Hyaluronan promotes CD44v3-Vav2 interaction with Grb2-p185(HER2) and induces Rac1 and Ras signaling during ovarian tumor cell migration and growth.

In this study we initially examined the interaction between CD44v3 (a hyaluronan (HA) receptor) and Vav2 (a guanine nucleotide exchange factor) in human ovarian tumor cells (SK-OV-3.ipl cell line). Immunological data indicate that both CD44v3 and Vav2 are expressed in SK-OV-3.ipl cells and that these two proteins are physically linked as a complex in vivo. By using recombinant fragments of Vav2 and in vitro binding assays, we have detected a specific binding interaction between the SH3-SH2-SH3 domain of Vav2 and the cytoplasmic domain of CD44. In addition, we have observed that the binding of HA to CD44v3 activates Vav2-mediated Rac1 signaling leading to ovarian tumor cell migration. Further analyses indicate that the adaptor molecule, growth factor receptor-bound protein 2 (Grb2) that is bound to p185(HER2) (an oncogene product), is also associated with the CD44v3-Vav2 complex. HA binding to SK-OV-3.ipl cells promotes recruitment of both Grb2 and p185(HER2) to the CD44v3-Vav2 complex leading to Ras activation and ovarian tumor cell growth. In order to determine the role of Grb2 in CD44v3 signaling, we have transfected SK-OV-3.ipl cells with Grb2 mutant cDNAs (e.g. Delta N-Grb2 that has a deletion in the amino-terminal SH3 domain or Delta C-Grb2 that has a deletion in the carboxyl-terminal SH3 domain). Our results clearly indicate that the SH3 domain deletion mutants of Grb2 (i.e. the Delta N-Grb2 (and to a lesser extent the Delta C-Grb2) mutant) not only block their association with p185(HER2) but also significantly impair their binding to the CD44v3-Vav2 complex and inhibit HA/CD44v3-induced ovarian tumor cell behaviors. Taken together, these findings strongly suggest that the interaction of CD44v3-Vav2 with Grb2-p185(HER2) plays an important role in the co-activation of both Rac1 and Ras signaling that is required for HA-mediated human ovarian tumor progression.

has been identified. Vav2 is a member of the Vav family of oncoproteins known to act as a GEF for RhoGTPases (e.g. Rac1, RhoA, and Cdc42) in a phosphotyrosine-dependent manner (27). The Vav molecule was first identified as an oncogene based on its ability to induce cell transformation following gene transfer using human tumor DNA (28). Vav1 is expressed exclusively in hematopoietic cells, whereas the homologous protein, Vav2, is now known to be expressed ubiquitously (29,30). Structurally, Vav2 is a 95-100-kDa protein, which contains numerous functional domains and structural motifs found in signal transduction proteins and oncoproteins. These motifs include aminoterminal calponin homology/leucine-rich and acid-rich domains (31), a Dbl homology domain (DH) (32,33), a pleckstrin homology domain (PH) (34), a cysteine-rich zinc-binding domain (CR) (31), and two src homology-3 (SH3) domains flanking a single SH2 domain (35,36). In particular, the sequence between aa 198 and 469 contains significant sequence homology to the DH of many proteins, which exhibit GDP/GTP exchange activity for specific members of the Ras superfamily of GTP-binding proteins (32,33). In vitro, Vav2 has been clearly shown to act as a GDP/GTP exchange protein for the Rho subfamily of GTPases, including Rac1, Cdc42, RhoA, RhoB, and RhoG (27,(37)(38)(39)(40). In addition, Vav2 contains one pleckstrin homology domain (PH), which is commonly detected in signaling molecules and cytoskeletal proteins (41,42). The PH domain may also mediate the association of Vav2 with the submembrane region of the cell via proteinprotein or protein-lipid interactions (41,42). The carboxylterminal region of Vav2 contains an SH2 domain surrounded by two SH3 domains. Although its function is still unclear, this region may facilitate interactions with submembrane signaling complexes.
It has been shown that overexpression of Vav2 is sufficient to activate its oncogenic potential (30). Vav2-induced cytoskeletal changes and cellular transformation are comparable, but not identical, to those induced by other members of the Dbl family of oncoproteins (30). Taken together, these sequence analyses of Vav2 suggest that Vav2 association with the transforming phenotype may be mediated via submembrane cytoskeletal regulation and/or activation of Rho family GTPases. Activation of Rac1, Cdc42, and RhoA has been shown to produce specific structural changes in the plasma membrane associated with cell movement such as lamellipodia, filopodia, and stress fiber formation, respectively (43). The coordinated activation of these GTPases is considered to be a possible mechanism underlying tumor cell motility and migration, an obvious prerequisite for metastasis (9 -11).
The HER2 oncogene (also called c-ErbB-2 or Neu) encodes an 185-kDa (p185 HER2 ) membrane protein, which contains a single transmembrane spanning region, two cystine-rich extracellular domains, and a tyrosine kinase-associated cytoplasmic domain (44). This protein belongs to the epidermal growth factor receptor subgroup of the receptor-linked tyrosine kinase superfamily (45). Overexpression or amplification of HER2 oncogene appears to correlate with poor survival rates of many known cancers including ovarian cancer (46,47). For example, the overall survival rate and time of relapse for patients with HER2 overexpression is significantly less than patients lacking HER2 overexpression (46,47). A number of published reports (46 -48) have demonstrated that HER2 overexpression is a causative factor (and not a consequence) of human cancer. Previous studies in transformed cells have also shown that guanidine nucleotide (GDP/GTP) exchange on Ras is stimulated by tyrosine phosphorylation of p185 HER2 and recruitment of the Sos exchange factor to the plasma membrane with the aid of the Grb2 adaptor protein (48 -54). In addition, dominantnegative mutants of Grb2 induced reversal of the transformed phenotype caused by point mutation-activated Rat HER-2/Neu (55). It appears that activation of p185 HER2 tyrosine kinase leads to Ras-mediated stimulation of a downstream kinase cascade, which at least includes Raf-1/MEK/MAPK and MEKK-1/JNKK/JNK pathways leading to cell growth (48 -55). Previously, we have determined that CD44 and p185 HER2 are physically linked to each other via interchain disulfide bonds in human ovarian tumor cells (SK-OV-3.ipl cell line) (2). Most importantly, HA binding to a CD44-associated p185 HER2 complex activates the p185 HER2 tyrosine kinase activity and promotes ovarian carcinoma cell growth (2). We believe that direct "cross-talk" between the two surface molecules, CD44 and the p185 HER2 , may be one of the most important signaling events in human ovarian carcinoma development.
In this study we have investigated a CD44v3-Vav2 interaction with Grb2-linked p185 HER2 in ovarian tumor cells in order to understand further the fundamental nature of tumorigenesis and metastasis in this lethal gynecological malignancy. By using a variety of biochemical and molecular biological techniques, we have found that both CD44v3 and Vav2 are overexpressed in human ovarian tumor cells (e.g. SK-OV-3.ipl cell line) and that CD44v3 is closely associated with Vav2 (in particular, the SH3-SH2-SH3 domain of Vav2) in SK-OV-3.ipl cells. Most importantly, we have established that the binding of HA to SK-OV-3.ipl cells stimulates CD44v3-specific Rac1 and Ras signaling and association of the linker molecule, Grb2, with CD44v3-Vav2 and p185 HER2 leading to ovarian tumor cell migration and growth. Furthermore, transfection of SK-OV-3.ipl cells with Grb2 mutant cDNAs (⌬N-Grb2cDNA and to a lesser extent the ⌬C-Grb2cDNA) significantly inhibits Grb2 association with CD44v3-Vav2 and p185 HER2 . Further analyses indicate that overexpression of the ⌬N-Grb2 mutant (and to a lesser extent the ⌬C-Grb2) not only significantly impairs both Rac1 and Ras signaling but also inhibits ovarian tumor cell migration and growth. These findings suggest that CD44v3-Vav2 interactions with Grb2-linked p185 HER2 are involved in the stimulation of both Rac1 and Ras signaling leading to the concomitant activation of HA-mediated ovarian tumor cell migration and growth.

