Src kinases mediate STAT growth pathways in squamous cell carcinoma of the head and neck.

Signal transducer and activator of transcription (STAT) proteins are constitutively activated in many malignancies, including squamous cell carcinoma of the head and neck (SCCHN). Previously, we reported that phosphorylation of the epidermal growth factor receptor (EGFR) is linked to activation of STATs 3 and 5 in SCCHN cells. The present study was undertaken to determine the role of Src family kinases in STAT activation and SCCHN growth. The Src family kinases c-Src, c-Yes, Fyn, and Lyn were expressed and activated by transforming growth factor-alpha stimulation in all four SCCHN cell lines examined but not in corresponding normal epithelial cells. In nine SCCHN cell lines tested, Src phosphotyrosine expression levels were highly correlated with activation levels of STATs 3 and 5. Co-immunoprecipitation analysis demonstrated interaction between c-Src and STATs 3 or 5 and EGFR in SCCHN cells, but no heterodimerization was detected between STAT3 and STAT5. SCCHN cells treated with either of two Src-specific inhibitors or transfected with a dominant-negative c-Src construct demonstrated decreased activation of STATs 3 and 5 and reduced growth rates in vitro. These results demonstrate a role for Src kinases in mediating activation of STATs 3 and 5 in concert with the EGFR in SCCHN cells. Strategies to target Src activation may contribute to the treatment of cancers that demonstrate increased levels of EGFR and STATs, including SCCHN.

Signal transducers and activators of transcription (STATs), 1 originally identified in the context of interferon signaling, can be activated by a number of cytokines and growth factors, including EGFR ligands (1,2). Activation of STATs involves phosphorylation of a single tyrosine at the C terminus, heteroor homodimerization, and translocation to the nucleus where STATs bind to consensus elements in the promoter regions of target genes (3,4). STAT proteins are involved in a variety of homeostatic cellular processes, including mitogenesis, differentiation, and apoptosis (5). Accumulating evidence also supports a role for STAT proteins in oncogenesis (6).
It is well established that TGF-␣, derived from either autocrine or paracrine sources, plays an essential role in the malignant transformation and progression of squamous cell cancers of the head and neck (SCCHN) by regulating the growth and survival of head and neck tumor cells (7)(8)(9). TGF-␣ induces intracellular signaling through stimulation of EGFR, which contains a cytoplasmic domain with intrinsic protein-tyrosine kinase activity. In response to ligand, EGFRs dimerize and become phosphorylated on multiple tyrosine residues. These phosphotyrosines, in turn, allow the activated receptor to recruit STATs (signal transducers and activators of transcription) to tyrosines Tyr-1068 and Tyr-1086 (10). Direct interaction between STAT protein src-homology 2 domains and the activated receptor leads to STAT phosphorylation dimerization, nuclear translocation, and gene regulation (11)(12)(13)(14)(15)(16).
Src kinases were initially implicated in STAT activation by studies examining the molecular mechanisms associated with v-Src-mediated transformation of fibroblasts and hematopoietic cell lines (17,18). Subsequent reports have demonstrated an essential role for Src kinases in mediating constitutive STAT activation in many human cancer cells (19,20). In general, Src appears to serve as an intermediate between tumorigenic protein-tyrosine kinases and STAT activation as demonstrated by studies in chronic myelogenous leukemia and breast cancer (21,22). STAT3 tyrosine phosphorylation and stimulation of DNA binding activity by five members of the Src kinase family (Src, Hck, Lyn, Fyn, and Fgr) in both normal and transformed cell types has also been reported (23). Although we have previously linked STAT activation in SCCHN to upstream stimulation of EGFR, the role of Src and/or JAK kinases in activation of STATs has not been determined in SCCHN. We and others have demonstrated that constitutive activation of STATs 3 and 5 are critical in head and neck carcinogenesis (24 -26). Potential mechanisms of STAT activation include activation of upstream receptor kinases (e.g. EGFR and PDGFR) and/or non-receptor kinases (e.g. Src and JAK). We have previously linked activation of STATs 3 and 5 to stimulation of EGFR in SCCHN cells. Targeting EGFR or STAT3 or STAT5 inhibited the growth of these cells in vitro (24,25). The present study was undertaken to determine the role of Src family kinases in mediating EGFR-STAT growth pathways in SCCHN. TGF-␣ stimulated activation of c-Src, Yes, Fyn, and Lyn in all cells tested, thus linking TGF-␣/EGFR autocrine signaling to Src kinase activation in SCCHN cells. Either of two Src-selective inhibitors or transfection of a dominant-negative c-Src mutant inhibited growth and activation of STATs 3 and 5 activation in SCCHN cells. These findings implicate Src family kinases in EGFR-mediated STAT activation and suggest that strategies to interfere with Src may demonstrate therapeutic efficacy in these cancers.

