Endoplasmic reticulum protein 29 (ERp29) confers radioresistance through the DNA repair gene, O6-methylguanine DNA-methyltransferase, in breast cancer cells

Resistance of cancer cells to radiotherapy is a major clinical problem in cancer treatment. Therefore, understanding the molecular basis of cellular resistance to radiotherapy and identification of novel targets are essential for improving treatment efficacy for cancer patients. Our previous studies have demonstrated a significant role of ERp29 in breast cancer cell survival against doxorubicin-induced genotoxic stress. We here reported that ERp29 expression in the triple negative MDA-MB-231 breast cancer cells significantly increased cell survival against ionizing radiation. Methylation PCR array analysis identified that ERp29 expression increased promoter hypomethylation of the DNA repair gene, O6-methylguanine DNA-methyltransferase (MGMT), by downregulating DNA methyltransferase 1. Knockdown of MGMT in the ERp29-transfected cancer cells increased radiosensitivity, leading to a decreased post-irradiation survival. In addition, radiation treatment in the MGMT-knockdown cells elevated phosphorylation of γ-H2AX and cleavage of caspase 3, indicating that depletion of MGMT facilitates DNA double strands breaks and increases cell apoptosis. Hence, our studies prove a novel function of ERp29\MGMT in cancer cell survival against radiation. Targeting ERp29\MGMT axis may be useful for providing better treatment efficacy in combination with radiotherapy in breast cancer.


ERp29 expression increases post-irradiation survival of MDA-MB-231 cells. To understand
whether ERp29 expression confers to radioresistance in MDA-MB-231 breast cancer cells, the ERp29transfected cells (MB-231/ERp29) and mock-transfected (MB-231/vector) control cells were exposed to ionizing radiation (0-6 Gy) and the post-irradiation survival of cells was analysed by clonogenic assay. The radiation dose, D37, which is required to reduce cell survival to 37%, was determined to evaluate radiosensitivity. As indicated in Fig. 1, ERp29 expression in MDA-MB-231 cells (clone B and E) resulted in highly increase of D37 compared to the mock-transfected control cells (clone B: 2.59 ± 0.15 Gy, clone E: 2.84 ± 0.13 Gy vs. control: 2.11 ± 0.14 Gy, p < 0.01).
To further substantiate the ERp29's role in radioresistance in this cell type, the exogenously expressed ERp29 was depleted by treatment with ERp29 siRNA (Suppl. Fig. 1A, Fig. 1B). In line with the reduction of ERp29 in clone B cells, the enhanced post-irradiation survival rate (D37) in clone B cells was attenuated from 2.68 ± 0.12 to 2.15 ± 0.08 (p < 0.05, Fig. 1b). In addition, the endogenous ERp29 in the parent MDA-MB-231 cells was also reduced by siRNA (Suppl. Fig. 1B; Fig. 1 C). Repression of ERp29 led to a decrease of post-irradiation survival (D37) to 1.68 ± 0.10 Gy, compared to the cells pre-treated with non-targeted siRNA (2.15 ± 0.11 Gy, p < 0.05, Fig. 1c). Therefore, ERp29 exerts a radioresistant function in MDA-MB-231 cells.

MDA-MB-231 cells.
Our previous studies showed that overexpression of ERp29 significantly increased the expression of tumour suppressors, such as E-cadherin (CDH1), at both mRNA and protein levels 22 . Given that expression of this tumour suppressor has been found to be regulated by epigenetic mechanism 23,24 , the role of ERp29 in epigenetic regulation was investigated using a Methyl-Profiler ™ DNA Methylation PCR Array in mock-transfected control cells and MB-231/ERp29 cells. Interestingly, we identified that over-expression of ERp29 remarkably enhanced promoter demethylation of tumour suppressor genes including p15, CDH1 and MGMT (Fig. 2a). For instance, the percentage of hypomethylation of CDH1 and MGMT promoters was increased from approximately 2% in mock-transfected control cells to 55-70% and 55-95%, respectively, in MB-231/ERp29 cells (clone B and E), resulting in increased mRNA and protein expression of CDH1 and MGMT (Fig. 2b). Moreover, it was also found ERp29 Scientific RepoRts | 5:14723 | DOi: 10.1038/srep14723 overexpression in MDA-MB-231 cells decreased promoter demethylation in several pro-oncogenes (e.g., Cox-2, Fig. 2a).
