Progesterone Receptor Inhibits Proliferation of Human Breast Cancer Cells via Induction of MAPK Phosphatase 1 (MKP-1/DUSP1)*

Background: To elucidate mechanisms for anti-proliferative actions of progesterone/progesterone receptor (PR) in human breast cancer cells, we focused on MKP-1, a dual-specificity MAPK phosphatase. Results: MKP-1 serves as a PR target gene that mediates progesterone/PR suppression of T47D cell proliferation. Conclusion: MKP-1 is a critical mediator of anti-proliferative and anti-inflammatory actions of PR in the breast. Significance: Findings suggest MKP-1 as a potential therapeutic target in breast cancer. The roles of progesterone (P4) and of progesterone receptor (PR) in development and pathogenesis of breast cancer remain unclear. In this study, we observed that treatment of T47D breast cancer cells with progestin antagonized effects of fetal bovine serum (FBS) to stimulate cell proliferation, whereas siRNA-mediated knockdown of endogenous PR abrogated progestin-mediated anti-proliferative effects. To begin to define mechanisms for the anti-proliferative action of P4/PR, we considered the role of MAPK phosphatase 1 (MKP-1/DUSP1), which catalyzes dephosphorylation and inactivation of MAPKs. Progestin treatment of T47D cells rapidly induced MKP-1 expression in a PR-dependent manner. Importantly, P4 induction of MKP-1 was associated with reduced levels of phosphorylated ERK1/2, whereas siRNA knockdown of MKP-1 blocked progestin-mediated ERK1/2 dephosphorylation and repression of FBS-induced cell proliferation. The importance of PR in MKP-1 expression was supported by findings that MKP-1 and PR mRNA levels were significantly correlated in 30 human breast cancer cell lines. By contrast, no correlation was observed with the glucocorticoid receptor, a known regulator of MKP-1 in other cell types. ChIP and luciferase reporter assay findings suggest that PR acts in a ligand-dependent manner through binding to two progesterone response elements downstream of the MKP-1 transcription start site to up-regulate MKP-1 promoter activity. PR also interacts with two Sp1 sites just downstream of the transcription start site to increase MKP-1 expression. Collectively, these findings suggest that MKP-1 is a critical mediator of anti-proliferative and anti-inflammatory actions of PR in the breast.

with progestin significantly increased breast cancer risk, as compared with use of estrogen alone (10,11). Depending on the physiological state and incubation conditions, P 4 can act either in a proliferative or differentiative capacity in cultured breast cancer cells (1). Moreover, both growth-stimulatory and -inhibitory actions of P 4 have been reported in cultured breast epithelial cells and on cancer development in animal tumor models (13,14). Studies using PR-positive mammary carcinoma cell lines as a model have demonstrated a biphasic cellular response to either P 4 or synthetic progestins (R5020 or ORG 2058), with an immediate proliferative burst followed by a sustained growth arrest (15)(16)(17). Moreover, treatment with P 4 has been reported to suppress cell proliferation in response to different mitogens, such as estrogens, serum, and insulin-like growth factor, alone or in combination (18). P 4 has been found to inhibit cell cycle progression of breast cancer cells by transient induction of the cyclin-dependent kinase inhibitors (CDKIs) p21 Cip1/WAF1 (p21) and p18 INK4c (p18), followed by a sustained induction of p27 Kip1 (p27), leading to association of these CDKIs with the different G 1 CDK complexes. CDKI⅐CDK complex formation results in down-regulation of CDK activity leading to decreased pRb phosphorylation and cell cycle arrest in the late G 1 phase (19 -22).
Mitogen-activated protein kinase (MAPK) activation has been implicated in breast cancer progression and metastasis. MAPK signal transduction pathways are evolutionarily conserved and play a central role in conveying information from the extracellular environment to the cytoplasm and finally into the nuclear compartment (23). There are at least three known MAPK signaling pathways, including the extracellular signalregulated protein kinase (ERK) pathway, the p38 MAPK pathway, and the c-Jun N-terminal kinase (JNK)/stress-activated protein kinase pathway (24). The ERK pathway plays a critical role in cell proliferation, and its activity is enhanced by growth factors, integrins, activation of G protein-coupled receptor systems, products of several proto-oncogenes (25), as well as by estrogen (26 -28) and P 4 (16,29). ERK1 and ERK2 protein kinases (ERK1/2) are activated by phosphorylation on threonine and tyrosine residues within a conserved Thr-X-Tyr motif via a kinase cascade involving Ras/Raf/Mek proteins. Activation of ERK1/2 commonly causes its translocation from the cytoplasm to the nucleus. Once inside the nucleus, the active ERK1/2 are capable of directly phosphorylating a number of transcription factors and cell cycle regulators (24).
Recently, a family of protein phosphatases, the dual-specificity phosphatases (DUSPs), were discovered to have the ability to interact and catalyze dephosphorylation of active MAPK (30). To date, ten members of the DUSP family have been identified in mammals (31,32). DUSP1 (also known as MAPK phosphatase 1 (MKP-1), CL100, 3CH134, Erp, and hVH-1) was observed to inactivate ERK1/2 by dephosphorylation of both threonine and tyrosine residues within the activation motif (33). MKP-1 has been shown to inhibit a number of cellular responses mediated by ERK1/2, JNK, and p38 MAPK (34 -36). Moreover, increasing MKP-1 protein expression has been shown to suppress growth of MCF-7 breast cancer cells (37). Recent studies have identified MKP-1 as a glucocorticoid receptor (GR) target gene, which mediates GR anti-inflamma-tory activity (38 -41). Progestins, acting through PR, also have been found to serve an anti-inflammatory role in both the uterus (42,43) and in the breast (44,45). We observed that the anti-inflammatory actions of PR are mediated by ligand-dependent and -independent mechanisms, resulting in an inhibition of NF-B activation with consequent down-regulation of hCOX-2, aromatase/hCYP19, and HER-2/neu expression (44). Notably, MKP-1 mRNA expression was observed to be induced by P 4 /PR in human breast cancer cells (22).
