Angioarrestin: An Antiangiogenic Protein with ... - Semantic Scholar

[CANCER RESEARCH 62, 3834 –3841, July 1, 2002]

Angioarrestin: An Antiangiogenic Protein with Tumor-inhibiting Properties Mohanraj Dhanabal,1,2 William J. LaRochelle,1 Michael Jeffers, John Herrmann, Luca Rastelli, William F. McDonald, Rajeev A. Chillakuru, Meijia Yang, Ferenc L. Boldog, Muralidhara Padigaru, Kelly D. McQueeney, Frank Wu, Stacey A. Minskoff, Richard A. Shimkets, and Henri S. Lichenstein CuraGen Corp., Branford, Connecticut 06405

ABSTRACT The angiopoietins comprise a family of proteins that have pro or antiangiogenic activities. Through a proprietary technology designed to identify transcripts of all expressed genes, we isolated a cDNA encoding an angiopoietin-related protein that we designate angioarrestin. The mRNA expression profile of angioarrestin was striking in that it was downregulated in many tumor tissues when compared with adjacent nontumor tissue, suggesting a role for this protein in tumor inhibition. To test this hypothesis, we ectopically expressed angioarrestin in HT1080 tumor cells and measured pulmonary tumor nodule formation in nude mice. HT1080 cells expressing angioarrestin showed a marked reduction in the number and size of tumor nodules. In vitro, the recombinant protein was systematically tested in a number of endothelial cell assays and found to block critical processes involved in the angiogenic cascade, such as vascular endothelial growth factor/basic fibroblast growth factor-mediated endothelial cell proliferation, migration, tubular network formation, and adhesion to extracellular matrix proteins. These findings reveal a novel function for angioarrestin as an angiogenesis inhibitor and indicate that the molecule may be a potential cancer therapeutic.

INTRODUCTION Angiogenesis plays an important role in embryogenesis and tumorigenesis (1). This statement is underscored by the fact that solid tumor growth beyond a few mm2 in diameter requires new blood vessel growth (2). The multistep angiogenesis process can be divided into induction and resolution phases. The former phase involves dynamic changes in endothelial cell-cell and cell-matrix interaction. This process includes degradation of basement membrane, proliferation, migration, and adhesion of endothelial cells (3). In the latter phase (resolution), perivascular-supporting cells are recruited, and a maturation process results in the assembly of fully functional new blood vessels (4, 5). All of these steps involve a wide variety of growth factors, receptors, proteases, adhesion molecules, and ECM3 components. A delicate balance between positive and negative regulatory signals controls each step of angiogenesis (6, 7). VEGF and bFGF are two of the best characterized angiogenesis agonists (8). Among the several types of angiogenesis inhibitors are proteolytic fragments of Type XVIII collagen and plasminogen, named endostatin and angiostatin, respectively (9 –11). Systemic administration of recombinant endostatin and angiostatin caused growth regression of a number of tumor types in murine xenograft models (10, 12). Moreover, in preclinical animal efficacy models, administration of endostatin can lead to stable tumor dormancy without the development of acquired drug resistance (12). Currently, endostatin and angiostatin are being evaluated in

clinical trials for a variety of human malignancies. Other antiangiogenic factors that have been evaluated preclinically include platelet factor-4 (13), thrombospondin-2 (14), the Mr 16,000 NH2-terminal fragment of prolactin (15), and maspin (16). The recent discovery of angiopoietins has provided additional insight into the molecular and cellular mechanisms of blood vessel formation (4, 17). Angiopoietins (Ang 1 and Ang 2) are Mr ⬃70,000 proteins that share considerable sequence homology. Each protein consists of a signal peptide, an NH2-terminal coiled-coil domain, a short linker peptide region, and a COOH-terminal FD. The coiled-coil region is responsible for dimerization of angiopoietin, and the FD binds to Tie2 receptors. Both Ang 1 and Ang 2 form dimers and oligomers (18, 19). In vivo analysis by targeted gene disruption reveals that Ang 1 recruits and sustains periendothelial support cells (4, 18, 20), whereas Ang 2 disrupts blood vessel formation in the developing embryo by antagonizing the effect of Ang 1 on the Tie2 receptor (17). Later, Ang 4 was shown to be a third protein capable of binding to the Tie2 receptor (21). Three additional proteins (ARP1, ARP2, and CDT6) with similarity to angiopoietins have also been discovered (22–25) that do not bind to the Tie2 or Tie1 receptor and do not possess a specific cysteine motif that is characteristic of angiopoietins. The molecular mechanism through which these proteins regulate angiogenesis has not been fully elucidated (21). Here, we used a homology-based gene mining approach to isolate an angiopoietin-related cDNA identical to ARP-1 that we designate as angioarrestin, because of its antiangiogenic properties both in vitro and in vivo. The mRNA expression profile of angioarrestin was strikingly down-regulated in tumor tissue in comparison to matched NAT. To functionally characterize angioarrestin, the cDNA was expressed in a eukaryotic expression system, and recombinant protein was purified. Systematic analysis in a variety of cell-based angiogenesis assays revealed for the first time that angioarrestin inhibited a number of angiogenic processes, including proliferation, migration, tube formation, and endothelial cell adhesion. These effects were endothelial cell specific and not observed in nonendothelial cells. In additional studies, nude mice were injected with HT1080 tumor cells engineered to ectopically express angioarrestin. Compared with control, these mice had a marked reduction in number and size of tumor nodules. In summary, these results demonstrate a novel function for angioarrestin as an inhibitor of angiogenesis and indicate its potential utility as a cancer therapeutic.

