Tyrosine Kinase p56lck Regulates Cell Motility and Nuclear Factor κB-mediated Secretion of Urokinase Type Plasminogen Activator through Tyrosine Phosphorylation of IκBα following Hypoxia/Reoxygenation*

Nuclear factor κB (NFκB) plays major role in regulating cellular responses as a result of environmental injuries. The molecular mechanism(s) by which hypoxia/reoxygenation (H/R) regulates p56lck-dependent activation of NFκB through tyrosine phosphorylation of IκBα and modulates the expression of downstream genes that are involved in cell migration in human breast cancer cells are not well defined. In this paper, we investigated the involvement of protein-tyrosine kinase p56lck in the redox-regulated activation of NFκB following H/R in highly invasive (MDA-MB-231) and low invasive (MCF-7) breast cancer cells. We demonstrated that H/R induces tyrosine phosphorylation of p56lck, nuclear translocation of NFκB, NFκB-DNA binding, and transactivation of NFκB through tyrosine phosphorylation of IκBα. Transfection of these cells with wild type Lck but not with mutant Lck F394 followed by H/R induces the tyrosine phosphorylation of inhibitor of nuclear factor κB (IκBα) and transcriptional activation of NFκB, and these are inhibited by Lck inhibitors. In vitro kinase assay demonstrated that immunoprecipitated p56lck but not Lyn or Fyn directly phosphorylate IκBα in presence of H/R. Pervanadate, H2O2, and H/R induce the interaction between Lck and tyrosine-phosphorylated IκBα, and this interaction is inhibited by Src homology 2 domain inhibitory peptide, suggesting that tyrosine-phosphorylated IκBα interacts with Src homology 2 domain of Lck. Luciferase reporter gene assay indicated that Lck induces NFκB-dependent urokinase type plasminogen activator (uPA) promoter activity in presence of H/R. Furthermore, H/R stimulates the cell motility through secretion of uPA. To our knowledge, this is the first report that p56lck in presence of H/R regulates NFκB activation, uPA secretion, and cell motility through tyrosine phosphorylation of IκBα and further demonstrates an important redox-regulated pathway for NFκB activation following H/R injury that is independent of IκB kinase/IκBα-mediated signaling pathways.

Nuclear factor B (NFB) plays major role in regulating cellular responses as a result of environmental injuries. The molecular mechanism(s) by which hypoxia/ reoxygenation (H/R) regulates p56 lck -dependent activation of NFB through tyrosine phosphorylation of IB␣ and modulates the expression of downstream genes that are involved in cell migration in human breast cancer cells are not well defined. In this paper, we investigated the involvement of protein-tyrosine kinase p56 lck in the redox-regulated activation of NFB following H/R in highly invasive (MDA-MB-231) and low invasive (MCF-7) breast cancer cells. We demonstrated that H/R induces tyrosine phosphorylation of p56 lck , nuclear translocation of NFB, NFB-DNA binding, and transactivation of NFB through tyrosine phosphorylation of IB␣. Transfection of these cells with wild type Lck but not with mutant Lck F394 followed by H/R induces the tyrosine phosphorylation of inhibitor of nuclear factor B (IB␣) and transcriptional activation of NFB, and these are inhibited by Lck inhibitors. In vitro kinase assay demonstrated that immunoprecipitated p56 lck but not Lyn or Fyn directly phosphorylate IB␣ in presence of H/R. Pervanadate, H 2 O 2 , and H/R induce the interaction between Lck and tyrosine-phosphorylated IB␣, and this interaction is inhibited by Src homology 2 domain inhibitory peptide, suggesting that tyrosinephosphorylated IB␣ interacts with Src homology 2 domain of Lck. Luciferase reporter gene assay indicated that Lck induces NFB-dependent urokinase type plasminogen activator (uPA) promoter activity in presence of H/R. Furthermore, H/R stimulates the cell motility through secretion of uPA. To our knowledge, this is the first report that p56 lck in presence of H/R regulates NFB activation, uPA secretion, and cell motility through tyrosine phosphorylation of IB␣ and further demonstrates an important redox-regulated pathway for NFB activation following H/R injury that is independent of IB kinase/IB␣-mediated signaling pathways.
Tumor invasion, malignant progression, and distant metastasis depend on complex multistep processes. One prerequisite is the ability of tumor cells to initiate extracellular proteolysis, which is required for the crossing of tissue barriers, cell migration, extracellular matrix invasion, tissue remodeling, and angiogenesis. There is abundant experimental evidence that urokinase type plasminogen activator (uPA), 1 a member of serine protease plays a major role in malignant progression and tumor metastasis (1). Up-regulations of uPA and urokinase type plasminogen activator receptor (uPAR) have been described in many human tumors. High levels of uPA and uPAR in tumor tissues are associated with poor prognosis of patients with cancer of the breast, lung, head and neck, uterine, cervix, bladder, and ovary (2). Most tumors are characterized by low extracellular pH, glucose depletion, high lactate levels, and regions with low oxygen tensions (3,4). Low oxygen tension in tumors have been associated with poor outcome (5,6), enhanced local or locoregional spread (7), and enhanced metastatic potential (8,9). Hypoxia is a key parameter, able to modulate the expression of a variety of genes that are involved in tumor angiogenesis, malignant progression, and distant metastasis (10 -12).
Lck, a member of the Src family non-receptor protein-tyrosine kinase, is mostly expressed in T cells and some B cells. Lck is also expressed in breast cancer tissues and cell lines (13,14). Lck binds to the cytoplasmic domain of CD4 and CD8 and plays an essential role in T cell activation and development (15,16). Lck is required for T cell receptor signaling in human Jurkat T cells and for antigen receptor-dependent cytolytic effector function in the CTLL-2 T cells (17,18). Moreover, mice lacking expression of functional Lck or overexpressing an inactive form of Lck have severely disrupted thymocyte development (19,20). Lck is a typical Src-like tyrosine kinase, and its activity is regulated by phosphorylation of a highly conserved tyrosine residue, Tyr-505, located near the carboxyl terminus (21,22). The protein-tyrosine kinase C-terminal Src kinase phosphorylates Lck and is a natural inhibitor of Lck kinase activity (23). In vitro, Lck undergoes autophosphorylation at Tyr-394 (24,25), and the extent of phosphorylation at Tyr-394 correlates with Lck activity and appears to be required for maximum catalytic activity (26,27). Studies of oxidative stress on Src family tyrosine kinases have yielded contradictory results. Nakamura et al. (28) reported that treatment of human peripheral blood T lymphocytes with oxidant diamide (1,1Јazobis(N,N-dimethyl formamide)) increased p56 lck catalytic activity and induced phosphorylation of p56 lck at both Tyr-394 and Tyr-505 residues (28). In contrast, H 2 O 2 increased the catalytic activity of Lck in most of the cell types (29) and phosphorylation of Tyr-394 is apparently required for H 2 O 2induced activation of Lck (30).
The NFB family consists of several members including p65, p50, RelB, and c-Rel molecules (31,32). NFB is a redox-sensitive transcription factor usually retained in the cytoplasm by its inhibitory protein termed IB␣ (33)(34)(35). The phosphorylation-dependent inactivation of IB␣ leads to the translocation of NFB into the nucleus, where it acts as a transcription factor. The phosphorylation of IB␣ has been extensively studied, which results in establishment of two distinct pathways; one involves serine-threonine phosphorylation and degradation of IB␣ through ubiquitination, and the other parallel and less explored pathway is through tyrosine phosphorylation of IB␣ (36,37). As compared with IKK-mediated serine phosphorylation of IB␣, tyrosine phosphorylation of IB␣ is capable of activating NFB in the absence of proteasome-dependent degradation of IB␣ and the exact molecular mechanism(s) of activation of NFB through tyrosine phosphorylation of IB␣ are not well established.
