| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 280, Issue 12, 11961-11972, March 25, 2005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From the Department of Medicine, University of California, and the Endocrine Unit, Veterans Affairs Medical Center, San Francisco, California 94121
Received for publication, October 22, 2004 , and in revised form, December 20, 2004.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
190 kDa) is expressed in SK-OV-3.ipl cells and that IQGAP1 interacts directly with Cdc42 in a GTP-dependent manner. Both IQGAP1 and Cdc42 were physically linked to CD44 in SK-OV-3.ipl cells following HA stimulation. Furthermore, the HA-CD44-induced Cdc42-IQGAP1 complex regulated cytoskeletal function via a close association with F-actin that led to ovarian tumor cell migration. In addition, the binding of HA to CD44 promoted the association of ERK2 with the IQGAP1 molecule, which stimulated both ERK2 phosphorylation and kinase activity. The activated ERK2 then increased the phosphorylation of both Elk-1 and estrogen receptor-
(ER), resulting in Elk-1- and estrogen-responsive element-mediated transcriptional up-regulation. Down-regulation of IQGAP1 (by treating cells with IQGAP1-specific small interfering RNAs) not only blocked IQGAP1 association with CD44, Cdc42, F-actin, and ERK2 but also abrogated HA-CD44-induced cytoskeletal function, ERK2 signaling (e.g. ERK2 phosphorylation/activity, ERK2-mediated Elk-1/ER phosphorylation, and Elk-1/ER-specific transcriptional activation), and tumor cell migration. Taken together, these findings indicate that HA-CD44 interaction with IQGAP1 serves as a signal integrator by modulating Cdc42 cytoskeletal function, mediating Elk-1-specific transcriptional activation, and coordinating "cross-talk" between a membrane receptor (CD44) and a nuclear hormone receptor (ER) signaling pathway during ovarian cancer progression. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The binding of HA to CD44 isoforms triggers direct "cross-talk" between two different tyrosine kinase (e.g. p185HER2 tyrosine kinase (5) and c-Src kinase (6))-linked signaling pathways (cell growth versus cell migration, respectively) in ovarian tumor cells. Recent studies have shown that HA-CD44 interaction in caveolin/cholesterol-enriched lipid raft microdomains plays an important role in promoting membrane-cytoskeleton association and stimulating several intracellular signaling pathways (e.g. Ca2+ mobilization, cellular pH changes, Rho signaling, phosphatidylinositol 3-kinase/Akt activation), leading to the onset of cytoskeletal function and ovarian tumor cell-specific behaviors (e.g. cell survival, growth, migration, and invasion) (15-17). These findings clearly demonstrate that CD44 plays a pivotal role in activating oncogenic signaling, leading to ovarian tumor cell function.
Members of the Rho subclass of the Ras superfamily (small molecular mass GTPases, e.g. RhoA, Rac1, and Cdc42) are known to be associated with changes in the membrane-linked cytoskeleton (18). For example, activation of RhoA, Rac1, and Cdc42 has been found to produce specific structural changes in the plasma membrane-cytoskeleton associated with membrane ruffling, lamellipodia, filopodia, and stress fiber formation (18). The rationale for our present focus on Rho GTPase-related signaling is based on previous reports suggesting that CD44-associated cytoskeletal proteins (e.g. ankyrin and ERM) and certain tumor cell-specific phenotypes are dependent on Rho GTPase activation (19-21). Furthermore, overexpression of Rho GTPases in human tumors often correlates with a poor cancer prognosis (22, 23). Consequently, coordinated Rho GTPase signaling is considered to be a possible mechanism underlying cell growth and migration, both of which are obvious prerequisites for metastasis (24, 25).
In the search for additional cellular targets of HA-CD44-activated Rho GTPases that correlate with metastatic behavior in ovarian tumor cells, a molecule identified as IQGAP1 (a 190-kDa protein, one of the downstream effectors of Cdc42) has been detected (26, 27). IQGAP1 inhibits the intrinsic GTPase activity of Cdc42, thereby significantly increasing the cellular levels of active Cdc42 (28-32). This molecule contains numerous functional domains and motifs found in many signal transduction proteins and oncoproteins. These structural domains include four putative calmodulin-binding (IQ) motifs, a calponin homology domain, a polyproline-binding (WW) domain, and a Ras GTPase-activating protein-related domain (26, 27). Some of these motifs are involved in IQGAP1 interaction with specific proteins, such as Cdc42 (28-34), Rac (33, 34), actin (30, 35, 36), calmodulin (29, 30), ERK2 (37), E-cadherin (38), and 
-catenin (39). The establishment of physical linkage between IQGAP1 and its binding partners has been shown to generate several important biological activities. For example, Cdc42-IQGAP1 binding mediates oncogenic signaling events that result in the activation of the actin cytoskeleton (28-36) and tumor cell migration/invasion (40). IQGAP1 binding to -catenin promotes both the rearrangement of the cadherin cell adhesion complex (41, 42) and transcriptional activation (39). Moreover, IQGAP1-ERK2 association modulates the Ras/MAPK pathways (in particular, ERK signaling) (37). These findings suggest that IQGAP1 not only serves as a scaffolding protein (mediating multiprotein complex assembly), but also directly participates in cytoskeleton activation and signaling coordination.