MATERIALS AND METHODS

Reagents and Cells. C-PAEs, 786-0, HT1080, NIH 3T3, CCD 1070, and HEK293T cells were obtained from American Type Culture Collection (Rockville, MD). HUVEC, HMVEC-d (Cascade Biologics, Portland, OR), and bovine aortic endothelial cells (Clonetics, Walkerville, MD) were used at passages 3–9. The eukaryotic expression vector pCEP4 was obtained from Invitrogen (San Diego, CA). Restriction enzymes and DNA polymerase were obtained from Boehringer Mannheim (Indianapolis, IN). Isolation and Cloning of Human Angioarrestin. A technology, SeqCalling, that provides DNA sequence information for the coding regions of expressed genes was used to identify different fragments encoding angioarrestin cDNA. Briefly, a partial angioarrestin cDNA was identified using a 3834

Received 2/12/02; accepted 4/29/02. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 M. D. and W. J. L. contributed equally to this manuscript. 2 To whom requests for reprints should be addressed, at CuraGen Corp., 322 East Main Street, Branford, CT 06405. Phone: (203) 871-4331; Fax: (203) 315-3301. 3 The abbreviations used are: ECM, extracellular matrix; VEGF, vascular endothelial growth factor; C-PAE, calf pulmonary arterial endothelial cell; NAT, nontumor adjacent tissue; FD, fibrinogen homology domain; FBS, fetal bovine serum; HUVEC, human umbilical vein endothelial cell; bFGF, basic fibroblast growth factor; HMVEC-d, human microvascular endothelial cells from dermis; RTQ, real-time quantitative; mAb, monoclonal antibody.

ANGIOARRESTIN IS AN ANTIANGIOGENIC PROTEIN

homology search of CuraGen’s SeqCalling database. The SeqCalling database was developed using a modification of the GeneCalling technology (26), except that digested fragments were size fractionated by high-performance liquid chromatography, PCR amplified, cloned, and then sequenced. All sequences were then collected into a database and collapsed into single gene assemblies based on sequence identity. Angioarrestin cDNA ends were generated by rapid amplification of cDNA ends using a cDNA pool from 28 separate tissues. On the basis of the predicted sequence, oligonucleotide primers were designed to amplify full-length angioarrestin cDNA by PCR. The primers used were as follows: forward primer, 5⬘-CCA CCA TGA AGA CTT TTA CTT TTA CCT GGA CCC-3⬘; reverse primer, 5⬘-GTC AAT AGG CTT GAT CAT CAT CTG AAC TG-3⬘. The PCR mix contained 4 ␮g of human heart cDNA, 75 pmol of primers, 5 ␮mol of deoxynucleotide triphosphates, 1 unit of Fidelity expand polymerase, and 5 ␮l of Fidelity expand buffer (Boehringer Mannheim). A touchdown PCR was used as per standard protocol (27). A single PCR product of 1.4 kb was obtained and cloned into pcDNA3.1V5his Topo vector (Invitrogen). A Hind III-Pme I fragment from pcDNA3.1/angioarrestin was ligated into pCEP4, and the resulting vector was named as pCEP4/angioarrestin. This vector contains an in-frame V5 and His6 tag at the 3⬘-terminus of the coding region. Expression, Purification, and Biochemical Characterization of Recombinant Angioarrestin. The pCEP4/angioarrestin vector was transfected into HEK293T cells using Lipofectamine Plus (Life Technologies, Inc., Rockville, MD). The conditioned medium was collected from transfected cells after 72 h, pooled, and loaded onto a Ni2⫹ affinity column (Qiagen, Valencia, CA). The column was washed with PBS (pH 7.4), containing 500 mM NaCl, followed by

the same buffer containing 5 mM imidazole. The bound protein was eluted with PBS (pH 7.4), containing 500 mM imidazole, pooled, and dialyzed overnight in PBS (pH 7.4). The protein was further purified by a second round of purification over a Ni2⫹ affinity column and dialyzed against PBS (pH 7.4). Protein concentration was determined using the Bradford reagent (Bio-Rad, Hercules, CA). Molarity was calculated using the molecular weight of the dimer. Protein purity was assessed by Silver staining after SDS-PAGE analysis using a 4 –15% Tris/glycine gradient gel. Western blot analysis was performed with anti-V5 tag mAb (1:5000; Ref. 28) conjugated to horseradish peroxidase, followed by enhanced chemiluminescence (Amersham Pharmacia Biotech, Piscataway, NJ). Angioarrestin was tested for endotoxin using the Gel Clot method (Cape Cod Associates, Cape Cod, MA) and found to have ⬍10 EU/mg protein. Purified protein was used in all in vitro assays. RTQ-PCR Expression Analysis. Human normal, tumor, or matched normal adjacent tumor tissues were obtained from National Cancer Institute’s Cooperative Human Tissue Network or The National Disease Research Institute. Tumor cell lines were derived from cultured cells obtained from American Type Culture Collection. RNA samples were prepared from the tissues using TRIzol reagent according to the manufacturer’s instructions (Life Technologies, Inc.). RTQ-PCR (29) was performed on an ABI Prism 7700 Sequence Detection System (PE Applied Biosystems, Foster City, CA) using TaqMan reagent (PE Applied Biosystems). RNAs were normalized using human ␤ actin and glyceraldehyde-3-phosphate dehydrogenase according to the manufacturer’s instructions. Equal quantity of normalized RNA was used as a template in PCR reactions with angioarrestin-specific reagents to obtain threshold cycle values. The following angioarrestin-specific primers and probe

Fig. 1. Genomic organization of the angioarrestin locus. Two exon/intron boundaries are indicated by vertical lines. The gene was mapped to chromosome 1 (1q24.2). The initiation and stop codons are in red. A line below the angioarrestin sequence denotes a putative signal sequence for protein secretion as determined by the PSORT computer program (40). Two ⴱ indicate the potential Nglycosylation sites as predicted by the PROSITE program (41). The coiled-coil region and FD protein region are delineated by orange and blue arrows, respectively.