EXPERIMENTAL PROCEDURES
Cells-Cell lines derived from SCCHN patients were grown in Dulbecco's modified Eagle's medium (Cellgro, Washington, D. C.) with 15% fetal bovine serum (Invitrogen, Grand Island, NY), plus 100 units/ml penicillin and 100 units/ml streptomycin (Invitrogen). Several SCCHN cell lines are part of a large collection established in the Department of Otolaryngology at the University of Pittsburgh (PCI-15b and PCI-37a) (27), in addition to UM-22B (28) and -1483 (29). SCCHN cell lines UPCI:SCC-25, UPCI:SCC-66, and UPCI:SCC-104 were gifts from Dr. Susanne M. Gollin (Graduate School of Public Health, University of Pittsburgh). The cell line OSC-19 was derived from a squamous cell carcinoma of the tongue (30). The cell line KB (ATCC CCL-17, American Type Culture Collection, Rockville, MD) was from a oral cavity tumor (31). Normal mucosal epithelial cells were obtained from primary cultures established from oropharyngeal mucosa harvested from patients undergoing non-oncological head and neck procedures as described by us previously (32). Sf-9 cells (Invitrogen) were cultured in Grace's complete insect cell medium containing 10% fetal bovine serum and 50 g/ml gentamicin (23). To generate recombinant baculoviruses and expression of proteins in Sf-9 insect cells, STAT3, STAT5a, or STAT5b cDNAs were subcloned into the baculovirus transfer vectors pVL1392 and pVL1393, respectively. The resulting constructs were used to make recombinant baculoviruses by co-transfection with BaculoGold DNA (BD Pharmingen) using the manufacturer's protocol. For protein expression, Sf-9 cells were grown to 50% confluence on 60-mm tissue culture plates and infected with recombinant baculoviruses for 1 h at 27°C. The virus was replaced with fresh medium, and the cells were incubated for 48 h prior to preparation of cytosolic extracts as described previously (33).

FIG. 1. Stimulation of Src family kinase activity by TGF-␣.
A, cellular lysates were prepared from representative serum-starved SC-CHN cell lines (PCI-15b, UM-22B, PCI-37a, and 1483) and stimulated with TGF-␣ (10 ng/ml) for 8 h, followed by immunoprecipitation with antisera against c-Src, c-Yes, Fyn, or Lyn and immunoblotting with antiphosphotyrosine antibody (PY20) or antisera to each Src family kinase to detect expression and phosphorylation levels. B, a time-course assay was performed to determine the kinetics of Src phosphorylation. SCCHN cells (1483) were serum-starved for 24 h and then treated with TGF-␣ (10 ng/ml). Cell lysates were prepared at 1, 2, 4, 8, and 12 h to determine c-Src expression and phosphorylation. To confirm that the kinetics of Src activation by TGF-␣ correlated with STAT activation, the same lysates were processed for immunoblotting with phosphotyrosine STAT3 antibody or for STAT5 EMSA.

Transfection of SCCHN Cells with Dominant-negative c-Src-1483
cells were transfected with a pUSEamp vector (Upstate Biotechnology, Inc., Lake Placid, NY) containing mutant Src (K296R/Y528F) cDNA using LipofectAMINE (Invitrogen, Grand Island, NY) according to the manufacturer's recommendations. Stably transfected clones were selected for resistance to the neomycin analogue, G418 (Invitrogen). Dominant-negative Src was generated by mutation of Lys-296 to Arg, which rendered the kinase domain incapable of binding ATP.
