Background
The selective estrogen-receptor α (ER) modulator, tamoxifen, is the recommended first-line adjuvant endocrine therapy for premenopausal women with ER-positive breast cancer, whereas postmenopausal women with ER-positive breast cancer will be offered an aromatase inhibitor. Although many patients benefit from the treatment,
de novo or acquired resistance occurs in approximately 30% of the patients, and is therefore a major clinical challenge [
1,
2]. Following relapse, many patients will benefit from treatment with the pure antiestrogen fulvestrant, a selective ER down regulator, which induces degradation of ER upon binding and subsequently abolishes ER signaling [
3,
4]. However, in spite of initial response, almost all patients with advanced disease eventually develop resistance against antiestrogen therapy [
1,
3,
5-
7].
Cell model systems are valuable tools to study the molecular mechanisms for endocrine resistant breast cancer. We have developed
in vitro cell culture models based on the ER-positive and estrogen responsive human breast cancer cell lines MCF-7 and T47D [
8-
11]. In line with other studies, we have shown that growth of breast cancer cell lines can switch from being ER-driven to being mediated by the HER receptors upon acquisition of resistance [
12-
18]. HER2 gene amplification or protein over expression in breast cancer is associated with a significantly shorter time to relapse, poor survival and reduced sensitivity to endocrine therapy [
19-
21]. We have previously shown that the expression of HER2 was increased in the T47D-derived fulvestrant resistant cell lines compared with the parental antiestrogen sensitive T47D breast cancer cells. However, resistant cell growth was not preferentially inhibited by knockdown of HER2 or by inhibition of HER receptor activity [
11]. These findings indicate that HER signaling presumably does not account for all cases of breast cancer resistance, emphasizing the need for continued investigations of the resistance mechanisms.
Tumor expansion depends on continued growth of tumor cells through mitotic cell division. A key mitotic regulator is the chromosomal passenger complex (CPC), composed of the catalytic component Aurora kinase B and the three regulatory and targeting components; inner centromere protein (INCENP), survivin and borealin. CPC is important for chromosome condensation, correction of erroneous kinetochore-microtubule attachments, activation of the spindle-assembly checkpoint and cytokinesis [
22]. The function of Aurora kinase B is linked to chromatin modification in relation to phosphorylation of histone H3 at Ser10 [
23]. The expression of Aurora kinase B is cell cycle regulated and the kinase is activated upon binding to INCENP, which is both a substrate and a positive regulator of Aurora kinase B [
24,
25]. Over expression of Aurora kinase B is evident in a range of primary cancers, such as prostate, head and neck, colon and thyroid cancers, and is associated with clinical aggressiveness [
26,
27].
To explore the molecular mechanisms driving antiestrogen resistant cell growth, we have utilized a large kinase inhibitor library comprising 195 kinase inhibitors on parental and fulvestrant resistant T47D breast cancer cell lines. We identified Aurora kinase B as a putative novel therapeutic target in fulvestrant and tamoxifen resistant breast cancer cells, and further explored its role in signaling and growth of fulvestrant resistant T47D cell lines by using the selective Aurora kinase B inhibitors, barasertib and hesperadin. Moreover, we investigated the clinical relevance of Aurora kinase B expression in primary tumors from breast cancer patients receiving tamoxifen as first-line adjuvant endocrine therapy.
