Prolactin regulates ZNT2 expression through the JAK2/STAT5 signaling pathway in mammary cells

Cell culture.

HC11 cells were a gift from Dr. Jeffery Rosen (Houston, TX) and used with permission of Dr. Bernd Groner (Institute for Biomedical Research, Frankfurt, Germany). Cells were routinely maintained in RPMI 1640 supplemented with 10% fetal bovine serum, epidermal growth factor (EGF, 10 ng/ml), insulin (5 μg/ml), and gentamycin (50 μg/ml). Cells were treated with PRL (1 μg/ml) and cortisol (2 μM) in serum-free RPMI 1640 and were pretreated with the specific inhibitors AG490 (10 μM) and PD-98059 (10 μM) for 4 h where indicated.

RNA isolation and real time PCR.

Total RNA was extracted from cells using TRIzol reagent (Invitrogen, Carlsbad, CA) and treated with DNAse I (Roche Applied Science, Indianapolis, IN). Complementary DNA was synthesized using the M-MLV Reverse Transcriptase (Applied Biosystems, Foster City, CA). Fluorescence real-time PCR was performed with iQ SYBR Green Supermix (Bio-Rad Laboratories, Hercules, CA) using the MJ Research Opticon 2 system (MJ Research, Waltham, MA). Pfaffl's method was applied for relative quantification of ZnT2 expression normalized to endogenous control (β-actin) gene (46). Specific primers for mZnT2 (5′-GGCCCGACGCAAACTCTATGT-3′, 5′-TGCACGGCCAGGTACACACGA-3′) and β-actin (5′-AGCCATGTACGTAGCCATCC-3′, 5′-CTCTCAGCTGTGGTGGTGAA-3′) were designed using Primer Express Software v2.0 (Applied Biosystems). PCR reactions were run in triplicate (95°C for 10 min, and 40 cycles of 95°C for 15 s, 58°C for 30 s and 72°C for 30 s).

Nuclear run-on assay.

Transcriptional regulation of the ZnT2 gene by prolactin was first confirmed following actinomycin D treatment (5 μg/ml). A nuclear run-on assay using biotin labeling and real-time quantitative PCR was performed by the method described previously (45). Briefly, cells were treated with PRL (1 μg/ml) for 24 h, after which the cells were trypsinized and resuspended in 4 ml lysis buffer (10 mM Tris·HCl, pH 7.4, 3 mM MgCl₂, 10 mM NaCl, 150 mM sucrose, 0.5% Nonidet P-40, and 1 mM PMSF). After 5 min incubation in ice, nuclei were then collected by centrifugation (170 g) and were kept in 100 μl freezing buffer (50 mM Tris·HCl, pH 8.3, 40% glycerol, 5 mM MgCl₂, and 0.1 mM EDTA). In vitro RNA synthesis from DNA was performed for 30 min at 37°C in transcription buffer [20 mM Tris·HCl, pH 8.0, 100 mM KCl, 2.5 mM MgCl₂, 2 mM dithiothreitol (DTT), 100 mM sucrose, and 10% glycerol, 2 mM each of ATP, GTP, and CTP, 1.3 mM UTP]. Biotin-16-UTP (Roche Applied Science) was included in the reaction mixture for detection of newly transcribed RNA and RNA was isolated with TRIzol reagent. The biotinylated transcripts were captured on streptavidin-coated magnetic beads and eluted in diethylpyrocarbonate (DEPC)-treated water. Total cDNA was then synthesized by random hexamer-primed reverse transcription of captured biotinylated nascent RNA. Quantitative real-time PCR was performed to monitor the transcript abundance of ZnT2.

Plasmid constructs.