Electrophoretic Mobility Shift Assay-Adherent cells (Ն10 6 ) were grown in 10-cm plastic tissue culture dishes and harvested by scraping with a rubber policeman. Nuclear extracts were prepared and EMSAs were performed on 4% native polyacrylamide gels as described previously (39,40). Briefly, the cells were centrifuged at 200 ϫ g for 5 min, washed with phosphate-buffered saline, and resuspended in 1 ml of ice-cold buffer containing 10 mM HEPES (pH 7.9), 1.5 mM MgCI 2 , 10 mM KCl, 1 mM NaF, 0.5 mM dithiothreitol (DTT), 0.2 mM phenylmethylsulfonyl fluoride, 1 g/ml pepstatin, 5 g/ml aprotinin, and 2 g/ml leupeptin. After 30-min incubation on ice, the cells were centrifuged at 10,000 ϫ g for 1 min. The pelleted nuclei were resuspended in 100 l of buffer containing 20 mM HEPES (pH 7.9), 25% glycerol, 420 mM NaCl, 1.5 mM MgCI 2 , 0.2 mM EDTA, 1 mM Na 3 VO 4 , 10 mM NaF, 0.5 mM DTT, 0.2 mM phenylmethylsulfonyl fluoride, 1 g/ml pepstatin, 5 g/ml aprotinin, and 2 g/ml leupeptin. Nuclear proteins were liberated during a 30-min incubation on ice. The supernatant was clarified by a 10-min centrifuge (10,000 ϫ g), and the extracts were aliquoted and stored at Ϫ80°C until used. Protein concentrations were determined using a protein assay (Bio-Rad). Probes were prepared by labeling doublestranded oligonucleotides (␤-casein oligonucleotide, high affinity serum-inducible element duplex oligonucleotide, or STAT4 gel shift oligonucleotides) with [␥-32 P]ATP and T4 polynucleotide kinase. Binding was performed in a 20-ml volume with 200 pg of radiolabeled probe, 10 mg of protein, 5 mM Tris (pH 7.9), 15 mM HEPES (pH 7.9), 80 mM KCl, 3.5 mM MgCl 2 , 5 mM EDTA, 0.1% Tween 20, 5 mM DTT, 10% glycerol, and 100 mg/ml poly(dI-dC). After incubation at room temperature for 20  (1483) with or without TGF-␣ stimulation (for 8 h) and SF-9 transfected with STAT5a were lysed, and the proteins were immunoprecipitated with anti-EGFR or c-Src or STAT5a Abs, respectively. The immunoprecipitated proteins were subjected to SDS-PAGE analysis using a gradient gel and analyzed by Western blot with anti-STAT5a Ab. B, the same lysed tumor cell proteins and lysates from SF-9-transfected cells expressing STAT5b were immunoprecipitated with anti-EGFR or c-Src or STAT5b Abs, respectively. The immunoprecipitated proteins were subjected to SDS-PAGE analysis by Western blot with anti-STAT5b Ab. C, the same lysed tumor cell proteins and lysates from SF-9 transfected cells expressing STAT3 were immunoprecipitated with anti-EGFR or c-Src or STAT3 Abs, respectively. The immunoprecipitated proteins were subjected to SDS-PAGE analysis by Western blot with anti-STAT3 Ab.  4) Abs, respectively. The immunoprecipitated proteins were separated by SDS-PAGE followed by immunoblotting with anti-STAT3 (lanes 1 and 2) or anti-STAT5a Abs (lanes 3 and 4). B, the same tumor cell lysates were immunoprecipitated with anti-STAT5b (lanes 1 and 3) or anti-STAT3 (lanes 2 and 4) Abs, respectively. The immunoprecipitated proteins were separated by SDS-PAGE followed by immunoblotting with anti-STAT3 (lanes 1 and 2) or anti-STAT5b Abs (lanes 3 and 4). C, Sf-9 insect cells were infected with STAT3 (3), STAT5a (5a), and c-Src (Src) baculoviruses either individually or in the combinations shown at the top. Uninfected cells were included as a negative control (Con). STAT5a was immunoprecipitated from clarified cell lysates and analyzed for associated STAT3 by immunoblotting. Panel a, recovery of immunoprecipitated STAT5a. Panel b, associated STAT3; note that low levels of Stat3 cross-react with the precipitating STAT5a antibody, but these are not increased in the presence of STAT5 and Src. Panels c and d, STAT5a and STAT3 immunoprecipitates were probed with antiphosphotyrosine antibodies to verify tyrosine phosphorylation by c-Src. Panels e-g, cell lysates were probed with STAT5a, STAT3, and Src antibodies to show equivalent protein expression in each culture. min, the products were electrophoresed on 4% polyacrylamide gels containing 2.5% glycerol and 0.5ϫ TBE (22 mM Tris, 22 mM boric acid, and 0.5 mM EDTA). A 50-fold excess of unlabeled double-stranded oligonucleotides was used as a cold competitor. For supershift studies, the nuclear proteins were incubated for 30 min at room temperature with polyclonal antisera (200 ng) raised against STAT5a and STAT5b (Upstate Biotechnology, Inc.) and STAT3 or STAT4 (Santa Cruz Biotechnology, Santa Cruz, CA) before addition of the probe (25).
