STAT1 facilitates oestrogen receptor α transcription and stimulates breast cancer cell proliferation

Abstract Oestrogen receptor α (ERα) is overexpressed in two‐thirds of all breast cancer cases and is involved in breast cancer development and progression. Although ERα ‐positive breast cancer can be effectively treated by endocrine therapy, endocrine resistance is an urgent clinical problem. Thus, further understanding of the underlying mechanisms involved in ERα signalling is critical in dealing with endocrine resistance in patients with breast cancer. In the present study, unbiased RNA sequence analysis was conducted between the MCF‐7 and MCF‐7 tamoxifen‐resistant (LCC2) cell lines in order to identify differentially expressed genes. The whole transcriptomic data indicated that the JAK‐STAT pathway is markedly up‐regulated, particularly the ISGF3 complex. As the critical effectors, STAT1 and IRF9 were up‐regulated 5‐ and 20‐fold, respectively, in LCC2 cells. The biological experiments indicated that STAT1 is important for ERα signalling. Depletion of STAT1 or inhibition of STAT1 function significantly decreased levels of ERα protein, ERα ‐target gene expression and cell proliferation in both the MCF‐7 and LCC2 cell lines. Chromatin immunoprecipitation revealed that ERα transcription is associated with STAT1 recruitment to the ERα promoter region, suggesting that transcriptional regulation is one mechanism by which STAT1 regulates ERα mRNA levels and ERα signalling in breast cancer cells. The present study reveals a possible endocrine‐resistant mechanism by which STAT1 modulates ERα signalling and confers tamoxifen resistance. Targeting of STAT1 is a potential treatment strategy for endocrine‐resistant breast cancers.

The present study reveals a possible endocrine-resistant mechanism by which STAT1 modulates ERα signalling and confers tamoxifen resistance. Targeting of STAT1 is a potential treatment strategy for endocrine-resistant breast cancers.

K E Y W O R D S
breast cancer, ERα, STAT1, transcription

| INTRODUCTION
Breast cancer is the most frequently diagnosed type of cancer in women worldwide. 1 A total of 60%-70% of breast cancer cases are ERα positive, which can be well-controlled by ERα selective antagonists such as tamoxifen. 2 Tamoxifen has a similar structure as E2; however, it has an extra chain that interferes with the conformational change of ERα protein into the active form. 3 Despite the effectiveness of tamoxifen treatment, a significant percentage of tumours with ERα expression develops endocrine resistance. 4 A number of different mechanisms have been shown to account for tamoxifen resistance. For example, ERα acquires constitutively active mutations at the ligand binding domain: Y537S and D538G.
These ERα mutant forms recruit the co-activators in the absence of oestrogen, while the affinity for ERα antagonists is also decreased. 5 ERα protein activity is also regulated by various post-translational modifications, such as phosphorylation and ubiquitination. [6][7][8] Several studies have shown that certain modifications at the ERα hinge domain enhance ERα transcriptional activity and confer tamoxifen resistance. 7,8 In addition to the active biological events in ERα signalling, a number of signalling pathways crosstalk with ERα via several effects. For example, numerous growth factor signalling kinases regulate ERα phosphorylation, including MAPK, RAS, AKT and PKA, [9][10][11] which subsequently enhance ERα stability or/and transcriptional activity and renders cells less sensitive to tamoxifen.
Although a number of possible and confirmed mechanisms have been shown to explain endocrine resistance in breast cancer, how endocrine resistance is generated in breast cancer remains unclear.
The LCC2 cell line, which was selected from the MCF-7 cell line for tamoxifen resistance in oophorectomized nude mice, is widely used as an evolutionary model for tamoxifen-resistant breast cancer. 12 This model was utilized in the present study in order to perform unbiased RNA sequencing. By comparing the transcriptomic profiles of the MCF-7 and LCC2 cell lines, JAK-STAT signalling was observed to be expressed at higher levels in LCC2 cells. As the main effectors of JAK-STAT signalling, STAT1 and IRF9 were markedly up-regulated in the LCC2 cells. STAT was shown to be elevated in breast cancer tumours, while its expression levels correlated with poor endocrine treatment outcome. The present study identified the involvement of STAT1 in facilitating ERα transcription in breast cancer cells.