The putative promoter of mouse ZnT2 (−988 bp upstream from the putative start codon) was amplified by PCR from the genomic BAC clone (RP23-204E6) with primers (forward: 5′-ATAGGTACCTCCGTTGAAGGTCCCACATC-3′, reverse: 5′-ATACTCGAGTCGGTGCGCCCTCGAAGTT-3′) and subcloned into the pGL3-basic vector (Promega, Madison, WI) via KpnI and XhoI restriction sites (pGL3-ZnT2–988). Two putative STAT5 binding elements (GAS elements) were identified in the promoter of ZnT2 (PROMO sever). Mutagenesis was performed using the Phusion Site-Directed Mutagenesis Kit (New England Biolab, Beverly, MA), and the following mutants were generated and sequence verified: pGL3-ZnT2-GAS1M [deletion of the putative GAS1 element (TTCCAGGAA) at position −674 to −665]; pGL3-ZnT2-GAS2M [deletion of the putative GAS2 element (TTCCAGGAA) at position −377 to −368]; and pGL3-ZnT2-GAS12M (deletion of both GAS1 and GAS2 elements). Plasmids pcDNA3/STAT5a and pcDNA3/PRLr, which express STAT5a and long-form PRLr, respectively, were constructed using PCDNA3.1/V5-His TOPO TA Expression Kit (Invitrogen). Expression vector for dominant-negative STAT5 (DNSTAT5) was generated as previously described (57).

Transient transfection and luciferase assay.

All plasmids were purified using the GenElute HP plasmid Midiprep Kit (Sigma) and quantitated by UV spectrophotometry. Cells were seeded in 24-well plates and transfected with either pGL3-ZnT2–988, pGL3-ZnT2-GAS1M, pGL3-ZnT2-GAS2M, pGL3-ZnT2-GAS12M, or pGL3-basic (vector control) using Lipofectamine 2000 (Invitrogen). Cells were transfected with either wtSTAT5a or the long-form of PRLr (or both) to amplify the luminescence where indicated. Cells were cotransfected with DNSTAT5 to inhibit the wtSTAT5-mediated signaling where indicated.

Twenty-four hours after transfection, cells were treated with PRL (0.3, 0.6, 1.2, 2.4, and 4.8 μg/ml) for 24 h. PRL (1 μg/ml) was found to be the optimal concentration for maximal ZnT2 promoter activation. The cells were harvested and lysed in passive lysis buffer (125 μl/well). Luciferase activity was measured using the Dual Luciferase Reporter Assay System (Promega). The relative firefly luciferase activity was normalized by Renilla luciferase activity. Each experiment was repeated at least three times.

Electrophoretic mobility shift assays.

Cells were cultured in serum-free media and then stimulated with PRL (1 μg/ml) for various times (1, 2, and 8 h) before harvesting nuclear protein extracts (NPE) as described previously (8). Oligonucleotides containing GAS1 (5′-ATCCAATTCCAGGAAGTCTCT-3′) and GAS2 (5′-ACAGGATTCCAGGAAGAGTGG −3′) were end-labeled using [γ-³²P]ATP (Perkin Elmer, Waltham, MA) and incubated in binding buffer (10 mM HEPES, 100 mM KCl, 0.05 mM EDTA, 2.5 mM MgCl₂, and 6% glycerol) with NPE (2 μg) and poly(dI-dC) (80 μg/ml) for 15 min at room temperature. The end-labeled, double-stranded oligonucleotide (5′-GCAGATTTCTAGGAATTCAAA-3′), harboring a consensus GAS element that exists in bovine β-casein promoter, was used as a positive control. Nonradioactive oligonucleotide competitor (GAS1: ATCCAATTCCAGGAAGTCTCT; GAS2: 5′-ACAGGATTCCAGGAAGAGTGG-3′) was generated by annealing unlabeled primers and added to binding reactions in 60-fold molar excess. The mutant competitor (GAS1: ATCCAATTATTTTAAGTCTCT; GAS2: 5′-ACAGGATTATTTTAAGAGTGG-3′), harboring mutation of the putative GAS element, was also generated and included in the binding reaction in the indicated group. For supershift assay, NPE were preincubated for 30 min with 2 μl STAT5 antibody (Cell Signaling Technology, Danvers, MA). Samples were electrophoresed through a 7.5% nondenaturing polyacrylamide gel. After electrophoresis, the gels were dried and exposed to autoradiography film at −70°C.