MATERIALS AND METHODS
Cell Culture-The SK-OV-3.ipl cell line was established from ascites that developed in a nu/nu mouse given an intraperitoneal injection of SK-OV-3 human ovarian carcinoma cell line (obtained from the American Type Culture Collection) as described previously (56). Cells were grown in Dulbecco's modified Eagle's medium/F-12 medium (Life Technologies, Inc.) supplemented with 10% fetal bovine serum. Cells were routinely serum-starved (and therefore deprived of serum HA) before adding HA. COS-7 cells were obtained from American Type Culture Collection and grown routinely in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum, 1% glutamine, 1% penicillin, and 1% streptomycin.
Antibodies and Various Reagents-Monoclonal rat anti-CD44 antibody (clone, 020; isotype, IgG 2b ; obtained from CMB-TECH, Inc., San Francisco, CA) used in this study recognizes a common determinant of the CD44 class of glycoproteins, including CD44v3 isoform (15,17,56). For the preparation of polyclonal rabbit anti-CD44v3 or rabbit anti-Vav2, specific synthetic peptides (Ϸ15-17 amino acids unique for either CD44v3 sequences or Vav2 sequences) were prepared, respectively, by the Peptide Laboratories of Department of Biochemistry and Molecular Biology using an Advanced Chemtech automatic synthesizer (model ACT350). Conjugated CD44v3 or Vav2 (to hemocyanin or polylysine) was injected into rabbits (or rat) to raise the antibodies. The anti-CD44v3 or anti-Vav2 sera were collected from each bleed and stored at 4°C containing 0.1% azide. Both rabbit anti-CD44v3 and rabbit anti-Vav2 IgGs were prepared using conventional DEAE-cellulose chromatography and were tested to be monospecific (by immunoblot assays). Monoclonal mouse anti-p185 HER2 antibody (c-neu-ab-3) and rat anti-Ras antibody (Y13-259) were purchased from Oncogene Science, Inc. Monoclonal mouse anti-Grb2 antibody was from Transduction Labora-tories. Monoclonal mouse anti-HA1 (hemagglutinin epitope) antibody (clone 12 CA5) and mouse anti-Xpress TM were obtained from Roche Molecular Biochemicals and Invitrogen, respectively. Both rooster comb H and fibronectin were purchased from Sigma. A highly polymerized form (ϳ10 6 daltons) of HA was purified by gel filtration column chromatography using Sephacryl S1000 column as described previously (5).
Reverse Transcriptase-PCR and One-step Cloning-Total RNA was extracted from SK-OV-3.ipl cells using the acid guanidinium thiocyanate-phenol-chloroform technique as described previously (57). Approximately 3 g of total RNA was used with an oligo(dT) primer in a reverse transcriptase system (Promega, Madison, WI) to synthesize cDNA at 42°C for 1 h, using avian myeloblastosis virus-reverse transcriptase. After synthesis of the first strand, PCR of the CD44 cDNA was done by initial melting of the RNA/cDNA at 94°C for 4 min, then 35 cycles of denaturation at 94°C for 30 s, annealing at 60°C for 30 s, and polymerization at 72°C for 1 min. The PCR primers were designed initially to look at and amplify the region involved with alternative splicing. Analysis of the presence of v3-containing isoforms was facilitated by designing PCR primers within exon 5 and v3 (exon 7). Specifically the primers were CD44 exon 5 (exon 5, 5Ј-GCACTTCAGGAGG-TTACATC-3Ј) and CD44 v3(exon 7) (exon 7, 5Ј-CTGAGGTGTCTGTC-TCTTTC-3Ј). The PCR products were one-step-cloned using TA cloning kit (Invitrogen, San Diego, CA) and sequenced by dideoxy sequencing method.
For the detection of Vav2 expression, ϳ3 g of total RNA isolated from SK-OV-3.ipl cells was used for RT-PCR analysis using SUPER-SCRIPT TM ONE-STEP RT-PCR with PLATINUM R Taq system (Life Technologies, Inc.). Two Vav2-specific primers (5Ј-GCGATGTACATC-AATGAAGTTAAACG-3Ј and 5Ј-GCCCTCCTCTTCTACGTACGTT-3Ј) were designed for this study. The reactions were cycled after an initial 2-min denaturation at 94°C followed by 35 cycles of denaturation at 94°C for 30 s, annealing at 60°C for 30 s, and polymerization at 72°C for 2 min. The PCR products were analyzed using a 2.0% agarose gel electrophoresis and visualized by ethidium bromide staining. As controls, RT-PCR was carried out in the absence of SUPERSCRIPT II reverse transcriptase or in the absence of template in the PCR mixture. No amplification products were detected in these control samples.
Cell Surface Labeling Procedures-SK-OV-3.ipl cells suspended in PBS were surface-labeled using the following biotinylation procedure. Briefly, cells (10 7 cells/ml) were incubated with sulfosuccinimidyl-6-(biotinamido)hexanoate (Pierce) (0.1 mg/ml) in labeling buffer (150 M NaCl, 0.1 M HEPES (pH 8.0)) for 30 min at room temperature. Cells were then washed with PBS to remove free biotin. Subsequently, the biotinylated cells were used for anti-CD44v3-mediated immunoprecipitation as described previously (56). These biotinylated materials precipitated by anti-CD44v3 antibody were analyzed by SDS-PAGE, transferred to the nitrocellulose filters, and incubated with ExtrAvidinperoxidase (Sigma). After an addition of peroxidase substrate (Pierce), the blots were developed using ECL chemiluminescence reagent (Amersham Biosciences) according to the manufacturer's instructions.
For detecting the recruitment of Grb2 and p185 HER2 into CD44v3 complex, SK-OV-3.ipl cells (either treated with HA (50 g/ml) or without any HA treatment) were solubilized by Nonidet P-40 (as described above) and immunoprecipitated with anti-CD44v3 antibody followed by anti-Grb2 and/or p185 HER2 -mediated immunoblot plus horseradish peroxidase-conjugated goat anti-mouse IgG (1:10,000 dilution) and ECL chemiluminescence reagent.
In some experiments, SK-OV-3.ipl cells (e.g. untransfected or transfected with HA1-tagged ⌬N-Grb2 cDNA or HA1-tagged ⌬C-Grb2 cDNA or vector only) were immunoblotted with various immunoreagents (e.g. rabbit anti-CD44v3 antibody, rabbit anti-Vav2 antibody, and mouse anti-p185 HER2 antibody) (5 g/ml) for 1 h at room temperature followed by incubation with horseradish peroxidase-conjugated goat anti-rabbit IgG or mouse IgG (1:10,000 dilution) at room temperature for 1 h. Some SK-OV-3.ipl cells (e.g. untransfected or transfected by HA1-tagged ⌬N-Grb2 cDNA or HA1-tagged ⌬C-Grb2 cDNA or vector only) were also immunoprecipitated with anti-HA1 antibody followed by immunoblotting with anti-CD44v3 or anti-Vav2 or p185 HER2 , respectively. These blots were then developed using ECL chemiluminescence reagent (Amersham Biosciences) according to the manufacturer's instructions. During these immunological analyses, an equal amount of cellular protein (50 g/ml) immunoprecipitated with the antibody was applied to SDS-PAGE followed by immunoblot analyses.