| Cell culture
MCF-7 and LCC2 cell lines were used in our previous study. 13 The cells were cultured in DMEM (Invitrogen, Carlsbad, CA) supplemented with 10% foetal bovine serum and 1% penicillin/streptomycin (Invitrogen) at 37°C in a humidified atmosphere of 5% CO 2 in air.

| RNA extraction and qPCR analysis
RNeasy kits (Qiagen, Beijing, China) were used to extract total RNA. qPCR was performed as previously described. 14 36B4 was used as an internal control. Primer sequences for qPCR are shown in Table 1.

| Quantification of cell viability
MCF-7 and LCC2 cells were transfected with siSTAT1 or siControl in 24-well plates. After 24 hours, the cells were seeded into 96-well plates. Cell numbers were determined using WST-1 cell proliferation reagent as previously described. 15

| Luciferase assay
The luciferase activity was performed with the Dual-Luciferase Reporter kit (Promega, Madison, WI, USA) . The ERE luciferase reporter was transfected together with the Renilla plasmid into the cells. Luciferase activity was measured after 24 hours.

| Chromatin immunoprecipitation (ChIP) assay
ChIP assay was performed in our previous study. MCF -

| RNA sequence analysis
The global gene expression analysis was based on the RNA sequencing platform from Beijing Genomic Institute. The RNA sequence data are deposited in the Gene Expression Omnibus database (accession number GSE118774). The analysis was performed for differentially expressed genes (P < 0.01 and fold change >2) using Ingenuity Pathway Analysis.

| Statistics
Student's t test and Pearson correlation coefficient were used for comparisons. For multiple group comparison, ANOVA (Analysis of Variance) was used for comparisons. Tukey's test was used as the posthoc test after ANOVA text. P < 0.05 was considered to be significant.

| The ISGF3 components STAT1 and IRF9 are up-regulated in tamoxifen-resistant cells and correlate with poor tamoxifen treatment outcome
Firstly, the tamoxifen resistance of LCC2 cells compared to MCF-

| STAT1 depletion decreases ERα mRNA and protein levels in breast cancer cells
The role of STAT1/IRF9 in oestrogen signalling was assessed.
Depletion of STAT1 via two different siRNAs decreased ERα protein levels in the MCF-7 cells ( Figure 2D), but IRF9 depletion did not change ERα protein levels ( Figure 2C).