Chromatin immunoprecipitation assays.

Cells were cultured in serum-free media and then stimulated with PRL (1 μg/ml) for 1, 2, and 8 h. Cells were then incubated with 1% formaldehyde for 20 min at room temperature to cross-link proteins to the DNA and washed with cold PBS buffer before lysis with SDS lysis buffer (50 mM Tris·HCl, pH 8.0, 1% SDS, 10 mM EDTA, 1 mM PMSF, and 1 mM Na₃VO₄). Cell lysates were sonicated on ice to shear chromatin and diluted in chromatin immunoprecipitation (ChIP) assay dilution buffer (16.7 mM Tris·HCl, pH 8.0, 0.01% SDS, 1% Triton X-100, 1.2 mM EDTA, and 167 mM NaCl). Fifty microliters of each diluted chromatin sample was collected as the input DNA. The rest of chromatin suspension was precleared with Protein A-agarose slurry containing salmon sperm DNA for 30 min and then incubated with either STAT5 antibody or normal rabbit IgG (control) overnight at 4°C with rocking. The agarose beads were collected by centrifugation and washed sequentially for 5 min with each of the following four buffers: low-salt wash buffer, high-salt wash buffer, LiCl wash buffer, and TE buffer. Immunoprecipitated DNA was eluted with 200 μl elution buffer (1% SDS and 0.1 M NaHCO₃). Cross-linking was reversed by heating at 65°C for 4 h, and samples were digested with protease K and purified by phenol-chloroform extraction and ethanol precipitation. Quantitative real-time PCR was used to quantitate both the input DNA and ChIP DNAs using ZnT2 promoter primers (5′-CTGTCACTCCCTTGCACCAA-3′; 5′-GTTGCTAAGTTAGGACTGGATGA-3′). ChIP data are expressed as percentage of input DNA (amount of ChIP DNA PCR product divided by that of the input DNA).

Statistical analysis.

Data represent means ± SE. Statistical significance for each experimental comparison was determined using Student's t-test or one-way ANOVA, with P < 0.05 considered a significant effect of treatment.

PRL induces ZnT2 expression in HC11 cells.

In this study, we used normal mammary epithelial cells that endogenously express PRLr to determine whether ZnT2 expression is regulated at the transcriptional level by PRL. Cells were treated with PRL (1 μg/ml) for up to 24 h, and ZnT2 mRNA level was quantified by real-time PCR. PRL treatment significantly increased ZnT2 mRNA level (fivefold) compared with untreated cells. (Fig. 1A). To evaluate whether the PRL-induced increase in ZnT2 mRNA level resulted from increased transcription of the ZnT2 gene or mRNA stability, we coincubated the cells with PRL (1 μg/ml) and actinomycin D (5 μg/ml) for 24 h. Treatment with actinomycin D, which blocks the function of the transcriptional machinery, completely abolished the PRL-induced upregulation of ZnT2 mRNA in HC11 cells, indicating that PRL regulates ZnT2 by a RNA polymerase-dependent transcriptional mechanism (Fig. 1B). To further determine whether the increase in ZnT2 mRNA level we detected in response to PRL stimulation resulted from increased transcription of the ZnT2 gene, we next performed a nuclear run-on assay in isolated nuclei from cells treated with PRL (1 μg/ml). Importantly, we detected a significant increase in the rate of ZnT2 gene transcription in cells treated with PRL (Fig. 1C). These data clearly indicated that PRL stimulation increased ZnT2 transcription and thus we next aimed to determine the signaling mechanisms responsible for this effect.

AG490 reduced effects of PRL treatment on ZnT2 mRNA level.