The Src family tyrosine kinase p56 lck is known to induce IB␣ tyrosine phosphorylation and NFB transactivation in T-lymphocyte upon pervanadate treatment (a tyrosine phosphatase inhibitor) (37,38). Loss of tyrosine kinase p56 lck in Jurkat cells abolished NFB activation and partially suppressed and delayed phosphorylation of Tyr-42 of IB␣ upon pervanadate treatment (38). Similarly, tyrosine phosphorylation of IB␣ is observed in bone marrow macrophage followed by TNF␣ treatment and this phosphorylation requires c-Src activity (39). Recent data demonstrate the involvement of c-Src kinase in H/R-induced tyrosine phosphorylation of IB␣. In this study, the triple knockout cell lines (c-Src Ϫ/Ϫ , Fyn Ϫ/Ϫ , and Yes Ϫ/Ϫ ) show residual IB␣ tyrosine phosphorylation upon treatment with H 2 O 2 or induction with H/R (40). These data clearly suggest that other tyrosine kinases are involved in these processes. Because most of the studies on IB␣ tyrosine phosphorylation has been carried out in hematopoietically derived T-cells, the functional significance of these on breast cancer-specific epithelial models remains unsolved.
uPA is a member of the serine protease family, which induces the conversion of plasminogen to plasmin. Plasmin regulates cell invasion by degrading matrix proteins such as fibronectin, type IV collagen, and laminin or indirectly by activating matrix metalloproteinases (1). Previous reports indicated that uPA plays a significant role in tumor growth and metastasis (1,2). The signaling pathways by which hypoxia/ reoxygenation controls uPA secretion through activation of NFB in breast cancer cells are not well defined.
In this study, we demonstrate the involvement of p56 lck in the redox-mediated nuclear translocation of NFB, NFB-DNA binding, transactivation of NFB, and NFB-dependent uPA promoter activity following H/R in MCF-7 and MDA-MB-231 cells. We also showed that H/R induces phosphorylation of p56 lck and stimulates Lck-mediated NFB activation through tyrosine phosphorylation of IB␣. These lead to the induction of uPA secretion, and all of these ultimately control the cell motility, invasiveness, and metastatic spread of breast cancer.
Cell Culture-The MDA-MB-231 and MCF-7 cells were purchased from ATCC (Manassas, VA). Both MDA-MB-231 and MCF-7 cells were cultured in Dulbecco's modified Eagle's medium. The media were supplemented with 10% fetal calf serum, 100 units/ml penicillin, 100 g/ml streptomycin, and 2 mM glutamine in a humidified atmosphere of 5% CO 2 and 95% air at 37°C.
Hypoxic Cultures-The MCF-7 and MDA-MB-231 cells grown to 50 -70% confluence were made hypoxic in evacuation chambers by intermittent application of vacuum and sparging with 95% N 2 , 5% CO 2 . Cells were analyzed at this point or maintained under hypoxic conditions in presence of 100 mM dithionate (an oxygen scavenger) at 37°C for 0 -24 h. Under these conditions, oxygen levels drop to Ͻ10 mm Hg. Cells were either harvested or reoxygenated for the indicated periods of time by replacing the medium with fresh medium and incubating the cultures in a humidified atmosphere of 5% CO 2 and 95% air at 37°C.
Plasmids and DNA Transfection-Wild type Lck and mutant Lck F394 (mutated at tyrosine 394 to phenylalanine) cDNAs in an expression vector (pCEP4) were a generous gift from Dr. Bartholomew M. Sefton (Salk Institute, La Jolla, CA). The luciferase reporter construct (pNFB-Luc) containing five tandem repeats of the NFB binding site was a kind gift from Dr. Rainer de Martin (University of Vienna, Vienna, Austria). Full-length human uPA promoter (Ϫ2062 to ϩ27) in luciferase reporter gene plasmid pGL2 basic was a generous gift from Dr. Ute Reuning (Universitaet Muenchen, Germany). Both MCF-7 and MDA-MB-231 cells were transiently transfected with each of the above cDNA construct using LipofectAMINE Plus according to the instructions from the manufacturer (Invitrogen). Briefly, cDNA (8 g) was mixed with Plus reagent, and then cDNA Plus reagent was incubated with LipofectAMINE. The LipofectAMINE Plus cDNA complex was added to the cells and incubated further at 37°C for 12 h. The control cells received LipofectAMINE Plus alone. The cell viability was detected by a trypan blue dye exclusion test. After incubation, medium was removed, and the cells were refed with fresh medium and maintained for an additional 12 h. These transfected cells were individually or in combination used for Lck kinase assay, interaction studies between Lck and tyrosine-phosphorylated IB␣, EMSA, luciferase reporter gene assay, cell migration assay, and detection of uPA expression by Western blot analysis.
Immunoprecipitation and in Vitro Kinase Assay-Both MCF-7 and MDA-MB-231 cells were individually induced by hypoxia for 0 -24 h and reoxygenated for 90 min. Cells were lysed in lysis buffer (1% Triton X-100 solution containing 1 mM phenylmethylsulfonyl fluoride, 20 g/ml leupeptin, and 2 mM EDTA). The protein concentrations in the cleared lysates were measured by Bio-Rad protein assay. The equal amount of total proteins in lysates were immunoprecipitated with rabbit polyclonal anti-Lck antibody, and half of the immunoprecipitated samples were resolved by SDS-PAGE and analyzed by Western blot using mouse anti-phosphotyrosine antibody. The remaining half of the samples was analyzed by Western blot using anti-Lck antibody. In separate experiments, both these cells were exposed with hypoxia for 3 h and then induced by H/R for 0 -135 min. Cells were lysed and immunoprecipitated with rabbit polyclonal anti-Lck antibody. The levels of phospho-and non-phospho-Lck in the immunoprecipitated samples were detected by Western blot analysis using mouse monoclonal anti-phosphotyrosine and rabbit polyclonal anti-Lck antibodies, respectively. As loading control, the expression of actin was analyzed by Western blot using anti-actin antibody.
In other experiments, these cells were transiently transfected with wild type Lck or mutant Lck F394 in pCEP4 in presence of Lipo-fectAMINE Plus. These cells were exposed with hypoxia for 3 h and then induced by H/R for 90 min. These transfected cells were immunoprecipitated with anti-Lck antibody, and tyrosine phosphorylation of Lck was detected by immunoblotting with anti-phosphotyrosine antibody. The same blots were reprobed with anti-Lck antibody. The expression of actin was detected by Western blot using anti-actin antibody as loading control.
In separate experiments, both these cells were exposed with hypoxia for 3 h and then induced by reoxygenation for 90 min. The cell lysates containing equal amount of total proteins were individually immunoprecipitated with anti-Lck, anti-Lyn, or anti-Fyn antibodies, and kinase assay was performed. Half of the immunoprecipitated samples were incubated with recombinant IB␣ and 2 Ci of [␥-32 P]ATP in kinase assay buffer (25 mM Hepes (pH 7.7) containing 0.2 mM Na 3 VO 4 , 10 M ATP, 5 mM MgCl 2 , 3 mM MnCl 2 , 2 M phenylmethylsulfonyl fluoride, 10 mg/ml aprotinin, 1 g/ml leupeptin, 1 g/ml pepstatin, and 1 mM dithiothreitol) at 30°C for 30 min. The kinase reaction was terminated by addition of SDS-sample buffer. The sample was resolved by SDS-PAGE, dried, and autoradiographed. In another experiments, these cells were individually pretreated with Lck inhibitors (16 M emodin or 4 nM pp2), induced by H/R, immunoprecipitated with anti-Lck antibody, and kinase assay was performed. In other experiments, cell lysates were immunoprecipitated with anti-Lck antibody and kinase assay was performed in presence of tyrosine phosphatase (TP). The other half of the immunoprecipitated samples were analyzed by Western blot using anti-Lck, anti-Lyn, or anti-Fyn antibody.