| STAT1 depletion inhibits breast cancer cell proliferation and sensitizes cells to the tamoxifen inhibition effect
In order to assess the effect of STAT1 on breast cancer cell proliferation, F I G U R E 2 STAT1 depletion decreases ERα protein levels in breast cancer cells. A, IRF9 and STAT1 mRNA level is elevated in LCC2 cells compared with MCF-7 cells. Total RNA was prepared and the expression of the endogenous IRF9 and STAT1 were determined by qPCR. ***P < 0.001 for target gene expression comparison. B, A subgroup of ERα target genes were increased in LCC2 cells compared with MCF-7 cells. Total RNA was prepared and the expression of the endogenous ERα target genes were determined by qPCR. C, Effect of IRF9 depletion induced by two different siRNA oligos. MCF-7 cells were transfected with siSTAT1 or siControl. After 48 hours, IRF9 and ERα protein levels were determined by western blot analysis. Actin was used as an internal control. The relative ERα protein levels were quantified with Image J. D, Effect of STAT1 depletion induced by two different siRNA oligos. MCF-7 cells were transfected with siSTAT1 or siControl. After 48 h, STAT1 and ERα protein levels were determined by western blot analysis. Actin was used as an internal control. The relative ERα protein levels were quantified with Image J. E, Effect of STAT1 depletion on ERα protein levels under vehicle or E2 treatment conditions. MCF-7 cells were transfected with siSTAT1 or siControl. After 48 hours, STAT1 and ERα protein levels were determined by western blot analysis. Actin was used as an internal control. The relative ERα protein levels were quantified with Image J. F, Effect of STAT1 depletion on ERα protein levels under vehicle or E2 treatment conditions. LCC2 cells were transfected with siSTAT1 or siControl. After 48 hours, STAT1 and ERα protein levels were determined by western blot analysis. Actin was used as an internal control. The relative ERα protein levels were quantified with Image J. G, Effect of STAT1 depletion on ERα protein levels under vehicle or E2 treatment conditions. T47D cells were transfected with siSTAT1 or siControl. After 48 h, STAT1 and ERα protein levels were determined by western blot analysis. Actin was used as an internal control. The relative ERα protein levels were quantified with Image J. H, STAT1 overexpression increases ERα mRNA level in MCF-7 cells. MCF-7 cells were seeded into 6-well plates. After 24 hours, 2 μg STAT1 plasmids were transfected into MCF-7 cells. After 48 hours, cells was harvested and ERα mRNA level was determined via QPCR HOU ET AL. F I G U R E 3 STAT1 depletion decreases ERα signalling in breast cancer cells. A, STAT1 depletion down-regulates ERα target genes in MCF-7 cells. MCF-7 cells were transfected with siSTAT1 or siControl. After 48 hours, cells were cultured in phenol red-free medium and treated with either ethanol or 10 nmol L −1 estradiol for 6 hours. Total RNA was prepared and the expression of the endogenous ERα target genes, PS2, GREB1 and PDZK1 were determined by qPCR. Results from three experiments are shown. ***P < 0.001 for target gene expression comparison. B, STAT1 depletion down-regulates ERα target genes in LCC2 cells. LCC2 cells were transfected with siSTAT1 or siControl. After 48 h, cells were cultured in phenol red-free medium and treated with either ethanol or 10 nmol L −1 estradiol for 6 h. Total RNA was prepared and the expression of the endogenous ERα target genes, PS2, GREB1 and PDZK1 were determined by qPCR. Results from three experiments are shown. ***P < 0.001 for target gene expression comparison. C, STAT1 depletion down-regulates ERα target genes in T47D cells. T47D cells were transfected with siSTAT1 or siControl. After 48 hours, cells were cultured in phenol red-free medium and treated with either ethanol or 10 nmol L −1 estradiol for 6 hours. Total RNA was prepared and the expression of the endogenous ERα target genes, PS2, GREB1 and PDZK1 were determined by qPCR. Results from three experiments are shown. ***P < 0.001 for target gene expression comparison. D, STAT1 depletion down-regulates ERα target genes in tamoxifen-treated condition in MCF-7 cells. MCF-7 cells were transfected with siSTAT1 or siControl. After 48 hours, cells were cultured in phenol red-free medium and treated with either ethanol or 1 μmol L −1 tamoxifen for 6 h. Total RNA was prepared and the