Abstract
Tamoxifen is a standard anti‐hormone treatment in estrogen receptor positive breast carcinoma patients. Unfortunately, about 50% of patients relapse during treatment. Promoter hypermethylation contributes to the epigenetic modulation of tamoxifen resistance‐related genes. To evaluate the contribution of DNMTs expression and their promoter methylation as diagnostic biomarkers in development of breast malignancy and tamoxifen resistance, the present study was designed and 107 breast tumors and normal breast tissues were recruited. Methylation‐specific high‐resolution melt curve analysis and quantitative RT‐PCR were performed to evaluate DNMTs promoter methylation and mRNA expression, respectively. Our results indicated that DNMT3A and DNMT3B promoters were demethylated in breast tumors as compared to control tissues. The mRNA expression levels of all three DNMTs were significantly increased in tumor specimens in comparison to control tissues (p < 0.05). Among tumor tissues, DNMT3A promoter methylation was significantly higher in tamoxifen sensitive patients (p = 0.001). Overexpression of DNMT3A (p = 0.037) and DNMT3B (p < 0.001) mRNA were observed in tamoxifen resistance group. Multivariate logistic regression analysis indicated that low methylation status of DNMT3A and overexpression of DNMT3B could be as independent predictors of disease recurrence. Multivariate Cox regression analysis, revealed that high methylation status of DNMT3A could be an independent and favorable predictor for disease free survival (p = 0.002) and overall survival (p = 0.026); high expression of DNMT1 (p = 0.03) remained significant and unfavorable predictive factor for overall survival. In conclusion, our data for the first time indicated that low methylation status of DNMT3A promoter and overexpression of DNMT3B could contribute to disease recurrence in tamoxifen‐treated breast cancer patients.
Keywords: breast cancer, DNA methyltransferase, high‐resolution melt curve analysis, promoter methylation, tamoxifen resistance
In order to find new diagnostic biomarkers for progression of tamoxifen resistance in breast cancer patients, 107 fresh frozen breast tumors and normal breast tissues were recruited and the present study was designed. Promoter methylation status and mRNA expression of DNMT1, DNMT3A, and DNMT3B were assessed in tamoxifen resistant and tamoxifen sensitive breast tumor tissues and normal breast tissue specimens. The results indicated that low methylation status of DNMT3A promoter and overexpression of DNMT3B could contribute to disease recurrence in tamoxifen‐treated breast cancer patients.
1. INTRODUCTION
Breast cancer is the most commonly occurring (31% of all female cancers) and the second cause of cancer related death in women (Tanic et al., 2012). Improvement of screening, detection of prognosis marker and different treatment options including radiotherapy, chemotherapy, hormone therapy, and targeted therapy leads to continuously decline in death rate in breast cancer patients. About 70% of breast tumors are Estrogen Receptor positive (ER+) (Lu et al., 2016). Estrogen plays important roles in initiation and progression of breast malignancies, therefore the common therapeutics on ER+ patients aimed to reduce estrogen level or blocking ERα signaling pathway (Jiang, Zheng, & Wang, 2013). Tamoxifen is one of the most effective hormone therapy choices which behaves as an estrogen antagonist agent and commonly included in treatment protocol of ER+ patients. Tamoxifen pharmacologically classify as a pioneer selective estrogen receptor modulator (SERM) (Wu et al., 2016). Although this drug saved innumerable patient lives for about four decades(Lin, Zhang, & Manson, 2011), however the most challenging issue with tamoxifen treatment is disease recurrence during treatment; unfortunately about 50% of patients experience resistance initially (de novo resistance) or over time (acquired resistance) (Schiff, Massarweh, Shou, & Osborne, 2003). The molecular mechanisms underlying tamoxifen resistance are not completely clear, however, two main mechanisms have been confirmed documentary including individual variability in the genetic (Rofaiel, Muo, & Mousa, 2010) and epigenetic (Dannenberg and Edenberg, 2006). Epigenetic processes mainly include DNA methylation, histone modifications, and microRNAs (miRNAs). Both genetic and epigenetic modifications could lead to initiation and progression of different malignancies and drug resistance during treatment; however epigenetic modifications are reversible which