ERKs, members of the MAPK family, are known to be activated by receptor tyrosine kinases, cytokine receptors, and G-protein-coupled receptors (43, 44). In particular, ERK phosphorylation of certain target proteins is directly involved in transcriptional activation by coordinating extracellular cues and intracellular signals (43, 44). For example, during epidermal growth factor (EGF) stimulation or integrin signaling, ERK activation is coupled with the phosphorylation of specific target molecules, including transcription factors (e.g. the ternary complex factor Elk-1) (45, 46) and the nuclear hormone receptor (e.g. estrogen receptor) (47, 48), resulting in cyclin D1 expression and cell cycle progression (45-49). Furthermore, growth factors have been found to activate ERK and to augment estradiol-mediated activation of nuclear hormone receptors/estrogen receptors during malignant transformation (47-49). Previous studies indicate that HA interaction with the hyaluronan receptor RHAMM regulates ERK activity (50). Most importantly, HA-activated ERK signaling participates in the up-regulation of certain transcription factors (e.g. NF-
B and AP-1) (50, 51) and cell growth and cell migration (50). Although ERK activation is closely associated with HA and RHAMM signaling, the regulatory mechanisms that modulate ERK-mediated transcriptional activation and cellular functions in HA-CD44-stimulated ovarian cancer progression are not well understood.
In this study, we have described a new HA-CD44-mediated oncogenic mechanism occurring in ovarian tumor cells. Specifically, we have focused on CD44 interactions with IQGAP1 during HA signaling in SK-OV-3.ipl ovarian tumor cells. Our results indicate that HA-CD44-stimulated IQGAP1 binds to both Cdc42 and ERK (in particular, ERK2). These interactions enhance F-actin binding and promote ERK activation, leading to the phosphorylation of Elk-1 and estrogen receptor (ER)-, Elk-1-specific transcriptional up-regulation, and estrogen-responsive element (ERE) reporter gene activation, as well as ovarian tumor cell migration. Down-regulation of IQGAP1 (by treating cells with IQGAP1-specific small interfering RNAs (siRNAs)) effectively blocks HA-CD44-stimulated cytoskeletal function, ERK signaling, and ovarian tumor cell behaviors. These findings suggest that HA-CD44-activated IQGAP1 plays an essential role in regulating Cdc42-cytoskeleton interaction and in modulating ERK activity required for Elk-1 and nuclear hormone receptor (ER) transcriptional up-regulation and ovarian cancer progression.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Antibodies and Reagents—Rat anti-CD44 monoclonal antibody (clone 020, isotype IgG2b; obtained from CMB-TECH, Inc., San Francisco, CA) recognizes a common determinant of the HA-binding region of CD44 isoforms, including CD44s, CD44E, and CD44 variant species. This rat anti-CD44 antibody was routinely used for HA-related blocking experiments. Both mouse anti-IQGAP1 and rabbit anti-phospho-ER antibodies were purchased from Upstate Cell Signaling Solutions (Lake Placid, NY). A number of immunoreagents, including rabbit anti-Elk-1, mouse anti-phospho-Elk-1, and goat anti-actin antibodies, were obtained from Santa Cruz Biotechnology Inc. (Santa Cruz, CA). Other immunoreagent such as mouse anti-Cdc42, rabbit anti-ERK, rabbit anti-ERK2, and rabbit anti-phospho-ERK antibodies were purchased from BD Biosciences, Oncogene Research Products (San Diego, CA), and Cell Signaling Technology (Beverly, MA). Mouse anti-ER antibody and recombinant ER were manufactured by Lab Vision Corp. (Freemont, CA) and Calbiochem, respectively. 17-[3H]Estradiol (specific radioactivity of 52 Ci/mmol) was purchased from Amersham Biosciences. Water-soluble -estradiol was obtained from Sigma. High molecular mass Healon HA polymers (500,000-Da polymers) were prepared by gel filtration chromatography using a Sephacryl S1000 column. The purity of the high molecular mass HA polymers used in our experiments was further verified by anion exchange high performance liquid chromatography. HA polymers were also digested by PH20 hyaluronidase. Both intact HA and PH20 hyaluronidase-treated HA fragments were used for analyzing their effects on biological activities as described under "Results."