Cloning, Expression, and Purification of CD44 Cytoplasmic Domain (CD44cyt) from Escherichia coli-The procedure for preparing the fusion protein of the cytoplasmic domain of CD44 was the same as described previously (12,13,16). Specifically, the cytoplasmic domain of human CD44 (CD44cyt) was cloned into pFLAG-AST using the PCRbased cloning strategy. By using human CD44 cDNA as template, one PCR primer pair (left, FLAG-EcoRI; right, FLAG-XbaI) was designed to amplify the complete CD44 cytoplasmic domain. The amplified DNA fragments were one-step-cloned into a pCR2.1 vector and sequenced. The DNA fragments were then cut out by double digestion with EcoRI and XbaI and subcloned into EcoRI/XbaI double-digested pFLAG-AST (Eastman Kodak Co.) to generate FLAG-pCD44cyt construct. The nucleotide sequence of FLAG/CD44cyt junction was confirmed by sequencing. The recombinant plasmids were transformed to BL21-DE3 to produce FLAG-CD44cyt fusion protein. The FLAG-CD44cyt fusion protein was further purified by anti-FLAG M2 affinity gel column (Kodak). The nucleotide sequences of primers used in this cloning protocol are as follows: FLAG-Ec-oRI, 5Ј-GAGAATTCGAACAGTCGAAGAAGGTGTCTCTTAAGC-3Ј; FLAG-XbaI, 5Ј-AGCTCTAGATTACACCCCAATCTTCAT-3Ј.
⌬N-Grb2 is an amino-terminal SH3 domain deletion of Grb2. ⌬C-Grb2 is a carboxyl-terminal SH3 domain deletion mutant of Grb2. The cDNAs encoding the ⌬N-Grb2 and ⌬C-Grb2 are driven by the cytomegalovirus promoter and preceded by the hemagglutinin epitope (HA1) tag in expression vector pCGN-Bam, which contains the hygromycinresistant gene as a selection marker (55). These HA1-tagged ⌬N-Grb2 cDNA or HA1-tagged ⌬C-Grb2 cDNA were then used for transient expression in SK-OV-3ipl cells as described below. The truncated Grb2 products such as HA1-tagged ⌬N-Grb2 or HA1-tagged ⌬C-Grb2 were expressed as 22 or 20 kDa, respectively, as detected by SDS-PAGE and immunoblot analyses as described above.
Cell Transfection-To establish a transient expression system, various cell lines (e.g. COS-7 cells or SK-OV-3.ipl.cells) were transfected with various plasmid DNAs (e.g. His-tagged SH3-SH2-SH3 cDNA, HA1-tagged ⌬N-Grb2, and ⌬C-Grb2 cDNA or vector alone) using electroporation methods according to those procedures described previously (58). Briefly, cells were plated at a density of 2 ϫ 10 6 cells per 100-mm dish and transfected with 25 g/dish plasmid cDNA using electroporation at 230 V and 960 microfarads with a Gene Pulser (Bio-Rad). Transfected cells were grown in the culture medium for at least 24 -48 h. Various transfectants were then analyzed for their protein expression (e.g. carboxyl-terminal Vav2 fragment and two truncated Grb2 proteins (⌬N-Grb2 or ⌬C-Grb2)) by immunoblot and immunoprecipitation as described above.
In Vitro Binding of CD44cyt to the SH3-SH2-SH3 Vav2 Fragment or Intact Vav2-Aliquots (0.5-1 ng of protein) of purified His-tagged SH3-SH2-SH3 Vav2 fragment (isolated from COS-7 or SK-OV-3.ipl cells) or intact Vav2 (purified from SK-OV-3.ipl cells)-conjugated Sepharose beads were incubated in 0.5 ml of binding buffer (20 mM Tris-HCl (pH 7.4), 150 mM NaCl, 0.1% bovine serum albumin and 0.05% Triton X-100) containing various concentrations (10 -800 ng/ml) of 125 I-labeled cytoplasmic domains of CD44 (CD44cyt) fusion protein (5,000 cpm/ng protein) at 4°C for 4 h. Specifically, equilibrium-binding conditions were determined by performing a time course (1-10 h) of 125 I-labeled CD44cyt binding to the SH3-SH2-SH3 Vav2 fragment or intact Vav2 at 4°C. The binding equilibrium was found to be established when the in vitro CD44-(SH3-SH2-SH3 Vav2 fragment or intact Vav2) binding assay was conducted at 4°C after 4 h. Following binding, the Vav2 fragment/intact Vav2-conjugated beads were washed extensively in binding buffer, and the beads-bound radioactivity was counted. Nonspecific binding was determined using a 50 -100-fold excess of unlabeled CD44cyt in the presence of the same concentration of 125 I-labeled CD44cyt. Nonspecific binding, which was ϳ20% of the total binding, was always subtracted from the total binding. Our binding data are highly reproducible. The values expressed in Fig. 5 represent an average of triplicate determinations of 3-5 experiments with an S.D. less than Ϯ5%.
In some cases, 0.1 g of surface-biotinylated CD44 was incubated with various Vav2-related proteins (e.g. purified intact Vav2 or His-SH3-SH2-SH3-coated beads) in the presence and absence of 100-fold excess amount of His-SH3-SH2-SH3 at room temperature in the binding buffer (20 mM Tris-HCl (pH 7.4), 150 mM NaCl, 0.1% bovine serum albumin and 0.05% Triton X-100) for 1 h. After binding, biotinylated CD44 bound to the beads was analyzed by SDS-PAGE, transferred to the nitrocellulose filters, and incubated with ExtrAvidin-peroxidase (Sigma). After an addition of peroxidase substrate (Pierce), the blots were developed using ECL chemiluminescence reagent (Amersham Biosciences) according to the manufacturer's instructions.
Vav2-mediated GDP/GTP Exchange For Rac1 Proteins-Purified E. coli-derived GST-tagged Rac1 or GST alone (20 pmol) was preloaded with GDP (30 M) in 10 l of buffer containing 25 mM Tris-HCl (pH 8.0), 1 mM dithiothreitol, 4.7 mM EDTA, 0.16 mM MgCl 2 , and 200 g/ml bovine serum albumin at 37°C for 7 min. In order to terminate preloading procedures, additional MgCl 2 was then added to the solution (reaching a final concentration of 9.16 mM) as described previously (16,17,59). Subsequently, 2 pmol of Vav2 (bound to anti-Vav2-conjugated Sepharose beads) isolated from SK-OV-3.ipl cells (untransfected or transfected with ⌬N-Grb2 cDNA or ⌬C-Grb2 cDNA or vector alone) grown in the presence or absence of HA (50 g/ml) (or fibronectin (50 g/ml) or control samples (nonspecific cellular material associated with preimmune serum-conjugated Sepharose beads)) was preincubated with 2.5 M [ 35 S]GTP␥S (Ϸ1,250 Ci/mmol) (in the presence or absence of 2.25 M GTP␥S) for 10 min followed by adding 2.5 pmol of GDPloaded GST-tagged Rac1 (or GDP-treated GST). At various time points, the reaction of each samples was terminated by adding ice-cold termination buffer containing 20 mM Tris-HCl (pH 8.0), 100 mM NaCl, and 10 mM MgCl 2 followed by filtering through nitrocellulose filters. The radioactivity associated with the filters was measured by scintillation fluid as described previously (16,17,59). The amount of [ 35 S]GTP␥S bound to Vav2 (bound to anti-Vav2-Sepharose beads) or control sample (preimmune serum-conjugated Sepharose beads) in the absence of GSTtagged Rac1 was subtracted from the original values. In some cases, Vav2-catalyzed GDT/GTP exchange reactions were measured using SK-OV-3.ipl transfectants (or untransfected cells) preincubated with monoclonal rat anti-CD44 antibody (or normal rat IgG) followed by HA treatment or fibronectin treatment (or no treatment). Data represent an average of triplicates from 3 to 5 experiments. The standard deviation was less than 5%.
