The mechanism involved in the regulation of phospholipase ... - Nature

ª

Oncogene (2002) 21, 6520 – 6529 2002 Nature Publishing Group All rights reserved 0950 – 9232/02 $25.00 www.nature.com/onc

The mechanism involved in the regulation of phospholipase Cg1 activity in cell migration Enza Piccolo1,4, Pasquale F Innominato1,4, Maria A Mariggio2, Tania Maffucci3, Stefano Iacobelli1 and Marco Falasca*,1,3 1

Department of Oncology and Neuroscience, Section of Medical Oncology, Universita ‘G. D’Annunzio’, Via dei Vestini 1, 66100 Chieti, Italy; 2Department of Drug Sciences, Laboratory of Cellular Physiology, Universita ‘G. D’Annunzio’, Via dei Vestini 1, 66100 Chieti, Italy; 3The Sackler Institute, University College London, 5, University Street, London WC1E 6JJ, UK

Activation of the enzyme phospholipase C (PLC) leads to the formation of second messengers inositol 1,4,5trisphosphate and diacylglycerol. Tyrosine kinase receptors activate this reaction through PLCg isoenzymes. PLCg activity involves its activation with, and phosphorylation by, receptor tyrosine kinases. Recently, it has been shown that phosphoinositide 3-kinase (PI 3-K) may regulate PLCg activity through the interaction of the PI 3-K product phosphatidylinositol 3,4,5-trisphosphate (PtdIns-3,4,5-P3) and the PLCg pleckstrin homology (PH) domain. In an effort to understand the signalling pathway that involves PI 3-K regulation of PLCg, we found that EGF induces a PI 3-K-dependent translocation of PLCg1 at the leading edge of migrating cells in a wound healing assay. Similarly, the isolated PH, but not the Src-homology (SH) domains, N-SH2 or SH3, of PLCg1, translocates at the leading edge. Our experiments also showed that stable PH PLCg1 expression blocks epidermal growth factor (EGF)- and seruminduced cell motility and increases cell adhesion in MDA-MB-231 cells. This may suggest that influence of PI 3-K on PLCg1 could be relevant in cell migration, where PLCg1 seems to play a key role by modulating a series of events involved in actin polymerization. Oncogene (2002) 21, 6520 – 6529. doi:10.1038/sj.onc. 1205821 Keywords: cell motility; epidermal growth factor; phosphoinositide 3-kinase; phospholipase C; pleckstrin homology domain Introduction Several studies have reported that phospholipase Cg1 (PLCg1) is highly expressed in colorectal cancer tissues, familial adenomatous polyposis, breast carcinoma, and highly metastatic colorectal tumour cell lines, suggesting that PLCg1 is implicated in the pathophysiology of

*Correspondence: M Falasca; E-mail: [email protected] 4 These authors contributed equally to this work Received 26 November 2001; revised 19 June 2002; accepted 28 June 2002

tumorigenesis (Soderquist et al., 1992; Nomoto et al., 1995; Smith et al., 1998). In addition, it has been shown that overexpression of PLCg1 in rat fibroblasts cells leads to malignant transformation (Chang et al., 1997). Growth factor-induced PLCg1 signalling is required for enhanced cell motility (Chen et al., 1994; Polk, 1998; Wells, 2000). In breast, prostate and glioblastoma multiforme cancer cell lines, pharmacological inhibition of PLCg1 signalling reduces cell invasiveness both in vivo and in vitro (Turner et al., 1997; Khoshyomn et al., 1999; Price et al., 1999). These findings suggest that the induction of motility via growth factor-induced PLCg1 signalling is generalizable to a variety of tumour types and may serve as a target to limit tumour progression. Considerable evidence indicates that the phosphoinositide 3-kinase (PI 3-K) signalling pathway plays a central role in the development of several features of malignancy. Recent data demonstrate that the a6b4 integrin promotes carcinoma invasion through a preferential and localized targeting of PI 3-K activity (Shaw et al., 1997). The a6b4-regulated PI 3-K activity is required for the formation of lamellae, dynamic sites of motility, in carcinoma cells (MDA-MB) (Shaw et al., 1997). In addition, it is noteworthy that the tumour suppressor protein, phosphatase and tensin homologue deleted on chromosome 10 (PTEN), whose gene is deleted or mutated in a wide variety of human cancers, possesses 3-phosphoinositide-phosphatase activity (Maehama and Dixon, 1999). The loss of PTEN correlates with the progression of tumours to a metastatic state (Li et al., 1997; Steck et al., 1997), consistent with its role in adhesion and motility. The expression of PTEN in human glioblastoma cells reduces the intracellular level of inositol 1,4,5-trisphosphate, suggesting that PTEN inhibits PLC activity (Morimoto et al., 2000). Much evidence indicates that PI 3-K plays a specific role in membrane translocation and activation of PLCg1. PLCg1 contains a pleckstrin homology (PH) domain in its N-terminal region, that is thought to mediate protein interaction with plasma membrane by binding to the lipid product of PI 3-K PtdIns-3,4,5-P3 (Falasca et al., 1998). It has been proposed that PI 3-K dependence of PLCg1 activation may be substantial at lower levels of tyrosine phosphorylation (Carpenter

