CD44 Interaction with Tiam1 Promotes Rac1 Signaling and Hyaluronic Acid-mediated Breast Tumor Cell Migration*

In this study we have explored the interaction between CD44 (the hyaluronic acid (HA)-binding receptor) and Tiam1 (a guanine nucleotide exchange factor) in metastatic breast tumor cells (SP1 cell line). Immunoprecipitation and immunoblot analyses indicate that both the CD44v3 isoform and the Tiam1 protein are expressed in SP1 cells and that these two proteins are physically associated as a complex in vivo. Using an Escherichia coli-derived calmodulin-binding peptide-tagged Tiam1 fragment (i.e. the NH2-terminal pleckstrin homology (PHn) domain and an adjacent protein interaction domain designated as PHn-CC-Ex, amino acids 393–738 of Tiam1) and an in vitrobinding assay, we have detected a specific binding interaction between the Tiam1 PHn-CC-Ex domain and CD44. Scatchard plot analysis indicates that there is a single high affinity CD44 binding site in the PHn-CC-Ex domain of Tiam1 with an apparent dissociation constant (K d ) of 0.2 nm, which is comparable with CD44 binding (K d = ∼0.13 nm) to intact Tiam1. These findings suggest that the PHn-CC-Ex domain is the primary Tiam1-binding region for CD44. Most importantly, the binding of HA to CD44v3 of SP1 cells stimulates Tiam1-catalyzed Rac1 signaling and cytoskeleton-mediated tumor cell migration. Transfection of SP1 cells with Tiam1cDNA promotes Tiam1 association with CD44v3 and up-regulates Rac1 signaling as well as HA/CD44v3-mediated breast tumor cell migration. Co-transfection of SP1 cells with PHn-CC-Ex cDNA and Tiam1 cDNA effectively inhibits Tiam1 association with CD44 and efficiently blocks tumor behaviors. Taken together, we believe that the linkage between CD44v3 isoform and the PHn-CC-EX domain of Tiam1 is required for HA stimulated Rac1 signaling and cytoskeleton-mediated tumor cell migration during breast cancer progression.

The transmembrane glycoprotein CD44 isoforms are all major hyaluronic acid (HA) 1 cell surface receptors that exist on many cell types, including macrophages, lymphocytes, fibroblasts, and epithelial cells (1)(2)(3)(4)(5)(6). Because of their widespread occurrence and their role in signal transduction, CD44 isoforms have been implicated in the regulation of cell growth and activation as well as cell-cell and cell-extracellular matrix interactions (1)(2)(3)(4)(5)(6)(7). One of the distinct features of CD44 isoforms is the enormous heterogeneity in the molecular masses of these proteins. It is now known that all CD44 isoforms are encoded by a single gene that contains 19 exons (8). Of the 19 exons, 12 exons can be alternatively spliced (8). Most often, the alternative splicing occurs between exons 5 and 15, leading to an insertion in tandem of one or more variant exons (v1-v10 (exon 6-exon 14) in human cells) within the membrane-proximal region of the extracellular domain (8). The variable primary amino acid sequence of different CD44 isoforms is further modified by extensive N-and O-glycosylations and glycosaminoglycan additions (9 -12). In particular, CD44v3-containing isoforms have a heparin sulfate addition at the membrane-proximal extracellular domain of the molecule that confers the ability to bind heparin sulfate-binding growth factors (9,10). Cell surface expression of CD44v isoforms changes profoundly during tumor metastasis, particularly during the progression of various carcinomas including breast carcinomas (13)(14)(15)(16)(17). In fact, CD44v isoform expression has been used as an indicator of metastasis.
It has been shown that interaction between the cytoskeletal protein, ankyrin, and the cytoplasmic domain of CD44 isoforms plays an important role in CD44 isoform-mediated oncogenic signaling (6,18,19). Specifically, the ankyrin-binding domain (e.g. NGGNGTVEDRKPSEL between amino acids 306 and 320 in the mouse CD44 (20) and NSGNGAVEDRKPSGL amino acids 304 and 318 in human CD44 (21)) is required for the recruitment of Src kinase and the onset of tumor cell transformation (21). Furthermore, HA binding to CD44 stimulates a concomitant activation of p185 HER2 -linked tyrosine kinase (linked to CD44s via a disulfide linkage) and results in a direct cross-talk between two different signaling pathways (e.g. proliferation versus motility/invasion) (22). In tumor cells, the transmembrane linkage between CD44 isoform and the cytoskeleton promotes invasive and metastatic-specific tumor phenotypes (e.g. matrix degradation (matrix metalloproteinases) activities (23,24), "invadopodia" formation (membrane projections), tumor cell invasion, and migration) (23). These findings strongly suggest that the interaction between CD44 isoform and the cytoskeleton plays a pivotal role in the onset of oncogenesis and tumor progression.
The Rho family proteins (e.g. Rho, Rac, and Cdc42) are members of the Ras superfamily of GTP-binding proteins structurally related to but functionally distinct from Ras itself (25,26). They are associated with changes in the membrane-linked cytoskeleton (26). For example, activation of RhoA, Rac1, and Cdc42 have been shown to produce specific structural changes in the plasma membrane-cytoskeleton reorganization leading to membrane ruffling, lamellipodia, filopodia, and stress fiber formation (26). The coordinated activation of these GTPases is considered to be a possible mechanism underlying cell motility, an obvious prerequisite for metastasis (27)(28)(29). In particular, Rac1 activation is known to initiate oncogenic signaling pathways that promote cell shape changes (33,34), influence actin cytoskeleton organization (33,34), and stimulate gene expression (35)(36)(37). The question of whether Rac1 activation is also involved in CD44v3-related cytoskeleton function that results in the metastatic phenotypes (e.g. tumor cell migration) of breast tumor cells remains to be answered.
