PMC:2246177 / 12259-30606 JSONTXT

{"target":"https://pubannotation.org/docs/sourcedb/PMC/sourceid/2246177","sourcedb":"PMC","sourceid":"2246177","source_url":"https://www.ncbi.nlm.nih.gov/pmc/2246177","text":"Results\n\nSTAT tyrosine phosphorylation mediated by breast tumor kinase\nPrevious studies have demonstrated that STAT5a and STAT5b are tyrosine phosphorylated by the EGFR and c-Src kinases, two kinases that are involved in breast cancer [12,26,27]. Given the recent evidence that Brk phosphorylates STAT3 [1], we investigated whether this kinase could also mediate the phosphorylation of the related STAT5a and STAT5b proteins. Mouse embryo fibroblasts from STAT5a/b knockout mice (MEF5-/-) were transfected with STAT3, STAT5a, or STAT5b along with various Brk expression vectors. MEF5-/- cells were chosen because these cells do not express endogenous Brk (Figure 1), STAT5a, or STAT5b. Immunoprecipitations were performed for the transfected STAT using specific antibodies and were analyzed for tyrosine phosphorylation by immunoblotting. Additionally, total lysates were analyzed for the expression of the Brk constructs.\nFigure 1 Breast tumor kinase mediates tyrosine phosphorylation of signal transducers and activators of transcription. Breast tumor kinase (Brk) mediates tyrosine phosphorylation of signal transducer and activator of transcription (STAT)3, STAT5a, and STAT5b. STAT5a/b knockout mouse embryo fibroblast (MEF5-/-) cells were transfected with (a) STAT3, (b) STAT5a, or (c) STAT5b along with either pRc (vector), Brk, K219M Brk (kinase inactive), or Y477F Brk (constitutively active). Immunoprecipitations (IPs) for STAT3, STAT5a, or STAT5b were performed using antibodies specifically directed against each STAT and were analyzed by immunoblotting with antibodies directed toward STAT3, STAT5a, STAT5b (middle), or phosphotyrosine (top). Total lysates were analyzed for the expression of the Brk construct by immunoblotting with anti-Brk antibody (bottom). In agreement with previous findings, Figure 1a demonstrates that wildtype Brk and constitutively active Brk (Y447F) were able to mediate STAT3 tyrosine phosphorylation, while the kinase inactive Brk (K219M) or vector (pRc) demonstrated no detectable tyrosine phosphorylation [1]. Importantly, Brk and the Y447F Brk, but not the K219M Brk, were also able to mediate STAT5a and STAT5b tyrosine phosphorylation (Figure 1b,c). These data demonstrate for the first time that Brk mediates tyrosine phosphorylation of STAT5a and STAT5b in addition to STAT3. Since our previous studies demonstrated that STAT5b, but not STAT5a, elicits a proproliferative effect in breast cancer cells [12], and other studies have shown that Brk also increases proliferation of breast cancer cells [22], we focused our efforts on Brk-mediated activation of STAT5b.\n\nPhosphorylation of STAT5b by breast tumor kinase\nWhile phosphorylation of Y699 is required for transcriptional activity, three tyrosines located in the transactivation domain can also be phosphorylated and modulate STAT5b activity [23]. These tyrosines are either positive (Y725) or negative (Y740 and Y743) regulators of STAT5b activity, and their phosphorylations are mediated by c-Src (Y699 and Y725) and the EGFR (Y699, Y725, Y740, and Y743) [12,27]. We investigated the ability of Brk to phosphorylate each of the four identified tyrosine sites, starting with the required activating tyrosine, Y699. MEF5-/- cells were transfected with STAT5b along with Brk constructs, and STAT5b immunoprecipitations were analyzed using the antiphospho-Y699 STAT5b-specific antibody.