Measurement of Cdc42 Activation—SK-OV-3.ipl cells (5.0 x 106 cells) were resuspended in buffer containing 118 mM KCl, 5 mM NaCl, 0.4 mM CaCl2, 1 mM EGTA, 1.2 mM magnesium acetate, 1.2 mM KH2PO4, 25 mM Tris-HCl (pH 7.4), and 20 mg/ml bovine serum albumin. An aliquot of the cell suspension was added to the electroporation cuvette and incubated at 4 °C for 5-10 min, followed by the addition of [35S]GTP
S (12.5 µCi). Subsequently, cells were electroporated at 25 microfarads and 2.0 kV/cm, followed by incubation either with HA (50 µg/ml) in the presence or absence of rat anti-CD44 antibody (50 µg/ml) or with PH20 hyaluronidase-treated HA fragments (50 µg/ml) or without any HA treatment at 37 °C for 10 min. Subsequently, [35S]GTPS-labeled cells were washed with phosphate-buffered saline (PBS; pH 7.4) and solubilized in 1.0% Nonidet P-40 with 1 mM GTP, 25 mM magnesium acetate, and protease inhibitors in PBS (pH 7.4). Nonidet P-40-solubilized cells were then incubated with mouse anti-Cdc42 IgG (5 µg/ml) plus goat anti-mouse antibody-conjugated beads. The amount of [35S]GTPS bound to Cdc42 associated with anti-Cdc42 antibody-conjugated Immunobeads was measured using a -counter. The values expressed in Table I represent an average of triplicate determinations of five experiments with S.D. < 5%.
|
Immunoblotting and Immunoprecipitation Techniques—SK-OV-3.ipl cells were serum-starved for 24 h, followed by incubation with or without 50 µg/ml HA or PH20 hyaluronidase-treated HA fragments for various time intervals (e.g. 0, 1, 3, 5, 7, 10, or 15 min) at 37 °C. Subsequently, cells were solubilized in 50 mM HEPES (pH 7.5), 150 mM NaCl, 20 mM MgCl2, 0.5% Nonidet P-40, 0.2 mM Na3VO4, 0.2 mM phenylmethylsulfonyl fluoride, 10 µg/ml leupeptin, and 5 µg/ml aprotinin and immunoblotted using various immunoreagents (e.g. rabbit anti-IQGAP1, anti-ERK, anti-ER, anti-phospho-ER, or anti-actin antibody; 5 µg/ml). In some experiments, immunoblot analyses of IQGAP1, Cdc42, ERK, Elk-1, ER, or CD44 in cells transfected with or without 50 pmol of IQGAP1 siRNA or scrambled siRNA were also carried out.
In some cases, SK-OV-3.ipl cells transfected with or without 50 pmol of IQGAP1 siRNA or scrambled siRNA were incubated with or without 50 µg/ml HA for various time intervals (e.g. 0, 1, 3, 5, 7, 10, or 15 min). These cells were extracted with Nonidet P-40 (as described above) and subjected to immunoprecipitation using rat anti-CD44 or rabbit anti-IQGAP1 antibody, followed by goat anti-rabbit IgG. The immunoprecipitated material was solubilized in SDS sample buffer, electrophoresed, and blotted onto nitrocellulose. After blocking nonspecific sites with 3% bovine serum albumin, the nitrocellulose filter was incubated with various antibodies (e.g. rabbit anti-IQGAP1, rat anti-CD44, rabbit anti-Cdc42, anti-ERK, or anti-phospho-ERK antibody; 5 µg/ml) for 1 h at room temperature, followed by incubation with horseradish peroxidase-conjugated goat anti-rabbit IgG (1:10,000 dilution) at room temperature for 1 h. The blots were then developed using ECL chemiluminescent reagent (Amersham Biosciences) according to the manufacturer's protocols.
IQGAP1 was purified at 4 °C from Nonidet P-40-solubilized cell extracts employing three sequential chromatographic steps by fast protein liquid chromatography and media from Amersham Biosciences as described previously (33). To demonstrate the Rho GTPase-binding properties of IQGAP1, purified IQGAP1 (isolated from SK-OV-3.ipl cells) was incubated with GDP- or GTP-loaded forms of Cdc42-GST-conjugated beads. Proteins associated with Cdc42-GST-conjugated beads were then analyzed by immunoblotting with anti-IQGAP1 antibody as described above.
ERK2-mediated Protein Phosphorylation in Vitro—ERK2 was prepared by anti-ERK2 antibody-conjugated immunoaffinity chromatography. The ERK2 kinase reaction was then carried out in 50 µl of kinase buffer containing 25 mM Tris-HCl (pH 7.5), 5 mM -glycerophosphate, 2 mM dithiothreitol, 0.1 mM Na3VO4, 10 mM MgCl2, 0.1% CHAPS, 0.1 µM calyculin A, 200 µM ATP, 100 ng of ERK2 (isolated from SK-OV-3.ipl cells transfected with or without 50 pmol of IQGAP1 siRNA or scrambled siRNA in the presence or absence of 50 µg/ml HA), and 1 µg of Elk-1 (obtained from anti-Elk-1 antibody-conjugated beads). After incubation for various time intervals (0, 10, 20, 30, 60, and 120 min) at 30 °C, the reaction mixtures were boiled in SDS sample buffer and subjected to SDS-PAGE. The protein bands were revealed by immunoblotting with anti-phospho-Elk-1 and anti-Elk-1 antibodies.