The effect of H/R on tyrosine phosphorylation of IB␣ was detected by exposing the cells with hypoxia for 0 -24 h and reoxygenated for 90 min. The cell lysates containing equal amounts of total protein were immunoprecipitated with anti-IB␣ antibody. Half of the immunoprecipitated samples were resolved by SDS-PAGE and analyzed by Western blot using anti-phosphotyrosine antibody. The same blots were reprobed with anti-IB␣ antibody. The remaining half of the immunoprecipitated samples was analyzed by Western blot using anti-phosphoserine-specific IB␣ antibody. As loading control, the expression of actin was analyzed by Western blot using anti-actin antibody. Similarly, the effect of reoxygenation on tyrosine phosphorylation of IB␣ was also analyzed by exposing the cells with hypoxia for 3 h, followed by reoxygenation for 0 -135 min. Cell lysates were immunoprecipitated with anti-IB␣ antibody, and half of the immunoprecipitated samples were immunoblotted with anti-phosphotyrosine antibody. The same blots were reprobed with anti-IB␣ antibody. The remaining half of the immunoprecipitated samples was analyzed by Western blot using antiphosphoserine-specific IB␣ antibody. The expression of actin was detected by Western blot analysis as loading control.
To delineate whether PMA or TNF␣ regulates the serine/threonine phosphorylation and degradation of IB␣ in breast cancer cells, both MCF-7 and MDA-MB-231 cells were treated with PMA (10 ng/ml) or TNF␣ (0.1 nM) for 0 -135 min. Cells were lysed in lysis buffer, and cell lysates containing equal amount of total proteins were individually analyzed by Western blot using mouse monoclonal anti-phosphoserine IB␣ antibody. The same blots were reprobed with rabbit polyclonal anti-IB␣ and anti-actin antibodies, respectively.
The role of Lck on tyrosine/serine phosphorylation of IB␣ was examined by transiently transfecting the cells with wild type Lck or Lck F394 in pCEP4 followed by induction with hypoxia for 3 h and reoxygenation for 90 min. These transfected cells were immunoprecipitated with anti-IB␣ antibody, and the levels of tyrosine-phosphorylated IB␣, serine-phosphorylated IB␣, and non-phosphorylated IB␣ in the immunoprecipitated samples were detected by Western blot analysis using their specific antibodies as described above.
The effect of Lck inhibitors (emodin or pp2) on tyrosine/serine phosphorylation of IB␣ was detected by treating the cells with various concentrations of emodin (0 -10 M) or pp2 (0 -4 nM). These cells were exposed with hypoxia for 3 h and reoxygenated for 90 min and lysed in lysis buffer. The cell lysates were immunoprecipitated with anti-IB␣ antibody and the levels of tyrosine-phosphorylated IB␣, serine-phosphorylated IB␣, and non-phosphorylated IB␣ were analyzed by Western blot using their specific antibodies. As loading control, the expression of actin was analyzed by Western blot using anti-actin antibody.
The direct interaction between tyrosine-phosphorylated IB␣ and Lck in these cells were examined by individually treating the cells with either 250 M pV for 20 min or 5 mM H 2 O 2 for 20 min or by inducing with hypoxia for 3 h followed by reoxygenation for 90 min. The cell lysates containing equal amount of total proteins were immunoprecipitated with rabbit polyclonal anti-Lck antibody and analyzed by Western blot using anti-IB␣ or anti-phosphotyrosine antibody. Similarly, to check the effect of SH2 domain inhibitory peptide on H/R-induced interaction between Lck and tyrosine-phosphorylated IB␣, both these cells were individually pretreated with different concentrations of this peptide (0 -150 nM) and then induced by H/R as described above. These cell lysates were immunoprecipitated with rabbit polyclonal anti-Lck antibody, and the levels of tyrosine-phosphorylated IB␣ and IB␣ were detected by Western blot analysis using anti-phosphotyrosine and anti-IB␣ antibodies, respectively.
To check whether tyrosine 394 of Lck is responsible for interaction with phosphorylated IB␣, both MCF-7 and MDA-MB-231 cells were individually transfected with wild type or mutant Lck F394 in presence of LipofectAMINE Plus and then induced by H/R. These cell lysates were immunoprecipitated with anti-Lck antibody and analyzed by Western blot using anti-IB␣ or anti-phosphotyrosine antibody. To delineate the effect of Lck inhibitors (emodin or pp2) on these interaction, both these cells were pretreated with different concentration of emodin (0 -10 M) or pp2 (0 -4 nM) and then induced by H/R as described above. These cells were immunoprecipitated with anti-Lck antibody and immunoblotted with anti-IB␣ antibody. As loading control, the expression of actin was analyzed by Western blot using anti-actin antibody.
Immunofluorescence Study-Both MCF-7 and MDA-MB-231 cells were grown in monolayer on silicon-coated glass slides and then induced by hypoxia for 3 h and reoxygenated for 0 -120 min. In separate experiments, cells were pretreated with various concentrations of emodin (0 -10 M) or pp2 (0 -4 nM) at 37°C for 6 h and then induced by hypoxia for 3 h followed by reoxygenation for 90 min. The cells were fixed in ice-cold methanol for 15 min, blocked with 5% bovine serum albumin for 30 min, and washed with phosphate-buffered saline (pH 7.4). The fixed cells were incubated with rabbit polyclonal anti-NFB, p65 antibody (1:100 dilution) at room temperature for 2 h. The cells were washed with phosphate-buffered saline (pH 7.4) and incubated with FITC-conjugated anti rabbit IgG for 1 h at room temperature. The cells were washed, mounted with coverslips, and analyzed under confocal microscopy (Zeiss).
NFB Luciferase Reporter Gene Assay-The semiconfluent MCF-7 and MDA-MB-231 cells grown in 6-well plates were individually transfected with a luciferase reporter construct (pNFB-Luc) containing five tandem repeats of the NFB-binding site (a generous gift from Dr. Rainer de Martin, University of Vienna, Vienna, Austria) using Lipo-fectAMINE Plus reagent. The transfection efficiency was normalized by cotransfecting the cells with pRL vector (Promega) containing a fulllength Renilla luciferase gene under the control of a constitutive promoter. These transfected cells were induced by hypoxia for 3 h and reoxygenated for 24 h. In separate experiments, both these cells transfected with pNFB-Luc were treated with varying concentrations of emodin (0 -16 M) or pp2 (0 -4 nM) at 37°C for 6 h and then induced by H/R as described above. In another experiments, these cells were transiently cotransfected with pNF-B-Luc and wild type Lck or Lck F394 in pCEP4 and then induced by H/R. Cells were harvested in passive lysis buffer (Promega). The luciferase activities were measured by luminometer (Lab Systems) using the dual luciferase assay system according to the instructions from the manufacturer (Promega). Changes in luciferase activity with respect to control were calculated.