Immunoblotting and Immunoprecipitation-Cells were grown to near confluence in 100-mm cultured dishes. Thereafter, cells were washed three times with ice-cold phosphate-buffered saline and harvested in 1 ml of lysis buffer (10 mM Tris-HCl, pH 7.6, 5 mM EDTA, 50 mM NaCl, 30 mM sodium pyrophosphate, 50 mM NaF, 1 mM Na 3 VO 4 , 1% Triton X-100, 1 mM phenylmethylsulfonyl fluoride, 5 mg/ml aprotinin, 1 mg/ml pepstatin A, and 2 mg/ml leupeptin). Aliquots from cleared lysates containing 50 mg of protein were subjected to electrophoresis on SDS-PAGE. Separated proteins were electroblotted to a nitrocellulose membrane (MSI, Westboro, MA). Phosphorylation of c-Src, c-Yes, Fyn, and Lyn was determined by immunoprecipitation with antiphosphotyrosine monoclonal antibody (PY20, Transduction Laboratories, Inc.), followed by immunoblotting with anti-c-Src, -c-Yes, -Fyn, and -Lyn antisera (Santa Cruz Biotechnology). Phosphorylation of c-Src and Lyn was determined by immunoprecipitation with anti-c-Src and -Lyn antisera (Santa Cruz Biotechnology), followed by immunoblotting with phosphospecific antibody that recognizes the tyrosine phosphorylation activation loop of these kinases (PY418, BioSource International, Inc., Camarillo, CA). JAK-2 phosphorylation was determined by immunoblotting with a phospho-specific JAK2 antibody (Y1007 and Y1008) (Upstate Biotechnology Inc.), and JAK-2 expression was determined using an antibody from Santa Cruz Biotechnology. Interaction of STAT5 with EGFR, c-Src, and STAT3 and interaction of STAT3 with EGFR, c-Src (Santa Cruz Biotechnology) was determined by immunoprecipitation with anti-EGFR, c-Src, and STAT3 monoclonal antibodies, followed by immunoblotting with anti-STAT5a, or STAT5b, or STAT3 antisera. (41,42). Src family kinases have not been previously examined in SCCHN. To determine the potential of EGFR ligand to induce activation of Src kinases in SCCHN cells, a series of SCCHN cell lines were serum-starved for 24 h followed by stimulation with recombinant TGF-␣ for 8 h and then immunoprecipitation with specific antisera against c-Src, c-Yes, Fyn, or Lyn and immunoblotting with antiphosphotyrosine antibody (PY20). As shown in Fig. 1A, TGF-␣ stimulated activation of c-Src, Yes, Fyn, and Lyn in all four SCCHN cell lines examined. A time course experiment demonstrated that 8 h of TGF-␣ was necessary under these conditions (Fig. 1B). Under these conditions, the timing of Src activation by TGF-␣ correlated with activation of STATs 3 and 5. These results implicate Src family kinases in a TGF-␣/EGFR autocrine signaling pathway that characterizes many malignancies, including SC-CHN. To determine if engagement of the EGFR is required for activation of Src family kinases, we examined the effects of an EGFR-specific tyrosine kinase inhibitor on Src activation. This EGFR-specific inhibitor has been previously reported not to directly inhibit Src activation (43). As shown in Fig. 2, inhibition of EGFR kinase activity was associated with decreased c-Src phosphorylation, suggesting that EGFR activity is required for Src activation in SCCHN cells.

