PTPMeg2 Interacts With STAT3 in Mammalian Cells
To search for negative regulators of STAT3, we examined the possibility of its interaction with different phosphatases, and PTPMeg2 was identified as a potential interacting protein. To confirm the interaction, Myc-PTPMeg2 and Flag-STAT3 were co-expressed in HEK 293T cells and co-immunoprecipitation and GST pull-down experiments were performed. The results showed that PTPMeg2 interacts with STAT3 in vitro (Figure 1A and 1B). Interestingly, we observed that PTPMeg2 preferentially interacted with STAT3 as it had either a weak or no interaction with STAT5 or STAT1 (Figure 1C). An in vivo interaction of endogenous PTPMeg2 and STAT3 proteins was observed in the mouse brain tissue (Figure 1D, left panel) and breast cancer MCF7 cells (Figure 1D, right panel). All these results suggested that PTPMeg2 interacts with STAT3 under physiological and pathological conditions.
PTPMeg2 interacts with STAT3. (A) PTPMeg2 was co-immunoprecipitated with STAT3 in mammalian cells. Lysates prepared from HEK293T cells expressing Flag-STAT3 and Myc-PTPMeg2 were precipitated with an anti-Flag antibody and the precipitants were blotted by an anti-Myc antibody. (B) STAT3 interacts with PTPMeg2 physically in vitro. GST or GST-STAT3 fusion proteins were mixed with the Myc-PTPMeg2 prepared from cells (C) PTPMeg2 preferentially interacts with STAT3. Myc-PTPMeg2 was co-expressed with Flag-tagged STAT3 (S3), STAT1 (S1) or STAT5a (S5a) in 293T cells. (D) PTPMeg2 interacts with STAT3 in vivo. Lysates from the mouse brain tissue and breast cancer MCF7 cells were used in co-immunoprecipitation assays to demonstrate the endogenous protein interaction with an anti-STAT3 antibody (C20) and anti-PTPMeg2 rabbit polyclonal antibody. (E) PTPMeg2 interacts with phosphorylated STAT3. A reciprocal immunoprecipitation assay was performed with an anti-Myc antibody or an anti-Flag antibody for the lysates of HEK293T cells under IL-6 stimulation for 30 min. (F) IL-6 induces the interaction of STAT3 and PTPMeg2 in vivo. An immunoprecipitation assay was performed using the endogenous proteins in MCF7 cells stimulated by IL-6 for 30 min. (G) PTPMeg2 decreases the accumulation of pSTAT3 in the nucleus. MCF7 cells transfected with STAT3 and/or PTPMeg2 WT/CS were treated without or with IL-6 for 30 min. Cells were immunostained with an anti-PTPMeg2 (FITC) and an anti-Flag antibody (TRITC, for Flag-STAT3). DAPI was used for nuclear staining. Scale, 10 μm.
PTPMeg2 Interacts With Both Phosphorylated and Unphosphorylated Forms of STAT3
To determine if the interaction of PTPMeg2 with STAT3 is regulated by cytokines, HEK 293T cells transfected with Flag-STAT3 and Myc-PTPMeg2 were stimulated by IL-6 for 30 min. A reciprocal immunoprecipitation experiment indicated that the interaction of PTPMeg2 and STAT3 was increased dramatically under stimulation of IL-6 (Figure 1E). Interestingly, we observed a strong band of phosphorylated STAT3 in a complex precipitated with an anti-Myc antibody (for Myc-PTPMEG2) (Figure 1E). Consistently we observed that IL-6 induced the interaction of endogenous STAT3 and PTPMeg2 in MCF7 cells (Figure 1F). These results suggest that PTPMeg2 interacts with the phosphorylated form of STAT3 (pSTAT3). Based on the observation that PTPMeg2 interacts with STAT3 in the absence of IL-6, we concluded that PTPMeg2 interacts with both the phosphorylated and unphosphorylated STAT3.