In some experiments, the kinase assays was conducted using ERK2 (isolated from anti-ERK2 antibody-conjugated beads using SK-OV-3.ipl cells transfected with or without 50 pmol of IQGAP1 siRNA or scrambled siRNA in the presence or absence of 50 µg/ml HA) in 1 ml of kinase buffer supplemented with 200 µM [32P]ATP and 18 pM recombinant ER at 30 °C for 30 min. Reactions were terminated by the addition of 20% cold trichloroacetic acid, and 2 mg/ml bovine serum albumin was then added as a carrier. Trichloroacetic acid-precipitated proteins were spotted onto Whatman No. 3MM filter papers, followed by extensive washing with 10% trichloroacetic acid. The radioactivity associated with trichloroacetic acid-precipitated materials was analyzed by liquid scintillation counting.
Luciferase Reporter Assays for Elk-1—Elk-1 transcriptional activity was determined using a construct encoding a fusion between the Gal4 DNA-binding domain and the transactivation domain of Elk-1 (Gal4-Elk-1) using the PathDetect® Elk-1 Trans-Reporting System (Stratagene, La Jolla, CA). Specifically, SK-OV-3.ipl cells pretreated with or without 50 pmol of IQGAP1 siRNA or scrambled siRNA were transfected with the fusion transactivator plasmid pFA2-Elk-1 (0.2 µg/well in 6-well plates) and the reporter plasmid pFR-luc (1 µg/well). As an internal control, these transfectants were also cotransfected with -galactosidase (1 µg/well) under the control of the cytomegalovirus promoter. After a 24-h incubation, these cells were washed with PBS and left untreated or treated with HA (50 µg/ml) or PH20 hyaluronidase-treated HA fragments (50 µg/ml) or pretreated with anti-CD44 antibody, followed by addition of HA (50 µg/ml). Subsequently, these cells were washed with PBS, lysed, and assayed for luciferase and -galactosidase activities using kits from Promega (Madison, WI). Data were normalized based on -galactosidase activity.
Detection of ER Activation—To analyze the activation of the ER, SK-OV-3.ipl cells pretreated with or without 50 pmol of IQGAP1 siRNA or scrambled siRNA were incubated in growth medium lacking phenol red (steroid-depleted medium) and containing 10% charcoal-treated fetal bovine serum for 4 days prior to transient transfection. These cells were then cotransfected with the luciferase reporter gene (3 µg/well; downstream from two EREs; kindly provide by Dr. P. J. Kushner, University of California at San Francisco) and -galactosidase (2 µg/well) under the control of the cytomegalovirus promoter (as an internal control) for 18 h. Subsequently, cells were washed twice with serum-free steroid-depleted medium and incubated with various reagents (e.g. no HA, HA (50 µg/ml), PH20 hyaluronidase-treated HA fragments (50 µg/ml), or anti-CD44 antibody (50 µg/ml) plus HA (50 µg/ml)). These cells were washed with PBS, lysed, and assayed for luciferase and -galactosidase activities using kits from Promega. Data were normalized based on -galactosidase activity.
Analyses of Estrogen Binding in SK-OV-3.ipl Cells—SK-OV-3.ipl cells grown in steroid-depleted medium and 10% charcoal-treated fetal bovine serum were plated overnight in 6-well plates at 40% confluence, followed by washing twice with serum-free steroid-depleted medium. These cells were then incubated at 37 °C for 45 min with 1 ml of serum-free steroid-depleted medium containing 0.091-54 nM 17-[3H]estradiol in the presence or absence of a 100-fold excess of nonradioactive 17-estradiol to measure nonspecific binding. Subsequently, the medium was removed, and the cells were washed twice with PBS. Cell-bound 17-[3H]estradiol was extracted by ethanol treatment, and radioactivity was analyzed by liquid scintillation counting.
F-actin Binding Assays—To study the binding interaction between CD44-associated Cdc42-IQGAP1 and F-actin, CD44-associated Cdc42-IQGAP1-conjugated beads (isolated from SK-OV-3.ipl cells transfected with or without 50 pmol of IQGAP1 siRNA or scrambled siRNA in the presence or absence of HA (50 µg/ml) or PH20 hyaluronidase-treated HA fragments (50 µg/ml) or pretreated with anti-CD44 antibody, followed by HA addition as described above)) in 50 µl of 50 mM Tris-HCl (pH 7.4), 134 mM KCl, and 1 mM MgCl2 were mixed with an equal volume of 8 µM 125I-labeled F-actin, followed by a 30-min incubation at room temperature. Following binding, the CD44-associated Cdc42-IQGAP1-conjugated beads were washed extensively with binding buffer and analyzed by liquid scintillation counting. The amount of F-actin that bound to the CD44-associated Cdc42-IQGAP1-conjugated beads isolated from untreated SK-OV-3.ipl cells (control) or from SK-OV-3.ipl cells treated with scrambled siRNA is designated as 100%. Each assay was set up in triplicate and repeated at least three times. All data were analyzed statistically using Student's t test, and statistical significance was set at p < 0.01.