PLCg1 and PI 3-K cross talk in cell motility E Piccolo et al

6521

and Ji, 1999). The focal adhesion kinase (FAK) is a non-receptor tyrosine kinase that is tyrosine phosphorylated in response to a number of stimuli. Interestingly, FAK plays a signalling role leading to activation of PLCg1 in response to integrin-mediated adhesion (Zhang et al., 1999; Carloni et al., 1997). FAK signalling enhances integrin-dependent cellular functions including cell spreading, migration and survival. Evidence indicating that growth factor stimulation of PLCg1 activity may result in part from membrane translocation via its PH domain and interaction with PtdIns-3,4,5-P3 raises the possibility that the activation of PI 3-K by FAK may also play a role in integrin stimulation of cellular PLCg1 activity (Falasca et al., 1998). In this study, we provide evidence for a major role of PI 3-K-dependent activation of PLCg1 in cell motility in MDA-MB-231 breast cancer cells in response to epidermal growth factor (EGF). In addition, the PH domain of PLCg1 may have a specific role in directional migration by inducing selective membrane translocation of PLCg1. Results EGF induces PLCg1 polarized translocation in a PI 3-K-dependent manner It has been shown that EGF promotes MDA-MB-231 breast cancer cell migration through a PI 3-K and PLC-dependent mechanism (Price et al., 1999). Therefore, to test the potential role of PLCg1 in cell migration, we carried out wound healing assays in MDA-MB-231 cells. After wounding, MDA-MB-231 were stimulated with 20 ng/ml EGF for 15 min, then fixed and stained for PLCg1 to study protein localization. In serum-starved cells, PLCg1 was dispersed in the cytoplasm (Figure 1A); upon wounding and treatment with EGF, the protein translocated to the cell edge exposed to the wounded area within 15 min (Figure 1B). To test the role of PI 3-K in PLCg1 translocation, MDA-MB-231 cells were pre-treated with 100 nM of the PI 3-K inhibitor wortmannin for 15 min before wounding and EGF stimulation. As shown in Figure 1C, pre-treatment with wortmannin inhibited the translocation of PLCg1. These data indicate that EGF induces PLCg1 translocation at the leading edge of migrating cells in a PI 3-K-dependent manner. PLCg1 and PH PLCg1 translocate at the leading edge in living cells Since PLCg1 translocation is PI 3-K dependent and the N-terminal PH domain of PLCg1 interacts with the PI 3-K lipid product PtdIns-3,4,5-P3, we hypothesize that the PH domain might provide a means to localize PLCg1 to the plasma membrane in a polarized manner. To test this hypothesis, we transiently co-transfected MDA-MB-231 cells with plasmids encoding the full length PLCg1 fused to the RFP and the isolated PH domain fused to GFP. The localization of the fusion

Figure 1 Polarized translocation of PLCg1 to the plasma membrane in MDA-MB-231 cells. Wound healing assay was performed on serum starved MDA-MB-231 cells (A). After wounding, untreated cells (B) or cells pre-treated with 100 nM wortmannin (C) were stimulated with EGF (20 ng/ml) for 15 min and stained for PLCg1. Dotted lines indicate the position of the wound. Arrows mark the polarized localization of the protein. Insets show PLCg1 localization in cells distant from the wound. Bar 10 mm Oncogene


6522

proteins was analysed using confocal fluorescence microscopy in living cells (Figure 2). In serum-starved cells no plasma membrane localization of RFP-PLCg1

and GFP-PH PLCg1 was observed and the fluorescence remained dispersed in the cytoplasm (0 min). The oriented translocation of the fusion proteins along the