To reveal the cellular location of the PTPMeg2/STAT3 complex, we performed an immunofluorescence staining assay in MCF7 cells transfected with STAT3 and PTPMeg2. The results showed that STAT3 was located in the cytoplasm under a quiet condition, but translocated into the nucleus after IL-6 stimulation (Figure 1G, b2-b3). When STAT3 was co-expressed together with PTPMeg2, a notable co-localization of the two proteins in the cytoplasm was observed (Figure 1G, c3). Interestingly, we observed that STAT3 remained in the cytoplasm under the stimulation of IL-6 when PTPMeg2 was co-expressed (Figure 1, 2, 3). This result suggests that PTPMeg2 blocks the translocation of STAT3 from the cytoplasm into the nucleus upon IL-6 stimulation. To support this notation, a mutant PTPMeg2CS, which lost the ability to dephosphorylate STAT3, failed to block STAT3 localization into the nucleus in response to IL-6 stimulation (Figure 1, 2, 3). These results suggest that STAT3 colocalizes with PTPMeg2 in the cytoplasm and overexpression of PTPMeg2 inhibits the translocation of STAT3 upon cytokine stimulation.
PTPMeg2 Enhances Dephosphorylation of STAT3
Our observation that over-expression of PTPMeg2 blocks STAT3 translocation implied that PTPMeg2 may regulate STAT3 phosphorylation. Since PTPMeg2 is a phosphatase, we determined to examine whether PTPMeg2 dephosphorylates STAT3. To this end, HEK293T cells were transfected with Flag-STAT3 and Myc-PTPMeg2 plasmids under IL-6 treatment for 30 min. The results showed that the level of pSTAT3 was decreased when PTPMeg2 was co-expressed with STAT3 (Figure 2A). In contrast, transfection of mutant PTPMeg2CS failed to decrease the level of pSTAT3 (Figure 2A). To examine whether the decreased level of pSTAT3 is induced by a dephosphorylation or protein degradation process, the level of pSTAT3 was examined after withdrawal of IL-6 and in the presence of MG132, an inhibitor of proteosome. Results showed that the level of pSTAT3 was decreased much more quickly when PTPMeg2 was over-expressed than that without PTPMeg2 (Figure 2B). At the same time, the level of pSTAT3 remained unchanged in the presence or absence of MG132 (Figure 2B). These dada indicated that PTPMeg2 induces dephosphorylation of pSTAT3 rather than its degradation. Furthermore, we showed that over-expression of PTPMeg2 promoted the dephosphorylation of STAT3 at the residue Tyr 705 but had no effect on the phosphorylation level of pSTAT3 at the residue Ser727 (Figure 2C). The role of PTPMeg2 on the dephosphorylation of pSTAT3 (Tyr705) was further confirmed in a dosage dependent experiment (Figure 2D). These results suggested that ectopic expression of PTPMeg2 regulates the tyrosine phosphorylation of STAT3.
PTPMeg2 enhances dephosphorylation of STAT3. (A) PTPMeg2 decreases the level of pSTAT3. Levels of pSTAT3(Tyr705) were examined in the HEK293T cells transfected with Flag-STAT3 and different forms of PTPMeg2 in the presence or absence of IL-6 (10 ng/ml) for 30 min. (B) PTPMeg2 increases the dephosphorylation of STAT3. HEK293T cells were treated with IL-6 (10 ng/ml) for 30 min, followed by a starvation (shown as withdrawal) for different times. (C) PTPMeg2 promotes the STAT3 dephosphorylation rate. HEK293T cells were treated with IL-6 (10 ng/ml) for 30 min, followed by starvation for different times. The dynamic changes of the pSTAT3 (Tyr705) levels demonstrates the dephosphorylation rate of STAT3. (D) PTPMeg2 mediates STAT3 dephosphorylation in a dose dependent manner. HEK293T cells were transfected with Flag-STAT3 (2 μg/well) and different amounts of PTPMeg2 in a 6 well plate. (E) PTPMeg2 dephosphorylates STAT3 in vitro. Different amounts of purified GST-PTPMeg2 (10, 5 and 2.5 μg/tube) was added to purified pSTAT3 in a PTPase buffer at 37°C for 30 min. pSTAT3 was prepared in HEK293T cells transfected with Flag-STAT3 for 48 h and then stimulated with IL-6 for 30 min. (F) The PTP domain of PTPMeg2 contributes to STAT3 dephosphorylation. Myc-PTP domain, Myc-SEC domain and different deletions of PTPMeg2 were co-transfected with Flag-STAT3 for 48 h before stimulation with 10 ng/ml IL-6 for 30 min.
