Aurora-A Mitotic Kinase Induces Endocrine Resistance through Down-Regulation of ERα Expression in Initially ERα+ Breast Cancer Cells

Development of endocrine resistance during tumor progression represents a major challenge in the management of estrogen receptor alpha (ERα) positive breast tumors and is an area under intense investigation. Although the underlying mechanisms are still poorly understood, many studies point towards the ‘cross-talk’ between ERα and MAPK signaling pathways as a key oncogenic axis responsible for the development of estrogen-independent growth of breast cancer cells that are initially ERα+ and hormone sensitive. In this study we employed a metastatic breast cancer xenograft model harboring constitutive activation of Raf-1 oncogenic signaling to investigate the mechanistic linkage between aberrant MAPK activity and development of endocrine resistance through abrogation of the ERα signaling axis. We demonstrate for the first time the causal role of the Aurora-A mitotic kinase in the development of endocrine resistance through activation of SMAD5 nuclear signaling and down-regulation of ERα expression in initially ERα+ breast cancer cells. This contribution is highly significant for the treatment of endocrine refractory breast carcinomas, because it may lead to the development of novel molecular therapies targeting the Aurora-A/SMAD5 oncogenic axis. We postulate such therapy to result in the selective eradication of endocrine resistant ERαlow/− cancer cells from the bulk tumor with consequent benefits for breast cancer patients.


Introduction
Approximately, 70% of human breast carcinomas fall into luminal subtypes, which are estrogen receptor alpha (ERa) positive [1]. ERa expression correlates with expression of progesterone receptor (PR), lower tumor grade, response to endocrine therapy, lower grade of aneuploidy, less frequent overexpression of HER-2 oncogene, bone metastases and slower rate of tumor recurrence [2]. Despite the clinical benefit of hormonal treatment in patients with ERa+ breast cancer, resistance to first and second-line endocrine therapy remains a major clinical problem [3,4]. The introduction of pure estrogen antagonists such as fulvestrant, to overcome the apparent disadvantage of tamoxifen with its partial agonist properties, did not resolve the endocrine resistance problem [5]. Second-line therapy with other endocrine agents such as aromatase inhibitors produces some beneficial effect but for the most part serves merely to delay onset of endocrine resistance [6]. In pre-clinical and clinical studies, development of endocrine resistance is associated with an aggressive behavior characterized by high frequency of distant metastases and poses a significant problem that affects negatively the disease-free and overall survival of breast cancer patients [7]. Response to one form of endocrine therapy after resistance to a previous therapy is a historically recognized observation that is the key to management of patients with metastatic disease [8]. Importantly, subsequent responses to serial endocrine therapy tend to be shorter, indicating a gradual shift from ''dependence on ERa signaling to alternative escape oncogenic pathways'' [9]. Several mechanisms of endocrine resistance have been proposed including: ERa mutations, altered expression of ERa coregulators, ligand-independent activation of ERa by growth factor receptor kinases and down-regulation/loss of ERa expression [3]. Importantly, whereas ''acquired resistance'' is predominantly the consequence of estrogen-independent activation of ERa in cancer cells still harboring an intact ERa signaling axis, ''intrinsic resistance'' is mostly likely due to down-regulation/loss of ERa expression leading to endocrine panresistance and tumor progression and represents a serious challenge for the treatment of ERa+ breast cancer patients [3]. For this reason, understanding the oncogenic pathways responsible for the development of resistance to endocrine therapy given before or after primary surgery is imperative to develop innovative therapeutic strategies aimed to suppress or at least delay recurrence and progression of initially ERa+ breast tumors.