\nBoth Brk and Y447F Brk were able to mediate the phosphorylation of Y699, while it was not detectable in vector cells or K219M Brk-transfected cells (Figure 2a). To determine whether Brk mediates the phosphorylation of tyrosines other than Y699, MEF5-/- cells were transfected with Brk along with STAT5b mutants in which all but one of the four tyrosines was mutated. Brk induced detectable tyrosine phosphorylation of wildtype STAT5b and the Y725/740/743F mutant (with intact Y699), confirming the results with the antiphospho-Y699-specific antibody. None of the other STAT5b mutant constructs, however, were phosphorylated by Brk (Figure 2b). These results demonstrate that Brk primarily mediates Y699 phosphorylation on STAT5b and not the three transactivation domain tyrosines.\nFigure 2 Signal transducer and activator of transcription 5b tyrosine phosphorylation and transcriptional activation. (a) Mouse embryo fibroblast (MEF5-/-) cells were transfected with signal transducer and activator of transcription 5b (STAT5b) along with pRc (vector), breast tumor kinase (Brk), K219M, or Y447F. Immunoprecipitations (IPs) were performed using antibodies specifically against STAT5b and were analyzed by immunoblotting with antibodies directed toward phospho-Y699 STAT5b (top) or STAT5b (bottom). (b) MEF5-/- cells were transfected with various STAT5b constructs along with constitutively active Y447F Brk. Immunoprecipitations (IPs) were performed as above and were analyzed via immunoblotting with antiphosphotyrosine antibody (top) or anti-STAT5b antibody (bottom). (c) MEF5-/- were transfected with the STAT5-specific Spi2.1 promoter luciferase construct and either His-vector, His-STAT5b, or His-Y699F with or without Brk. Luciferase activity was measured and normalized to total protein. Values from four independent experiments performed in triplicate reported as the fold induction over His, pRc ± standard error: His, pRc (1.00 ± 0.00) and Brk (1.05 ± 0.39); STAT5b, pRc (28.89 ± 5.51) and Brk (75.48 ± 25.79); Y699F, pRc (1.17 ± 0.14) and Brk (0.93 ± 0.27). Student's t test was used to determine statistical significance between STAT5b pRc and STAT5b Brk. *P = 0.0123.\n\nBreast tumor kinase mediates STAT5b transcriptional activity\nSince Y699 phosphorylation is required for the transcriptional function of STAT5b, we analyzed the ability of Brk to mediate STAT5b transcriptional activity. MEF5-/- cells were transfected with the STAT5-specific luciferase reporter plasmid (Spi2.1-luc) along with His-vector, STAT5b, or Y699F plus or minus Brk. Since the MEF5-/- cells express endogenous STAT1 and STAT3, the specificity of this reporter was demonstrated by the inability of the His-vector to mediate luciferase activity (Figure 2c). The expression of STAT5b increased Spi2.1-mediated luciferase activity, and the cotransfection of Brk significantly increased this transcriptional activity further. As seen previously, the Y699F STAT5b mutant was unable to mediate STAT5b transcriptional activity, and the addition of Brk had no effect. These results demonstrate not only that Brk mediates STAT5b transcriptional activity, but that Y699 is required.\n\nBreast tumor kinase directly phosphorylates Y699 of STAT5b in vitro\nTo determine whether Brk could mediate the phosphorylation and activation of STAT5b directly, or whether other tyrosine kinases known to activate STAT5b (such as EGFR and c-Src) were required, an in vitro kinase assay was performed. Purified recombinant Brk was incubated with purified recombinant GST, GST-STAT5b, or GST-Y699F. The in vitro kinase reaction was analyzed via silver stain (Figure 3a) and by immunoblotting with the antiphospho-Y699 STAT5b-specific antibody (Figure 3b). The silver stain illustrates that GST, GST-STAT5b, GST-Y699F, and Brk were all expressed (Figure 3a). Furthermore, the specificity of the phospho-Y699 STAT5b antibody was confirmed by its inability to recognize the recombinant GST-Y699F protein (Figure 3b). Phospho-Y699 STAT5b antibody, however, detected phosphorylated Brk in this assay. Immunoblotting with this antibody demonstrated that the GST-STAT5b was phosphorylated on Y699 in the presence of, but not in the absence of, Brk. This result demonstrates that Brk is able to directly phosphorylate Y699 on STAT5b in vitro.