Figure 2 Translocation of PLCg1 and PH PLCg1 to the plasma membrane in living cells. MDA-MB-231 cells were transiently cotransfected with plasmids encoding RFP-PLCg1 and GFP-PH PLCg1. After wounding and EGF stimulation, the translocation of the fusion proteins was followed by laser-scanning confocal microscopy. Bar 10 mm Oncogene


6523

leading edge of the cells in the direction of the wound appeared after 6 min and became pronounced after 10 min of EGF stimulation. No changes were observed in MDA-MB-231 cells transfected with the vector alone, pEGFP-C1, with or without EGF or serum stimulation (data not shown). Wortmannin pre-treatment inhibited RFP-PLCg1 and GFP-PH PLCg1 plasma membrane translocation (Falasca et al., 1998; data not shown). These data indicate that the PH domain may mediate PLCg1 translocation to the leading edge of migrating cells. PH domain but not N-SH2 and SH3 domains translocates at the leading edge of migrating cells To study the specific role of the PH domain in PLCg1 polarized migration, wound healing assays were performed in MDA-MB-231 cells transiently transfected with a plasmid encoding the GFP-PH PLCg1 or the pEGFP-C1 vector alone, as a control. As shown in Figure 3, GFP appeared to be mostly cytosolic (A – D) whereas GFP-PH PLCg1 localized at the leading edge of migrating cells, being polarized towards the direction of the wound (E – H). Since it had been shown that FAK is involved in the directionality of cell migration (Gu et al., 1999), these cells were stained with anti-FAK antibody and with rhodamine-labelled phalloidin to visualize the actin cytoskeleton. As shown in Figure 3, PH PLCg1 co-localized with FAK and actin at the leading edge. To test whether other domains are involved in PLCg1 polarized translocation, wound healing assays were performed in MDAMB-231 cells transiently transfected with plasmids encoding the N-SH2 or the SH3 domain of PLCg1 fused to the GFP. As shown in Figure 4, after wounding and EGF stimulation, neither N-SH2 (A – D) nor SH3 domain (E – H) showed the polarized pattern observed with PH PLCg1. These data confirm that PH domain specifically mediates the polarized translocation of PLCg1. Stable PH PLCg1 expression blocks EGF- and serum-stimulated migration of MDA-MB-231 To investigate the effect of stable PH PLCg1 overexpression on cell migration, MDA-MB-231 stable cell lines expressing the GFP-PH PLCg1 or the GFP alone were selected. Control GFP, GFP-PH PLCg1 clones and wild type MDA-MB-231 cells were grown to confluence and then wounded with a pipette tip. After 30 h in the presence of serum, control and wild type cells had migrated into the wounded area while GFPPH PLCg1 clone failed to repopulate the wounded area (Figure 5A). Similar results were obtained using EGF after wounding and at least three different GFP-PH PLCg1 clones (data not shown). A quantitative analysis of this experiment was obtained by using dishes with grids and counting the number of cells migrated in the wounded area (Figure 5B). GFP-PH PLCg1 clone migration was 50% lower than control GFP clone and wild type MDA-MB-231. No effect on cell growth was

observed in GFP-PH PLCg1 clone, compared to GFP clone and wild type cells (data not shown). Pretreatment with wortmannin inhibited the EGF-induced migration of wild type MDA-MB-231 by 23+11%. Alternatively, the effect of overexpression of GFP-PH PLCg1 on cell migration was quantified by using modified Boyden chamber migration assay in the presence of EGF (Figure 6). The number of GFP-PH PLCg1 cells migrated into the membrane was 65% lower than control GFP clone and wild type MDAMB-231. Complete inhibition of EGF-induced migration was observed upon wortmannin pre-treatment. This decrease in cell migration was consistent with that obtained by wound healing assay. These results show that PH PLCg1 overexpression inhibits MDA-MB-231 cell migration. Stable PH PLCg1 expression increases cell adhesion in MDA-MB-231 Since cell locomotion events include cell/substratum adhesion at the leading edge, we performed adhesion assays on GFP and GFP-PH PLCg1 stable transfectants, on different substrates and for different plating times (Figure 7). PH PLCg1 clone had statistically higher levels of adhesion than GFP when laminin (Figure 7A) or fibronectin (Figure 7B) was used as a substrate. The difference in cell adhesion between GFPPH PLCg1 and GFP clone did not change when higher concentrations of fibronectin or longer plating times were used (data not shown). No difference in cell adhesion was observed when cells were plated on BSA. Similar results were obtained using at least three different GFP-PH PLCg1 clones. These data indicate that PH PLCg1 overexpression may affect cell adhesion through an integrin receptor-mediated mechanism. Discussion Receptor tyrosine kinase (RTK) signalling downstream of PLCg1 activates different biological responses such as DNA synthesis, transformation, cell movement and differentiation. RTK phosphorylation of PLCg1 is mediated via recognition of phosphotyrosine-containing sequences in the receptor by the SH2 domains of PLCg1. Several other mechanisms of activation of PLCg1 isoenzymes in mammalian cells have also been proposed. The N-terminal PH domain of PLCg1 interacts with the PI 3-K lipid product PtdIns-3,4,5-P3 and might provide a means to localize PLCg1 to the plasma membrane. However, little is known about the function of the influence of PI 3-K on PLCg1 activation. In an effort to understand the signalling pathway that involves PI 3-K regulation of PLCg1, we found that EGF induces a PI 3-K-dependent translocation of PLCg1 at the leading edge of migrating cells in a wound healing assay. This translocation is likely to be driven by the PH domain since PH PLCg1, but not NSH2 and SH3 domains of PLCg1, translocates at the leading edge. In addition, stable PH PLCg1 expression Oncogene