To further confirm the role of PTPMeg2 on dephosphorylation of STAT3, purified GST-PTPMeg2 and GST-PTPMeg2CS fusion proteins were used to incubate with pSTAT3 prepared from mammalian cells for an in vitro phosphatase activity experiment. The results showed that the tyrosine phosphorylation level of STAT3 was dramatically reduced when GST-Meg2 protein was added in a dose dependent manner (Figure 2E). As controls, addition of GST or GST-PTPMeg2CS had no effect on the level of pSTAT3 (Figure 2E, lane 1 to 2). This result indicated that STAT3 is a substrate of PTPMeg2. To address whether the PTP domain of PTPMeg2 has the phosphatase activity, the SEC domain, PTP domain and mutations of different deletions (ΔSEC, ΔPTP) were generated to examine the effect on the level of pSTAT3. A Western blot result showed that both PTP domain and ΔSEC domain had the ability to dephosphorylate pSTAT3 (Figure 2F, lanes 4 and 5). These data indicated that the PTP domain is responsible for the phosphatase activity of PTPMeg2, which is in consistency with the role of the PTP domain in other phosphatases.
PTPMeg2 Suppresses the Transcriptional Activation of STAT3
We questioned whether PTPMeg2 regulates the transcriptional activity of STAT3 based on its interaction with STAT3. To this end, we used an APRE luciferase reporter, which responds to STAT3 activation, to examine the effect of PTPMeg2 on STAT3 mediated transcriptional activity. The results showed that over-expression of PTPMeg2 in MCF7 cells resulted in a decrease of the luciferase activity in response to over-expressed STAT3 and stimulation of IL-6 (Figure 3A). The inhibitory role of PTPMeg2 on the STAT3-mediated luciferase activity was dose dependent (Figure 3B, left columns). Interestingly, when the mutant PTPMeg2CS was increasingly expressed the STAT3-mediated luciferase activity was increased (Figure 3B, right columns). These results suggest that the mutant PTPMeg2CS acts as a dominant negative antagonist of endogenous PTPMeg2 in regulating STAT3 phosphorylation. In consistence, depletion of the PTP domain impaired the activity of the phosphatase (Figure 3C). Finally, we showed that depletion of PTPMeg2 by three shRNAs increased the luciferase activity mediated by STAT3 (Figure 3D) while these shRNAs dramatically recovered the phosphorylation of the endogenous STAT3 protein (Figure 3E). In contrast, over-expression of PTPMeg2 had no effect on the transcriptional activity of STAT1 in response to INF-gamma stimulation (Figure 3F). These results indicate that PTPMeg2 inhibits STAT3 activation with certain specificity.
PTPMeg2 decreases the transcription activity of STAT3. (A) PTPMeg2 inhibits IL-6-stimulated STAT3 transcriptional activity. Luciferase assays were performed using MCF7 cells with transient expression of Myc-PTPMeg2 and Flag-STAT3, transfected with APRE-Luc reporter and pRL-TK (as an internal control). Relative luciferase activities were normalized with the internal control. Results are presented as mean ± SD from three independent experiments. (B) PTPMeg2 inhibits STAT3-mediated transcriptional activity in a dose dependent manner. Different amounts of PTPMeg2 were co-expressed with STAT3 in the presence or absence of IL-6 stimulation. Different amounts of PTPMeg2CS were also co-expressed with STAT3. (C) The PTP domain of PTPMeg2 inhibits the STAT3 transcriptional activity. Different domains and deletions of PTPMeg2 were co-expressed with Flag-STAT3 in the MCF7 cells and luciferase assays were performed as above. (D) Depletion of PTPMeg2 results in an enhanced STAT3 transcriptional activity. An shRNA targeting PTPMeg2 was co-expressed with STAT3 in MCF7 cells transfected with APRE-Luc reporter and pRL-TK under stimulation of IL-6. Luciferase activities were assayed as above. (E) Levels of pSTAT3(Tyr705) and STAT3 were examined in the MCF7 cells overexpressed and depleted PTPMeg2 in separated membranes from the same lysates. (F) PTPMeg2 has no effect on the STAT1 transcriptional activity. M67 luciferase reporters were co-transfected into MCF7 cells with STAT1 and PTPMeg2, in the presence or absence of IFN-gamma. Luciferase activities shown are mean ± SD of three independent experiments.