3835


Fig. 2. Analysis of angioarrestin transcript using RTQ-PCR. RTQ-PCR analysis was performed as described in “Materials and Methods” using angioarrestin-specific probes. A representative expression profile in normal (A) and tumor tissue samples (B) is depicted with black and red bars, respectively. RNA samples from matched NAT and tumor were normalized as described. Transcript expression is presented as a percentage of the sample exhibiting the highest level of expression and varied by ⬍15% among individual experiments.

were used: forward primer 5⬘-GCT GGG AGG TAA CGA GAT TCA G-3⬘, probe primer Fam 5⬘-ATC CAG GTT ATC CCA GAG ATT TAA TGC CAC C-3⬘ Tamra, and reverse primer 5⬘-TGG GAG AAG TTG CCA GAT CAG-3⬘. Endothelial Cell BrdUrd Incorporation Assay. The effect of angioarrestin on proliferation was assessed using HUVEC, HMVEC-d, and C-PAE. The cells were plated in 96-well flat-bottomed plates precoated with Attachment Factor (Cascade Biologics) at 3 ⫻ 104 cells/well in 100 ␮l of Medium 200 (Cascade Biologics) containing 0.5% FBS. After 24 h of starvation at 37°C, the cells were washed two times with serum-free medium and then fed with fresh medium containing 1% FBS with VEGF165 and bFGF (10 ng/ml; R&D Systems, Minneapolis, MN) with and without angioarrestin protein. The BrdU assay was performed according to the manufacturer’s specification (Roche Molecular Biochemicals, Indianapolis, IN). Migration Assay. To determine the ability of recombinant angioarrestin to block migration of HUVEC and HMVEC-d toward VEGF165, 24-well transwell (BD Biosciences, Bedford, MA) migration chambers having an 8-␮m pore size were used. The transwells were coated with 10 ␮g/ml Type I collagen (BD Biosciences) from rat tail for 1 h at 37°C. After washing with PBS, the wells were seeded with HUVEC suspended at 2 ⫻ 105 cells/ml in Medium 200 containing 1% BSA (Sigma Chemicals, St. Louis, MO). The bottom chambers (600 ␮l) were filled with Medium 200 containing 1% FBS supplemented with 10 ng/ml recombinant VEGF165. The top chamber was seeded with 4 ⫻ 104

cells/well in 200 ␮l containing different angioarrestin concentrations. Cells were allowed to migrate for 4 h at 37°C. After incubation, cells on the top surface of the membrane (nonmigrated cells) were scraped with a cotton swab. Cells on the bottom side of the membrane (migrated cells) were stained with 0.2% Crystal Violet dye (Fisher Scientific, Springfield, NJ) in 70% ethanol for 30 min. The cells were then destained in PBS (pH 7.4), and the membrane was left to air dry at room temperature. Migrated cells were counted using a Zeiss Axiovert 100 inverted microscope. Three independent areas per filter were counted, and the mean number of migrated cells was calculated. RGD control peptide (Invitrogen; Cat. No. 12135-018) having the sequence GRGDSP was used as a positive control. Each experiment was done in duplicate and repeated four times. Endothelial Cell Tube Formation Assay. Endothelial tube formation assays using HUVEC and HMVEC-d were performed as described previously (30). Briefly, HUVECs were trypsinized, washed with PBS (pH 7.4), counted, and resuspended in Medium 200 containing 2% FBS supplemented with 10 ng/ml VEGF165 in the presence or absence of different concentrations of angioarrestin. Cells (25,000) were seeded per well onto solidified Matrigel (BD Biosciences) in 96-well plates. The cells were incubated at 37°C for 18 –20 h. Tubular networks were visualized with an inverted phase contrast Zeiss Axiovert 100 inverted microscope. Each dilution was done in duplicate, and the assay was repeated three times. The tubular network patterns were

3836


Fig. 3. Expression and purification of recombinant angioarrestin. HEK293T cells were transiently transfected with pCEP4/angioarrestin. In A, the recombinant protein secreted into the medium was harvested and purified as described in “Materials and Methods.” The purified protein was analyzed by silver staining using 4 –15% SDS-PAGE. Purified protein (100 ng) was analyzed in the presence (Lane 2) or absence (Lane 3) of 5 mM DTT. Lane 1, molecular weight standards (Amersham Pharmacia Biotech). B, Western blot analysis of recombinant angioarrestin expressed from HEK293T-transfected cells. Purified protein (100 ng) was resolved by 4 –15% SDS-PAGE in the presence (Lane 1) or absence (Lane 2) of 5 mM DTT and subjected to Western blot analysis using anti-V5 tag mAb.

