ERα is required for suppressing OCT4‐induced proliferation of breast cancer cells via DNMT1/ISL1/ERK axis

Abstract Objective POU5F1 (OCT4) is implicated in cancer stem cell self‐renewal. Currently, some studies have shown that OCT4 has a dual function in suppressing or promoting cancer progression. However, the precise molecular mechanism of OCT4 in breast cancer progression remains unclear. Materials and Methods RT‐PCR and Western blot were utilized to investigate OCT4 expression in breast cancer tissues and cells. Cell proliferation assays and mouse models were applied to determine the effects of OCT4 on breast cancer cell proliferation. DNMT1 inhibitors, ChIP, CoIP, IHC and ERα inhibitors were used to explore the molecular mechanism of OCT4 in breast cancer. Results OCT4 was down‐regulated in breast cancer tissues, and the overexpression of OCT4 promoted MDA‐MB‐231 cell proliferation and inhibited the proliferation of MCF‐7 cells in vitro and in vivo, respectively. Two DNMT1 inhibitors (5‐aza‐dC and zebularine) suppressed OCT4‐induced MDA‐MB‐231 cell proliferation through Ras/Raf1/ERK inactivation by targeting ISL1, which is the downstream of DNMT1. In contrast, OCT4 interacted with ERα, decreased DNMT1 expression and inactivated the Ras/Raf1/ERK signalling pathway in MCF‐7 cells. Moreover, ERα inhibitor (AZD9496) reversed the suppression of OCT4‐induced proliferation in MCF‐7 cells via the activation of ERK signalling pathway. Conclusions OCT4 is dependent on ERα to suppress the proliferation of breast cancer cells through DNMT1/ISL1/ERK axis.

metastasis, 3,4 and OCT4 could suppress the metastatic potential of breast cancer cells (BCCs). 5 OCT4 has been used as a marker of cancer stem cells and has distinct functions in different pathways in cancer cells. Some studies have revealed that aberrant expression of OCT4 is linked to a variety of human cancers. [6][7][8] Furthermore, epigenetic mechanisms, such as DNA modification, is involved in tumour-propagating phenotype induced by OCT4. 9 However, the role and underlying mechanism of OCT4 in BC progression remain elusive.
Human cancers have been found to be associated with aberrant DNA hypermethylation at CpG islands of tumour suppressor genes, most of which are unmethylated in normal cells. 10 DNMT1 is a major DNA methyltransferase that is responsible for maintaining the methylation status during DNA replication. 11 DNMT3a and DNMT3b mainly perform de novo methylation of either unmethylated DNA or hemimethylated DNA to assist in maintenance. 12 Previous studies demonstrated that the stem cell pluripotent markers OCT4 and Nanog could regulate DNA methylation during differentiation in embryonic stem cells, and changes in DNA methylation patterns result in altered cell proliferation. 13 Additionally, DNA methylation is associated with tumour proliferation and metastasis through the extracellular signal-regulated kinase (ERK) signalling pathway. 14,15 ERK1/2, also known as p42/44 mitogenactivated protein kinase (MAPK), can be activated by a variety of growth factors and has many substrates. ERK signalling promotes the activated GTP-bound Ras proteins to activate the Raf-MEK-ERK kinase cascade by a series of phosphorylation events of the kinases. 16 The Ras-ERK pathway mediates various cellular processes, including cell growth, proliferation, differentiation, survival and migration. 17,18 Our recent study using the next-generation sequencing (NGS) showed that overexpression of OCT4 in human hair follicle mesenchymal stem cells up-regulated the expression of 1181 genes, including KRAS gene, which is the upstream of ERK signalling pathway. 19 Therefore, the correlation of the stem cell pluripotent marker OCT4, DNA methylation and ERK signalling pathway in breast cancer proliferation should be examined. However, the present studies demonstrate that OCT4 exerts dual effects in breast cancer, 5,20 which may be related to the multiple intrinsic genes involved in different breast cancer subtypes, especially estrogen receptor (ER), progesterone receptor (PR) and human epidermal growth factor 2 (HER2). Estrogen receptor alpha-positive (ERα+) subtype accounts for approximately 80% of all breast cancers, which is the most common cancer in women. 21 Up to 50% of ERα+ primary BC lose ERα expression in recurrent tumours, conferring resistance to tamoxifen therapy. 22 Inactivation of ESR1 gene via methylation strongly correlates with poor prognosis as well as an aggressive phenotype in TNBC. 22 Additionally, ERα can be complexed with OCT4 to promote tamoxifen resistance in breast cancer cells. 23 In the current study, we provide evidence that OCT4 is down-regulated in invasive breast cancer, which plays a key role in BCC proliferation. However, OCT4 can function as an oncogene as well as tumour suppressor gene in TNBCs and luminal A subtype cells. Therefore, we elucidated the mechanism by which OCT4 exerts its tumour-suppressive function, showing that OCT4 is dependent on ERα to suppress the proliferation of breast cancer cells through DNMT1/ISL1/ERK axis, and this axis will be a novel potential target for improving the diagnosis, therapy and prognosis of breast cancer patients.

