Promoter Methylation of the MGRN1 Gene Response to Chemotherapy of Epithelial Ovarian Cancer Patients

Objective: Aberrant DNA methylation is considered to play a critical role in the chemoresistance of epithelial ovarian cancer (EOC). In this study, we explored the relationship between hypermethylation of the Mahogunin Ring Finger 1 (MGRN1) gene promoter and primary chemoresistance in EOC patients. Methods: Hypermethylation of the MGRN1 promoter region in the cancer tissues of platinum-resistant EOC patients was observed by genome-wide methylation level analysis. Matrix-assisted laser desorption/ionization time-of-ight (MALDI-TOF) mass spectrometry was used to analyze the methylation level of the MGRN1 promoter region. MGRN1 mRNA and protein expression were examined using RT-qPCR and IHC assays, respectively. The effect of MGRN1 methylation on the cellular response to cisplatin was detected by knockdown assays in SKOV3 cells. Additionally, we performed transcriptome analysis using RNA-seq and explored the possible mechanism by which MGRN1 expression affects the resistance of ovarian cancer cells to platinum. Results: The RRBS assay showed that the upstream region of MGRN1 from -1148 to -1064 was signicantly hypermethylated in chemoresistant EOC patients (P=1.78×10 −7 ). The MALDI-TOF mass spectrometry assays revealed a strong association between hypermethylation of the MGRN1 upstream region and platinum resistance in patients with EOC. Spearman’s correlation analysis revealed a signicantly negative connection between the methylation level of MGRN1 and its expression in EOC. In vitro analysis demonstrated that knockdown of MGRN1 reduced the sensitivity of cells to cisplatin and that expression of EGR1 was signicantly decreased in SKOV3 cells with low levels of MGRN1 expression. Similarly, EGR1 mRNA expression was lower in platinum-resistant EOC patients and was positively correlated with MGRN1 mRNA expression. Conclusion: The hypermethylation of the MGRN1 promoter region and low expression of MGRN1 were associated with platinum resistance in EOC patients.