Analysis of Ras-bound Guanine Nucleotides-The procedures used in this assay were the same as those described previously (55, 60) with slight modifications. SK-OV-3.ipl cells (ϳ1 ϫ 10 4 cells) were first incubated with phosphate-free medium supplemented with 5% dialyzed serum followed by labeling with [ 32 P]orthophosphate (400 Ci/ml) in the same medium for 12 h. After washing with phosphate-buffered saline (PBS, pH 7.2), labeled cells were then incubated with HA (50 g/ml) for various time intervals (e.g. 0, 2, 5, 10, 15, 20, 30, and 60 min). Subsequently, cells were solubilized in 50 mM HEPES (pH 7.5), 150 mM NaCl, 20 mM MgCl 2 , 0.5% Nonidet P-40, 0.2 mM Na 3 VO 4 , 0.2 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, and 5 g/ml aprotinin followed by anti-Ras-mediated immunoprecipitation plus agarose bead-coated goat anti-rat IgG. The beads were extensively washed with lysis buffer, and proteins were solubilized in 1% SDS at 68°C. The bound guanine nucleotides were chromatographed on polyethyleneimine-cellulose plates in 1.3 M LiCl. The GTP/GDP ratio was determined by a Betascope 603 Blot Analyzer (Betagen, Boston). The percentage of GTP-bound Ras in untransfected SK-OV-3.ipl cells (with no HA treatment control) was designated as 100%. Each assay was set up in triplicate and repeated at least 3 times. All data were analyzed statistically using the Student's t test and statistical significance was set at p Ͻ 0.01.

Binding of [ 3 H]Hyaluronan or 125 I-Fibronectin to SK-OV-3.ipl Cells-[ 3 H]HA was prepared from rat fibrosarcoma cells by D-[ 3 H]glucosamine
labeling as described previously (5). Fibronectin (ϳ2 g of protein) was 125 I-labeled to a specific activity of 60,000 cpm/ng protein using the methods described by Frank and Speck (61). To test ligand binding, SK-OV-3.ipl cells (ϳ4 ϫ 10 5 cells) grown in 6-well culture plates were incubated with various concentrations (0.3-4.8 g/ml) of [ 3 H]HA (1 ϫ 10 5 dpm/g) or 125 I-fibronectin (1 ϫ 10 5 dpm/g) in binding buffer (PBS containing 0.2% bovine serum albumin) at 4°C for 4 h. In some cases, cells were first treated with monoclonal rat anti-CD44 antibody (50 g/ml) at room temperature for 30 min followed by [ 3 H]HA or 125 Ifibronectin binding. Following binding, the cells were washed three times in the binding buffer and solubilized in 20 mM Tris-HCl (pH 7.4) solution containing 0.1% SDS. The radioactivity associated with solubilized cell extracts was counted in a liquid scintillation counter. The nonspecific binding was determined in presence of 100-fold excess of unlabeled HA (or unlabeled fibronectin) and was subtracted from the total binding.
In Vitro Tumor Cell Growth Assays-SK-OV-3.ipl cells (untransfected or transfected with various truncated Grb2cDNAs (e.g. ⌬N-Grb2cDNA or ⌬C-Grb2cDNA) or vector alone) (5 ϫ 10 3 cells/well) were treated with HA (50 g/ml) (or pre-treated with rat anti-CD44 followed by HA treatment (50 g/ml) or untreated). These cells were then plated in 96-well culture plates in 0.2 ml of Dulbecco's modified Eagle's medium/F-12 medium supplement (Life Technologies, Inc.) containing either 0.5% fetal bovine serum or no serum for 24 h at 37°C in 5% CO 2 , 95% air. In each experiment, a total of 5 plates (10 wells/ treatment (e.g. HA treatment or rat anti-CD44 plus HA treatment or no treatment)/plate) was used. Experiments were repeated three times. The in vitro growth of these cells were analyzed by measuring increases in cell number using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assays (CellTiter 96 R non-radioactive cell proliferation assay according to the procedures provided by Promega) (2). Subsequently, viable cell-mediated reaction products were recorded by a Molecular Devices (Spectra Max 250) enzyme-linked immunosorbent assay reader at a wavelength of 450 nm.
Tumor Cell Migration Assays-Twenty four transwell units were used for monitoring in vitro cell migration as described previously (16,17). Specifically, the 5-m porosity polycarbonate filters (CoStar Corp., Cambridge, MA) were used for the cell migration assay. SK-OV-3.ipl cells (ϳ1 ϫ 10 4 cells/well in PBS (pH 7.2) (in the presence or absence of rat anti-CD44 antibody (50 g/ml)) were placed in the upper chamber of the transwell unit. In some cases, cells were transfected with various truncated Grb2cDNAs (e.g. ⌬N-Grb2cDNA and ⌬C-Grb2cDNA) or vector alone. The growth medium containing high glucose Dulbecco's modified Eagle's medium supplemented with 200 g/ml hyaluronic acid was placed in the lower chamber of the transwell unit. After 18 h of incubation at 37°C in a humidified 95% air, 5% CO 2 atmosphere, vital stain 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (Sigma) was added at a final concentration of 0.2 mg/ml to both the upper and the lower chambers and incubated for additional 4 h at 37°C. Migrative cells at the lower part of the filter were removed by swabbing with small pieces of Whatman filter paper. Both the polycarbonate filter and the Whatman paper were placed in dimethyl sulfoxide to solubilize the crystal. Color intensity was measured in 450 nm. Cell migration processes were determined by measuring the cells that migrate to the lower side of the polycarbonate filters by standard cell number counting methods as described previously (16,17). The CD44-specific cell migration was determined by subtracting nonspecific cell migration (i.e. cells migrate to the lower chamber in the presence of rat anti-CD44 antibody treatment) from the total migrative cells in the lower chamber. The CD44-specific cell migration in vector-transfected cells (control) is designated as 100%. Each assay was set up in triplicate and repeated at least 3 times. All data were analyzed statistically using the Student's t test, and statistical significance was set at p Ͻ 0.01.

Expression of CD44v3 and Vav2 In Human Ovarian
Tumor Cells CD44v3 Expression-The expression of CD44 variant (CD44v) isoforms is known to be closely correlated with onco-CD44v3-Vav2 and Grb2-HER2 Interaction in Ovarian Tumor Cells genic signaling and metastatic behavior of a variety of tumor cells including human ovarian tumor cells (2,11,19). The question of what CD44v isoform(s) is expressed in human ovarian tumor cells (e.g. SK-OV-3.ipl cells) is addressed in this study. By using a PCR primer pair to amplify between exon 5 and v3 (exon 7) by RT-PCR, we have detected the presence of one major v3-containing PCR product of Ϸ400 bp (Fig. 1A, lane 2, based on markers indicated in Fig. 1A, lane 1) in SK-OV-3.ipl cells. This v3-containing PCR product was then "one-stepcloned" into the pCRII vector from Invitrogen Corp. (San Diego, CA) and sequenced. The nucleotide sequence confirms that this band belongs to the v3 exon insertion (data not shown). We believe that the RT-PCR is specific since no amplified fragment can be detected in samples incubated without reverse transcriptase (Fig. 1A, lane 3).
