Inhibition of Angiogenesis by Growth Factor Receptor Bound Protein 2-Src Homology 2 Domain Binding Antagonists

Molecular Cancer Therapeutics 1289

Inhibition of angiogenesis by growth factor receptor bound protein 2-Src homology 2 domain binding antagonists

Jesus V. Soriano,1 Ningfei Liu,4 Yang Gao,2 characterized role of Grb2 in signaling cell cycle progres- Zhu-Jun Yao,2 Toshio Ishibashi,3 sion together with our present findings suggests that Charles Underhill,4 Terrence R. Burke, Jr.,2 Grb2-SH2 domain binding antagonists have the potential and Donald P. Bottaro1 to act as anticancer drugs that target both tumor and vascular cell compartments. [Mol Cancer Ther 2004;3(10): 1 2 Laboratories of Cellular and Molecular Biology and Medicinal 1289–99] Chemistry, National Cancer Institute, NIH, Bethesda, Maryland; 3Department of Otolaryngology, Social Insurance Central General Hospital, Tokyo, Japan; and 4Department of Cell Biology, Georgetown University, Washington, District of Columbia Introduction Angiogenesis, the sprouting of new capillaries from existing blood vessels, consists of a complex cascade of events in- Abstract cluding endothelial cell–mediated degradation and inva- Growth factor receptor bound protein 2 (Grb2) is an sion of the extracellular matrix, endothelial cell migration, intracellular adaptor protein that participates in the signal proliferation, differentiation, basement membrane deposi- transduction cascades of several angiogenic factors, tion, and the organization of endothelial cords into capil- including hepatocyte growth factor, basic fibroblast lary structures. This sequence of events is regulated by growth factor, and vascular endothelial growth factor. soluble angiogenic growth factors such as fibroblast growth We described previously the potent blockade of hepato- factor (FGF)-2, vascular endothelial growth factor (VEGF), cyte growth factor–stimulated cell motility, matrix inva- and hepatocyte growth factor (HGF) operating through an sion, and epithelial tubulogenesis by synthetic Grb2-Src intricate network of tyrosine kinase receptor–mediated homology 2 (SH2) domain binding antagonists. Here, we signaling pathways (1–5). The intracellular signal trans- show that these binding antagonists block basic morpho- duction pathways activated by various receptor tyrosine genetic events required for angiogenesis, including hep- kinases share a subset of common events and mediators. atocyte growth factor–, vascular endothelial growth On ligand binding, receptor kinases undergo autophos- factor–, and basic fibroblast growth factor–stimulated phorylation on multiple cytoplasmic tyrosine residues. endothelial cell proliferation and migration, as well as Phosphorylated tyrosine residues on receptors or receptor phorbol 12-myristate 13-acetate–stimulated endothelial substrates then become binding sites for Src homology 2 cell migration and matrix invasion. The Grb2-SH2 domain (SH2) domain and/or phosphotyrosine binding domain– binding antagonists also impair angiogenesis in vitro,as containing effector proteins. Growth factor receptor bound shown by the inhibition of cord formation by macrovas- protein 2 (Grb2; refs. 6–8) is an adaptor protein that binds cular endothelial cells on Matrigel. We further show that a phosphotyrosyl proteins through its SH2 domain and representative compound inhibits angiogenesis in vivo as interacts through its two SH3 domains with proteins measured using a chick chorioallantoic membrane assay. containing proline-rich motifs (9–11). These results suggest that Grb2 is an important mediator Grb2 was first found to link growth factor receptor ty- of key proangiogenic events, with potential application rosine kinases to the Ras signal transduction pathway to pathologic conditions where neovascularization con- (6, 12) as well as to other intracellular signaling proteins tributes to disease progression. In particular, the well- including Gab family members and phosphatidylinositol 3 (P1-3)-kinase, leading ultimately to cell cycle progression. In the epidermal growth factor (EGF) signaling pathway, receptor autophosphorylation allows Grb2 binding via its SH2 domain, bringing SH3 domain–bound SOS-1 into the Received 5/19/04; revised 7/14/04; accepted 8/11/04. EGF receptor (EGFR) signaling complex. SOS-1 then fa- Grant support: Swiss National Research Foundation Fellowship for Advanced Scientists (J.V. Soriano). cilitates guanine nucleotide exchange and activation of The costs of publication of this article were defrayed in part by the Ras and subsequently a cascade of kinase activation steps payment of page charges. This article must therefore be hereby marked involving Raf and mitogen-activated protein kinases advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. (MAPKs), which in turn stimulate the transcription of key Note: J.V. Soriano is presently at American Type Culture Collection, cell cycle regulators (13). SH2 domain–mediated recogni- P.O. Box 1549, Manassas, VA 20108. tion by Grb2 is specific for the phosphopeptide motif Requests for reprints: Donald P. Bottaro, Urologic Oncology Branch, pYXNX (where pY is phosphotyrosine, N is aspara- Center for Cancer Research, National Cancer Institute, Building 10, gine, and X is any residue), which is found in several Room 2B47, 9000 Rockville Pike, Bethesda, MD 20892-1501. Phone: 301-402-6499; Fax: 301-402-0922. tyrosine kinases and their substrates. Thus, ligand-inde- E-mail: [email protected] pendent EGFR activation, such as growth hormone– Copyright C 2004 American Association for Cancer Research. induced EGFR tyrosine phosphorylation by JAK2, also