Expression and Activation of Src Family Kinases in SC-CHN by EGFR Ligand-Parallel activation of c-Src and EGFR and interaction between Src kinases and EGFR has been identified in several human cancers
Src Activation Levels Are Correlated with Constitutive Activation Levels of STATs 3 and 5-We previously reported that constitutive activation levels of STAT3 are correlated with EGFR expression levels in SCCHN cell lines (44). To determine if Src activation was linked to STAT activation, we determined levels of Src phosphorylation by immunoblotting with a phosphospecific antibody that recognizes the tyrosine phosphorylation activation loop of these Src family kinases (PY418). In the absence of phospho-specific antisera that can distinguish STAT5 isoforms, STAT5 activation was determined by EMSA and STAT3 activation was measured by immunoblotting with a phosphotyrosine-specific antibody. In all nine cell lines examined, Src activation levels were highly correlated with constitutive STAT5 activation levels (p Ͻ 0.0001, r ϭ 0.989). Similarly, Src activation levels were also correlated with constitutive STAT3 activation levels in the same SCCHN cell lines (p ϭ 0.0196, r ϭ 0.833) (Fig. 3). To determine the specificity of Src and STAT activation correlations, we examined the potential association of STAT4 with Src kinase levels in SCCHN cells. EGFR inhibition does not abrogate STAT4 activation (data not shown). As shown in Fig. 3, Src activation was not correlated with constitutive STAT4 activation levels in the same SCCHN cell lines (p ϭ 0.9349, r ϭ 0.001. These results suggest that Src activation is specifically linked to activation of STATs 3 and 5 in SCCHN cells. Interaction of c-Src with EGFR and STATs in SCCHN Cells-EGFR has been reported to associate with STATS 1, 3, and 5 in both normal and transformed epithelial cells (45,46). To determine if EGFR and/or STATs interact with c-Src in SCCHN cells, we performed co-immunoprecipitation studies and found that c-Src formed a complex with STAT5a, STAT5b, and STAT3. Moreover, we found that EGFR also interacted with STAT5a, STAT5b, and STAT3 suggesting a transmembrane complex involving EGFR, c-Src, and STATs 3 and 5 in  (Fig. 4). Although EGFR activation induces both STAT3 and STAT5 activation in SCCHN cells, we were unable to detect STAT3-STAT5 heterodimers (Fig. 5, A and B). To confirm this observation in a defined expression system, the possibility of STAT3-STAT5 heterodimer formation was also tested in the Sf-9 cell overexpression system. STAT3 and STAT5a were expressed individually and together in the presence or absence of c-Src as an activating tyrosine kinase.
STAT5a was then immunoprecipitated and tested for the presence of associated STAT3 by immunoblotting. As shown in Fig.  5C, STAT5a-3 heterodimers were not observed in this system, despite high level expression of the STAT proteins and tyrosine phosphorylation in the presence of c-Src. This finding is consistent with the observation in SCCHN cells and suggests that STAT3 and STAT5 function independently in the EGFR/c-Src signaling pathway.  (Neo1-4) were immunoprecipitated with anti-c-Src Ab following immunoblotting with antiphosphotyrosine PY418 against the activating tail loop of Src family kinases or anti-Lyn Ab. B, nuclear extracts were prepared, followed by EMSA for STAT5 activation in four representative clones (DNc-Src1-4) compared with vector-transfected controls (Neo1-4). C, nuclear extracts were prepared, followed by EMSA for STAT3 activation in five representative clones (DNc-Src1-5) compared with vectortransfected controls (Neo1-2). D, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium assay demonstrating decreased growth in a representative DN Src clone compared with vector-transfected control.
