Elevated CRB3 expression suppresses breast cancer stemness by inhibiting β‐catenin signalling to restore tamoxifen sensitivity

Abstract Tamoxifen is a first‐line drug for hormone therapy (HT) in oestrogen receptor‐positive breast cancer patients. However, 20% to 30% of those patients are resistant to tamoxifen treatment. Cancer stem cells (CSCs) have been implicated as one of the mechanisms responsible for tamoxifen resistance. Our previous study indicated that decreased expression of the CRB3 gene confers stem cell characteristics to breast cancer cells. In the current investigation, we found that most of the breast cancer patient tissues resistant to tamoxifen were negative for CRB3 protein and positive for β‐catenin protein, in contrast to their matched primary tumours by immunohistochemical analysis. Furthermore, expression of CRB3 mRNA and protein was low, while expression of β‐catenin mRNA and protein was high in tamoxifen resistance cells (LCC2 and T47D TamR) contrast to their corresponding cell lines MCF7 and T47D. Similarly, CRB3 overexpression markedly restored the tamoxifen sensitivity of TamR cells by the MTT viability assay. Finally, we found that CRB3 suppressed the stemness of TamR cells by inhibiting β‐catenin signalling, which may in turn lead to a decrease in the breast cancer cell population. Furthermore, these findings indicate that CRB3 is an important regulator for breast cancer stemness, which is associated with tamoxifen resistance.

Long-term clinical observations have revealed that HT has several advantages such as low toxicity, easy treatment and comparable therapeutic efficacy over chemotherapy. Currently, there are three main types of drugs used in HT: selective oestrogen receptor modulators (SERMs), aromatase inhibitors (AIs) and selective oestrogen receptor down-regulators (SERDs). 5 Tamoxifen, one of the SERMs, is a first-line treatment drug for HT in oestrogen-positive breast cancer. It competes with oestrogen for ER in vivo and in turn inhibits the biological functions of oestrogen, and reduces mortality and recurrence rates of breast cancer, as demonstrated by the 5-year survival rate of ER(+) patients of breast cancer rising to 85% in recent years. 6 However, approximately 20% 30% of ER(+) breast cancer patients acquired tamoxifen resistance. 6 As tamoxifen resistance drastically enhances malignancy, metastasis and mortality of breast cancer, it has become a limitation in the therapeutic application of tamoxifen and a barrier that ultimately affects the prognosis of breast cancer patients. Therefore, it is important to identify the major molecular targets that lead to tamoxifen resistance, molecularly predict tamoxifen resistance, assess more effective treatment strategies and design a potential rational regimen for combination therapy.
CRB3, the human orthologue of CRB, is a tumour suppressor in mammalian epithelial cells. Loss of CRB3 was observed in mice which were initially injected with a non-tumorigenic kidney cell line and later developed tumours spontaneously through in vivo selection, and re-expressing CRB3 restored cell-cell junctions, cell polarity and contact inhibition. 7 CRB3 is also able to relay cell density information to regulate proliferation of mouse mammary cells. 8 Ben Margolis et al reported that transmembrane protein CRB3 also localized to a membrane compartment at the mitotic spindle poles and played a crucial role in cell division via unique interactions with importin b-1. 9 Our previous study showed that CRB3 expression was associated with poor prognosis in 136 pairs of human ccRCC (clear cell renal cell carcinoma) and adjacent normal kidney tissues by Immunohistochemistry. 10 Altogether, decreased expression of the CRB3 gene activated the Wnt signalling pathway, resulting in the non-tumorigenic, immortalized breast epithelial cell line MCF 10A acquiring cancer stem cell (CSC) properties. 11 It has been postulated that tamoxifen resistance is attributed to the expansion of cancer stem cells. 12 The IHC staining intensity was scored as 0 (negative), 1 (weakly positive), 2 (moderately positive) or 3 (strongly positive). The extent of staining was defined as the percentage of stained cells per field and scored as 0 (<10%), 1 (10%-40%), 2 (40%-70%) or 3 (>70%). The staining score for each field was calculated as the combination of the intensity and the extent of the staining.

| Real-time PCR
The primers were designed by TaKaRa. Measurements were performed in triplicate and normalized to GAPDH levels. Primer pairs used in real-time PCR are listed in Table 2.