In order to characterize CD44 protein expression in the hu-man ovarian tumor cells (SK-OV-3.ipl cell line), we have utilized surface biotinylation techniques followed by immunoprecipitation with a monoclonal rat anti-CD44 antibody (recognizing a common determinant of the CD44 class of glycoproteins, including various variant isoforms) (Fig. 1B, lane 1) or a rabbit anti-CD44v3 antibody (recognizing a v3-specific sequence located at the membrane proximal region of the extracellular domain of CD44) (Fig. 1B, lane 3). Our results indicate that a single surface-biotinylated polypeptide (Ϸ85 kDa) displaying immunological cross-reactivity with CD44 is preferentially expressed on the cell surface of human ovarian tumor cells (Fig. 1B, lanes 1 and 3). In particular, an anti-CD44v3 antibody capable of immunoprecipitating the 85-kDa surface-biotinylated protein (Fig. 1B, lane 3) further confirms that the CD44v3 protein is expressed on the surface of SK-OV-3.ipl cells. No CD44-containing material is observed in control samples when normal rat IgG or preimmune rabbit IgG is used in these experiments (Fig. 1B, lane 2 or lane 4).
To determine whether CD44v3 is involved in HA binding, SK-OV-3.ipl cells were preincubated with a monoclonal rat anti-CD44 antibody that recognizes a common determinant of the CD44 class of glycoproteins including CD44v3 isoform followed by the addition of [ 3 H]HA. As shown in Table I, monoclonal rat anti-CD44 antibody (50 g/ml) (but not normal rat IgG control (50 g/ml)) significantly inhibits [ 3 H]HA binding. In controls, SK-OV-3.ipl cells were also incubated with 125 Ifibronectin. Our data also show that neither rat anti-CD44 antibody nor normal rat IgG is capable of blocking the interaction between certain surface molecule(s) and fibronectin (Table  I). These findings support the notion that CD44v3 is the major HA receptor (but not the fibronectin receptor) on ovarian tumor cells. The v3 (or exon 7) insertion in the CD44v3 isoform has also been shown to contain heparin sulfate addition sites (62) required for the binding of heparin-binding growth factors, cytokines, and chemokines that promote tumor progression (11,63). These observations prompted our laboratory to examine CD44v3-associated oncogenic signaling in human ovarian tumor cells.
Vav2 Expression-Several lines of evidence support the notion that the Vav family of GEFs (dbl molecules) is involved in RhoGTPase activation, cytoskeleton function, and cellular transformation (30,(37)(38)(39)(40). To examine the expression of Vav2 transcripts in human ovarian tumor cells, total RNA from SK-OV-3.ipl cells was isolated and analyzed by RT-PCR using Vav2-specific primer pairs. Employing ethidium bromide staining of agarose gels, we have identified a major PCR product of 1500 bp corresponding to Vav2 (Fig. 2A, lane 2, based on markers indicated in Fig. 2A, lane 1) in SK-OV-3.ipl cells. No amplified fragment is detected in samples incubated without reverse transcriptase (Fig. 2A, lane 3), indicating that the  a SK-OV-3.ipl cells were either untreated (control) or pretreated with rat anti-CD44 IgG (50 g/ml) or normal rat IgG (50 g/ml) at room temperature for 30 min followed by incubating with either [ 3 H]HA (1 g/ml) or 125 I-fibronectin (1 g/ml) at 4°C for 2 h in PBS. The nonspecific binding was determined in the presence of unlabeled HA (100 g/ml), or unlabeled fibronectin (100 g/ml) was subtracted. The specific binding observed in no treatment (control) is designated as 100%.
RT-PCR is specific. Our data clearly demonstrate that the Vav2 transcript is expressed in human ovarian tumor cells.
Since gene expression at the mRNA level does not always correlate with cellular protein expression, it is important to determine whether Vav2 protein expression occurs in the human ovarian tumor cells. Immunoblot analysis, utilizing an anti-Vav2 antibody designed to recognize a Vav2-specific sequence, reveals a single polypeptide (Ϸ95 kDa) (Fig. 2B, lane 2). We have also demonstrated that the Vav2 protein detected by anti-Vav2mediated immunoblot is specific since no protein is detected in these cells incubated with preimmune rabbit IgG (Fig. 2B, lane  1). These results clearly indicate that Vav2 is expressed at both the transcript and protein levels in SK-OV-3.ipl cells.

Association between CD44v3 and Vav2 In Human
Ovarian Tumor Cells Next we addressed the question of whether there is a direct interaction between CD44v3 and Vav2 in human ovarian tu-mor cells. First, we carried out both anti-CD44v3-mediated and anti-Vav2-mediated precipitation followed by anti-Vav2 immunoblot (Fig. 2B, lane 3) or anti-anti-CD44v3 immunoblot (Fig.  2B, lane 4), respectively. Our results demonstrate that the Vav2 band is revealed in anti-CD44v3-immunoprecipitated materials (Fig. 2B, lane 3), and the CD44v3 band is detected in the anti-Vav2-immunoprecipitated materials (Fig. 2B, lane 4). These findings suggest that CD44v3 and Vav2 are physically associated in ovarian tumor cells.
Previous studies have indicated that the SH3-SH2-SH3 domain of Vav2 is involved in protein-protein interactions (64 -67) during growth factor receptor-mediated cellular signaling (68). To determine whether there is a direct interaction between CD44 and Vav2, we have prepared a purified recombinant SH3-SH2-SH3 fragment of Vav2 and a FLAG-tagged cytoplasmic domain of CD44 (FLAG-CD44cyt) fusion protein for in vitro binding analysis with these two molecules. Specifically, we incubated 125 I-labeled FLAG-CD44cyt with the SH3-SH2-SH3 fragment of Vav2 (or an intact Vav2) under equilibrium binding conditions. Scatchard plot analyses presented in Fig. 3 indicate that the SH3-SH2-SH3 fragment of Vav2 binds to the cytoplasmic domain of CD44 (CD44cyt) at a single site with high affinity (an apparent dissociation constant (K d ) of Ϸ1.87 nM) (Fig. 3A). This interaction between CD44 and the SH3-SH2-SH3 fragment of Vav2 is comparable in affinity to CD44 binding to intact Vav2 (K d Ϸ1.98 nM) (Fig. 3B). These findings clearly establish the fact the SH3-SH2-SH3 domain of Vav2 contains the CD44-binding site.

Vav2-catalyzed Rac1 Activation in SK-OV-3.ipl Cells
It has been reported that Vav2 functions as an exchange factor for Rho-like GTPases such as Rac1 (34, 40 -42). To confirm that the Vav2-like molecule in this study functions as a GDP/GTP exchange factor (or a GDP-dissociation stimulator (GDS) protein) for RhoGTPases, we first isolated Vav2 from SK-OV-3.ipl cells using anti-Vav2-conjugated Sepharose beads. As indicated in Fig. 4, the isolated Vav2 activates GDP/GTP exchange on GST-tagged Rac1 (Fig. 4). The initial onset of the exchange reaction on GST-tagged Rac1 occurs within 1-2 min after the addition of Vav2, and the reaction reaches its maximum ϳ10 min after Vav2 addition (Fig. 4b). Most importantly, we have observed that the addition of HA to CD44v3-containing SK-OV-3.ipl cells stimulates both the rate of Vav2-mediated GDP/GTP exchange reaction and the maximal amount of bound [ 35 S]GTP␥S to GST-Rac1 (at least a 1.5-fold increase) (Fig. 4, curve a) as compared with Vav2 isolated from untreated SK-OV-3.ipl cells (Fig. 4, curve b). In control samples, a low level of [ 35 S]GTP␥S-bound material was detected in GST alone under the same GDP/GTP exchange reaction using Vav2 isolated from SK-OV-3.ipl cells in the presence (Fig. 4c) or absence (Fig. 4d) of HA treatment. Therefore, we conclude that Vav2 in SK-OV-3.ipl cells functions as a GDP/GTP exchange factor for Rho-like GTPases such as Rac1GTPase.