| Reverse transcription PCR
Total RNA was collected using TRIzol reagent (Invitrogen). Reverse transcription PCR (RT-PCR) was conducted according to our previous protocol. 24 GAPDH was used as an endogenous control. The PCR primers are shown in Table 1 and Table S1. The reaction products were resolved on 1.5% agarose gels and visualized by staining with ethidium bromide. The image was observed and photographed under a viltalight lamp using a Gel Imaging System (Bio-Rad Laboratories, Inc, Hercules, CA). The results were analysed by Quantity One 4.4.1 software (Bio-Rad Laboratories, Inc).

| Plate colony formation assay
In 6-well plate, cells were seeded into each well with 2 mL DMEM supplemented with 10% FBS. After 2 weeks, the resulting colonies were fixed with methanol at room temperature for 15 minute and then stained with Giemsa for 20 minute. Colonies were counted.
The colony formation index was defined as the ratio of colony numbers to the initial numbers of the cells plated in each well (100 cells/well).

| Soft agar colony formation assay
Cell suspensions were mixed with 1.2% soft agar in DMEM containing 20% FBS. In 6-well plate, cells were seeded into each well with 2 mL DMEM supplemented with 10% FBS. Next day, the cell suspension was removed. After 2 weeks, the colonies were counted. The colony formation index was defined as the ratio of colony numbers to the initial numbers of the cells plated in each well (100 cells/well).

| Immunofluorescence
Cells were seeded on small coverslips. After washing three times with PBS, the cells were fixed with 4% paraformaldehyde for 10 minute at room temperature. The cells were incubated with 0.1% Triton X-100 and BSA for 1 hour and then incubated with Ki67 antibody (1:200; Abcam, ab15580) at 4°C overnight. After washing three times with PBS, the cells were incubated with secondary antibody (1:1000; CST, #8889) for 1 hour. After washing three times with PBS, cells were stained with DAPI for nuclei (D8417; Sigma).  Tumour volumes were determined according to the following formula:

| Animal model
(length × width 2 )/2. After the last measurement of tumour volume, the mice were sacrificed under anaesthesia with 5 mg/100 g body weight sodium pentobarbital, and tumour tissues were removed.

| Co-Immunoprecipitation
Cells were lysed in IP Lysis/Wash Buffer supplemented with phenylmethane sulfonyl fluoride (PMSF) and protease inhibitor cocktail (PIC) at 4°C for 30 minute and then centrifuged at 13 000 g for 20 minute at 4°C. Lysate was immunoprecipitated with the TA B L E 1 PCR primers sequences

| Immunohistochemistry
All samples were fixed in 4% paraformaldehyde overnight at

| Chromatin immunoprecipitation
Chromatin immunoprecipitation (ChIP) was performed using the EZ

| Statistical analysis
Differences between means of independent groups were tested using unpaired Student's t -test. Analyses were carried out using GraphPad Prism 7 (GraphPad software). P values < 0.05 were considered statistically significant. Each experiment was repeated three times.

