Isobavachalcone sensitizes cells to E2‐induced paclitaxel resistance by down‐regulating CD44 expression in ER+ breast cancer cells

Abstract Oestrogen receptor (ER) is expressed in approximately 60%‐70% of human breast cancer. Clinical trials and retrospective analyses have shown that ER‐positive (ER+) tumours are more tolerant to chemotherapeutic drug resistance than ER‐negative (ER−) tumours. In addition, isobavachalcone (IBC) is known as a kind of phytoestrogen with antitumour effect. However, the underlying mechanism of IBC in ER+ breast cancer needs to be elucidated further. Our in vitro experiments showed that IBC could attenuate 17β‐estradiol (E2)‐induced paclitaxel resistance and that E2 could stimulate CD44 expression in ER+ breast cancer cells but not in ER− cells. Meanwhile, E2 could promote ERα expression to render ER+ breast cancer cells resistant to paclitaxel. Furthermore, we established paclitaxel‐resistant breast cancer cell lines and determined the function of ERα in the enhancement of paclitaxel resistance via the regulation of CD44 transcription. IBC down‐regulated ERα and CD44 expression and thus inhibited tumour growth in paclitaxel‐resistant xenograft models. Overall, our data demonstrated for the first time that IBC could decrease CD44 expression level via the ERα pathway and make ER+ breast cancer cells sensitive to paclitaxel treatment.

However, the underlying mechanism of chemotherapeutic drug resistance in ER+ breast cancer has not yet been elucidated.
17b-Estradiol (E 2 ) can regulate and maintain a series of physiological processes, such as reproduction, CNS development and metabolic balance. 8 Many cellular responses to E 2 are mediated by oestrogen receptors (ERs), including ERa and ERb. Although clinical and experimental data have confirmed the key role of ERa in breast cancer, the role of ERb in breast cancer remains controversial. 8,9 Recent studies have demonstrated that expression of ERa can prevent paclitaxel-induced apoptosis in breast cancer MCF-7 cells. 10,11 ERa can act as a nuclear receptor and regulate target gene transcription via the "classical" model or the "nonclassical" model. 12 In addition to its role in the "classical" model of gene regulation through oestrogen response elements (EREs) in promoter regions, ERa can form protein complexes with other transcription factors, such as Sp1, Ap1 and NF-jB, to regulate gene transcription involved in the "nonclassical" model. [12][13][14] These observations suggest that ERa can regulate a large variety of genes that are associated with the development of chemotherapy resistance in ER+ breast cancer cells.
Psoralen, a traditional Chinese herb, is isolated from dried ripe fruits of leguminous Psoralea corylifolia L. [15][16][17] It is warm natured and pungent flavoured, with the effect of enriching the kidney and strengthening yang. 18 Recent studies have shown that psoralen has some biological functions, such as blood vessel dilatation, myocardial contractility enhancement and antifungal, anticancer and oestrogenlike effects. 19 Modern pharmacological studies have also shown that isobavachalcone (IBC), an important component of psoralen, has strong antibacterial, antioxidant, anti-reverse transcriptase, antitubercular and anticancer abilities. 20,21 Previous studies have reported that IBC inhibits tumour formation in mouse skin cancer and induces apoptosis in neuroblastoma. 22,23 However, the potential functions of IBC in cancer-related treatment need further study.
CD44 and CD24 are characteristic of the cancer stem cell phenotype, and these molecules are closely associated with poor prognosis and chemotherapy resistance in cancer. [24][25][26][27] Recently, natural substances from plants have been documented as effective intervention agents in the down-regulation of CD44/CD24 expression in experimental breast carcinoma. 28    for each dose of paclitaxel tested.

| Cell viability assay
ZR-75-1 and MCF-7 cells were seeded at 5000 cells per well in 96well plates and then treated with the indicated concentrations of paclitaxel (72 hours) or E 2 /IBC (48 hours). Subsequently, the cells were treated with 10 lL MTT (5 mg/mL) at 37°C for 4 hours followed by 150 lL dimethyl sulphoxide, and cell viability was determined by measuring the absorbance at 570 nm using a microplate reader (Bio-Rad, California, USA).
After incubation with appropriate secondary antibodies for 1 hour at room temperature, the membranes were detected by Tanon 5200 Imaging System (Shanghai, China).