Tiam1 (T lymphoma invasion and metastasis 1) has been identified as an oncogene because of its ability to activate Rho-like GTPases during malignant transformation (38,39). Specifically, Tiam1 is capable of activating Rac1 in vitro as a guanine nucleotide exchange factor and inducing membrane cytoskeleton-mediated cell shape changes, cell adhesion, and cell motility (34, 40 -42). It also acts as a Rac-specific guanine nucleotide exchange factor in vivo and induces an invasive phenotypes in lymphoma cells (40). These findings have prompted several research groups to investigate the mechanisms involved in the regulation of Tiam1. For example, addition of certain serum-derived lipids (e.g. sphingosine-1-phosphate and lysophosphatidic acid) to T-lymphoma cells promotes Tiam1-mediated Rac1 and Cdc42 signaling and T-lymphoma cell invasion (43). Tiam1 has also been found to be phosphorylated by protein kinase C in Swiss 3T3 fibroblasts stimulated by lysophosphatidic acid (44) and platelet-derived growth factor (45). Most recently, Exton and co-workers (46) demonstrate that phosphorylation of Tiam1 by Ca 2ϩ /calmodulin-dependent protein kinase II (but not protein kinase C) regulates Tiam1catalyzed GDP/GTP exchange activity in vitro. These findings support the notion that posttranslational modifications of Tiam1 by certain serine/threonine kinase(s) during surface receptor-mediated activation may play an important role in Tiam1-Rac1 signaling. Tiam1 transcript has been detected in breast cancer cells (39). However, it is not known at the present time whether there is any structural and functional relationship(s) between Tiam1-Rac1 signaling and CD44v3-mediated invasive and metastatic processes of breast cancer cells.
In this paper, using a variety of biochemical, molecular biological, and immunocytochemical techniques, we have found that the cell adhesion molecule, CD44v 3 isoform, which binds directly to HA, is closely associated with Tiam1 (in particular, the NH 2 -terminal pleckstrin homology (PHn), a putative coiled coil region (CC), and an additional adjacent region (Ex), designated as PHn-CC-Ex domain of Tiam1) in SP1 breast tumor cells. Most importantly, HA binding to CD44v 3 isoform stimulates Tiam1-specific GDP/GTP exchange for Rho-like GTPases such as Rac1 and promotes cytoskeleton-mediated tumor cell migration. These findings suggest that a transmembrane interaction between CD44v3 and Tiam1 plays an important role in promoting oncogenic signaling and tumor cell-specific phenotypes required for HA-mediated breast tumor cell migration.

MATERIALS AND METHODS
Cell Culture-Mouse breast tumor cells (e.g. SP1 cell line) (provided by Dr. Bruce Elliott, Department of Pathology, and Biochemistry, Queen's University, Kingston, ON, Canada) were used in this study. Specifically, SP1 cell line was derived from a spontaneous intraductal mammary adenocarcinoma that arose in a retired female CBA/J breeder in the Queen's University animal colony. These cells were capable of inducing lung metastases by sequential passage of SP1 cells into mammary gland (47). These cells were cultured in RPMI 1640 medium supplemented with 5-7% fetal calf serum, folic acid (290 mg/ liter), and sodium pyruvate (100 mg/liter). 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 Reagents-For the preparation of polyclonal rabbit anti-Tiam1 antibody or rabbit anti-CD44v3 antibody, specific synthetic peptides (Ϸ15-17 amino acids unique for the COOH-terminal sequence of Tiam1 or the CD44v3 sequence) were prepared by the Peptide Laboratories of Department of Biochemistry and Molecular Biology using an Advanced Chemtech automatic synthesizer (model ACT350). These Tiam1-related or CD44v3-related polypeptides were conjugated to polylysine and subsequently injected into rabbits to raise the antibodies, respectively. The anti-Tiam1-specific or anti-CD44v3-specific antibody was collected from each bleed and stored at 4°C containing 0.1% azide. The anti-Tiam1 IgG or anti-CD44v3 IgG fraction was prepared by conventional DEAE-cellulose chromatography, respectively. Mouse monoclonal anti-HA (hemagglutinin epitope) antibody (clone 12 CA5) was purchased from Roche Molecular Biochemicals. Mouse monoclonal anti-green fluorescent protein (GFP) was purchased from PharMingen. Escherichia coli-derived GST-tagged Rac1 was kindly provided by Dr. Richard A. Cerione (Cornell University, Itheca, NY).
Cell Surface Labeling Procedures-SP1 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 (23). These biotinylated materials precipitated by anti-CD44v3 antibody were analyzed by SDS-polyacrylamide gel electrophoresis, 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 Pharmacia Biotech) according to the manufacturer's instructions.
Immunoprecipitation and Immunoblotting Techniques-SP1 cells were solubilized in 50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1% Triton X-100 buffer and immunoprecipitated using rabbit anti-CD44v3 antibody or rabbit anti-Tiam1 antibody followed by goat anti-rabbit IgG, respectively. The immunoprecipitated material was solubilized in SDS sample buffer, electrophoresed, and blotted onto the nitrocellulose. After blocking nonspecific sites with 3% bovine serum albumin, the nitrocellulose filter was incubated with rabbit anti-Tiam1 antibody (5 g/ml) or rabbit anti-CD44v3 antibody (5 g/ml), respectively, for 1 h at room temperature followed by incubation with horseradish peroxidaseconjugated goat anti-rabbit IgG (1:10,000 dilution) at room temperature for 1 h. The blots were developed using ECL chemiluminescence reagent (Amersham Pharmacia Biotech) according to the manufacturer's instructions.
In some experiments, SP1 cells or COS cells (e.g. untransfected or transfected by various Tiam1 cDNAs including the full-length mouse Tiam1cDNA (FL1591) or HA-tagged NH 2 -terminally truncated C1199 Tiam1cDNA or GFP-tagged PHn-CC-ExcDNA or C1199Taim1cDNA plus GFP-tagged PHn-CC-ExcDNA (as co-transfection) or vector only) were immunoblotted with mouse anti-HA antibody (5 g/ml) or anti-GFP antibody (5 g/ml), respectively, for 1 h at room temperature followed by incubation with horseradish peroxidase-conjugated goat anti-mouse IgG or goat anti-mouse IgG (1:10,000 dilution) at room temperature for 1 h. The blots were developed using ECL chemiluminescence reagent (Amersham Pharmacia Biotech) according to the manufacturer's instructions.