The discovery that breast carcinomas contain a sub-population of cells harboring stem-like properties (cancer initiating cells) has generated excitement because these cancer initiating cells may represent the source of therapeutic failures, tumor recurrence and poor clinical outcome [10]. It has also been demonstrated that cancer cells that undergo through epithelial to mesenchymal transition (EMT) acquire a basal CD44 + /CD24 low/2 cancer stemlike phenotype with increased capacity for self-renewal, invasion, drug resistance, and tumor progression [11]. Moreover, breast cancer initiating cells display down-regulation of ERa and clonal expansion of these CD44 + /CD24 low/2 /ERa low/2 cancer cells may be responsible for tumor recurrence and development of distant metastases of initially ERa+ breast carcinomas [12,13]. Aberrant activation of HER-2/MAPK and TGFb/SMAD oncogenic signalings induces EMT and plays an important role in the maintenance of breast cancer initiating cells [14][15][16]. However, the underlying molecular mechanisms responsible for ERa downregulation by aberrant activation of HER-2/MAPK and TGFb/ SMAD pathways remain elusive and are the subject of ongoing investigation.
The mitotic kinase Aurora-A plays a key role in breast cancer progression through the development of centrosome amplification and chromosomal instability (CIN) [17][18][19]. Aurora-A is overexpressed in human breast tumors and is associated with an invasive ERa low/basal-like phenotype and poor-prognosis [20,21]. Moreover, it has been demonstrated that estrogen is causally linked via ERa to Aurora-A overexpression, centrosome amplification, CIN, and aneuploidy leading to breast tumors in susceptible mammary gland cells [22]. Nonetheless, the causal role of aberrant Aurora-A kinase activity in the development of endocrine resistance and breast cancer progression through molecular mechanisms that are independent from its mitotic function and CIN remains elusive. Herein we demonstrate a significant and novel non-mitotic role of Aurora-A kinase in the induction of tumor progression of ERa+ breast cancer xenografts through activation of EMT and the genesis of CD44 + /CD24 low/2 cancer initiating cells [23]. Moreover, these studies revealed a non canonical cross-talk between Aurora-A kinase and SMAD5 oncogenic signaling in promoting EMT and invasiveness. We demonstrate for the first time the causal role of Aurora-A/SMAD5 oncogenic axis in the development of endocrine resistance through down-regulation of ERa expression in initially ERa+ breast cancer cells. This contribution is important for the treatment of endocrine refractory breast carcinomas, because it may lead to the development of novel molecular therapies targeting the Aurora-A/SMAD5 oncogenic axis. We postulate such therapy to result in selective eradication of endocrine resistant ERa low/2 cancer cells from the bulk tumor thereby delaying tumor progression with consequent benefits on the progression-free and overall survival of breast cancer patients.

Results and Discussion
Because development of endocrine resistance and progression of ERa+ breast tumors is frequently characterized by aberrant activation of MAPK signaling [24], we employed ERa+ MCF-7 cells over-expressing a constitutive active Raf-1 oncoprotein (vMCF-7 DRaf-1 ) as previously described [23,25]. In our previous studies we have showed that vMCF-7 DRaf-1 cells display MAPK hyper-phosphorylation compared to parental MCF-7 cells, demonstrating a constitutive activation of Raf/MAPK oncogenic signaling. Importantly, constitutive activation of Raf/MAPK signaling conferred higher migratory properties of MCF-7 cells in vitro, predictive of an invasive phenotype that was validated in vivo through the development of distant metastases in tumor xenograft models. To investigate the extent to which metastatic lesions derived from vMCF-7 DRaf-1 xenografts displayed ERa down-regulation, we established murine MCF-7 and vMCF-7 DRaf-1 xenografts. Tumor xenografts were surgically removed 12 weeks after implantation without sacrificing the animals to monitor the development of distant metastases as previously described [25]. As expected, 8 weeks following surgical removal, only vMCF-7 DRaf-1 xenografts developed frank distant metastases (lung and spleen). Importantly, vMCF-7 DRaf-1 metastatic lesions showed ERa downregulation resulting in