¹Department of Biological Chemistry, The Alexander Silberman Institute of Life Sciences, The Hebrew University of Jerusalem, Jerusalem 91904, Israel

²Molecular Oncology Program, H Lee Moffitt Cancer Center and Research Institute and Department of Biochemistry and Molecular Biology, University of South Florida College of Medicine, Tampa, Florida, FL 33612, USA

Correspondence to: Alexander Levitzki, Department of Biological Chemistry, The Alexander Silberman Institute of Life Sciences, The Hebrew University of Jerusalem, Jerusalem 91904, Israel

Tumors that overexpress HER-2/neu receptor or exhibit enhanced EGFR signaling have been reported to possess constitutively activated Src family kinases, especially pp60^c-Src. High levels of pp60^c-Src activity have also been reported for cell lines that overexpress the EGFR or the chimeric EGFR-HER-2 receptor. It has therefore been suggested that Src kinases may contribute significantly to the oncogenic phenotype of these cells and to the degree of malignancy of tumors that overexpress EGFR family receptors. In this study we show that the induced expression of c-SRC antisense RNA or the application of a selective Src kinase inhibitor induces growth arrest, programmed cell death and reverses the transformed properties of cells that overexpress EGFR or HER-2 receptors. We show that inhibition of Src kinase expression or activity results in the reduction of Stat3 tyrosine phosphorylation, decline of Bcl-X_L expression, and induction of cell death. Using a construct in which the promoter of Bcl-X, which possesses putative Stat3 sites, is tethered to the luciferase reporter gene, we show that inhibition of Src activity or expression induces a decline in Bcl-X expression. We also show that the expression of activated Src induces activation of the Bcl-X promoter. This activation is inhibited by the expression of kinase dead Src or of Stat3, the dominant-negative form of Stat3. Taken together, these results support the hypothesis that Src positively regulates the transformed phenotype of cells overexpressing EGFR family kinases. Furthermore, these results also suggest that Src positively regulates Bcl-X_L expression via Stat3 activation and thus acts not only as a potent mitogenic signaling element, but also as an anti-apoptotic signaling protein. The combination of both activities probably confers upon activated Src its oncogenic activity. Since Src kinase is activated in many tumors, pp60^c-Src kinase inhibitors may prove useful as anti-cancer agents for many types of cancer.

c-Src, Bcl-X, EGFR, HER-2

Introduction

Src family kinases, especially pp60^c-Src (c-Src), cooperate with receptor tyrosine kinases and components of the extracellular matrix to regulate normal cell growth and cell movement (Parsons and Parsons, 1997; Erpel and Courtneidge, 1995). pp60^c-Src (c-Src) has been shown to associate directly with the platelet derived growth factor receptor (PDGFR) (Broome and Hunter, 1996) and CSF-1 receptor (Courtneidge et al., 1993). Src was also reported to interact with EGFR (Sato et al., 1995; Stover et al., 1995), and to phosphorylate the HER-2/neu and IGF-1 receptors (Stover et al., 1995; Peterson et al., 1996). In these cases, c-Src phosphorylates unique phosphorylation sites on these receptors that serve as docking sites for downstream signal transducers like p85, the regulatory subunit of PI 3'-kinase and Shc (Stover et al., 1995). Thus, activation of Src may bypass the need for activation of growth factor receptors by growth factor.

The cooperation of Src family kinases with growth factor receptors also seems to play an important role in oncogenesis and tumor maintenance. Overexpression of EGFR or HER-2 in tumors correlates with increased c-Src kinase activity (Verbeek et al., 1996; Brunton et al., 1997; Ottenholff-Kalff et al., 1992; Osherov and Levitzki, 1994), and EGFR acts synergistically with c-Src to increase cell transformation and tumor invasiveness (Maa et al., 1995). In addition, activation of the EGFR or HER-2/neu receptor in cell lines causes elevation in the activity of Src family kinases, and a direct correlation between the level of receptor expression and the level of Src kinase activity has been noted (Osherov and Levitzki, 1994). Furthermore, in A431 cells, which overexpress EGFR and express the autocrine ligand TGF (Van de Vijver et al., 1991), Src kinase was found to be persistently active (Osherov and Levitzki, 1994). Similarly, NIH3T3 cells that overexpress HER-2 also possess persistently activated Src kinase (Levitzki, 1996).