In consideration of the potential role of MKP-1 as an important PR target gene in the breast that mediates some of its antiinflammatory/anti-proliferative actions, in the present study, we investigated the mechanisms whereby P 4 /PR modulates MKP-1. We observed that the PR acts in a ligand-dependent manner to suppress serum-induced T47D cell proliferation and that these anti-proliferative actions were associated with PR induction of MKP-1 expression. In addition, P 4 /PR induction of MKP-1 promoter activity was mediated via PR binding to PREs in DNA and by PR-Sp1 interactions. Finally, using an siRNA approach, we verified that MKP-1 serves as a PR target gene that mediates P 4 repression of ERK1/2 activation by serum growth factors and the subsequent increase in cell proliferation.

MATERIALS AND METHODS
Reagents and Cell Culture-T47D breast cancer cells and HEK293 cells were obtained from the American Type Culture Collection (ATCC, Manassas, VA). T47D cells were maintained in RPMI 1640 medium (Invitrogen, Carlsbad, CA) with phenol red and supplemented with 7.5% fetal bovine serum (FBS) plus antibiotic-antimycotic solution (Sigma). HEK293 cells were maintained in Dulbecco's modified Eagle's medium (DMEM, Invitrogen) with phenol red and supplemented with 5% FBS plus antibiotic-antimycotic solution. Cells were cultured and grown in an air-carbon dioxide (95:5) atmosphere at 37°C. For transient transfection studies, cells were seeded in medium without phenol red and supplemented with 2.5% FBS stripped with dextran-coated charcoal (Invitrogen). For RNA and protein expression experiments, cells were seeded in maintenance medium; the next day cells were changed to serum-free medium without phenol red for another 24 h before treatment. For treatment with various reagents, cells were incubated in serum-free medium without phenol red for times indicated. Progesterone (Sigma), Mifepristone (RU486, Sigma), and all other chemicals were the highest quality available from commercial sources.
Cloning and Plasmids-The cDNA for human MKP-1 was purchased from Origene (Rockville, MD) and subcloned into pcDNA3 expression vector (Invitrogen). The pMKP1-A-Luc plasmid, which contains Ϫ403 bp of sequence upstream and ϩ490 bp downstream of the transcription start site (TSS) of the human MKP-1 gene was amplified from human genomic DNA and cloned into pGL4 vector (Promega, Madison, WI). pMKP1-B (Ϫ403/ϩ216), pMKP1-C (Ϫ403/ϩ113), and pMKP1-D (Ϫ403/ϩ18) were made by PCR amplification using pMKP1-A as template and subcloned into pGL4 vector. Sitedirected mutagenesis was performed using a QuikChange II site-directed mutagenesis kit (Stratagene, La Jolla, CA), according to the manufacturer's protocol.
Transient Transfection, RNA Interference, and Reporter Assay-For MKP-1 overexpression experiments, T47D cells were transfected with pcDNA3 or MKP-1 expression vector using Neon® Transfection System (Invitrogen) according to the manufacturer's recommendations. After transfection, cells were seeded in 6-well plates with growth medium for 24 h and then placed in fresh RPMI 1640 medium without phenol red or FBS. For RNA interference (RNAi) experiments, small inhibitory RNA (siRNA) oligonucleotides against PR-A and PR-B (43,46), human MKP-1 (Invitrogen), and silencer-negative control oligonucleotides (Ambion, Austin, TX) were transfected using the Neon® Transfection System (Invitrogen). For luciferase reporter assays, T47D and HEK 293 cells were seeded in 24-well plates and transfected using FuGENE® HD transfection reagent (Roche Applied Science) with MKP-1 reporter constructs (100 ng), PR-B expression vectors (100 ng), and Renilla luciferase plasmid (20 ng, Promega). One day after transfection, cells were treated with DMSO or P 4 (100 nM) for 24 h in medium without phenol red or FBS. Cells from each experiment were then harvested in 100 l of 1ϫ Passive Lysis Buffer (Promega). Firefly luciferase and Renilla luciferase activities were assayed using a Dual-Luciferase assay system (Promega). Relative luciferase activities were calculated by normalizing Firefly luciferase activity to Renilla luciferase activity in the same samples to correct for transfection efficiencies.
Cell Proliferation-For cell proliferation assays, ethynyl-2Јdeoxyuridine (EdU) incorporation assays were performed. T47D cells were seeded in 4-well chamber slides, and cells were placed in medium without serum and phenol red on the next day. After 24-h incubation, cells were treated with DMSO or P 4 (100 nM) in the absence or presence of 5% FBS in phenol redfree medium for another 24 h. EdU (10 M, Invitrogen) was added to the medium for the last 4 h of the treatment. EdU incorporation was assayed using a Click-iT EdU Alexa Fluor® 488 cell proliferation assay kit (Invitrogen). Cell fixation, permeabilization, and EdU detection were performed according to the manufacturer's instructions. 4Ј,6-Diamidino-2-phenylindole (DAPI) staining was used to identify nuclei for determination of cell number. EdU and DAPI signals were captured with a Zeiss Axiovert 100M fluorescence microscope, and captured images were processed and analyzed with ImageJ software (National Institutes of Health). The EdU and DAPI signals for each sample were analyzed in six different fields. The relative intensity of the EdU signal was calculated by normalizing the fluorescence signal of EdU with the DAPI staining.