Blockade of Src Kinases Inhibits Activation of STATs 3 and 5 and SCCHN Cell Growth-We have previously reported that EGFR blockade abrogates both growth and constitutive activation of STATs 3 and 5 in SCCHN cells (25,47). Others have demonstrated a necessary role for Src family kinases in STAT activation downstream of other oncogenic tyrosine kinases, such as Bcr-Abl (21,48). To determine the role of Src kinases in mediating cell growth and STAT activation in this tumor system, SCCHN cells were treated with either of two pyrrolopyrimidine Src kinase inhibitors (A419259 and PD0180970) (21,49) followed by determinations of STAT activation and proliferation assays. PD1890970 has been shown not to inhibit other kinases, including EGFR, PDGFR, or basic fibroblast growth factor receptor (50). A419259 has been reported to specifically inhibit several Src family kinases, including c-Src and Lyn (21). As shown in Fig. 6, both Src inhibitors blocked phosphorylation of both Lyn and c-Src in SCCHN cells. Blockade of Src kinases was associated with decreased activation of STATs 3 and 5 as well as growth inhibition (Figs. 6, C and D). To verify that EGFR activation was not affected by treatment with Src-selective inhibitors, SCCHN cells (1483) were treated with increasing concentrations of PD180970 or A419259 followed by immunoprecipitation of EGFR and immunoblotting with antiphosphotyrosine antibody (PY99) (Fig. 6E). JAK kinases have also been implicated in constitutive STAT activation in cancer cells (20).
To determine the role of JAK in SCCHN, cells were treated with a JAK-2-specific inhibitor (AG490) (51) at a dose demonstrated to block JAK-2 phosphorylation (100 nM). Treatment of SCCHN cells with AG490 did not abrogate cell growth or constitutive activation of STATs 3 or 5 in either of two SCCHN cell lines tested (Fig. 7).
Expression of Dominant-negative Src Inhibits Constitutive STAT Activation and Cell Growth in SCCHN Cells-The use of pharmacological inhibitors is limited by the difficulty of demonstrating specificity at precise dosing levels. Therefore, we also employed a dominant-negative strategy to determine the role of Src kinases. SCCHN cells were stably transfected with a dominant-negative construct containing an kinase-inactive form of c-Src. Following isolation of stable clones, cells were immunoprecipitated with c-Src antibody followed by immunoblotting with a phosphospecific antibody that recognizes the tyrosine phosphorylation activation loop of these Src family kinases (PY418) or the same c-Src antibody (as a control) to determine the effects of mutant Src on expression and activation levels of c-Src. As shown in Fig. 8A, SCCHN cells expressing mutant Src demonstrated decreased expression of phosphorylated c-Src. In contrast, expression and activation levels of the other Src family kinases (Fyn, Lyn, and c-Yes) were not modulated by stable expression of mutant c-Src. When analyzed by gel shift assay, lower levels of constitutive STAT5 and STAT3 activation levels were detected in the dominant-negative transfectants compared with levels in vector-transfected control clones (Fig. 8, B and C). Growth assays also demonstrated a 34% decrease proliferation in the dominant-negative c-Src transfectants compared with vector-transfected controls (Fig. 8D). These results demonstrate that c-Src activation plays a role in activation of growth pathways mediated by STATs 3 and 5 in SCCHN. DISCUSSION The present study investigated the role of Src family kinases in mediating EGFR-STAT growth pathways in SCCHN. The EGFR ligand TGF-␣ stimulated activation of c-Src, Yes, Fyn, and Lyn in all SCCHN cells tested, thus linking TGF-␣/EGFR autocrine signaling to Src kinase activation in SCCHN cells. Src activation was highly correlated with constitutive activa-tion of STATs 3 and 5 but not with STAT4 activation in SC-CHN. Co-immunoprecipitation studies demonstrated interaction of EGFR and c-Src with STAT5a/b or STAT3 in SCCHN cells but no interaction between STAT5a or STAT5b with STAT3 was detected. Either of the 2 Src-selective inhibitors (PD0180970 and A419259) or a dominant-negative c-Src mutant decreased growth and activation of STATs 3 and 5 in SCCHN cells, but AG490, a JAK-2 specific inhibitor, had no effect on cell proliferation or STAT activation.