To further verify the regulatory role of ERp29 in methylation/demethylation, MS-PCR was used to analyse the methylation status of the promoter region (e.g., 3 rd CpG islands) of MGMT in these cell models. Compared to mock-transfected control cells, the MB-231/ERp29 cells showed a significant reduction of methylation and increase of demethylation of MGMT promoter, similar to those observed in MDA-MB-231 cells treated with 5′ -aza-dC (Fig. 2c). These results suggest that ERp29 expression is able to re-activate MGMT transcription and expression by epigenetic regulation in MDA-MB-231 cells.
ERp29 regulates MGMT promoter methylation via DNMT1 in MDA-MB-231 cells. Since DNA methyltransferase is responsible for increase of DNA methylation, the expression of DNMT1, DNMT3A and DNMT3B was analysed in mock-transfected control cells and MB-231/ERp29 cells. As indicated in Fig. 3a, relative to control cells, ERp29 overexpression in MDA-MB-231 cells significantly inhibited the expression of DNMT1, rather than the expression of DNMT3A or 3B. The role of DNMT1 in epigenetic regulation of MGMT expression was further supported by the fact that DNMT1 knockdown by siRNA in MDA-MB-231 cells (Suppl. Fig. 1C) led to an increase of MGMT expression compared to the cells treated with non-targeted control siRNA (Fig. 3b). MS-PCR analysis showed that DNMT1 knockdown in MDA-MB-231 cells enhanced demethylation and reduced methylation of MGMT promoter relative to the cells treated with control siRNA (Fig. 3c). These data indicate a critical role of DNMT1 in ERp29-mediated inhibition of MGMT promoter methylation.
MGMT is a downstream target regulated by ERp29. To further understand whether MGMT is a downstream target of ERp29, the MB-231/ERp29 cells (clone B) were respectively treated for 48 h with MGMT siRNA, or ERp29 siRNA, or the non-targeted control siRNA. We showed that depletion of ERp29 in MB-231/ERp29 cells reduced the level of MGMT compared to those treated with control siRNA (Fig. 4a). However, depletion of MGMT was unable to affect the level of total ERp29 (endogenously and exogenously expressed) in these cells (Fig. 4a). This is reflected by the fact that the overall ERp29 level in the MGMT siRNA-treated MB-231/ERp29 cells was similar to that expressed in the untreated or control siRNA-treated MB-231/ERp29 cells. Because MDA-MB-231 parent cells showed weak expression of MGMT as assessed by Western blot (Fig. 2b), knockdown of endogenous MGMT was performed in MCF-7 cells which express high levels of both ERp29 and MGMT. Knockdown efficiency of ERp29 and MGMT by siRNA in MCF-7 cells were assessed (Suppl. Fig. 1D and E). As demonstrated in Fig. 4b, reduction of endogenous ERp29 by siRNA decreased the level of MGMT. Nevertheless, repression of endogenous MGMT by siRNA was unable to reduce ERp29 expression in MCF-7 cells. These results support that MGMT is a downstream target of ERp29.
MGMT mediates ERp29-induced radioresistance in MDA-MB-231 cells. We showed that ERp29 overexpression resulted in radioresistance in MDA-MB-231 cells (Fig. 1). To understand whether the upregulated MGMT by ERp29 is involved in radioresistance, the MB-231/ERp29 cells (clone B) were treated for 48 hours with MGMT siRNA to reduce MGMT expression or with control siRNA. These cells were then irradiated (0-6 Gy) and the post-irradiation survival assessed using the clonogenic assay. Relative to MB-231/ERp29 cells treated with control siRNA (D37 = 2.72 ± 0.09 Gy), MGMT knockdown in these cells caused a significantly decrease of D37 (1.82 ± 0.13 Gy, p < 0.05) (Fig. 5a). These data indicate that MGMT depletion re-sensitizes to radiation in MB-231/ERp29 cells. The role of MGMT in protecting cells from radiation-induced cell death was further investigated in MCF-7 cells which showed relative high expression of endogenous MGMT (Fig. 4b). The endogenously expressed MGMT in MCF-7 cells was markedly repressed by treatment with MGMT siRNA for 48 hours (Suppl. Fig. 1D) and this caused a significant reduction of the post-irradiation cell survival rate (D37) from 2.82 ± 0.09 to 2.01 ± 0.11 (p < 0.05, Fig. 5b). These data further suggest MGMT is critical for radioresistance in MCF-7cancer cells.