ChIP-Chromatin immunoprecipitation (ChIP) was carried out, as described in detail previously (43). ChIP assays were performed using a ChIP kit (CHIP, Upstate), according to the manufacturer's recommendations. Briefly, confluent 100-mm dishes of T47D cells were treated with DMSO or P 4 (100 nM) for 1 h and then cross-linked with 1% formaldehyde. Cells were washed with ice-cold 1ϫ PBS, lysed, and sonicated on ice to produce sheared soluble chromatin. The soluble chromatin was precleared with Protein A/G Plus agarose beads (60 l) at 4°C for 1 h and then incubated with antibodies for PR (R&D Systems), Sp1 (Upstate), or with non-immune IgG (Santa Cruz Biotechnology, Santa Cruz, CA) at 4°C overnight. Immune complexes were collected on protein A/G-agarose and eluted. Cross-linking of immunoprecipitated chromatin complexes and of input controls was reversed by heating at 65°C for 4 h, followed by proteinase K (Invitrogen) treatment. The purified DNA was subjected to PCR amplification and analyzed by electrophoresis on an ethidium bromide-stained agarose gel. The purified DNA also was analyzed by quantitative PCR.
Statistical Analysis-Statistical significance was determined by analysis of variance and Student's t test, and the levels of probability were noted. The data were expressed as means Ϯ S.E. for at least three separate (replicate) experiments for each treatment.

RESULTS
Progesterone Acting via PR Induces MKP-1 mRNA Expression in T47D Cells-As noted above, MKP-1 is a known target of GR and has been shown to be induced by P 4 in breast cancer cells (47). The mechanism whereby P 4 modulates MKP-1 expression and the role of MKP-1 in mediating effects of P 4 to alter cell proliferation remain undefined. To analyze the temporal induction of MKP-1 expression by progestin/PR, real-time PCR was used to measure mRNA in T47D cells cultured in the absence or presence of the synthetic progestin, R5020 (Fig. 1A). Treatment of T47D cells with R5020 induced MKP-1 mRNA expression within 4 h and reached maximal levels after 24 h of treatment. In association with mRNA induction, MKP-1 protein levels were rapidly induced following P 4 treatment (Fig.   1B). Furthermore, upon knockdown of PR levels using a siRNA targeting PR (Fig. 1C) or by decreasing PR function by co-treatment with the PR antagonist RU486 (Fig. 1D), P 4 induction of MKP-1 mRNA expression was significantly impaired. Taken together, these data suggest that P 4 induction of MKP-1 expression in T47D cells is mediated by PR.
To examine the relationship between PR and MKP-1 further in vivo, we analyzed expression levels of MKP-1, PR, and GR in 30 human breast cancer cell lines. As shown in Fig. 1E, there was a positive correlation between MKP-1 and PR mRNA levels in these cell lines (p Ͻ 0.0001; r ϭ 0.515; Pearson correlation). Interestingly, no correlation between MKP-1 and GR mRNA was observed. Previously, we reported that levels of PR mRNA in these human breast cancer cell lines were correlated with FIGURE 1. Progesterone induces MKP-1 expression in breast cancer cells. T47D cells cultured in phenol red-free medium in the absence of FBS were treated with DMSO (V), R5020, or progesterone (P 4 , 100 nM) for 2-36 h. A, RNA was isolated and reverse-transcribed, and levels of MKP-1 mRNA were assessed by real-time qPCR (qRT-PCR). Relative levels of MKP-1 mRNA were calculated by normalizing against ribosomal RNA. Data are the mean Ϯ S.E. of values from four independent experiments and are expressed as arbitrary units. *, significantly (p Ͻ 0.05) increased as compared with vehicle (V). B, whole cell lysates were analyzed by immunoblotting for levels of MKP-1 and Sp1 (loading control) as a function of time after initiation of R5020 treatment. C, T47D cells were transfected with siRNA targeting both PR-B and PR-A, then treated with vehicle or P 4 (100 nM) for 6 h. qRT-PCR was used to analyze levels of MKP-1 mRNA. Data are the mean Ϯ S.E. from three replicate samples and are expressed as -fold induction by P 4 treatment over vehicle control. *, significantly (p Ͻ 0.05) increased as compared with vehicle (V). D, T47D cells were treated with vehicle or P 4 (100 nM) with or without RU486 (100 nM) for 12 h. MKP-1 mRNA levels were analyzed by qRT-PCR. Data are the mean Ϯ S.E. of three replicate samples and are expressed as -fold induction by P 4 treatment over vehicle control. *, significantly (p Ͻ 0.05) increased as compared with vehicle (V); **, significantly (p Ͻ 0.05) decreased as compared with progesterone. E, RNA from 30 human breast cancer cell lines was extracted and reverse-transcribed, and mRNA levels of PR, GR, MKP-1, and IB␣ were assessed by qRT-PCR. The relative levels of all four mRNA transcripts were calculated by normalizing against h36B4 mRNA and analyzed using Pearson's coefficient of correlation. mRNA levels of the NF-B inhibitor IB␣ (44), which also is a PR target gene in breast cancer cells (44,48). Interestingly, the correlation between MKP-1 and IB␣ mRNA in the breast cancer cell lines also was found to be highly significant (Fig. 1E).
We also searched the Oncomine 4.4 database for breast cancer datasets to interrogate the relationship between MKP-1 mRNA expression with PR status. The Richardson breast cancer dataset shows that MKP-1 mRNA levels were higher in PR(ϩ) breast carcinoma samples, as compared with samples with negative PR status (49). Other datasets also indicated that MKP-1 mRNA levels were higher in PR(ϩ) breast cancer cells and tissue samples than in those negative for PR (50 -53).