Introduction
Due to the lack of initial symptoms and sensitive screening methods, approximately 70% of women with epithelial ovarian cancer (EOC) are diagnosed at an advanced stage of disease [1,2]; EOC is the most lethal gynecologic malignancy in China [3,4]. Currently, the treatment strategy for patients with advanced EOC is platinum-based chemotherapy following primary debulking surgery [5][6][7]. Although the majority of EOC patients respond well to rst-line chemotherapy, most of these patients relapse and develop platinum resistance within 2 years [7,8]. In addition, nearly 20% of patients do not respond at the beginning of chemotherapy [9][10][11]. Therefore, chemotherapy resistance has become an important cause of high mortality among EOC patients.
Resistance to chemotherapeutics, including intrinsic and acquired resistance, is based on highly complex and individually variable biological mechanisms [12]. Abnormal methylation of DNA has been considered to play an important role during the development of acquired chemoresistance in EOC patients [13].
However, there are currently very few studies about the effect of DNA methylation on the development of intrinsic resistance in EOC patients. We observed abnormal hypermethylation in the promoter region of the Mahogunin Ring Finger 1 (MGRN1) gene in the ovarian cancer tissues of EOC patients with intrinsic resistance by reduced representation bisul te sequencing (RRBS) (Tian et al. 2020). MGRN1 is an intracellular C3HC4 RING nger domain protein that exhibits E3 ubiquitin ligase activity and plays critical roles in the control of protein degradation [14]. Ubiquitin-mediated proteolysis has played a crucial role in controlling protein level homeostasis and regulating the cell cycle, cell proliferation, apoptosis and DNA damage responses, which are involved in tumorigenesis, tumor development, prognosis and drug resistance [15,16]. However, the role of MGRN1 in tumorigenesis, tumor progression, and drug responses is not currently well understood. A study by Dugué et al suggests that hypomethylation of MGRN1 CpG sites in peripheral blood DNA is associated with the development of sporadic and familiar breast cancer [17].
Based on the results of RRBS, this study investigated the role of hypermethylation of the MGRN1 upstream region in platinum resistance in EOC patients. First, we examined the level of MGRN1 methylation and expression in platinum-resistant and platinum-sensitive EOC patients. Furthermore, we also investigated the possible role and mechanism of decreased MGRN1 expression in ovarian cancer cells in the response to cisplatin in vitro. To the best of our knowledge, this is the rst study to investigate the role of the methylation status of the MGRN1 promoter region in platinum resistance in EOC patients.  Table 1. The inclusion criterion for cases was histologically con rmed primary EOC. The exclusion criteria was a history of chemotherapy therapy before surgery. According to the NCCN guidelines, recurrent disease was identi ed clinically (i.e., pelvic pain and weight loss), biochemically (i.e., elevated CA-125 levels), and/or with imaging [18]. Based on the platinum-free interval (PFI), which was calculated from the date of the last platinum compound treatment to the date of disease progression, all the study participants were divided into a platinum-sensitive group (n=92) and a platinum-resistant group (n=62). PFI of less than 6 months is widely used to clinically de ne platinum-resistant disease, whereas a PFI greater than 6 months is often used to de ne platinum-sensitive disease [19]. The participants were regularly followed-up for 5 years. Total RNA was isolated from 96 epithelial ovarian cancer tissue samples using the TRIzol-chloroform extraction method (Generay Biotech Co., Ltd., Shanghai, China), as described by the manufacturers. The total cDNA was reverse-transcribed using the Revert Aid First-Strand cDNA Synthesis Kit (Thermo Scienti c, USA). The speci c primers for the target genes that were used in RT-qPCR were designed using Primer Premier 5.0 and produced by Sangon Biotech Co., Ltd. (Shanghai, China). GAPDH was used as the housekeeping gene. The primer sequences for PCR amplification were as follows: MGRN1 forward, 5'-TACAAAGACGATG CCGACAG-3'; MGRN1 reverse, 5'-GCCTGGCAGTAGATGGTGAT-3'; GAPDH forward, 5'-AATCCCATCACCATCTTCCA-3'; and GAPDH reverse, 5'-TGGACTCCACGACGTACTCA -3'. The reactions were run with the QuantiNova TMSYBR® Green PCR Kit (Qiagen, Hilden, Germany) in an Mx3005P instrument. The comparative quanti cation of each target gene was performed based on the cycle threshold (Ct) and normalized to GAPDH using the 2 -ΔCt method.

MGRN1 immunohistochemical (IHC) study of the clinical samples
Of the 96 epithelial ovarian cancer samples, 86 para n-embedded epithelial ovarian cancer tissue samples collected in the pathology department of the Fourth Hospital of Hebei Medical University were used for immunohistochemical (IHC) staining of MGRN1. MGRN1 immunostaining was performed using a primary antibody, namely rabbit antihuman MGRN1 (RNF156, 1:500 dilution; Proteintech, China).
Briefly, 4-μm thick sections were dewaxed in xylene and dehydrated through a graded series of ethanol. After blocking endogenous peroxidase and non-specific binding, the sections were incubated overnight at 4°C with primary antibodyand then with biotinylated secondary antibody and streptavidin-peroxidase complex. After the sections were washed in PBS, they were incubated with DAB reagent and counterstained with haematoxylin. Negative control sections were incubated with PBS instead of primary antibody. The immunohistochemical staining was evaluated using a previously reported scoring method [21]. The immunoreactivity of MGRN1 was considered to be positive in tumor cells showing cytoplasmic staining without nuclear staining. The sections were independently examined by two pathologists, who were blinded to the clinicopathological information.

Cell culture
The human serous ovarian cancer SKOV3 cell line was purchased from the iCell Bioscience Inc.
(Shanghai, China). The SKOV3 cell line was cultured in Dulbecco's modi ed Eagle's medium (DMEM) (Gibco; Thermo Fisher Scienti c, Inc.) The medium was always supplemented with 10% (w/v) fetal bovine serum, 100 U penicillin, and 100 µg/L streptomycin (Gibco; Thermo Fisher Scienti c, Inc.). The cells were maintained in a 95% humidi ed and 5% CO 2 atmosphere at 37°C. All the experiments were performed in triplicate.