These findings have suggested that Src plays a key role in tumorigenesis and should therefore be targeted for anti-cancer drugs (Levitzki, 1996). To study this hypothesis further, we repressed c-Src expression with SRC antisense RNA. In addition, we also used a complementary approach in examining the role of Src in transformation by employing the selective Src family kinase inhibitor, PP1 (Hanke et al., 1996). PP1 potently inhibits kinase activity without affecting Src protein levels, enabling us to examine the contribution of Src kinase activity to the transformed phenotype.

Results

Expression of SRC antisense reduces the levels of pp60^{c-Src and induces growth inhibition and morphological changes}

Two cell lines were stably transfected with plasmids containing a fragment of c-Src cDNA in the antisense orientation into inducible systems (Gossen and Bujard, 1992) (see Materials and methods). This was to avoid inducing the expression of redundant Src family kinases such as Fyn and Yes, which appear to compete for c-Src substrates in c-SRC knockout systems (Imamoto et al., 1994). Two cell lines were used: (i) DHER14 cells are NIH3T3 cells overexpressing EGFR in which pp60^c-Src is activated in the presence of EGF (Osherov and Levitzki, 1994). DHER14 clones contain a fragment of the c-Src cDNA inserted into a Lac-Switch plasmid; SRC antisense RNA is expressed by exposure of cells to IPTG. (ii) CSH12 cells are NIH3T3 cells overexpressing the EGFR_(out)-HER-2_(in) chimera, which possess high pp60^c-Src activity (Levitzki, 1996). CSH12 clones contain a fragment of the c-Src cDNA inserted into a Tet-off-regulated plasmid, and SRC antisense RNA is expressed when cells are exposed to tetracycline-free medium.

Induction of the c-SRC antisense RNA in the DHER14 cell line inhibited the growth (Figure 1A), and reduced c-Src protein levels by 75% at 32 h (Figure 1B). In addition, the IPTG-treated cells adopted the morphology of quiescent cells, becoming large and flat (Figure 1C), while the uninduced cells appear to have the morphology of transformed cells (Figure 1D). Furthermore, expression of c-SRC antisense RNA also reduced the ability of DHER14 cells to form colonies in soft agar (Figure 2). The number and size of colonies on soft agar decreased over 50% (Figure 2A). Similar results were obtained with the selective inhibitor of Src kinase PP1 (Figure 2b and c; see also below).

c-Src inhibitor reduces Stat3 phosphorylation

Recent reports suggest that Stat3 is required for cellular transformation by Src (Bromberg et al., 1998; Turkson et al., 1998). To examine the involvement of Stat3 tyrosine phosphorylation by Src (Yu et al., 1995), we utilized specific anti-P-Y-Stat3 antibodies. In CSH 12 cells there is a significant level of constitutive Stat3 phosphorylation (Figure 3A) as compared to DHER14 cells (Figure 3C). Application of PP1 decreased Stat3 tyrosine phosphorylation in both CSH12 and DHER14 in a dose dependent manner.

c-SRC antisense reduces Bcl-X_L expression

To further investigate the effect of of c-SRC antisense RNA on Src-mediated signaling, we looked for changes in Bcl-X_L, an anti-apoptotic protein, knowing that v-Src induces the expression of a newly discovered Bcl-2 family gene, NR-13, in avian cells (Gillet et al., 1995; Mangeney et al., 1996). Induced expression of SRC antisense RNA resulted in a concomitant decline in the levels of Bcl-X_L (Figure 4). The decline in Bcl-X_L levels followed the decline in c-Src levels in both the DHER14 and the CSH12 cell lines expressing SRC antisense RNA. The decline in Bcl-X_L expression in CSH12 and DHER14 cells (Figure 4A - C) was over 50%. Depletion of c-Src and Bcl-X_L was specific, as levels of actin and Erk-2 were unchanged (Figure 4A - F). The amounts of Src and Bcl-X_L are normalized to the amount of Erk-2.

Application of Src kinase inhibitor reduces colony formation in soft agar, reduces Bcl-X_L expression, and causes apoptosis

In order to examine whether inhibiting Src activity is sufficient for inducing the effects observed by reduction of the Src protein we employed the selective Src inhibitor PP1 (Hanke et al., 1996). This inhibitor is a poor blocker of the EGF and IGF-1 receptors in cell-free systems as well as in cellular assays; PP1 does not inhibit significantly the kinase activity of these receptors in intact cells up to 50 M (Hanke et al., 1996 and our unpublished results).