6524

Figure 3 PH PLCg1 localization in a wound healing assay. MDA-MB-231 cells were transiently transfected with plasmids encoding GFP (A – D) or GFP-PH PLCg1 (E – H). After wounding and EGF stimulation, cells were stained for actin (A,E) and FAK (C,G). (D) and (H) show the merged images. Bar 10 mm

blocks EGF- and serum-induced cell motility and increases cell adhesion in MDA-MB-231 cells. Taken together, these data may suggest that influence of PI 3Oncogene

K on PLCg1 could be relevant in cell migration, where PLCg1 seems to play a key role by modulating a series of events involved in actin polymerization. In particular,


6525

Figure 4 N-SH2 and SH3 PLCg1 localization in a wound healing assay. MDA-MB-231 cells were transiently transfected with plasmids encoding the N-SH2 (A – D) and SH3 (E – H) domains of PLCg1 fused to GFP. After wounding, cells were stimulated with EGF and stained for actin (A,E) and FAK (C,G). (D) and (H) show the merged images. Bar 10 mm

the PH domain-mediated membrane targeting could play a specific role in the activation of PLCg1 in growth factor-induced cell motility. This is in agreement with

recent findings that the PH domain may not be involved in the initial translocation to the EGF receptor, indicating that it could be important for some other Oncogene


6526

Figure 5 Effect of stable overexpression of GFP-PH PLCg1 on cell motility in a wound healing assay. In vitro wound healing assay was performed on wild type MDA-MB-231 cells, GFP and GFP-PH PLCg1 stable transfectants. (A) shows representative phase contrast images of the wounded cell monolayers. (B) shows a quantitative analysis of cell migration. Data are presented as the number of migrated cells and are representative of three independent experiments performed in triplicate

Figure 6 Effect of stable overexpression of GFP-PH PLCg1 in a cell migration assay. Wild type MDA-MB-231 cells, GFP and GFP-PH PLCg1 stable transfectants were tested for their ability to migrate in a transwell assay in response to EGF. Data are presented as the percentage of migrated cells compared to wild type (100%) and are representative of four independent experiments performed in triplicate Oncogene

Figure 7 Effect of stable overexpression of GFP-PH PLCg1 on cell adhesion. Cell adhesion assay was performed on MDA-MB231 GFP and GFP-PH PLCg1 clones. Cells were seeded on 96well plates coated with BSA (A,B) as a control, and laminin (A) or fibronectin (B) and incubated for 15, 30 and 60 min. Relative number of adherent cells was determined by staining cells with Crystal Violet and measurements of absorbance values at 570 nm. Dye levels are directly proportional to the number of cells. Data are representative of four independent experiments performed in triplicate

steps involved in the stimulation of PLCg1 (Matsuda et al., 2001). In addition, our observation that PH PLCg1 overexpression did not affect MDA-MB-231 cell growth indicates that the PH domain could have a role in cell migration but not proliferation. Previous data have shown that pharmcological inhibition of PI 3-K and PLCg pathways blocks cell migration in MDA-MB-231 cells. The data presented here suggest that PLCg1 is a downstream effector of PI 3K involved in the regulation of integrin-dependent signalling by growth factors. Cell motility is the result of different physical processes that are well coordinated. Cell locomotion events include lamellipodia extension, cell/substratum adhesion at the leading edge, contraction to move the cell forward and release of adhesion at the uropodium. A very important step in cell migration is the formation of a leading edge on front-to-rear asymmetry. Recent data suggest that this cell polarization could be generated by the activation of PI 3-K (Haugh et al., 2000). In fibroblasts, it has been shown that a local


