin was significantly decreased (43-74%). This decrease can be considered impressive given that the affinity of soluble _v3 receptor protein for tumstatin may be much weaker and the binding likely inefficient as compared with membrane-bound _v3 integrin. This evidence strongly suggests that the _v3 binding site is located within amino acids 54-132. When soluble ₅₁ integrin protein was used instead of _v3 protein, the recovery effect was not observed, strongly suggesting that the effect was specifically dependent on the blockade of tumstatin binding to _v3 integrin. When endothelial cells were incubated with RGD-4C peptide, cell attachment to tumstatin-coated plates was not inhibited. Since the RGD-4C peptide binds to _v3 integrin receptors and blocks vitronectin binding to endothelial cells, we speculate that tumstatin may bind to a site on the _v3 integrin receptor, which is different from RGD binding site. Soluble _v3 integrin was able to inhibit endothelial cell attachment to tumstatin, possibly by binding to tumstatin and masking it's cell binding domain. These results again demonstrate the strong interaction of tumstatin with _v3 integrin receptor. Also, neutralizing the RGD-dependent _v3 integrin site by cyclic RGD peptide did not inhibit tumstatin binding to _v3 integrin receptor on endothelial cells, again strongly suggesting that tumstatin binds to a RGD-independent and distinct site on _v3 integrin receptor. Although, the possibility of an additional receptor for tumstatin cannot be ruled out due to the experimental strategy used in the current study.

Our results of _v3 integrin involvement for tumstatin's anti-angiogenic activity is consistent with the notion that VEGF up-regulates the expression of _v3 on endothelial cells (32, 33). Since angiogenesis depends on specific endothelial cell adhesive events mediated by _v3 integrin (3, 4), it is possible that the anti-angiogenic effect of tumstatin is mediated by disrupting the interaction of proliferating endothelial cells to the matrix component such as vitronectin and fibronectin, events considered as important anti-apoptotic signal (34). Whether tumstatin functions by directly suppressing the activity of VEGF and/or bFGF remains to be elucidated. Collectively, our results reveal two RGD-independent sites on tumstatin. The previously identified site (11) and our discovery of a second site reveal no sequence homology, and presumably they may bind to distinct sites on _v3 integrin receptor or bind to the same site by requiring different co-receptor protein for functionality. Interestingly, although the 185-203 sequence containing tum-4 binds to endothelial cells, and this binding can be inhibited by anti-_v3 integrin neutralizing antibody, this interaction is not sufficient to input an anti-angiogenic property to this molecule. Hence, these results strongly suggest that although both tumstatin sites have the capacity to bind to _v3 integrin on the cell surface of both cell types, additional cell-specific ligand-receptor interactions may be necessary for the specific effects on either endothelial or melanoma cells (35). In this regard, many recent studies suggest that _v3-ligand interaction may need other cell surface proteins to facilitate appropriate downstream signaling into the cell interior (36). The cell-specific regulation of _v3 integrin by different ligands (such as the two different sequence of tumstatin) may be facilitated by additional binding protein on the cell surface, such as CD47 (36-38). The presence of two novel RGD-independent _v3 sites on tumstatin, facilitating two unique anti-tumor activities, makes it a valuable therapeutic agent for inhibition of tumor growth.

	ACKNOWLEDGEMENTS

RGD, CNGRC, and T1 peptides were gifts from Ilex oncology, Inc. Also, we thank Dr. Meenhard Herlyn for providing us with WM-164 melanoma cells.

	FOOTNOTES

^* This work was supported in part by Grants DK-51711 and DK-55001 from the National Institutes of Health (to R. K.), 1998 Hershey Prostate Cancer Research Award (to R. K.), 1998 American Society of Nephrology Carl Gottschalk Research Award (to R. K.), and research funds from the Beth Israel Deaconess Medical Center.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Recipient of the 1999 Research Award for Young Scientists from the Inoue Foundation for Science of Japan.