Overexpression of MKP-1 in T47D Cells Suppresses Phosphorylation of ERK1/2 and Cell Proliferation-To examine the potential anti-proliferative function of MKP-1 in breast cancer cells, T47D cells were transiently transfected with MKP-1 expression vector, and EdU incorporation was employed to assess cell proliferation. After 24 h of serum (5% FBS) stimulation, EdU incorporation was significantly increased as compared with cells maintained in serum-free medium ( Fig. 2A). However, overexpression of MKP-1 in T47D cells significantly inhibited the increase in cell proliferation caused by FBS, as compared with vector control (Fig. 2A). The anti-proliferative activity of MKP-1 also was observed using BrdU and thymidine incorporation assays (data not shown). It should be noted that the relative intensity of fluorescence signal observed in the EdU assay was determined by normalizing the EdU signal to the cell number, as determined by DAPI staining. This excludes the possibility that the reduced incorporation of EdU observed upon MKP-1 overexpression was caused by the cytotoxicity and cell death.
The effect of FBS to increase T47D cell proliferation was associated with increased phosphorylation of ERK1 and -2 ( Fig.  2B), whereas the action of overexpressed MKP-1 to inhibit cell proliferation was associated with an inhibition of FBS-induced ERK1/2 phosphorylation. Taken together, these results suggest that increased MKP-1 expression suppressed T47D cell proliferation in part, due to its action to decrease levels of phospho-ERK1/2.
Progesterone and Progesterone Receptors Suppress T47D Cell Proliferation and Repress ERK1/2 Phosphorylation by Induction of MKP-1 Expression-In light of these collective findings, we postulated that P 4 exerts an inhibitory effect on cell proliferation in breast cancer cells via its effect to up-regulate MKP-1 expression. To explore the effects of P 4 /PR on T47D cell proliferation and the role of MKP-1, T47D cells were transfected with non-targeting siRNA, or with siRNAs for PR-A and PR-B or for MKP-1 to knock down the endogenous levels of these proteins. The cells were then synchronized in RPMI 1640 medium without FBS for 24 h and then treated with DMSO (vehicle (V)) or P 4 (100 nM) with or without 5% FBS for another 24 h. Cell proliferation was analyzed using the EdU staining assay. As can be seen in Fig. 3A, when T47D cells were incubated with non-targeting control siRNA, FBS significantly increased the rate of cell proliferation, whereas treatment with P 4 in FBS-containing medium significantly inhibited the cell proliferation rate, as compared with cells cultured with FBS alone. However, when cells were transfected with siRNAs to knock down PR-A and PR-B and cultured in medium containing FBS plus P 4 , the inhibitory effect of P 4 on cell proliferation was abrogated. This suggests that PR mediates the inhibitory effect of P 4 on T47D cell proliferation. In cells transfected with FIGURE 2. Overexpression of MKP-1 represses proliferation of T47D breast cancer cells and inhibits ERK1/2 phosphorylation. T47D cells were transiently transfected with pcDNA3 (empty vector) or MKP-1 expression vector, and then cultured in RPMI 1640 medium with or without 5% FBS for 24 h. A, cell proliferation was assayed by assessing EdU relative to DAPI staining. EdU and DAPI signals were captured using a Zeiss Axiovert 100M fluorescence microscope and analyzed with ImageJ software (NIH). The EdU and DAPI signals for each sample were analyzed in six different fields. The relative intensity of the EdU signal was calculated by normalizing the fluorescence signal of EdU with the DAPI staining. Data are expressed as mean Ϯ S.E. for each treatment group (n ϭ 6 fields per sample). *, significantly (p Ͻ 0.05) increased as compared with vehicle (V); **, significantly (p Ͻ 0.05) decreased as compared with cells cultured in FBS and transfected with empty vector. B, overexpression of MKP-1 represses phosphorylation of ERK1/2 induced by FBS. T47D cells were transiently transfected with pcDNA3 (empty vector) or MKP-1 expression vector and cultured with or without 5% FBS for 8 h. Whole cell lysates were assessed by immunoblot analysis for levels of Sp1 (loading control), MKP-1, phospho-ERK1/2 (P-ERK1/2), and ERK1/2. Shown is a representative immunoblot of at least three replicate experiments. DECEMBER 16, 2011 • VOLUME 286 • NUMBER 50

JOURNAL OF BIOLOGICAL CHEMISTRY 43095
siRNA for MKP-1, we also observed that co-treatment with P 4 did not inhibit cell proliferation induced by FBS, indicating that MKP-1 expression is crucial for the anti-proliferative action of P 4 in T47D cells. Moreover, when these cells were transfected with siRNA targeting PR-A and PR-B or MKP-1 and cultured in FBS-containing medium, cell proliferation was significantly increased over that of cells transfected with non-targeting siRNA (Fig. 3A). This suggests that both PR and MKP-1 play a role to maintain the cell proliferation rate.