Stable cell lines
MGRN1 expression plasmids and lentiviral packaging reagents and shRNA were purchased from Genecopoeia Inc. (MD, USA). The designed three target sequences in the MGRN1 gene were 5′-GGAAACTACTTTGCTTCGCAC-3′ (shRNAa); 5′-GCGTGTTTCCAGTAGTCATC C-3′ (shRNAb) and 5′-GGCATTGAGAACAAGAACAAC-3′ (shRNAc). The most effective construct, recombinant plasmid inserted with MGRN1 gene shRNA expression vector shRNAa was selected for the study. A random sequence of shRNA (shNC) was used as the negative control. SKOV3 cells were seeded in a six-well plate at a density of 4×10 5 cells/mL in a volume of 2mL/well. When the SKOV3 cells reached 70-80% confluence, they were transfected with shRNA. Transfection of the SKOV3 cell line was performed according to the manufacturer's protocol.

Detection of changes in MGRN1 via Western blotting
Proteins were isolated using RIPA lysis buffer. The total proteins were extracted, and a BCA protein assay kit (Thermo) was used to quantify protein concentration. Rabbit anti-human MGRN1 antibody (RNF156, Proteintech, China) and β-actin (ab8226, Abcam, Cambridge, UK) were used as the primary antibody. Antirabbit IgG was used as the secondary antibody (diluent ratio of 1:5000; Proteintech, China). The antigenantibody reaction was visualized by detection with an odyssey assay (ECL, Millipore, Billerica, MA).

Cell viability assays
The cells were inoculated in 96-well microplates in medium containing 10% fetal bovine serum and penicillin/streptomycin. After overnight incubation, the cells were treated with cisplatin (P zer), returned to the incubator for 24 h, and then analyzed. Cell Counting Kit-8 (CCK-8) was used to measure cell activity. Ten microliters of CCK-8 was added to each well and incubated for 3 h (37°C; 5% carbon dioxide). Then, the absorbance was measured at 492 nm with a microplate reader. Each experiment was repeated three times.

apoptosis assays
The cells were collected after 24h of cisplatin treatment. The Annexin V Apoptosis Detection kit I (BD Biosciences, Franklin Lakes, NJ, USA) was used to analyze the apoptosis of the SKOV3 cells. Brie y, cells were seeded into 6-well plates. After treatment with drugs, the adherent cells were trypsinized without EDTA and collected by centrifugation. After washing with PBS two times, the cells were resuspended in 100μL of 1×binding buffer and were subsequently incubated with 5μL of Annexin V staining solution at room temperature for 30min in the dark. Then, 400μL of 1×binding buffer was added, and the uorescence intensity was evaluated on a FACS Aria™ (BD Biosciences) ow cytometer. Each assay was performed in triplicate.

RNA sequencing
This experiment was conducted at Differential Gene Technology Co., Ltd. (Anhui, China). EdgeR was used to identify the top ten enriched annotation terms among the differentially expressed genes (1.5-fold in either direction, P<0.05) between the SKOV3 sh-NC group and the SKOV3 sh-MGRN1 group.

Statistical analysis
The statistical analyses were performed using SPSS 21.0 (Chicago, IL, USA). The Wilcoxon rank sum test was used to compare the methylation level and mRNA expression of MGRN1 between the two groups. The χ2 test was used to compare the protein expression of MGRN1 in each group. Spearman correlation analysis was performed to analyze the correlation between MGRN1 expression and methylation status. A t test was used to analyze the cell activity and apoptosis data.