Cloning, Expression, and Purification of CD44 Cytoplasmic Domain (CD44cyt) from E. coli-The procedure for preparing the fusion protein of the cytoplasmic domain of CD44 was the same as described previously (48). Specifically, the cytoplasmic domain of human CD44 (CD44cyt) was cloned into pFLAG-AST using the PCR-based cloning strategy. Using human CD44 cDNA as template, one PCR primer pair (left, FLAG-EcoRI; right, FLAG-XbaI) was designed to amplify complete CD44 cytoplasmic domain. The amplified DNA fragments were one-step cloned into a pCR2.1 vector and sequenced. Then the DNA fragments were cut out by double digestion with EcoRI and XbaI and subcloned into EcoRI/XbaI double-digested pFLAG-AST (Eastman Kodak Co., Rochester, NY) 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 (Eastman Kodak Co.). The nucleotide sequence of primers used in this cloning protocol are: FLAG-EcoRI, 5Ј-GAGAATTCGAACAGTCGAA-GAAGGTGTCTCTTAAGC-3Ј, and FLAG-XbaI, 5Ј-AGCTCTAGATTA-CACCCCAATCTTCAT-3Ј.
Expression Constructs-Both the full-length mouse Tiam1cDNA (FL1591) and the NH 2 -terminally truncated Tiam1cDNA (C1199) were kindly provided by Dr. John G. Collard (The Netherlands Cancer Institute, Amsterdam, The Netherlands). Specifically, the full-length Taim1 (FL1591) cDNA was cloned into the eukaryotic expression vector, pMT2SM. The NH 2 -terminally truncated C1199 Tiam1 cDNA (carrying a HA epitope tag at the 3Ј end) was cloned into the eukaryotic expression vector, pUTSV1 (Eurogentec, Belgium). The Tiam1 fragment, PHn-CC-Ex domain was cloned into calmodulin-binding peptide (CBP)tagged vector (pCAL-n) (Stratagen) using the PCR-based cloning strategy. Using human Tiam1 cDNA as a template, PHn-CC-Ex domain was amplified by PCR with two specific primers (left, 5Ј-AACTCGAGAT-GAGTACCACCAACAGTGAG-3Ј, and right, 5Ј-AAAAAGCTTTCAGC-CATCTGGAACAGTGTCATC-3Ј) linked with specific enzyme digestion site (XhoI or HindIII). PCR product digested with XhoI and HindIII was purified with QIAquick PCR Purification Kit (Qiagen). The PHn-CC-Ex domain cDNA fragment was cloned into pCAL-n vector digested with XhoI and HindIII. The inserted PHn-CC-Ex domain sequence was confirmed by nucleotide sequencing analyses. The recombinant plasmids were transformed to BL21-DE3 to produce CBP-tagged PHn-CC-Ex fusion protein. This fusion protein was purified from bacteria lysate by calmodulin affinity resin column (Sigma). The PHn-CC-Ex domain cDNA fragment was also cloned into pEG-FPN1 vector (CLONTECH) digested with XhoI and HindIII to create GFP-tagged PHn-CC-Ex cDNA. The inserted PHn-CC-Ex domain sequence was confirmed by nucleotide sequencing analyses. This GFPtagged PHn-CC-Ex domain cDNA was then used for transient expression in SP1 cells as described below. The molecular mass of the GFPtagged PHn-CC-Ex is expressed as 68 kDa in SP1 or COS-7 cells by SDS-polyacrylamide gel electrophoresis and immunoblot analyses.
Cell Transfection-To establish a transient expression system, SP1 cells (or COS-7 cells) were transfected with various plasmid DNAs (e.g. HA-tagged C1199 Tiam1cDNA, GFP-tagged PHn-CC-ExcDNA, or HAtagged C1199Tiam1cDNA plus GFP-tagged PHn-CC-ExcDNA (as cotransfection) or vector alone) using electroporation methods according to those procedures described previously (74). Briefly, SP1 cells were plated at a density of 2 ϫ 10 6 cells/100-mm dish and transfected with 25 g/dish plasmid cDNA using electroporation at 230 V and 960 microfaraday 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. Tiam1-related proteins) by immunoblot, GDP/GTP exchange reaction on Rac1, and tumor cell migration assays as described below.
In Vitro Binding of Tiam1/Tiam1 Fragment to CD44 -Aliquots (0.5-1 ng of protein) of purified FLAG-CD44cyt fusion protein bound to Anti-FLAG M2 antibody immunoaffinity 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 intact Tiam1 (purified from SP1 cells) (5000 cpm/ng protein) or 125 I-labeled recombinant Tiam1 fragment (CBP-tagged PHn-CC-Ex) at 4°C for 4 h. Specifically, equilibrium binding conditions were determined by performing a time course (1-10 h) of 125 I-labeled Tiam1 (or CBP-tagged PHn-CC-Ex) binding to CD44 at 4°C. The binding equilibrium was found to be established when the in vitro Tiam1 (or PHn-CC-Ex)-CD44 binding assay was conducted at 4°C after 4 h. Following binding, the immunobeads were washed extensively in binding buffer, and the bead-bound radioactivity was counted. Nonspecific binding was determined using a 50 -100-fold excess of unlabeled Tiam1 (or PHn-CC-Ex) in the presence of the same concentration of 125 I-labeled Tiam1 or 125 I-labeled CBPtagged PHn-CC-Ex. Nonspecific binding, which was approximately 20% 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 three to five experiments with a standard deviation less than Ϯ 5%.
In some cases, 0.1 g of surface biotinylated CD44v3 was incubated with various Tiam1-related proteins (e.g. purified intact Tiam1, HAtagged C1199, CBP-PHn-CC-Ex, or HA/CBP-coated beads) in the presence and absence of 100-fold excess amount of CBP-PHn-CC-Ex 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 CD44v3 bound to the beads was analyzed by SDS-polyacrylamide gel electrophoresis, 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 Pharmacia Biotech) according to the manufacturer's instructions.