The application of PP1 to DHER14 cells severely reduced their ability to form colonies in soft agar in a dose dependent manner as depicted in Figure 2(B and C) (IC₅₀ of 3 M). Application of PP1 to DHER14 and CSH12 cells resulted in a decline in Bcl-X_L levels in a dose dependent manner (Figure 5). This effect of PP1 was specific, inasmuch as the levels of Erk-2 and Src were not affected (Figure 5A - C). These results are corroborated by the finding that the PP1 induced decline in Bcl-X_L expression was followed by apoptosis (Figure 6), as measured by Annexin-V binding (Figure 6A, B and C) or DNA fragmentation (data not shown). Significant apoptotic cell death was already noted at 5 M PP1 (Figure 6B and C) in both cell lines. No significant necrosis was observed: cells were not stained by propidium iodide (Figure 6A and B (d - f)). Transfection of SRC antisense plasmid (pOPI3-CRS) into DHER14 cells also resulted in cell death, as compared to the empty vector (pOPI3). As a positive control, a similar plasmid coding for the pro-apoptotic protein HA-BAD was used (Figure 6D).

c-SRC antisense, kinase-dead c-Src, Src inhibitor and Stat3 reduce transcriptional activation of the Bcl-X promoter

We used two constructs, in which 3.7 kb upstream of the start of transcription (`full-length' promoter) or a shorter fragment of 0.6 kb (`short promoter') including six repeats of Stat3 binding sites (Catlett-Falcone et al., 1999), were tethered to a luciferase reporter gene. We examined transcription from the Bcl-X full-length promoter in transient co-transfections with c-SRC antisense plasmid (pOPI3-CRS), or in the presence of Src inhibitor PP1. As shown in Figure 7A, reporter expression was suppressed by c-SRC antisense and by PP1. We then examined transcription from the short promoter-luciferase construct. Figure 7B shows that the transient transfection of activated Src (Src F527) into CSH12 cells significantly elevated the expression of the short fragment of Bcl-X promoter-reporter gene whereas the expression of the kinase-dead mutant Src reduced its expression even below control levels. The co-expression of Stat3, the dominant negative form of Stat3 (Turkson et al., 1998), strongly inhibited the expression of the Bcl-X reporter gene (Figure 7B). The selective Src inhibitor PP1 had a similar effect (Figure 7C).

Discussion

Elevated Src kinase activity has been shown in tumors that overexpress growth factor receptors, such as EGFR and HER-2 (Brunton et al., 1997; Ottenholff-Kalff et al., 1992; Osherov and Levitzki, 1994; Maa et al., 1995). We have shown previously that the level of Src kinase activity is proportional to the level of EGFR expression in DHER14 cells (Osherov and Levitzki, 1994). Src kinase is constitutively active in A431 cells, which overexpress EGFR as well as its ligand TGF (Van de Vijver et al., 1991). We have also shown that cells that overexpress Her-2/neu possess constitutive Src kinase activity (Osherov and Levitzki, 1994).

In the present study we examined more directly the possible involvement of c-Src in the oncogenic phenotype of cells that overexpress the EGFR (DHER14 cells) and the HER1-2/neu chimeric receptor (CSH12). The decrease in c-Src levels after induction of c-SRC antisense RNA led to decreased levels of c-Src protein and growth inhibition of DHER14 cells (Figure 1). The morphological changes in DHER14 cells subsequent to antisense RNA induction were reminiscent of changes seen during reversion of the transformed phenotype; the cells became large and flat, like quiescent cells (Figure 1). These changes might point to the involvement of c-Src in adhesion signaling pathways (Sabe et al., 1997; McGill et al., 1997). Growth in soft agar is a measure of the ability of cells to grow without attachment to the substratum, which is typical of tumor cells. Colony formation in soft agar, which correlates with invasiveness, was also diminished after c-SRC antisense RNA induction or addition of Src specific inhibitor (Figure 3). Thus, c-Src kinase seems to play an important role in the transformed phenotype of these cells. It has indeed been pointed out already that expression of v-Src confers upon cells the ability to grow in suspension and therefore may allow such cells to metastasize (Bromberg et al., 1998; Turkson et al., 1998; McGill et al., 1997). We considered the possibility that activated Src may play a role not only in stimulating mitogenic pathways but also in elevating the anti-apoptotic robustness of the cancer cell. We found that a reduction in Src expression (Figure 4A - F) or Src kinase activity (Figure 5) resulted in a concomitant decline in both the phosphorylation state of Stat3 (Figure 3) and the level of Bcl-X_L protein (Figure 5).