MGRN1 promoter methylation levels in the platinum-resistant group and platinum-sensitive group
In our previous study, we used the RRBS assay to compare the differences in the genome-wide methylation patterns between 8 platinum-resistant EOC patients and 8 platinum-sensitive EOC patients. The results showed that a region from − 1148 to -1064 within the promoter of MGRN1 was signi cantly hypermethylated in the platinum-resistant group compared to the platinum-sensitive group (Fig. 1A). To further con rm the results of the RRBS assay, MALDI-TOF mass spectrometry was used to examine the methylation levels of this region in 18 platinum-resistant EOC patients and 22 platinum-sensitive EOC patients. In the present study, we tested the methylation levels of ve CpG sites (-1148, -1118, -1107, -1097 and − 1064 from the transcription start site) within this region. The analysis revealed that methylation levels of two CpG sites (-1107 and − 1097) were signi cantly higher in the tumor tissues of platinum-resistant EOC patients than in those of platinum-sensitive EOC patients (P = 0.01, 0.04, Fig. 1B).
MGRN1 expression in the platinum-resistant group and platinum-sensitive group RT-qPCR was used to determine the mRNA levels of MGRN1 in the tumor tissues from 41 platinumresistant EOC patients and 55 platinum-sensitive EOC patients. The results showed that the mRNA level of MGRN1 in platinum-resistant EOC patients was 1.20-fold lower than that in platinum-sensitive EOC patients (P < 0.01, Fig. 1C). Furthermore, IHC analysis was conducted to examine the protein expression of MGRN1 in 36 platinum-resistant EOC patients and 50 platinum-sensitive EOC patients. The analysis results showed that the frequency of positive MGRN1 expression in platinum-resistant EOC patients was signi cantly lower than that in platinum-sensitive EOC patients (P = 0.02, Table 2). IHC staining showed that the MGRN1 protein was mainly expressed in the cytoplasm of EOC tissues (Fig. 1D). Silencing of MGRN1 expression in serous ovarian cancer cells by shRNA To investigate the role of MGRN1 expression in the sensitivity of ovarian cancer cells to cisplatin, SKOV3 cells were transfected with shRNAa-MGRN1, shRNAb-MGRN1, shRNAc-MGRN1 plasmid or shNC plasmid, respectively. After transfection for 48 hours, the expression of MGRN1 was con rmed by RT-qPCR. As shown in Fig. 2A, shRNAa-MGRN1 could effectively decrease MGRN1 expression in SKOV3 cells, as compared with shNC groups. We also con rmed the expression of MGRN1 in shRNAa-MGRN1 group by Western blot. Therefore, we established MGRN1 stable knockdown cell lines using shRNAa-MGRN1 (P < 0.05, Fig. 2A, B).
Effect of MGRN1 knockdown on the cellular response to cisplatin CCK-8 assays were used to compare the cell proliferation of the shRNA-MGRN1 group and shNC group. The proliferation rate was signi cantly higher in the shRNA-MGRN1 group than in the shNC group after treatment with cisplatin at several concentrations for 24 h (P < 0.05, Fig. 2C). In addition, ow cytometry analysis demonstrated that the percentage of apoptotic cells in the shRNA-MGRN1 group was signi cantly lower than that in the shNC group after exposure to 10 µM cisplatin (P = 0.03, Fig. 2D).

RNA-seq analysis reveals that EGR1 expression is differentially regulated by MGRN1 in ovarian cancer cells
To gain a better understanding of the differential regulation of transcription between the shNC transfection group and shRNA-MGRN1 transfection group, we performed an RNA-seq analysis of the total RNA harvested from the shNC group and shRNA-MGRN1 group. The top ten annotated, protein-coding genes that were differentially regulated in the shNC group compared to the shRNA-MGRN1 group are shown in Table 3 (P < 0.05). Of these genes, EGR1 was the most differentially expressed, with 4.26-fold lower expression in the shRNA-MGRN1 group than in the shNC group (P = 8.65E − 07 ). EGR1 is a key gene involved in regulating cell proliferation and apoptosis in a variety of cancer tissues, and knockdown of EGR1 has been shown to promote resistance to cisplatin. Thus, we further validated the mRNA levels of EGR1 in cells and tissues by quantitative reverse-transcription PCR (RT-qPCR). The results showed that EGR1 mRNA expression was reduced by 72% in the SKOV3 shRNA-MGRN1 group compared with the SKOV3 shNC group (P < 0.01, Fig. 3A). The mRNA level of EGR1 in 41 platinum-resistant EOC patients was 1.15-fold lower than that in 55 platinum-sensitive EOC patients (P = 0.02, Fig. 3B). Spearman's correlation analysis showed that MGRN1 mRNA expression was signi cantly positively correlated with EGR1 mRNA expression (r = 0.379, P = 0.01).