and resuspended at 1.5 ⫻ 105 cells/ml in serum-free medium containing 1% BSA. The cells were then mixed with different concentrations of angioarrestin in 100-␮l volumes containing 2 ⫻ 104 cells/treatment for 15 min at room temperature. After incubation, the cell suspension was then added to each well, and the plates were incubated at 37°C for 45 min in 5% CO2. At the end of the incubation period, unattached cells were removed by washing three times with serum-free medium, and attached cells were counted using a Cytofluor 4000 fluorometer (PE Applied Biosystems). The number of attached cells was represented as a percentage of endothelial cell adhesion, as determined by the ratio of attached cells in the presence or absence of factor. Inhibition of Tumor Nodule Formation. The plasmid containing angioarrestin (pcDNA3/angioarrestin) was used to transfect HT1080 cells with Lipofectamine Plus reagent according to the manufacturer’s protocol. Cells were supplemented with 10% FBS 5 h after transfection. The transfectants were grown in DMEM (Life Technologies, Inc.) containing 10% FBS supplemented with 500 ␮g/ml G-418 (Life Technologies, Inc.), and angioarrestin expression was confirmed by Western blot analysis. As control, HT1080 cells were transfected with the appropriate empty vector and selected as described above. After 2 weeks of culture, pools of transfected cells were subcultured, washed with PBS (pH 7.4), and used as described below. Athymic nude mice (five per group) were injected i.v. with 0.5 ⫻ 106 cells of HT1080 control pool or HT1080 angioarrestin in a volume of 200 ␮l. Four weeks after injection, the animals were sacrificed, and lung tissues were harvested and infused with formalin. The entire lung tissues were embedded, and step sections (400 ␮m) were prepared and stained with H&E using a standard protocol (O’Reilly et al., 1994a). The tumor number and area were assessed by morphometric analysis on three stepwise sections from each animal. For calculating mean tumor area (␮m2), the length and width of the tumor were microscopically measured using an ocular ␮m.

RESULTS Cloning and Analysis of Human Angioarrestin. The process of gene mining was used to identify a cDNA related to angiopoietin family members. The gene sequence (Fig. 1) encodes a protein of 491 amino acids that was found to be 100% identical to ARP-1. Because of its antiangiogenic properties in both the in vitro and in vivo analysis detailed below, we propose changing the name of ARP-1 to angioarrestin. Angioarrestin has 58.6, 44.7, 31.9, 31.0, and 29.4% overall amino acid identity to ARP2, CDT6, Ang 1, Ang 2, and Ang 4,

Fig. 4. Effect of angioarrestin on endothelial cell BrdUrd incorporation. Purified angioarrestin was tested for its ability to inhibit endothelial cell DNA synthesis using a BrdUrd incorporation assay. Cells were synchronized by serum starvation overnight in minimal medium and stimulated with VEGF and bFGF at 10 ng/ml in the presence of 1% FBS. Different concentrations of purified angioarrestin were added along with growth factors. Each value is a mean of duplicate cultures from a representative experiment. BrdUrd incorporation with growth factors alone was considered as maximum incorporation, and the percentage of inhibition was calculated. The error bars represent the percentage of inhibition ⫹/⫺SD. F, HUVEC; Œ, HMVEC-d; f, C-PAE.

captured using a Kodak digital camera and analyzed using a PhotoShop 5.5 program. Fig. 5. Effect of angioarrestin on endothelial cell migration. A total of 4 ⫻ 104 Endothelial Cell Adhesion Assay. Untreated 96-well flat-bottomed tissue cells/well was seeded into the top chamber of a Boyden transwell apparatus. The culture plates (Fisher Scientific) were used in the cell adhesion assay. The chemotactant VEGF (10 ng/ml) was added to the bottom chamber in the presence of 1% plates were coated with 10 ␮g/ml different ECM proteins (Type I collagen, FBS. Endothelial cells were allowed to migrate for 4 h at 37°C. The number of cells Type IV collagen, fibronectin, vitronectin, laminin, and Matrigel) overnight at migrating in the presence or absence of angioarrestin was counted in three independent 4°C. The remaining protein binding sites were blocked with 1% BSA in PBS fields. The number of cells that migrated in the presence of 1% FBS with VEGF (10 (pH 7.4) for 2 h at 37°C. HUVECs were grown to subconfluence (70 – 80%) in ng/ml) was considered maximum migration, and the percentage of inhibition was calculated. The error bars represent the percentage of inhibition ⫹/⫺SD. Each value is a mean Medium 200. The cells were labeled with Calcein-AM fluorophore (Molecular from a representative experiment. F, HUVEC; Œ, HMVEC-d. The experiment was Probes, Eugene, OR) as described (31). The cells were trypsinized, washed, repeated three times with similar results. 3837


Fig. 6. Effect of angioarrestin on endothelial cell tube formation. Well-formed tubes were observed in 2% FBS containing VEGF, whereas recombinant angioarrestin inhibited VEGF-induced tube formation of HUVECs on Matrigel in a dose-dependent manner. The tubular network formation was visualized in an inverted Axiovert 100 microscope. A representative tube formation from an independent field is shown. The experiment was repeated three times with similar results.

respectively. Like all angiopoietins or related proteins, angioarrestin has a signal peptide, coiled-coil region, and a FD. However, angioarrestin has only one of three closely spaced conserved cysteines found in the FD of angiopoietin family members (21). Genomic DNA analysis showed that angioarrestin is localized to human chromosome 1 (1q24.2) and contains four exons. Angioarrestin mRNA Expression in Human Normal and Tumor Tissues. The mRNA expression profile of angioarrestin was examined using RTQ-PCR (29). The primer/probe used was designed in such a way as to amplify only angioarrestin and not other known angiopoietin family members. The result from a representative experiment demonstrates that angioarrestin has a broad spectrum of expression in normal tissues (Fig. 2A) and is most abundant in adrenal gland, placenta, and small intestine. Interestingly, a unique expression profile of angioarrestin was noted in tumor and matched NAT (Fig. 2B). These data demonstrate that tumor tissue from prostate, lung, kidney, thyroid, and bladder cancer showed dramatically less expression when compared with normal adjacent tissue. Expression and Characterization of Recombinant Angioarrestin. To characterize the biological function of angioarrestin, recombinant protein was expressed, purified from the conditioned medium of HEK293T cells, and analyzed by SDS-PAGE. Angioarrestin migrated as a doublet of Mr ⬃60,000 and 64,000 under reducing conditions (Fig. 3A, Lane 2). In addition, a minor species was observed at Mr ⬃120,000. Under nonreducing conditions (Fig. 3A, Lane 3), a predominant protein of Mr ⬃120,000 was observed along with minor higher molecular weight products. Proteins of corresponding molecular weight were also detected by Western blot analysis with a mAb that detects the V5 tag engineered in the COOH terminus of angio-