EMSA-Electrophoretic mobility shift assay was performed as described previously (41). Briefly, MCF-7 and MDA-MB-231 cells were induced by hypoxia for 3 h followed by reoxygenation for 90 min. In separate experiments, cells were transiently transfected with wild type Lck or Lck F394 in presence of LipofectAMINE Plus or pretreated with Lck inhibitors (emodin and pp2) and then induced by H/R as described above. The nuclear extracts were prepared, and the extracts (10 g) were incubated with 16 fmol of 32 P-labeled double-stranded NFB oligonucleotide (5Ј-AGT TGA GGG GAC TTT CCC AGG-3Ј) in binding buffer (25 mM Hepes (pH 7.9), 0.5 mM EDTA, 0.5 mM dithiothreitol, 1% Nonidet P-40, 5% glycerol, and 50 mM NaCl) containing 2 g of salmon sperm DNA. The DNA-protein complex was resolved by a native polyacrylamide gel and analyzed by autoradiography. For super shift assay, the nuclear extracts from MDA-MB-231 cells were incubated with anti-p65 antibody for 30 min at room temperature and analyzed by EMSA.
uPA Promoter Construct and Luciferase Reporter Gene Assay-The full-length human uPA-promoter (Ϫ2062 to ϩ 27) was cloned into the luciferase reporter gene plasmid, pGL2-Basic. Both MCF-7 and MDA-MB-231 cells were individually cotransfected with uPA promoterdriven luciferase reporter plasmid pGL2-basic and pRL vector containing full-length Renilla luciferase gene under the control of constitutive promoter. In separate experiments, cells transfected with uPA promoter-driven pGL2-basic were cotransfected with wild type Lck or Lck F394 in presence of LipofectAMINE Plus. In another experiments, transfected cells were pretreated with various concentrations of Lck inhibitors (0 -16 M emodin or 0 -4 nM pp2) or NFB inhibitory peptides (100 g/ml SN-50 or SN-50M). Cells were induced by hypoxia for 3 h and reoxygenated for 24 h. Cells were harvested in passive lysis buffer (Promega). The luciferase activities were measured by luminometer (Lab Systems) using dual luciferase assay system according to the instructions from the manufacturer (Promega). Changes in luciferase activity with respect to control were calculated.
Western Blot Analysis-The effect of H/R on uPA secretion was examined by Western blot analysis. The cells were either induced by hypoxia for 3 h and reoxygenation for 0 -16 h or pretreated with emodin (0 -16 M), pp2 (0 -4 nM), SN-50 (100 g/ml), and SN-50 M (100 g/ml) and then induced by H/R. In separate experiments, the cells were transiently transfected with wild type Lck or mutant Lck F394 in presence of LipofectAMINE Plus and then induced by H/R. The cells were lysed in lysis buffer, and the level of uPA was detected by Western blot analysis using anti-uPA antibody. The blots were reprobed with anti-actin antibody as loading controls.
Cell Migration Assay-The migration assay was conducted using a Transwell cell culture chamber according to the standard procedure as described (42)(43)(44). Briefly, both these cells were individually pretreated with emodin (0 -16 M), pp2 (0 -4 nM), SN-50 (100 g/ml), SN-50M (100 g/ml), or anti-uPA antibody (10 g/ml). In another experiments, cells were also transiently transfected with wild type Lck or mutant Lck F394 in presence of LipofectAMINE Plus. These cells were induced by hypoxia for 3 h and reoxygenated for 16 h. Cells were harvested using trypsin-EDTA and collected by centrifugation at 800 ϫ g for 10 min. The cell suspension (5 ϫ 10 5 cells/well) was added to the upper chamber of the prehydrated polycarbonate membrane filter. The lower chamber was filled with fibroblast conditioned medium, which acted as chemoattractant. The cells were incubated in a humidified incubator with 5% CO 2 and 95% air at 37°C for 16 h. The non-migrating cells on the upper side of the filter were scraped and washed. The migrating cells on the reverse side of the filter were stained with Giemsa. The migrating cells on the filter were counted, and a photomicrograph was taken under an Olympus inverted microscope. The number of cells migrated under normoxic condition were considered as 100%. Preimmune IgG were used as nonspecific control.

H/R Induces Tyrosine Phosphorylation of Lck and Lck Kinase Activity in Breast
Cancer Cells-We have investigated the effects of H/R on Lck-dependent NFB-mediated uPA secretion through tyrosine phosphorylation of IB␣ in breast cancer cells. Accordingly, we first checked whether H/R is able to induce tyrosine phosphorylation of Lck in highly invasive (MDA-MB-231) and low invasive (MCF-7) breast cancer cells. Both MCF-7 and MDA-MB-231 cells were individually induced by hypoxia for 0 -24 h and reoxygenated for 90 min. Cells were lysed, and equal amounts of total proteins from the cell lysates were subjected to immunoprecipitation with rabbit polyclonal anti-Lck antibody. The immunocomplex was resolved by SDS-PAGE and detected by Western blot analysis using anti-phosphotyrosine antibody. Our data revealed that maximum tyrosine phosphorylation of Lck was observed when both MCF-7 (Fig. 1A, upper panel, lanes 1-7) and MDA-MB-231 (Fig. 1B, upper panel, lanes 1-7) cells were exposed with hypoxia for 3 h followed by reoxygenation for 90 min. The remaining half of the immunoprecipitated samples were resolved by SDS-PAGE and immunoblotted with anti-Lck antibody. The level of non-phospho-Lck was unchanged in both these cells (middle panels, lanes 1-7). To check the effect of reoxygenation on tyrosine  A and B, lanes 1-7), whereas the expression of non-phospho-Lck was unchanged (middle panels of A and B, lanes 1-7). Actin was used as loading control (lower panels of A and B). C and D, both MCF-7 (panel C) and MDA-MB-231 (panel D) cells were exposed with hypoxia for 3 h and reoxygenated for 0 -135 min. Equal amounts of total proteins from cell lysates were immunoprecipitated with rabbit polyclonal anti-Lck antibody. Half of the immunoprecipitated samples were analyzed by Western blot using anti-phosphotyrosine antibody (upper panels of C and D, lanes [1][2][3][4][5][6][7][8][9], and the remaining half of the samples were immunoblotted with anti-Lck antibody (middle panels of C and D, lanes [1][2][3][4][5][6][7][8][9]. Note that maximum tyrosine phosphorylation of Lck is observed when cells are reoxygenated for 75 min and sustained up to 105 min (upper panels in C and D, lanes [5][6][7]. The expression of non-phospho-Lck under these conditions remained the same (middle panels of C and D, lanes [1][2][3][4][5][6][7][8][9]. Actin was used as loading control (lower panels of C and D). All of these bands were quantified by densitometric analysis, and the values of -fold changes are calculated. The results shown here represent three experiments exhibiting similar effects. phosphorylation of Lck in a time-dependent manner, both these cells were induced by hypoxia for 3 h and reoxygenated for 0 -135 min. Cell lysates were immunoprecipitated with rabbit polyclonal anti-Lck antibody, and immunoprecipitated samples were analyzed by Western blot using anti-phosphotyrosine and anti-Lck antibodies, respectively. The results showed that maximum tyrosine phosphorylation of Lck was observed at 75 min and sustained up to 105 min in both MCF-7 (Fig. 1C) and MDA-MB-231 (Fig. 1D) cells (upper panels, lanes 1-9). The expression of non-phospho-Lck remained unchanged (middle panel, lanes [1][2][3][4][5][6][7][8][9]. Actin was used as loading controls (Fig. 1, A-D, lower panels). The bands were quantified by densitometric analysis and normalized with respect to actin. The values of -fold changes are indicated.