\nFigure 3 In vitro Y699 phosphorylation by breast tumor kinase. In vitro Y699 phosphorylation of signal transducer and activator of transcription 5b (STAT5b) by breast tumor kinase (Brk). Purified recombinant GST, GST-STAT5b, or GST-Y699F were incubated with (+) or without (-) purified recombinant Brk in reaction buffer containing 20 nM ATP at 30°C for 30 minutes. An equal volume of 2 × Laemmli buffer was added to end the reaction and the samples were run on SDS-PAGE. Gels were analyzed via (a) silver stain and (b) by immunoblotting with an antiphospho-Y699 STAT5b-specific antibody.\n\nBreast tumor kinase expression in human breast cancer cell lines\nWhile studies in the MEF5-/- cells allowed us to characterize the role of Brk in STAT5b activation, we continued our studies in breast cancer cell lines as a means to investigate the biological relevance of Brk-mediated STAT5b activation. We determined the relative expression of Brk by analyzing mRNA and protein levels in a panel of frequently studied breast cancer cell lines. The total mRNA was isolated from each cell line, cDNA was generated, and primers specific for either Brk or β-actin were used for amplification [25]. In each sample, the relative expression of Brk mRNA was determined by normalizing to the amount of β-actin. The expression of Brk mRNA in the MDA-MB-468 cell line was set to 1 (Figure 4a) because this cell line expressed the lowest detectable amount of Brk mRNA. While BT-20 cells and T47D cells had the highest mRNA levels, SKBr3 cells and MDA-MB-468 cells expressed moderate levels, and BT-549 cells had no detectable Brk mRNA.\nFigure 4 Breast tumor kinase expression in breast cancer cell lines. (a) Total RNA was isolated from the breast cancer cell lines using the RNeasy kit (Qiagen), and the cDNA was generated using iScript cDNA synthesis (Bio-Rad). The cDNA was amplified using primers specifically for breast tumor kinase (Brk) or β-actin. The relative level of Brk mRNA was determined using densitometric analysis and was normalized to the amount of β-actin. The level of Brk mRNA in the MDA-MB-468 cell line was set to 1; numbers indicate the fold increase compared with the 468 cells. (b) Whole-cell lysates were prepared from the breast cancer cell lines. Lysates were immunoblotted with anti-Brk antibody (top) and with anti-Ran antibody (bottom). The relative expression of Brk protein was determined using densitometric analysis and was normalized to the amount of Ran. The expression of Brk protein in the MDA-MB-468 cell line was set to 1; numbers indicate the fold increase compared with the 468 cells. The amount of Brk protein in detergent cell lysates from each cell line was determined by immunoblotting. The relative expression of Brk protein was determined by normalizing to the nuclear import protein Ran, shown previously as a valid normalization control in breast cancer cell lines [28]. Again the amount of Brk protein in the MDA-MB-468 cell line was set to 1 (Figure 4b) because this cell line had the lowest detectable amount of Brk protein. Brk was highly expressed in BT-20 cells and T47D cells, was moderately expressed in SKBr3 cells and MDA-MB-468 cells, and was not detectable in BT-549 cells. Relatively high levels of Brk expression were seen in both the estrogen-receptor-positive T47D cells and the estrogen-receptor-negative BT-20 cells. Furthermore, while MDA-MB-468 cells, BT-20 cells, and BT-549 cells all express the EGFR, they expressed different amounts of Brk.\n\nBreast tumor kinase mediates STAT5b activity in breast cancer cells\nTo examine the ability of Brk to mediate the transcriptional activity of endogenous STAT5b in a breast cancer cell line, the BT-549 cells, which have no detectable endogenous Brk, were utilized. The STAT5-specific Spi2.1 luciferase reporter with or without Brk was transfected into the BT-549 cells. While there was a detectable basal level of transcriptional activity, Brk significantly increased endogenous STAT5b transcriptional activity, as did the constitutively active Brk (Y447F) (Figure 5a). Although not nearly as effective, the kinase inactive Brk (K219M) did significantly increase the STAT5b luciferase activity compared with vector alone. This result is not necessarily unexpected since a role of the kinase inactive Brk in other cell models has been reported (see Discussion).