made it attractive research field to find novel treatment based on epigenetic therapy. Recent epigenetic studies have proven that CpG islands promoters’ hypermethylation could be included as a common process which contributed to switch off or downregulation of tumor related genes. DNA methylation could provide important biomarkers for cancer detection, diagnosis, prognosis prediction, and treatment response (Chen et al., 2015). Aberrant DNA methylation is associated with disease aggressiveness and resistance progression in tamoxifen‐treated breast cancer patients with the involvement of hormone receptor promoters and tumor suppressor genes promoters (Altundag, Altundag, & Gunduz, 2004; Phuong et al., 2011). Furthermore, tamoxifen is a prodrug and requires bioactivation to exhibit therapeutic effects. The pharmacologically active metabolites of tamoxifen are 4‐hydroxy‐tamoxifen and 4‐hydroxy‐N‐desmethyltamoxifen (endoxifen). Promoter hypermethylation which results in downregulation of drug‐metabolizing enzymes (cytochrome P450) also could lead to tamoxifen resistance (Widschwendter et al., 2004). In addition, methylation analyses have revealed that tamoxifen resistant cell lines share 3,000 hypermethylated CpGs (Williams, Anderton, Lee, Pentecost, & Arcaro, 2014). DNA methyltransferases (DNMTs) are a group of enzymes which are responsible for maintenance (DNMT1) and de novo (DNMT3A, DNMT3B) DNA methylation (Das & Singal, 2004). Regarding the presence of CpG islands in the promoter region of tamoxifen resistance‐related genes and susceptibility for switch off by DNMTs mediated promoter methylation; we proposed that overexpression of DNMTs (DNMT1, DNMT3A, and DNMT3B) in primary tumors may be induce promoter hypermethylation and downregulation of tamoxifen resistance‐related genes. Besides, the high CpG content of DNMTs promoters' points to methylation‐mediated regulation of DNMTs expression. To our knowledge, there are no reports that compare the DNMTs promoter methylation and expression in tamoxifen‐treated breast cancer patients. Therefore, the present study was designed to assess DNMTs mRNA expression in TAM‐R and TAM‐S and normal breast tissues to understand if overexpression of DNMTs associated with progression of breast cancer and resistance incidence during tamoxifen treatment period. The promoter methylation of DNMTs also measured to find out if there is any relation between DNA methylation pattern and mRNA expression of DNMTs. Finally, relation of the obtained data with DFS (disease free survival) and OS (overall survival) were analyzed to indicate if DNMTs expression and promoter methylation status are contributed to disease progression or not.
2. MATERIALS AND METHODS
2.1. Tissue selection and patients’ characteristics
Surgical pathology records of the breast cancer patients of Iran National Tumor Bank who experienced breast surgery with lymph node dissection from 2005 to 2014 were reviewed. All patients with complete assessed clinicopathological data were selected. ER‐positive ductal carcinoma breast patients, who had undergone of surgery, radiotherapy, chemotherapy, and finally received adjuvant tamoxifen as anti‐hormone treatment for 6 months to 5 years or more, were included. ER‐negative breast patients and patients with prior neoadjuvant therapy were excluded from the present research. A retrospective case–control study which was comprised of 72 tumors and 30 corresponding normal adjacent (15 adjacent normal of tamoxifen resistant and 15 adjacent normal of tamoxifen sensitive patients) and five normal breast tissues (which was resected from patients who underwent breast cosmetic plastic surgery) was designed. Recruitment of two kinds of normal breast tissues help us to understand if they are basically different regarding DNMTs expression and promoter methylation or not. Fresh‐frozen breast tissue samples and their matched formalin‐fixed paraffin‐embedded (FFPE) tissue specimens were obtained from Iran National Tumor Bank. All frozen tissues were kept at liquid nitrogen except during grossing and sectioning, and immediately returned to liquid nitrogen when possible. One section from each fresh tissue sample was retrieved and stored at −70 °C for nucleic acids extraction. All FFPE specimens of the same tumors were stained with Hematoxlyn and Eosin (H&E) and rechecked by two independent pathologists and histological diagnosis was confirmed. Furthermore, all tissue specimens were undergone immunohistochemical analysis and ER, PR, HER‐2, and p53 scoring were elucidated previously.