| Cell viability assay
The cells were seeded in 48-well plates at a density of 10 4 cells/ well. One day post-seeding, various concentrations of tamoxifen were added for 24, 48 or 72 hour. DMSO was used as a control.

| Animals, xenotransplantation and treatments
All animal experiments were approved by the Institutional Animal Care and Use Committee of Xi'an Jiaotong University. A total of 2.5 9 10 6 cells resuspended in 100 lL of PBS were injected subcutaneously into the mammary fat pads of 6-week-old female SCID/ Beige mice (Laboratory Animal Center of Xi'an Jiaotong University, China). Tamoxifen (5 mg/kg in peanut oil) was administered daily by gavage as previously described. 17 Tumour volume was calculated using the following formula: (long axis 9 short axis 2 )/2.

| Statistical analysis
Statistical analyses were performed in GraphPad Prism, version 7.00. The statistical significance between two groups was compared by unpaired t test, nonparametric Spearman's correlation or Wilcoxon signed-rank test; three groups were compared by oneway ANOVA with Dunnett's multiple comparisons test or two-way ANOVA with Sidak's multiple comparisons test. All statistical tests were two-sided. All data were from experiments performed at least three times with similar results. All results are expressed as mean AE SEM (n = 3, *P < .05, **P < .01, ***P < .001, ****P < .0001). Immunohistochemistry analysis showed that CRB3 expression was reduced in tamoxifen-resistant tissues compared to their matched primary tumour counterparts ( Figure 1A). As our previous study demonstrated that decreased CRB3 expression leads to activation of the Wnt signalling pathway, 11 we also examined b-catenin expression in tamoxifen-resistant tissues by IHC. As expected, b-catenin expression was elevated in resistant tissues ( Figure 1A).  Figure 1D and E). These expression patterns strongly suggested that CRB3 and b-catenin might be involved in tamoxifen resistance of breast cancer.

| CRB3 regulates tamoxifen sensitivity of breast cancer cells
To study the role of CRB3 in tamoxifen sensitivity, we decreased properties of breast cancer cells whose CRB3 expression was altered.
CD44 high /CD24 low and ALDH immunophenotypical cells represent a tumour cell population with limited stem cell-like potential (21,26). We hence evaluated a tumour cell population with these markers upon changes of CRB3 expression levels. The FACS analysis revealed that the CD44 high /CD24 low (the FMO and isotype control data are shown in Figure S1) and ALDH subpopulations were significantly increased as CRB3 was knocked down and were decreased as CRB3 was overexpressed ( Figure 3A-D). The EGF-supplemented serum-free mammosphere formation is a standard assay of CSC self-renewal. 18 Similar to the observations in changes of the CD44 high /CD24 low and ALDH subpopulations, CRB3-knockdown MCF7 and T47D cells displayed increased size and number of mammospheres ( Figure 4A and B), while CRB3 overexpression resulted in a decrease in size and number of mammospheres in LCC2 ( Figure 4C) and T47D TamR cells ( Figure 4D).
In aggregate, these results demonstrated that elevated CRB3 expression inhibits stemness of tamoxifen-resistant breast cancer cells.

| CRB3 inactivates b-catenin in tamoxifenresistant breast cancer cells
The result in Figure 1a

| CRB3 prevents proliferation of tamoxifenresistant cells
As the Wnt/b-catenin signalling pathway plays a crucial role in regulating cell proliferation, 19 we examined the effect of CRB3 on proliferation. As shown in Figure 6A, the growth of CRB3-downregulated MCF7 cells was increased compared with the parental cells, and CRB3-overexpressing LCC2 cells exhibited a decreased proliferation rate, which were consistent to cell cycle division analysis by flow cytometry ( Figure 6B). CRB3-knockdown MCF7 cells appeared to mainly accumulate in the S phase, whereas the number of cells in the G1 phase decreased significantly ( Figure 6C and D).
However, CRB3-overexpressing LCC2 cells accumulated in the G1 phase. Furthermore, CRB3 up-regulation decreased cyclin D1 expression but increased p21 and p27 expression, three of which are cell cycle regulatory molecules ( Figure 6E). Overall, these results demon-

CONFLI CT OF INTEREST
The authors declare no conflict of interest.