Src kinases have been reported to regulate a variety of cell functions, including cell cycle progression, growth, survival, and migration (52,53). Studies using human tumor tissues and tumor-derived cell lines have demonstrated that enhanced tumorigenicity in vivo is correlated with elevated c-Src expression and tyrosine phosphorylation (54,55). The only report on Src family kinases in SCCHN showed that c-Src was overexpressed in areas of hyperproliferation, dysplastic epithelium, benign papillomas, and inflamed normal tissue indicating that c-Src contributes to the increased protein-tyrosine kinase activity in SCCHN carcinogenesis (56). In the present study, we demonstrated other Src family kinases, including c-Yes, Fyn, and Lyn, in addition to c-Src, are also expressed in SCCHN. Furthermore, activation of these Src family kinases in this tumor system can be linked to upstream stimulation and engagement of EGFR.
Several reports have demonstrated an essential role for Src kinases (generally c-Src) in mediating constitutive STAT activation in human cancer cells (19,20). Abrogation of Src has been shown to lead to diminished tyrosine phosphorylation of STAT5 in K562 leukemia cells, where STAT5 phosphorylation was reduced in the presence of Src inhibitors (57). To date, Src-mediated STAT activation has not been reported in SC-CHN. Our results demonstrate that Src activation was highly correlated with constitutive activation of STATs 3 and 5 in SCCHN. Abrogation of Src kinase activity using pharmacological inhibitors or abrogation of c-Src by a dominant-negative mutant specifically decreased constitutive activation of STAT3 and STAT5 in SCCHN suggesting that Src kinases contribute to the regulation of STAT activation in this tumor system. Similar results have been reported in cell lines derived from human breast cancer and vulvar squamous carcinoma (20,45) Potential mechanisms of STAT activation in oncogenesis include activation of upstream receptor kinases (e.g. EGFR and PDGFR) and/or non-receptor kinases (e.g. Src and JAK). Targeting EGFR kinase activity or STAT3 or STAT5 inhibited the growth of SCCHN cells in vitro (25,47,58). The differential effects of Src/Abl kinases on the nuclear translocation of STAT5b and STAT5a suggested that Src activation was able to result in tyrosine phosphorylation and DNA binding of both STAT5a and STAT5b. However, Src-induced nuclear translocation of only STAT5B but not STAT5A was observed (59). STAT5 activation by Src may occur by a mechanism distinct from that implicated in cytokine activation of the JAK/STAT pathway, resulting in the selective nuclear translocation of STAT5b (59). Our results suggest that Src kinases are linked to upstream activation of EGFR by ligand and downstream activation of STATs 3 and 5 in SCCHN. Co-immunoprecipitation studies demonstrated interaction of EGFR and c-Src with STAT5a/b or STAT3 in SCCHN cells, but no interaction of STAT5a or STAT5b with STAT3 could be detected. Src family kinases may serve to facilitate interaction between STATs and EGFR or form a multiple molecular kinase complex to transduce signals from EGFR to STATs. It is also likely that there are other pathways, in addition to Src family kinases, that contribute to STAT activation.
Janus tyrosine kinases (JAKs) have been reported to play an essential role in mediating constitutive STAT activation in many human cancer cells (20, 60 -63). However, in vulvar squamous carcinoma cells (A431), EGF-and neu differentiation factor-induced STAT activation was found to be dependent on Src but not JAK kinases (45). Upon EGF stimulation, c-Src was rapidly recruited to STAT/ErbB receptor complexes. Pharmacological Src kinase inhibitors and a dominant-negative c-Src ablated both STAT and JAK tyrosine phosphorylation, but transfection of a dominant-negative JAK construct did not affect EGF-induced STAT phosphorylation (45). Activation of JAKs was not involved in interleukin-3-induced activation of STATs in myeloid cell proliferation (64). In the present investigation, JAK2 kinase activity was shown to play a non-essential role in STAT activation and cell growth in SCCHN. Our cumulative findings suggest that several Src family kinases, including c-Src and Lyn, contribute to EGFR-mediated STAT growth pathways in SCCHN where strategies to interfere with Srcs may demonstrate therapeutic efficacy.