Discussion
In this study, we con rmed that MGRN1 gene promoter hypermethylation is associated with platinum resistance in patients with EOC based on the following ndings: 1) the upstream region of MGRN1 was signi cantly hypermethylated and lower expression in the cancer tissues of platinum-resistant patients with EOC, 2) MGRN1 mRNA expression was signi cantly negatively correlated with the methylation level of the MGRN1 promoter region, 3) knockdown of MGRN1 expression could desensitize SKOV3 ovarian cancer cells to cisplatin, 4) knockdown of MGRN1 expression in SKOV3 cells could signi cantly reduce EGR1 mRNA expression, which signi cantly correlated with the resistance in platinum-treated cancer patients.
In the analysis of genome-wide methylation levels in tissue samples from 8 platinum-resistant and 8 platinum-sensitive EOC patients, we found that the methylation level of the MGRN1 upstream region (-1148 to -1064) was signi cantly higher in the platinum-resistant group. Mass spectrometry analysis of an expanded EOC sample size showed that hypermethylation of the MGRN1 promoter region was associated with platinum resistance in EOC patients. We also discovered that the expression levels of MGRN1 mRNA and protein in platinum-resistant EOC patients were signi cantly lower than those in platinum-sensitive EOC patients. Correlation analysis indicated that the methylation level of the MGRN1 promoter region was associated with MGRN1 mRNA expression. Furthermore, knockdown of MGRN1 expression could increase proliferation and decrease apoptosis in SKOV3 cells challenged with cisplatin. These ndings suggested that lower MGRN1 expression due to hypermethylation of its promoter region might induce platinum resistance in EOC.
MGRN1, an E3 ubiquitin ligase of the Really Interesting New Gene (RING) nger family, is involved in many biological and cellular mechanisms [22]. However, there is no study of the role of MGRN1 in chemotherapy resistance in cancer patients to date. In the current study, microarray analysis of total RNA showed that knockdown of MGRN1 expression in SKOV3 cells resulted in signi cant downregulation of multiple genes, including early growth response protein 1 (EGR1). EGR1 is a transcription factor that can be induced by a variety of stimuli or stressors, including growth factors, hormones, ionizing radiation, and chemotherapy drugs [23][24][25], and plays essential roles in cell proliferation and apoptosis [26][27][28].
Knockdown of EGR1 expression can decrease cisplatin-induced apoptosis in a variety of cancer cells [29][30][31], while Kim overexpression of this gene sensitizes ovarian cancer cells to cisplatin-induced apoptosis [30]. He et al found that EGR1 expression levels were signi cantly higher in ovarian cancer tissues with low ERCC1 expression than in ovarian cancer tissues with high ERCC1 expression, suggesting that EGR1 expression is positively correlated with potential cisplatin-sensitive ovarian cancer, since ERCC1 is widely accepted as a biomarker of platinum resistance [30]. In our study, it was also observed that the expression of EGR1 in EOC patients with platinum resistance was signi cantly downregulated and was positively correlated with the expression of MGRN1. Therefore, we speculate that MGRN1 may affect the platinum resistance of ovarian cancer by regulating the expression of EGR1. Of course, the molecular mechanism by which MGRN1 regulates EGR1 requires further study.
In summary, our study demonstrated that the hypermethylation of MGRN1 may be predictive of platinum resistance in EOC patients. Considering that DNA methylation may be used as a molecular marker for ovarian cancer chemotherapy, we believe that our ndings warrant con rmation in a larger patient cohort and could facilitate patient selection for chemotherapy.