Background
Breast cancer is the most common invasive female cancer worldwide and this is also true for Taiwan [
1,
2]. Rapid advances made using genomic approaches, molecular analysis using immunohistochemistry, measurement of proliferative capacity, and gene expression profiles have allowed the further categorization of breast cancers into four types; these are the luminal A, luminal B, HER2 and basal-like subtypes [
3‐
6]. Biological subtyping not only helps to provides feasible treatment strategies that correlate well with clinical outcome, but also helps to predicts whether the primary tumor has a preferential distant metastasis site [
7].
The epithelial–mesenchymal transition (EMT), which is characterized by a loss of cell–cell adhesion, epithelial polarity, and the gain of mesenchymal properties, such as migration and invasiveness, is mechanism well known to be involved in tumor metastasis and invasion [
8,
9]. Among many EMT regulators, which include Slug, ZEB, E47, Snail and TWIST1, over-expression of the latter two correlates well with tumor recurrence and a poor prognosis for breast cancer [
10‐
13]. Recently, breast cancer subtypes have been identified as having a propensity to give rise to their first distant metastases at certain preferred body sites. For examples, luminal A breast cancers have a tendency to give rise first to bone metastases, while basal type cancers metastasize to liver and brain and HER2-enriched cancers give rise to liver and lung metastases [
7]. Moreover, a recent investigation has suggested that mutational activation of ESR1 plays an important role in acquired endocrine resistance during breast cancer therapy [
14]. However, the role of the estrogen receptor (ER) in the preferential metastasizing of Luminal A breast cancer to bone remains unexplored.
The Src family is a group of nonreceptor tyrosine kinases. Accumulating evidence has suggested that Src plays an important role in bone resorption and osteoclast activation, which seems to lead to the preferential bone metastasis of prostate cancer and the occurrence of late-onset bone metastasis in breast cancer [
15,
16]. Recent clinical studies have been carried out to evaluate the efficacy of Src inhibitors in the treatment of solid tumors and breast cancer with bone metastasis [
17,
18]. Nevertheless, the role of ER-Src signaling during Snail over-expression bone metastasis remains to be elucidated.
RhoGTPases and GTPase activating proteins (GAPs) contain a highly conserved GTPase activating domain that targets small GTPases; these latter proteins regulate intracellular cytoskeletal (actomyosin and microtubules) dynamics via their down-stream Rho kinase [
19]. Among the many RhoGAPs, p190RhoGAP (p190) is a protein that has critical phosphorylation and protein interaction motifs in its middle domain [
20]. A previous study has shown that p190 plays an important role in regulating cytoskeleton dynamics via inactivation of Rho signaling [
21]. In addition, Src phosphorylates p190 at tyrosine 1105 (Y1105) allows enhancement of p190’s RhoGAP activity and Rho inactivation [
22]. However, information on the relationship between the ER-Src-p190 RhoGAP axis and bone metastasis of breast cancer is not available.
Clinically, it was observed during our daily practice that breast cancer with a luminal A subtype would seem to have a significant incidence of bone recurrence as the first distant metastatic site. Experimentally, animal studies have shown that MCF-7 cells that are over-expressing Snail (MCF-7 o/e Snail) have up-regulated expression of
Snail,
N-
cadherin and
Vimentin and down-regulated expression of
E-
cadherin compared to MCF-7 cells, suggesting an EMT process is occurring. In contrast, MCF-7 o/e Snail cells in bone tissue have up-regulated
E-
cadherin, but down-regulated
Snail,
N-
cadherin and
Vimentin. These findings suggest in the latter case a mesenchymal–epithelial transition (MET) process is occurring. Ingenuity pathway analysis (IPA) of the differential gene expression between MCF-7 cells, MCF-7 o/e Snail cells and MCF-7 o/e Snail bone metastasis tissue demonstrates that ERα played a central role in the bone metastasis process (Additional file
1).