To further confirm whether tyrosine 394 of Lck is involved in H/R-induced tyrosine phosphorylation; these cells were transfected with wild type Lck and mutant Lck F394, induced by hypoxia for 3 h, and reoxygenated for 90 min. These cells were lysed and immunoprecipitated with anti-Lck antibody. Half of the immunoprecipitated samples were immunoblotted with anti-phosphotyrosine antibody, and the remaining half of the samples were immunoblotted with anti-Lck antibody. The data revealed that MCF-7 ( Fig. 2A)  anti-Fyn (lane 7) antibody. These results demonstrated that Lck plays significant role in H/R-induced tyrosine phosphorylation of IB␣ in these cells. To delineate the specificity of Lck on phosphorylation of IB␣, both these cells were pretreated with emodin or pp2 (Lck inhibitors) and induced by H/R, and then kinase assay was performed. The data revealed that both emodin and pp2 blocked the H/R-induced Lck kinase activity using IB␣ as substrate in MCF-7 and MDA-MB-231 cells (lanes 4 and 5). To further confirm the specificity of tyrosine phosphorylation of IB␣ by Lck, both these cell lysates were immunoprecipitated with anti-Lck antibody and immunoprecipitated samples were used for kinase assay in presence of TP, which specifically inhibits the tyrosine phosphorylation. The data indicated that TP inhibited the H/R-induced Lck-dependent tyrosine phosphorylation of IB␣ in these cells (lane 8). As expected, no phosphorylation of IB␣ was observed in cells grown under normoxic condition (lane 1) or cells induced by H/R but without immunoprecipitation with anti-Lck antibody (lane 3). Half of the immunoprecipitated samples (using anti-Lck, anti-Lyn, or anti-Fyn antibody) containing equal amount of total proteins were resolved by SDS-PAGE and analyzed by Western blot using anti-Lck, anti-Lyn, or anti-Fyn antibody to ensure the equal amount of loading. These results showed that Lck, Lyn, and Fyn are expressed in these cells and an equal amount of total protein was used for kinase assay (Fig. 2, C and  D, lower panel, lanes 1-8). These data further demonstrated that H/R induces the tyrosine phosphorylation of IB␣ by inducing the Lck but not Lyn or Fyn kinase activity in these cells.
H/R Stimulates Tyrosine but Not Serine Phosphorylation of IB␣-To delineate the role of H/R on tyrosine phosphorylation of IB␣, both MCF-7 and MDA-MB-231 cells were exposed with hypoxia for 0 to 24 h and reoxygenated for 90 min. Cell lysates were immunoprecipitated with rabbit polyclonal anti-IB␣ antibody, and the immunoprecipitated samples were analyzed by Western blot using mouse monoclonal anti-phosphotyrosine antibody. The tyrosine phosphorylation of IB␣ was detected after 1 h of exposure with hypoxia, whereas maximum tyrosine phosphorylation was found at 3 h of exposure with hypoxia followed by reoxygenation in both MCF-7 (Fig. 3A, panel a,  lanes 1-7) and MDA-MB-231 (Fig. 3B, panel a, lanes 1-7) cells. Previous reports indicated that various cytokines and other factors play major roles in serine phosphorylation and degradation of IB␣; therefore, we sought to determine whether H/R has any effect on serine phosphorylation and degradation of IB␣ in breast cancer cells. Accordingly, cells were induced by H/R and cell lysates were immunoprecipitated with anti-IB␣ antibody. The immunoprecipitated samples were immunoblotted with mouse monoclonal anti-phosphoserine IB␣ antibody. Same blots were reprobed with rabbit polyclonal anti-IB␣ antibody. H/R had no effect on serine phosphorylation and degradation of IB␣ in MCF-7 (Fig. 3A, panels b and c,  lanes 1-7) and MDA-MB-231 (Fig. 3B, panels b and c, lanes  1-7) cells. Actin was used as loading controls (panel d, lanes  1-7). The bands were quantified by densitometric analysis and normalized with respect to actin. The values of -fold changes are indicated.
To examine the effect of reoxygenation on IB␣ tyrosine phosphorylation in a time-dependent manner, both MCF-7 and MDA-MB-231 cells were induced by hypoxia for 3 h and reoxygenated for 0 -135 min. Cell lysates containing equal amount of total proteins were subjected to immunoprecipitation with rabbit polyclonal anti-IB␣ antibody. The immunocomplex was resolved by SDS-PAGE and detected by Western blot analysis using mouse monoclonal anti-phosphotyrosine antibody (Fig. 3,  C and D, panels a, lanes 1-9). The data revealed that maximum tyrosine phosphorylation of IB␣ started at 75 min and was sustained up to 90 min upon reoxygenation in case of MCF-7 cells (Fig. 3C, panel a, lanes 5 and 6) and initiated at 60 min and remained up to 135 min in MDA-MB-231 cells (Fig. 3D,  panel a, lanes 4 -9). To check whether H/R had any effect on serine phosphorylation and degradation of IB␣, both these cells were induced by H/R and cell lysates were immunoprecipitated with anti-IB␣ antibody. The immunoprecipitated samples were separated by SDS-PAGE and immunoblotted individually with mouse monoclonal anti-phosphoserine IB␣ and anti-IB␣ antibodies, respectively. These data suggested that H/R had no effect on serine phosphorylation and degradation of IB␣ in both these cells (Fig. 3, C and D, panels b and c,  lanes 1-9). Actin was used as loading controls (Fig. 3, C and D,  panel d, lanes 1-9). To further confirm whether H/R regulates tyrosine phosphorylation but not serine phosphorylation and degradation of IB␣, both these cells were treated with PMA (10 ng/ml) or TNF␣ (0.1 nM), which are known to induce serine phosphorylation and degradation of IB␣. Cell lysates containing equal amount of total proteins were individually analyzed by Western blot using anti-phosphoserine IB␣ antibody. The same blots were reprobed with rabbit polyclonal anti-IB␣ antibody. The data showed that PMA (Fig. 3E) and TNF␣ (Fig.  3F) independently induce serine phosphorylation and degradation of IB␣ in MDA-MB-231 (panels a and b, lanes 1-9) cells. Maximum PMA or TNF␣ induced serine phosphorylation of IB␣ was observed at 15 min, whereas maximum IB␣ degradation was noted at 30 min in these cells. Similar results were obtained in MCF-7 cells (data not shown). Actin was used as loading controls (panel c, lanes 1-9). All of these bands were quantified densitometrically, and the values of -fold changes are calculated. These results clearly suggested that H/R induces tyrosine phosphorylation but not serine phosphorylation and degradation of IB␣ in these cells.
H/R Enhances Lck-dependent Tyrosine Phosphorylation of IB␣-To delineate whether Lck is involved in tyrosine phosphorylation of IB␣ in presence of H/R, both MCF-7 and MDA-MB-231 cells were transiently transfected with wild type or mutant Lck F394 and induced by hypoxia for 3 h followed by reoxygenation for 90 min. Cell lysates were immunoprecipitated with anti-IB␣ antibody. The immunoprecipitated samples were detected by Western blot analysis using anti-phosphotyrosine antibody. The data showed that cells transfected with wild type Lck induced the tyrosine phosphorylation of IB␣ at least 3-fold in MCF-7 cells (Fig. 4A, panel a, lane 3) and at least 2-fold in MDA-MB-231 cells (Fig. 4B, panel a,

lane 3) compared with cells transfected with LipofectAMINE Plus alone (lane 2) or mutant Lck F394-transfected cells (lane 4) in
presence of H/R. As expected, in the absence of H/R, no tyrosine phosphorylation of IB␣ was observed (lane 1). The serine phosphorylation and degradation of IB␣ were also detected by Western blot using anti-phosphoserine-IB␣ and anti-IB␣ antibodies, respectively. The results indicated that transfection of these cells with wild type or mutated Lck followed by treatment with H/R have no effect on serine phosphorylation and degradation of IB␣ (Fig. 4, A and B, panels b and c, lanes 1-4). Actin-specific band was shown as loading controls (panel d ,  lanes 1-4).