Mol Cancer Ther 2004;3(10). October 2004

leads to Grb2-mediated MAPK pathway activation and c- fos expression (14, 15). Similarly, mechanical stress leading to increased angiotensin II production and transactivation of EGFR and other intracellular kinases implicates Grb2 recruitment in cardiac hypertrophy and myocardial remodeling (16, 17). Many of the signaling pathways in which Grb2 functions are critical for vasculogenesis, angiogenesis, and lymphan- giogenesis. The VEGF receptors VEGFR2 and VEGFR3 bind Grb2 directly, and Grb2 is also recruited to receptors through its interaction with Shc (18–21). Similarly, Grb2 participates in the signaling pathways of FGF-2–, angio- poietin-1-, and HGF-stimulated vasculogenesis and angiogenesis via direct association with activated receptors and/ or through intermediaries such as FGF receptor substrate 2, Gab1, and Shc (22–25). As a mediator of signals downstream of potent epithelial and endothelial growth and motility factors, it is not surprising that Grb2 has been implicated in tumor progression and metastasis. For example, Grb2 is critical in metastatic signaling downstream of ErbB-2/Neu in a mouse model of mammary tumorigenesis (26), and the critical role of Grb2 in linking Figure 1. Chemical structures of Grb2-SH2 domain binding antagonists. HGF-stimulated c-Met activation with Ras and Rac- X, position of phosphotyrosine (pY) in the basic peptidomimetic structure regulated cell migration in normal development seems to recognized by the Grb2-SH2 domain, which is replaced with phosphonomethyl phenylalanine in compound C126 or with two carboxylic acid extend to c-Met-driven cell transformation and tumor moieties in C90, as shown below the backbone structure. Both moieties metastasis as well (27–30). confer resistance to cellular phosphatases. While all of these properties make Grb2 an excellent target for the design of therapeutic antitumor agents, Grand Rapids, MI). The HGF isoform HGF/NK1, which evaluating Grb2-SH2 domain binding antagonist phospho- has the same biological properties as full-length HGF, was peptides using preclinical models has been hampered by produced in a bacterial expression system and then their relative inability to cross the cell membrane, resulting purified and refolded as described previously (38). The in poor bioavailability (31, 32). The binding antagonists Grb2-SH2 domain binding antagonists designated C126 used in this study (33, 34) are potent tripeptide mimetic and C90 were synthesized and purified as described inhibitors that seem to have significantly improved po- (33, 34). Human recombinant basic FGF (bFGF) and human tential for cell membrane penetration (33); their chemical recombinant VEGF were from R&D Systems (Minneapolis, structures are shown in Fig. 1. Replacement of phosphotyr- MN). Phorbol 12-myristate 13-acetate (PMA) was from osine at position X (Fig. 1) with phosphonomethyl phe- Sigma Chemical Co. (St. Louis, MO). nylalanine (compound C126) or related structures protects Cultured Cell Lines this moiety from phosphatases (35). In a further evolution Human dermal microvascular (HMEC-1) endothelial of this concept, compound C90 has two carboxylic acid cells (39) were grown in RPMI 1640 (Biofluids, Rockville, moieties as a phosphomimetic group. These moieties confer MD) containing 10% fetal bovine serum (Biofluids) and resistance to cellular phosphatases and are the most potent 2 mmol/L glutamine (Sigma Chemical). Human micro- non-phosphorous-containing phosphotyrosine mimetics vascular endothelial cells (HMVEC) from neonatal dermis yet reported for the Grb2-SH2 domain (36, 37). We have were purchased from Cascade Biologicals (Portland, OR) reported previously that these synthetic inhibitors of the and cultured in medium 131 containing MVGS (media Grb2-SH2 domain potently block HGF-stimulated cell supplement) and 1% glutamine as indicated by the manu- motility, matrix invasion, and branching morphogenesis facturer. Human umbilical vein endothelial cells (HUVEC) but not HGF-stimulated mitogenesis in epithelial and were isolated from freshly delivered cords as reported hematopoietic cells. In this study, we have investigated previously (40) and grown on Nunclon dishes (Nunc, the role of Grb2 in angiogenesis in vitro and in vivo and Roskilde, Denmark) in RPMI 1640 supplemented with evaluated the potential use of Grb2 binding antagonists as 20% bovine calf serum (Hyclone Laboratories, Logan, UT), inhibitors of key steps of this process regulated by well- 50 Ag/mL gentamycin, 2.5 Ag/mL amphotericin B (fungi- characterized angiogenic factors. zone, Life Technologies, Gaithersburg, MD), 5 units/mL sodium heparin (Fisher Scientific, Pittsburgh, PA), and Materials and Methods 200 Ag/mL endothelial cell growth supplement (Collabo- Reagents rative Research, Bedford, MA) and were used between Full-length human HGF protein was a gift from Dr. passages 3 and 6. HMEC-1 and HUVEC were gifts from George Vande Woude (Van Andel Research Institute, Dr. Hynda Kleinman (NIDCR, NIH, Bethesda, MD).

Mol Cancer Ther 2004;3(10). October 2004

Cell Migration Assays (5% FCS and 150 Ag/mL endothelial cell growth supple- The migration of HUVEC or HMEC-1 was measured in ment). At confluence, cells were pretreated with or without modified Boyden chambers adapted from procedures Grb2 binding antagonists for 24 hours after which time described previously (41, 42). In brief, Biocoat Cell En- medium was replaced with fresh medium supplemented vironment control inserts (8 Am pore size, Becton Dick- with or without the indicated concentration of binding inson, Bedford, MA) were coated with 0.1% gelatin (Sigma antagonists and/or 20 ng/mL HGF. Medium and com- Chemical) for at least 1 hour at 37jC and air dried. Lower pounds were changed every day. After 5 days, the cultures chambers contained 0.7 mL RPMI plus 0.1% bovine serum were fixed in situ in 2.5% glutaraldehyde in 0.1 mol/L albumin, to which 20 ng/mL HGF and/or 300 nmol/L Grb2-SH2 domain binding antagonists were added. Cells sodium cacodylate buffer (pH 7.4) and five randomly were pretreated with the indicated concentrations of Grb2 selected optical fields (measuring 1 Â 1.4 mm) per ex- binding antagonists for 18 to 24 hours, trypsinized, washed perimental condition in each of three separate experiments twice in RPMI plus 0.1% bovine serum albumin, added to were digitally recorded under phase-contrast microscopy upper chambers (5 Â 104 cells/well) with or without using a 20Â objective by focusing 20 Am beneath the sur- binding antagonists in a final volume of 0.5 mL, and face of the gel. Invasion was quantitated using IPLab incubated for 4 hours at 37jC. Cells on the upper surface of software (Scanalytics, Fairfax, VA) by measuring the total each filter were removed with a cotton swab, whereas cells length of all cellular structures that had penetrated be- that had traversed to the bottom surface of the filter were neath the surface monolayer either as apparently single fixed and stained using Diff-Quik (Dade Diagnostics, cells or in the form of cell cords as described (44). Values Deerfield, IL) and counted by bright-field microscopy were compared using Student’s unpaired t test and a sig- using a 10Â objective. Mean values from four randomly selected fields (1 Â 1.4 mm) were calculated for each of nificant value was taken as P < 0.001. Results were ex- triplicate wells per experimental condition. Values are pressed as mean total length (Am) per field. expressed either as the ratio of growth factor–treated cells HUVECCordFormationAssayonMatrigel to control migrating cells designated on the y axis as The HUVEC tube formation assay was performed as ‘‘Migration (Fold Increase)’’ or as the mean number of cells described previously (41, 45). Briefly, 96-well plates were per optical field. coated with 90 AL Matrigel (10 mg/mL, Collaborative Cell Proliferation Assays Research) and incubated at 37jC for 30 minutes to promote HUVECs or HMVECs were seeded in triplicate into type gelling. HUVECs (10,000 cells/sample) were resuspended I collagen-coated 48-well plates (3,000 cells/well, Biocoat) in reduced growth medium (serum concentration 1% and in 500 AL complete EGM-Bullet Kit medium (BioWhittaker, 5 units/mL heparin) and added to each well with the indicated reagents in a final volume of 100 AL. After 18 hours, Walkersville, MD), allowed to attach for 4 hours, and the plates were fixed with Diff-Quik and at least four incubated for 16 hours with or without the indicated randomly selected fields per experimental condition were concentrations of binding antagonists. The cultures were digitally recorded under bright-field illumination using a rinsed twice in serum-free medium and incubated with 10Â objective. The mean additive length of all of the endothelial cell basal medium (BioWhittaker) supple- cords present in each optical field was measured using A A mented with 50 g/mL heparin, 50 g/mL ascorbic acid, IPLab software and compared with controls using the 10% FCS, and the indicated concentrations of binding Student’s unpaired t test. antagonists. After either 4 days (HUVECs) or 5 days Chick Chorioallantoic Membrane Assay (HMVECs), cells were harvested with trypsin-EDTA and To assess in vivo angiogenesis, a chick chorioallantoic counted with a hemocytometer. The mean number of cells membrane (CAM) assay was performed using a modifi- per milliliter was calculated from three independent cation of a procedure described previously (46, 47). To this measures for each of triplicate wells per experimental con- end, apical holes were drilled in 10-day-old chicken eggs to dition and compared with controls using the Student’s expose the CAMs, and filter paper discs (0.5 cm in unpaired t test. Results are described as the ratio of growth diameter) containing 20 ALofa1or3Amol/L C90 in PBS factor–treated cells to control proliferating cells designated were placed on the surface of each CAM (day 0). The holes on the y axis as ‘‘Proliferation (Fold Increase).’’ were covered with parafilm and the eggs were incubated in Extracellular Matrix Invasion Assay a humidified atmosphere at 37jC for 24 hours after which Invasion of collagen gels by HMEC-1 was adapted from a 100 ALof1or3Amol/L C90 was given i.v. using a 30-gauge procedure described previously (43, 44). Briefly, 1.5 mg/mL needle. Three days later (day 4), the discs and associated type I collagen (Cohesion Technologies, Palo Alto, CA) CAM were excised and immediately fixed in 3.7% form- was mixed with 10Â MEM and sodium bicarbonate (11.76 aldehyde. For computer-assisted image analysis, the discs mg/mL) at a ratio of 8:1:1 (v/v/v) on ice and 0.4 mL from 10 eggs per experimental condition in each of two aliquots were dispensed into 16-mm tissue culture wells separate experiments were divided into quadrants using (Nunc) and allowed to gel at 37jC for 20 minutes. Cells fine forceps, the blood vessels in each quarter were digitally were seeded onto collagen gels (2.5 Â 104 cells/well) and photographed, and the mean vessel area was normalized grown to confluence (1 week) in reduced growth medium to the total quadrant area and was quantitated using