To further elucidate the underlying mechanisms for the antiproliferative activity of P 4 /PR and of MKP-1, T47D cells were transfected with a non-targeting siRNA or with siRNAs for PR-A/-B or MKP-1 and serum-starved for 24 h, as above. The cells were then cultured in medium containing FBS with or without P 4 for 2, 6, and 12 h, and ERK1/2 expression and phosphorylation state, as well as PR-A/B and MKP-1, were analyzed by immunoblotting. As shown in Fig. 3B, in T47D cells transfected with non-targeting control siRNA, incubation in medium containing FBS markedly induced phosphorylation of ERK1/2 at 2, 6, and 12 h. Co-treatment with P 4 reduced the levels of ERK1/2 phosphorylation in FBS-treated cells after 6 and 12 h of treatment. Importantly, this decrease in ERK1/2 phosphorylation was associated with P 4 induction of MKP-1 protein expression at these time points. On the other hand, siRNA knockdown of PR-A/B expression caused an apparent increase in ERK1/2 phosphorylation as compared with FBS-treated cells trans-FIGURE 3. Anti-proliferative effect of P 4 /PR on T47D cells is mediated by induction of MKP-1. A, effect of P 4 /PR to inhibit T47D cell proliferation is prevented by knockdown of PR-A/-B or MKP-1. T47D cells were synchronized in phenol red-free RPMI 1640 medium without FBS for 24 h and then treated with DMSO (V) or P 4 (100 nM) in phenol red-free RPMI 1640 medium with or without 5% FBS for another 24 h. Cell proliferation was determined by using the EdU staining assay, as above. Data are the mean Ϯ S.E. (n ϭ 6). *, significantly (p Ͻ 0.05) increased as compared with corresponding vehicle (V) controls; **, significantly (p Ͻ 0.05) decreased as compared with corresponding cells cultured with FBS without P 4 ; (#) significantly (p Ͻ 0.05) increased as compared with corresponding treatment in cells transfected with non-targeting siRNA. B, siRNA-mediated knockdown of PR-A/B and MKP-1 prevents P 4 inhibition of ERK phosphorylation. Transient transfection of specific siRNAs targeting MKP-1 and PR-A/B reduced endogenous MKP-1 and PR protein expression, respectively, in T47D cells. After transfection with 20 nM non-targeting siRNA, PR-A/B, or MKP-1 siRNA, cells were synchronized in phenol red-free RPMI 1640 medium without FBS for 24 h and then treated with DMSO (V) or P 4 (100 nM) in phenol red-free RPMI 1640 medium with or without 5% FBS for times indicated. Whole cell lysates were assayed by immunoblotting for phospho-ERK1/2, Sp1, PR-A/B, and MKP-1. Shown are representative immunoblots of experiments repeated at least three times with comparable results. fected with non-targeting siRNA, impaired the capacity of P 4 to reduce ERK1/2 phosphorylation, and blocked the action of P 4 to induce MKP-1 expression (Fig. 3B). Interestingly, incubation of T47D cells transfected with non-targeting or PR-A/B siRNA in FBS-containing medium caused a marked and rapid decline in MKP-1 expression (Fig. 3B). The effect of FBS to markedly reduce MKP-1 protein levels is likely due to activated ERKmediated phosphorylation and subsequent ubiquitination of MKP-1, resulting in its proteolytic degradation (54). Notably, this effect of FBS was antagonized by P 4 treatment in cells transfected with non-targeting siRNA but not in cells transfected with PR-A/B siRNA.
In cells transfected with MKP-1-specific siRNA, the ability of P 4 to repress ERK1/2 phosphorylation also was blocked; this was correlated with a loss of P 4 induction of MKP-1 expression (Fig. 3B). Taken together, these findings suggest that the effect of P 4 /PR to suppress FBS-induced cell proliferation is mediated by induction of MKP-1 expression.

Progesterone Up-regulates MKP-1 Expression via Recruitment of PR to PRE and Sp1
Response Elements-To further elucidate the molecular mechanisms for P 4 induction of MKP-1 expression, we performed in silico analysis to define putative response elements within the regulatory regions of the hMKP-1 gene. Using the MatInspector program from Genomatix, we identified four potential PREs scattered throughout a 4-kb region surrounding the hMKP-1 TSS (Fig. 4A). ChIP assays were then employed to assess PR occupancy of MKP-1 genomic regions containing the four PREs in cultured T47D cells. As can be seen in Fig. 4B, PRs were recruited to two of the PREs, PRE1 and PRE2, located at ϩ192 and ϩ434 bp downstream of TSS, respectively. PR also was bound to a region (P0), which lies proximal to the TSS and does not contain a predicted PRE. To quantify in vivo binding of PR to these three sites, we used ChIP-qPCR (right panel). Treatment of the T47D cells with P 4 for 60 min significantly increased the recruitment of PR to the PRE1 and PRE2 sites but had no effect to increase PR binding to the P0 site (Fig. 4C).
To determine the functional roles of the identified PR-binding regions in P 4 induction of MKP-1, we generated a luciferase reporter construct containing a genomic region of ϳ0.9 kb from Ϫ403 bp upstream to ϩ490 bp downstream of the MKP-1 gene TSS (pMKP1-A). In HEK293 cells co-transfected with pMKP1-A and an expression vector containing PR-B, P 4 treatment significantly induced luciferase activity compared with vehicle, suggesting that PREs are contained within this region (Fig. 5A). In HEK293 cells co-transfected with PR-B and reporter constructs containing deletions of this MKP-1 reporter construct, including pMKP1-B (Ϫ403/ϩ216) and pMKP1-C (Ϫ403/ϩ113), P 4 retained the ability to induce luciferase activity. However, after further deletion of the MKP-1 first exon from ϩ113 to ϩ18, P 4 lost its ability to induce luciferase activity. This suggests that the region between ϩ18 and ϩ113 contains critical progesterone-responsive elements. As mentioned, using ChIP, this region was observed to bind endogenous PR in the T47D cells (Fig. 4, B and C), although no consensus PREs were present. On the other hand, within the ϩ18 to ϩ113 bp genomic region, we identified two putative Sp1 sites at ϩ20 and ϩ52 bp. We considered that these may be responsible for P 4 induction of MKP-1 promoter activity, because PR-Sp1 interaction has been reported to mediate rapid effects of P 4 on PR transactivation (2,55). As indicated in Fig.  5B, point mutations introduced individually into either of the two Sp1 sites in the MKP-1C-luciferase reporter did not affect P 4 /PR-B responsiveness of MKP-1 promoter activity in co-transfected HEK293 cells. However, when both Sp1 sites were mutated, both basal and P 4 induction of MKP-1 promoter activity were greatly reduced, suggesting that these Sp1 sites act together to mediate basal and PR induction of MKP-1 expression. Furthermore, in T47D cells co-transfected with PR-B and the MKP-1C-luciferase reporter, mithramycin A, an inhibitor  hMKP-1 promoter. A, schematic of hMKP-1 gene (Ϫ1000 to ϩ3000 bp) and genomic regions amplified for ChIP analysis. "ϩ1" is the TSS; the filled boxes and circles indicate coordinates of regions amplified in the ChIP assay. T47D cells were treated with DMSO (V) or P 4 (100 nM) for 60 min before cross-linking with formaldehyde. Chromatin was isolated and immunoprecipitated with anti-PR-A/-B antibody or non-immune mouse IgG, as indicated. B, to analyze recruitment of PR to putative PREs, cross-linked chromatin was isolated from P 4 -treated cells and immunoprecipitated with antibody recognizing PR-A/-B; the five genomic regions were amplified by PCR, electrophoresed in an agarose gel, and stained with EtBr. Input was used as positive control. PR binding was observed only for P0, PRE1, and PRE2. C, to quantify P 4 induction of PR binding, qPCR was performed using the indicated primers and calculated relative to the corresponding IgG control. The -fold recruitment by P 4 treatment is expressed relative to the corresponding vehicle treated. Data are the mean Ϯ S.E. of three replicate samples. As can be seen, P 4 treatment increased binding of endogenous PR to PRE1 and PRE2, whereas no induction of PR binding to P0 was observed.