expression of the endogenous ERα target genes, PS2, GREB1 and PDZK1 were determined by qPCR. Results from three experiments are shown. ***P < 0.001 for target gene expression comparison. E, STAT1 depletion down-regulates ERα target genes in tamoxifen-treated condition in LCC2 cells. LCC2 cells were transfected with siSTAT1 or siControl. After 48 h, cells were cultured in phenol red-free medium and treated with either ethanol or 1 μmol L −1 tamoxifen for 6 hours. Total RNA was prepared and the expression of the endogenous ERα target genes, PS2, GREB1 and PDZK1 were determined by qPCR. Results from three experiments are shown. ***P < 0.001 for target gene expression comparison. F, STAT1 depletion affects ERE-luciferase activity in MCF-7 cells. MCF-7 cells were transfected with siSTAT1 or siControl together with a ERE luciferase reporter plasmid. Cells were treated with 10 nmol L −1 estradiol or vehicle. Luciferase activity was measured 48 hours after transfection. Results from three experiments are shown. ***P < 0.001 for luciferase activity comparison. G, STAT1 depletion affects ERE-luciferase activity in LCC2 cells. LCC2 cells were transfected with siSTAT1 or siControl together with a ERE luciferase reporter plasmid. Cells were treated with 10 nmol L −1 estradiol or vehicle. Luciferase activity was measured 48 hours after transfection. Results from three experiments are shown. ***P < 0.001 for luciferase activity comparison. H, STAT1 depletion affects EREluciferase activity in T47D cells. T47D cells were transfected with siSTAT1 or siControl together with a ERE luciferase reporter plasmid. Cells were treated with 10 nmol L −1 estradiol or vehicle. Luciferase activity was measured 48 hours after transfection. Results from three experiments are shown. ***P < 0.001 for luciferase activity comparison F I G U R E 4 STAT1 depletion inhibits breast cancer cell proliferation and sensitizes cells to the tamoxifen inhibition effect. A, WST-1 assay was used to determine the cellular metabolic activity at the indicated time points after transfection. MCF-7 cells were transfected with siSTAT1 and siControl. After 24 hours, cells were seeded into 96-well plates. These experiments were performed in triplicates. All values are presented as the mean ± standard deviation (n = 3, ***P < 0.001). B, WST-1 assay was used to determine the cellular metabolic activity at the indicated time points after transfection. LCC2 cells were transfected with siSTAT1 and siControl. After 24 hours, cells were seeded into 96well plates. These experiments were performed in triplicates. All values are presented as the mean ± standard deviation (n = 3, ***P < 0.001). C, WST-1 assay was used to determine the cellular metabolic activity at the indicated time points after transfection. MCF-7 cells were transfected with siSTAT1 and siControl. After 24 hours, cells were seeded into 96-well plates treated with vehicle or 1 μmol L −1 tamoxifen. These experiments were performed in triplicates. All values are presented as the mean ± standard deviation (n = 3, ***P < 0.001). D, WST-1 assay was used to determine the cellular metabolic activity at the indicated time points after transfection. LCC2 cells were transfected with siSTAT1 and siControl. After 24 hours, cells were seeded into 96-well plates treated with vehicle or 1 μmol L −1 tamoxifen. These experiments were performed in triplicates. All values are presented as the mean ± standard deviation (n = 3, ***P < 0.001). E, STAT1 depletion sensitized MCF-7 cells to the tamoxifen inhibition effect. MCF-7 cells were transfected with siSTAT1 and siControl. After 24 hours, cells were seeded into 96-well plates. Cells were treated with the indicated concentration of tamoxifen for 48 hours and the number of cells was quantified using a WST-1 assay. These experiments were performed in triplicates. All values are presented as the mean ± standard deviation (n = 3, ***P < 0.001). F, STAT1 depletion sensitized LCC2 cells to the tamoxifen inhibition effect. LCC2 cells were transfected with siSTAT1 and siControl. After 24 hours, cells were seeded into 96-well plates. Cells were treated with the indicated concentration of tamoxifen for 48 hours and the number of cells was quantified using a WST-1 assay. These experiments were performed in triplicates. All values are presented as the mean ± standard deviation (n = 3, ***P < 0.