This observation supports recent reports that activated Stat3 is required for cellular transformation by Src, and that Stat3-dependent Bcl-X regulation may be important in maintaining cell survival in cells in which Src is activated (Bromberg et al., 1998; Turkson et al., 1998; Catlett-Falcone et al., 1999). Indeed it has been recently reported that v-Src induces the expression of a newly discovered Bcl-2 family gene, NR-13, in avian cells, and that this gene product acts as a potent anti-apoptotic agent (Gillet et al., 1995; Mangeney et al., 1996). As expected, the decline in Bcl-X_L expression was followed by induction of programmed cell death, as depicted in Figure 6. CSH12 cells were more sensitive than DHER14 cells to Src inhibition, with faster induction of apoptosis already at 24 h (data not shown). This sensitivity may reflect the contribution of Src activation by the HER-2 receptor to cell survival in CSH12 cells. Since expression of v-Src in NIH3T3 cells constitutively activates Stat3 and induces Stat3-dependent transcription (Turkson et al., 1998), and the Bcl-X_L promoter possesses putative Stat3 binding sites (Catlett-Falcone et al., 1999), we examined the possible involvement of c-Src in the regulation of transcription from the Bcl-X promoter. We found that application of the Src kinase inhibitor PP1 or transfection of SRC anti-sense plasmid inhibits luciferase production from a Bcl-X promoter-luciferase construct (Figure 7). Transfection of activated Src (Src F527) to CSH12 cells elevated the expression of a Bcl-X reporter gene whereas the expression of the kinase-dead mutant Src reduced its expression even below control levels. Co-expression of the dominant negative form of Stat3 profoundly inhibited the expression from the Bcl-X reporter gene (Figure 7B). These data suggest that Stat3 mediates c-Src dependent Bcl-X_L transcription. Furthermore, PP1 treatment leads to inhibition of Stat3 phosphorylation, which is the primary mechanism of Stat3 activation (Turkson et al., 1998; Yu et al., 1995). The activation and association of Stat3 with v-Src has already been established in a number of systems (Turkson et al., 1998; Yu et al., 1995; Cao et al., 1996). It seems therefore that Stat3 is an important mediator of Src-dependent Bcl-X_L expression. Our observations suggest that receptor tyrosine kinases can activate anti-apoptotic pathways through activation of Src, which in turn induce elevation of Bcl-X_L expression. One such receptor may be the CSF-1 receptor which is known to activate c-Src (Courtneidge et al., 1993), and to activate Stat3 (Cao et al., 1996). This conclusion does not negate the possibility of direct phosphorylation of Stat3 by growth factor receptors (Fu and Zhang, 1993).

Another possible mechanism by which Src activity may confer anti-apoptotic robustness to the cancer cell is by its interaction with receptors that transduce anti-apoptotic signals. For example, it has been shown that activation of cells with IGF-1 rapidly induces the expression of Bcl-X_L (Parrizas and LeRoith, 1997), and that c-Src phosphorylates the IGF-1 receptor, leading to receptor activation even in the absence of IGF-1 (Peterson et al., 1996). Thus, Src may induce Bcl-X_L expression via IGF-1R. The observation that cells lacking IGF-1R could not be transformed by Src (Peterson et al., 1996) supports the existence of such a pathway. It is also possible that c-Src may mediate crosstalk between other growth factor receptors and the IGF-1 receptor, thus driving an anti-apoptotic signal in the absence of IGF-1.

In summary, our results support the hypothesis (Levitzki, 1996) that c-Src is a major signaling element in the transformation machinery of cells that express the EGFR family receptors at high levels. Src seems to function as an anti-apoptotic element through the positive regulation of Bcl-X_L expression, thus elevating the oncogenic robustness of the transformed cell. These results strengthen the hypothesis that Src family kinases, especially pp60^c-Src, should be targeted for the development of anti-cancer agents that may show efficacy in a wide range of malignancies.

Materials and methods

Cell culture

DHER14 cells were grown in Dulbecco's modified Eagles medium (DMEM) supplemented with 10% calf serum (CS). CSH12 cells were grown in DMEM supplemented with 10% fetal calf serum (FCS). All media were supplemented with penicillin and streptomycin. For routine growth of clones, antibiotic selection was maintained. All experiments were performed in the absence of antibiotic selection. Transcription from the Lac Switch plasmid pOPI3-CRS was induced by adding 5 mM isopropyl--D-thio-galactopyranoside (IPTG). Transcription from the tTA-regulated plasmid pNRTIS-CRS was induced by replacing the medium with a doxicyclin free medium.

c-SRC antisense mRNA-construction of plasmids

For the Lac Switch system (Stratagene) a 175 bp fragment from the 5' region of the mouse neuronal c-SRC coding region was synthesized by the Polymerase Chain Reaction (PCR) from plasmid PVN 1.8. The fragment was inserted in the antisense orientation into plasmid pOPI3CAT, generating plasmid pOPI3-CRS, in which the SRC antisense sequence was transcribed under control of the bacterial lac repressor. The tTA-regulated plasmid pNRTIS-CRS was constructed by inserting a 200 bp PCR fragment from the 5' part of the c-SRC coding region into plasmid pNRTIS-21 (gift of Dr T Tenev and Dr F Böhmer, University of Jena, Germany) in the antisense orientation.