6527

generation of 3-phosphoinositides is a key power for spatial sensing governing directional migration. These data indicate that 3-phosphoinositides play a role in directed migration via both G protein- and tyrosine kinase-mediated signalling processes. Our study suggests that PLCg1 is also recruited via its PH domain at the leading edge of migrating cells in a PI 3-K-dependent manner. Therefore, PLCg1 recruitment through a polar gradient of 3-phosphoinositides may play a central role in the first step of cell locomotion, leading to the secondary formation of calcium gradients and protein kinase C activation. The non-receptor protein-tyrosine kinase FAK is a key mediator of integrin signalling events controlling cellular responses to the extracellular matrix, including spreading, migration, proliferation and survival (Schlaepfer et al., 1999). FAK has been found to play a specific role in the link between EGF and PDGF receptors and b1 integrins during fibroblasts chemotactic migration (Sieg et al., 2000). In addition, it has been demonstrated that FAK activates PLCg1 in response to integrin-mediated cell adhesion (Zhang et al., 1999). Evidence indicates that growth factor stimulation of PLCg1 activity may be mediated by plasma membrane translocation via its PH domain interacting with the lipid product of PI 3-K PtdIns3,4,5-P3 (Falasca et al., 1998). This raises the possibility that the ability of FAK to activate PI 3-K may lead to integrin stimulation of PLCg1 activity. A similar mechanism has been recently proposed for the FAKrelated protein tyrosine kinase 2 (PYK2) (Nakamura et al., 2001). It has been demonstrated that PLCg1 is a common mediator for adhesion and growth factor signals in prefusion osteoclasts and that adhesion or macrophage stimulating factors induce direct interaction of PLCg1 with PYK2 (Nakamura et al., 2001). Interestingly, these effects were blocked by PI 3-K inhibitors, indicating that PLCg1 is a downstream effector of PI 3-K and that this pathway can mediate cell spreading and migration in response to growth factor and cytokines. This mechanism represents a particularly attractive target for cancer therapy since its downstream activation may be the common postreceptor mechanism mediating cell motility and invasion that is induced by many growth factors. It remains to be answered whether the PI 3-Kdependent membrane translocation of the PH domain is sufficient for PLCg1 polarized targeting. In fact, although it has been clearly demonstrated that the membrane translocation of PH PLCg1 is PI 3-Kdependent, the interaction with PtdIns-3,4,5-P3 appears to be less specific compared to other PH domains (Kavran et al., 1998). Most PH domains bind to phosphoinositides with a broad range of specificity and affinity (Kavran et al., 1998). A weak and low specific interaction may be necessary but not sufficient to recruit the host protein to the membrane thus probably requiring alternative, co-operative mechanisms (Maffucci and Falasca, 2001). One of these mechanisms has been proposed for the EGF-induced recruitment of the docking protein Grb2-