Mol Cancer Ther 2004;3(10). October 2004

Optima 5 software. Statistical significance was determined Blockade of Growth Factor ^ Induced Endothelial Cell by Student’s t test. Values are expressed as mean F SD Proliferation by Grb2 Binding Antagonist C90 area in pixels. Ten eggs were used for each experimental To investigate the activity of compound C90 during condition. growth factor–induced endothelial cell proliferation, we cultivated HUVECs and HMVECs on type I collagen- Results coated wells in partially supplemented endothelial culture Grb2-SH2 Domain Binding Antagonists Inhibit Endo- medium as described in Materials and Methods. Under thelialCellMotility these stringent culture conditions, HGF (25 ng/mL), VEGF We initially assessed the effect of C126 on the chemo- (10 ng/mL), and bFGF (5 ng/mL) induced significant kinetic response of immortalized HMEC-1 and primary (P < 0.0001) increases of 2.3-, 2-, and 4.1-fold, respectively, HUVECs to well-characterized angiogenic factors. Com- in the mean number of macrovascular HUVECs per mil- pound C126 (300 nmol/L) did not significantly alter the liliter (Fig. 4A, open columns). Similar significant (P < 0.001) basal levels of endothelial cell motility (Fig. 2A and B, open increases in cell numbers were observed in HMVECs columns). However, it reduced to near-basal levels by (Fig. 4B, open columns). Addition of Grb2 inhibitor C90 5.5- and 3.5-fold increases in cell motility induced by 20 (30 and 300 nmol/L) resulted in significant, but markedly ng/mL HGF on HMEC-1 and HUVECs, respectively. different, levels of inhibition of proliferation in HUVECs Interestingly, this inhibitory activity was not limited to HGF stimulation as shown by the complete blockade of bFGF-induced HUVEC migration by 300 nmol/L C126 (Fig. 2A and B, solid columns). To further characterize the potential antiangiogenic effect of the Grb2 binding antagonists, we assessed neonatal HMVEC migration in the presence or absence of increasing concentrations of either HGF (0–50 ng/mL) or bFGF (0.2–50 ng/mL) and compound C90 (300 nmol/L). Compound C90 abolished (P < 0.001) cell migration induced by HGF (Fig. 2C, left) and significantly (P < 0.001) inhibited the biphasic stimulatory activity of bFGF (Fig. 2C, right). Notably, the effect of the optimal concentration of bFGF (5 ng/mL) was reduced by 72% (Fig. 2C, right). We next assessed the effect of C90 on the activity of VEGF, the most potent angiogenic molecule known to date. When incubated in modified Boyden chambers in the presence of VEGF, HMEC-1, HMVEC, and HUVEC underwent different, albeit significant, de- grees of migration (Fig. 3). However, the addition of compound C90 significantly (P < 0.001) reverted the effect of VEGF in all three cell lines, although the degree of inhibition of VEGF activity differed among the endothelial cell types. Similar results were observed with binding antagonist C126 (Fig. 3 and data not shown). Grb2 Binding Antagonist C90 Inhibits Endothelial Cell Migration Induced by PMA To determine whether the effect of Grb2 binding antagonists is restricted to the inhibition of natural angiogenic factors or might also block the effect of artificial proangiogenic molecules, we assessed the effect of C90 on the Figure 2. Grb2 binding antagonists inhibit HGF- and bFGF-induced migratory properties of HUVECs in the presence of PMA, a endothelial cell migration. HUVECs were pretreated with or without 300 potent tumor and angiogenesis promoter. Upon exposure to nmol/L C126 for 18 hours; after which time, cell migration in the presence PMA in vitro, both microvascular and macrovascular endo- of inhibitor and/or HGF (20 ng/mL) or bFGF (50 ng/mL) was quantified using modified Boyden chambers. A and B, open and solid columns, thelial cells undergo a vascular morphogenetic program by migration of cells in the absence or presence of HGF, respectively. invading the surrounding extracellular matrix and subse- Columns, ratio of migrating cells in growth factor – treated wells to control quently forming extensive network of capillary-like tubular wells (CTRL). B, gray columns, migration of cells in the presence of bFGF. Results from one representative experiment. C, effect of Grb2 structures (43). Addition of PMA (40 ng/mL) to cultures binding antagonist C90 on HGF- and bFGF-induced HMVEC migration. of HUVECs in modified Boyden chambers resulted in a 3- Cells were pretreated for 18 hours with or without 300 nmol/L C90; after fold increase in cell migration. This proangiogenic activity, which time, the experiments were done in the presence or absence of which cannot be mimicked by the maximal effective inhibitor and the indicated concentrations of growth factors. Columns, mean number of cells per optical field from triplicate wells per experimental concentration (5 ng/mL) of bFGF, was reduced to near- condition; bars, SE. Where no bars are visible, the error is too small to basal levels in the presence of 300 nmol/L C90 (Table 1). be shown.