Knockdown of MGMT reduces DNA repair capacity and enhances DNA damage after irradiation. Phosphorylation of H2AX is a marker for the cellular response to DNA-DSBs 25 . Upon DNA damage in cells, histone H2AX is phosphorylated on serine 139 to generate γ -H2AX 26 . Hence, γ -H2AX expression after ionizing radiation reflects the formation of double strands DNA breaks and DNA repair capacity. To investigate the effect of the upregulated MGMT by ERp29 on the expression of γ -H2AX, the MB-231/ERp29 cells (clone B) were transfected with non-target control siRNA or MGMT siRNA for 48h. These cells, together with vector-transfected control cells, were mock-irradiated or irradiated with 4 Gy. After 12 hours of incubation, the expression of γ -H2AX was assessed. As indicated in Fig. 6a, irradiation treatment significantly increased the expression of γ -H2AX in vector-transfected control cells relative to the mock-irradiation treatment (column 2 vs. 1). However, compared to the irradiation-treated vector-transfected control cells, irradiation treatment only moderately increased the expression of   Cells were treated with control or MGMT siRNA (#3) for 48 hours and then exposed to irradiation at the indicated doses. Expression of MGMT was efficiently repressed by siRNA in the ERp29-overexpressed clone B cells. (b) Depletion of endogenous MGMT by siRNA in MCF-7 cells sensitized to radiation treatment. MCF-7 cells were treated with control or MGMT siRNA (#3) for 48 hours and then exposed to irradiation at the indicated doses. The expression of MGMT in MCF-7 cells was efficiently reduced by siRNA. (c) Re-expression of MGMT in the MB-231/ERp29 siRNA cells restores radioresistance. Cells were transfected with pcDNA-MGMT or pcDNA for 24 hours and exposed to radiation treatment. MGMT expressed was examined by immunoblot. Post-radiation survival rate was assessed by clonogenic assay as described in Fig. 1 Scientific RepoRts | 5:14723 | DOi: 10.1038/srep14723 γ -H2AX in MB-231/ERp29 cells (column 3 vs. 2), indicating a protective role of ERp29 from the irradiation treatment. Nevertheless, knockdown of the ERp29-upregulated MGMT by siRNA markedly increased radiation-induced expression of γ -H2AX, relative to MB-231/ERp29 cells pre-treated with control siRNA (column 4 vs. 3; Fig. 6a). Similarly, knockdown of the endogenously expressed MGMT in MCF-7 cells led to significant increase of γ -H2AX expression, compared to the cells treated with control siRNA after radiation treatment (Fig. 6b).
To understand the effect of MGMT depletion on the kinetics of γ -H2AX after irradiation, the MB-231/ERp29 cells were treated with non-target control siRNA or MGMT siRNA for 48 h and then irradiated at 4 Gy. The level of γ -H2AX was analysed at 0, 2, 12 and 24 hours post-irradiation (Fig. 6c). For the MB-231/ERp29 cells treated with control siRNA, the level of γ -H2AX was rapidly increased within 2 hours and then decreased after irradiation. In contrast, MGMT knockdown in MB-231/ERp29 cells resulted in a sustained high expression and slow reduction of γ -H2AX within 24 hours. These results suggest that loss of MGMT by siRNA leads to increased double strand DNA breaks and impairment of DNA repair capacity in these cells.

Knockdown of MGMT enhances cell apoptosis in the ERp29-transfected MDA-MB-231 cells.
To examine whether the ERp29-induced activation of MGMT is responsible for antagonizing radiation-induced apoptosis, the MB-231/ERp29 cells were treated with MGMT siRNA or control siRNA for 48 hours and exposed to radiation treatment (4 Gy). After 48 hours post-incubation, these cells were used for examining the expression of cleaved caspase 3 and cell viability. In parallel, the vector-transfected control cells were also exposed to radiation treatment (4 Gy). As shown in Fig. 7a, radiation treatment significantly increased the expression of cleaved caspase-3 (column 2 vs. 1) in the To further establish the role of MGMT in cell survival after irradiation, these cells were applied for viability analysis using the MTT assay. Our data showed that irradiation caused > 50% reduction of viable cells in the vector-transfected control cells (column 2 vs. 1; Fig. 7b). In MB-231/ERp29 cells, MGMT depletion alone or treated with control siRNA/radiation was unable to significantly decrease the cell viability compared to the untreated vector-transfected control cells (column 3 vs. 1, and 4 vs. 1; Fig. 7b). However, when the MB-231/ERp29 cells were exposed to a combinatory treatment with MGMT siRNA and radiation, the cell viability was remarkably reduced by > 40% compared to those treated with control siRNA/radiation (column 5 vs. 4; Fig. 7b). Therefore, depletion of MGMT in the MB-231/ERp29 cells significantly increases radiation-induced cell death.