associated protein 1 (Gab1). The PH domain of Gab1 binds PtdIns-3,4,5-P3, and is likely to act co-operatively with the Met-binding domain. Concerted action of both domains may be required to achieve stable membrane recruitment of Gab1 of sufficient duration to enhance its tyrosine phosphorylation and association with downstream signalling molecules (Rodrigues et al., 2000). Alternatively, a co-operative mechanism could occur within the same PH domain as recently proposed for PLCb1 that has been found to bind both PtdIns-3-P and G protein bg subunits (Razzini et al., 2000). It has been proposed that PLCg1 targeting to the plasma membrane was the result of a synergistic action of the SH2 domains (mediating the association to the receptor) with the PH domain, which could stabilize the interaction by binding to the membrane via PtdIns3,4,5-P3 (Falasca et al., 1998; Bae et al., 1998). Recent work demonstrates that all determinants required for the association of PLCg1 to the EGF receptor are found within a region unique for the PLCg1 family (Matsuda et al., 2001). Furthermore, the association of PLCg1 to the EGF receptor is unaffected by mutations in the PH domain or inhibition of PI 3-K, thus indicating that the PH domain-mediated translocation does not contribute to the stability of the complex (Matsuda et al., 2001). An intriguing hypothesis at present under investigation is that a specific cytoskeletal protein may co-operate with PtdIns-3,4,5-P3 in the membrane recruitment of PLCg1. The last question is how the isolated PH PLCg1 domain increases cell adhesion and decreases cell motility. These effects could play a role in favouring the metastatic spread of breast cancer cells. Thus, PH domain could play a role in a concerted action with the SH domains in the recruitment of PLCg1 to focal adhesion complexes in integrin-mediated cell adhesion. In fact, the C-terminal SH2 domain of PLCg1 has been found to interact with FAK and seems to be involved in PLCg1 activation in response to cell adhesion (Zhang et al., 1999) whereas the SH3 could mediate interactions with the cytoskeleton (Bar-Sagi et al., 1993). Therefore, we can hypothesize that the PH domain interaction with plasma membrane has a role both in cell-oriented migration and in the architecture of focal adhesion complexes. In summary, our results show that PLCg1 recruitment through a PI 3-K-mediated activation provides a link between integrin- and growth factor-mediated signalling pathways to regulate cell motility in MDAMB-231 breast cancer cells. These findings add a new insight to the understanding of the mechanisms involved in cell invasiveness and contributes to the development of possible new therapeutic strategies. Materials and methods Materials and DNA constructs Dulbecco’s modified Eagle’s medium (DMEM) and foetal bovine serum heat inactivated (FBS-HI) were obtained from Oncogene


6528

Gibco, Life Technologies, Inc. Mouse anti-PLCg1 monoclonal antibody was from Santa Cruz Biotechnology (Santa Cruz, USA), rabbit anti-FAK polyclonal antibody was from Sigma. Anti mouse Alexa 594 was from Molecular Probes, anti-rabbit CY5 was from Oregon Green. All other biochemicals were of the highest possible purities and were obtained from Sigma. The cDNA fragment corresponding to the PH domain from rat PLCg1 was subcloned into the green fluorescent protein (GFP) expression vector pEGFP-C1 (Clontech) as previously described (Falasca et al., 1998). The pGEX-2TK (Pharmacia) plasmids containing the N-terminal Src homology (SH) -2 or the SH3 domains of PLCg1 were a kind gift of Dr S-G Rhee (National Institute of Health, Bethesda, USA); for expression of these domains fused to the GFP, the BamHI/EcoRI fragments were subcloned into the BglII/ EcoRI site of pEGFP-C1. The full length PLCg1 fused to the red fluorescent protein (RFP) was a kind gift of Dr M Katan (The Institute of Cancer Research, London, UK). Cell culture and transfection MDA-MB-231 breast cancer cells were maintained in DMEM containing 10% FBS-HI and supplemented with sodium pyruvate, penicillin and streptomycin. For transfection, cells were seeded onto glass coverslips in a six-well plate and transfected with 1 mg of GFP fusion protein vectors. LipofectAMINE (Life Technologies, Inc.) was used according to the manufacturer’s instructions. After 24 h, cells were serum-starved overnight in 0.5% FBS-HI. For co-transfection experiments, 0.5 mg of GFP-PH PLCg1 and 1.5 mg of RFP-PLCg1 were used and cells were serum-starved after 48 h. The MDA-MB-231 stable cell lines expressing the GFPPH PLCg1 or the GFP alone (as a control) were selected with 500 mg/ml G418. Single cell clones were screened by Western blotting and fluorescence microscopy and maintained in medium containing 250 mg/ml G418.

experimental procedure. The images were stored in a Compaq OS/2 Warp processor with 24 bit colour allowing visualization of two colours collected simultaneously or three colours collected sequentially, and displayed separately or in a merged image. The acquisition and processing of the images were controlled by LaserSharp software (BioRad). Cell migration assay Cell migration was tested by in vitro wound healing assays. Wild type MDA-MB-231 cells, GFP and GFP-PH PLCg1 clones were grown to confluence, serum deprived for 18 h and then wounded with a linear scratch by a sterile pipette tip. After washing, cells were incubated for 30 h in DMEM with or without 10% FBS in the presence of 0.5 mg/ml mitomycin-C (Sigma). At the indicated time, cells were photographed in phase contrast with a Nikon microscope. For a quantitative analysis, the same experiment was performed using cell culture dishes with 2 mm grid (Nunc). After 30 h, cells were fixed with 4% paraformaldehyde and the migration was quantified by counting the number of cells per grid square (4 mm2). Alternatively, cell migration was performed in Transwell chambers (tissue culture treated, 6.5 mm diam., 8 mm pores, Costar Corp.) as previously described (Mizushima et al., 1997). Briefly, serum-starved wild type MDA-MB-231, GFP and GFP-PH PLCg1 stable transfectants, were removed from culture dishes with HBSS containing 0.01% trypsin and then resuspended in DMEM containing 1% BSA. 100 000 cells/100 ml were then added to the top of each migration chamber and allowed to migrate into the membrane for 4 h in the presence of EGF (10 ng/ ml). EGF was added in the lower chamber or alternatively in both chambers. Cells attached to the bottom surface of the membrane were fixed with 4% paraformaldehyde, stained with 0.5% Crystal Violet/20% methanol and counted using an inverted microscope. Cell adhesion assay