To check the effect of Lck inhibitors (emodin and pp2) on H/R-induced tyrosine phosphorylation of IB␣, these cells were individually pretreated with different concentrations of emodin or pp2 and then induced by H/R as described above. Cell lysates were immunoprecipitated with anti-IB␣ antibody and detected by Western blot analysis using mouse monoclonal antiphosphotyrosine antibody. The results revealed that H/R-induced tyrosine phosphorylation of IB␣ (Fig. 4, C and D, panel  a, lane 2) was inhibited when increasing concentrations of emodin (panel a, lanes 3 and 4) or pp2 (panel a, lanes 5 and 6) were used. As expected, no tyrosine-phosphorylated IB␣ specific band was detected in non-H/R-induced cells (lane 1). Both emodin and pp2 had no effects on serine phosphorylation (Fig.  4, C and D, panel b, lanes 1-6) and degradation (panel c, lanes 1-6) of IB␣ in presence of H/R in these cells. Actin was used as loading control (panel d, lanes 1-6). These data clearly demonstrated that Lck plays significant role in H/R-induced tyrosine phosphorylation but not serine phosphorylation of IB␣ in these cells. All of these bands were quantified by densitometric analysis, and the values of -fold changes are indicated.
H/R Induces the Interaction between SH2 Domain of Lck and Tyrosine-phosphorylated IB␣-To delineate the role of H/R or other agents in regulation of direct interaction between SH2 domain of Lck and tyrosine-phosphorylated IB␣, both these cells were individually treated with pV (250 M) for 20 min, H 2 O 2 (5 mM) for 20 min, or induced by hypoxia for 3 h and reoxygenated for 90 min. Cell lysates were immunoprecipitated with rabbit polyclonal anti-Lck antibody, separated by SDS-PAGE, and analyzed by Western blot using anti-IB␣ or anti-phosphotyrosine antibody. The results showed that H/R induces the interaction between tyrosine-phosphorylated IB␣ and Lck in both MCF-7 (Fig. 5A, upper and middle panels, lane 4) and MDA-MB-231 (Fig. 5B, upper and middle panels, lane 4) A and B, lanes 1-7). Serine phosphorylation of IB␣ and non-phospho IB␣ were remained unchanged (panels b and c of A and B, lanes 1-7). Actin was used as loading control (panels d of A and B). C and D, both MCF-7 (C) and MDA-MB-231 (D) cells were exposed with hypoxia for 3 h and reoxygenated for 0 -135 min. Cell lysates were immunoprecipitated with rabbit polyclonal anti-IB␣ antibody. The immunocomplex was individually analyzed by Western blot using mouse monoclonal anti-phosphotyrosine (panels a of C and D, lanes 1-9), mouse monoclonal anti-phosphoserine IB␣ (panels b of C and D, lanes [1][2][3][4][5][6][7][8][9], and rabbit polyclonal anti-IB␣ (panels c of C and D, lanes 1-9) antibodies. Note that H/R induces maximum tyrosine phosphorylation at 75-90 min, but there were no changes in serine phosphorylation and degradation of IB␣ in these cells. Actin was used as loading control (panels d of C and D). E and F, PMA (E) and TNF␣ (F) induce serine phosphorylation and degradation of IB␣. MDA-MB-231 cells were treated with PMA (10 ng/ml) or TNF␣ (0.1 nM) for 0 -135 min. Cell lysates containing equal amount of total proteins were individually analyzed by Western blot using mouse monoclonal anti-phosphoserine IB␣ antibody (panels a of E and F, lanes [1][2][3][4][5][6][7][8][9]. The same blots were reprobed with anti-IB␣ (panels b of E and F, lanes 1-9) and anti-actin (panels c of E and F, lanes 1-9) antibodies, respectively. Note that both PMA and TNF␣ induce serine phosphorylation of IB␣ at 15 min (panels a of E and F, lane 2), whereas maximum degradation was occurred at 30 min (panels b of E and F, lane 3) in these cells. All these bands were quantified densitometrically, and the values of -fold changes are calculated. The results shown here represent three experiments exhibiting similar effects.
Western blot using anti-IB␣ or anti-phosphotyrosine antibody. The data showed that H/R-induced interaction between tyrosinephosphorylated IB␣ and Lck (Fig. 5, C and D, upper and middle panels, lane 2) was drastically reduced by SH2 domain inhibitory peptide in a dose-dependent manner (upper and middle panels, lanes 3 and 4), suggesting that SH2 domain of Lck is involved in this interaction. No IB␣-specific bands were detected in absence of H/R (upper and middle panels, lane 1).
To delineate the effect of genetic (Lck F394) and pharmacological (emodin or pp2) inhibitors of Lck on H/R-induced interaction between SH2 domain of Lck and tyrosine-phosphorylated IB␣, both these cells were either transfected with wild type or mutant Lck F394 and then induced by H/R. Similarly, these cells were pretreated with different concentrations of emodin (0 -10 M) or pp2 (0 -4 nM) and induced by H/R. Cell lysates were immunoprecipitated with anti-Lck antibody and detected by Western blot using anti-IB␣ or anti-phosphotyrosine antibody. The data suggested that wild type Lck but not Lck F394 induces the interaction between Lck and tyrosinephosphorylated IB␣ (Fig. 5, E and F, upper and (Fig. 5, G and H,  upper panels, lanes 2-6). No band was observed in the absence of H/R (lane 1). Actin was used as loading controls (Fig. 5, A-H,  lower panels). All these bands were quantified by densitometric analysis, and the values of -fold changes are indicated.

H/R Stimulates Nuclear Translocation and Transactivation of NFB through Lck-mediated Pathway-
To examine the effect of H/R on NFB nuclear translocation, both these cells were grown on glass slides, induced by hypoxia for 3 h, and reoxygenated for 0 -120 min. Cells were fixed and incubated with rabbit polyclonal anti-NFB p65 antibody. These cells were further incubated with FITC-conjugated anti-rabbit IgG. Similarly, to check the effect of Lck inhibitors (emodin or pp2) on H/R-induced NFB nuclear translocation, both these cells were treated with these inhibitors and then induced by H/R. These cells were incubated with anti-NFB p65 antibody and then with FITC-conjugated anti-rabbit IgG. The data showed that H/R-induced nuclear translocation of p65 started at 60 min and was sustained up to 90 min, whereas it shuttled back into the cytoplasm at 120 min in MCF-7 (Fig. 6, panel A, a-d) and in MDA- MB-231 (panel B, a-d) cells. Both emodin and pp2 dose-dependently suppressed the H/R-induced translocation of p65 into the nucleus (panels A and B, e-h). These results clearly suggested that Lck and tyrosine-phosphorylated IB␣ regulate the H/R induced translocation of NFB into the nucleus in these cells.