ERa +/2 cell heterogeneity compared to MCF-7 and vMCF-7 DRaf-1 primary tumors ( Figure 1A). These findings indicate that ERa low/2 cancer cells display more invasive properties over ERa+ cancer cells in vivo and their clonal expansion may induce tumor progression. To investigate whether vMCF-7 DRaf-1 primary tumors carried a singular sub-population of cancer cells harboring an ERa low/phenotype that was mostly observed in the metastatic lesions described above, we re-cultured cells from primary vMCF-7 DRaf-1 tumor xenografts (referred to as first generation derived from xenografts, 1GX). Significantly, vMCF-7 DRaf-1 1GX cells showed down-regulation of ERa expression due to loss of ERa in ,28% of bulk cancer cells ( Figure 1B-C). These findings demonstrate that cancer cells harboring an ERa low/2 phenotype were already present in vMCF-7 DRaf-1 primary tumors and their clonal expansion may promote the onset of distant metastases during tumor progression. Next we investigated whether down-regulation of ERa expression was causally linked to development of endocrine resistance in vMCF-7 DRaf-1 1GX cells. Parental MCF-7 and variant cells were treated in vitro with 17-b Estradiol alone or in combination with the anti-estrogen 4-OHtamoxifen and endocrine sensitivity was determined by analyzing the percentage of cancer cells in the S phase of the cell cycle. vMCF-7 DRaf-1 1GX cells displayed the highest resistance to 4-OH tamoxifen compared to parental MCF-7 and vMCF-7 DRaf-1 cells indicating that down-regulation of ERa induces ''intrinsic resistance'' to conventional endocrine therapy ( Figure 1D). Since we have previously shown that tumor progression of vMCF-7 DRaf-1 1GX xenografts is causally linked to aberrant Aurora-A kinase activity [23], we tested Aurora-A expression in MCF-7 and variant cells. As demonstrated before, Aurora-A was over-expressed in vMCF-7 DRaf-1 1GX cells compared to parental cells (Figure 2A). Next we investigated the causal role of Aurora-A kinase activity in the development of endocrine resistance by employing alisertib, a novel Aurora-A kinase small molecule inhibitor currently being tested in oncology clinical trials [26]. We have previously demonstrated that treatment of breast cancer cells with 1 mM alisertib selectively inhibits Aurora-A kinase activity [23]. Specifically, our study showed that the transcriptome profile of breast cancer cells treated with alisertib or shRNA targeting Aurora-A displayed ,90% gene expression overlapping demonstrating the high specificity of 1 mM alisertib for targeting Aurora-A kinase activity. vMCF-7 Raf1 1GX cells that displayed strong 4-OH tamoxifen resistance were treated with fulvestrant (a selective ERa down-regulator that increases ERa degradation and inhibits estrogen signaling) alone or in combination with alisertib. Combination of fulvestrant with alisertib induced a stronger effect on inhibition of cell proliferation measured by Real-Time Cell Proliferation Assay ( Figure 2B). Based on our previous results viewing a novel cross-talk between Aurora-A kinase and SMAD5 oncogenic signaling in the development of EMT and breast cancer progression [23], we also analyzed SMAD5 nuclear phosphorylation in vMCF-7 Raf1 1GX cells treated with fulvestrant and/or alisertib. Importantly, restoration of endocrine sensitivity was mechanistically linked to suppression of SMAD5 nuclear phosphorylation ( Figure 2C), demonstrating the causal role of Aurora-A/SMAD5 oncogenic axis in the development of endocrine resistance in initially ERa+ and hormone sensitive breast cancer cells. Significantly, these findings were also validated by performing immunoblotting analysis of vMCF-7 Raf1 1GX cells treated with fulvestrant and/or alisertib showing a selective alisertib-induced down-regulation of SMAD5 phosphorylation ( Figure 2D). Next we wanted to investigate the causal role of Aurora-A/SMAD5 oncogenic axis in the development of ''intrinsic resistance'' to conventional endocrine therapy through down-regulation of ERa expression in initially ERa+vMCF-7 Raf1 cancer cells. We employed a lenti-vector engineered to over-express Aurora-A kinase and we analyzed Aurora-A phosphorylation/activation, ERa and p,SMAD5 nuclear localization in vMCF-7 Raf-1 and vMCF-7 Raf-1/Aurora-A cells ( Figure 3A-B). Our studies showed that aberrant Aurora-A kinase activity induced down-regulation of ERa nuclear