HIF-1α stimulates aromatase expression driven by prostaglandin E2 in breast adipose stroma

Abstract

Introduction: The majority of postmenopausal breast cancers are estrogen-dependent. Tumor-derived factors, such as prostaglandin E2 (PGE2), stimulate CREB1 binding to cAMP response elements (CREs) on aromatase promoter II (PII), leading to the increased expression of aromatase and biosynthesis of estrogens within human breast adipose stromal cells (ASCs). Hypoxia inducible factor-1α (HIF-1α), a key mediator of cellular adaptation to low oxygen levels, is emerging as a novel prognostic marker in breast cancer. We have identified the presence of a consensus HIF-1α binding motif overlapping with the proximal CRE of aromatase PII. However, the regulation of aromatase expression by HIF-1α in breast cancer has not been characterized. This study aimed to characterize the role of HIF-1α in the activation of aromatase PII.

Methods: HIF-1α expression and localization were examined in human breast ASCs using quantitative PCR (QPCR), Western blotting, immunofluorescence and high content screening. QPCR and tritiated water-release assays were performed to assess the effect of HIF-1α on aromatase expression and activity. Reporter assays and chromatin immunoprecipitation (ChIP) were performed to assess the effect of HIF-1α on PII activity and binding. Treatments included PGE2 or DMOG ((dimethyloxalglycine), HIF-1α stabilizer). Double immunohistochemistry for HIF-1α and aromatase was performed on tissues obtained from breast cancer and cancer-free patients.

Results: Results indicate that PGE2 increases HIF-1α transcript and protein expression, nuclear localization and binding to aromatase PII in human breast ASCs. Results also demonstrate that HIF-1α significantly increases PII activity, and aromatase transcript expression and activity, in the presence of DMOG and/or PGE2, and that HIF-1α and CREB1 act co-operatively on PII. There is a significant increase in HIF-1α positive ASCs in breast cancer patients compared to cancer-free women, and a positive association between HIF-1α and aromatase expression.

Conclusions: This study is the first to identify HIF-1α as a modulator of PII-driven aromatase expression in human breast tumor-associated stroma and provides a novel mechanism for estrogen regulation in obesity-related, post-menopausal breast cancer. Together with our on-going studies on the role of AMP-activated protein kinase (AMPK) in the regulation of breast aromatase, this work provides another link between disregulated metabolism and breast cancer.

Figure 1

Effect of PGE₂ on HIF-1α expression and nuclear localization in primary human breast ASCs. PGE₂ caused a significant increase in HIF-1α transcript (A) and nuclear protein expression (B). Confocal microscopy demonstrated that HIF-1α (green) is mainly perinuclear in breast ASCs under basal conditions (C, top left) and that PGE₂ stimulates the translocation of HIF-1α to the nucleus (C, top right). Treatment with DMOG (C, bottom left) and DMOG with PGE2 (C, bottom right) caused a much higher HIF-1α staining in the nucleus. The merged lamin B1+B2 nuclear stain (red) and HIF-1α are found as insets at the bottom right of each image. The percentage of cells positive for nuclear HIF-1α was also shown to be significantly increased with PGE₂ treatment (D). vc = vehicle control, n = 3, repeated twice. Confocal images are representative of the majority of cells examined.

Figure 2

Role of HIF-1α in aromatase regulation. (A) ChIP demonstrated that PGE₂ and DMOG significantly stimulate the endogenous binding of HIF-1α to aromatase PII. Treatment of ASCs with PGE₂ or DMOG caused a significant increase in aromatase transcript expression (B) and activity (C). (D) HIF-1α overexpression in MCF-7 cells significantly increased aromatase activity in the presence of PGE₂. vc = vehicle control, n = 3, repeated twice.

Figure 3

HIF-1α is necessary and acts cooperatively with CREB1 to induce PII activity in response to PGE₂. (A) A putative HRE (italic) was found to overlap with the proximal CRE (boxed) of aromatase PII. (B) Reporter assays demonstrated that mutation of the proximal CRE of aromatase PII inhibited the HIF-1α/DMOG-mediated effect on PII activity. (C) PII activity was significantly increased in HIF-1α transfected cells, treated with DMOG and PGE₂, and co-transfection with CREB1 resulted in a further increase in PII activity compared to cells transfected with either HIF-1α or CREB1 alone. Knockdown of HIF-1α (E) using siRNA significantly reduced PII activity (D) and aromatase transcript expression (F) and suppressed the PGE₂-mediated effect on PII activity (D). β-gal, β-galactosidase activity; RLU, relative luciferase units. N = 3, all experiments repeated twice.

Figure 4

HIF-1α and aromatase expression in ASCs from tumor-bearing breast tissue compared to cancer-free using immunohistochemistry. (A) Percentage of HIF-1α positive ASCs from tumor-bearing breast tissue was significantly increased compared to tissue from cancer-free women. (B) Double-positive and single-positive ASCs for HIF-1α and aromatase were shown to be significantly increased in tumor-bearing compared to cancer-free tissues. Double negative ASCs for HIF-1α and aromatase were significantly reduced in tumor-bearing compared to cancer-free tissues. n = 10 for tumor-bearing breast tissue; n = 10/cancer-free breast tissue.

Figure 5

Model of the PGE₂-mediated regulation of aromatase expression by HIF-1α in breast ASCs. Tumor-derived factor PGE₂ increases HIF-1α transcript expression and nuclear localization. HIF-1α dimerizes with HIF-1β and then translocates to nucleus where it interacts with the proximal CRE of aromatase PII. HIF-1α together with CREB, CRTCs, CBP and p300 act to increase the PII-driven expression of aromatase.