Tumor tissues were divided into two groups: tamoxifen resistant (TAM‐R) and tamoxifen sensitive (TAM‐S) regarding disease recurrence during tamoxifen treatment period. All TAM‐R patients experience disease recurrence (local or regional recurrence, distant metastasis e.g., bones, liver, lungs metastasis or death) while receiving tamoxifen treatment (at least 6 months), median time to recurrence was 25 months. All patients in TAM‐S group had received standard adjuvant tamoxifen treatment for 5 years or more without any signs of disease recurrence. Median follow up for tamoxifen sensitive patients was 85 months and patients' general health conditions were evaluated every 6 months. All of the patients were followed up until August, 2016. Among TAM‐R patients, 17/36 (47%) experience metastasis and 19/36 (53%) died during their tamoxifen‐treatment period.
The informed consents have been obtained from each recruited patient for medical record review and tissue sample donation before surgery by Iran Tumor Bank previously. This study was approved by the Ethics Committee of Mashhad University of Medical Sciences (MUMS). The clinicopathological features of patients are displayed in Table 1. Patients in TAM‐S and TAM‐R groups were matched for clinicopathological features including age at diagnosis time, histologic grade, T stage, estrogen, and progesterone receptors, HER‐2 and P53 expression status. However, TAM‐R patients would more likely have a lymph node metastasis (p = 0.021) and perineural invasion (PNI) (p = 0.031) in comparison to TAM‐S group.
Table 1.
Clinicopathological characteristics of tamoxifen‐treated breast cancer patients
| Feature | Categories | Number of tissues | TAM‐S | TAM‐R | p‐value |
|---|---|---|---|---|---|
| Mean age at diagnosis time ± SD a | 72 | 44.28 ± 8.46 | 48.21 ± 10.54 | 0.118 | |
| Histologic grade (MBR b ) | |||||
| Grade I | 25 | 16 (44.4) | 9 (25.0) | 0.21 | |
| Grade II | 34 | 15 (41.7) | 19 (52.8) | ||
| Grade III | 13 | 5 (13.9) | 8 (22.2) | ||
| T stage | |||||
| T1 | 9 | 4 (11.1) | 5 (13.9) | 0.494 | |
| T2 | 44 | 25 (69.4) | 19 (52.8) | ||
| T3 | 17 | 6 (16.7) | 11 (30.6) | ||
| T4 | 2 | 1 (2.8) | 1 (2.8) | ||
| N stage | |||||
| N0 | 22 | 14 (38.9) | 8 (22.2) | 0.021 | |
| N1 | 21 | 12 (33.3) | 9 (25) | ||
| N2 | 33 | 9 (25) | 9 (25) | ||
| N3 | 11 | 1 (2.8) | 10 (27.8) | ||
| ER‐ status | |||||
| Positive | 72 | 36 (50) | 36 (50) | − | |
| Negative | 0 | ||||
| PR status | |||||
| Positive | 47 | 24 (66.7) | 23 (63.9) | p > 0.99 | |
| Negative | 25 | 12 (33.3) | 13 (36.1) | ||
| HER‐2 status | |||||
| Positive | 19 | 11 (30.6) | 8 (22.2) | 0.594 | |
| Negative | 53 | 25 (69.4) | 28 (77.8) | ||
| P53 status | |||||
| Positive | 23 | 14 (38.9) | 9 (25.0) | 0.312 | |
| Negative | 49 | 22 (61.1) | 27 (75.0) | ||
| Ductal Carcinoma In Situ (DCIS) histology | |||||
| Comedo type | 9 | 4 (11.1) | 5 (13.9) | 0.5 | |
| Non‐Comedo type | 63 | 32 (88.9) | 31 (86.1) | ||
| Nipple Involvement | |||||
| Present | 13 | 6 (16.7) | 6 (16.7) | p > 0.99 | |
| Absent | 59 | 30 (83.3) | 30 (83.3) | ||
| Lymphatic invasion | |||||
| Present | 55 | 25 (69.4) | 30 (83.3) | 0.267 | |
| Absent | 17 | 11 (30.6) | 6 (16.7) | ||
| Perineural invasion (PNI) | |||||
| Present | 30 | 10 (27.8) | 20 (55.6) | 0.031 | |
| No | 42 | 26 (72.2) | 16 (44.4) | ||
| Extracapsular Nodal Extension (ECE) | |||||
| Present | 15 | 4 (11.1) | 11 (30.6) | 0.079 | |
| Absent | 57 | 32 (88.9) | 25 (69.4) |
Standard Deviation.