Tiam1-mediated GDP/GTP Exchange for Rac1 Proteins-Purified E. coli-derived GST-tagged Rac1 (20pmol) 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. To terminate preloading procedures, additional MgCl2 was then added to the solution (reaching a final concentration of 9.16 mM) as described previously (40,49). Subsequently, 2 pmol of Tiam1 (in anti-Tiam1 (or anti-HA or anti-CD44v3)-Sepharose bead-conjugated forms) isolated from COS-7 cells (transfected with either the full-lengh Tiam1cDNA or NH 2 -terminally truncated Tiam1cDNA) or SP1 cells (transfected with various plasmid DNAs such as HA-tagged C1199 Tiam1cDNA, GFP-tagged PHn-CC-ExcDNA or HA-tagged C1199Tiam1cDNA plus GFP-tagged PHn-CC-ExcDNA (as co-transfection) or vector alone, grown in the presence or absence of hyaluronic acid (100 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 GDP-loaded GST-tagged Rac1GTPase as described previously) (49). The amount of [ 35 S]GTP␥S bound to Tiam1 (conjugated to anti-Tiam1-Sepharose beads) or control sample (preimmune serum-conjugated Sepharose beads) in the absence of Rac1GTPase was subtracted from the original values. Data represent an average of triplicates from three to five experiments. The standard deviation was less than 5%.
Double Immunofluorescence Staining-SP1 cells (transfected with various plasmid DNAs such as HA-tagged C1199 Tiam1cDNA, GFPtagged PHn-CC-ExcDNA, or HA-tagged C1199Tiam1cDNA plus GFPtagged PHn-CC-ExcDNA (as co-transfection) or vector alone) were first washed with PBS buffer (0.1 M phosphate buffer (pH 7.5) and 150 mM NaCl) and fixed by 2% paraformaldehyde. Subsequently, SP1 transfectants were stained with rhodamine (Rh)-labeled rabbit anti-CD44v3 antibody. In some cases, Rh-labeled cells were then rendered permeable by ethanol treatment followed by incubating with FITC-conjugated mouse anti-HA IgG. To detect nonspecific antibody binding, Rh-CD44v3-labeled cells were incubated with FITC-conjugated normal mouse IgG. No labeling was observed in such control samples. The fluorescein-and rhodamine-labeled samples were examined with a confocal laser scanning microscope (MultiProbe 2001 Inverted CLSM system, Molecular Dynamics, Sunnyvale, CA).
Cell Adhesion Assay-SP1 cells were metabolically labeled with Tran 35 S label (20 Ci/ml) as described above. After labeling, the cells were washed in PBS and incubated in PBS containing 5 mM EDTA at 37°C to obtain a nonadherent single cell suspension. Labeled cells (Ϸ9.1 ϫ 10 5 cpm/10 5 cells) (in the presence or absence of anti-CD44v3 antibody) were plated on the HA-coated plates at 4°C for 30 min. Following incubation, the wells were washed three times in PBS, the adherent cells were solubilized in PBS containing 1% SDS, and the well bound radioactivity was determined by liquid scintillation counting. Nonspecific binding was determined by including 300 g/ml soluble HA during the incubation of cells on HA-coated wells. The nonspecific binding was 10 -15% of the total well-associated radioactivity and has been subtracted.
Tumor Cell Migration Assays-Twenty-four transwell units were used for monitoring in vitro cell migration as described previously (23). Specifically, the 5-m porosity polycarbonate filters (CoStar Corp., Cambridge, MA) were used for the cell migration assay. SP1 cells (Ϸ1 ϫ 10 4 cells/well in PBS, pH 7.2) (in the presence or absence of anti-CD44v3 antibody (50 g/ml)) were placed in the upper chamber of the transwell unit. In some cases, cells were transfected with either C1199Tiam1cDNA, PHn-CC-ExcDNA, C1199Tiam1cDNA plus PHn-CC-ExcDNA, 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, the 3-(4,5-dimethyl thiazol-2-yl)-2,5-diphenyl tetrazolium bromide (Promega) was added at a final concentration of 0.2 mg/ml to both the upper and the lower chambers and incubated for an 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 (23,49). The CD44-specific cell migration was determined by subtracting nonspecific 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 vector-transfected 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.

Identification of CD44 Variant Isoform(s) as HA Receptor(s) in SP1 Cells-The expression of CD44 variant isoforms such as
CD44v3 is known to be closely correlated with metastatic and proliferative behavior of a variety of tumor cells including various carcinomas such as human breast tumor cells (14 -19). Immunoblotting with anti-CD44v3 antibody (recognizing the v3-specific sequence located at the membrane-proximal region of the extracellular domain of CD44) indicates that a single CD44v3 protein (molecular mass ϭ ϳ260 kDa) is expressed in SP1 cells (Fig. 1, lane 1). Furthermore, we have utilized surface biotinylation techniques and anti-CD44v3-mediated immunoprecipitation to characterize this CD44v3 molecule. Our results show that the 260-kDa CD44v3 molecule can be surface-biotinylated and is located on the surface of SP1 cells (Fig. 1, lane 3). No CD44v 3 -containing material is observed in control samples when preimmune rabbit serum is used in either immunoblot (Fig. 1, lane 2) or immunoprecipitation experiments (Fig. 1,  lane 4). Further analyses using reverse transcriptase-PCR, cloning, and nucleotide sequence techniques indicate that this CD44v3 belongs to the CD44v 3,8 -10 isoform in SP1 cells (data not shown). This CD44v 3,8 -10 variant exon structure was previously identified in human breast carcinoma samples (14 -19), and its molecular mass (expressed at the protein level) has been shown to be Ϸ260 kDa (9).