Methods
Cell lines, culture condition and reagents
The T47D cell line was originally obtained from the Human Cell Culture Bank (Mason Research Institute, Rockville, MD, USA) and maintained as previously described [
11]. The fulvestrant resistant cell lines; T47D/182
R-1 (182
R-1) and T47D/182
R-2 (182
R-2) were established from T47D grown with 5% fetal calf serum (FCS) and long term treated with 100 nM fulvestrant (Tocris, Avonmouth, Bristol, UK) as described in [
11]. To enable ER-mediated growth inhibition by tamoxifen, the T47D cell line was adapted to grow in medium (RPMI, 8 μg/ml insulin and 2 mM glutamax) supplemented with only 2% FCS (T47D/S2). This cell line was used for establishment of the tamoxifen resistant cell lines T47D/TR-1 (TR-1) and T47D/TR-2 (TR-2) [
28]. The fulvestrant and tamoxifen resistant cell lines were maintained in the same growth medium as their parental T47D cell lines plus 100 nM fulvestrant or 1 μM tamoxifen (Sigma-Aldrich, St. Louis, MO, USA), respectively. For experiments, 2.5 × 10
5 U penicillin and 31.25 μg/l streptomycin (Gibco, Life Technologies, Carlsbad, CA, USA) were added to the growth medium. Barasertib was purchased from Selleck Chemicals (Houston, TX, USA). Stock solutions of 10
−3 M fulvestrant were dissolved in 96% ethanol, whereas stock solutions of 10 mM barasertib were dissolved in dimethyl sulfoxide (DMSO).
Kinase inhibitor screen
The kinase inhibitor library comprising 195 different kinase inhibitors was purchased from Selleck Chemicals and the experiment was performed as previously described [
28]. In brief, cells were seeded in triplicate in 96-well plates in their standard growth medium and allowed to adhere for 2 days before 5 days treatment with 1 μM of the kinase inhibitors. Vehicle-treated (0.1% DMSO) controls (6–10 wells/plate) were included in each plate. Cell viability was assayed using CellTiter-Glo Luminescent Cell Viability Assay (Promega, Madison, WI, USA) and luminescence was measured using Varioscan Flash platereader (Thermo Fisher Scientific, Waltham, MA, USA).
Cell growth assays
Dose–response growth experiments were performed in 96-well plates. Cells were seeded in their standard growth medium and allowed to adhere for 2 days before 5 days treatment with barasertib or JNJ-7706621 (Selleck Chemicals) at indicated concentrations. Cell number was determined using a crystal violet colorimetric assay as described previously [
29]. All experiments were performed at least twice with similar results. Data represent mean values of 6 wells ± SD from one representative experiment.
Western blot analysis
To investigate the effect of barasertib on protein expression and phosphorylation of Aurora kinase A, Aurora kinase B and INCENP, as well as PARP cleavage, parental and fulvestrant resistant T47D cell lines were treated for 4–96 hours with 50 nM barasertib (Selleck Chemicals). Cell lysis and western blot analyses were performed as previously described [
11]. Antibodies targeting the following proteins were used: Aurora kinase B (1:1000, #AJ1069a, Nordic Biosite, Copenhagen, Denmark), pThr288-Aurora kinase A/pThr232-Aurora kinase B/pThr198-Aurora C (1:1000, #2914, Cell Signaling Technology, Danvers, MA, USA), INCENP (1:2000, #ab12183, Abcam, Cambridge, MA, USA), Hsp70 (1:500,000, #MS-482-PO, Thermo Fisher Scientific), and PARP1 (1:1400, #6639GR, BD, Franklin Lakes, NJ, USA). All experiments were performed using at least two independent sets of lysates with similar results. Quantification was performed using Image J. The protein expression level of the specific proteins was measured relative to the respective Hsp70 loading control. The level in parental and untreated cells was set to 1.0.
Flow cytometry
For cell cycle analysis, cells were fixed in 70% ethanol and incubated for 30 min with 20 μg/ml propidium iodide (Sigma-Aldrich, Copenhagen, Denmark) and 40 μg/ml RNase A (Roche, Basel, Switzerland) [
30]. To detect the fraction of phospho-Histone H3 Ser10 positive cells, cells were fixed in 2% formaldehyde (37°C, 10 min), permeabilized in 90% ethanol (−20°C, overnight) and incubated 1 hour at 37°C with AlexaFluor488-conjugated phospho-S10-Histone H3 antibody (1:50, #3465, Cell Signaling Technology) before staining with 20 μg/ml propidium iodide (Sigma-Aldrich). Cell death was measured utilizing a SYTOX green assay, as previously described [
31]. Briefly, cells were incubated with 0.5 μmol/L SYTOX green nucleic acid stain (Life Technologies) (37°C, 15 min), harvested in AccuMax (Thermo Fisher Scientific) and pooled with cells from the growth medium. Samples were subsequently resuspended in 1% FBS in PBS and kept on ice. All samples were analyzed using FACSort flow cytometer and CellQuest Pro (BD).