of Sp1 transcription factor binding to GC-rich elements in DNA (56), blocked P 4 induction of MKP1-C-luciferase reporter expression (Fig. 5C). Also, treatment of T47D cells with mithramycin A significantly reduced P 4 induction of endogenous MKP-1 (Fig. 5D). Treatment with mithramycin A, by contrast, did not affect the induction of SGK mRNA expression by P 4 in T47D cells (supplemental Fig. S1). These collective findings suggest that Sp1 may play a role in P 4 /PR-mediated MKP-1 expression.
To determine whether the putative PREs at ϩ197 and ϩ434 bp are required for P 4 -induced MKP-1 expression, these were mutated singly and together in the context of the MKP-1Aluciferase reporter construct containing the MKP-1 genomic region spanning Ϫ403 to ϩ490 bp. These mutated constructs were co-transfected with a PR-B expression vector into HEK293 cells cultured with or without P 4 . Constructs containing mutations in both Sp1 sites, with or without mutations in both PREs, also were tested. In contrast to our findings using  Fig. 4 to bind endogenous PR. B, to assess the role of these putative Sp1 sites, HEK293 cells were transiently co-transfected with PR-B expression vector and the hMKP-1C-Luc reporter construct (Ϫ403 to ϩ113) containing mutations in either or both of these sites. Whereas mutation of the individual putative Sp1 response elements had little or no effect, mutation of both sites prevented P 4 induction of hMKP-1 promoter activity. A and B, data are the mean Ϯ S.E. of three replicate determinations for each treatment group. *, significantly (p Ͻ 0.05) increased as compared with vehicle (V). C, T47D cells were co-transfected the MKP-1C luciferase reporter construct and the PR-B expression plasmid. After transfection, cells were treated with DMSO (V) or P 4 (50 nM) alone or in combination with the Sp1 DNA-binding inhibitor, mithramycin A (MA, 100 nM), for 12 h. D, T47D cells were treated with vehicle or progesterone (P 4 , 100 nM) alone or in combination with MA (100 nM) for 6 h. qRT-PCR were used to analyze endogenous levels of MKP-1. In C and D, data are the mean Ϯ S.E. of three replicate determinations for each treatment group. *, significantly (p Ͻ 0.05) increased as compared with vehicle (V); **, significantly (p Ͻ 0.05) decreased as compared with progesterone alone (P 4 ).
HEK293 cells transfected with the MKP-1C-luciferase reporter construct, in which mutation of both Sp1 sites markedly reduced basal and P 4 induction of promoter activity, mutations in both Sp1 sites in the MKP-1A construct had little effect on basal or P 4 stimulation of MKP-1 promoter activity (Fig. 6A). Although mutagenesis of one or both PREs resulted in a reduction in basal and P 4 -induced MKP-1 promoter activity, P 4 responsiveness was partially retained. On the other hand, when both PREs and both Sp1 sites were mutated, P 4 induction of MKP-1 promoter activity was greatly reduced. This suggests that both Sp1 sites and the two PREs cooperatively act to mediate P 4 induction of MKP-1 promoter activity. To further study the mechanism of P 4 /PR induction of MKP-1 expression, HEK 293 cells were transiently transfected with a PR-B wild type (WT), a PR-B S345A mutant (2), which abrogates PR-Sp1 interaction, or with a PR-B DNA binding-defective mutant (mDBD) containing three point mutations in the DNA-binding domain (G585E-S586G-V589A). In HEK293 cells transfected with equivalent amounts of either the PR-B S345A mutant or the PR-B mDBD mutant expression vectors (supplemental Fig. S2), P 4 partially retained the ability to induce endogenous MKP-1 expression, as compared with PR-B WT (Fig. 6B). However, in parallel HEK293 cell transfection studies, the PR-B S345A mutant was unable to induce expression of p21, cyclin-dependent kinase inhibitor 1 (Fig. 6B), whereas, PRB-mDBD mediated P 4 induction of p21 expression at a level comparable to that of PRB-WT. Notably, p21 has been shown to be up-regulated by P 4 /PR via PR interaction with Sp1 (2). Importantly, the PRB-mDBD expression vector failed to mediate P 4 induction of a mouse mammary tumor virus-luciferase reporter (MMTV-Luc), which was markedly induced by P 4 in the presence of co-transfected PRB-WT and PRB-S345A expression vectors (supplemental Fig. S2). Taken together, these results suggest that PR acts in a ligand-dependent manner through binding to two PREs (ϩ197 and ϩ434) downstream of the MKP-1 TSS to up-regulate MKP-1 promoter activity. PR also acts via nonclassical mechanisms to increase levels of MKP-1 expression by interaction with Sp1 transcription factor bound to two putative Sp1 response elements just downstream of the TSS.