Mol Cancer Ther 2004;3(10). October 2004

Table 1. Grb2 binding antagonist C90 inhibits PMA-induced HUVEC migration

Treatment Group Migrating Cells per Field

Control 43 F 9 PMA 148 F 15 PMA + C90 58 F 8 bFGF 86 F 14

NOTE: HUVECs pretreated with or without C90 (300 nmol/L) for 18 hours were seeded in modified Boyden chambers in the presence or absence of either 40 ng/mL PMA or 5 ng/mL bFGF as described in Materials and Methods. Values are the mean F SE number of cells per optical field; four randomly selected fields were quantitated for each of triplicate wells per experimental condition. and HMVECs. Whereas C90 inhibited serum-dependent (P < 0.001) and bFGF-dependent (P < 0.0001) HUVEC proliferation at concentrations of 30 nmol/L, it failed to significantly inhibit HGF- and VEGF-induced proliferation at this concentration (Fig. 4A, gray columns). Only higher concentrations of compound (300 nmol/L) were able to induce significant (P < 0.001) reductions in cell counts under these conditions (Fig. 4A, solid columns). In contrast, no significant inhibition of HMVEC proliferation was observed with 30 nmol/L compound C90 in either the presence or the absence of growth factors, and generally less but significant (P < 0.001) inhibition of HMVEC growth was observed with 300 nmol/L compound (Fig. 4B). Figure 4. Effects of Grb2 binding antagonist C90 on endothelial cell Grb2 Binding Antagonists Block HGF-Induced Matrix proliferation. HUVECs (A) or HMVECs (B) were grown in 48-well plates, pretreated 16 hours without (open columns)orwith30(gray columns) Invasion by Microvascular Endothelial Cells or 300 (solid columns) nmol/L compound C90, grown for either 4 days We investigated the effect of Grb2 binding inhibitors (HUVECs) or 5 days (HMVECs), harvested with trypsin/EDTA, and on extracellular matrix invasion using HMEC-1 grown to counted. The mean number of cells per milliliter was calculated from three independent measures for each of triplicate wells per experimental condition and compared with controls using the Student’s unpaired t test. Columns, ratio of growth factor – treated cells to control proliferating cells (Fold Increase).

confluence on the surface of collagen type I gels as described in Materials and Methods. When monolayers of HMEC-1 were grown under control conditions for 5 days, they discretely invaded the underlying collagen matrix as single cells (Fig. 5A, top). Addition of 20 ng/mL HGF to the cultures induced a 6-fold increase in collagen invasion by either single cells or short cell cords devoid of lumen (Fig. 5A, middle, and B). Coaddition of C126 (30-300 nmol/L) and HGF to the cultures suppressed collagen invasion induced by HGF (Fig. 5, bottom). Invasion was quantitated by measuring the total length of all cellular structures that had penetrated 20 Am beneath the surface monolayer either as apparently single cells or in the form of cell cords as described previously (43). Significant (P < 0.001) decreases Figure 3. Grb2 binding antagonists inhibit VEGF-induced endothelial cell migration. The migration of HMEC-1, HMVECs, and HUVECs was in the total length of the cords penetrating the surface of measured using modified Boyden chambers. Cells were pretreated for the gel matrix were observed with concentrations of C126 18 hours with (filled symbols) or without (open symbols) 300 nmol/L as low as 30 nmol/L, and while no effect was observed on C90; after which time, the experiments were performed in the presence or control levels of invasion, HGF-induced HMEC-1 matrix absence of inhibitor and the indicated concentrations of VEGF. Points, mean number of cells per optical field from triplicate wells per experimental invasion was generally reversed in the presence of the Grb2 condition; bars, SE. binding antagonist (Fig. 5B).

Mol Cancer Ther 2004;3(10). October 2004

Grb2 Binding Antagonist C126 Impairs Endothelial Cell Tubulogenesis In vitro In an in vitro model of angiogenesis (41, 44, 45), we seeded HUVECs onto Matrigel gel layers and immediately treated the cells with the indicated agents (Fig. 6). Under control conditions, the cells migrated on the surface of the

Figure 5. Grb2 binding antagonist C126 inhibits HGF-induced matrix invasion by HMEC-1. A, cells seeded on collagen gels were grown to confluence prior to incubation with or without inhibitors and/or HGF (20 ng/mL) for 24 hours as described in Materials and Methods. In the three panels, invasion was assessed by focusing 20 Am beneath the surface of the gel. HMEC-1 incubated under control conditions remain primarily on the surface of the collagen gel, which appears blurred in the background, and only a few cells have invaded the underlying matrix (top). Cells Figure 6. Grb2 binding antagonist C126 inhibits HGF-stimulated incubated with HGF invade the gel extensively either as whole cells or as angiogenesis in vitro. HUVECs were plated on Matrigel-coated 96-well cellular projections (middle). Coaddition of HGF and the Grb2 binding plates and incubated with or without 300 ng/mL HGF/NK1 and or 1 Amol/L antagonist C126 results in a significantly reduced number of invading cells C126 for 18 hours as described in Materials and Methods. Under control invading the matrix (bottom). B, quantitative analysis of collagen matrix conditions (top), cells form irregular aggregates on the surface of the invasion. Open columns and solid columns, matrix invasion by HMEC-1 Matrigel and short, discontinuous cell cords were visualized by dark-field in the absence or presence of HGF, respectively. Columns, mean length of microscopy (Â40). Cells treated with HGF/NK1 form an extensive network all projections or single cells invading the collagen at 20 Am beneath the of anastomosing, thick endothelial cords (middle). Cells treated with surface of the gel; bars, SE. Results are from at least three independent HGF/NK1 and C126 have formed only a few thin and discontinuous cords, experiments. similar to controls (bottom).