Tumor Cell Migration Assays—Twenty-four Transwell units were used for monitoring in vitro cell migration as described previously (16, 17). Specifically, 5-µm porosity polycarbonate filters (Costar Corp., Cambridge, MA) were used for the cell migration assay. SK-OV-3.ipl cells (transfected with or without 50 pmol of IQGAP1 siRNA or scrambled siRNA in the presence or absence of rat anti-CD44 antibody (50 µg/ml)) were placed in the upper chamber of the Transwell unit. Medium lacking or containing HA (50 µg/ml) or PH20 hyaluronidase-treated HA fragments (50 µg/ml) was placed in the lower chamber of the Transwell unit. After an 18-h incubation at 37 °C in a humidified 95% air and 5% CO2 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 migrated to the lower side of the polycarbonate filters by standard cell number counting methods as described previously (74, 75). The CD44-specific cell migration was determined by subtracting nonspecific cell migration (i.e. cells that migrated to the lower chamber in the presence of anti-CD44 antibody treatment). Each assay was performed in triplicate and repeated at least five times. The number of ovarian tumor cells that migrated in untreated SK-OV-3.ipl cells (control) or in SK-OV-3.ipl cells treated with scrambled siRNA is designated as 100%. All data were analyzed statistically by Student's t test, and statistical significance was set at p < 0.01.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
S binding assay, we determined that Cdc42 isolated from SK-OV-3.ipl human ovarian tumor cells displays specific guanine nucleotide-binding activity (Table I). In particular, we demonstrated that the addition of HA to CD44-containing SK-OV-3.ipl cells caused almost a 3-fold increase in the binding of [35S]GTPS to Cdc42 compared with the amount of binding present in untreated SK-OV-3.ipl cells (Table I) or in SK-OV-3.ipl cells pretreated with anti-CD44 antibody, followed by HA treatment (Table I). To rule out the possibility that Cdc42 activation was mediated by potential contaminants in our HA preparation, SK-OV-3.ipl cells were also incubated with PH20 hyaluronidase-treated HA fragments. Our results show that no significant Cdc42 activation was detected in cells treated with PH20 hyaluronidase-digested HA fragments compared with intact HA (Table I). These findings indicate that intact HA (not contaminants in the intact HA preparation or in hyaluronidase-treated HA fragments) and CD44 are directly involved in Cdc42 signaling in human ovarian tumor cells.
A large number of Cdc42-associated proteins are known to play important roles in signal transduction and various cellular functions (18). One of the known binding partners for the GTP-bound (activated) form of Cdc42 is IQGAP1 (28-34). Using a specific anti-IQGAP1 antibody-mediated immunoblot technique, we found that significant amounts of IQGAP1 (molecular mass of 190 kDa) were expressed in SK-OV-3.ipl cells (Fig. 1A, lane 2). We believe that the IQGAP1 detected in SK-OV-3.ipl cells is specific since no protein was detected in these cells using preimmune rabbit IgG (Fig. 1A, lane 1). Several lines of evidence have indicated that IQGAP1 binds to Rho GTPases such as Cdc42 (28-34). In this study, we incubated purified IQGAP1 (isolated from SK-OV-3.ipl cells) with GDP- or GTP-loaded forms of Cdc42-GST-conjugated beads. Proteins associated with Cdc42-GST-conjugated beads were then analyzed by immunoblotting with anti-IQGAP1 antibody (Fig. 1B). Our results show that a large amount of IQGAP1 associated with GTP-bound Cdc42-GST-conjugated beads (Fig. 1B, lane 2), whereas a lower level of IQGAP1 bound to GDP-bound Cdc42 (lane 1). These results suggest that IQGAP1 interacts directly with Cdc42 in a GTP-dependent manner.
|
Moreover, Fig. 1C shows that HA was capable of promoting an accumulation of endogenous Cdc42 (panel b, lane 2) in a complex with IQGAP1 (panel a, lane 2) and CD44 (panel c, lane 2) in SK-OV-3.ipl cells, whereas only a small amount of endogenous Cdc42 (panel b, lanes 1 and 3) was co-precipitated with IQGAP1 (panel a, lanes 1 and 3) and CD44 (panel c, lanes 1 and 3) in untreated cells (lane 1) or in cells pretreated with anti-CD44 antibody, followed by HA treatment (lane 3). In addition, we noticed that the Cdc42 association (Fig. 1C, panel b, lane 4) with IQGAP1 (panel a, lane 4) and CD44 (panel c, lane 4) was relatively low in cells incubated with PH20 hyaluronidase-treated HA fragments compared with intact HA (panels a and c, lanes 4). These observations suggest that intact HA (not contaminants in the intact HA preparation or in hyaluronidase-treated HA fragments) promotes Cdc42 binding to the CD44-IQGAP1 complex in ovarian tumor cells. Furthermore, the results of an in vitro F-actin binding assay indicate that the HA-CD44-induced Cdc42-IQGAP1 complex was capable of forming a stable ternary structure containing Cdc42, IQGAP1, and F-actin (Table II). Very little F-actin binding to these Cdc42-IQGAP1 complexes (isolated either from untreated cells or from cells pretreated with anti-CD44 antibody, followed by HA addition) was detected using these binding assays (Table II). We observed that F-actin binding to the Cdc42-IQGAP1 complexes (isolated from cells incubated with PH20 hyaluronidase-treated HA fragments) was also very low (Table II). These observations suggest that the HA-CD44-induced Cdc42-IQGAP1 complex is involved in the F-actin binding required for microfilament organization.