^§ To whom correspondence should be addressed: Nephrology Division, Dept. of Medicine, RW 563a, Beth Israel Deaconess Medical Center, 330 Brookline Ave., Boston, MA 02215. Tel.: 617-667-0445; Fax: 617-975-5663; E-mail: rkalluri@caregroup.harvard.edu.

Published, JBC Papers in Press, June 2, 2000, DOI 10.1074/jbc.C000186200

	ABBREVIATIONS

The abbreviations used are: NC1, noncollagenous 1; FCS, fetal calf serum; PBS, phosphate-buffered saline; C-PAE, bovine pulmonary arterial endothelial cells; HUVEC(s), human umbilical vein endothelial cell(s); BSA, bovine serum albumin; PAGE, polyacrylamide gel electrophoresis; DMEM, Dulbecco's modified Eagle's medium; VEGF, vascular endothelial growth factor; bFGF, basic fibroblast growth factor.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				R. Silva, G. D'Amico, K. M. Hodivala-Dilke, and L. E. Reynolds Integrins: The Keys to Unlocking Angiogenesis Arterioscler. Thromb. Vasc. Biol., October 1, 2008; 28(10): 1703 - 1713. [Abstract] [Full Text] [PDF]

				H. P. Eikesdal, H. Sugimoto, G. Birrane, Y. Maeshima, V. G. Cooke, M. Kieran, and R. Kalluri Identification of amino acids essential for the antiangiogenic activity of tumstatin and its use in combination antitumor activity PNAS, September 30, 2008; 105(39): 15040 - 15045. [Abstract] [Full Text] [PDF]

				E. D. Karagiannis and A. S. Popel A systematic methodology for proteome-wide identification of peptides inhibiting the proliferation and migration of endothelial cells PNAS, September 16, 2008; 105(37): 13775 - 13780. [Abstract] [Full Text] [PDF]

				M. N. Rudzinski, L. Chen, and M. R. Hernandez Antiangiogenic Characteristics of Astrocytes From Optic Nerve Heads With Primary Open-angle Glaucoma Arch Ophthalmol, May 1, 2008; 126(5): 679 - 685. [Abstract] [Full Text] [PDF]

				C. S. Boosani, A. P. Mannam, D. Cosgrove, R. Silva, K. M. Hodivala-Dilke, V. G. Keshamouni, and A. Sudhakar Regulation of COX-2 mediated signaling by {alpha}3 type IV noncollagenous domain in tumor angiogenesis Blood, August 15, 2007; 110(4): 1168 - 1177. [Abstract] [Full Text] [PDF]

				E. Guerriero, J. Chen, Y. Sado, R. R. Mohan, S. E. Wilson, J. L. Funderburgh, and N. SundarRaj Loss of Alpha3(IV) Collagen Expression Associated with Corneal Keratocyte Activation Invest. Ophthalmol. Vis. Sci., February 1, 2007; 48(2): 627 - 635. [Abstract] [Full Text] [PDF]

				T. Kawaguchi, Y. Yamashita, M. Kanamori, R. Endersby, K. S. Bankiewicz, S. J. Baker, G. Bergers, and R. O. Pieper The PTEN/Akt Pathway Dictates the Direct {alpha}V{beta}3-Dependent Growth-Inhibitory Action of an Active Fragment of Tumstatin in Glioma Cells In vitro and In vivo Cancer Res., December 1, 2006; 66(23): 11331 - 11340. [Abstract] [Full Text] [PDF]

T. Miyoshi, S. Hirohata, H. Ogawa, M. Doi, M. Obika, T. Yonezawa, Y. Sado, S. Kusachi, S. Kyo, S. Kondo, et al. Tumor-specific expression of the RGD-{alpha}3(IV)NC1 domain suppresses endothelial tube formation and tumor growth in mice FASEB J, September 1, 2006; 20(11): 1904 - 1906.