Inside Pages

[Cancer Biology & Therapy 2:4, e67-e72: EPUB Ahead of Print, http://www.landesbioscience.com/journals/cbt/abstract.php?id=471, July/August 2003]; ©2003 Landes Bioscience

Research Paper

Maspin Regulates Different Signaling Pathways for Motility and Adhesion in Aggressive Breast Cancer Cells ABSTRACT Previous studies from our laboratory and others have demonstrated that treatment of breast cancer cells with exogenous maspin led to a significant decrease in cell motility, and an increase in cell adhesion to human fibronectin. However, the signaling mechanisms by which maspin, a putative tumor suppressor gene, might regulate cell motility and adhesion have not been previously addressed. In this study, we hypothesized that maspin could inhibit cell motility through the Rho GTPase pathway, specifically by affecting Rac activity. To test this intriguing hypothesis we utilized an experimental approach where invasive and metastatic MDA-MB-231 breast cancer cells were either treated exogenously with recombinant maspin protein, or stably transfected with maspin. The data revealed decreased Rac1 activity within 4 h, and a decrease in the Rac1 effector, PAK1, within 12 h. In addition, an increase in PI3K and ERK1/2 activities within 1 h of recombinant maspin (rMaspin) treatment was observed, which returned to baseline level after 12 h. ERK activity was shown to be downstream of PI3K, as pretreatment with the PI3K inhibitor, LY294002, inhibited the stimulation of ERK activity by rMaspin. Furthermore, rMaspintreated cells displayed approximately a 30% increase in cell adhesion which was abrogated by pretreatment with LY294002. Increased focal adhesions and stress fibers were observed after 12 h of rMaspin treatment, when the cells were least motile and had reverted to a more epithelial-like phenotype. These data suggest that maspin may inhibit cell motility by regulating Rac1 and subsequently PAK1 activity, and promote cell adhesion via PI3K/ERK pathways. This study provides new insights into the diverse signaling pathways affected by maspin to suppress the metastatic phenotype, and could contribute to novel therapeutic approaches for the treatment of invasive and metastatic breast cancer.

Cancer Biol Ther 2003; 4: http://www.landesbioscience.com/journals/cbt/abstract.php?id=471.

Once the issue is complete and page numbers have been assigned, the citation will change accordingly.

KEY WORDS

fo

ot

N

This manuscript has been published online, prior to printing, for Cancer Biology & Therapy Volume 2, Issue 4. Definitive page numbers have not been assigned. The current citation for this manuscript is:

e.

Received 06/05/03; Accepted 06/25/03

nc

*Correspondence to: Mary J.C. Hendrix; Department of Anatomy and Cell Biology; 51 Newton Road; 1-100 Bowen Science Building, Carver College of Medicine, Iowa City, Iowa 52242-1109 USA; Tel: 319.335.7755; Fax: 319.335.7770; Email: [email protected]

ie

Center at The University of Iowa; Carver College of Medicine; Iowa City, Iowa USA

sc

1The Department of Anatomy and Cell Biology and 2Holden Comprehensive Cancer

rd

is t

rib ut io n.

Valerie A. Odero-Marah1 Zhila Khalkhali-Ellis1 Jirapat Chunthapong1 Sumaira Amir1 Richard E.B. Seftor1,2 Elisabeth A. Seftor1,2 Mary J.C. Hendrix1,2,*