Both the Jak2/STAT5 and MAPK signaling pathways have been reported to be important for PRL signal transduction in cells of mammary epithelial origin (13, 48). To determine which signaling pathway plays a role in the transcriptional regulation of the ZnT2 gene, HC11 cells were pretreated with the Jak2 inhibitor AG490 (10 μM) or the MAPK inhibitor PD-98059 (10 μM) for 4 h before PRL treatment to inhibit Jak2 or MAPK signaling. After the 4-h pretreatment, the HC11 cells were treated with PRL for an additional 24 h. The results showed that chemical inhibition of Jak2 eliminated the PRL-induced increase in ZnT2 mRNA abundance (Fig. 2). In contrast, no effect of MAPK inhibition was observed. These findings strongly suggested that the transcriptional activation of ZnT2 in mammary cells by PRL is Jak2/STAT5 dependent.

PRL induced the transcriptional activity of ZnT2 promoter through involvement of STAT5.

To determine the mechanisms through which Jak2/STAT5 mediates the activation of ZnT2 transcription, we explored ZnT2 promoter activation by PRL using a luciferase reporter assay. A 988-bp fragment of the wild-type ZnT2 promoter was subcloned into the pGL3-basic luciferase reporter vector. Cells were transfected with the ZnT2 full-length promoter construct (pGL3-ZnT2–988) or with the empty vector as a control and then treated with PRL (1 μg/ml) for 24 h. Activation of the ZnT2 promoter was measured by the luciferase activity, and the result indicated that the promoter was only slightly increased in HC11 cells treated with PRL. To amplify the luminescence, we utilized an artificial cell system in which we overexpressed either the long-form of PRLr or STAT5a (or both) to exacerbate the response to PRL. The results showed that the introduction of either long-form PRLr or STAT5a significantly enhanced the PRL-induced luciferase activity, and cotransfection of both PRLr and STAT5a resulted in the most robust increase in luciferase activity (P < 0.05) (Fig. 3A).

To verify a role for STAT5 in this response we next generated a dominant negative STAT5 construct, which is a COOH-terminal truncated mutant (STAT5aΔ713) that acts as a dominant negative mutant for STAT5 by forming inactive heterodimers with wild-type STAT5 proteins (55). Coexpression of DNSTAT5 with the ZnT2 promoter construct resulted in both decreased basal and PRL-induced luciferase activity (Fig. 3B). Together, these results indicate that the enhanced transcription of the ZnT2 gene resulting from PRL stimulation results from activation of the Jak/STAT signaling pathway through STAT5a.

Identification of two functional STAT5 binding sites in the ZnT2 promoter.

We next aimed to determine how STAT5 elicited the activation of the ZnT2 promoter. Genes that respond directly to PRL often possess canonical STAT5 response elements (GAS), most often located in the region upstream of the transcription start site (1, 2). The canonical GAS element is composed of the consensus sequence 5′-TTCNNNGAA-3′. In the presence of PRL, PRLr signaling can activate STAT5, which then form homo- or heterodimers that translocate into the nucleus where they bind to the GAS element in the promoters of target genes. Analysis of the 5′-untranslated region of the ZnT2 gene revealed that the ZnT2 promoter contains two putative GAS elements within this region, a distal GAS1 element (from −674 to −665) and a proximal GAS2 element (from −377 to −368) (Fig. 4A).

To verify the functional relevance of these putative GAS elements, both GAS1 and GAS2 elements were mutated as illustrated in Fig. 4A, and cells were transfected with either the wild-type promoter construct ZnT2PFL (pGL3-ZnT2-988) or the mutated promoter constructs GAS1M (pGL3-ZnT2-GAS1M) or GAS2M (pGL3-ZnT2-GAS2M) or GAS12M (pGL3-ZnT2-GAS12M). PRLr and STAT5a constructs were cotransfected to enhance PRL signaling. Cells were then treated with varying doses of PRL (0, 0.3, 0.6, 1.2, 2.4, and 4.8 μg/ml) and luciferase activity was measured after 24 h treatment. As illustrated in Fig. 4B, addition of PRL increased luciferase activity in a dose-dependent manner only in cells transfected with the wild-type promoter. In comparison, mutation of either the GAS1 or GAS2 element resulted in a substantial reduction in PRL-induced ZnT2 promoter activity. Moreover, the promoter harboring the GAS1 mutation was able to respond to very high PRL concentrations. Elimination of both GAS1 and GAS2 elements completely eliminated PRL-stimulated ZnT2 promoter activity, suggesting that a cooperative relationship exists between the GAS1 and GAS2 elements for PRL-induced activation. It is interesting to note that the proximal GAS2 element appears more important to PRL stimulation as the GAS2 mutant resulted in a complete abrogation in activity beyond that of the distal GAS1 element.