CD44 is the major hyaluronan cell surface receptor (50), and a cellular adhesion molecule in many different cell types (51). Specific HA-binding motifs have been identified and localized in the extracellular domain of all CD44 isoforms (52,53). To determine whether HA promotes cell adhesion, breast tumor cells (SP1 cell line) were incubated with plastic dishes coated with HA. As shown in Table I, SP1 cells adhere to the HAcoated dishes very well. In addition, because preincubation with anti-CD44v3 antibody blocks the adhesion of SP1 cells to HA-coated dishes, these data clearly indicate that CD44v3 isoform involves a specific binding interaction with the extracellular matrix component such as HA and is a cell surface adhesion molecule in SP1 cells.
Analysis of a Complex Formed between CD44v3 and Tiam1 in SP1 Cells in Vivo-Both CD44v isoforms (14 -19) and Tiam1 (39) have been detected in a variety of tumor cells. In this study we have addressed the question of whether there is an interaction between CD44v3 isoform and Tiam1 in breast tumor cells (e.g. SP1 cells). First, we have analyzed Tiam1 expression (at the protein level) in breast tumor cells such as SP-1 cell line. Immunoblot analysis, utilizing anti-Tiam1 antibody designed to recognize the specific epitope located at the COOH-terminal sequence of Tiam1 reveals a single polypeptide (molecular mass ϭ ϳ200 kDa) (Fig. 2, lane 2). We have demonstrated that Tiam1 detected in SP1 cells revealed by anti-Tiam1-mediated immunoblot is specific because no protein is detected in these cells using preimmune rabbit IgG (Fig. 2, lane 1). Furthermore, we have carried out anti-CD44v3-mediated and anti-Tiam1mediated precipitation followed by anti-Tiam1 immunoblot (Fig. 2, lane 3) or anti-anti-CD44v3 immunoblot (Fig. 2, lane 4), respectively, using SDS-polyacrylamide gel electrophoresis analyses. Our results clearly indicate that the Tiam1 band is revealed in anti-CD44v3-immunoprecipitated materials (Fig.  2, lane 3). The CD44v3 band can also be detected in the anti-Tiam1-immunoprecipitated materials (Fig. 2, lane 4). These findings clearly establish the fact that CD44v3 and Tiam1 are closely associated with each other in vivo in breast tumor cells.
In Vitro Binding Between Tiam1 (or PHn-CC-Ex Domain) and CD44 -Previous studies indicate that Tiam1 membrane localization (through its NH 2 -terminal PHn domain and an adjacent protein interaction domain (designated as PHn-CC-Ex, a sequence between amino acids 393-738 of Tiam1)) (Fig. 3, A and C) is required for the activation of Rac1 signaling pathways leading to membrane ruffling and c-Jun NH 2 -terminal kinase activation (37,54). To test whether CD44 is one of the membrane proteins involved in the direct binding to Tiam1, we have used purified CBP-tagged PHn-CC-Ex fusion protein (Figs. 3C and 4, lane 1) and the FLAG-tagged cytoplasmic domain of CD44 (FLAG-CD44cyt) fusion protein (Fig. 4, lane 2) to identify the CD44 binding site on the Tiam1 molecule. Specifically, we have tested the binding of FLAG-CD44cyt to 125 Ilabeled CBP-PHn-CC-EX (or 125 I-labeled intact Tiam1) under equilibrium binding conditions. Scatchard plot analyses presented in Fig. 5 indicate that PHn-CC-Ex binds to the cytoplasmic domain of CD44 (CD44cyt) at a single site (Fig. 5A) with high affinity (an apparent dissociation constant (K d ) of Ϸ0.2 nM). This interaction between PHn-CC-Ex and CD44 is comparable in affinity with CD44 binding (K d ϭ ϳ0.13 nM) to intact Tiam1 (Fig. 5B). These findings clearly indicate that Tiam1 (in particular, PHn-CC-Ex domain) contains the CD44 binding site.
Further analyses using an in vitro binding assay show that surface biotinylated CD44v3 (isolated from SP1) specifically binds to Tiam1 (including intact Tiam1 (Fig. 6, lane 1), HAtagged C1199 Tiam1 (Fig. 6, lane 2) or Tiam1 fragment (PHn-CC-Ex) (Fig. 6, lane 3))-coated beads. In the presence of an excess amount (Ϸ100-fold) of recombinant PHn-CC-Ex Tiam1 fragment, the binding interaction between CD44v3 and these-Tiam1-related proteins is readily abolished (Fig. 6, lanes 4 -6). These observations suggest that (i) the breast tumor cell-spe-  cific CD44v3 is also capable of interacting with Tiam1 (e.g. intact Tiam1 (Fig. 6, lane 1), HA-tagged C1199 Tiam1 (Fig. 6, lane 2), or Tiam1 fragment (PHn-CC-Ex) (Fig. 6, lane 3)); and (ii) the Tiam1 fragment such as PHn-CC-Ex acts as a potent competitive inhibitor for Tiam1 binding to CD44v3 in vitro (Fig. 6, lanes 4 -6). Tiam1-catalyzed Rac1 Activation in SP1 Cells-Rac1 GTPase becomes activated when bound GDP is exchanged for GTP by a process catalyzed by guanine nucleotide (GDP-GTP) exchange factors or GDP dissociation stimulator proteins (i.e. promoting GTP binding to RhoA by facilitating the release of GDP) (25,  26). Tiam1 is known to function as an exchange factor for the Rho-like GTPases such as Rac1 (34, 40 -42). To investigate whether Tiam1 complexed with CD44v3 acts as a GDP/GTP exchange factor (or a GDP dissociation stimulator protein) for E. coli-derived GST-Rac1, we have isolated Tiam1 complexed with CD44v3 from SP1 cells using anti-Tiam1-conjugated Sepharose beads. Our data show that Tiam1 complexed with CD44v3 from SP1 cells causes the exchange of preloaded GDP for [ 35 S]GTP␥S on GST-Rac1 in a time-dependent manner (Fig.  7, lines a and b). Most importantly, addition of HA to CD44v 3 containing SP1 cells stimulates the total amount of bound [ 35 S]GTP␥S to GST-Rac1 (Fig. 7, line b) (at least 1.5-fold increase) as compared with Tiam1 isolated from untreated SP1 cells (Fig. 7, line b) or HA-treated SP1 cells in the presence of anti-CD44v3 antibody (data not shown). No [ 35 S]GTP␥S-bound material was detected in these samples containing GST alone under the same GDP/GTP exchange reaction using Tiam1 isolated from SP1 cells (in the presence (Fig. 7, line c) or absence (Fig. 7, line d) of HA treatment). These findings suggest that the HA interaction with CD44v 3 isoform-containing SP1 cells promotes Tiam1 activation of Rac1.