Bases on the above, we propose a hypothesis whereby ERα modulates Snail-regulated bone metastasis via the Src-p190 signaling pathway. The aim of this study was to demonstrate the correlation between ER (+) status and bone metastasis as the preferred site clinically and to investigate the role of ERα-Src signaling in MCF-7 o/e Snail cells using an in vivo bone metastasis model.
Methods
Subjects
Between Jan. 2004 and Dec. 2008, 1701 patients with breast cancer were admitted to Taipei Veterans General Hospital. Under the approval of Institutional Review Board (# 2014-11-001AC) of this hospital, their records were retrospectively reviewed in order to investigate the correlation between their sites of distant metastasis and various factors such as the expression of receptors (ER, PR) and their HER2/neu status. The first metastatic site of these patients was defined as metastasis involving a single organ, such as bone, lung, liver, brain and so on. Those who were firstly diagnosed as having multiple organs metastasis (≥2) were excluded. The mean follow up time was >60 months. The positivity of receptor status was defined as positive when ER or PR was ≥10% while, on the other hand, those having ER or PR <10% were classed as negative. The positivity of HER2/neu was defined as positive when staining was +++ by immunohistochemistry or amplified gene expression by FISH examination.
Cell lines and reagents
The human breast cancer cell line MCF-7 (ER+, HER2-low) was obtained from the Food Industry Research and Development Institute (Taiwan, R.O.C.), while the Snail-over-expressed MCF-7 (MCF-7 o/e Snail) line obtained from Professor Hong SC, Department of Orthopedics, Taipei Veterans General Hospital. Both cell lines were routinely screened to show they were free from mycoplasma contamination. They were maintained in DMEM supplemented with 10% fetal bovine serum (FBS), 2 mM l glutamine, 1.5 g/L NaHCO3, 0.1 mM NEAA, 1.0 mM sodium pyruvate, penicillin/streptomycin (Invitrogen, NY, USA) and 10% fetal calf serum (Chemicon, CA, USA).
Western blotting analysis
Cultured cells were lysed in a buffer containing 150 mM KCl, 10 mM Tris pH 7.4 and 1% Triton X-100 together with phosphatase inhibitor and protease inhibitors cocktail (Complete Mini; Roche, Mannheim, Germany). The protein concentrations of the cell homogenates were measured using the Bradford method [
23]. Thirty microgram of proteins were separated using 10% SDS-PAGE and then transferred to a nitrocellulose membrane (Hybond-C; Amersham Biosciences, NJ, USA). The membrane was blocked with 5% bovine serum albumin, which was followed by probing with various specific antibodies such as Src antibody, C-term (GTX61220, GeneTex, Texas, USA), p190-B RhoGAP antibody (GTX61259, GeneTex, Texas, USA), ERα (GTX100634, GeneTex, San Antonio, Texas), anti-α-tubulin 1A (GTX109832, GeneTex, San Antonio, Texas), and anti-β-actin (GTX109639, GeneTex, San Antonio, Texas), N-cadherin (C-terminus clone EPR1792Y. #04-1126 Merck Millipore, Billerica, Massachusetts), E-Cadherin E (#3195) and Vimentin (GTX100619, GeneTex, Texas, USA) and Snail (#AP2054a, ABGENT, San Diego, CA) antibodies. These were purchased commercially (Cell signaling, Danvers, Massachusetts).