To confirm that these two GAS elements in the ZnT2 promoter are targets for STAT5 binding, we performed gel electrophoretic mobility assays using nuclear proteins isolated from cells that had been treated with PRL for 1, 2, 8, and 24 h (Fig. 5A). As illustrated in Fig. 5A, a strong protein-DNA complex was detected following incubation with a GAS2-specific oligonucleotide probe (Fig. 5A, lane 7 to lane 12). In contrast, a very weak binding complex was detected using a GAS1-specific oligonucleotide probe (Fig. 5A, lane 1 to lane 6). Because it was very difficult to visualize the GAS1-protein complex at the same exposure level as the GAS2-protein complex, we extended the exposure time to visualize GAS1 binding (Fig. 5A, inset). This confirmed that the distal GAS1 element has the capability to form a protein-DNA complex similar in size to the GAS2 complex, although the binding affinity was much lower than for GAS2. To verify that the GAS2-DNA complex was specific, Fig. 5B illustrates that the GAS2 complex was competed by 60-fold molar excess of unlabeled GAS2 oligonucleotide probe (Fig. 5B, lane 3) but not by the mutant GAS element (Fig. 5B, lane 4). A supershift assay was also performed to investigate the specific binding of STAT5 protein to the GAS2 element. These data clearly illustrated that the mobility of the nuclear protein complex was retarded following preincubation with STAT5 antibody (Fig. 5B, lane 5). Similarly, the specificity of the GAS1-DNA complex was verified under the same conditions except these experiments required a longer exposure time (Fig. 5C). Our data further revealed that a time lapse in DNA binding to GAS1 does not occur in response to PRL activation. STAT5 binding activity could be detected 2 h following PRL addition (Fig. 5A, lane 10) and reached a maximum at 8 h (Fig. 5A, lane 11). Longer treatment resulted in a decrease of STAT5-DNA binding activity (Fig. 5A, lane 12). In contrast, STAT5-DNA binding in complex with GAS1 has almost the same response pattern but with much less sensitivity (Fig. 5A, inset).

PRL induced STAT5 recruitment to ZnT2 promoter in vivo.

Results from the in vitro gel shift experiments demonstrated a specific interaction between STAT5 and both GAS elements in the ZnT2 promoter following treatment with PRL. To verify that STAT5 binds specifically to the ZnT2 promoter in vivo, we stimulated cells with PRL for different periods of time (1, 2, and 8 h) and performed ChIP analysis using STAT5 antibody or normal rabbit IgG. Genomic DNA fragments, which were immunoprecipitated with STAT5 antibody or normal rabbit IgG, were then analyzed by quantitative real-time PCR to assess the binding capability of STAT5 to DNA in vivo. As Fig. 6 illustrates, binding of STAT5 to the ZnT2 promoter was specific as immunoprecipitation with normal rabbit IgG was extremely low compared with immunoprecipitation with STAT5-specific antibody. Our data indicated that STAT5 indeed binds to the ZnT2 promoter following PRL stimulation in vivo. Although a modest increase on STAT5 recruitment was detected in as little as 1 h following PRL stimulation, this difference was not statistically significant until 8 h (Fig. 6). Taken together, results from this study clearly indicate that PRL increases the transcription of ZnT2 through the stimulation of the Jak2/STAT5 signaling cascade.