|
|
|
In addition, we measured Elk-1-driven gene transcription using a construct encoding a fusion between the Gal4 DNA-binding domain and the transactivation domain of Elk-1 (Gal4-Elk-1). Specifically, SK-OV-3.ipl cells were transfected with Gal4-Elk-1 fusion and firefly luciferase reporter constructs. Fig. 3C shows that Elk-1 transactivation activity was significantly stimulated in cells treated with HA (bar b) compared with Elk-1 transcriptional activity in untreated cells (bar a) or in cells pretreated with anti-CD44 antibody, followed by HA treatment (bar c). In the presence of PH20 hyaluronidase-treated HA fragments, the transcriptional activation of Elk-1 detected in SK-OV-3.ipl cells was very low (Fig. 3C, bar d). These observations strongly indicate that CD44-associated IQGAP1-ERK signaling is involved in HA-dependent Elk-1 phosphorylation and Elk-1-regulated transactivation in ovarian tumor cells.
HA-CD44-induced IQGAP1 Signaling in the Regulation of ERK-mediated ER Phosphorylation and Transcriptional Activation in SK-OV-3.ipl Cells—The action of the ER (in particular, ER) is known to be regulated by the binding of its ligand (estrogen) or by input from signal transduction cascades (47, 52). Fig. 4a shows a representative example of a saturation binding curve involving 17-[3H]estradiol and ERs in SK-OV-3.ipl cells (in vivo). Saturation binding occurred at a ligand concentration of 55 nM. Moreover, the Scatchard plot analysis presented in Fig. 4b indicates that 17-[3H]estradiol bound to the ER at a single site with high affinity (an apparent dissociation constant (Kd) of 
7.0 nM). Using a specific anti-ER antibody-mediated immunoblot technique, we determined that ER (molecular mass of 67 kDa) was the major ER expressed in SK-OV-3.ipl cells (Fig. 4c). EGF stimulation of MAPKs such as ERK1 and ERK2 has been shown to phosphorylate ER at Ser118 in the N-terminal domain and to enhance transcriptional activity (47, 48, 54). The question of whether HA-CD44-induced IQGAP1 signaling and ERK2 activation are capable of causing ER phosphorylation and transcriptional activation in ovarian tumor cells has not been addressed previously.
|
function, we demonstrated that IQGAP1-linked ERK2 isolated from SK-OV-3.ipl cells treated with HA was capable of phosphorylating ER in vitro (Fig. 5A, bar b). A much lower level of ER phosphorylation was detected in untreated cells (Fig. 5A, bar a). These findings indicate that ER serves as a cellular substrate for HA-CD44-activated and IQGAP1-linked ERK2 in vitro.
|
phosphorylation in vivo (as detected by anti-phospho-ER antibody) was also greatly enhanced in SK-OV-3.ipl cells treated with HA. The amount of HA-mediated ER phosphorylation was similar to that of estrogen-induced ER phosphorylation (Fig. 5B, panels a and b, lanes 5). In contrast, ER phosphorylation was very low in untreated SK-OV-3.ipl cells (Fig. 5B, panels a and b, lanes 1) or in SK-OV-3.ipl cells pretreated with anti-CD44 antibody, followed by HA treatment (panels a and b, lanes 3). Treatment of cells with PH20 hyaluronidase-digested HA fragments did not reveal any significant stimulation of ER phosphorylation (Fig. 5B, panels a and b, lanes 4). Therefore, we believe that ER phosphorylation also occurs in vivo (via IQGAP1-ERK activation events) during HA-CD44 interaction similar to estrogen binding in ovarian tumor cells such as SK-OV-3.ipl cells.
Next, we examined the potential impact of ER phosphorylation on the regulation of ERE gene activation. Fig. 5C shows that the level of ERE gene activity was low in untreated cells (bar a) or in cells pretreated with anti-CD44 antibody, followed by HA treatment (bar c). ERE gene activity was greatly enhanced in SK-OV-3.ipl cells treated with intact HA (Fig. 5C, bar b). In contrast, cells treated with PH20 hyaluronidase-digested HA fragments did not show any increase in ERE gene activation (Fig. 5C, bar d). The level of HA-stimulated ERE transcriptional activation (Fig. 5C, bar b) appeared to be comparable with that detected following estradiol treatment (bar e). These findings clearly indicate that intact HA induces ER-mediated nuclear activation and/or transcriptional up-regulation in a CD44-dependent manner in SK-OV-3.ipl cells.