MBR: Modified‐Bloom‐Richardson.
2.2. Genomic DNA extraction and sodium bisulfite modification
DNA extraction was performed using QIAmp DNA mini kit (Qiagen, Hilden, Germany) according to the manufacturer's protocol. The purity and concentration of extracted DNA was analyzed by NanoDrop 2000C Spectrophotometer (Thermo Fisher Scientific, Waltham, MA). The absorbance ratio at 260/280 and 260/230 of each DNA sample was around 1.8 and 1.8–2.1, respectively. Samples with final concentrations more than 200 ng/μl were accepted for further experiments. Performance of agarose gel electrophoresis indicated the integrity of each DNA sample. EZ DNA Methylation‐GoldTM Kit (Zymo Research, Irvine, CA) was used for Bisulfite conversion of genomic DNA according to the manufacturer's instruction. All unmethylated cytosines convert to uracils during bisulfite treatment while methylated cytosines remain unaffected. Bisulfite converted DNA was stored at −20 °C until methylation specific HRM experiments.
2.3. Methylation‐specific high‐resolution melt curve analysis (MS‐HRM)
DNA methylation status was evaluated using a semi‐quantitative method (MS‐HRM). Fully methylated human DNA control was diluted with different concentrations of unmethylated bisulfite‐converted DNA control (EpiTect PCR control DNA Set, Qiagen, Hilden, Germany) to provide controls with different methylation percentages (0, 25, 50, 75, and 100%). The promoter methylation status of each DNA sample was evaluated using comparison of its melting curve with melting curves of control DNAs. HRM primer pairs were designed according to MS‐HRM primer design protocol which expanded by Wojdacz, Borgbo, and Hansen (2009) to specifically targeting bisulfite converted methylated or unmethylated DNA. MethPrimer (http://www.urogene.org/methprimer/index1.html) (a free online tool which specifically designs primer for bisulfite converted sequences in methylation studies) was used for primer design. MS‐HRM primer sequences were as follow: DNMT1 (NC_000019.10) fwd: (5′‐ GTTGGGGTTGGGGTTGTTAGT‐3′) and rev: (5′‐AAAAAACCCTCCRTACAACTCC‐3′); DNMT3A (NC_000002.12) fwd: (TGCGTTTTGTTTGTTTTAGTATTGG‐3′) and rev: (5′‐TCAAAATACRAAAAACTCACTCC‐3′); DNMT3B (NC_000020.11) fwd: (5′‐GCGGTGTCRATTTTTTTTGTAG‐3′) and rev: (5′‐ CCCCCAAAATAATTTCACTAACC‐3′). The amplified sequences were 212, 209 and 226 base pair for DNMT1, DNMT3A, and DNMT3B, respectively.