| Clonogenic cell survival assay
Cells suspended in fresh culture medium were plated at a density of 400 cells/well onto 6-cm plates and incubated for 24 hours and then cultured in fresh medium containing 5 nmol/L paclitaxel. At the end of 2 weeks of incubation, the cells were fixed with cold methanol and stained with crystal violet (Beyotime, China) for 20 minutes at room temperature. Thereafter, the plates were gently washed with water and allowed to air-dry. All experiments were performed in triplicate.

| Flow cytometry assay
Approximately 10 6 cultured cells were harvested and fixed in 75% ethanol diluted in PBS at 20°C overnight. The cells were then incubated in PBS containing 100 lg/mL propidium iodide (PI), 100 lg/ mL RNase and 0.1% Triton X-100 at room temperature for 30 minutes before flow cytometry analysis. Cell cycle distribution and DNA content were determined using a BD FACSCalibur system (Becton, Dickinson and Company, USA).

| Tumour xenograft and immunohistochemistry assay
Animal experiments were approved by the Animal Care Committee of Nanjing First Hospital, Nanjing Medical University (Approval No. SYXK20160006). Approximately 5 9 10 6 MCF-7/R cells were injected into the mammary fat pads of 5-week-old female athymic nude mice. The mice were randomly assigned into two groups of six.
When the size of the xenograft reached approximately 50 mm 3 , IBC was intraperitoneally administered daily at 100 mg/kg into the nude mice. Tumour volumes were calculated at the indicated time-points with the formula: p/6 9 length 9 width2. Tumour samples from the mice were deparaffinized and rehydrated. After blocking the endogenous peroxidase activity, the sections incubated with primary monoclonal anti-ERa and anti-CD44 antibodies at a dilution of 1:500 overnight at 4°C. The sections were incubated with appropriate HRP-conjugated secondary antibody for 1 hour at room temperature. Colour was developed with 3-amino-9-ethylcarbazole solution.
Image pro plus (IPP) was used as the method of quantitative image analysis on immunohistochemical staining, evaluating the expressing protein of ERa and CD44 gene by IBC treatment. Firstly, the positive area (ERa or CD44 protein staining) in the image was taken as AOI (area of interest). Then the value of SA (sum area) and IOD (integrated optical density) was measured by IPP software, and the value of MOD (mean optical density) was calculated as the following formula: MOD = IOD/SA.

| Statistical analysis
Each of the experiments was repeated at least three times. Values are expressed as the mean AE SD of triplicate measurements unless otherwise noted. Student's paired t test was used to analyse differences between the sample of interest and its control. P < .05 was considered statistically significant.

| IBC decreased resistance of ER+ breast cancer to paclitaxel
Although oestrogen can maintain biological functions in breast cancer cells, its roles in the stimulation of drug resistance in breast cancer cells are still controversial. Therefore, we analysed the viable number of breast cancer cells after treatment with different concentrations of E 2 . Our data indicated that the number of viable cells displayed an increasing tendency in ER+ ZR-75-1 and MCF-7 cells but not in ER-MDA-MB-231 cells ( Figure 1A). After 2 weeks of incubation with 5 nmol/L paclitaxel, the number of colonies formed by ZR-75-1 and MCF-7 cells in the E 2 group was higher than that in the control group ( Figure 1B), suggesting that E 2 could promote the proliferation of ER+ breast cancer cells to make them resistant to paclitaxel.
IBC is well-known as a kind of phytoestrogen, but its therapeutic mechanism in breast cancer remains unclear. To investigate the effect of IBC on breast cancer cells, we carried out cell viability assays in the presence of 1 lmol/L E 2 . Our results demonstrated that the viability significantly decreased in ZR-75-1 and MCF-7 cells treated with 8 lmol/L IBC ( Figure 1C). In addition, fewer colonies of ZR-75-1 and MCF-7 cells were formed in the E 2 +IBC group than in the E 2 group ( Figure 1B). These findings indicated that E 2 could increase the resistance of ER+ breast cancer cells to paclitaxel, while IBC could attenuate E 2 -induced paclitaxel resistance.

| CD44 was up-regulated by E 2 in ER+ breast cancer cells
Although CD44 and CD24 expression is a poor prognostic marker for different tumours including breast cancer, 24,30 the role of CD44 and CD24 in paclitaxel resistance of breast cancer cells stimulated by E 2 needs further study. Therefore, we analysed CD44 and CD24 expres-  Figure 2H). Colony formation assay showed that fewer paclitaxel-resistant colonies were formed in the si-CD44 group, which was, than in the control group ( Figure 2I). These data suggested that CD44 expression might be involved in the resistance to paclitaxel-mediated E 2 stimulation.