io

INTRODUCTION Maspin, a putative tumor suppressor gene, is expressed at high levels in normal mammary epithelial and myoepithelial cells1,2 and silenced in metastatic breast and prostate cancers.3-5 It belongs to the serine-protease inhibitor (serpin) family, though some members like ovalbumin have non-inhibitory functions.6,7 To date, its function has been primarily identified as inhibiting invasion, motility and metastasis.1,3 Treatment of aggressive MDA-MB-231 breast cancer cells (which do not make maspin) with recombinant maspin (rMaspin), and transfection of maspin negative MDA-MB-435 cells with the maspin gene results in significant reduction of motility.1,8 These results suggest the involvement of the Rho GTPase family of proteins which are known to play an important role in regulating cell migration.9 The Rho family are small GTP-binding proteins that include RhoA, Rac, and Cdc42. RhoA induces stress fiber formation;10 Rac stimulates formation of lamellipodia;11 and Cdc42 affects formation of filopodia and regulates cell polarity.11,12 In addition, RhoA, Rac1 and Cdc42 have been shown to be overexpressed in human colon, breast and lung tumors when compared to normal tissue, with the most dramatic differences being observed in breast tissue.13 The Rho proteins can interact with multiple downstream effectors to trigger a complex array of signaling pathways. Rac1 and Cdc42 are known to be potent stimulators of c-jun N terminal kinase (JNK) and the p38 family of proteins.14,15 Rac1 and Cdc42 can also activate the serine/threonine kinase p21-activated kinase (PAK). Activated PAK1 leads to accumulation of F-actin and the formation of lamellipodia and filopodia. In fact, activated PAK1 colocalizes with F-actin at the leading edge of motile cells and has also been shown to activate JNK.16 Apart from the PAK-JNK pathway, there are other effectors for PAK1. Activated PAK1 has been shown to phosphorylate myosin light chain kinase (MLCK) and LIM-motif containing kinase (LIMK) proteins, which results in increased cell migration.17,18

B

maspin, Rac GTPase, P13K, ERK, breast cancer

©

20 03

La

nd

es

This research was funded by NIH/NCI CA75681, and the Marilyn Rozeboom Endowment from the Order of the Eastern Star (to M.J.C.H), and IN-122V for the American Cancer Society, administered through the Holden Comprehensive Cancer Center at The University of Iowa (to Z.K.E)

e67

Cancer Biology & Therapy

2003; Vol. 2 Issue 4

MASPIN REGULATES CELL MOTILITY AND ADHESION

It is widely held that cell-cell and cell-matrix adhesion is required for development and proper functioning of multicellular organisms. Integrin engagement during cell adhesion regulates various pathways including gene expression and cell growth, and there is generally cross-talk between integrin and growth factor signaling.19 For example, cell adhesion to fibronectin and presence of mitogen can synergistically signal to activate the extracellular signal regulated kinase (ERK)-type mitogen-activated protein kinase (MAPK).20 Though phosphoinositide 3 kinase (PI3K) has been mostly associated with cell motility, there are some reports indicating that it can also participate in cell adhesion.20 Small adhesions or focal complexes induced by Rac at the leading edge of cells drive rapid cell migration;21 whereas focal adhesions that are larger, more mature, and stable tend to inhibit cell migration.22,23 Previous studies from our laboratory and others demonstrated the ability of maspin to inhibit cell motility and stimulate cell adhesion to fibronectin.1,8,24 We hypothesized that maspin’s ability to regulate cell motility and adhesion are mediated through two distinct signaling pathways. In the current study, we utilized rMaspin-treated and stable maspin transfectants of the highly invasive and metastatic MDA-MB-231 cells and examined changes in Rho GTPase, Rac1, and PAK1 activities, which are associated with altered motility. In addition, we tested the same cells for the ability of maspin to affect PI3K/ERK activity and downstream cell adhesion. The results demonstrated a reduction in Rac1 and its effector, PAK1, and showed that cell adhesion was increased via upregulation of PI3K/ERK activities. This is the first evidence revealing the signaling pathways underlying the molecular regulation of cell motility and cell adhesion by maspin, and may provide new insights into novel therapeutic strategies.