It is also noted that the amount of [ 35 S]GTP␥S-bound material detected in these samples isolated from SK-OV-3.ipl cells pretreated with rat anti-CD44 (but not normal rat IgG) followed by HA treatment was significantly reduced (Table  II). Since other extracellular matrix components such as fibronectin fail to promote CD44-specific Vav2-mediated Rac1 activation (Table II), we believe that the binding of HA to CD44v3-containing SK-OV-3.ipl cells is directly involved in Vav2-mediated Rac1 activation.

CD44v3-Vav2 Interaction with the Grb2-p185 HER2 Complex
The adaptor molecule, Grb2, contains a single SH2 domain and two SH3 domains (Fig. 6A) and is a key component in Ras

CD44v3-Vav2 and Grb2-HER2 Interaction in Ovarian Tumor Cells
signaling activated by receptor tyrosine kinases including p185 HER2 (48 -54). Several studies also show that Grb2 is capable of binding to Vav directly (69,70). To examine the possible relationship between the CD44v3-Vav2 complex and other important signaling molecules such as p185 HER2 and Grb2, we have analyzed the anti-CD44v3-mediated immunoprecipitates from cell lysate by immunoblotting with anti-p185 HER2 (Fig. 5B, a) or anti-Grb2 (Fig. 5B, b) antibody, respectively. Our results demonstrate that both p185 HER2 (Fig.  5B, a, left lane) and Grb2 (Fig. 5B, b, left lane) are complexed with the CD44v3 (together with Vav2) (Fig. 5B, c and d, left  lanes). Furthermore, we have observed that HA treatment of SK-OV-3.ipl cells stimulates p185 HER2 tyrosine kinase activity (Fig. 5A) and causes a significant increase in the amount of p185 HER2 (Fig. 5B, a, right lane), Grb2 (Fig. 5B, b, right lane), and Vav2 (Fig. 5B, c, right lane), recruited into the CD44v3associated (Fig. B, d, right lane) signaling complex leading to Ras activation (Table III), tumor cell growth (Table IV), and migration (Table V). These observations strongly suggest that HA interaction with CD44v3-Vav2 complex is capable of assembling other signaling molecules (e.g. p185 HER2 and Grb2) into a large complex for the activation of multiple signaling pathways including both Rac1 and Ras pathways during ovarian tumor cell functions.
Furthermore, we have demonstrated that transfection of SK-OV-3.ipl cells with ⌬C-Grb2 cDNA does not cause any change in HA-dependent, Vav2-catalyzed GDP/GTP exchange reaction on Rac1 (Fig. 7, curves b and d) and Ras activation (Table III) as well as the ovarian tumor cell-specific properties (e.g. cell growth (Table IV) and migration (Table V)), as compared with vector-transfected cells (Fig. 7, curves a and c ; Tables III-V). However, overexpression of ⌬N-Grb2 (by transfection of SK-OV-3.ipl cells with ⌬N-Grb2cDNA) significantly impairs HAmediated Vav2-Rac1 signaling (Fig. 7, curves e and f) and Ras activation (Table III) as well as ovarian tumor cell growth (Table IV) and migration (Table V). These data support the conclusion that CD44-Vav2 interaction with the Grb2-p185 HER2 complex plays an important role in promoting HAdependent Rac1 activation and Ras signaling leading to the concomitant stimulation of human ovarian tumor cell migration and growth. Following binding, beads were washed extensively in binding buffer, and the bead-bound radioactivity was counted. As a control, 125 I-labeled FLAG-CD44cyt was also incubated with uncoated beads to determine the binding observed due to the nonspecific binding of the ligand. Nonspecific binding, which represented ϳ20% of the total binding, was always subtracted from the total binding. Our binding data are highly reproducible. The values expressed under "Results" represent an average of triplicate determinations of 3-5 experiments with an S.D. less than Ϯ5%. A, Scatchard plot analysis of the equilibrium binding data between 125 I-labeled FLAG-CD44cyt and the SH3-SH2-SH3 fragment of Vav2. B, Scatchard plot analysis of the equilibrium binding data between 125 I-labeled FLAG-CD44cyt and intact Vav2.

DISCUSSION
CD44, which exists in a number of different isoforms, has been detected on the surface of ovarian cancer cells causing very strong adhesion to peritoneal mesothelium (24). Overexpression of certain CD44 isoforms on the tumor cell surface is closely associated with the progression of various carcinomas including ovarian cancers (2,19,71,72). By using both RT-PCR (Fig. 1A) and immunoblot analyses (Fig. 1B), we have found that one of the major CD44 isoforms expressed on the surface of human ovarian tumor cells (e.g. SK-OV-3.ipl cell line) is the CD44v3 isoform. It has been speculated that the CD44v3 isoform on epithelial tumor cells may act as surface modulators to facilitate unwanted growth factor receptor-growth factor interactions (11,62,63) and subsequent tumor formation. It is also possible that the CD44v3 isoform interacts with extracellular matrix materials (e.g. hyaluronan (HA)) such that the epithelial tumor cells undergo abnormal growth and migration. Consequently, CD44v3 appears to play an important role in oncogenic signaling leading to tumor cell-specific phenotypes in a variety of human solid neoplasms, particularly those of gynecologic origin such as ovarian cancers (71,72).
CD44v3-mediated oncogenesis is frequently regulated by activation of RhoGTPases such as Rac1 and RhoA (15,17). In particular, Rac1 activation initiates oncogenic signaling pathways that are closely associated with cell shape changes (78,79), actin cytoskeleton reorganization (78, 79), gene expression (80,81), and tumor cell invasion/migration (16,17). It has been well established that HA binding to CD44 promotes Rac1 acti-  a Procedures for measuring CD44v3-associated Vav2-catalyzed GDP/ GTP exchange reaction on GST-Rac1 were described under "Materials and Methods." Each assay was set up in triplicate and repeated at least three times. All data were analyzed statistically by Student's t test, and statistical significance was set p Ͻ 0.01. The values expressed in this table represent an average of triplicate determinations of three to five experiments with an S.D. of less than 5%.

TABLE II Effect of monoclonal rat anti-CD44 antibody on CD44v3-associated Vav2-mediated Rac1 activation in ovarian tumor cells (SK-OV-3.ipl cells)
vation (17,18). In particular, the interaction between the CD44v3 isoform and one of the guanine nucleotide exchange factors, Tiam1, up-regulates Rac1 signaling and cytoskeletonmediated metastatic breast tumor progression (17). In this study our aim was to identify CD44v3-linked molecules in human ovarian tumor cells whose expression correlates with HA-mediated metastatic behavior.