Immunofluorescence staining MDA-MB-231 cells were treated as indicated, then fixed with 4% paraformaldehyde in PBS for 15 min at room temperature and permeabilized with 0.2% Triton X-100 in PBS for 5 min at room temperature. The coverslips were washed with PBS, blocked with PBS containing 0.1% BSA for 30 min at room temperature and then incubated for 2 h at room temperature with different primary antibodies (mouse antiPLCg1 monoclonal antibody, rabbit anti-FAK polyclonal antibody). Rhodamine-labelled phalloidin was used to visualize actin cytoskeleton. After washing three times with PBS, the coverslips were incubated for 1 h at room temperature with different secondary antibodies (anti mouse Alexa 594, anti-rabbit CY5). Control stainings were performed without primary or secondary antibodies. Confocal microscopy Microscopy was performed by using a Bio-Rad MRC-1024 confocal system (BioRad Laboratories) with a KryptonArgon mixed gas laser (excitation lines: 488, 568, and 647 nm) and an Axiovert 100 microscope equipped with a 63X/1.25 PLAN NEOFLUAR oil immersion objective lens (Zeiss, Jena, Germany). Fixed cell images were taken using slow speed scanning and a Kalman filter (n=4), while in vivo cell images were collected, at every 60 s, using normal speed scanning and a direct filter. The laser potency, photomultiplier, and pin-hole size were kept constant during the entire Oncogene

Cell adhesion assays were performed by incubating cells with the coated substrates at 378C for the indicated times, and quantified by measuring absorbance at 570 nm for Crystal Violet staining. Exponentially growing MDA-MB-231 cells were harvested through trypsinization, washed and incubated on a wheel rotor at 378C in DMEM containing 0.5% BSA at a concentration of 16106 cells/ml for 1 h. A 96-well plate was precoated with 100 ml of either fibronectin (10 mg/ml from human plasma, Calbiochem), or laminin (20 mg/ml, from EHS mouse sarcoma, Boehringer Mannheim), or BSA 1%, in sterile PBS, at 48C overnight, and then saturated with DMEM containing 1% BSA at 378C for 1 h. 100 000 cells were plated onto each coated well and incubated at 378C for the indicated times. After washing, cells were fixed with formaldehyde 2.5% for 20 min at room temperature and stained with 0.5% (w/v) Crystal Violet/20% methanol for 30 min at room temperature. After extensive washes with PBS, plates were air dried and dye was solubilized by adding 100 ml of 0.1 M sodium citrate. OD was calculated by an ELISA plate reader (Mios, Merck) at 570 nm.

Acknowledgements We would like to thank Prof P Innocenti and Prof G Fano for their support, Prof MA Horton for critical reading of the manuscript, Dr S-G Rhee and Dr M Katan for the gift


6529

of constructs. M Falasca is supported by an endowment from the Dr Mortimer and Mrs Theresa Sackler Trust. Publication (3) from the Sackler Institute for Muscular

Skeletal Research, UCL. This work was supported in part by ‘Trenta ore per la vita’ and ‘Lega Italiana per la Lotta contro i Tumori’, Sezione di Pescara.