arrestin (Fig. 3B, Lanes 1 and 2). These results suggest that angioarrestin exists predominantly as a disulfide-linked dimer. Angioarrestin Inhibits Endothelial Cell BrdUrd Incorporation. The ability of angioarrestin to affect the growth factor-induced proliferation of HUVEC, HMVEC-d, and C-PAE was examined in an in vitro BrdUrd incorporation assay. As shown in Fig. 4, angioarrestin inhibited VEGF/bFGF-induced BrdUrd incorporation in a dosedependent manner. The IC50 was 75, 90, and 250 ng/ml (0.6, 0.8, and 2 nM) for HUVEC, HMVEC-d, and C-PAE, respectively. Angioarrestin did not inhibit the bFGF-stimulated proliferation of nonendothelial cells, such as NIH 3T3 and CCD1070 (data not shown). These results suggest that angioarrestin inhibits the proliferation of various endothelial cells derived from different vessel types. Angioarrestin Inhibits Endothelial Cell Migration. We assessed the chemotactic response of HUVEC and HMVEC-d to increasing concentrations of angioarrestin in the presence of VEGF using a Boyden chemotaxis chamber assay. Angioarrestin significantly inhibited the VEGF-induced migration of endothelial cells in a dosedependent manner (Fig. 5). Inhibition (68%) observed with the highest dose (100 ng/ml, 0.8 nM) tested was comparable with that seen with a control RGD peptide (50 and 70% inhibition at 1 and 10 ␮g/ml, respectively; data not shown). No effect was seen on nonendothelial cell lines (NIH 3T3 and CCD1070 treated with bFGF) at angioarrestin concentrations ⱕ10 ␮g/ml (data not shown). These results demonstrate that angioarrestin inhibits the VEGF-mediated migration of endothelial cells. Angioarrestin Inhibits Endothelial Cell Tube Formation. The endothelial cell tube formation assay comprehensively measures important steps in the angiogenic cascade that occur during the forma-

3838


addition, angioarrestin also inhibited adhesion of HMVEC-d with a similar inhibitory profile (Fig. 7B). Additional experiments demonstrated that angioarrestin did not induce apoptosis in adherent HUVECs over a 12–15 h treatment (data not shown). Together, these findings show that angioarrestin inhibits endothelial cell adhesion to different ECM proteins. Ectopic Expression of Angioarrestin Inhibits Tumor Nodule Formation. Because tumor growth beyond a few mm2 in diameter is dependent on new blood vessel formation (2), we tested whether ectopic expression of angioarrestin in HT1080 fibrosarcoma cells would affect tumor formation in vivo. Angioarrestin protein was secreted in the conditioned medium of angioarrestin-HT1080 transfectants, and it was demonstrated that the in vitro growth rates of the cells were comparable with empty vector-transfected HT1080 (data not shown). However, when angioarrestin-transfected HT1080 cells were injected i.v. into nude mice and pulmonary tumor nodules were measured, a statistically significant (P ⬍ 0.05) reduction in the overall mean tumor number relative to control HT1080 cells was observed (Fig. 8A). In addition, tumors from angioarrestin-transfected HT1080 cells were significantly smaller (P ⬍ 0.05) when compared with control (Fig. 8B). As shown in Fig. 8C, tumor tissue from angioarrestin-transfected HT1080 cells had fewer blood vessels (arrows) that were markedly smaller and less mature than control. Thus, the ectopic expression of angioarrestin in HT1080 fibrosarcoma cells inhibited tumor nodule formation and decreased tumor size in comparison to control HT1080 transfectants, an effect that was likely attributable to an inhibition of tumor-associated angiogenesis. DISCUSSION

Fig. 7. Effect of angioarrestin on endothelial cell adhesion to different ECM components. HUVEC (A) or HMVEC-d (B) were labeled with Calcein AM (Molecular Probes) at 5 ␮M final concentration in serum-free medium for 30 min at 37°C. Typically, ⬎90% of cells incorporated the Calcein AM fluorescence dye. Ten ␮g/ml purified fibronectin (Œ), Matrigel (f), or vitronectin (F) were used to coat untreated tissue culture plates. Cells were allowed to adhere, and the number of cells bound in a solid phase binding assay was quantified using a CytoFluor 4000 fluorometer as described in “Materials and Methods.” Each value represents the percentage of inhibition of cell adhesion from duplicate cultures. Bars, ⫹/⫺SD.

tion of new blood vessels (3). HUVECs cultured on Matrigel, a solid gel of mouse basement membrane protein, rapidly align and form hollow tube-like structures. As expected, HUVECs treated with 2% FBS and VEGF showed organized endothelial cell tube formation. Angioarrestin at 10 ␮g/ml (80 nM) selectively inhibited endothelial tube formation (Fig. 6). A dose-dependent inhibition of tubular network formation was observed in the range of 1–10 ␮g/ml (8 – 80 nM) angioarrestin. A similar dose response was seen when HMVEC-d were used for tube formation (data not shown). Thus, angioarrestin inhibits tube formation in large and small vessel endothelial cells. Angioarrestin Inhibits Endothelial Cell Adhesion to Different ECM Proteins. To address whether angioarrestin inhibits endothelial cell processes in part through effects on cell adhesion, HUVECs were plated onto wells of untreated or ECM-treated tissue culture plates. HUVECs adhered to all of the ECM proteins, with maximum binding seen to fibronectin-coated plates. In the presence of different concentrations of angioarrestin, there was a dose-dependent inhibition of cell adhesion to ECM-coated plates (Fig. 7A). At 10 ␮g/ml (80 nM), there was a 58% reduction of HUVECs in binding to fibronectin-treated plates. Similar inhibitory profiles were also observed with collagen (Type I and IV) and laminin ECM proteins (data not shown). In