Mol Cancer Ther 2004;3(10). October 2004

Figure 7. Grb2 binding antagonist C90 blocks VEGF- and bFGF-induced angiogenesis in vitro. HUVECs were plated on Matrigel- coated 96-well plates with or without C90 (300 nmol/L) and VEGF (10 ng/mL) or bFGF (5 ng/mL) for 18 hours. In the presence of VEGF, HUVECs form an extensive network of interconnecting cords on the surface of the Matrigel visualized by phase-contrast microscopy (Â40; top left). Simultaneous addition of C90 inhibits VEGF-driven cord formation, resulting in the formation of short, discontinuous cell cords and cell aggregates on the surface of the Matrigel (top right). In the presence of 5 ng/mL bFGF, HUVECs develop a network of anastomosing endothelial cords similar to that induced by VEGF (bottom left); coaddition of C90 to bFGF-treated cultures results in partial inhibition of cord formation (bottom right).

gel, established contact with each other, and after 12 to 18 alone were placed onto the CAM of 10-day-old eggs, which hours formed irregular ridges or cords, a process reminis- were subsequently given a single i.v. injection of PBS, a cent of the early steps of angiogenesis (Fig. 6, top). Addition large number of vessels occupied the CAM area covered of 20 to 50 ng/mL HGF (data not shown) or 300 ng/mL by the paper discs (Fig. 8A, left). In striking contrast, when HGF/NK1 to the cultures resulted in the formation of a the discs were treated with compound C90 (3 mmol/L), the continuous, extensive network of thick cords. In contrast, number of vessels underlying the paper discs was sig- coaddition of HGF/NK1 and compound C126 (1 Amol/L) nificantly reduced (Fig. 8A, right). Morphometric analy- markedly reduced the extent of cord formation (Fig. 6, sis showed a significant (P < 0.05) and dose-dependent compare middle and bottom panels). A quantitative analysis decrease in mean vessel area in CAM treated with C90 of mean length of the cords per optical field showed a relative to control-treated CAM samples (Fig. 8B). Similar significant (P < 0.001) 3-fold increase induced by HGF/ results were obtained when measuring mean vessel length NK1 (3,217.0 F 166.5 Am in HGF/NK1-treated cultures (data not shown). We also observed that chick embryos versus 1,189.5 F 166 Am in controls) and a significant treated with compound C90 were alive and showed the (P = 0.001) 30% decrease when coaddition of inhibitor same extent of motility as embryos from PBS-treated eggs C126 and HGF/NK1 was compared with HGF/NK1 alone at the end of the experiment. (3,217.0 F 166.5 Am in HGF/NK1 alone versus 2,502.8 F 108 Am in HGF/NK1 plus C126). Qualitatively similar results were observed when HGF or HGF/NK1 were Discussion substituted by 5 ng/mL bFGF or 10 ng/mL VEGF (Fig. 7, The ability of Grb2-SH2 domain binding antagonists to compare left and right panels). block HGF-stimulated epithelial cell migration and inva- Grb2 Binding Antagonist C90 Inhibits Angiogenesis sion has been described (48, 49). Our prior studies using in the Chick CAM an in vitro model of mammary epithelial morphogenesis Chick CAM assays performed on 10-day-old eggs dem- revealed that these binding antagonists could block HGF- onstrated that compound C90 inhibits basal in vivo induced formation of branching duct-like structures (49) angiogenesis (Fig. 8). When paper discs treated with PBS and prompted us to determine whether they might also

Mol Cancer Ther 2004;3(10). October 2004

Figure 8. Grb2 binding antagonist C90 inhibits angiogenesis in the chick CAM. Chick CAM assays were used to assess the impact of Grb2-SH2 domain antagonism on angiogenesis in vivo. Apical holes were drilled in 10-day-old chicken eggs to expose the CAM, and filter paper discs (0.5 cm diameter) containing 20 ALofa1or3Amol/L solution of C90 were placed on the CAM surface. After 24 hours, 100 ALof1or3Amol/L C90 was given again i.v. Three days later, the discs and associated CAM were excised, fixed, and photographed by bright-field microscopy. Representative filters from control (left) and C90-treated (right) CAM are in A. Discs from 10 eggs per experimental condition in each of two separate experiments were digitally photographed and mean vessel area was normalized to total area using Optima 5 software. B, columns, expressed as a mean area in pixels; bars, SD. *, P < 0.001, statistical significance from control, Student’s t test.

modulate the morphogenetic response of endothelial cells to and/or adaptors (9, 25, 48–52). VEGF and bFGF are among various angiogenic stimuli, including HGF. Here, we show the most potent regulators of angiogenesis but share that these Grb2-SH2 domain binding antagonists also block intracellular signaling mediators with a variety of angio- HGF-stimulated matrix invasion and branching morphogenesis signaling pathways (4). In porcine aortic endothe- genesis by vascular endothelial cells as well as HGF-, VEGF-, lial cells, for example, VEGF-A-induced phosphorylation of and bFGF-stimulated vascular endothelial cell proliferation, VEGFR2 induces the recruitment of Shc, Grb2, and Nck and migration, and the formation of capillary-like structures on formation of the Shc-Grb2 complex (21). The FLT4L kinase, reconstituted extracellular matrices. an alternatively spliced form of the VEGFR3/FLT4 recep- Grb2-SH2 domain binding blockade was associated with tor, binds Grb2 directly through its cytoplasmic region and inhibition of growth stimulated by HGF, bFGF, and VEGF recruits a second pool of Grb2 by phosphorylating tyrosine more significantly in HUVECs than in HMVECs. These residues within appropriate sequence motifs in Shc (18, 20). results contrast with our previous findings where little Grb2 acting at this level is believed to link receptor inhibition of HGF-stimulated epithelial cell proliferation activation to cell cycle progression and/or cell motility was observed (49) and suggest that the roles of Grb2 in via the Sos/Ras and Rac1/Rho pathways, respectively facilitating cell cycle progression and other activities are (1, 18, 22, 26, 27, 29, 48, 52–55). We and others have cell type dependent and may vary among different vas- reported direct evidence of the ability of Grb2-SH2 domain cular beds. The results presented here are also consistent binding antagonists to block HGF-stimulated Grb2 signal- with the phenotypic heterogeneity generally observed ing at this level, with primary impact on cell motility and among endothelial cells from different tissues and different motility-dependent processes (48, 49). levels of the vasculature. Most significant in the present In addition to its role as a receptor-proximal adaptor study, and consistent with our prior report, was potent protein, Grb2 participates directly in the regulation of actin Grb2-binding antagonist-induced inhibition of motility filament formation and actin-based cell motility. Grb2 is observed in every cell line tested. In addition, and perhaps important in T-cell receptor signaling (56) and is a critical as a consequence of motility inhibition, matrix invasion link between Wiskott-Aldrich Syndrome protein (WASp) and cord formation were also potently blocked. and the actin cytoskeleton; WAS patients show defects in While the molecular mechanisms by which SH2 binding T-cell polarization and migration in response to physiologic antagonists inhibit motility, invasion, and cord formation stimuli, resulting in thrombocytopenia, eczema, and immu- are not yet fully defined, published evidence indicates that nodeficiency (57–59). Studies of WASp function and the Grb2 may regulate cell motility at two levels: as a receptor intracellular motility of invasive microbial pathogens such proximal adaptor and as a direct participant in actin as Listeria monocytogenes and Vaccinia virus have elucidated filament based motility. On ligand stimulation, Grb2 is an important role for Grb2 in directly promoting actin- recruited via its SH2 domain to growth factor receptors or based motility (58, 60, 61). These microbes invade eukary- to tyrosine-phosphorylated receptor proximal substrates otic cells and use a limited number of microbial surface