| ERa induced by E 2 enhanced CD44 expression in breast cancer cells
Our previous study showed that ERa was essentially involved in chemoresistance, which may be attributed to its role as a nuclear transcriptional factor in regulating some important tumour drugresistant genes. 10 Thus, we detected the expression of ERa in ER+  paclitaxel-sensitive cells transfected with ERa (Fig. 3C). Moreover, the viability of ZR-75-1 and MCF-7 cells transfected with ERa was higher than that of cells transfected with control plasmids ( Figure 3D and E). Interestingly, IBC could clearly reverse the increase in expression of ERa and CD44 stimulated by E 2 as shown by real-time PCR and Western blotting ( Figure 3F and G). Taken together, these results suggested that IBC could decrease CD44 expression in an ERa-dependent manner.

| Establishment of paclitaxel-resistant ZR-75-1/ R and MCF-7/R cell lines
To study the roles of ERa and CD44 in the acquisition of resistance to anticancer agents, we selected ERa+ breast cancer cells to estab-  Figure 1 were treated with paclitaxel (5 nmol/L) for 14 d before being subjected to colony formation assay. *P < .05, **P < .01 P-gp expression in the paclitaxel-resistant cell models. Compared with that in paclitaxel-sensitive cells, the expression level of P-gp was significantly increased in paclitaxel-resistant cells ( Figure 4E), which indicated that ZR-75-1/R and MCF-7/R cells were chemotherapeutic drug-resistant breast cancer cells. Interestingly, compared with those in paclitaxel-sensitive cells, the expression levels of ERa and CD44 were also increased in paclitaxel-resistant cells ( Figure 4E and F), suggesting that ERa and CD44 are up-regulated by paclitaxel and might be involved in the acquisition of the drug resistance phenotype in ERa+ breast cancer cells. This was consistent with our previous findings and strongly suggested that ERa expression was significantly positively correlated with CD44 expression in paclitaxelresistant breast cancer cells.

| Knockdown of ERa significantly decreased CD44 gene expression to render breast cancer cells sensitive to paclitaxel
To evaluate the function of IBC in the paclitaxel resistance phenotypes of ERa+ breast cancer cells, we detected the protein expression of ERa and CD44 in paclitaxel-resistant breast cancer cells with IBC treatment. Indeed, compared with that in control cells, the expression of ERa and CD44 protein was down-regulated in paclitaxel-resistant cells treated with IBC ( Figure 5A). Paclitaxel blocks the G2/M phase of cell cycle and induces cancer cell apoptosis. 33 As shown in Figure 5B Figure 5D). Real-time PCR showed that compared with that in cells transfected with control plasmids, the expression of CD44 mRNA was down-regulated in paclitaxel-resistant cells transfected with ERa-shRNA ( Figure 5E). In addition, knockdown of ERa expression with ERa-shRNA significantly decreased the number of viable ZR-75-1/R and MCF-7/R cells ( Figure 5F and G), indicating that IBC could down-regulate CD44 expression by the ERa pathway to sensitize of resistant breast cancer cells to paclitaxel.

ERa and CD44 expression in a paclitaxel-resistant mouse xenograft model
To evaluate the clinical implications of our findings, we also studied the functional effect of IBC in vivo in established paclitaxel-resistant breast cancer xenograft models. Treatment with IBC reduced tumour growth in the mouse xenograft model ( Figure 6A). In addition, the levels of ERa and CD44 were lower in the IBC group than in the control group by immunohistochemical (IHC) staining and IPP software analysis ( Figure 6B and C), and a positive correlation was observed between ERa and CD44 in MCF-7/R mouse xenograft tumours treated with IBC. These results suggested that IBC displayed an antitumour activity in MCF-7/R xenograft tumour models.