\nFigure 5 Breast tumor kinase mediates endogenous signal transducer and activator of transcription 5b transcriptional activity. (a) BT-549 cells were transfected with the Spi2.1 promoter luciferase construct along with pRc, breast tumor kinase (Brk), K219M, or Y447F expression vectors. Luciferase activity was measured and normalized to total protein. Values from four independent experiments carried out in triplicate were reported as average luciferase/protein ± standard error: pRc (1,859.37 ± 592.24); Brk (10,031.94 ± 1,359.68); K219M (3,109.82 ± 777.57), Y477F (12,878.57 ± 2,464.26). Student's t test was used to determine statistical significance between pRc and Brk, and between pRc and K219M. *P \u003c 0.0001, • P = 0.0430. (b) BT-549 cells were transfected with the Spi2.1 promoter luciferase construct along with pRc, Brk, or K219M, and either pcDNA, c-Src, or K-Src expression vectors. Luciferase activity was measured and normalized to total protein. Values from four independent experiments carried out in triplicate were reported as the average fold induction over pcDNA, pRc ± SE: pcDNA, pRc (1.00 ± 0.00), Brk (2.92 ± 0.39), and K219M (1.36 ± 0.13); c-Src, pRc (1.10 ± 0.12), Brk (5.01 ± 0.38), and K219M (1.24 ± 0.18); K-Src, pRc (0.99 ± 0.19) and Brk (2.76 ± 0.89). Student's t test was used to determine statistical significance between pcDNA pRc and pcDNA Brk, between pcDNA pRc and pcDNA K219M, and between pRc Brk and c-Src Brk. *P = 0.0027, • P = 0.0322, ◆ P \u003c 0.0086. Since Brk and c-Src are both able to phosphorylate STAT5b on Y699, the potential cooperation between Brk and c-Src was investigated. As previously, Brk dramatically enhanced STAT5b transcriptional activity, and the K219M Brk was also able to significantly enhance this activity (Figure 5b). Unexpectedly, exogenous overexpression of c-Src did not enhance STAT5b transcriptional activity (Figure 5b). c-Src overexpression enhanced Brk-induced STAT5b transcriptional activity, however, indicating that Brk and c-Src together can enhance STAT5b transcriptional activity. The kinase activity of Brk was necessary for this collaboration as the addition of c-Src to K219M Brk did not enhance STAT5b activity. Likewise, the kinase inactive c-Src (K-Src) failed to significantly enhance Brk-induced STAT5b transcriptional activity, suggesting that the kinase activity of c-Src is also necessary for this cooperation between Brk and c-Src. While Brk and c-Src individually have different effects on STAT5b transcriptional activity, these two tyrosine kinases together increase STAT5b transcriptional activity. These data suggest that Brk and c-Src do not merely substitute for one another in activating STAT5b, but rather that Brk can function independently as well as together with c-Src.\n\nSTAT5b and breast tumor kinase involvement in DNA synthesis\nTo determine the biological significance of the Brk-mediated increase in STAT5b activity, the role of Brk and STAT5b, individually and together, on DNA synthesis in breast cancer cell lines was investigated. Since Brk signals downstream of the EGFR [20,21,30,30] and HER2 [29] tyrosine kinases, the EGFR/HER2-overexpressing SKBr3 breast cancer cell line and the EGFR-overexpressing BT-20 breast cancer cell line were chosen for these studies. Both cell lines express the c-Src and Brk tyrosine kinases as well as STAT5b. To investigate the effect on proliferation, siRNA technology was used to knockdown Brk or STAT5b protein levels, individually or together, and then the effect on DNA synthesis was measured by BrdU incorporation. Lysates were analyzed via immunoblotting to determine significant and specific knockdown of Brk and STAT5b.