Transfections

Stable transfections were by electroporation. One pulse of 250 volts, capacitance - 960 Farad was applied to 2.5´10⁶ cells in a Bio-Rad 0.4 cm cuvette. DHER14 cells were transfected with plasmid pOP3'SS, which leads to constitutive expression of Lac repressor. Following electroporation cells were plated on 10 cm culture dishes and grown in medium containing hygromycin (300 g/ml; Boehringer Mannheim). Ten hygromycin-resistant clones were isolated for each cell type. The clones were probed with polyclonal anti Lac repressor antibody (Stratagene). Clones which were found to express high levels of the Lac repressor were chosen for transfection with the antisense expression plasmid pOPI3-CRS. After transfection with plasmid pOPI3-CRS, cells were plated on 10 cm culture dishes and grown in a double selection medium containing 300 g/ml hygromycin and 400 g/ml G-418 (GIBCO-BRL). Plasmid pNRTIS-CRS was transfected into CSH12 cells by electroporation as described above. The selective medium contained G-418 (400 g/ml) and doxicyclin (100 ng/ml Sigma). In transient transfections cells were transfected with Fugene 6 (Boehringer Mannheim) according to the manufacturer's instructions. Cells were seeded at 7´10⁴ cells/well in a 6 well plate 24 h prior to transfection. Total DNA for transfection was 2.5 g per well including 1 g of luciferase reporter constructs, 0.5 g of -galactosidase (-gal) internal control vector and 1 g of the specific construct, as indicated in results. The Bcl-X promoter-luciferase plasmids (pGL2-3.2, pGL2-0.6R) have been described (Grillot et al., 1997). pUSE-WT-Src, pUSE-KD-Src and pUSE-activated-Src constructs were obtained from Upstate Biochemistry. c-SRC antisense was co-transfected with Lac-Z gene containing plasmid under the CMV promoter to DHER14 cells. Cells were seeded in 6 well plates (7´10⁴ cells/well). Twenty-four and 72 h after transfection cells were stained with X-Gal (GIBCO-BRL) (Sanes et al., 1986), and blue cells were counted in ten different fields. The decline in blue cell number was taken as a measure of antisense (or BAD) induced cell death. We attributed the differences in the numbers of blue cells at 24 h to differences in transfection efficiencies, because we assumed that cell death was minimal at that time. We therefore normalized the 72 h cell counts to the 24 h counts. If some cells had already died at 24 h, this would only heighten the significance of the results.

Growth curves and cell staining

Cell growth was determined by the microculture methylene blue assay (Ben-Bassat et al., 1995). Cells were fixed in glutaraldehyde, 0.05% final concentration, for 10 min at room temperature. After washing, the microplates were stained with 1% methylene blue in 0.1 M borate buffer pH 8.5 for 60 min at room temperature. The plates were then washed extensively and rigorously to remove excess dye and dried. The dye taken up by cells was eluted in 0.1 N HCl for 60 min at 37°C, and absorbance monitored at 650 nm. Each point of the growth curve experiments was calculated from 6 - 8 wells. The cells were photographed after fixation and methylene blue staining as described.

Immunoblotting

Cells were washed three times with PBS, then lyzed with sample buffer (10% glycerol, 0.2 M Tris pH 6.8, 5% -mercaptoethanol, 3% SDS) and boiled for 5 min. The whole cell lysate was then subjected to SDS - PAGE and transferred to nitrocellulose. The membranes were blocked in TBST (10 mM Tris HCl pH 7.5, 50 mM NaCl and 0.1% Triton X-100) containing 5% low fat (1%) milk for 30 min. Membranes were then incubated for 1.5 h with primary antibody (indicated in Figure legends). Membranes were washed extensively with TBST and immunoreactive proteins were detected by incubation with horseradish peroxidase-conjugated anti-mouse IgG (Jackson Immuno Research 1 : 10 000) for detection of monoclonal antibodies or horseradish peroxidase-conjugated anti-rabbit IgG (Jackson Immuno Research 1 : 10 000) for detection of polyclonal antibodies. Proteins were visualized using enhanced chemiluminescence (ECL).