In the present study, we isolated a cDNA encoding an angiopoietinrelated protein and characterized its novel antiangiogenic properties. Recombinant angioarrestin inhibited a number of angiogenic processes, including endothelial cell proliferation, migration, tube formation, and adhesion. The observed effects were observed in endothelial cells and not in nonendothelial cell types. In vivo, ectopic expression of angioarrestin in HT1080 tumor cells resulted in a significant reduction of tumor nodule number and size. Using a sensitive RTQ-PCR assay, angioarrestin mRNA was found to be lower in tumor tissues relative to matched NAT. Ample experimental and clinical correlates exist for suppression of antiangiogenic factors during tumor progression. This concept was illustrated when reconstitution of a wild-type p53 tumor suppressor in a BT549 breast carcinoma cell line resulted in secretion of an angiogenesis inhibitor (32). Down-regulation of the retinoic acid receptor ␤2, maspin, and E-cadherin genes have correlated with cancer progression and increased tumor invasion or metastasis (16, 33, 34). Invasiveness, vessel density, and poor clinical outcome have been reported to statistically correlate with down-regulation of thrombospondin-1 in advanced cervical cancer (35). Likewise, a significant stepwise decrease in maspin expression parallels breast cancer progression from ductal carcinoma in situ, to locally invasive cancer, and finally to lymph node metastasis (36). Decreased expression of maspin in prostate cancer has also been reported to correlate inversely with the development of local recurrence or systemic tumor progression (37). Additional experiments will be needed to conclusively demonstrate a tumor suppressive role for angioarrestin and whether the decrease in angioarrestin expression results in tumor formation. Angioarrestin inhibits angiogenic processes through effects on both induction and resolution phases of the angiogenic cascade. In the induction phase of angiogenesis, angioarrestin inhibited BrdUrd incorporation, migration, and endothelial cell adhesion to ECM proteins. The activity of angioarrestin (IC50 ⫽ 0.8 nM) to block endothe-

3839


Fig. 8. Effect of ectopic expression of angioarrestin on mean tumor number and size. HT1080 cells were transfected with a plasmid containing the coding region of angioarrestin and injected (0.5 ⫻ 106 cells) i.v. into athymic nude mice as described in “Material and Methods.” Four weeks after injection, the animals were sacrificed and evaluated for tumor nodule growth. The mean tumor nodule number (A) was counted using a microscope. Bars, ⫹/⫺SE. The tumor size was measured from individual animals using an ocular ␮m. Tumor area (B) was expressed as length ⫻ width (␮m2). Each point represents the average of mean tumor number at three different levels in the tumor section. Bars, ⫹/⫺SE. f, HT1080 control; f, HT1080-Angioarrestin. H&E-stained tumor tissues (C) from HT1080 control and HT1080 angioarrestin mice were examined in high-power magnification (⫻400). Animals injected with control cells had numerous easily identifiable blood vessels (arrows) containing erythrocytes. In contrast, tumor tissues from animals injected with HT1080 angioarrestin had fewer vessels.

lial cell BrdUrd incorporation compares favorably with other antiangiogenic molecules, including angiostatin (IC50 ⫽ 140 nM) and endostatin (IC50 ⫽ 20 nM; Refs. 9 and 10). Angioarrestin was also more potent (IC50 ⫽ 0.3 nM) in the migration assay when compared with published data for endostatin (IC50 ⫽ 3 nM; Refs. 9 and 10) and angiostatin (IC50 ⫽ 10 nM; Ref. 11). With respect to the resolution phase, angioarrestin inhibited endothelial cell tube formation at ⬃80 nM, whereas previous reports (9, 10) have shown that angiostatin was active in this assay at 200 nM. The potent in vitro activity of angioarrestin across all phases of the angiogenic process was confirmed in vivo. Ectopic expression of angioarrestin in HT1080 cells suppressed tumor growth when compared with control. This activity was comparable with endostatin, and angiostatin tested in similar models (38, 39). Additional studies will be required to determine whether angioarrestin can suppress tumor formation when administered systemically or through gene therapy. Two additional molecules related to angioarrestin, ARP2 and

CDT6, with significant homology to angioarrestin have also been described (23, 25). The possibility exists that these three molecules comprise a family of antiangiogenic proteins. Although this is the first report demonstrating that angioarrestin/ARP1 possesses antiangiogenic properties, evidence suggests that CDT6 may also inhibit angiogenesis (24). Surprisingly, the only activity (22) observed for ARP1 and ARP2 was a modest endothelial cell-sprouting activity. It is difficult to rationalize the observed effect on sprouting considering the potent antiangiogenic activity that is demonstrated by angioarrestin in the present study. Interestingly, none of the three angiopoietin-related molecules bind to Tie2, the receptor through which angiopoietins function, nor do they interact with the closely related Tie1 receptor (23, 24). This raises the possibility that angioarrestin, ARP2, and CDT6 all interact with unidentified receptor(s) on endothelial cells. The mechanism by which angioarrestin exerts its antiangiogenic properties is currently under investigation.

3840


ACKNOWLEDGMENTS We thank Cooperative Human Tissue Network and National Disease Research Institute for their support of high-quality tissues and pathological information. We thank Catherine E. Burgess for gene mining; CuraGen’s SeqCalling and full-length cloning groups for cDNA isolation; Alison Bendele for performing the nude mouse tumor assays; Beth Rittman, Jennifer Zmijewski, Xiaohong Liu, Christie Peperato, Lisa Deegler, and Shanna Davis for technical assistance; and Medhi Mesri and Craig Hackett for helpful discussions.