Hoechst stain and fluorescence imaging
T47D, 182R-1 and 182R-2 cells were seeded in SlideFlask Chambers and treated with 0.1% DMSO (control) or 50 nM barasertib. The cells were fixed in 4% formaldehyde, permeabilized by 0.2% Triton X-100, stained with Hoechst 33342 (Life Technologies, 1:40,000) and mounted using Fluorescence mounting medium (Dako, Glostrup, Denmark). Pictures were captured using Zeiss AX10 Imager A2 microscope (Carl Zeiss Microscopy, LLC, Thornwood, NY, USA).
Patients
The cohort included 268 high-risk ER-positive postmenopausal breast cancer patients diagnosed between 1989 and 2001. The patients had received tamoxifen as first-line adjuvant endocrine treatment according to the guidelines from the Danish Breast Cancer Cooperative Group (DBCG) [
32]. The standard clinico-pathological parameters have previously been published [
33]. The biomarker study was approved by the local ethics committee for Region South Denmark, S-VF-20040064, the Ethical Committee waived the requirement for informed consent from the participants.
Immunohistochemistry (IHC)
IHC was conducted on tissue microarrays (TMAs) using a standard immunoperoxidase procedure [
33]. In brief, TMA sections, comprising two 2 mm cores from each patient, were dewaxed and rehydrated through graded ethanol. Antigen retrieval was performed by heat-induced epitope retrieval (microwaving) for 15 minutes in 10 mM Tris, 0.5 mM EDTA, pH 9. Endogenous peroxidase activity was quenched by 3% hydrogen peroxide and non-specific binding blocked by Serum-free protein block (Dako). Aurora kinase B antibody (1:500, #AJ1069a, Nordic Biosite) was applied over night at 4°C. EnVision (Dako) was used for signal amplification and positive staining was visualized using 3.3-diaminobenzidine tetrahydrochloride (DAB; Dako). Nuclei were counterstained with haematoxylin before mounting in Pertex (Histolab, Göteborg, Sweeden). Aurora kinase B expression was scored as percentage positive tumor cells, blindly and without reference to the patient history.
Statistics
In the kinase inhibitor screen, one-tailed Student’s t-test was performed on triplicate values comparing the growth inhibitory effect in parental and resistant cell lines. In the remaining experiments, group comparisons were done using a two-tailed t-test with Bonferroni adjusted p-values for multiple testing. In the biomarker study, uni- and multivariate analyses were performed. The multivariate analysis included tumor grade, size, nodal status and age as standard covariates. Kaplan-Meier life tables with log-rank testing were generated to assess the association between the percentage of Aurora kinase B positive tumor cells, and disease-free and overall survival. The statistical analysis on the clinical data was performed in R version 3.0.1, with the R package “rms”. For all experiments, P < 0.05 were considered statistically significant.