PTPMeg2 Inhibits Breast Cancer Cell Proliferation and Tumor Growth in Nude Mice
Since STAT3 phosphorylation is highly related to tumorigenesis, we attempted to examine whether PTPMeg2 could affect tumor progression. For this purpose, we used two human breast cancer cell lines MCF7 and MDA-MB-231. We observed that the endogenous PTPMeg2 protein level was low in MDA-MB-231 cells but high in MCF7 cells while the level of endogenous pSTAT3 displayed a reversed trend (Figure 4A). Therefore we determined to establish a gain-of-function model in MDA-MB-231 cells and a loss-of-function model in MCF7 cells.
PTPMeg2 inhibits breast cancer cell proliferation and tumor growth in nude mice. (A) PTPMeg2 expression correlates with the pSTAT3 level in breast cancer cells. The levels of pSTAT3, STAT3 and PTPMeg2 are shown. (B) Depletion of PTPMeg2 results in an increased level of pSTAT3 and expression of downstream gene expression. An shRNA targeting PTPMeg2 was stably expressed in MCF7 cells (MCF7/shPTPMeg2). Two downstream proteins CyclinD1 and Bcl-xL were examined. (C) Silencing PTPMeg2 increases the MCF7 cell proliferation. MCF7 stable cell lines were examined by an MTT assay. (D-F) Depletion of PTPMeg2 increases growth of tumors from MCF7 cells in nude mice. 1 × 107 MCF7 cells stably expressing shPTPMeg2 were injected s.c. into the right flanks of 6-week-old female nude mice (n = 6). Tumor weights (D), xenograft tumors (E), tumor sizes (F) are shown. (G) Over-expression of PTPMeg2 inhibits pSTAT3 and expression of its targeted genes. MDA-MB-231 cells were infected with Ad/PTPMeg2 or Ad/GFP. (H) Over-expression of PTPMeg2 inhibits MDA-MB-231 cell proliferation. MDA-MB-231 cells were examined by an MTT assay. (I-K) Over-expression of PTPMeg2 represses tumor growth. Tumor formation was observed in nude mice injected with 2 × 106 MDA-MB-231cells infected with Ad/PTPMeg2 or Ad/GFP. Xenograft tumors (I), tumor weights (J) and tumor volumes (K) are shown. All the data were obtained from 6 Balb/c-nu nude mice. Tumor formation was examined in nude mice injected with 6 × 106 MDA-MB-231cells infected with pMSCV/PTPMeg2 or pMSCV/vector virus. Xenograft tumors (L), tumor weights (M) and tumor volumes (N) are shown.
To address the increased pSTAT3 was the cause of decreased PTPMeg2, we stably depleted PTPMeg2 by using an shRNA targeting PTPMeg2 in MCF7 cells. A Western blot analysis showed that pSTAT3 was dramatically increased when PTPMeg2 was depleted (Figure 4B, right lane). Intriguingly, a cell proliferation experiment result showed that the growth of MCF7 cells was increased when PTPMeg2 was depleted (Figure 4C). An in vivo tumor growth experiment in a xenograft tumor model in mice showed that MCF7 cells with stable depletion of PTPMeg2 formed larger tumors than mock transfected cells (Figure 4D-E) and grew more rapidly (Figure 4F). On the other hand, we over-expressed PTPMeg2 in MDA-MB-231 cells using an adenovirus expression system. The results showed that MDA-MB-231 cells infected with the adenovirus expressing PTPMeg2 had a lower level of endogenous pSTAT3 than the cells infected with a control adenovirus (GFP) (Figure 4G, right lanes). And these cells grew much more slowly (Figure 4H) and the cells formed smaller sized tumors (Figure 4I-J) and had slower tumor growth rate (Figure 4K), lower tumor weight (Figure 4J) and slower tumor growth (Figure 4K). To further confirm the inhibitory role of PTPMeg2 on tumor growth in a moderate expression system, we used the retroviral system to ectopically express PTPMeg2 (pMSCV/PTPMeg2) in MDA-MB-231 cells. The results were similar to that using the adenovirus expression system (Figure 4L-N). All these results indicated that PTPMeg2 inhibits STAT3 phosphorylation directly and PTPMeg2 is a tumor suppressor.