Vav is one of the GEFs that are identified as an oncogene by its ability to up-regulate RhoGTPase activity during malignant transformation (27,(37)(38)(39)(40). Vav was first isolated in the course of gene transfer assays using human tumor DNA (28) and has been shown to activate multiple RhoGTPases (e.g. Rac1, Cdc42, RhoA, RhoB, and RhoG) (27,(37)(38)(39)(40). The expression of Vav (also called Vav1) is restricted to hematopoietic system, whereas Vav2 (a homologue of Vav1) is widely expressed in a number of different cell types (29,30). Our laboratory has demonstrated that Vav2 is expressed at both the RNA ( Fig. 2A) and protein (Fig. 2B) levels in human ovarian tumor cells such as SK-OV-3.ipl cells. Specifically, Vav2 is identified as a 95-kDa protein (Fig. 2B) that is capable of carrying out GDP/GTP exchange on Rac1 (Fig. 4) in a manner similar to Vav2 described in other cell types (34, 40 -42, 65, 66). In contrast, the initial rate of Vav2-catalyzed GDP/GTP exchange on GSTtagged Cdc42 and GST-tagged RhoA is significantly slower than that detected on Rac1 (data not shown). Therefore, we believe that Vav2 isolated from SK-OV-3.opl cells functions as an exchange factor for the Rho-like GTPases such as Rac1 (and to a lesser extent for RhoA and Cdc42).
The Vav2 molecule contains several functional domains including an amino-terminal leucine-rich and acid-rich domain (31), a DH domain (32,33), a PH domain (34), a cysteine-rich zinc-binding domain (31), and two SH3 domains flanking a single SH2 domain (the carboxyl-terminal SH3-SH2-SH3 domain) (35,36). In particular, the DH domain of Vav2 by itself exhibits GDP/GTP exchange activity for specific members of the Ras superfamily of GTP-binding proteins (32, 33) and plays  a Procedures for measuring Ras activation in SK-OV-3.ipl cells transfected with either ⌬C-Grb2cDNA or ⌬N-Grb2cDNA or vector alone were described under "Materials and Methods." Each assay was set up in triplicate and repeated at least 3 times. All data were analyzed statistically by Student's t test and statistical significance was set at p Ͻ 0.01. The values expressed in this table represent an average of triplicate determinations of 3-5 experiments with an S.D. of less than Ϯ 5%. a Procedures for measuring ovarian tumor cell growth in SK-OV-3.ipl cells transfected with either ⌬C-Grb2cDNA or ⌬N-Grb2cDNA or vector alone were described under "Materials and Methods." Each assay was set up in triplicate and repeated at least 3 times. All data were analyzed statistically by Student's t test, and statistical significance was set at p Ͻ 0.01. The values expressed in this table represent an average of triplicate determinations of 3-5 experiments with an S.D. less than Ϯ 5%. a SK-OV-3.ipl cells (Ϸ1 ϫ 10 4 cells/well in phosphate-buffered saline (PBS), pH 7.2) were placed in the upper chamber of the transwell unit. In some cases, SK-OV-3.ipl cells were transfected with either ⌬C-Grb2cDNA-transfected cDNA, ⌬N-Grb2cDNA-transfected cDNA, or vector alone. After 18 h of incubation at 37°C in a humidified 95% air, 5% CO 2 atmosphere, cells on the upper side of the filter were removed by wiping with a cotton swap. Cell migration processes were determined by measuring the cells that migrate to the lower side of the polycarbonate filters containing HA (or no HA) by standard cell number counting assays as described under "Materials and Methods." The CD44-specific cell migration was determined by subtracting non-specific cell migration (i.e. cells migrate to the lower chamber in the presence of anti-CD44v3 antibody treatment) from the total migrative cells in the lower chamber. The CD44-specific cell migration in vectortransfected cells (control) is designated as 100%. Each assay was set up in triplicate and repeated at least three times. All data were analyzed statistically using the Student's t test, and statistical significance was set at p Ͻ 0.01. The values expressed in this table represent an average of triplicate determinations of 3-5 experiments with an S.D. less than Ϯ 5%. an important role in RhoGTPase signaling and cellular transformation (30). A number of laboratories have reported that the carboxyl-terminal SH3-SH2-SH3 domain interacts with a diverse array of proteins (e.g. Grb2 (69,70), heterogeneous ribonuclear proteins K and C (65), nuclear protein Ku-70 (66), and the cytoskeletal protein Zyxin (67)). This sequence is also required for Vav2 targeting to the plasma membrane and activation of c-Jun NH 2 -terminal kinase I (82). At the present time, identification of the membrane protein(s) involved in Vav2 binding remains to be accomplished.
In this study we have presented new evidence for a close, physical interaction between Vav2 and the transmembrane glycoprotein, CD44v3 isoform. First, we have shown that both Vav2 and CD44v3 are physically associated as a complex in vivo (Fig. 2B). By using two recombinant proteins, the carboxylterminal SH3-SH2-SH3 domain of Vav2 and the cytoplasmic domain of CD44, we have demonstrated that the SH3-SH2-SH3 domain of Vav2 is directly involved in the binding to the cytoplasmic domain of CD44 (Fig. 3A). In fact, the binding affinity of CD44 to the carboxyl-terminal SH3-SH2-SH3 fragment is comparable to CD44 binding to intact Vav2 (Fig. 3B). These findings suggest that the SH3-SH2-SH3 fragment of Vav2 is responsible for the recognition of CD44 in vitro. In addition, we have detected that the binding of HA to CD44v3 promotes Vav2-catalyzed Rac1 activation (Fig. 4) and tumor cell migration (Table V). Therefore, we believe that CD44v3 is both a cell membrane attachment site and an activator for Vav2 function. Clearly, these two proteins (CD44v3 and Vav2) are not only structurally linked but also functionally coupled.
Recent studies (83,84) have demonstrated that receptorinduced formation of macromolecular signaling complexes is often mediated by molecular scaffolds and adaptors. In this regard, Vav2 is known to interact with adaptor proteins, such as Grb2 (69,70), that mediate binding to the oncogene product p185 HER2 in transformed cells (49 -52, 55). The question of whether the CD44v3-Vav2 complex is also involved in the communication with other oncogenic signaling components (e.g. p185 HER2 and/or Grb2) in ovarian tumor cells is addressed in this study. The HER-2/Neu (also known as ErbB-2) protooncogene encodes a 185-kDa transmembrane glycoprotein with intrinsic tyrosine activity homologous to the epidermal growth factor receptor (44,45). Amplification/overexpression of the HER-2/Neu oncogene is often associated with human ovarian cancers and poor clinical prognosis in patients (46,47). Previously, we have found that CD44 expression is concomitantly up-regulated with p185 HER2 in a mouse fibroblast cell line (85) and in human ovarian tumor cells (2). Up-regulation of surface CD44 in those p185 HER2 -overexpressed cells results in a dramatic enhancement of HA-mediated cell adhesion (85) and tumor cell growth (2). In this study we have also observed that p185 HER2 is co-expressed with CD44v3 in human ovarian tumor cells (Figs. 5 and 6) and that a close association exists between CD44v3 and p185 HER2 in SK-OV-3.ipl cells (Figs. 5B and 6C). Our previous study has shown that CD44 and p185 HER2 are closely associated with each other in a complex involving interchain disulfide bonds in ovarian tumor cells (2). Several cysteine residues including 6 -7 cysteine residues at the external domain, 1 Cys (aa 286) in the transmembrane region, and 1 Cys (aa 295) have been identified in the cytoplasmic domain of CD44 (86). It is likely that some of these cysteine residues of CD44 are involved in the disulfide linkage with one or more cysteine residues in p185 HER2 . The question of which cysteine residues in human CD44 are directly responsible for forming disulfide linkages with p185 HER2 remains to be determined. The association of p185 HER2 with other surface molecules via disulfide linkages has also been reported previously (87). Although CD44 and p185 HER2 are physically linked via interchain disulfide bonds (2), it is also possible that these two membrane proteins can be complexed together through other linker molecules (e.g. Grb2).