localization that was functionally linked to increased SMAD5 nuclear phosphorylation ( Figure 3B-C). Importantly, treatment of vMCF-7 Raf-1/Aurora-A cells with alisertib inhibited Aurora-A phosphorylation, restored ERa nuclear expression and suppressed SMAD5 nuclear phosphorylation ( Figure 3B-C). Moreover, the causal role of Aurora-A kinase in the down-regulation of ERa nuclear expression through activation of SMAD5 was validated by employing a dominant negative (DN) Aurora-A construct that abrogated Aurora-A kinase activity ( Figure 3B-C). Significantly, over-expression of Aurora-A in parental MCF-7 cells resulted in a similar phenotype, although vMCF-7 Raf-1/Aurora-A cells displayed a stronger phosphorylation of nuclear SMAD5 and down-regulation of ERa expression, demonstrating a synergistic cross-talk between MAPK pathway and aberrant Aurora-A kinase activity in the abrogation of ERa signaling ( Figure S1). Finally, to investigate the role of SMAD5 as a down-stream target of Aurora-A-induced ERa down-regulation, we engineered MCF-7 over-expressing SMAD5 ( Figure 4A). Notably, over-expression of SMAD5 in MCF-7 cells induced down-regulation of ERa nuclear localization demonstrating the causal role of SMAD5 nuclear signaling in the development of breast cancer cells harboring an ERa low/2 phenotype ( Figure 4B-C). Taken together, these results indicate that the partial response of endocrine resistant vMCF-7 Raf1 1GX cancer cells to fulvestrant is likely due to its effect on the ERa+ sub-population that still retains a functional ERa signaling. Inhibition of Aurora-A kinase activity by alisertib impairs SMAD5 nuclear activation and restores ERa expression in the subpopulation of ERa low/2 cancer cells leading to restoration of endocrine sensitivity. Furthermore, our findings are supported by recent studies that have demonstrated that downregulation of ERa expression is responsible for endocrine therapeutic failure and tumor progression of initially ERa+ breast tumors [27]. Although it has already been demonstrated that key transcription factors involved in the development of EMT such as Snail and Slug induce down-regulation of ERa expression in breast cancer cells [28,29], our findings establish a previously unrecognized oncogenic signaling involved in the abrogation of ERa function through activation of the Aurora-A/SMAD5 oncogenic axis. Moreover, vMCF-7 DRaf-1 breast cancer xenografts represent an innovative and useful pre-clinical model to investigate the mechanisms leading to endocrine panresistance and tumor progression based on the clonal expansion of ERa low/2 cancer cells. Importantly, although these studies are based on the vMCF- Based on these findings and our published data [23,25], we propose a novel model of endocrine resistance in initially ERa+ and hormone sensitive breast cancer cells: aberrant activation of MAPK signaling promotes phosphorylation and stabilization of Aurora-A kinase that in turn induces down-regulation/loss of ERa expression through activation of SMAD5 nuclear signaling leading to endocrine resistance and tumor progression ( Figure 5). Because our previous studies have demonstrated the causal role of aberrant Aurora-A kinase activity in the development of CD44 + /CD24 low/2 cancer initiating cells [23], we speculate that Aurora-A/SMAD5 oncogenic axis may induce ERa down-regulation by promoting the clonal expansion of CD44 + /CD24 low/cancer initiating cells that display low levels of ERa. Moreover, because Aurora-A is downstream of MAPK, we believe that molecular targeting of Aurora-A in endocrine resistant breast cancer cells will be more effective than targeting MAPK pathway due to Aurora-A direct effects on the clonal expansion of CD44 + /CD24 low/2 /ERa low/2 cancer initiating cells [23]. However, we don't exclude that patients with endocrine resistant breast tumors could benefit from the combination of MAPK and Aurora-A small molecule inhibitors.
In conclusion, our studies are of important translational relevance because they will lay the groundwork for innovative clinical trials employing small molecule inhibitors of Aurora-A kinase to suppress the Aurora-A/SMAD5 oncogenic axis, restore ERa expression and sensitivity to endocrine therapy for a subset of hormone refractory breast tumors with anticipated benefits on the progression-free and overall survival of breast cancer patients.