REFERENCES 1. Folkman, J. Tumor angiogenesis. Adv. Cancer Res., 19: 331–358, 1974. 2. Folkman, J. Tumor angiogenesis: therapeutic implications. N. Engl. J. Med., 285: 1182–1186, 1971. 3. Madri, J. A. Extracellular matrix modulation of vascular cell behaviour. Transpl. Immunol., 5: 179 –183, 1997. 4. Davis, S., Aldrich, T. H., Jones, P. F., Acheson, A., Compton, D. L., Jain, V., Ryan, T. E., Bruno, J., Radziejewski, C., Maisonpierre, P. C., and Yancopoulos, G. D. Isolation of angiopoietin-1, a ligand for the TIE2 receptor, by secretion-trap expression cloning. Cell, 87: 1161–1169, 1996. 5. Davis, S., and Yancopoulos, G. D. The angiopoietins: Yin and Yang in angiogenesis. Curr. Top. Microbiol. Immunol., 237: 173–185, 1999. 6. Hanahan, D. A. Flanking attack on cancer. Nat. Med., 4: 13–14, 1998. 7. Hanahan, D., and Folkman, J. Patterns and emerging mechanisms of the angiogenic switch during tumorigenesis. Cell, 86: 353–364, 1996. 8. Kuwano, M., Fukushi, J., Okamoto, M., Nishie, A., Goto, H., Ishibashi, T., and Ono, M. Angiogenesis factors. Intern. Med., 40: 565–572, 2001. 9. O’Reilly, M. S., Boehm, T., Shing, Y., Fukai, N., Vasios, G., Lane, W. S., Flynn, E., Birkhead, J. R., Olsen, B. R., and Folkman, J. Endostatin: an endogenous inhibitor of angiogenesis and tumor growth. Cell, 88: 277–285, 1997. 10. O’Reilly, M. S., Holmgren, L., Shing, Y., Chen, C., Rosenthal, R. A., Cao, Y., Moses, M., Lane, W. S., Sage, E. H., and Folkman, J. Angiostatin: a circulating endothelial cell inhibitor that suppresses angiogenesis and tumor growth. Cold Spring Harb. Symp. Quant. Biol., 59: 471– 482, 1994. 11. O’Reilly, M. S., Holmgren, L., Shing, Y., Chen, C., Rosenthal, R. A., Moses, M., Lane, W. S., Cao, Y., Sage, E. H., and Folkman, J. Angiostatin: a novel angiogenesis inhibitor that mediates the suppression of metastases by a Lewis lung carcinoma. Cell, 79: 315–328, 1994. 12. Boehm, T., Folkman, J., Browder, T., and O’Reilly, M. S. Antiangiogenic therapy of experimental cancer does not induce acquired drug resistance. Nature, 390: 404 – 407, 1997. 13. Jouan, V., Canron, X., Alemany, M., Caen, J. P., Quentin, G., Plouet, J., and Bikfalvi, A. Inhibition of in vitro angiogenesis by platelet factor-4-derived peptides and mechanism of action. Blood, 94: 984 –993, 1999. 14. Streit, M., Riccardi, L., Velasco, P., Brown, L. F., Hawighorst, T., Bornstein, P., and Detmar, M. Thrombospondin-2: a potent endogenous inhibitor of tumor growth and angiogenesis. Proc. Natl. Acad. Sci. USA, 96: 14888 –14893, 1999. 15. Ferrara, N., Clapp, C., and Weiner, R. The 16K fragment of prolactin specifically inhibits basal or fibroblast growth factor stimulated growth of capillary endothelial cells. Endocrinology, 129: 896 –900, 1991. 16. Zhang, M., Volpert, O., Shi, Y. H., and Bouck, N. Maspin is an angiogenesis inhibitor. Nat. Med., 6: 196 –199, 2000. 17. Maisonpierre, P. C., Suri, C., Jones, P. F., Bartunkova, S., Wiegand, S. J., Radziejewski, C., Compton, D., McClain, J., Aldrich, T. H., Papadopoulos, N., Daly, T. J., Davis, S., Sato, T. N., and Yancopoulos, G. D. Angiopoietin-2, a natural antagonist for Tie2 that disrupts in vivo angiogenesis. Science (Wash. DC), 277: 55– 60, 1997. 18. Suri, C., Jones, P. F., Patan, S., Bartunkova, S., Maisonpierre, P. C., Davis, S., Sato, T. N., and Yancopoulos, G. D. Requisite role of angiopoietin-1, a ligand for the TIE2 receptor, during embryonic angiogenesis. Cell, 87: 1171–1180, 1996. 19. Procopio, W. N., Pelavin, P. I., Lee, W. M., and Yeilding, N. M. Angiopoietin-1 and -2 coiled coil domains mediate distinct homo-oligomerization patterns, but fibrinogen-like domains mediate ligand activity. J. Biol. Chem., 274: 30196 –30201, 1999.