Soft agar assay

Each well of a 6-well culture dish was coated with 2 ml bottom agar mixture (DMEM, 10% FCS or CS, 1% agar). After the bottom layer had solidified, 2 ml top agar mixture (DMEM, 10% FCS or CS, 0.3% agar) containing the cells were added. After this layer had solidified, the protein tyrosine inhibitors (at concentrations indicated in figure legends) or IPTG were overlaid in an additional 2 ml of medium (DMEM, 10% FCS or CS). IPTG was used at a final concentration of 5 mM. Plates were incubated at 37°C. On day 10, colonies from 10 - 20 different fields were counted and the average number of colonies per well was calculated. The wells were overlaid with 2 ml MTT (1 mg/ml), incubated for an additional 4 h in 37°C, and then photographed under light microscope magnification ´100.

Apoptosis assays

For FACS analysis, cells were seeded on 90 mm dishes and treated with PP1 for 24 and 48 h. Cells in medium and adherent cells were collected after trypsinization, and stained with Annexin-V-FLOUS+PI according to the manufacturer's instructions (Boehringer Mannheim). For DNA fragmentation analysis, cells were fixed with 70% ethanol for 1 h on ice, and then incubated with PI solution (50 g/ml PI, 0.1% sodium citrate, 0.1% Triton X-100, 100 g/ml RNAse A, in PBS). For fluorescent microscopy, cells were grown on 6-well plates and treated as written above. Cells were stained with Annexin-PI and photographed under MRC-1024 confocal microscope (BioRad). The photographs show FITC (525/40 nm) and PI (655/90 nm) channels together.

Src inhibitor

PP1 (Hanke et al., 1996), was synthesized as described (Schindler et al., 1999).

Acknowledgements

We would like to thank Dr Ricardo Martinez from SUGEN, Inc. for providing us with the mouse SRC cDNA. We also thank Dr Shoshana Klein and Dr Allan Bar-Sinai from our laboratory for their help preparing the manuscript. This study was partially supported by The James S McDonnel Foundation (USA) and partially by The Israel Science Foundation of the Israel Academy of Sciences and Humanities, Jerusalem, Israel.

Ben-Bassat H, Vardi DV, Gazit A, Klaus SN, Chaouat M, Hartzstark Z and Levitzki A. (1995). Exp. Dermatol. 4, 82-88. MEDLINE

Broome MA and Hunter T. (1996). J. Biol. Chem. 271, 16796-16806.

Bromberg JF, Horvath CM, Besser D, Lathem WW and Darnell Jr. JE. (1998). Mol. Cell Biol., 18, 2553-2558. MEDLINE

Brunton VG, Ozanne BW, Paraskeva C and Frame MC. (1997). Oncogene 14, 283-293. MEDLINE

Cao X, Tay A, Guy GR and Tan YH. (1996). Mol. Cell Biol. 16, 1595-1603. MEDLINE

Catlett-Falcone R, Landowski TH, Oshiro MM, Turkson J, Levitzki A, Savino R, Ciliberto G, Moscinski L, Fernandez-Luna JL, Nunez G, Dalton WS and Jove R. (1999). Immunity 10, 105-115. MEDLINE

Courtneidge SA, Dhand R, Pilat D, Twamley GM, Waterfield MD and Roussel M. (1993). EMBO J. 12, 943-950. MEDLINE

Erpel T and Courtneidge S. (1995). Curr. Opin. Cell 7, 176-182.

Fu XY and Zhang JJ. (1993). Cell 74, 1135-1145. MEDLINE

Gillet G, Guerin M, Trembleau A and Brun G. (1995). EMBO J. 14, 1372-1381. MEDLINE

Gossen M and Bujard H. (1992). Proc. Natl. Acad Sci. USA 89, 5547-5551. MEDLINE

Grillot DA, Gonzalez-Garcia M, Ekhterae D, Duan L, Inohara N, Ohta S, Seldin MF and Nunez G. (1997). J. Immunol. 158, 4759-4757.

Hanke JH, Gardner JP, Dow RL, Changelian PS, Brissette WH, Weringer EJ, Pollok BA and Connelly PA. (1996). J. Biol. Chem. 271, 695-701. MEDLINE

Imamoto A, Soriano P and Stein PL. (1994). Curr. Opin. Genet. Dev. 4, 40-46. MEDLINE

Levitzki A. (1996). Anti-Cancer Drug Design. 8, 239-244.