To confirm the inhibitory role of PTPMeg2 on tumor growth is depended on regulation of STAT3 phosphorylation, we used v-Src transformed NIH3T3 fibroblasts in a xenograft tumor model. The result showed that v-Src transformed cells had a much higher STAT3 phosphorylation level than non-transformed cells and over-expression of PTPMeg2 significantly decreased the level of pSTAT3 (Figure 5A). Consistent with the decreased level of pSTAT3, the tumor size (Figure 5B), weight (Figure 5C) and tumor growth (Figure 5D) from v-Src transformed cells were decreased when PTPMeg2 was forcedly expressed. These data implied a correlation of PTPMeg2-reduced tumor growth and the decreased level of pSTAT3. To address whether pSTAT3 is a key target by PTPMeg2, we examined the cell proliferation ability in the STAT3 KO cells. A MTT experiment indicated that overexpression of PTPMeg2 inhibited the cell growth dramatically in wild-type cell but had no effect in the STAT3 KO cells (Figure 5E-F), suggesting that the inhibitory role of PTPMeg2 on the cell proliferation is depended on STAT3. Together with the biochemical data, these results suggested that the ability of PTPMeg2 to inhibit the tumor growth and cell proliferation is depending on its role of regulation of phosphorylated STAT3.
The correlation of the PTPMeg2 expression and the pSTAT3 level. (A-D) Over-expression of PTPMeg2 decreases the tumor growth of v-Src transformed fibroblasts. (A) v-Src activated pSTAT3 was decreased by Ad/PTPMeg2. A Western blot was used to show the indicated proteins. STAT3 and pSTAT3 were examined in separated membranes from the same lysates. (B-D) 5 × 105 v-Src/NIH3T3 cells infected with Ad/PTPMeg2 or Ad/GFP were used to inject s.c. into the right (Ad/PTPMeg2) and left (Ad/GFP) flanks of 6-week-old male nude mice (n = 5). Xenograft tumors (B), tumor weights (C) and tumor volumes (D) are shown. (E) Over-expression of PTPMeg2 decreases the STAT3+/+ cell proliferation. Mouse hepatic STAT3+/+ and STAT3+/+cells infected with the indicated adenovirus were used to examine the proliferation rate by an MTT assay. Results were averaged from three repeats. (F) A Western blot was used to show the indicated proteins. (G) A high level of pSTAT3 is correlated with a low expression of PTPMeg2 in the breast carcinoma. Representative images of immunohistochemical staining of PTPMeg2, pSTAT3 and STAT3 are shown. All the tumor tissues were derived from surgery sections. Peri-tumor (P) or tumor (T) tissues are marked with dotted lines. Scale bars, 50 μm.
In consistence with the tumor growth experiment, we examined the expression of STAT3 downstream genes. The result showed that both Bcl-xL and cyclin D1 were dramatically increased when PTPMeg2 was depleted in MCF7 cells (Figure 4B). In contact, Bcl-xL and cyclin D1 were decreased when PTPMeg2 was over-expressed in MDA-MB-231 cells (Figure 4G). We observed no alteration of STAT3 expression but the phosphorylated STAT3 level was changed with either over-expression or depletion of PTPMeg2 (Figure 4B, G). These results suggest that PTPMeg2 regulates STAT3 phosphorylation and thereafter the downstream gene expression.
To address whether PTPMeg2 regulates STAT3 dephosphorylation in human tumors, we examined the correlation of pSTAT3 level and expression of PTPMeg2 in human breast cancers. The result showed that expression of PTPMeg2 was in a strong positive status in peritumoral tissues (90% = 37/41) and in a negative status in paired tumor tissues (83% = 59/71) (Figure 5E, Table 1). In contrast, pSTAT3 remained at a low (or negative) level in the peritumoral tissues but at a high level (71% = 52/73) in the paired tumor tissues (Figure 5E). We observed a negative correlation between PTPMeg2 expression and the pSTAT3 level (spearman's correlation coefficient was -3.33, p = 0.004, Table 2) from a Spearman's correlation test. The analysis also revealed that the increased STAT3 level was correlated with reduced PTPMeg2 expression in the breast carcinoma (correlation coefficient is -2.65, p = 0.023, Table 2). These data indicated that PTPMeg2 might be an important regulator of STAT3 dephosphorylation in tumors.
Breast Cancer Res. 2012;14(2):R38 © 2012 BioMed Central, Ltd.
Copyright to this article is held by the author(s), licensee BioMed Central Ltd. This is an Open Access article: verbatim copying and redistribution of this article are permitted in all media for any purpose, provided this notice is preserved along with the article's original citation.