Mol Cancer Ther 2004;3(10). October 2004

proteins to harness host cell actin and associated regulatory preliminary pharmacokinetic studies, we have not observed molecules for propulsion through the host cytoplasm. toxicity in normal adult mice after 21 days of i.p. injections Unraveling these pathogenic events helped to define the of millimolar doses of binding antagonist.6 Further studies roles of several eukaryotic proteins in normal actin-based of these compounds in other animal models are underway. motility. In most mammalian cells, the WASp family Despite the modest performance of angiogenesis inhib- member neural WASp interacts with the ARP2/3 complex itors as single agents in early human clinical trials (66), more and G-actin to stimulate actin polymerization. Neural WASp recent trials showing better efficacy have resulted in activity is enhanced by other effectors such as Nck, Cdc42, approval of the first rationally designed antiangiogenic and Grb2; disruption of Grb2-SH3 or -SH2 domains dimin- drug (67). In parallel, the finding that angiogenesis in- ishes actin polymerization and thus actin-based motility hibitors can enhance the potency of conventional chemo- (58, 60). The activity of the Grb2-SH2 domain binding therapy and radiotherapy has brought widespread support antagonists described here at this level of Grb2 function to multidrug clinical trials where conventional and anti- is currently under investigation. angiogenic therapies are combined (68). Combination trials, Grb2 also acts as a critical downstream intermediary in however, are inherently more complex with regard to their several angiogenic signaling pathways not tested here, design, execution, and interpretation, in addition to po- including platelet-derived growth factor (PDGF), ELK, and tential obstacles related to the approval and/or commer- Eph, suggesting that these pathways may also be suscep- cialization of drug combinations (69). These circumstances tible to blockade by Grb2-SH2 domain binding antagonists. make single agents that target both tumor and vascular PDGF is implicated in different biological processes such compartments very attractive from the earliest phases of as vascular remodeling, wound healing, and cancer (54, 62). rational drug design and development. Together with our Consistent with the spectrum of pathways disrupted by previous studies, the results presented here suggest that Grb2 antagonism, we have also observed that the Grb2 potent simultaneous disruption of Grb2-SH2 domain– binding antagonist C90 inhibits PDGF-BB–induced mediated signaling downstream of HGF, VEGF, and bFGF NIH3T3 cell migration.5 In microvascular endothelial cells, and potentially other angiogenic regulators in the tumor membrane-bound LERK-2 stimulates ELK tyrosine phos- microenvironment by a single agent may be a clinically phorylation and recruitment of Grb2 and Grb10 via their viable strategy to inhibit tumor cell invasion into the SH2 domains, signaling the assembly of microvascular surrounding tissue, tumor metastasis, and the recruitment of endothelial cells into capillary-like cords (52). Interestingly, new blood vessels needed to sustain these pathologic processes. natively expressed Eph-related receptors, including ELK, Mek4, and Eck, do not signal proliferative responses on Acknowledgments ligand binding (52, 63), reinforcing the hypothesis that We thank Dr. Hynda Kleinman for endothelial cell lines; Dr. George Vande motility and morphogenesis are primary cellular responses Woude for full-length HGF protein; Sauda Ayub for assistance with downstream of Grb2. We also show that the Grb2-SH2 illustrations; and Diane Breckenridge, Dr. Angelina Felici, and Dr. Nese Atabey for helpful suggestions. domain binding antagonists inhibit PMA-induced cell migration. PMA is an artificial activator of PKC; PKC- References mediated migration plays an important role not only during angiogenesis but also during tumor development, 1. Gerwins P, Skoldenberg E, Claesson-Welsh L. Function of fibroblast growth factors and vascular endothelial growth factors and their receptors tumor invasion, and metastasis. Signaling through the in angiogenesis. Crit Rev Oncol Hematol 2000;34:185 – 94. VEGF receptors FLT-1 and FLK-1 is thought to occur via 2. Ferrara N. Role of vascular endothelial growth factor in regulation of PKC activation and the MAPK and PI-3-kinase pathways, physiological angiogenesis. Am J Physiol Cell Physiol 2001;280:1358 – 66. ultimately leading to capillary formation in vitro and to 3. Pepper MS. Positive and negative regulation of angiogenesis: from cell vasculogenesis and angiogenesis in vivo (64, 65). biology to the clinic. Vasc Med 1996;1:259 – 66. The widespread role of Grb2 in most, if not all, an- 4. Folkman J. Angiogenesis and angiogenesis inhibition: an overview. giogenesis signaling pathways together with our observa- EXS 1997;79:1 – 8. tions that Grb2-SH2 domain binding antagonists blocked 5. Rosen EM, Lamszus K, Laterra J, Polverini PJ, Rubin JS, Goldberg ID. HGF/SF in angiogenesis. Ciba Found Symp 1997;212:215 – 26. endothelial cell migration, matrix invasion, proliferation, 6. Lowenstein EJ, Daly RJ, Batzer AG, et al. The SH2 and SH3 domain- and cord formation in vitro suggested that they might containing protein GRB2 links receptor tyrosine kinases to ras signaling. also inhibit angiogenesis in vivo. Results obtained using the Cell 1992;70:431 – 42. chick CAM assay support this contention: SH2 domain 7. Clark SG, Stern MJ, Horvitz HR. C. elegans cell-signaling gene sem-5 binding antagonists displayed potent local inhibition of encodes a protein with SH2 and SH3 domains. Nature 1992;356:340 – 4. CAM vasculogenesis. We noted that these treatments did 8. Matuoka K, Shibata M, Yamakawa A, Takenawa T. Cloning of ASH, a ubiquitous protein composed of one Src homology region (SH) 2 and not seem to disrupt vasculogenesis generally and that chick two SH3 domains, from human and rat cDNA libraries. Proc Natl Acad Sci embryos displayed grossly normal development and U S A 1992;89:9015 – 9. motility throughout the time course of the experiments. In 9. Buday L. Membrane-targeting of signaling molecules by SH2/SH3