Maa MC, Leu TH, McCarley DJ, Schatzman RC and Parsons SJ. (1995). Proc. Natl. Acad Sci. USA 92, 6981-6985. MEDLINE

Mangeney M, Schmitt JR, Leverrier Y, Thomas J, Marve J, Brun G and Gillet G. (1996). Oncogene 13, 1441-1446. MEDLINE

McGill G, Shimamura A, Bates RC, Savage RE and Fisher DE. (1997). J. Cell Biol. 138, 901-911. MEDLINE

Ottenholff-Kalff AE, Rijksen G, Van Beurden EA, Hennipman A, Michels A and Staal GE. (1992). Cancer Res. 52, 4773-4778. MEDLINE

Osherov N and Levitzki A. (1994). Eur. J. Biochem. 225, 1047-1053. MEDLINE

Parrizas M and LeRoith D. (1997). Endocrinology 138, 1355-1358. MEDLINE

Parsons JT and Parsons SJ. (1997). Curr. Opin. Cell Biol. 9, 187-192. MEDLINE

Peterson JE, Kulik G, Jelinek T, Reuter WM, Shannon JA and Weber MJ. (1996). J. Biol. Chem. 271, 31562-31571. MEDLINE

Sabe H, Hamaguchi M and Hanafusa H. (1997). Oncogene 14, 1779-1788. MEDLINE

Sanes JR, Rubenstein JL and Nicolas JF. (1986). EMBO J. 5, 3133-3142. MEDLINE

Sato K, Sato A, Aoto M and Fukami Y. (1995). Biochem. Biophy. Res. Comm. 210, 844-851.

Schindler T, Sicheri F, Pico A, Gazit A, Levitzki A and Kuriyun J. (1999). Mol. Cell in press.

Stover DR, Becker M, Liebetanz J and Lydon NB. (1995). J. Biol. Chem. 270, 15591-15597. MEDLINE

Turkson J, Bowman T, Garcia R, Caldenhoven E, de Groot RP and Jove R. (1998). Mol. Cell Biol. 18, 2545-2552. MEDLINE

Van de Vijver MJ, Kumar R and Mendelsohn J. (1991). J. Biol. Chem. 266, 7503-7508. MEDLINE

Verbeek BS, Vroom TM, Adriaansen-Slot SS, Ottenhoff-Kalff AE, Geertzema JG, Hennipman A and Rijksen G. (1996). J. Pathol. 180, 383-388. Article MEDLINE

Yu CL, Meyer DJ, Campbell GS, Larner AC, Carter-Su C, Schwartz J and Jove R. (1995). Science 269, 81-83. MEDLINE

Figure 1 Growth inhibition and morphological changes induced by c-SRC antisense expression in DHER14 cells. (A) DHER14 cells were grown on 96 well plates (3000 cells seeded per well) in the presence or absence of 5 mM IPTG. Every 12 h one 96 well plate was stained with methylene blue and quantified (see Materials and methods). (B) DHER14 cells were grown on 90 mm culture dishes and treated with IPTG for the times indicated, cells were lyzed and after Western blotting probed with the monoclonal anti-Src antibody 327. (C and D) DHER14 cells seeded (3000 cells/well) on 96 well plates were grown for 48 h in the presence (C) or absence (D) of IPTG. Cells were stained with methylene blue (see Materials and methods) and photographed under a light microscope (magnification ´250)

Figure 2 Inhibition of DHER14 colony formation in soft agar after antisense induction or PP1 treatment. (A) DHER14 cells from Src antisense clone (50 000 cells per well) were seeded on a 6 well plate in soft agar (see Materials and methods), in the absence (a) or presence (b) of 5 mM IPTG. (see Materials and methods). (B) DHER14 cells (10⁵ cells per well) were seeded on a 6 well plate in soft agar 20 nM EGF were added to both agar and medium. PP1 was added to the upper medium at the indicated concentrations. (C) Colonies were grown for 10 days, counted and photographed under the microscope (magnification ´100) after MTT staining