\nAs shown in Figure 6a,c, knockdown of Brk did not effect the expression of STAT5b and knockdown of STAT5b did not effect the expression of Brk. Furthermore, knockdown of Brk and/or STAT5b did not effect the expression of HER2, EGFR, c-Src, STAT3, or STAT1 (data not shown). Knockdown of Brk (siBrk) inhibited DNA synthesis in both SKBr3 cells and BT-20 cells by 47% and 31%, respectively (Figure 6b,d). Knockdown of STAT5b (siSTAT5b) significantly decreased DNA synthesis of both cell lines to an extent similar to that seen with Brk knockdown. Importantly, the knockdown of Brk and STAT5b together did not further decrease DNA synthesis compared with either single knockdown, suggesting that signaling mediated by Brk and STAT5b are contributing to the same proproliferative pathway.\nFigure 6 Breast tumor kinase and signal transducer and activator of transcription 5b roles in DNA synthesis. Knockdown of breast tumor kinase (Brk) and/or signal transducer and activator of transcription 5b (STAT5b) in SKBr3 cells (a, b) or BT-20 cells (c, d) was performed using the Dharmacon siGenome SMARTpool duplex for each protein as per the manufacturer's instructions. (a, c) Seventy-two hours after transfection of the siRNA, cells were lysed. Total lysates were analyzed via immunoblotting with anti-STAT5b (top), anti-Brk (middle), or anti-β-actin (bottom) antibodies. (b, d) Sixty-six hours after transfection of the siRNA, bromodeoxyuridine (BrdU) was added to the medium for 6 hours. The cells were fixed and stained with a fluorescent antibody against BrdU. Cells were scored for BrdU incorporation and graphed. Between 160 and 200 cells were counted for each treatment group. (b) Average percentage of BrdU incorporation ± standard error from three independent experiments performed in triplicate for the SKBr3 cells: media (21.54 ± 0.16), siLuc (24.98 ± 2.78), siBrk (12.37 ± 1.65), siSTAT5b (12.37 ± 0.51), siBrk/siSTAT5b (12.39 ± 0.21). Student's t test was utilized to determine statistical significance between media and siBrk, siSTAT5b, or siBrk/siSTAT5b, and between siLuc and siBrk, siSTAT5b, or siBrk/siSTAT5b. *P ≤ 0.0050, • P \u003c 0.0025. (d) Average percentage of BrdU incorporation ± standard error from three independent experiments performed in triplicate for the BT-20 cells: media (25.28 ± 0.55), siLuc (26.87 ± 0.52), siBrk (18.97 ± 0.32), siSTAT5b (18.87 ± 0.03), siBrk/siSTAT5b (17.37 ± 0.88). Student's t test was utilized to determine statistical significance between media and siBrk, siSTAT5b, or siBrk/siSTAT5b, and between siLuc and siBrk, siSTAT5b, or siBrk/siSTAT5b. *P ≤ 0.0016, • P \u003c 0.0007.\n\nDi","divisions":[{"label":"Title","span":{"begin":0,"end":7}},{"label":"Section","span":{"begin":9,"end":2618}},{"label":"Title","span":{"begin":9,"end":70}},{"label":"Figure caption","span":{"begin":923,"end":1778}},{"label":"Section","span":{"begin":2620,"end":5576}},{"label":"Title","span":{"begin":2620,"end":2668}},{"label":"Figure caption","span":{"begin":4175,"end":5575}},{"label":"Section","span":{"begin":5575,"end":6553}},{"label":"Title","span":{"begin":5575,"end":5635}},{"label":"Section","span":{"begin":6555,"end":8280}},{"label":"Title","span":{"begin":6555,"end":6622}},{"label":"Figure caption","span":{"begin":7688,"end":8279}},{"label":"Section","span":{"begin":8279,"end":11187}},{"label":"Title","span":{"begin":8279,"end":8343}},{"label":"Figure caption","span":{"begin":9304,"end":10299}},{"label":"Section","span":{"begin":11189,"end":14821}},{"label":"Title","span":{"begin":11189,"end":11256}},{"label":"Figure caption","span":{"begin":12048,"end":13540}},{"label":"Section","span":{"begin":14823,"end":18346}},{"label":"Title","span":{"begin":14823,"end":14882}},{"label":"Figure caption","span":{"begin":16509,"end":18345}}],"tracks":[]}