Total RNA extraction and reverse transcription-PCR
Total RNA was isolated using a modified single-step guanidinium thiocyanate method [
24] (TRI REAGENT, T-9424, Sigma Chem. Co., St. Louis, MO, USA). Complementary DNA (cDNA) was prepared from the total RNA using a First Strand cDNA Synthesis Kit (Invitrogen, CA, USA). Changes in de novo gene synthesis in each treatment group was detected by reverse transcriptase-polymerase chain reaction (RT-PCR). The gene expression of p190 RhoGAP, Src, and SNAIL were elucidated using commercially available primers. The primers for GAPDH were Forward-5′-GGAGCGAGATCCCTCCAAAAT-3′, Reverse-5′-GGCTGTTGTCATACTTCTCATGG-3′; for Src were Forward, 5′-GAACCCGAGAGGGACCTTC-3′, Reverse-5′-GAGGCAGTAGGCACCTTTTGT-3′; for E-cadherin were, Forward-5′-CGAGAGCTACACGTTCACGG-3′, Reverse-5′-GGGTGTCGAGGGAAAAATAGG-3′; for N-Cadherin were Forward-5′-CAACTTGCCAGAAAACTCCAGG-3′, Reverse-5′-ATGAAACCGGGCTATCTGCTC-3′; for p190 RhoGAP were Forward-5′-GCACAACTCGACCTTCTTTGG-3′, Reverse-5′-CGAAATAGAGTATGTCGGCCTTG-3′; for ESR1 were Forward-5′-GCACCCTGAAGTCTCTGGAA-3′ Reverse-5′-TGGCTAAAGTGGTGCATGAT-3′; and for hSnail1 were Forward-5′-GCTCCACAAGCACCAAGAGT-3′ Reverse-5′-ATTCCATGGCAGTGAGAAGG-3‘.
Any possible contamination of the various PCR components was excluded by performing a PCR reaction with these components in the absence of RT product for each set of experiments (non-template control, NTC). Quantification of the RNA transcripts was carried out and the relative mRNA expression levels of the various specific genes were normalized against the expression level of GAPD in the same RNA extract. All samples were analyzed in triplication.
Short hairpin RNA (shRNA) transfection
Short hairpin RNA (shRNA) was used to silence the ESR1 gene and the shRNA was obtained from Academia Sinica. One day after the Snail-over-expressing MCF-7 (MCF-7 o/e Snail) cell line was subcultured, the cells (30–40% confluent) were transfected for 24 h with shRNA against ESR1 or a non-silencing control; this was carried using GenePORTER 2 transfection reagent (Genlantis, San Diego, CA, USA) dissolved in Optimem (Invitrogen) at a final concentration of 80 nM. Next the MCF-7 o/e Snail cells were allowed to recover before further experiments were carried out. After several passages, a ∆ESR1 o/e Snail MCF-7 cell line was established by puromycin selection. The transfection efficiency was validated by Western blot analysis.
An in vivo bone metastasis model was established using either a tumor cell xenograft or intratibial injection of cancer cells. Study protocols, which involved experiments using mice, were approved (#1041118) by the Institutional Animal Committee of Yang-Ming University. All animals were treated under the regulations of the “Guide for the care and use of laboratory animals” [DHHS publication No. (NIH), revised 1996] and the “Improving bioscience research reporting for animal research”. Immunodeficient NU-Foxn1nu mice were obtained from National Laboratory Animal Center (Taipei, Taiwan, ROC). They were given ad libitum access to food and water and were maintained in a specific pathogen-free environment with 12 h light–dark cycle at 22–24 °C and 50% humidity under the regulations of animal care committee of National Yang-Ming University. The mice were used for the experiments at 8 weeks of age. MCF-7 o/e Snail cells and MCF-7 o/e Snail knocked down ESR1 (MCF-7 o/e Snail, ∆ESR1) cells were injected into the right tibia via the knee joint. In total 1 × 10
7 cells/0.1 mL PBS were used for each mouse, and this gave rise to a noticeable solid tumor around the injection site at day 14. The left site of the same mouse was used as the negative control [
25,
26]. After the tumor had formed, the mice were sacrificed when the tumor size was less than 2% of body weight and tissue samples were then collected for further analysis.
Statistic analysis
Correlation between ER positivity and metastatic site was analyzed using Fisher’s exact test. Experimental results are expressed as the mean ± SEM. Differences between two groups were analyzed by student t test or Mann–Whitney U test while those between groups at each time point were identified by one-way ANOVA, followed by Dunnett’s post hoc test. A p value of <0.05 is considered statistically significant compared to vehicle or no treatment group.