References Bae YS, Cantley LG, Chen CS, Kim SR, Kwon K and Rhee SG. (1998). J. Biol. Chem., 273, 4465 – 4469. Bar-Sagi D, Rotin D, Batzer A, Mandiyan V and Schlessinger J. (1993). Cell, 74, 83 – 91. Carloni V, Romanelli RG, Pinzani M, Laffi G and Gentilini P. (1997). Gastroenterology, 112, 522 – 531. Carpenter G and Ji Q-S. (1999). Exp. Cell Res., 253, 15 – 24. Chang JS, Noh DY, Park IA, Kim MJ, Song H, Ryu SH and Suh PG. (1997). Cancer Res., 57, 5465 – 5468. Chen P, Xie H, Sekar MC, Gupta K and Wells A. (1994). J. Cell Biol., 127, 847 – 857. Falasca M, Logan SK, Lehto VP, Baccante G, Lemmon MA and Schlessinger J. (1998). EMBO J., 17, 414 – 422. Gu J, Tamura M, Pankov R, Danen EHJ, Takino T, Matsumoto K and Yamada KM. (1999). J. Cell Biol., 146, 389 – 403. Haugh JM, Codazzi F, Teruel M and Meyer T. (2000). J. Cell Biol., 151, 1269 – 1279. Kavran JM, Klein DE, Lee A, Falasca M, Isakoff SJ, Skolnik EY and Lemmon MA. (1998). J. Biol. Chem., 273, 30467 – 30508. Khoshyomn S, Penar PL, Rossi J, Wells A, Abramson DL and Bhushan A. (1999). Neurosurgery, 44, 568 – 578. Li J, Yen C, Liaw D, Podyspanina K, Bose S, Wang SI, Puc J, Miliaresis C, Rodgers L, McCombie R, Bigner SH, Giovanella BC, Ittmann M, Tycko B, Hibshoosh H, Wigler MH and Parsons R. (1997). Science, 275, 1876 – 1878. Maehama T and Dixon JE. (1999). Trends Cell Biol., 9, 125 – 128. Maffucci T and Falasca M. (2001). FEBS Lett., 506, 173 – 179. Matsuda M, Paterson HF, Rodriguez R, Fensome AC, Ellis MV, Swann K and Katan M. (2001). J. Cell Biol., 153, 599 – 612. Mizushima H, Takamura H, Miyagi Y, Kikkawa Y, Yamanaka N, Yasumitsu H, Misugi K and Miyazaki K. (1997). Cell Growth Differ., 8, 979 – 987.

Morimoto AM, Tomlinson MG, Nakatani K, Bolen JB, Roth RA and Herbst R. (2000). Oncogene, 19, 200 – 209. Nakamura I, Lipfert L, Rodan GA and Duong LT. (2001). J. Cell Biol., 152, 361 – 373. Nomoto K, Tomita N, Miyake M, Xhu DB, LoGerfo PR and Weinstein IB. (1995). Mol. Carcinogenesis, 12, 146 – 152. Polk DB. (1998). Gastroenterology, 114, 493 – 502. Price JT, Tiganis A, Agarwal A, Djakiew D and Thompson EW. (1999). Cancer Res., 59, 5475 – 5478. Razzini G, Brancaccio A, Lemmon MA, Guarnieri S and Falasca M. (2000). J. Biol. Chem., 275, 14873 – 14881. Rodrigues GA, Falasca M, Zhang Z, Ong SH and Schlessinger J. (2000). Mol. Cell. Biol., 20, 1448 – 1459. Schlaepfer DD, Hauck CR and Sieg DJ. (1999). Prog. Biophys. Mol. Biol., 71, 435 – 478. Shaw LM, Rabinovitz I, Wang HH-F, Toker A and Mercurio AM. (1997). Cell, 91, 949 – 960. Sieg DJ, Hauck CR, Ilic D, Klingbeil CK, Schaefer E, Damsky CH and Schlaepfer DD. (2000). Nat. Cell Biol., 2, 249 – 256. Smith MR, Court DW, Kim HK, Park JB, Rhee SG, Rhim JS and Kung HF. (1998). Carcinogenesis, 19, 177 – 185. Soderquist AM, Todderud G and Carpenter G. (1992). Cancer Res., 52, 4526 – 4529. Steck PA, Pershouse MA, Jasser SA, Yung WK, Lin H, Ligon AH, Langford LA, Baumgard ML, Hattier T, Davis T, Frye C, Hu R, Swedlund B, Teng DH and Tavtigian SV. (1997). Nat. Genet., 15, 356 – 362. Turner T, Van-Epps-Fung M, Kassis J and Wells A. (1997). Clin. Cancer Res., 3, 2275 – 2282. Wells A. (2000). Adv. Cancer Res., 78, 31 – 101. Zhang X, Chattopadhyay A, Ji QS, Owen JD, Ruest PJ, Carpenter G and Hanks SK. (1999). Proc. Natl. Acad. Sci. USA, 96, 9021 – 9026.

Oncogene