We have also checked whether H/R induces the NFB transcriptional activity and whether Lck plays any role in H/Rinduced NFB transactivation in these cells. Accordingly, both these cells were transfected with NFB luciferase reporter construct (pNFB-Luc) in presence of LipofectAMINE Plus and NFB reporter gene assay was performed. In separate experiments, these cells were transiently transfected with wild type or mutant Lck F394 followed by transfection with pNFB-Luc.  A and  B, lanes 1-4), anti-phosphoserine IB␣ (panels b in A and B, lanes 1-4) and anti-IB␣ (panels c in A and B, lanes 1-4) antibodies. Note that maximum tyrosine phosphorylation of IB␣ was observed when both cells were transfected with wild type Lck (panels a in A and B, lane 3). Actin was used as loading control (panels d in A and B). C and D, effect of Lck inhibitors (emodin and pp2) on H/R-induced tyrosine phosphorylation of IB␣. Both MCF-7 (C) and MDA-MB-231 (D) cells were treated with various concentrations of Lck inhibitors (0 -10 M emodin and 0 -4 nM pp2). Cells were exposed with hypoxia for 3 h and reoxygenated for 90 min. Cell lysates were immunoprecipitated with anti-IB␣ antibody and analyzed by Western blot using anti-phosphotyrosine (panels a in C and D, lanes 1-6), anti-phosphoserine IB␣ (panels b in C and D, lanes 1-6), and anti-IB␣ (panels c in C and D, lanes [1][2][3][4][5][6] antibodies. Note that both emodin and pp2 dose-dependently suppressed the H/R-induced tyrosine phosphorylation of IB␣ in these cells (panels a in C and D, lanes 1-6). Actin was used as loading control (panels d in C and D). All of these bands were quantified by densitometric analysis, and the values of -fold changes are indicated. The results shown here represent three experiments exhibiting similar effects.
In other experiments, cells transfected with pNFB-Luc were treated with different doses of Lck inhibitors (emodin or pp2). All these cells were induced by hypoxia for 3 h and reoxygenated for 24 h. Cells were harvested in passive lysis buffer and used to measure the luciferase activity. The data demonstrated that wild type Lck enhanced the NFB transcriptional activity compared with cells transfected with LipofectAMINE Plus alone or Lck F394-transfected cells in presence of H/R (Fig. 6,  C and D). Both emodin and pp2 inhibited the H/R-induced NFB transactivation in a dose-dependent manner (Fig. 6, C  and D). The values were normalized to Renilla luciferase activity. The -fold changes were calculated, and the results are expressed as the means Ϯ S.E. of three determination. The values were also analyzed by Student's t test (p Ͻ 0.001).
H/R Induces Lck-dependent NFB-DNA Binding and uPA Promoter Activity-In previous figures (Fig. 6), we have demonstrated that H/R induces Lck-dependent NFB nuclear translocation and transactivation. Therefore we sought to determine the role of Lck on H/R-induced NFB-DNA binding. Accordingly, both MCF-7 (panel A) and MDA-MB-231 (panel B) cells were transiently transfected with wild type Lck or mutant Lck F394 plasmid or pretreated with Lck inhibitors (emodin or pp2) and then induced with hypoxia for 3 h and reoxygenation for 90 min. The nuclear extracts were prepared and analyzed by EMSA. The results demonstrated that H/R induces NFB-DNA binding (Fig. 7, panels A and B, lane 2) 5 and 6). These results clearly demonstrated that H/R induces Lck-dependent NFB-DNA binding in these cells. To further confirm that the band obtained (panels A and B) by EMSA in H/R-induced cells  A and B, lanes 1-4) and anti-phosphotyrosine (middle panels in A and B, lanes 1-4) antibodies. Note that pV,  G and  H, lanes 1-6). Actin was used as loading controls (lower panels of A-H). All these bands were normalized with actin, and -fold changes were calculated. The results shown here represent three experiments exhibiting similar effects. is indeed NFB, the nuclear extracts from MDA-MB-231 cells were incubated with anti-p65 antibody and analyzed by EMSA. Fig. 7C showed the shift of NFB-specific band to higher molecular weight when the nuclear extracts were treated with anti-p65 antibody (lanes 1 and 2).
Because NFB-responsive element is present in the promoter region of uPA, we sought to determine the effect of H/R on Lck-dependent NFB-mediated uPA promoter activity by using full-length uPA promoter-driven luciferase reporter gene assay. Both MCF-7 and MDA-MB-231 cells were transiently transfected with full-length uPA promoter plasmid pGL2-Basic. The transfection efficiency was normalized by cotransfecting the cells with pRL vector (Promega) containing a full-length Renilla luciferase gene under the control of a constitutive promoter. In separate experiments, these transfected cells were individually co-transfected with wild type Lck or Lck F394 or treated with increasing concentrations of Lck inhibitors (emodin and pp2). Cells were then induced with H/R for 24 h and harvested in passive lysis buffer (Promega). The luciferase activity was measured. Results demonstrated that H/R induces uPA promoter activity compared with cells grown under normoxic condition (Fig. 7, panels D and E). uPA promoter activity was much higher in wild type Lck-transfected cells compared with cells transfected with Lck F394 or induced with H/R alone. Both emodin and pp2 dose-dependently suppressed H/R induced uPA promoter activity (panels D and E). These results clearly demonstrated that Lck regulates NFB-dependent uPA promoter activity upon H/R induction.
To further confirm whether H/R-induced uPA promoter activity is NFB-dependent, both these cells were transfected with full-length uPA promoter-driven luciferase reporter plasmid pGL2-Basic and then treated with NFB inhibitory pep- tides (SN-50 or SN-50M). The cells were induced with H/R for 24 h, and luciferase activity was measured. The results shown that NFB inhibitory peptide SN-50 inhibited H/R-induced uPA promoter activity in both these cells. These data suggested that NFB-responsive elements but not regulatory elements are essential for H/R-induced Lck-dependent uPA promoter activity in these cells.
H/R Stimulates Lck-dependent NFB-mediated uPA Secretion-Because it is documented that NFB-responsive element is present in the promoter region of uPA and we have shown in Fig. 7 that H/R stimulates NFB-dependent uPA promoter activity, we sought to determine the impact of H/R on NFBmediated uPA secretion in MCF-7 and MDA-MB-231 cells. Accordingly, these cells were subjected to hypoxia for 3 h and reoxygenation for 0 -16 h. Cell lysates containing equal amount of total proteins were resolved by SDS-PAGE and detected by Western blot analysis using mouse monoclonal anti-uPA anti-body. The data indicated that there were at least 7.2-and 7.8-fold increases in uPA secretion in MCF-7 (Fig. 8A, upper  panel, lanes 1-9) and MDA-MB-231 (Fig. 8B, upper panel, lanes  1-9) cells, respectively, when induced by H/R for 16 h compared with control. However, the rate of induction of overall uPA secretion was significantly higher in MDA-MB-231 cells compared with MCF-7 cells.
To examine the role of Lck on H/R-induced uPA expression, these cells were individually pretreated with different concentrations of emodin (0 -16 M) or pp2 (0 -4 nM). Both these cells were then induced by H/R, and the level of uPA in the cell lysates was detected by Western blot analysis using anti-uPA antibody. The data indicated that both emodin and pp2 dosedependently inhibited H/R-induced uPA secretion in MCF-7 (Fig. 8C, upper panels, lanes 1-8) and MDA-MB-231 (Fig. 8D,  upper panels, lanes 1-8) 1-4).