Discussion
In this study, we demonstrated that there is a correlation between an ER (+) status and bone metastasis as the preferred metastatic site clinically. We then investigated the role of ERα-Src signaling in MCF-7 cells with Snail over-expression using a bone metastasis model in vivo. Although many clinical studies have shown a correlation between breast cancer subtypes and their patterns of distant metastasis [
7,
27], we are the first to incorporate translational research by elucidating the role of ER-Src signaling in the mechanisms of bone metastasis, both in vitro and in vivo.
There is consensus that an experimental bone metastasis model is not the same situation as that observed in humans. Nonetheless, a bone metastasis-like model involving intratibial injection of cancer cells is generally accepted as a suitable in vivo model when studying the carcinogenesis process related to bone sarcoma [
25], the mechanisms of bone metastasis [
26,
28], and the efficacy evaluation of a therapeutic agent [
29].
The epithelial–mesenchymal transition (EMT) and its reverse process (the mesenchymal–epithelial transition or MET) are critical during embryogenesis [
30]. Among many EMT inducers, which include Snail1, Slug, Twist, FOXC2 etc., Snail is well known to bind to the promoter of E-cadherin and repress its transcription, leading to loss of E-cadherin expression, a fundamental event in the EMT [
31]. A previous study has demonstrated that Snail over-expression represses the de novo synthesis of ER-alpha and does this via the interaction of Snail with regulatory sequences at the ESR1 locus; this leads to increased cell invasiveness [
32]. Our results from the in vivo bone metastatic model, where tumors occurred after injection of MCF-7 o/e Snail cells, but after not injection of MCF-7 cells, suggests that the EMT is a critical event during the bone metastatic process. It should be noted that there are reciprocal changes in the various EMT markers, namely E-cadherin, N-cadherin, Vimentin, and Snail, between MCF-7 o/e Snail cells and MCF-7 o/e Snail tissue (bone tissue), leading us to speculate that a reverse EMT, that is a MET, occurs during seeding of the tumor cells into the bone. The complexity of the bone environment might result in a range of different ligand-receptor interaction. For example, it is known that the administration of 17 beta-estradiol (E2) decreases Slug expression and increases E-cadherin in ERα over-expressing ERα(−) cell lines [
33]. In addition, recent investigations have suggested that mutations that activate ESR1 play an important role in acquired resistance to hormonal therapy [
14] and the preference towards bone metastasis among breast cancer patients (ASCO, annual meeting, 2016). Our results are in agreement with the above-mentioned studies.
The proto-oncogene c-src, a non-receptor tyrosine kinase, has been demonstrated with its cooperating partners to play an important role in the development of many cancers [
34]. Src activation is highly associated with bone metastasis by prostate cancer [
35] and by late-onset bone metastasis from breast cancer. C-Src overexpression is an independent predictor of a poor outcome among breast cancer patients with bone metastasis [
36] and is a potential therapeutic target for patients with secondary bone metastasis [
37].
Recent evidence has demonstrated that Src plays an important role in the signaling and cross talk between growth-related signaling pathways such as the ER pathways [
38] and that this can occurs in either a direct or an indirect manner [
39]. Interestingly, using microarray and ingenuity pathway analysis (IPA), differential gene expression between MCF-7 cells, MCF-7 o/e Snail cells and MCF-7 o/e Snail bone metastasis tissue demonstrated that ERα has direct interaction with Src pathway (Additional file
1). Furthermore, Src is able to phosphorylate p190 at tyrosine 1105 (Y1105), which has been shown to enhance p190’s RhoGAP activity and thus Rho inactivation [
22]; this in turn leads to an increase in proliferative potential and a decrease in migratory activity, which is subsequently followed by colonization of bone tissue. Our results from the knock down of ESR1 shows that there is down-regulation of the de novo synthesis of Src and p190 RhoGAP at both the protein and the transcriptional level; this finally results in a decrease in the incidence of bone metastasis. Thus the above findings help to explain the mechanisms by which breast cancer leads to bone metastasis.
Authors’ contributions
JHC and CSW writing the manuscript and providing clinical data. JYW handling the animals. CYH reading pathology. SCH providing the cell line. LMT conceiving the concept and providing the clinical data. YMS providing clinical and facilities supports. All authors read and approved the final manuscript.