MATERIALS AND METHODS Reagents and Antibodies. G418 was purchased from Mediatech Inc., Herndon, VA. Human fibronectin (FN) and Lipofectamine Plus transfection reagent were obtained from Invitrogen Life Technologies, Inc., Carlsbad, CA. LY294002, leupeptin, aprotinin, and phenylmethylsulfonyl fluoride (PMSF) were purchased from Sigma Aldrich Chemical Co., St Louis, MO. Sodium orthovanadate was from Fisher Scientific, Pittsburg, PA. Enhanced chemiluminescence (ECL) detection reagents were from PerkinElmer Life Sciences, Inc., Boston, MA. BCA protein estimation kit was from Pierce, Rockford, IL. Purified recombinant human maspin was produced in Saccharomyces cerevisiae as previously described.25 The Prolong Antifade kit and the Texas Red-X-phalloidin were from Molecular Probes, Eugene, OR. The Rac Activation Assay Kit, mouse monoclonal antiAkt/PKB, rabbit polyclonal anti-MAP kinase 1/2, and anti-JNK/SAPK1 antibodies were purchased from Upstate Biotechnology, Lake Placid, NY. Rabbit polyclonal anti-Akt/PKB, anti-ERK1/2 and anti-JNK1/2 phosphospecific antibodies were obtained from Biosource International, Inc., Camarillo, CA. Myelin basic protein and mouse monoclonal anti-Rac1 antibody were obtained from BD Transduction Laboratories, San Diego,CA. The rabbit polyclonal anti-PAK antibody was obtained from Santa Cruz Biotechnology, Inc, Santa Cruz, CA. Vinculin ascites monoclonal antibody was a kind gift from Dr. Galen Schneider (Department of Prosthodontics, The Universtiy of Iowa). Mouse monoclonal anti-human maspin antibody was from BD Pharmingen International, San Diego, CA. Mouse anti-actin antibody was obtained from Chemicon International, Temecula, CA. HRP-conjugated goat anti-mouse and HRP conjugated goat anti-rabbit were from Jackson Immunoresearch Technologies Inc., Bar Harbor, ME. FITC-conjugated goat anti-mouse antibody was purchased from ICN Biomedicals, Inc., Aurora, OH. Cell Lines and Cell Culture Conditions. The highly invasive and metastatic MDA-MB-231 breast carcinoma cell line (ATCC, Rockville, MD) was routinely maintained in RPMI 1640 containing 10% FBS and

www.landesbioscience.com

0.1% gentamycin. A human papiloma virus 16E6 (HPV16E6) immortalized normal mammary epithelial cell line, 1436N1, was a gift from Dr Shijie Sheng, Wayne State University, Detroit, MI, and was maintained in D-complete medium (1:1 a-MEM: Ham’s F-12, containing 1% FCS, 1 ng/ml cholera toxin, 10 mM Hepes pH 7.7, 50 µM ascorbic acid and Mito+ serum free supplement). Generation of MDA-MB-231 Stable Transfectants. MDA-MB-231 cells were grown to 90% confluence in a 6-well culture dish and then transfected with 1 mg of either maspin cDNA or the neomycin resistance selection marker in pcDNA3.1 vector only (Invitrogen, San Diego, CA). Stable transfectants were selected by treatment with 500 mg/ml G418 and selected clones were tested for maspin expression by Western blot analysis. The membrane invasion culture system assay, MICS,26 was used to test for functionality. The maspin transfectants displayed an approximate 25% decrease in invasion as compared to the control, thus proving that the construct was functional (data not shown). Western Blot Analysis. Confluent MDA-MB-231 cells were harvested and replated into T-75 flasks coated with 25 µg/ml of human fibronectin. The cells were allowed to attach overnight, then treated with fresh media containing 40 µg/ml rMaspin protein for various time points. This concentration of rMaspin has previously been shown to be effective in inhibiting invasion in vitro.1,8,24 Treated and untreated control cells were lysed in a modified RIPA buffer (50 mM Tris, pH 7.6, 150 mM NaCl, 1% NP-40, 0.5% Deoxycholate) containing leupeptin (50 µg/ml), aprotinin (10 µg/ml), EDTA (2 mM), phenylmethylsulfonyl fluoride (1 mM), and sodium orthovanadate (1mM). The cell lysates were centrifuged, and the supernatants were collected and protein content quantified using a micro BCA reagent assay. Cell lysates (20-30 mg total protein) were resolved on a 10% SDS-PAGE (12% gel for Rac assays), followed by transblotting onto nitrocellulose membrane (Schleicher & Schuell, Keene, NH). The membranes were blocked in TBS-TB (TBS with 0.05% Tween-20 and BSA) and 5% milk, and then incubated with appropriate dilutions of antibody in blocking buffer. After washing, membranes were incubated with peroxidase-conjugated goat anti-mouse IgG, washed, and visualized using an enhanced chemiluminescence system. Blots were stripped in stripping buffer (62.5 mM Tris-HCl, pH 6.7 containing 100 mM β-mercaptoethanol, 2% SDS), prior to reprobing with a different antibody. Rac and PAK Activity Assays. For the Rac activity assay, 300-500 µg of the protein cell lysates were incubated with 10 µg of PAK-PBD agarose beads for 1 h at 4˚C to immunoprecipitate active Rac. The beads were washed 3 times with RIPA buffer, resuspended in 2X SDS sample buffer, boiled, and resolved on a 12% SDS-PAGE gel, followed by blotting and probing with anti-Rac1 antibody. To assay for PAK activity, 300 µg of total cell lysate was incubated with 5 µg anti-PAK1 antibody for 1 h at 4˚C. The immune complex was then adsorbed to protein A sepharose beads and washed 3 times with RIPA buffer, followed by 3 washes in kinase buffer (20 mM Hepes, 50 mM NaCl, 10 mM MnCl2, 0.1% TritonX100, 10% glycerol and 2 mM NaF). The immune complex was then incubated with kinase buffer containing 3 mg myelin basic protein substrate and 3 µCi [32p]-γ-ATP for 30 min at 30˚C. SDS-PAGE sample buffer was then added to the beads, boiled for 5 min, and proteins were resolved on a 10% SDSPAGE, transferred onto nitrocellulose membrane, followed by assessment of 32p incorporation by autoradiography. The blot was subsequently probed with anti-PAK1 antibody. Cell-ECM Adhesion Assay. Cell adhesion assays were performed in triplicate as described previously.27 Briefly, MDA-MB-231 cells were pretreated with rMaspin (40 µg/ml) for 24 h in the presence or absence of PI3K inhibitor, LY294002 (20 µM) at 37˚C, 5% CO2. The inhibitor was added 30 min prior to rMaspin addition. The cells were harvested by treatment with 2 mM EDTA and counted. 1 X 106 cells were then plated onto fibronectin-coated 6-well dishes for 1 h at 37˚C, followed by several washes with PBS to remove non-adherent cells. Adherent cells were trypsinized and counted using a hemacytometer. Immunofluorescence. 5 x 104 cells were cultured on fibronectin-coated coverslips in12-well culture dishes and treated with rMaspin (40 µg/ml) for various time points. The cells were then fixed with methanol/acetone (1:1), washed with phosphate-buffered saline, blocked, and incubated with the