CD44v3-Tiam1 Interaction in Rac1 Signaling and Cytoskeleton-mediated Tumor Cell Migration-Previous studies indicate that the invasive phenotype of tumor cells characterized by an invadopodia structure (or membranous projections) (56,57) and tumor cell migration (28,29) is closely associated with CD44v 3,8 -10 -linked cytoskeleton function (23). In this study we have transiently transfected breast tumor cells (e.g. SP-1 cells) with HA-tagged NH 2 -terminally truncated C1199 Tiam1 cDNA (Fig. 3B). Our results show that the C1199 Tiam1 is expressed as a 150-kDa protein (Fig. 8A, lane 1) detected by anti-HAmediated immunoblot in CD44v 3 -positive breast tumor cells (SP1 cells). No protein band was detected in vector-transfected SP1 cells by anti-HA-mediated immunoblotting (Fig. 8A, lane  3). Using anti-CD44v3 immunoprecipitation of SP1 cellular protein followed by immunoblotting with anti-HA antibody, we have found that the 150-kDa C1199 Tiam1 is co-precipitated with CD44v3 (Fig. 8A, lane 2). In control samples, immunoblotting of rabbit preimmune IgG-precipitated material using anti-HA antibody does not reveal any protein associated with this material (Fig. 8A, lane 4). Double immunofluorescence staining data also confirms the close association between CD44v3 (Fig.  9A) and the C1199 Taim1 (Fig. 9B) in the plasma membranes and long membrane projections. In contrast, vector-transfected cells expressing CD44v3 on the surface (Fig. 9, inset a) (with no detectable C1199 Tiam1 by anti-HA label (Fig. 9, inset b)) fail to display long membrane projections. Furthermore, we have demonstrated that transfection of SP1 cells with C1199 Tiam1 cDNA stimulates CD44v3-associated Tiam1-catalyzed GDP/ GTP exchange on Rac1 (Table II) and induces a significant amount of increase in CD44v3-specific and HA-mediated breast tumor cell migration (Table II) compared with vector-transfected SP1 transfectants (Table II). These results are consistent with previous findings indicating that transfection of NIH3T3 cells with the NH 2 -terminally truncated C1199 Tiam1 cDNA confers potent oncogenic properties (42). Treatment of SP1 cells (e.g. untransfected cells or transfected cells) with certain agents (e.g. cytochalasin D (a microfilament inhibitor)) causes a remarkable inhibition of CD44v3/HA-specific tumor cell migration (Table II). These observations suggest that CD44v3-associated Tiam1 signaling and cytoskeleton-mediated tumor cell motility are closely coupled.
Moreover, we have found that SP1 cells transfected with GFP-tagged PHn-CC-Ex Tiam1 cDNA express a 68-kDa pro- tein as detected by anti-GFP antibody (Fig. 8B, lane 1). In vector-transfected SP1 cells, we are not able to detect any protein band by anti-GFP-mediated immunoblotting (Fig. 8B,  lane 3). Using anti-CD44v3 immunoprecipitation of SP1 cellular protein followed by immunoblotting with anti-GFP antibody, we have found that the 68-kDa PHn-CC-Ex Tiam1 fragment is co-precipitated with CD44v3 (Fig. 8B, lane 2). No protein band was found when immunoblotting of rabbit preimmune IgG-precipitated materials with anti-GFP antibody was used (Fig. 8B, lane 4). It is also noted that both GFP-tagged PHn-CC-Ex domain (Fig. 9D) and CD44v3 are co-localized in the plasma membranes (Fig. 9C). However, no significant stimulation of long membrane projection was observed in these cells (Fig. 9, C, D, and insets c and d). Furthermore, we have demonstrated that CD44v3 staining detected in these SP1 transfectants revealed by anti-CD44v3-mediated immunostaining is specific because no surface label (Fig. 9, inset c) is detected in these GFP-PHn-CC-Ex overexpressed cells (Fig. 9, inset d) using preimmune rabbit IgG (Fig. 9, inset c). Additionally, we have demonstrated that overexpression of GFP-tagged PHn-CC-Ex domain in SP1 transfectants does not cause any significant changes of breast tumor cell properties (e.g. CD44v3associated Tiam1-Rac1 signaling or HA-mediated tumor cell migration (Table II)) Finally, we have conducted co-transfection of SP1 cells with HA tagged C1199 Tiam1 cDNA and GFP-tagged PHn-CC-Ex cDNA. Our results indicate that C1199 Tiam1 and PHn-CC-Ex Tiam1 fragment are co-expressed as a 150-kDa protein (Fig.  8C, row a, lane 1) and a 68-kDa protein (Fig. 8C, row b, lane 1), respectively, in SP1 cells. No protein band was revealed in vector-transfected SP1 cells by anti-HA (Fig. 8C, row a, lane 3) or anti-GFP-mediated (Fig. 8C, row b, lane 3) immunoblotting. Using anti-C44v3 antibody immunoprecipitation of SP1 cell lysate followed by immunoblotting with anti-GFP antibody and anti-HA, respectively, we have found that the 68-kDa PHn-CC-Ex Tiam1 fragment (Fig. 8C, row b, lane 2) (but not 150-kDa C1199 Tiam1 (Fig. 8C, row a, lane 2)) is co-precipitated with CD44v3. In control samples, immunoblotting of rabbit preimmune IgG-precipitated material using anti-HA antibody (Fig. 8C, row a, lane 4) or anti-GFP antibody (Fig. 8C, row b,  lane 4) does not reveal any protein associated with this material. Immunocytochemical staining results confirm that the PHn-CC-Ex Tiam1 fragment (Fig. 9, inset e) is co-localized with CD44v3 ( Fig. 9, inset f) in the plasma membranes of SP1 transfectants. In contrast, the C1199 Tiam1 (Fig. 9E) fails to display plasma membrane localization as the PHn-CC-Ex domain does (Fig. 9F). Co-expression of PHn-CC-Ex domain and C1199 Tiam1 also efficiently blocks CD44v3-associated Tiam1-Rac1 activation and CD44v3-dependent and HA-mediated breast tumor cell migration (Table II). These results are consistent with a previous report showing that co-transfection of COS-7 cells with PHn-CC-Ex cDNA and C1199 Tiam1 cDNA results in an inhibition of C1199 Tiam1-induced Rac1 signaling and membrane ruffling (54). These findings suggest that the NH 2 -terminal PHn domain and an adjacent protein interaction domain (PHn-CC-Ex) play an important role in regulating Tiam1 localization to the plasma membrane proteins such as CD44v3 isoforms and for oncogenic signaling during extracellular matrix component (e.g. hyaluronic acid)-regulated breast tumor cell invasion and migration.