MS‐HRM was used to detect the methylation levels of DNMTs promoters. Each MS‐HRM experiment involved of standard curves with known methylation ratios including 100, 75, 50, 25, and 0%. PCR amplification and HRM analyses were performed using LightCycler ® 96 Instrument‐Roche. Each reaction involved of 2 μl of bisulfite converted DNA, 1 μl of each primer (10 pmol), 4 μl of master mix (5× HOT FIREPol® EvaGreen® qPCR Mix Plus (ROX)); at final volume of 25 μl. The condition of MS‐HRM amplification was as follow: pre‐incubation at 95 °C for 12 min which followed by 42 cycles of denaturation at 95 °C (30 s), appropriate annealing temperature for each primer pairs (30 s) and final extension at 72 °C (30 s). Finally, HRM analysis was performed with one step of 95 °C for 60 s, 40 °C for 60 s, 60 °C for 1 s, and continuous acquisition to 97 °C at one acquisition per 0.07 °C.
2.4. RNA extraction and cDNA synthesis
Total RNA was extracted with RiboEx Total RNA (301–001) in clean RNase‐free tubes according to the manufacturer's instruction. Isolated RNAs were eluted in DEPC (Diethylpyrocarbonate) treated water. The concentration and purity of extracted RNA were analyzed by NanoDrop 2000C Spectrophotometer (Thermo Scientific). The absorbance ratio at 260/280 and 260/230 of each RNA sample was around 2 and 1.9–2.2, respectively. In agarose gel electrophoresis 28S, 18S, and 5S ribosomal RNA were visible which indicated RNA integrity. The extracted RNA samples were treated with DNase I, RNase free enzyme, Thermo Scientific, (EN0521). Complementary DNA (cDNA) was synthesized with random hexamer primers using RevertAid First Strand cDNA Synthesis Kit, Fermentas Life Science, (K1622) using 2 µg of total RNA in a final reaction volume of 20 µl.
2.5. Assessment of specific primers for real time PCR experiments
Most of human genes included in DNMTs are alternatively spliced which translate into different protein isoforms (de la Grange, Dutertre, Correa, & Auboeuf, 2007). All of the splicing variants of each gene were detected in NCBI (National Center for Biotechnology Information, Bethesda, MD), as follow: DNMT1 (NM_001318731.1, NM_001318730.1, NM_001130823.2, NM_001379.3); DNMT3A (NM_001320893.1, NM_153759.3, NM_022552.4, NM_175629.2); DNMT3B (NM_001207056.1, NM_001207055.1, NM_175850.2, NM_175849.1, NM_175848.1, NM_006892.3); afterward the common exons were selected for further study and primer design experiments. Specific primers were designed using Beacon Designer 8 software, in a way that to target all of the isoforms of desired genes. Primer pairs were designed specifically to span exon‐exon junctions, finally, a BLAST search (http://blast.ncbi.nlm.nih.gov/Blast.cgi) and in silico PCR analysis (https://genome.ucsc.edu/cgi‐bin/hgPcr), and amplicon secondary structure analysis (http://unafold.rna.albany.edu/?q=mfold/DNA‐Folding‐Form) were performed to detect primer specificity and the feasibility of Real Time PCR experiments. All primers were synthesized by Metabion (Germany) and their sequences were as follow: DNMT1 fwd: (5′‐ GCAAACCACCATCACATCTCAT‐3′), DNMT1 rev: (5′‐ GTCTAGCAACTCGTTCTCTGGA‐3′); DNMT3A fwd: (5′‐ ACCACGACCAGGAATTTGACC‐3′), DNMT3A rev: (5′‐ CAATGTAGCGGTCCACCTGAA‐3′), DNMT3B fwd: (5′‐ TTGGAATAGGGGACCTCGTGTG‐3′), DNMT3B rev: (5′‐ AGAGACCTCGGAGAACTTGCCATC‐3′), β‐actin fwd: (5′‐ TCATGAAGTGTGACGTGGACATC‐3′) and β‐actin rev: (5′‐ CAGGAGGAGCAATGATCTTGATCT‐3′). The amplified sequences were 158, 150, 152, and 156 base pair for DNMT1, DNMT3A, DNMT3B, and β‐actin, respectively.