Discussion
Previous studies have demonstrated a significant role of ERp29 in the resistance to DOX in breast cancer cells 12 . The present work has further established a critical function of ERp29 in radioresistance through a mechanism involved in epigenetic regulation of the DNA repair gene, MGMT. Our data suggest that the ERp29\MGMT axis could be a potential target for the treatment of radioresistant cancer cells.
Radiotherapy is a typical approach for cancer treatment. However, some cancer cells develop resistant capacity against irradiation by various mechanisms, such as increased expression of anti-apoptotic proteins and increased ability of DNA repair 27 . Because cancer cell resistance to radiation is a major hurdle of radiotherapy, understanding the molecular mechanisms of radioresistance and identifying the potential targets to enhance radiosensitivity are fundamental to improve the efficacy of radiotherapy. It was reported that ERp29 knockdown attenuated radioresistance and enhanced cell apoptosis in nasopharyngeal cancer cells 11 . In support of this, we showed that ERp29 expression significantly increased the post-irradiation survival and depletion of ERp29 sensitized to radiation, leading to reduced post-irradiation survival in the basal-like MDA-MB-231 cells (Fig. 1). In addition, previous studies indicated that ERp29 expression conferred resistance to DOX, an agent used for killing cancer cells by causing DNA damage and cell apoptosis 28 , in cancer cells 12 . The current studies further provide evidence to support that ERp29 plays a protective role for DNA integrity from the genotoxic stress induced by DOX or irradiation.
In this study, we investigate the molecular base underlying ERp29-mediated radioresistance in MDA-MB-231 breast cancer cells. Using the methylation PCR array, we have demonstrated that overexpression of ERp29 resulted in hypomethylation of CpG islands in tumour suppressor genes such as CDH1, p16 and MGMT, thereby leading to upregulation of these genes. Further studies revealed that ERp29 expression reduced the expression of DNMT1, a key enzyme causing DNA methylation. To our knowledge, this is for the first time to report a novel function of ERp29 in epigenetic regulation in cancer cells, although the mechanisms of action remain to be fully elucidated. Epigenetic regulation is a complex process involves DNA methylation and histone modification. DNA methylation at CpG islands is carried out by one of 3 DNMTs (DNMT-1, − 3a & − 3b) 29 and is considered to be a very stable epigenetic modification. However, histone modifications are more labile and regulated by histone acetyltransferase and lysine methyltransferases 30 . Generally, modification of histone H3 at lysine residues 9 or 27 (H3K9 or H3K27) is associated with gene silencing, whereas H3K4 methylation correlates with gene activation 31 . Although our current studies demonstrated an involvement of DNMT1 in the reactivation of MGMT by ERp29, other molecular regulation, e.g, histone modifications, should not be excluded.
MGMT is an important DNA repair protein by transferring the methyl group from guanine to its own cysteine residue 32 . Loss of MGMT is unable to remove O 6 -methylguanine (O 6 MeG) which induces DNA-DSBs to trigger apoptosis via signalling involved in ATR/ATM, Chk1 and p53 21 . As such, MGMT plays a significant role in DNA repair and cell survival. We demonstrated that MGMT is a critical player in ERp29-mediated radioresistance in MDA-MB-231 cells. Our data showed that depletion of MGMT by siRNA in MB-231/ERp29 cells remarkably reduced the post-irradiation survival of cells. This is further supported by the fact that reduction of MGMT led to an enhanced expression of γ -H2AX, an indicator of DNA-DSBs 33,34 , in MB-231/ERp29 cells. In line with this, reduction of MGMT also resulted in an increased expression of cleaved caspase 3 and a reduction of cell viability. Our studies confirm that MGMT knockdown reduced the DNA repair capacity and promotes cell apoptosis in MB-231/ERp29 cells after irradiation treatment. Taken together, the ERp29-upregulated MGMT is a key player in preventing radiation-induced DNA damage in MB-231/ERp29 cells.