Clinical studies indicate that a number of CD44v isoforms have been detected at high levels on the surface of tumor cells during tumorigenesis and metastasis (13)(14)(15)(16)(17). As the histologic grade of each of the tumors progresses, the percentage of lesions expressing an associated CD44v isoform increases. In particular, the CD44v3-containing isoforms are detected preferentially on highly malignant breast carcinoma tissue samples. In fact, there is a direct correlation between CD44v3 isoform expression and increased histologic grade of the malignancy (14,17,57).
It has been speculated that some of these CD44v3 isoforms on epithelial cells may act as surface modulators to facilitate unwanted growth factor receptor-growth factor interactions (9, 10) and subsequent tumor formation. The CD44-related glycoproteins are also known to mediate cell adhesion to extracellular matrix components (e.g. HA) and to function as the major hyaluronate receptor (50). In this study we have demonstrated that a 260-kDa CD44v3 isoform is expressed on the surface of breast tumor cells (SP1 cell line) (Fig. 1) and that it interacts with extracellular matrix HA as an adhesion receptor (Table I). Furthermore, addition of HA to SP1 cells stimulates tumor cell migration in a CD44v3-specific and cytoskeleton-dependent manner (Table II). These findings are consistent with previous findings that CD44v isoforms expressed in tumor cells often display enhanced hyaluronic acid binding, which increases cell migration capability (58,59).
The invasive phenotype of CD44v3-mediated breast tumor cells, characterized by invadopodia formation (23), matrix metalloproteinase-9 activation (23, 24), and tumor cell motility (23,48) has been linked to cytoskeletal function, a process in which the small GTP-binding proteins such as RhoA and Rac1 are shown to play important roles. Tsukita and co-workers (60) have reported that Rho-like proteins participate in the interaction between the CD44 and the ERM cytoskeletal proteins. Our recent study determined that RhoA is physically linked to CD44v3 isoform (e.g. CD44v 3, 8 -10 ) in breast tumor cells (48). Rho-kinase stimulated by activated RhoA (GTP-bound form of RhoA) appears to play a pivotal role in promoting CD44v 3,8 -10ankyrin interaction during membrane-cytoskeleton function and metastatic breast tumor cell migration (48). Signaling to the RacGTPase known to regulate actin assembly associated with membrane ruffling, pseudopod extension, cell motility, and cell transformation (33)(34)(35)(36)(37) has been shown to be abnormal in breast tumor cells as compared with normal breast epithelial cells (61). The fact that Rac1 induces stress fiber formation in a Rho-dependent manner suggests that cross-talk occurs between the Rho and Rac1 signaling pathways (33). The question of whether the activation of Rac1 signaling is involved in CD44v3-cytoskeleton-mediated breast tumor-specific events remains to be answered.
Tiam1, which was identified by retroviral insertional mutagenesis and selected for its invasive cell behavior in vitro, has been shown to regulate Rac1 activation (38,39). This molecule is largely hydrophilic and contains several functional domains including a Dbl homology domain (38,62,63), a Discs large homology region (38,64), and two pleckstrin homology (PH) domains (e.g. PHn (the PH domain located at the NH 2 -terminal region of the molecule; and PHc (the PH domain located at the COOH-terminal region of the molecule)) ( Fig. 3) (38). In particular, the Dbl homology domain of Tiam1 exhibits GDP/GTP exchange activity for specific members of the Ras superfamily of GTP-binding proteins (62,63) and plays an important role in Rac1 signaling and cellular transformation (33)(34)(35)(36)(37). In breast tumor cells (e.g. SP1 cells), Tiam1 is detected as a 200-kDa protein (Fig. 2) that is capable of carrying out GDP/GTP exchange for Rac1 (Fig. 7), similar to Tiam1 described in other cell types (34, 40 -42, 65, 66). Other functional domains such as Discs large homology region have been implicated in the binding of membrane protein networks (38,64). The PH domain may mediate association with the submembrane region of the cell via protein-protein or protein-lipid interactions (67). Based on mutational analyses and immunofluorescence staining, Collard and co-workers (37,54) report that the NH 2 -terminal PHn domain (but not PHc) and an adjacent protein interaction domain (e.g. PHn-CC-Ex domain) (Fig. 3) are required for Tiam1 targeting to the plasma membrane and Rac1 activation in fibroblasts. At the present time, identification of the membrane protein(s) involved in Tiam1 binding has not been established.
In this study we have presented new evidence that a close interaction occurs between Tiam1 and certain plasma membrane proteins such as CD44v3 isoform. Using two recombinant proteins (CBP-tagged PHn-CC-Ex domain (Fig. 4, lane 1) and FLAG-tagged CD44 cytoplasmic domain (FLAG-CD44cyt) (Fig. 4, lane 2)), we have demonstrated that the PHn-CC-Ex domain of Tiam1 is directly involved in the binding to the cytoplasmic domain of CD44 (Figs. 5 and 6). In fact, the binding affinity of the PHn-CC-Ex domain of Tiam1 to CD44 is comparable with the intact Tiam1 binding to CD44 (Figs. 5 and 6). In the presence of PHn-CC-Ex, the binding between Tiam1 and CD44 (e.g. CD44v3) is greatly reduced (Fig. 6). The ability of PHn-CC-Ex to effectively compete for Tiam1 binding to the plasma membrane proteins such as CD44v3 (Fig. 6) strongly suggests that the PHn-CC-Ex of Tiam1 is responsible for the recognition of CD44 in vitro.