Discussion
Although endocrine therapy targeting estrogen signaling through ER has clearly improved survival of breast cancer patients, treatment resistance is complex and a major clinical challenge. To explore the molecular mechanisms behind antiestrogen resistance, we have developed cell lines resistant to the antiestrogens fulvestrant and tamoxifen based on the estrogen responsive T47D breast cancer cell line, and utilized a kinase inhibitor screen to identify kinases involved in growth of the fulvestrant resistant T47D cell lines. We found that the Aurora kinase B specific inhibitor barasertib preferentially inhibited growth of the fulvestrant resistant T47D breast cancer cell lines compared with growth of the parental fulvestrant sensitive T47D cells. To verify the role of Aurora kinase B for fulvestrant resistant cell growth, siRNA-mediated knock-down experiments were performed with several siRNA constructs. However, we did not obtain significant knock-down with any of the constructs, including the construct which in our MCF-7 cells reduced Aurora kinase B expression [
38]. Therefore, another Aurora kinase B specific inhibitor hesperadin was tested and the observed preferential growth inhibition of the fulvestrant resistant cell lines supports the role of Aurora kinase B for growth of fulvestrant resistant T47D cells. The inhibition with hesperadin at 1 μM was less pronounced than for 1 μM barasertib, which may be explained by the difference in potency of the two inhibitors, IC
50 for barasertib is 0.37 nM and for hesperadin 250 nM (see
www.selleckchem.com). Aurora kinase B is a key regulator of mitosis and essential for cell proliferation. It is the catalytic component of the chromosomal passenger complex (CPC), composed of Aurora kinase B, survivin, INCENP and borealin [
22]. The complex is critical for accurate chromosomal segregation, cytokinesis, and regulation of the mitotic checkpoint [
22]. We show that phosphorylation of Aurora kinase B and INCENP was increased in the fulvestrant resistant cell lines compared to the level in the parental T47D cells, indicating active CPC complex and suggesting that activation of Aurora kinase B protein is of particular importance for the fulvestrant resistant T47D cell lines. Over expression of Aurora kinase B has previously been found to interfere with chromosome bi-orientation and the spindle-assembly checkpoint due to enhanced disruption of kinetochore-microtubule attachments and sister chromatid cohesion [
39]. In addition, Aurora kinase B over expression also caused abnormal cytokinesis resulting in chromosome segregation errors [
39,
40]. Although we do not find overexpression of Aurora kinase B in the fulvestrant resistant cell lines, our finding of increased level of the active form of Aurora kinase B and the active form of the downstream targets INCENP and mitosis specific histone H3, indicates that Aurora kinase B plays a major role for fulvestrant resistant cell growth, and the increased cell death, as measured by SYTOX-positive cells in the resistant cell lines, compared to the untreated parental T47D cell line, could possibly be caused by impaired cytokinesis.
In this study, barasertib preferentially inhibited fulvestrant resistant cell growth and phosphorylation of both Aurora kinase B and its targets INCENP and mitosis specific histone H3 in the resistant cells. Moreover, barasertib obstructed proper chromosome segregation and induced arrest of the fulvestrant resistant cells in the G2-phase. The increased proportion of cells in SubG1 together with PARP cleavage in the fulvestrant resistant cell lines indicates that barasertib-induced cell death is mediated by the apoptotic death pathway, as previously described in a panel of human myeloma cell lines [
41]. Parental T47D cells treated with barasertib also displayed erroneous chromosome segregation and multinucleated cells. However, in contrast to the resistant cells, progression through the cell cycle was not prevented, rather the cells reentered a new cell cycle without cytokinesis, resulting in polyploid cells (>4N) and cell survival, at least for a period. Collectively, our results show that the fulvestrant resistant cells are more vulnerable to disturbance of proper cell cycle progression induced by treatment with barasertib, and suggest that the fulvestrant resistant cells are more dependent of Aurora kinase B for survival, resulting in cell death upon inhibition of Aurora kinase B. The primary goal of the study was to disclose the growth promoting pathways in fulvestrant resistant T47D cells. However; we also found significant and preferential growth inhibition of two tamoxifen resistant T47D cell lines with barasertib, but the effect was less pronounced in the tamoxifen resistant T47D cell lines compared to fulvestrant resistant cell lines. We have recently shown that Aurora kinase A and ER are major players in growth of tamoxifen resistant cell lines, including the tamoxifen resistant T47D cell lines [
28], and Aurora kinase A has been found to confer tamoxifen resistance by activating ER by phosphorylation [
37]. Thus, whereas Aurora kinase B appears to play a major role in the ER-negative fulvestrant resistant T47D cell lines [
11], Aurora kinase B may play a minor role in the ER-positive tamoxifen resistant T47D cell lines. Major importance of Aurora kinase A has also been found in ER-positive aromatase inhibitor resistant cell lines, and Aurora kinase B also contributes to aromatase inhibitor resistant cell growth [
38]. These data support the importance of Aurora kinases for growth of endocrine resistant breast cancer cells and whereas Aurora kinase B has a major role in ER-negative cells, Aurora kinase A appears to have a major role in ER-positive breast cancer cells.