Upon activation of the cells, HER-2/neu is phosphorylated at tyrosine residues in its intracellular domain that provides docking sites for Src homology 2 (SH2) and/or protein tyrosine binding domain-containing proteins (44,45). One of the signaling molecules known to be recruited to the tyrosine phosphorylation sites of HER-2/neu is Grb2, which was originally identified as a growth factor receptor-bound protein (known as a mammalian homolog of Caenorhabditis elegans Sem5 and Drosophila Drk) (88 -90). Grb2 is a 24-kDa adaptor protein containing an SH2 domain flanked by two SH3 domains (Fig. 6A). Through the SH3 domains, Grb2 is constitutively associated with Sos (named for the Sons of Sevenless gene), a 150-kDa GEF for Ras, by targeting the proline-rich motif as its carboxyl terminus (48 -55). Furthermore, membrane relocalization of Sos by recruitment of the Grb2-Sos complex causes Ras activation (48 -55). Ras activation, in turn, stimulates a downstream kinase cascade, the Raf-1/MEK/MAPK pathway leading to cell growth (48 -54). Ras-activated MAPK (in particular, ERK1 and ERK2) also plays an important role in modulating cytoskeleton function and/or influencing cell motility (91,92). Thus, the Ras-MAPK signaling pathways appear to be involved in the onset of multiple functions including growth and/or motility. The fact that a dominant-negative mutant of Grb2, which impairs the association of Grb2 with p185 HER2 , inhibits Ras activation and induces reversal of the transformed phenotypes caused by p185 HER2 suggests that Grb2 is responsible for the p185 HER2 -induced oncogenesis (55). We have now demonstrated that Grb2 is also expressed in SK-OV-3.ipl cells (Fig.  5B) and that it is associated with a CD44v3 and p185 HER2containing signaling complex in the ovarian tumor cells (Fig.  5B). HA binding to CD44v3 stimulates p185 HER2 tyrosine kinase ( Fig. 5A) which, in turn, recruits Grb2 into the CD44v3-p185 HER2 signaling complex (Fig. 5B). Grb2 has been shown to bind directly to p185 HER2 (55). Therefore, it is possible that HA/CD44-activated p185 HER2 is involved in the recruitment of Grb2 into the CD44v3-p185 HER2 signaling complex (Fig. 5B). A number of other Ras GEFs have also been reported. For example, C3G (known as Crk SH3-binding GEF) can activate Ras in yeast via the binding of its proline-rich domain to the aminoterminal domain of the adaptor protein, Crk (93,94). The Crk-C3G complex may thus mimic the Grb2-Sos complex and couple the oncogenic signal of the mutation-activated p185 HER2 to Ras (93,94). Therefore, we cannot rule out the possibility that Crk-C3G complex is also involved in Ras activation during HA-stimulated CD44v3/p185 HER2 signaling events. Identification of the immediate upstream activator(s) for Ras during CD44v3/p185 HER2 -activated oncogenic events is currently under investigation in our laboratory.
It has been reported that Grb2 may have other cellular functions in addition to linking receptor tyrosine kinases to the Ras signaling pathway. Microinjection studies with anti-Grb2 antibodies in rat kidney cells suggest that Grb2 may play a role in signaling from receptor tyrosine kinase to the small GTPbinding protein Rac (93). Also, the Grb2-SH3 domains have been shown to bind several proteins besides Sos1, including Vav (69, 70), p85 (94), Sos2, and other proteins (p228, p140, p55, and p28) (95-98). The multiple Grb2-SH3 domain inter-actions may mediate novel cellular functions of this adaptor protein. In this study we have observed that endogenous Grb2 is complexed with CD44v3-Vav2 and p185 HER2 in a signaling complex and that HA treatment induces recruitment of Grb2 (together with p185 HER2 ) into CD44v3-containing multimolecular complexes (Fig. 5B) leading to the co-activation of Rac1 (Figs. 4 and 7) and Ras signaling (Table III) and ovarian tumor cell growth (Table IV) and migration (Table V). These observations strongly suggest that Grb2 not only plays an important role in receptor tyrosine kinase (p185 HER2 )-mediated Ras signaling pathway (Table III) (15,16,17,18), but also participates in CD44v3-Vav2-mediated Rac1 signaling pathway (Figs. 4 and  7). Although Grb2 is known to bind to Vav2 (69,70), the important question of whether Grb2 interacts with CD44 directly awaits further investigation. The fact that transfection of SK-OV-3.ipl cells with ⌬N-Grb2cDNA (but not ⌬C-Grb2cDNA) not only abolishes Grb2 association with CD44v3-Vav2 complexes (Fig. 6), but also inhibits its binding to p185 HER2 (Fig. 6), suggests that the amino-terminal SH3 domain of Grb2 (not the carboxyl-terminal SH3 domain of Grb2) is directly involved in the assembly of these diverse signaling molecules containing CD44v3-Vav2 and p185 HER2 . Most importantly, overexpression of the dominant-negative mutant protein, ⌬N-Grb2 (and to a lesser extent the ⌬C-Grb2), inhibits HA-dependent tumor cell behaviors (e.g. Rac1 (Fig. 7), Ras activation (Table III), tumor cell growth (Table IV), and migration (Table V)). These observations further support the conclusion that interfering with the interaction between these diverse signaling molecules (containing CD44v3-Vav2 and p185 HER2 and the endogenous Grb2) by ⌬N-Grb2 impairs oncogenic signaling (e.g. Rac1 and Ras activation) and ovarian tumor cell behaviors (e.g. tumor cell migration and growth). Our results are consistent with previous FIG. 8. A proposed model for the interaction between CD44v3-Vav2 and Grb2-p185 HER2 during oncogenic signaling and ovarian tumor progression. The binding of HA to CD44v3 isoform (containing the v3 exon-encoded structure) induces CD44v3 interaction with Vav2 which, in turn, activates RhoG-TPase (Rac1) signaling leading to cytoskeleton activation and tumor cell migration. HA binding to the CD44v3-p185 HER2 complex (formed by a disulfide linkage) results in p185 HER2 tyrosine kinase (TK) activation, which recruits the adaptor molecule, Grb2, into the CD44v3-p185 HER2 complex. The ability of Grb2 (complexed with CD44v3 and p185 HER2 ) to stimulate Ras signaling, and to interact with Vav2 for Rac1 signaling, promotes the concomitant onset of tumor cell growth and migration leading to ovarian tumor progression. findings indicating that dominant-negative mutants of Grb2 lacking the amino-(⌬N-Grb2) or carboxyl-(⌬C-Grb2) terminal SH3 domains can reverse p185-activated cell transformation (55).
In summary, we would like to propose that CD44v3-Vav2 interaction with Grb2-p185 HER2 forms an important signaling complex that plays a pivotal role in promoting direct cross-talk between the Rac1 signaling pathway and the Ras signaling pathway leading to the concomitant onset of ovarian tumor cell migration and growth during ovarian cancer progression (Fig.  8). Under physiologic conditions, hyaluronan often exists as a high molecule polymer (Ͼ10 6 daltons). However, under the condition of tumorigenesis, generation of hyaluronan fragments may occur. Both HA and certain size HA fragments (3-25 disaccharide units) have been found to promote CD44specific cell proliferation (5) and interleukin-8 gene expression (76). The question of whether certain small size HA fragments are capable of promoting the same CD44v3-Vav2 and Grb2-p185 HER2 oncogenic signaling and ovarian tumor cell migration/growth as the native large molecular weight hyaluronan used in this study is currently undergoing investigation in our laboratory.