2.6. Quantitative Real‐Time Polymerase Chain Reaction (qRT‐PCR)
qRT‐PCR was performed on Applied Biosystems® StepOneTM instrument using SYBR‐Green SYBR® Premix Ex TaqTM II (Tli RNaseH Plus), Bulk, Takara. qRT‐PCR reaction was performed in a total volume of 25 µl; mixture containing 12.5 µl of SYBR‐Green master mix, 0.8 µl of each primer (10 pmol), 2 µl cDNA, and 8.9 µl of DNase, RNase free water. In non‐template control tube, 2 µl of water was added instead of cDNA. qRT‐PCR experiments were initiated with a 5 min denaturation at 95 °C followed by 40 cycles of denaturation at 95 °C for 30 s and annealing‐ extension at 60 °C for 1 min. In order to determine the specificity of the reaction and target amplification, melt curve analysis was performed at the end of each Real‐time PCR experiment by increasing the temperature from 60 °C to 95 °C with a temperature transition rate of 0.5 °C/s. Comparative (relative) Ct method was used to analysis DNMTs expression, each Ct value was normalized to β‐actin as housekeeping gene.
2.7. Survival analysis
Kaplan–Meier survival analysis was performed to detect the impact of DNMTs promoter methylation status and mRNA expression levels on patients' survival separately. Cox regression analysis was applied to determine whether promoter methylation and expression of DNMTs had predictive value when added to the base model of other clinicopathological features. Disease‐free survival (DFS) was defined as the time between the date of surgery to the date of first confirmed disease recurrence. In disease free survival analysis, local or regional disease recurrence or distant metastasis considered as disease recurrence or event. Overall survival (OS) was evaluated by the duration between the date of surgery to death date. In OS survival analysis, death was defined as event. The patients had no other diseases than breast cancer influencing survival.
2.8. Statistical analysis
Statistical analyses were performed using SPSS software version 20. Clinical parameters as well as DNMTs' promoter methylation status and mRNA expression were collected. Mann‐Whitney U Test was used to assess the differential expression of nonparametric data of mRNA expression and DNA methylation between TAM‐R, TAM‐S and normal tissue specimens. Spearman correlation coefficient was employed to evaluate the association between DNMTs promoter methylation and mRNA expression. The crude and adjusted odds ratios (ORs) and 95% confidence intervals (CIs) were calculated by performance of univariate and multivariate logistic regression analyses in order to define relation between DNMTs’ expression levels and promoter methylation statuses on disease outcome in tamoxifen‐treated breast carcinoma patients. Kaplan–Meier survival analysis was performed to evaluate the effects of DNMTs promoter methylation and mRNA expression on patients′ survival; statistical comparison between the two groups was estimated using log‐rank test. Cox proportional hazards regression model was performed to evaluate the hazard ratios of independent predictor variables, after adjustment of other significant clinicopathological features. HR was represented with 95% confidence intervals (95%CI). Kaplan–Meier survival analysis, logistic regression analysis and Cox proportional hazards regression were performed using DNMTs promoter methylation and mRNA expression as categorical variables. We assessed >50% methylation as hypermethylation status. Tumors with mRNA expression levels below the median were classified as the low expression group. p < 0.05 was considered statistically significant.