We have also investigated the effect of NFB inhibitory peptide (SN-50) on H/R-induced uPA expression in these cells. Both these cells were either treated with SN-50 or SN-50M, induced by H/R, and uPA level was detected by Western blot analysis. The data indicated that SN-50 but not control peptide SN-50M suppressed the H/R-induced uPA secretion in MCF-7 (Fig. 8G, upper panel, lanes 1-4) and MDA-MB-231 (Fig. 8H,  upper panel, lanes 1-4) cells. All these blots were reprobed with anti-actin antibody as loading controls (Fig. 8, A-H (Fig. 9, C and D). These data are analyzed statistically using Student's t test and were statistically significant (p Ͻ 0.002). These data demonstrated that H/R induces the Lck-mediated NFB activation through tyrosine phosphorylation of IB␣ and subsequently stimulates uPA secretion, and all of these ultimately control the motility of breast cancer cells (Fig. 10). DISCUSSION In this report, we present evidence for the involvement of protein-tyrosine kinase, p56 lck in the redox-regulated activation of NFB through tyrosine phosphorylation of IB␣, uPA secretion and cell motility following H/R in highly invasive (MDA-MB-231) and low invasive (MCF-7) breast cancer cells. We demonstrated that H/R induces tyrosine phosphorylation of p56 lck and nuclear translocation of NFB, NFB-DNA binding, and transactivation of NF-B through tyrosine phosphorylation of IB␣ in these cells. Transient transfection of these cells with wild type Lck but not with mutant Lck F394 followed by H/R induces the tyrosine phosphorylation of IB␣ and transcriptional activation of NFB, and these processes are inhibited by Lck inhibitors (emodin and pp2). In vitro kinase assay indicated that immunoprecipitated p56 lck but not Lyn or Fyn directly phosphorylates IB␣ in presence of H/R, and this activity is blocked by Lck inhibitors. Pervanadate, H 2 O 2 , and H/R induce the interaction between Lck and tyrosine-phosphorylated IB␣, and this interaction is inhibited by SH2 domain inhibitory peptide indicating that SH2 domain of Lck is involved in this interaction. H/R stimulates NFB-dependent uPA promoter activity and cell motility and subsequently induces uPA expression in these cells. These data demonstrated that H/R induces cell motility and tyrosine kinase, p56 lck -dependent NFB activation, and uPA expression through tyrosine phosphorylation of IB␣ in MCF-7 and MDA-MB-231 cells.
Previous results suggested that Lck acts as a proto-oncogene. The overexpression of wild type p56 lck reproducibly developed thymic tumors, indicating that p56 lck contribute to the pathogenesis of human neoplastic diseases (45). Earlier data demonstrated that H 2 O 2 induced p56 lck catalytic activity by phosphorylating Tyr-394 and Tyr-505. Thus dephosphorylation of Tyr-505 is not a prerequisite for either phosphorylation of Lck at Tyr-394 or catalytic activation of kinase. These data suggested that activation of Lck by phosphorylation of Tyr-394 is dominant over any inhibition induced by phosphorylation of Tyr-505 (30). However, it is unclear whether phosphorylation of Lck induced by H/R treatment is the result of activation of kinase or inhibition of a phosphatase or by both. It has been also indicated that treatment of cells with pervanadate or diamide induces Lck kinase activity and phosphoryl- ation at Tyr-394 (28). It is possible that these agents may act through similar mechanisms. Our data reveals that reactive oxygen species generated by hypoxia/reoxygenation induces phosphorylation of p56 lck at Tyr-394 and its catalytic activity.
pV induces multiple signaling pathways through Lck, Fyn, and Zap-70 in Jurkat cells (46). We have recently reported that pV induces NFB activation through tyrosine phosphorylation of IB␣ in breast cancer cells (44). However, pV had no effect on tyrosine phosphorylation of IB␣ in Lck-deficient Jurkat variants, indicating that IB␣ could be phosphorylated by Lck. pV induces tyrosine phosphorylation of IB␣ without degradation (37). There are evidences that tyrosine-phosphorylated IB␣ at position 42 binds to SH2 domain-containing proteins like phosphatidylinositol 3Ј-kinase and c-Src upon treatment with pV (47). Recent report also demonstrated that H 2 O 2 , pV, or H/R induce tyrosine phosphorylation of IB␣ and NFB activation directly through c-Src-dependent pathways (40). However, the study also demonstrated that c-Src, Fyn, and Yes triple knockout cells able to tyrosine-phosphorylate IB␣ upon pV treatment. This data clearly suggest that there are other tyrosine kinase(s), which are involved in this process. In our study we have demonstrated that H/R induces Lck tyrosine phosphorylation, which leads to induction of Lck kinase activity and subsequently phosphorylates tyrosine residue of IB␣ without degradation and then activates NFB in breast cancer cells.
The activation of transcription factor NFB is regulated by variety of stimuli in different cell types (31,32). The NFB/Rel system has the capacity of specifically responding to various stimuli by inducing the tyrosine phosphorylation of IB␣. Because other IB family proteins such as IB␤ and IB␥ lack tyrosine residue at position 42, involvement of these other proteins in tyrosine phosphorylation and NFB activation is out of the question. Our finding suggests that oxidative stress (H/R)-induced NFB activation does not require degradation of IB␣. Although IB␣ is not degraded, it is most likely that it may modify during oxidative stress through another unknown mechanism, which is prerequisite for NFB dissociation and nuclear translocation. Although pV, a tyrosine phosphatase inhibitor, is known to activate tyrosine phosphorylation of IB␣, the physiological significance of this phenomenon is remained unclear. H/R also induces IB␣ tyrosine phosphorylation and NFB nuclear translocation during tumor progression. Our studies have further demonstrated an IKKindependent pathway that regulates NFB through p56 lck activation. Lck activated by H/R is able to phosphorylate IB␣ in vitro, and further induces NFB nuclear translocation and transcriptional activity. These findings clearly suggest that NFB activated by H/R is functionally active. In the present report, we have delineated the involvement of IB␣/p56 lck (proteintyrosine kinase) pathway in mediating NFB transcriptional activation in breast cancer cells following hypoxia/reoxygenation. uPA is a serine protease that plays a major role in tumor invasion, malignant progression, and distant metastasis by converting plasminogen into plasmin. There is abundant experimental evidence that uPA plays a significant role in malignant progression and tumor metastasis (1). The secretion of uPA is the prerequisite for proteolytic degradation of extracellular matrix, extracellular matrix invasion, and cell migration. An up-regulation of uPA and uPAR has been described in many human tumors. High levels of uPA and uPAR in tumor tissues are associated with poor prognosis of patients with cancer of the breast, lung, head and neck, uterine cervix, bladder, and ovary (2). It has been demonstrated that uPA is a downstream target molecule of NFB, because NFB-responsive element is present in uPA. Therefore, we sought to determine the level of cellular uPA after induction of breast cancer cells by H/R. We have demonstrated that H/R induced NFB activity and uPA secretion through tyrosine phosphorylation of IB␣ in these cells. The induction of uPA secretion by H/R is comparatively higher in MDA-MB-231 cells than in MCF-7 cells.
In summary, we have demonstrated for the first time that H/R induces the tyrosine phosphorylation of p56 lck and its kinase activity. Lck is able to directly tyrosine-phosphorylate IB␣, and this phosphorylation event is required for NFB activation and uPA secretion following H/R in MCF-7 and MDA-MB-231 cells. H/R also enhances the interaction between SH2 domain of Lck and IB␣ through inducing tyrosine phosphorylation of IB␣, and this interaction is blocked by SH2 domain inhibitory peptide in these cells. Transfection of these cells with wild type Lck but not with Lck F394 stimulates the tyrosine phosphorylation of IB␣ and NFB transactivation in the presence of H/R, indicating that tyrosine 394 is involved in these processes. H/R induces uPA secretion and cell migration. Finally, these data demonstrated that H/R regulates p56 lckmediated NFB activation, NFB-dependent uPA promoter activity, and cell motility through expression of uPA in breast cancer cells (Fig. 10). These findings may be useful in designing novel therapeutic interventions that block redox-regulated p56 lck -dependent NFB activation, resulting in reduction of uPA secretion and consequent blocking of cell motility, invasiveness, and metastatic spread of breast cancer. five tandem repeats of NFB binding site. The full-length human uPA promoter (Ϫ2062 to ϩ27) in luciferase reporter gene plasmid pGL2 basic was the gift of Dr. Ute Reuning.