5 J. Soriano and D. Bottaro, unpublished observations. 6 D. Bottaro, unpublished observations.

Mol Cancer Ther 2004;3(10). October 2004

domain-containing adaptor proteins. Biochim Biophys Acta 1999;1422: 31. Dunican DJ, Doherty P. Designing cell-permeant phosphopeptides 187 – 204. to modulate intracellular signaling pathways. Biopolymers 2001;60: 45– 60. 10. Rozakis-Adcock M, Fernley R, Wade J, Pawson T, Bowtell D. The 32. Fretz H, Furet P, Garcia-Echeverria C, Schoepfer J, Rahuel J. SH2 and SH3 domains of mammalian Grb2 couple the EGF receptor to the Structure-based design of compounds inhibiting Grb2-SH2 mediated Ras activator mSos1. Nature 1993;363:15 – 6. protein-protein interactions in signal transduction pathways. Curr Pharm Des 2000;6:1777 – 96. 11. Pawson T, Scott JD. Signaling through scaffold, anchoring, and adaptor proteins. Science 1997;278:2075 – 80. 33. Yao ZJ, King CR, Cao T, et al. Potent inhibition of Grb2 SH2 domain 12. Rozakis-Adcock M, McGlade J, Mbamalu G, et al. Association of the binding by non-phosphate-containing ligands. J Med Chem 1999;42:25 – 35. Shc and Grb2/Sem5 SH2-containing proteins is implicated in activation of 34. Gao Y, Luo J, Yao ZJ, et al. Inhibition of Grb2 SH2 domain binding the Ras pathway by tyrosine kinases. Nature 1992;360:689 – 92. by non-phosphate-containing ligands. 2. 4-(2-Malonyl)phenylalanine as a 13. Furge KA, Zhang YW, Vande Woude GF. Met receptor tyrosine potent phosphotyrosyl mimetic. J Med Chem 2000;43:911 – 20. kinase: enhanced signaling through adapter proteins. Oncogene 2000; 19:5582 – 9. 35. Burke TR Jr, Smyth MS, Nomizu M, Otaka A, Roller PP. Preparation of fluoro- and hydroxy-4-phosphonomethyl-D,L-phenylalanine suitably pro- 14. Yamauchi T, Ueki K, Tobe K, et al. Growth hormone-induced tected for solid phase synthesis of peptides containing hydrolytically stable tyrosine phosphorylation of EGF receptor as an essential element leading analogues of O-phosphotyrosine. J Org Chem 1993;58: 1336 – 40. to MAP kinase activation and gene expression. Endocr J 1998;45 Suppl:27 – 31. 36. Bandyopadhyay G, Sajan MP, Kanoh Y, et al. Glucose activates mitogen-activated protein kinase (extracellular signal-regulated kinase) 15. Schlaepfer DD, Hauck CR, Sieg DJ. Signaling through focal adhesion through proline-rich tyrosine kinase-2 and the Glut1 glucose transporter. kinase. Prog Biophys Mol Biol 1999;71:435 – 78. J Biol Chem 2000;275:40817 – 26. 16. Zhang S, Weinheimer C, Courtois M, et al. The role of the Grb2-p38 MAPK signaling pathway in cardiac hypertrophy and fibrosis. J Clin Invest 37. Shi ZD, Wei CQ, Lee K, et al. Macrocyclization in the design of non- 2003;111:833 – 41. phosphorus-containing Grb2 SH2 domain-binding ligands. J Med Chem 2004;47:2166 – 9. 17. Yamazaki T, Komuro I, Yazaki Y. Role of the renin-angiotensin system in cardiac hypertrophy. Am J Cardiol 1999;83:53 – 7H. 38. Stahl SJ, Wingfield PT, Kaufman JD, et al. Functional and biophysical characterization of recombinant human hepatocyte growth factor isoforms 18. Pajusola K, Aprelikova O, Pelicci G, Weich H, Claesson-Welsh L, produced in Escherichia coli. Biochem J 1997;326:763 – 72. Alitalo K. Signaling properties of FLT4, a proteolytically processed receptor tyrosine kinase related to two VEGF receptors. Oncogene 1994;9: 39. Ades EW, Candal FJ, Swerlick RA, et al. HMEC-1:establishment of an 3545 – 55. immortalized human microvascular endothelial cell line. J Invest Dermatol 19. Saaristo A, Veikkola T, Tammela T, et al. Lymphangiogenic gene 1992;99:683 – 90. therapy with minimal blood vascular side effects. J Exp Med 2002;196: 40. Jaffe EA, Nachman RL, Becker CG, Minick CR. Culture of human 719 – 30. endothelial cells derived from umbilical veins. Identification by morpho- 20. Fournier E, Blaikie P, Rosnet O, Margolis B, Birnbaum D, Borg JP. Role logic and immunologic criteria. J Clin Invest 1973;52:2745 – 56. of tyrosine residues and protein interaction domains of SHC adaptor in VEGF receptor 3 signaling. Oncogene 1999;18:507 – 14. 41. Malinda, KM, Nomizu M, Chung M, et al. Identification of laminin a1 and h1 chain peptides active for endothelial cell adhesion, tube formation, 21. Kroll J, Waltenberger J. The vascular endothelial growth factor and aortic sprouting. FASEB J 1999;13:53 – 62. receptor KDR activates multiple signal transduction pathways in porcine aortic endothelial cells. J Biol Chem 1997;272:32521 – 7. 42. Murohara T, Witzenbichler B, Spyridopoulos I, et al. Role of endothelial nitric oxide synthase in endothelial cell migration. Arterioscler 22. Shono T, Kanetake H, Kanda S. The role of mitogen-activated pro- Thromb Vasc Biol 1999;19:1156 – 61. tein kinase activation within focal adhesions in chemotaxis toward FGF- 2 by murine brain capillary endothelial cells. Exp Cell Res 43. Montesano R, Orci L. Tumor-promoting phorbol esters induce 2001;264:275 – 83. angiogenesis in vitro. Cell 1985;42:469 – 77. 23. Stoletov KV, Ratcliffe KE, Terman BI. Fibroblast growth factor 44. Pepper MS, Vassalli JD, Orci L, Montesano R. Biphasic effect of receptor substrate 2 participates in vascular endothelial growth factor- transforming growth factor-h1onin vitro angiogenesis. Exp Cell Res induced signaling. FASEB J 2002;16:1283 – 5. 1993;204:356 – 63. 24. Fournier TM, Kamikura D, Teng K, Park M. Branching tubulogenesis 45. Grant DS, Kinsella JL, Fridman R, et al. Interaction of endothelial cells but not scatter of Madin-Darby canine kidney cells requires a functional with a laminin A chain peptide (SIKVAV) in vitro and induction of Grb2 binding site in the Met receptor tyrosine kinase. J Biol Chem 1996; angiogenic behavior in vivo. J Cell Physiol 1992;153:614 – 25. 271:22211. 46. Brooks PC, Silletti S, von Schalscha TL, Friedlander M, Cheresh DA. 25. Birchmeier C, Birchmeier W, Gherardi E, Vande Woude GF. Met, Disruption of angiogenesis by PEX, a noncatalytic metalloproteinase metastasis, motility and more. Nat Rev Mol Cell Biol 2003;4:915 – 25. fragment with integrin binding activity. Cell 1998;92:391 – 400. 26. Dankort D, Maslikowski B, Warner N, et al. Grb2 and Shc adapt- 47. Liu N, Lapcevich RK, Underhill CB, et al. Metastatin: a hyaluronan- er proteins play distinct roles in Neu (ErbB-2)-induced mammary tumori- binding complex from cartilage that inhibits tumor growth. Cancer Res genesis: implications for human breast cancer. Mol Cell Biol 2001; 2001;61:1022 – 8. 21:1540 – 51. 27. Ridley AJ, Comoglio PM, Hall A. Regulation of scatter factor/ 48. Gay B, Suarez S, Weber C, et al. Effect of potent and selective hepatocyte growth factor responses by Ras, Rac, and Rho in MDCK cells. inhibitors of the Grb2 SH2 domain on cell motility. J Biol Chem 1999; Mol Cell Biol 1995;15:1110 – 22. 274:23311 – 5. 28. Cheng AM, Saxton TM, Sakai R, et al. Mammalian Grb2 regulates 49. Atabey N, Gao Y, Yao ZJ, et al. Potent blockade of hepatocyte multiple steps in embryonic development and malignant transformation. growth factor-stimulated cell motility, matrix invasion and branching Cell 1998;95:793 – 803. morphogenesis by antagonists of Grb2 Src homology 2 domain interactions. J Biol Chem 2001;276:14308 – 14. 29. Maina F, Casagranda F, Audero E, et al. Uncoupling of Grb2 from the Met receptor in vivo reveals complex roles in muscle development. Cell 50. Carter-Su C, Smit LS. Signaling via JAK tyrosine kinases: growth 1996;87:531 – 42. hormone receptor as a model system. Recent Prog Horm Res 1998;53:61 – 82. 30. Saucier C, Papavasiliou V, Palazzo A, Naujokas MA, Kremer R, Park M. Use of signal specific receptor tyrosine kinase oncoproteins reveals that 51. Ridyard MS, Robbins SM. Fibroblast growth factor-2-induced signal- pathways downstream from Grb2 or Shc are sufficient for cell trans- ing through lipid raft-associated fibroblast growth factor receptor formation and metastasis. Oncogene 2004;21:1800 – 11. substrate 2 (FRS2). J Biol Chem 2003;278:13803 – 9.