DISCUSSION
The role of P 4 and PR in the development and pathogenesis of breast cancer remains controversial. On the one hand, as mentioned above, P 4 treatment of human breast cancer cell lines has been reported to cause a rapid and transient increase in cell proliferation via up-regulation of MAPK signaling and increased cell cycle progression (15)(16)(17). On the other hand, the presence of PR in a breast tumor serves as an independent predictor for benefit from adjuvant endocrine therapy and of disease-free survival (8,9). Moreover, breast tumors that are PR(Ϫ) have a much higher proliferation rate and are more likely to manifest increased expression of the tumorigenic prognostic indicators, HER-2/neu and EGFR, than PR(ϩ) tumors (57)(58)(59)(60). In postmenopausal women with advanced metastatic breast cancer that had progressed during tamoxifen therapy, medroxyprogesterone acetate (MPA) inhibited disease progression (61)(62)(63)(64). In addition, in postmenopausal women with early stage (I-IIB) breast cancer, MPA significantly decreased the recurrence rate after a median follow-up of 37 months (65). Further, using clinical data obtained from human breast tumor specimens, we observed that PR expression levels serve as an important predictor of a lower stage of breast cancer at diagnosis, a measure of breast tumor progression (66). Using human breast cancer cell lines as a model, it was demonstrated that, in ER(Ϫ)/PR(Ϫ) MDA-MB-231 cells, overexpression of PR inhibited cell growth and induced spreading and adherence (67). Moreover, in ER(ϩ)/PR(ϩ) T47D breast cancer cells, P 4 treatment suppressed cell proliferation stimulated by different mitogens (18) and inhibited cytokine activation of cyclooxygenase 2 (43).
The objective of the present study was to elucidate the underlying mechanisms for the anti-proliferative actions of P 4 /PR in human breast cancer cells. We focused on the role of MKP-1/ DUSP1, a dual-specificity MAPK phosphatase (33), which has been reported to be up-regulated by glucocorticoids (38 -41) and progestins (47) in a number of cell types. Herein, we observed that treatment with P 4 repressed proliferation of ER ϩ / PR ϩ T47D cells induced by serum. The anti-proliferative action of P 4 was associated with a rapid induction of mRNA and protein expression of MKP-1. Importantly, P 4 stimulation of MKP-1 was associated with reduced levels of phosphorylated ERK1 and -2. Overexpression of MKP-1 in these cells also antagonized the stimulatory effect of serum on T47D cell proliferation and inhibited ERK1/2 phosphorylation, whereas serum induction of T47D proliferation was enhanced by knockdown of endogenous MKP-1. The role of PR was supported by the finding that the P 4 -mediated induction of MKP-1 was abrogated in T47D cells by PR knockdown or upon treatment with the PR antagonist, RU486. Moreover, serum induction of T47D proliferation was enhanced by knockdown of endogenous PR.
Our findings of an anti-proliferative effect of P 4 in T47D cells differ from those of Lange and colleagues who observed a biphasic effect of P 4 , with stimulation of cell proliferation at time points of 18 and 24 h, followed by a subsequent decline (16). We suggest that these disparities are primarily due to differences in the culture conditions utilized. In their studies (16), T47D cells that stably expressed PR-B were serum-starved, then cultured with or without progestin in medium with 5% charcoal-stripped FBS. In our study, T47D cells were cultured with or without P 4 in medium containing 5% FBS (not charcoalstripped). After 24 h, we observed such a pronounced stimulatory effect of FBS on EdU incorporation in the absence of P 4 (Fig. 3A) that any stimulatory effect of P 4 /PR on EdU incorporation was likely masked. Under these conditions, the predominant effect of P 4 was to antagonize the effect of FBS on cell proliferation. Moreover, the marked stimulatory effect of FBS on cell proliferation was due, in part, to its combined actions to increase ERK1/2 phosphorylation and to decrease MKP-1 expression. This inhibitory effect on MKP-1 was previously reported to be due to ERK-mediated phosphorylation of MKP-1 followed by MKP-1 degradation (54). However, when cells were incubated with P 4 plus FBS, progesterone/PR inhibited ERK phosphorylation and blocked the decline in MKP-1 protein levels (Fig. 3B).
Our data suggest that PR acts to increase MKP-1 expression and to inhibit MAPK activation in T47D cells both by classical and non-classical mechanisms. Findings from ChIP and luciferase reporter assays suggest that PR acts in a ligand-dependent manner through binding to two PREs (ϩ197 and ϩ434) downstream of the MKP-1 TSS to up-regulate MKP-1 promoter activity. These putative PREs do not correspond in sequence or location to GR half-sites observed to mediate glucocorticoid induction of the mouse MKP-1 gene (68). PR also acts via nonclassical mechanisms to increase levels of MKP-1 expression by interaction with two putative Sp1 response elements just downstream of the TSS. Binding of endogenous PR to the Sp1 sites appears to occur in a ligand-independent manner. These alternative mechanisms of PR-mediated activation of MKP-1 promoter activity are supported by the findings that P 4 induction of MKP-1 expression is only partially reduced upon transient transfection of HEK293 cells with a PR-B S345A mutant (2), which abrogates PR-Sp1 interaction, or with a PR-B DNA binding-defective mutant (mDBD) containing three point mutations in the DNA-binding domain (G585E-S586G-V589A).