20. Sato, A., Iwama, A., Takakura, N., Nishio, H., Yancopoulos, G. D., and Suda, T. Characterization of TEK receptor tyrosine kinase and its ligands. Angiopoietins, in human hematopoietic progenitor cells. Int. Immunol., 10: 1217–1227, 1998. 21. Valenzuela, D. M., Griffiths, J. A., Rojas, J., Aldrich, T. H., Jones, P. F., Zhou, H., McClain, J., Copeland, N. G., Gilbert, D. J., Jenkins, N. A., Huang, T., Papadopoulos, N., Maisonpierre, P. C., Davis, S., and Yancopoulos, G. D. Angiopoietins 3 and 4: diverging gene counterparts in mice and humans. Proc. Natl. Acad. Sci. USA, 96: 1904 –1909, 1999. 22. Kim, I., Kwak, H. J., Ahn, J. E., So, J. N., Liu, M., Koh, K. N., and Koh, G. Y. Molecular cloning and characterization of a novel angiopoietin family protein, angiopoietin-3. FEBS Lett., 443: 353–356, 1999. 23. Kim, I., Moon, S. O., Koh, K. N., Kim, H., Uhm, C. S., Kwak, H. J., Kim, N. G., and Koh, G. Y. Molecular cloning, expression, and characterization of angiopoietinrelated protein. Angiopoietin-related protein induces endothelial cell sprouting. J. Biol. Chem., 274: 26523–26528, 1999. 24. Peek, R., Kammerer, R. A., Frank, S., Otte-Holler, I., and Westphal, J. R. The angiopoietinlike factor CDT6 is a putative morphogen for human cornea. J. Biol. Chem., 277: 686 – 693, 2002. 25. Peek, R., van Gelderen, B. E., Bruinenberg, M., and Kijlstra, A. Molecular cloning of a new angiopoietinlike factor from the human cornea. Investig. Ophthalmol. Vis. Sci., 39: 1782–1788, 1998. 26. Shimkets, R. A., Lowe, D. G., J. T., T., Sehl, P., Jin, H., Yang, R., Predki, P. F., Rothberg, B. E. G., Murtha, M. T., Roth, M. E., Shenoy, S. G., Windemuth, A., Simpson, J. W., Simons, J. F., Daley, M. P., Gold, S. A., McKenna, M. P., Hillan, K., Went, G. T., and Rothberg, J. M. Gene expression analysis by transcript profiling coupled to a gene database query. Nat. Biotechnol., 17: 798 – 803, 1999. 27. Sambrook, J., Fritsch, E. F., and Maniatis, T. (eds.). Molecular Cloning: A Laboratory Manual, Ed. 3, Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press, 1989. 28. Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A., and, and Struhl, K. (eds.). Current Protocols in Molecular Biology. New York: Greene Publishing Associates and Wiley-Interscience, 1994. 29. Heid, C. A., Stevens, J., Livak, K. J., and Williams, P. M. Realtime quantitative PCR. Genome Res., 6: 986 –994, 1996. 30. Kamphaus, G. D., Colorado, P. C., Panka, D. J., Hopfer, H., Ramchandran, R., Torre, A., Maeshima, Y., Mier, J. W., Sukhatme, V. P., and Kalluri, R. Canstatin, a novel matrix-derived inhibitor of angiogenesis and tumor growth. J. Biol. Chem., 275: 1209 –1215, 2000. 31. Senger, D. R., Claffey, K. P., Benes, J. E., Perruzzi, C. A., Sergiou, A. P., and Detmar, M. Angiogenesis promoted by vascular endothelial growth factor: regulation through ␣1␤1 and ␣2␤1 integrins. Proc. Natl. Acad. Sci. USA, 94: 13612–13617, 1997. 32. Volpert, O. V., Stellmach, V., and Bouck, N. The modulation of thrombospondin and other naturally occurring inhibitors of angiogenesis during tumor progression. Breast Cancer Res. Treat., 36: 119 –126, 1995. 33. Widschwendter, M., Berger, J., Muller, H. M., Zeimet, A. G., and Marth, C. Epigenetic downregulation of the retinoic acid receptor-␤2 gene in breast cancer. J. Mammary Gland Biol. Neoplasia, 6: 193–201, 2001. 34. Mareel, M., Bracke, M., and Van Roy, F. Cancer metastasis: negative regulation by an invasion-suppressor complex. Cancer Detect. Prev., 19: 451– 464, 1995. 35. Kodama, J., Hashimoto, I., Seki, N., Hongo, A., Yoshinouchi, M., Okuda, H., and Kudo, T. Thrombospondin-1 and -2 messenger RNA expression in epithelial ovarian tumor. Anticancer Res., 21: 2983–2987, 2001. 36. Maass, N., Teffner, M., Rosel, F., Pawaresch, R., Jonat, W., Nagasaki, K., and Rudolph, P. Decline in the expression of the serine proteinase inhibitor maspin is associated with tumour progression in ductal carcinomas of the breast. J. Pathol., 195: 321–326, 2001. 37. Machtens, S., Serth, J., Bokemeyer, C., Bathke, W., Minssen, A., Kollmannsberger, C., Hartmann, J., Knuchel, R., Kondo, M., Jonas, U., and Kuczyk, M. Expression of the p53 and Maspin protein in primary prostate cancer: correlation with clinical features. Int. J. Cancer, 95: 337–342, 2001. 38. Cao, Y., O’Reilly, M. S., Marshall, B., Flynn, E., Ji, R. W., and Folkman, J. Expression of angiostatin cDNA in a murine fibrosarcoma suppresses primary tumor growth and produces long-term dormancy of metastases. J. Clin. Investig., 101: 1055–1063, 1998. 39. Yoon, S. S., Eto, H., Lin, C. M., Nakamura, H., Pawlik, T. M., Song, S. U., and Tanabe, K. K. Mouse endostatin inhibits the formation of lung and liver metastases. Cancer Res., 59: 6251– 6256, 1999. 40. Nakai, K., and Kanehisa, M. A knowledge base for predicting protein localization sites in eukaryotic cells. Genomics, 14: 897–911, 1992. 41. Bairoch, A., Bucher, P., and Hofmann, K. The PROSITE database, its status in 1997. Nucleic Acids Res., 25: 217–221, 1997.

3841