| Reduction of STAT1 levels reduces recruitment of STAT1 to the ERα promoter, which is a potential mechanism for ERα signalling regulation
ChIP assays were conducted in order to detect the possible association between STAT1 and ERα (data not shown). The IP assay using MCF-7/LCC2 cells did not indicate the association between STAT1 and ERα. As ERα mRNA levels were also markedly decreased (Figure Figure 5B). 16 ChIP assay was performed in order to detect STAT1 binding to ERα promoter regions. As ERα has been shown to bind to its own gene promoter regions, ERα antibody-based ChIP was used as the positive control. The ChIP assay showed that STAT1 binds to the ERα promoter E2 but not to promoter A, while ERα binds to all three promoters ( Figure 5C). Transfection with siRNA targeting STAT1 resulted in significantly decreased levels of binding at promoter E2 ( Figure 5D). However,   Figure S2A and S2B). Coupled with the data that show that STAT1 depletion significantly decreases ERα mRNA levels, it indicates that STAT1 binding to the ERα promoter region is a potential mechanism by which STAT1 facilitates ERα transcription and ERα signalling ( Figure 6).

| DISCUSSION
The JAK-STAT signalling pathway consists of three main components: the cell surface receptor, JAK and STAT proteins. 17 Among the transcription factors, STATs are the major effectors in regulation of target gene expression. For example, the ISGF3 complex, which consists of STAT1-STAT2-IRF9 proteins, binds to specific nucleotide sequences and is activated by interferon signalling. 18 Previous studies showed that STAT1 plays both oncogenic and tumour-suppression roles in various types of cancer, which may depend on the cancer cell background. 19,20 For example, STAT1 promoted oesophageal cancer invasion in the presence of p53 mutation, 21 while STAT1 also induced cell cycle inhibition via interaction with cyclin D1 and CDKs. 22 In breast cancer, STAT1 signalling correlates with poor endocrine treatment outcome, while the molecular mechanism is not clear. 23 In the present study, STAT1 was found to be elevated in human breast cancer compared to normal breast tissues using a publicly available database. STAT1 is necessary to maintain ERα signalling in breast cancer cells, probably by regulating ERα gene expression. The present study offers a possible mechanism by which the JAK-STAT pathway component STAT1 is involved in regulating oestrogen signalling activity and modulating tamoxifen sensitivity in breast cancer cells.
Since the development of endocrine therapy, tamoxifen has been used to treat patients with breast cancer for more than 40 years. This has resulted in a marked reduction in the mortality rate and remains one of the most effective treatments against breast cancer. Intensive research has been conducted in the past decades in order to investigate the underlying mechanism of endocrine resistance. In addition to the hyper-activation of ERα signalling, either by a mutation for ERα constitutive activation or elevated ERα co-activators, 5,9 the crosstalk between ERα signalling and other pathways also plays an important role in mediating tamoxifen resistance. A previous study found that MCF-7 cells transfected with HER2 acquired tamoxifen resistance in xenograft mice models. 24 Further studies have shown that ERα interacts with several other signalling pathways, including the HER2, EGFR and NFKB pathways. 25,26 In the present study, novel crosstalk between ERα signalling and the JAK-STAT pathway was identified. As an important transcription factor, STAT1 may not only mediate JAK-STAT activation, but also transactivate oestrogen signalling via modulation of ERα gene expression.
Along with the extensive studies of tamoxifen resistance in breast cancer, a number of tamoxifen-resistant breast cancer cell lines have been derived, with the majority of which from tamoxifen-sensitive MCF-7 cells. 12,27,28 Among them, LCC2 cells are the most frequently used tool for investigating the mechanism of tamoxifen resistance. 12 In the present study, whole genomic expression profiles were compared between LCC2 and MCF-7 cells. The pathway enrichment analysis showed higher expression levels of JAK-STAT components, including STAT1. The data indicate that STAT1 is an important component in the regulation of ERα transcription in ERα -positive cancer cells. As modulation of ERα levels is one feasible approach to target oestrogen signalling and cell proliferation, STAT1 is a potential drug target for ERα -positive breast cancers.

ACKNOWLEDG EMENTS
We thank all the members of Henan Key Laboratory of Immunology and Targeted Drugs for sharing valuable material and research support.

AUTHORS' CONTRI BUTIONS
YXH, XL and QHL contributed to the manuscript writing. YXH, XL, QHL, HJY, MX, KM, GSW, XML and SZ contributed to the molecular and cellular biology experiments. JTX and GU contributed to the clinical data analysis and RNA sequence data analysis. TZ and HW contributed to scientific design, manuscript revising and the funding support for this study.

AVAILABILITY OF DATA AND MATERIALS
Additional data and materials may be requested from the corresponding author on reasonable request.