e68


A

B

Figure 1. rMaspin down-regulates Rac1 and PAK1 activity in MDA-MB-231 cells. MDA-MB-231 cells were treated with 40 mg/ml rMaspin for various time points (0–24 h). (A) Active Rac (Rac-GTP) was immunoprecipitated from whole cell lysates using PAK-PBD agarose beads, followed by SDS-PAGE and immunoblotting with anti-Rac1 antibody. Rac1, maspin, and β-actin levels were assayed in whole cell lysates using anti-Rac1,anti-maspin and antib-actin antibodies, respectively. (B) Whole cell lysates from rMaspin-treated MDA-MB-231 cells were immunoprecipitated with antibody to PAK1 followed by incubation with kinase buffer containing myelin basic substrate protein and [32p]-ATP. The reaction was terminated with addition of 2X SDS sample buffer, resolved on an SDS polyacrylamide gel, transferred onto nitrocellulose membrane, and revealed by autoradiography. The blot was subsequently probed with anti-PAK1 antibody. PAK1 activity levels were normalized to PAK1 protein levels, and presented as % of PAK1 activity versus time (0, 12, 24 h). Experiments were done four times and the significance was determined using a paired t-test, P=0.03.

vinculin antibody for 1 h, followed by incubation with FITC-conjugated secondary antibody for 1 h. For filamentous actin staining, Texas red-conjugated phalloidin was used. Slides were mounted using the Prolong Antifade kit, and the images captured with an Axioskop 2 (Carl Zeiss, Inc., Thornwood, NY) and Spot camera (Diagnostic Instrument, Inc., Sterling Heights, MI) using the Zeiss Axiovision 2.0.5 software (Carl Zeiss, Inc.)

RESULTS Reexpression of Maspin in MDA-MB-231 Cells Causes Suppression of Rac1 and PAK1 Activity. Previous results have demonstrated that the highly invasive and metastatic MDA-MB-231 cells (that are normally devoid of maspin) treated exogenously with rMaspin undergo a significant reduction in cell motility (refs. 1 and 8; video available at http://www.anatomy. uiowa. edu/maspin). To determine whether this could be due to disruption of the Rho GTPase signaling pathway, we treated MDA-MB-231 cells plated on human fibronectin with 40 µg/ml rMaspin for various time points. Whole cell lysates were prepared for assessment of Rac1 activity using PAK-PBD agarose beads to pull down active Rac1 (Fig. 1). Within 4 h of rMaspin