| OCT4 expression is down-regulated in breast cancer tissues
Previous analyses of OCT4 expression in human somatic tumour cell lines showed that OCT4 expression was absent in MCF-7 and HeLa cells compared with nTera cells. 25 In the Finak Breast data set, we found that the POU5F1 mRNA level was significantly decreased in invasive breast cancer compared to the breast samples with all four probes ( Figure 1A). Consistently, OCT4 expression declined in 40 human breast cancer tissues relative to 10 normal breast tissues, as shown by using Western blot and IHC analyses ( Figure 1B,C). These results indicate that OCT4 might play an important role in suppressing breast cancer progression.

| OCT4 plays an important role in breast cancer proliferation in vitro and in vivo
To assess the role of OCT4 in BCCs, we established stable OCT4- Collectively, these findings demonstrate that OCT4 plays opposite roles in the tumour-propagating capacity of TNBC and luminal subtype cells in vitro and in vivo.

| OCT4 effects on DNMT1 and Ras/Raf1/ ERK1/2 signalling pathway
As OCT4 is involved in epigenetic regulation in embryonic stem cells,  Figure 4D and Figure S4), whereas OCT4 inactivated the pathway in MCF-7 cells ( Figure 4E). These findings indicate that OCT4 may affect the proliferation of BCCs through DNMT1 and ERK signalling pathway.

| 5-aza-dC and zebularine inhibit MDA-MB-231-OCT4 cell proliferation via inactivation of ERK signalling pathway
To investigate whether OCT4 affects cell proliferation by regulating DNMT1 and ERK signalling pathway, we treated

| ISL1 expression is associated with overall survival in breast cancer
We analysed DNMT1 downstream target ISL1, which is considered as a tumour suppressor gene and is hypermethylated in cancer. 30 As ISL1 is a downstream target of DNMT1 and correlated with cells treated with 5-aza-dC and zebularine ( Figure 6B). Additionally, we found that ISL1 interacted with Ras in MCF-7-OCT4 cells ( Figure 6C). These findings reveal that DNMT1 inactivates ERK signalling pathway by targeting ISL1 in MDA-MB-231 and MCF-7 cells.
Next, we evaluated ISL1 expression in 10 samples of normal breast tissues and 66 samples of different subtypes of breast cancer tissues. IHC analysis revealed that ISL1 was localized predominantly in the nucleus ( Figure 6D). We further investigated the relationship between ISL1 expression and clinical parameters. We found that  (Table 2). Moreover, we constructed Kaplan-Meier curves for the overall survival (OS), which indicated that patients with ISL1-positive tumours had a higher OS rate ( Figure 6E). These results demonstrate that ISL1 is a tumour suppressor gene in BC and may be associated with tumorigenesis and progression in BC.

| OCT4 interacting with ERα suppresses cell proliferation of MCF-7 cells
To elucidate the mechanism of the dual functions of OCT4 in proliferation of different subtypes of BCC, we hypothesized that the opposing roles of OCT4 in cell proliferation might be associated with the intrinsic characteristics of BCCs. Previous studies showed that DNMT1 was negatively correlated with ERα in breast cancer. 31 Therefore, we tested ERα expression in MDA-MB-231-OCT4 and MCF-7-OCT4 cells by using PCR and Western blot analyses.
These results demonstrate that OCT4 overexpression with the loss of ERα promoted the proliferation of breast cancer cells, indicating that OCT4 is dependent on ERα to suppress the proliferation of breast cancer cells through DNMT1/ISL1/ERK axis ( Figure 7H).

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

AUTH O R CO NTR I B UTI O N
C.Q., X.J. and Y.Li. were involved in the study conception and design.
drafted and approved the article. C.Q. revised the article critically.
All authors have read and approved the manuscript for publication.