In addition, we have detected that Taim1 and CD44v3 are physically linked to each other as a complex in vivo (Figs. 2, 8,  SO) or Me 2 SO alone) were placed in the upper chamber of the transwell unit. In some cases, SP1 cells were transfected with either HA-tagged C1199 Tiam1 cDNA or GFP-tagged PHn-CC-ExcDNA or HA-tagged C1199 Tiam1 cDNA plus GFP-tagged PHn-CC-ExcDNA 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 swab. Cell migration processes were determined by measuring the cells that migrate to the lower side of the polycarbonate filters containing HA by standard cell number counting assays as described under "Materials and Methods." Procedures for measuring CD44v 3 -associated Tiam1catalyzed 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 at p Ͻ 0.01. In these experiments Ϸ30 -40% of input cells (Ϸ1 ϫ 10 4 cells/well) undergo in vitro cell invasion and migration in the control samples.
b The values expressed in this table represent an average of triplicate determinations of three to five experiments with a standard deviation of less than 5%. and 9) and that HA binding to CD44v3 promotes Tiam1-catalyzed Rac1 activation ( Fig. 7 and Table II) and tumor cell migration (Table II). Our data also indicate that overexpression of Tiam1 (by transfecting SP1 cells with C1199 Tiam1cDNA) (Figs. 8 and 9) not only promotes C1199 Tiam1 association with CD44v3 (Figs. 8 and 9) but also enhances the metastatic capability of tumor cells (e.g. Rac1 activation and tumor cell migration (Table II)). These results suggest that Tiam1 and CD44v3 are not only structurally linked but also functionally coupled. Previously, it has been shown that Tiam1-activated Rac1 initiates oncogenic signaling cascades that involve activation of c-Jun NH 2 -terminal kinase (37,54) and a novel family of serine/threonine kinases, Paks (p-21 activated kinases) (68,69). However, the identification of CD44v3-Tiam1-mediated downstream targets (e.g. c-Jun NH 2 -terminal kinase and/or Paks activities) during HA-mediated breast tumor progression and metastasis remains to be answered.
Furthermore, we have found that co-transfection of SP1 cells with PHn-CC-Ex cDNA and C1199 Tiam1cDNA (Figs. 8 and 9) effectively blocks tumor cell-specific behaviors (e.g. C1199 Tiam1 association with CD44v3 (Figs. 8 and 9), Rac1 signaling (Table II), and tumor cell migration (Table II)). These findings further support our conclusion that PHn-CC-Ex acts as a potent competitive inhibitor that is capable of interfering with C1199 Tiam1-CD44v3 interaction in vivo. Recently, we have also identified a unique sequence residing within the PHn-CC-Ex domain as the putative cytoskeletal binding site of Tiam1 (70). Most importantly, interaction between Tiam1 and the cytoskeleton up-regulates the GDP/GTP exchange activity of Rho-like GTPases and stimulates breast tumor cell invasion/ migration (70). These observations clearly suggest that the PHn-CC-Ex fragment of Tiam1 is one of the important regulatory domains required for Tiam1 function.
In fibroblasts, Tiam1-induced membrane ruffling is dependent on Rac1 (but not RhoA) activity (71). The fact that Tiam1 is involved in both Rac1-and RhoA-mediated pathways during neurite formation in nerve cells suggests that the balance between two Tiam1-activated Rho-like GTPases (e.g. Rac1 and RhoA) determines a particular biological activity (65). Tiam1-Rac1 signaling is also implicated in promoting integrin-mediated cell-cell and cell-extracellular matrix interaction and lymphoid cell invasion (34,65). In addition, the laminin receptor, ␣ 6 ␤ 1 integrin appears to require Rac1 as a downstream of Tiam1 signaling in neuroblastoma cell activation (65). In epithelial Madin-Darby canine kidney cells, fibronectin and/or laminin1-induced Tiam1-Rac1 signaling up-regulates E-cadherin-mediated adhesion and plays an invasion-suppressor role in Ras-transformed Madin-Darby canine kidney cells (66). However, if Madin-Darby canine kidney cells were grown on different collagen substrates, the expression of Tiam1 or constitutively activated Rac1 (V12Rac) in these cells is able to inhibit the appearance of E-cadherin adhesion and promote cell migration (72,73). Our studies show that approximately 60% (63 Ϯ 4%, n ϭ 5) of the GDP dissociation activity can be detected in the guanine nucleotide exchange assay using Tiam1 isolated from untreated breast tumor cells ( Fig. 7 and Table II). We have also observed that approximately 90% (92 Ϯ 5%, n ϭ 5) of Rac1 is exchanging GDP for GTP in the presence of Tiam1 isolated from either HA-treated cells (Fig. 7) or C1199 Tiam1 cDNA-transfected breast tumor cells (Table II). These findings suggest that Tiam1-catalyzed Rac1 activation is tightly regulated by various signals. Apparently, different responses by Tiam1-catalyzed Rho-like GTPases are controlled by specific upstream activators (in particular, cell adhesion receptors (e.g. CD44, integrin, or E-cadherin, etc.) or extracellular matrix components (e.g. HA, collagens, laminin, or fibronectin, etc.)), which may result in selective Tiam1-activated Rho-like GTPases and distinct biological outcome. In summary, we believe that Tiam1-CD44v3 interaction plays a pivotal role in regulating oncogenic signaling required for RhoGTPase activation and cytoskeleton function during HA-mediated metastatic breast tumor cell progression. This could be one of the critical steps in CD44 variant isoform-mediated breast tumor spread and metastasis.