e69

Figure 2. Maspin reexpression upregulates PI3K and ERK activities, while JNK is unaffected, in MDA-MB-231 cells. Whole cell lysates from rMaspin treated MDA-MB-231 cells, neo (MB-231neo3) and maspin transfectants (MB-231masp6) were prepared. Phosphorylated AKT (p-AKT) was measured with a phospho-specific anti-AKT antibody to assess PI3K activity, after which the blot was stripped and reprobed with an anti-AKT antibody. ERK activity (p-ERK1/2) was measured using a phospho-specific anti-ERK1/2 antibody, and the blot was stripped and reprobed with anti-ERK antibody. rMaspintreated MDA-MB-231 cells as well as the sham and maspin transfectants were assayed for JNK activity using phospho-specific anti-JNK antibody. The blot was subsequently stripped and reprobed with anti-JNK, anti-maspin, and finally anti-actin antibody. 1436N1, an immortalized normal mammary epithelial cell line served as a positive control.

treatment, a significant decrease in Rac1 activity (Rac-GTP) was observed, and was sustained up to 24 h; whereas Rac1 levels (measured from whole cell lysates) remained constant (Fig. 1A). SDS-PAGE and Western blot analysis of the cell lysates from control and maspin-treated cells indicated that rMaspin quickly enters the cytoplasmic compartment of MDA-MB-231 cells (approximately 1 h following treatment) and the maspin levels decreased slightly by 24 h, possibly due to degradation (Fig. 1A). To determine whether a downstream effector of Rac1 could be affected by maspin, we tested rMaspin-treated MDA-MB-231 cells for their PAK1 activity. PAK1 was immunoprecipitated from total cell lysates prepared from rMaspin-treated cells, and incubated with myelin basic substrate (MBP) in kinase buffer containing [32p]-ATP. Results from autoradiography analysis revealed a decrease in PAK1 activity within 12-24 h of rMaspin treatment, indicating that in addition to Rac1, PAK1 activity was also affected by rMaspin (Fig. 1B). Thus, reexpression of maspin downregulates the activity of Rac1, and its effector PAK1 in MDA-MB-231 breast cancer cells. Phosphoinositide 3 Kinase (PI3K) and ERK1/2 Are Activated, While JNK Remains Unaffected in Response to rMaspin. To investigate other signaling molecules which may be downstream of maspin and Rac in the regulation of cell motility, we assessed PI3K activity by measuring phosphorylation of its downstream effector AKT. Rac activity has been shown previously to be regulated by PI3K.28 PI3K activity, as determined by phosphorylated