| DISCUSSION
Drug resistance is one of the major obstacles limiting the success of cancer chemotherapy. In recent years, both clinical observations and experimental studies have shown that the therapeutic efficacy of anticancer drugs can be altered by steroid hormones and their receptors, but the potential molecular mechanisms remain unclear. [34][35][36] In this study, using our established paclitaxel-resistant breast cancer cells, we explored whether and how IBC influenced paclitaxel resistance in breast cancer cells via the ERa-dependent pathway. First, we verified that IBC can reverse E 2 -induced paclitaxel resistance in ER+ breast cancer cells. Furthermore, we confirmed a positive correlation between the expression level of ERa and CD44 in our F I G U R E 6 IBC down-regulated CD44 expression mediated by the ERa pathway and inhibited tumour growth in MCF-7/R xenograft models. A, MCF-7/R cells were subcutaneously injected into the right flank of nude mice (6 mice per group) on day 0. When the tumour volume reached 50 mm 3 , the mice were intraperitoneally administered with IBC (100 mg/kg) daily for 24 d. Tumour volume was measured at the indicated time-points. B, Representative immunohistochemistry images of ERa and CD44 staining in tumour tissues of nude mice. C, Quantitative image analysis utilized IPP software on immunohistochemical staining described in Figure 6B of ERa and CD44 protein expression. *P < .05, **P < .01 established paclitaxel-resistant ER+ breast cancer cells. Moreover, knockdown of ERa expression by shRNA or IBC increased the sensitivity of cells with a decreased expression of CD44 to paclitaxel.
Finally, using xenograft tumour models, we confirmed that IBC could reduce tumour growth in ER+ xenografts by inhibiting ERa and CD44 expression.
De novo anti-oestrogen resistance is observed mainly in ER-/PRtumours, but cancer cells can sometimes switch the phenotype from ER+ to ERÀ. 37,38 Based on failure to endocrine therapy, the patients can be treated with chemical therapeutic agents. Paclitaxel, a firstline clinical chemotherapy drug, is used in the treatment of various human solid tumours, including breast cancer. 39,40 However, some types of breast tumours are resistant to paclitaxel. Cumulative data from clinical studies and retrospective studies indicate that the ER status might affect the therapeutic efficacy of paclitaxel because paclitaxel is less effective in patients with ER+ breast tumours than in patients with ER-breast tumours. 5,41 However, most of these data were obtained from comparative studies in tumour cell lines extracted from different patients. Thus, it is difficult to reach a consensus on cellular and molecular mechanisms in these cell lines.
Therefore, the comparison of established paclitaxel-resistant breast cell lines and parental breast cell lines may provide us with a valuable model system to explore the mechanism underlying ERamediated resistance to paclitaxel and other chemotherapeutic agents in breast tumour cells.
IBC is a traditional Chinese medicine that has been used as an antibacterial and antitumour agent. Recent studies support the notion that IBC can inhibit human tongue and liver cancer cells by blocking the Akt signalling pathway. 21  In this study, for the first time, we correlated ERa and CD44 expression, revealing a potential mechanism that ERa could enhance paclitaxel resistance in ER+ breast cancer cells by up-regulating CD44 expression. In our study, we proposed the following drug resistance model: during paclitaxel therapy for breast cancer, paclitaxel can active ERa expression and then ERa can transcriptionally regulate the CD44 gene, which is responsible for resistance to chemotherapy agents. Therefore, ERa can be developed as a therapeutic target in the context of CD44 activation to overcome drug resistance in the clinic.
CD44 is closely associated with chemotherapeutic drug resistance in cancer cells. Previous studies have revealed that CD44 is highly expressed in triple-negative breast cancer (TNBC) and correlated with poor survival of TNBC patients. 42 Of note, CD44 also plays a key role in ER+ breast cancer, resulting in poor prognosis and radiotherapy resistance. 42 However, the role of CD44 in drug resistance of ER+ breast cancer is worth studying in depth. Interestingly, in this study, a high expression of CD44 was positively correlated with ERa existed in our established paclitaxel-resistant ER+ breast cancer cells. It is worth investigating whether the high expression of CD44 increased the resistance of cells. The present study proved that the high expression of CD44 mediated by E 2 -ERa is an important factor rendering cells resistant to paclitaxel. This is in agreement with the notion that IBC down-regulates CD44 expression by the ERa pathway to sensitize breast cancer cells to paclitaxel. In conclusion, the present study provided a potential antitumour mechanism of IBC.
In summary, the present study demonstrated that the ER status might play a significant role in determining the sensitivity of breast cancer to paclitaxel. The results obtained from this study may provide valuable information for improving the understanding of CD44-

CONFLI CTS OF INTEREST
The authors confirm that there are no conflicts of interest.