Effects of IQGAP1 siRNA on HA-CD44-mediated Cdc42-Actin Binding, ERK Activation, and Ovarian Tumor Cell Migration—Transfection of mammalian cells with synthetic siRNAs (21-23 nucleotides in length) specifically suppresses expression of endogenous and heterologous genes by RNA interference (58-60). To confirm that IQGAP1 is a signaling regulator responsible for specific downstream effector functions (e.g. recruiting CD44, Cdc42, and ERK2 into a multimolecular complex; mediating Cdc42-actin binding; activating ERK signaling; and stimulating Elk-1/ER-mediated transcriptional activities and tumor cell migration) during HA-CD44 interaction, we transfected SK-OV-3.ipl cells with a specific siRNA sequence targeting human IQGAP1. Fig. 6A (panel a, lane 2) clearly demonstrates that the IQGAP1 siRNA target sequence used successfully suppressed the expression of IQGAP1 in SK-OV-3.ipl cells. In control samples, no IQGAP1 down-regulation was observed in SK-OV-3.ipl cells treated with transfection reagents containing scrambled sequences (Fig. 6A, panel a, lane 1). Because other cellular proteins such as Cdc42 (Fig. 6A, panel b, lane 2), ERK (panel c, lane 2), Elk-1 (panel d, lane 2), ER (panel e, lane 2), and CD44 (panel f, lane 2) were expressed at comparable levels in the IQGAP1 siRNA-treated cells compared with control cells containing scrambled sequences (panels b-f, lanes 1), we conclude that the selective down-regulation of IQGAP1 expression by this IQGAP1 siRNA is specific.
|
phosphorylation in vitro (Fig. 7D, bar d) and in vivo (Fig. 7E, lane 4), and Elk-1/ERE transcriptional activation) in cells incubated with HA (Fig. 7C, bar d; and Fig. 7F, bar d) compared with those signaling events detected in cells treated with scrambled siRNA (Fig. 7A, lanes 1 and 2; Fig. 7B, panels a and b, lanes 1 and 2; Fig. 7C, bars a and b; Fig. 7D, bars a and b; Fig. 7E, panels a and b, lanes 1 and 2; and Fig. 7F, bars a and b).
|
-mediated transcriptional activation, and tumor cell migration.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
HA is the major glycosaminoglycan found in the extracellular matrix of mammalian tissues. It is well known that extracellular matrix components such as HA are also present in large amounts in the mesothelial lining of the peritoneum (8, 9, 13, 14). Furthermore, hyaluronan often accumulates at sites of ovarian tumor cell attachment (8, 9, 13, 14). CD44 is considered to be one of the major HA receptors (4). Specific HA-binding motifs have been identified and localized in the extracellular domain of CD44 (10-12). The binding of HA to CD44 is known to be involved in the onset of a variety of biological activities, including Ca2+ mobilization; receptor redistribution; cell adhesion, proliferation, migration, and aggregation; gene expression, angiogenesis; and tumor metastasis (5-7, 15-19, 24, 25, 62-68).
Previous studies have indicated that the intracellular domain of CD44 binds to certain cytoskeletal proteins such as ankyrin and the ERM proteins (exrin, radixin, and moesin) (19-21). Post-translational modification of the CD44 cytoplasmic domain by acylation (69), serine/threonine kinase-mediated phosphorylation (19, 70, 71), or GTP binding (72) enhances the binding between CD44 and cytoskeletal proteins. Thus, the transmembrane interaction between CD44 isoforms and ankyrin/ERM provides a direct link between the extracellular matrix and the cytoskeleton.
Accumulating evidence indicates that certain Rho GTPase-related signaling molecules are structurally and functionally coupled with the cytoplasmic domain of CD44 (10, 25). Rho GTPases have two interconvertible forms: GDP-bound inactive forms and GTP-bound forms. When the Rho GTPases are in the GTP-bound form, they often interact with specific downstream target(s) and initiate a series of biological activities (18). Previous studies have indicated that HA-CD44 interaction induces activation of Rho GTPases (such as RhoA and Rac1), resulting in cytoskeletal reorganization, actomyosin-based cytoskeletal function, and tumor progression (7, 17, 19). In particular, selective interaction of the CD44 cytoplasmic domain with RhoA-activated Rho kinase (19) and Rho/Rac1-specific guanine nucleotide exchange factors (p115RhoGEF (17), Tiam1 (62), and Vav2 (7)) has been shown to play an important role in promoting invasive and metastatic tumor phenotypes.