Human Breast Cancer Cell Lines
The human breast cancer cell line MCF-7 was obtained from ATCC (Manassas, VA, USA). The MCF-7 cells over-expressing the Raf-1 oncoprotein were generated as previously described [23,25]. All cell lines were maintained in EMEM medium containing 5 mM glutamine, 1% penicillin/streptomycin, 20 microgram insulin/ml and 10% FBS at 37C in 5% CO2 atmosphere.

Tumor Xenografts and Immunohistochemistry
Procedures established by the Institutional Animal Care and Use Committee based on US NIH guidelines for the care and use of laboratory animals were followed for all experiments. Fourweek-old non-ovariectomized female NCR/Nu/Nu nude mice were anesthetized by exposure to 3% isoflurane and five mice per each group were injected subcutaneously with 2610 6 MCF-7 or vMCF-7 DRaf-1 cancer cells suspended in 50 ml of 50% Matrigel (BD Bioscience, Bedford, MA, USA). Tumor localization and growth was monitored using the IVS imaging system from the ventral view 10 min after luciferin injection. After 12 weeks, mice were killed and xenograft tumors were processed for histology, and immunohistochemistry analyses. Paraffin-embedded tumor tissues were stained with ERa antibody (Abcam, Cambridge, Massachusetts, USA) as previously described [23]. To re-establish cultures from 1GX explants, primary tumors tissues were excised from killed animals, minced using sterile scissors, transferred to complete culture medium and fibroblast-free tumor cell lines were established by serial passages in culture. Animals were examined everyday and body weight and primary tumor size were measured at least 1-2 times per week. Consistent distress and potential pain (.1 day) were alleviated by euthanasia. If some of the animals were loosing greater than 10% of their body weight, if blood was consistently observed in the urine or around the genitals of the mice, the mice were appropriately euthanized. When typical signs of distress including labored breathing and inactivity were consistently observed for .1 day, the animals were appropriately euthanized. When the primary tumor was .2 cm, the animals were sacrificed. Animals were euthanized using Pentobarbital (IP 100 mg/kg) followed by cervical dislocation. The Mayo Clinic Institutional Animal Care and Use Committee (IACUC) approved this study.

Fluorescence Microscopy
Cells were fixed in absolute methanol at 220uC for 10 min, blocked in 5% normal goat serum, 1% glycerol, 0.1% BSA, 0.1% fish skin gelatin, 0.04% sodium azide and incubated with primary antibodies. Primary antibodies against the proteins ERa (Santa Cruz Biotechnology, Delaware Avenue Santa Cruz, CA, USA) and P,SMAD5 (Cell Signaling Technology, Boston, MA, USA) were followed by secondary antibodies conjugated with Alexa 488 or Alexa 568 (Molecular Probes, Eugene, OR, USA). Images were digitally recorded at multiple focal planes using a Zeiss Axiovert 200 M fluorescence microscope and analyzed as maximum projections. Results are derived from three independent experiments. Nuclear staining counting was performed employing the ImageJ Software, experiments were performed in triplicate.

Endocrine Resistance Studies
For endocrine resistance studies, breast cancer cells were cultured in phenol-red free medium with 5% FBS for 48 hours. Following starvation from 17-b estradiol, cells were treated with 17-b estradiol (10 210 M) alone or in combination with 4-OHtamoxifen (10 27 M) for 48 hours. Results are derived from three independent experiments (+/2 s.d.). Cell cycle profile was performed by FACS as earlier described [23]. For In Vitro real time cell proliferation assay we employed the xCELLigence technology (ACEA). Following 48 hours starvation from 17-b estradiol, cells were treated with Fulvestrant (50 nM) and/or Alisertib (50 nM). Results are derived from three independent experiments.

Immunoblot and Lenti-Vector Expression Studies
Immunoblot and lenti-vector expression studies were performed as previously described [23]. Antibodies employed for the immunoblot analysis were the followings: Aurora-A and P,SMAD5 (Cell Signaling Technology, Boston, MA, USA). The Dominant Negative (DN) Aurora-A vector was kindly provided by Dr. Lomberk (Mayo Clinic, Rochester, MN).