Mol Cancer Ther 2004;3(10). October 2004

52. Stein E, Cerretti DP, Daniel TO. Ligand activation of ELK receptor 60. Scaplehorn N, Holmstrom A, Moreau V, Frischknecht F, Reckmann I, tyrosine kinase promotes its association with Grb10 and Grb2 in vascular Way M. Grb2 and Nck act cooperatively to promote actin-based motility of endothelial cells. J Biol Chem 1996;271:23588 – 93. Vaccinia virus. Curr Biol 2002;12:740 – 5. 53. Jones N, Master Z, Jones J, et al. Identification of Tek/Tie2 binding 61. Carlier MF, Le Clainche C, Wiesner S, Pantaloni D. Actin-based partners. Binding to a multifunctional docking site mediates cell survival motility: from molecules to movement. BioEssays 2003;25:336 – 45. and migration. J Biol Chem 1999;274:30896 – 905. 62. Gendron RL. A plasticity for blood vessel remodeling is defined by 54. Bornfeldt KE, Raines EW, Graves LM, Skinner MP, Krebs EG, Ross R. pericyte coverage of the preformed endothelial network and is regulated by Platelet-derived growth factor. Distinct signal transduction pathways PDGF-B and VEGF. Surv Ophthalmol 1999;44:184 – 5. associated with migration versus proliferation. Ann N Y Acad Sci 1995; 63. Beckmann MP, Cerretti DP, Baum P, et al. Molecular characterization 766:416 – 30. of a family of ligands for eph-related tyrosine kinase receptors. EMBO J 55. Boerner JL, Danielsen AJ, Lovejoy CA, et al. Grb2 regulation of the 1994;13:3757 – 62. actin-based cytoskeleton is required for ligand-independent EGF receptor- 64. Ilan N, Mahooti S, Madri JA. Distinct signal transduction pathways mediated oncogenesis. Oncogene 2003;22:6679 – 89. are utilized during the tube formation and survival phases of in vitro 56. Zhang W, Samelson, LE. The role of membrane-associated adaptors angiogenesis. J Cell Sci 1998;111:3621 – 31. in T cell receptor signaling. Semin Immunol 2000;12:35 – 41. 65. Holash J, Wiegand SJ, Yancopoulos GD. New model of tumor 57. She HY, Rockow S, Tang J, et al. Wiskott-Aldrich syndrome protein is angiogenesis: dynamic balance between vessel regression and growth associated with the adapter protein Grb2 and the epidermal growth factor mediated by angiopoietins and VEGF. Oncogene 1999;18:5356 – 62. receptor in living cells. Mol Cell Biol 1997;8:1709 – 21. 66. Mousa SA, Mousa AS. Angiogenesis inhibitors: current and future directions. Curr Pharm Des 2004;10:1 – 9. 58. Carlier MF, Nioche P, Broutin-L’Hermite I, et al. GRB2 links signaling to actin assembly by enhancing interaction of neural Wiskott-Aldrich 67. Goodman L. Persistence-luck-Avastin. J Clin Invest 2004;113:934. syndrome protein (N-WASp) with actin-related protein (ARP2/3) complex. 68. Kerbel R, Folkman J. Clinical translation of angiogenesis inhibitors. J Biol Chem 2000;275:21946 – 52S. Nat Rev Cancer 2002;2:727 – 39. 59. Wada T, Jagadeesh GJ, Nelson DL, Candotti F. Retrovirus-mediated 69. Millar A. Limitations of combination anti-angiogenesis and chemo- WASP gene transfer corrects Wiskott-Aldrich syndrome T-cell dysfunc- therapy. Nat Rev Cancer 2003 [published online 6 Jan 2003]. Available tion. Hum Gene Ther 2002;13:1039 – 46. from: http://www.nature.com/nrc/journal/v2/n10/corres/nrc905_fs.html

Mol Cancer Ther 2004;3(10). October 2004

Downloaded from mct.aacrjournals.org on September 26, 2021. © 2004 American Association for Cancer Research. Inhibition of angiogenesis by growth factor receptor bound protein 2-Src homology 2 domain bound antagonists

Jesus V. Soriano, Ningfei Liu, Yang Gao, et al.

Mol Cancer Ther 2004;3:1289-1299.

Updated version Access the most recent version of this article at: http://mct.aacrjournals.org/content/3/10/1289

Cited articles This article cites 62 articles, 16 of which you can access for free at: http://mct.aacrjournals.org/content/3/10/1289.full#ref-list-1

Citing articles This article has been cited by 3 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/3/10/1289.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/3/10/1289. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.