In summary, our studies demonstrate a novel function of ERp29\MGMT in maintaining DNA integrity and radioresistance in MDA-MB-231 cells. Hence, targeting the ERp29\MGMT could be an effective strategy to overcome resistance to radiotherapy in cancer. To this end, further studies will be undertaken to investigate the radioresistant function of ERp29\MGMT in other cancer cell models and to investigate the association of this ERp29\MGMT axis with radioresistant phenotype in clinical tumour specimens.

Materials and Methods
Antibodies and reagents. The following antibodies were used in this study: rabbit anti-human Irradiation and clonogenic assay. Clonogenic cell survival after irradiation treatment was determined using colony formation assay. Briefly, cells were plated in 6-well plates at appropriate dilutions. After 24 h post-incubation, cells were irradiated with single dose of 1, 2, 4, or 6 Gy at 37 °C with a dose rate 1.7 Gy/min using a Gamma Cell 1000 Elite irradiator (Theratronics, Ottawa, Canada) with a 137 Cs source. Mock-irradiated cells were used as control. The plates were incubated at 37 °C for 10 days to allow colony formation. Colonies were fixed with methanol and stained with crystal violet. Colonies with > 50 cells were counted manually. Clonogenic fractions of irradiated cells were normalized to the plating efficiency of non-irradiated control group. The value of radiation dose D37, a dose required to reduce the number of clonogenic cells to 37%, was used to evaluate radiosensitivity.
Cell viability. Cells were seeded at 3 × 10 3 cells per well in 96 well plates and incubated at 37 °C in a cell culture incubator. After 15 hours post-incubation, the cells were exposed to irradiation (4 Gy), followed by incubation for 48 hours. The viable cells were examined using the Vybrant ® MTT Cell Proliferation Assay Kit (Life Technologies).
Methyl-Profiler ™ DNA Methylation PCRh. Promoter methylation/demethylation was performed using the Methyl-Profiler ™ DNA Methylation PCR Array (SuperArray Bioscience Corporation, Frederick, MD). Briefly, 0.5 μg of total genomic DNA was digested with enzyme A and/or B and the PCR reactions were set up according to the provided protocol. PCR was carried out with the 96-well plate format and the ABI 7500 Fast Real-Time PCR System (Applied Bio systems, Singapore). Gene amplification was detected with SYBR, and data analysis was carried out using the Δ Ct method (http://www.sabiosciences.com/methylationdataanalysis.php). The "Results" worksheet displays the relative percentage of hypermethylated, intermediately methylated and unmethylated DNA in each target promoter sequence. Data represent a mean of percentage of unmethylation level from triplicate experiments.

DNA extracthion, bisulfite treatment and methylation specific PCR (MS-PCR). Genomic
DNA was isolated from cells using the QIAamp DNA Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. The final yield and quality of extracted DNA was measured using Nanodrop 2000c (Thermo Scientific, Wilmington, DE). The DNA (2.5 μ g) was then converted by sodium bisulfite treatment using as per standard protocol using the EZ DNA Methylation ™ kit (Zymo Research Cooperation, Irvine, CA).

RNA interference, MGMT vector and transfection.
For knockdown of target genes by siRNA, For construction of MGMT expression vector, MGMT human cDNA (Origene) were amplified using forward primer 5′ -CGGGATCCGG CACCATGGAC AAGGATTGTG AAATGAAACG-3′ with BamHI; and reverse primer 5′ -CTCTAGATCA GTTTCGGC CAGCAGGCGG GGAGC-3′ with XbaI. MGMT was then subcloned into pCDNA3.1 (Invitrogen ™ ) after digestion with BamHI/XbaI. For transient transfection, MDA-MB-231/ERp29 siRNA cells incubated in a 6-well plate were transfected with pcDNA-MGMT or pcDNA at 2 μ g per well for 24 hours using lipofectamine ™ 2000 (Invitrogen ™ ) and the expression of MGMT was examined by immunoblot. In parallel, the pcDNA-MGMT and pCDNA-transfected cells were used for radiation treatment and clonogenic assay. Statistical analyses. All data shown are the results of at least three independent experiments and are expressed as the mean ± standard deviation (SD). Student's t tests (two-tailed) were used to analyse the significance of differences between groups. P < 0.05 was considered as significant.