2003; Vol. 2 Issue 4


AKT (p-AKT), was increased within 1 h of rMaspin treatment with maximal activity by 4 h which was followed by a decrease in activity thereafter (Fig. 2). PI3K has also been associated with cell adhesion and spreading in conjunction with ERK1/2,20 so we next examined whether ERK1/2 activity (p-ERK) is affected by rMaspin treatment. The data revealed a concomitant increase in ERK1/2 activity (Fig. 2). These observations were further confirmed using an MDA-MB-231 stable maspin transfectant (MB231masp6), which exhibited stronger PI3K and ERK activities compared to the sham transfected control (Fig. 2). We also tested JNK, another downstream effector of Rac and PAK. Treatment of MDA-MB-231 cells with rMaspin did not affect JNK activity, as measured by phospho-specific JNK antibody (Fig. 2). Furthermore, the stable maspin transfectant of MDAMB-231 (MB-231masp6) showed similar JNK activity to MB-231neo3, the neo transfectant control (Fig. 2), thus establishing that JNK is not part of the maspin-Rac1-PAK1 pathway in our experimental model. The 1436N1 immortalized normal mammary epithelial cells that express high levels of maspin exhibited high PI3K activity, very low levels of ERK activity and similar JNK activity as compared to the MDA-MB-231 maspin transfected cells (Fig. 2). These data indicate that maspin reexpression upregulates PI3K and ERK1/2 activities, but does not alter JNK activity in the aggressive breast cancer cells. PI3K Acts Upstream of ERK to Affect Cell Adhesion, and Is Not Involved in Rac Signaling in rMaspin-Treated Cells. To gain insight into the possible signal transduction mechanisms that might involve ERK activation, we utilized the pharmacological inhibitor of PI3K, LY294002. As demonstrated in Figure 3A, addition of LY294002 inhibited PI3K activity (p-AKT) and abrogated the rMaspin-mediated ERK1/2 activation (p-ERK1/2) in MDA-MB-231 cells. To further confirm that PI3K is not part of the Rac signaling cascade, we assayed rMaspin-treated MDA-MB-231 cells that had been pretreated with the PI3K inhibitor, LY294002, for Rac activity. The data revealed a decrease in Rac1 activity (Rac-GTP) within 12 h, similar to that observed in the absence of inhibitor (Fig. 3A). This suggests that PI3K is not functioning through the Rac signaling pathway in rMaspin-treated MDA-MB-231 cells. We next examined the possible involvement of PI3K in cell adhesion to fibronectin in the MDA-MB-231 cells. Cells were pretreated for 24 h with 40 mg/ml rMaspin, plated onto fibronectin-coated dishes, and adherent cells counted after 1 h. rMaspin-treated cells showed an approximate 30% increase in cell adhesion as compared to control untreated cells (Fig. 3B). Treatment with the PI3K inhibitor, LY294002, led to a slight decrease in cell adhesion. Treatment with both the PI3K inhibitor and rMaspin blocked the ability of maspin to increase cell adhesion (Fig. 3B), which suggests that maspin acts via PI3K to promote cell adhesion. Therefore, rMaspin leads to activation of ERK1/2 downstream of PI3K, and its effect on Rac signaling is not mediated by PI3K. rMaspin Treatment Results in Increased Focal Adhesions and Stress Fiber Formation. To investigate whether increased cell adhesion is associated with augmented focal adhesion and stress fiber formation, immunofluorescence microscopy was utilized to track vinculin in focal adhesions and phalloidin to visualize stress fiber formation in rMaspin-treated MDA-MB231 cells (Figs. 4A and 4B). The results show that more focal adhesions were observed after 12-24 h in cells pretreated with rMaspin (Fig. 4A), which correlates with increased cell adhesion (shown in Fig. 3B). Stress fibers were formed after 12 h and were most prominent after 24 h (Fig.4B), in response to rMaspin treatment. These results show that rMaspin treatment leads to increased focal adhesions and well formed stress fibers, as the less motile MDA-MB-231 cells revert to a more epithelial-like phenotype. (Movie available at http://www.anatomy.uiowa.edu/maspin)

DISCUSSION Our investigation describes for the first time that maspin acts via two distinct signaling pathways in the regulation of breast cancer cell motility and adhesion, as presented by the hypothetical model in Figure 5. The data demonstrated that maspin reexpression leads to decreased motility by down-regulating Rac1 and subsequently PAK1 activity. In addition, maspin’s effects on Rac signaling are not mediated www.landesbioscience.com

A

B

Figure 3. PI3K is upstream of ERK signaling in rMaspin-treated MDA-MB-231 cells, and does not appear to affect Rac activity. To examine the ERK signaling pathway in relation to maspin effects, whole cell lysates were made from MDA-MB-231 cells or cells pretreated for 30 min with PI3K inhibitor, LY294002, prior to addition of rMaspin for various amounts of time (0–24 h). (A) PI3K activity was measured using phospho-specific anti-AKT antibody; the blot was then stripped and reprobed with anti-AKT antibody. ERK activity was determined using phospho-specific anti-ERK antibody followed by reprobing of the stripped blot with anti-ERK antibody. To determine whether PI3K was part of the Rac signaling cascade, MDA-MB-231 cells were either untreated or pretreated for 30 min with the LY294002 inhibitor, before rMaspin was added for various amounts of time. Total cell lysates were prepared and examined for Rac activity by immunoprecipitating active Rac with PAK-PBD agarose beads, resolving samples on an SDS polyacrylamide gel, transferring onto nitrocellulose membrane, and probing with anti-Rac1 antibody. Rac1 levels were assayed from total cell lysates with anti-Rac1 antibody. (B) The functional relevance of PI3K upregulation by rMaspin was examined by performing cell adhesion assays. MDA-MB-231 cells were either untreated, pretreated for 24 h with LY294002 alone, rMaspin alone, or maspin plus LY294002. 1 x 106 cells were then plated onto fibronectin-coated dishes for 1 h; non-adherent cells were washed away with PBS and adherent cells trypsinized and counted using a hemacytometer. Experiments were performed in triplicate and the experimental significance determined using a paired t-test, control vs. LY294002, P