Studies by Lange and colleagues using T47D cells suggested that progestin/PR rapidly activates EGF receptor (EGFR), c-Src, and MAPK signaling, resulting in increased phosphorylation of PR on Ser-345 (2). PR phosphorylated on Ser-345, in turn, tethers to Sp1 sites within specific PR-regulated promoters (e.g. EGFR), resulting in a rapid burst of cell proliferation (2). Based on the present findings, we suggest that Pϳ345-PR interaction with Sp1 response elements within the MPK-1 promoter may serve a critical and protective role in the breast by inactivating MAPKs and switching off this proliferative signal. The tethering of PR to promoter-bound Sp1 may further enhance and stabilize PR binding to the two downstream PREs to increase the magnitude and duration of MKP-1 induction and maintain a reduced rate of cell proliferation.
In the present study, we observed that the increase in T47D cell proliferation and in the levels of phosphorylated ERK1/2 caused by siRNA-mediated knockdown of endogenous PR-A and PR-B occurred in the absence of exogenous P 4 . Our previous findings also revealed that, in T47D cells, siRNA-mediated knockdown of endogenous PR-A and -B caused a marked induction of cyclooxygenase 2, aromatase/CYP19, and Her-2/ neu expression in the absence of P 4 (43,44). This was associated with a pronounced increase in nuclear levels of NF-B p65 (44), suggesting that decreased PR expression promotes activation of NF-B. Thus, PR likely suppresses NF-B activation, in part, via a ligand-independent mechanism that may involve direct interaction with p65 (3). Our previous findings suggest that PR also inhibits NF-B activation in a ligand-dependent manner via P 4 /PR induction of IB␣ expression. However, in other studies using T47D cells, it was observed that, although the progestin R5020 increased IB␣ expression, it had no effect on p65 nuclear localization (45).
In studies using T47D cells, the progestin R5020 was observed to inhibit estradiol-induced cell proliferation in a PR-dependent manner (69). The action of estrogen to stimulate breast cancer growth is mediated, in part, by non-genomic mechanisms involving interaction of ER␣ with growth factor receptors and signaling cascades that originate at the plasma membrane. These include activation of MAPK and phosphatidylinositol 3-kinase (PI3K) pathways, with enhanced estrogenindependent phosphorylation and activation of nuclear and membrane-associated ER␣ (70). Interestingly, the ER␣ co-activator SRC-3 (AIB1), which is up-regulated in metastatic human breast cancer, is phosphorylated in response to E 2 , MAPK (71), and Her-2/neu activation (72). SRC-3 phosphorylation promotes its nuclear translocation, increased interaction with ER, and enhancement of ER transcriptional activity (71). The HER-2/neu proto-oncogene is amplified in ϳ30% of invasive ductal carcinomas of the breast (57). Her-2/neu overexpression triggers activation of MAPK, PI3K, and NF-B signaling pathways (73,74). Moreover, MAPK overexpression in breast cancer cells promotes activation/phosphorylation of NF-B (75), ER␣ (76), and AP-1 (77), which enhance breast cancer initiation, growth, and survival. Previously, we observed that Her-2/neu expression was markedly inhibited by P 4 treatment of T47D cells and by up-regulation of PR expression in MCF-7 cells (44). Thus, in addition to its action to inhibit MAPK activation via induction of MKP-1, P 4 /PR may also suppress expression of Her-2/neu, a major activator of MAPK.
Besides its potential action to inhibit breast cancer cell proliferation, MKP-1 also likely plays an anti-inflammatory role. Overexpression of MKP-1 in macrophages blocks lipopolysaccharide (LPS)-induced production of pro-inflammatory cytokines by inhibiting p38 MAPK (78). Conversely, MKP-1 Ϫ/Ϫ mice manifest an exacerbated response to LPS, manifesting elevated cytokine levels in serum and prolonged activation of p38 MAPK and JNK in macrophages (79). The anti-inflammatory actions of glucocorticoids in a variety of cell types have been attributed, in part, to their actions to increase expression of MKP-1 and decrease MAPK activation, resulting in decreased expression of genes encoding inflammatory mediators, such as cyclooxygenase 2. Although MKP-1 inhibition of p38 MAPK activity can cause decreased NF-B and AP-1 activation of proinflammatory gene transcription, decreased p38 activity also has been found to cause destabilization of pro-inflammatory mRNAs (80).
Despite the anti-inflammatory, immunosuppressive, and anti-proliferative actions of glucocorticoids, trials of glucocorticoid therapy in breast cancer have lacked efficacy (81). Interestingly, estradiol has been reported to inhibit GR transactivation activity by protein phosphatase 5-mediated dephosphorylation of GR at Ser-211 (82), a ligand-induced phosphorylation site associated with GR transcriptional activation (12). Consequently, treatment of human breast cancer cells with estradiol inhibited glucocorticoid induction MKP-1 and serum glucocorticoid kinase (SGK) expression (82). On the other hand, we observed that, whereas estradiol treatment of T47D cells blocked dexamethasone induction of MKP-1 and SGK, it failed to inhibit P 4 induction of MKP-1 and SGK expression (supplemental Fig. S3). Notably, the Ser-211 phosphorylation site present in GR is not conserved in PR. These intriguing findings suggest that P 4 /PR signaling is not affected by estrogen-medi-ated inhibition. Notably, whereas, expression levels of MKP-1 and PR were positively correlated in 30 human breast cancer cell lines, no correlation between the levels of MKP-1 and GR were observed (Fig. 1E). Similarly, we previously found that expression levels IB␣ and PR were positively correlated in these breast cancer cell lines, whereas no correlation was found between levels of IB␣ and GR (44). These collective findings suggest that MKP-1 is an important target of P 4 /PR in human breast cancer cells and may play a significant role in its antiproliferative and anti-inflammatory actions.