Figure 3 Src kinase inhibition reduces Stat3 tyrosine phosphorylation. DHER14 and CSH12 cells were seeded in 6 well plates (7´10⁴ cells/well) and after 24 h were treated with PP1 at the indicated concentrations for 4 h. DHER14 cells were activated with 20 nM EGF (10 min 37°C) after the 4 h incubation with PP1. Cells were lyzed with sample buffer and the lyzates were then run on SDS - PAGE. The Western blot was probed with specific anti-P-Y-Stat3 (1 : 1000 dilution; New England Biolabs). As a control, the blots were probed with anti-Erk-2 (1 : 1000 dilution; Santa Cruz) or anti-Stat3 (1 : 1000 dilution; Transduction Laboratories). Stat3 tyrosine-phosphorylation levels were quantified using the NIH-image program. (A) CSH12 cells. (B) Normalized density of CSH12. (C) DHER14 cells. (D) Normalized density of DHER14 cells

Figure 4 The decrease in c-Src expression correlates with a decrease in Bcl-X_L levels. Clones with the Src antisense mRNA induction systems Lac Switch or Tet were grown and induced as described in Materials and methods at the times indicated. (A and D). Cells were lyzed and after Western blotting probed with the monoclonal anti-Src antibody 327. The blots were also probed with polyclonal antibodies against Bcl-X_L (1 : 1000 dilution; Transduction Laboratories). As a control, the blots of Lac Switch system clones were probed with anti-actin (1 : 2000 dilution; Santa Cruz). The blots of Tet clones were probed with anti-Erk-2 (1 : 1000 dilution; Santa Cruz). (B, C, E and F). c-Src, actin, Erk-2 and Bcl-X_L levels were quantified using the NIH-image program. Bcl-X_L and c-Src levels were normalized to Erk-2 levels. At the 24 h time point (CSH12) less protein was applied to the gel which led to lesser amount of Erk-2. However, the amounts of Src and Bcl-X_L were normalized to the amount of Erk-2. (A - C). DHER14 cells. (D - F). CSH12 cells

Figure 5 Inhibition of Src kinase activity correlates with a decrease in Bcl-X_L levels. CSH12 (A and B) and DHER14 cells (C - F) were grown in 6 well plates (Nunc) in the presence of PP1 at the indicated concentrations for 24 (A and D) or 48 h (E and F). Cells were lyzed with sample buffer and the lysates were then run on SDS - PAGE. The Western blot was probed with specific antibodies and the c-Src, Erk-2 and Bcl-X_L levels were quantified using the NIH-image program. Bcl-X_L levels were normalized to Erk-2 levels. The data shown is representative of at least two experiments

Figure 6 SRC antisense and PP1 induce apoptotic cell death. DHER14 and CSH12 cells were seeded in 6 well plates (7´10⁴ cells/well) or 90 mm plates (5´10⁵) (FACS analysis), and after 24 h were treated with PP1 at the indicated concentrations for 24 or 48 h. Adherent and suspended cells were collected, stained with Annexin-V-FLOUS+Propidium Iodide (PI) solution (Boehringer Mannheim), and photographed in a confocal microscope. The data is representative of three different experiments. c-SRC antisense was co-transfected with Lac-Z gene containing plasmid under the CMV promoter to DHER14 cells. Cells were seeded in 6 well plates (7´10⁴ cells/well). Twenty-four and 72 h after transfection cells were stained with X-Gal (see Materials and methods) and blue cells were counted in ten different fields. The number of blue cells in the 72 h point was normalized to the 24 h point (see Materials and methods). (A) DHER14 cells. a - c: 0, 10, 20 M PP1, phase contrast. (d - f) Annexin-V+PI staining (525+655 nm). The white arrow indicates necrotic cell. g - i: FACS analysis of Annexin-V binding. (B) CSH12 cells. a - c: 0, 5, 20 M PP1, phase contrast. d - f: Annexin-V+PI staining (525+655 nm). g - i: FACS analysis of Annexin-V binding (C). Per cent of apoptotic cells at different PP1 concentrations measured by FACS. (D) Number of blue cells after co-transfection of pOPI3, pOPI3-CRS or HA-BAD together with CMV-Lac-Z plasmid, (a representative experiment out of few)

Figure 7 Transfections of c-SRC anti-sense, kinase-dead c-Src, Stat3 or Src kinase inhibition by PP1 reduces transcription from the Bcl-X promoter. Cells were seeded in 6 well plates (7´10⁴ cells/well) and after 24 h were transfected (See Materials and methods) with the plasmids indicated in the results. After 48 h cells were lyzed with luciferase lysis buffer (Promega) and luciferase and -galactosidase activity were measured. Luciferase activity was normalized to bägalactosidase activities. Data shown is representative of three experiments (A) CSH12 and DHER14 cells, as indicated. (B - C) CSH12 cells