As part of our continuing effort to identify CD44 isoform-linked Rho GTPase signaling components that correlate with certain metastatic behaviors, a new candidate molecule named IQGAP1 (a Cdc42/Rac1-binding protein) has been identified. The IQGAP family of proteins has been identified in yeast, amoebas, and mammals (26, 27). There are at least three different isoforms of IQGAP (IQGAP1, IQGAP2, and IQGAP3) detected in mammals (27). At the present time, IQGAP1 is the best characterized isoform, whereas very little information is currently known regarding either IQGAP2 or IQGAP3. Therefore, we have focused our investigation on IQGAP1, which has been shown to contain several protein-protein interaction domains (e.g. a calponin homology domain, four IQ motifs, and a Ras GTPase-activating protein-related domain) and to be associated with a number of diverse target molecules (26, 27). Using an IQGAP1-specific antibody, we have confirmed the presence of IQGAP1 in SK-OV-3.ipl human ovarian tumor cells (Fig. 1). IQGAP1 and the cell-surface molecules CD44 and Cdc42 were closely associated in a complex in human ovarian tumor cells following HA stimulation (Fig. 1). Several lines of evidence indicate that IQGAP1 has cross-linking activity for F-actin that is augmented by GTP-bound active Cdc42 (30, 35, 36). We have also demonstrated that the HA-induced CD44-Cdc42-IQGAP1 complex is capable of binding to F-actin (Table II). These findings are consistent with previous reports showing that Cdc42-IQGAP1 plays an important role in controlling the dynamics of actin assembly (30, 35, 36). The physiological significance of the in vitro F-actin-binding properties of the purified CD44-Cdc42-IQGAP1 complex was further bolstered by the observation that the down-regulation of IQGAP1 by treating cells with IQGAP1 siRNA resulted in the disappearance of the CD44-Cdc42-IQGAP1 complex (Fig. 6), the reduction of F-actin binding to the CD44-Cdc42-IQGAP1 signaling complex (Table II), and the inhibition of HA-CD44-mediated ovarian tumor cell migration (Table III). These results are consistent with our findings indicating that disruption of the F-actin structure by treating cells with cytochalasin D caused a significant reduction in HA-CD44-mediated ovarian tumor cell migration (Table III). These observations further support the likelihood that CD44-Cdc42-IQGAP1 complex formation is one of the critical steps in regulating F-actin binding required for HA-CD44-mediated cytoskeletal function and ovarian tumor cell migration.
Previously, we reported that HA binding to the CD44-p185HER2 complex promotes Ras signaling, which may be one of the important signaling events in human ovarian carcinoma development (5). Ras stimulation is known to promote a downstream kinase cascade, including Raf-1/MEK/ERK pathways (43, 44). The MAPKs ERK1 and ERK2 are ubiquitous protein kinases that transmit and integrate cell-surface signals by phosphorylating target proteins throughout the cell. In particular, ERK signaling is closely associated with ovarian cancer progression (55-57). Therefore, an understanding of how ERK activation is regulated will provide important insight into the disease mechanisms related to cancer progression. ERK kinase signaling is a multistep process involving self-phosphorylation, which is required for the enzyme (kinase) to be able to phosphorylate other key substrates and to influence cell behaviors. During Ras signaling, the MAPKs ERK1 and ERK2 are phosphorylated and activated by the MAPK kinases MEK1 and MEK2, which are in turn phosphorylated and activated by ERK/MAPK kinase kinases, which are members of the Raf-1 family (43, 44).
Although ERK phosphorylation and activation of the Ras/Raf-1/MEK/ERK signaling cascade are important, it is also possible that other signaling pathway(s) may be involved in the activation of ERK. In fact, a recent study indicated that IQGAP1 plays an important role in ERK binding, phosphorylation, and activation during EGF/insulin-like growth factor signaling in breast tumor cells (e.g. MCF cell lines) (37). In this study, we have demonstrated that IQGAP1 is capable of binding to ERK2 (and to a lesser extent, ERK1) and inducing marked phosphorylation of ERK2 in ovarian tumor cells (Fig. 2). Moreover, both ERK2 phosphorylation and kinase activity (associated with IQGAP1) (Fig. 2) could be stimulated by HA and blocked by cells pretreated with anti-CD44 antibody during HA stimulation (Fig. 2 and Table I). Thus, it appears that IQGAP1-associated ERK2 phosphorylation and activity are closely coupled with HA-mediated CD44 activation in ovarian tumor cells. Additional analyses indicated that reduction of IQGAP1 expression and IQGAP1-ERK2 complex formation resulted in inhibition of ERK2 phosphorylation (Fig. 6) and activity (Fig. 7). These findings strongly support the conclusion that IQGAP1 is directly involved in ERK2 binding and ERK phosphorylation and activity during HA-mediated CD44 signaling in ovarian tumor cells.
The ternary complex factors such as Elk-1 belong to a family of extracellular signal-regulated transcription factors (45, 46). ERK activation often induces phosphorylation of Elk-1 at multiple serine and threonine phosphorylation motifs (45, 46). Functional studies have shown that Elk-1 phosphorylation potentiates transcriptional ac