Administration of barasertib has been shown to potently inhibit growth of colon, lung, hematologic and breast tumor xenografts as well as a panel of human breast cancer cell lines [
35,
42,
43]. In clinical studies, no objective tumor response was observed in any of the patients with solid malignant tumors treated with barasertib. However, they generally tolerated barasertib well, and 23-25% of the barasertib-treated patients achieved prolonged disease stabilization [
44,
45] (see
www.clinicaltrials.gov). Barasertib was also well tolerated in patients with acute myeloid leukemia enrolled in clinical phase I and I/II studies. These trials showed an overall response rate of 19-25% [
46,
47]. Collectively, these findings suggest that breast cancer patients resistant to antiestrogens could benefit from treatment targeting Aurora kinase B, e.g. barasertib.
At present, we do not know the mechanisms whereby the activity of Aurora kinase B is up regulated in our fulvestrant resistant cell lines or how Aurora kinase B is involved in resistant cell growth. The findings in this study suggest that the protein harbors key functions needed for the resistant cell lines to survive upon development of resistance. Although we do not find over expression of Aurora kinase B in the fulvestrant resistant T47D cell lines, over expression of Aurora kinase B may be important for tumor cell growth. Noteworthy, over expression of Aurora kinase B in Chinese hamster embryonic diploid fibroblasts results in aneuploidy cells capable of forming aggressive tumors in nude mice [
40] and Aurora kinase B is found over expressed in several solid cancers, including breast, colorectal, kidney, lung, and prostate cancer [
27]. Additionally, a correlation between Aurora kinase B expression and poor survival has been demonstrated in several cancers including glioblastomas, head and neck squamous cell cancer and lung cancer [
27,
48,
49]. Here, immunohistochemical analysis performed on primary tumors from ER-positive breast cancer patients, receiving first-line adjuvant endocrine therapy with tamoxifen, revealed that high percentage (above median) of Aurora kinase B positive tumor cells was a marker for reduced disease-free and overall survival in this patient cohort. Thus, Aurora kinase B may be a marker for resistance to antiestrogens. However, since Aurora kinase B is an important cell cycle regulator, we cannot exclude that it is a proliferation marker, as resistant cells may have high proliferative activity. This is to the best of our knowledge the first study showing a link between high Aurora kinase B and reduced benefit from tamoxifen treatment. Based on our
in vitro studies with the antiestrogen resistant T47D breast cancer cell lines, analysis of the association between Aurora kinase B and survival of breast cancer patients treated with fulvestrant would also be of great interest. Archival breast cancer tissue from such patients is unfortunately not available for us. Our previous findings of the importance of Aurora kinase A and ER for growth of tamoxifen and aromatase inhibitor resistant breast cancer cell lines [
28,
38] and for high Aurora kinase A expression as a marker for reduced response to tamoxifen therapy [
28], indicate that both Aurora kinase A and B may be useful markers in endocrine resistant breast cancer and also targets for treatment.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
SLL and TK were the principal investigators responsible for study design, experimental work, interpretation of data and writing the manuscript. SLL and CWY performed the experiments with flow cytometry, AVL, BBR and MB were responsible for tumor collection and AVL, BBR, SLL and TK for biomarker evaluation, AKDH performed the statistical analysis of the biomarker study, and AEL contributed to interpretation of data and writing of the manuscript. All authors critically read, commented and approved the final manuscript.