3. RESULTS
DNMTs’ promoter methylation and mRNA expression were analyzed in all tumors and normal breast tissues. As mentioned before, evaluation of DNMTs promoter methylation was performed using a series of completely methylated to entirely unmethylated standard DNA samples in MS‐HRM experiments. It was found that 34.7% (25/72), 31.9% (23/72), and 12.5% (9/72) of tumor tissues have low promoter methylation status for DNMT1, DNMT3A, and DNMT3B, respectively. In adjacent normal breast tissues these ratios were 26.6% (8/30), 3.3% (1/30), and 6.6% (2/30) while in normal breast tissue specimens which were resected from healthy persons these ratios were 20% (1/5), 0% and 0%, respectively. The DNMTs transcripts were expressed in all specimens, regardless of whether they were TAM‐R, TAM‐S, or adjacent normal or normal breast tissues. Our findings revealed that there was no significant difference between promoter methylation statuses of DNMT1 (p = 0.086), DNMT3A (p = 0.598), and DNMT3B (p = 0.873) in normal adjacent tumor tissues in comparison to normal breast tissues from healthy individuals; similarly no significant changes were found in mRNA expression of DNMT1 (p = 0.567), DNMT3A (p = 0.395), and DNMT3B (p = 0.321) between these two kinds of normal breast tissues. Figure 1 showed the normalized melting curves of standard controls and breast tissue samples.
Figure 1.
Methylation Specific‐High Resolution Melt curves analysis (MS‐HRM) graphs; some representative graphs for HRM on DNMT1 (a), DNMT3A (c) and DNMT3B (e) studied samples beside controls. Standards melting curves composed of 0, 25, 50, 75, and 100% methylated human control DNA (a, c, and e). Methylation level of each studied sample calculated by comparison of melting curve pattern to controls. Tumors and normal studied samples exhibit different status of methylation in comparison to 0, 50, and 100% methylated controls (b, d, and f). Most of tumors exhibited low methylation status, less than 50%; in contrast normal breast tissues almost exhibited high methylation status, more than 50%, representative graphs for DNMT1 (b), DNMT3A (d), DNMT3B (f)
3.1. DNMTs promoter methylation and mRNA expression in tumor and adjacent normal tissue samples
Comparison of tumor tissues (N = 72) and adjacent normal tissues (N = 30) showed no significant difference in DNMT1 promoter methylation (p = 0.649), however the promoters of DNMT3A (p = 0.002) and DNMT3B (p = 0.021) were significantly demethylated in tumor tissues as compared to the adjacent normal. In addition all three types of DNMTs were significantly overexpressed in tumor tissues compared to adjacent normal tissues DNMT1 (p < 0.001), DNMT3A (p = 0.046) DNMT3B (p = 0.007). Mean fold increase in the mRNA expression of DNMTs in tumor tissues compared to adjacent normal breast tissue specimens were 4.02, 2.16, and 3.74 for DNMT1, DNMT3A, and DNMT3B, respectively (Supplementary Table S1).
3.2. DNMTs promoter methylation and mRNA expression in TAM‐R and TAM‐S breast tumor tissues
In tamoxifen resistant (N = 36) tumors compared to the sensitive (N = 36) group, no significant difference was observed in promoter methylation status of DNMT1 (p = 0.563) and DNMT3B (p = 0.066). However, the amount of DNMT3A promoter methylation was significantly higher in TAM‐S (p = 0.001). mRNA expression of DNMT3A (p = 0.037) and DNMT3B (p < 0.001) were significantly higher in TAM‐R patients, in contrast no significant difference was observed for DNMT1 mRNA expression (p = 0.076). Mean fold increase in the mRNA expression of DNMTs in tamoxifen resistant tumors compared to tamoxifen‐sensitive ones were 2.64, 2.44, and 4.68 for DNMT1, DNMT3A, and DNMT3B, respectively (Supplementary Table S2). Spearman correlation coefficient analysis showed that there was a negative correlation between DNMTs mRNA expression and promoter methylation, although this negative association was only statistically significant about DNMT3A (r:‐0.318 p = 0.007), however, that was not significant for DNMT1 (r:‐0.192 p = 0.107) and DNMT3B (r:‐0.157 p = 0.188).
3.3. Correlation between clinicopathological characteristics and tamoxifen treatment outcome
Considering that the different subtypes of breast cancer show different prognostic value, univariate and multivariate logistic regression analysis were performed to evaluate the effects of promoter methylation and expression of DNMTs on treatment outcome after adjusting for other confounding factors. To investigate the impact of each clinicopathological character on resistance progression, univariate logistic regression analysis was performed and finally predictive factors which were statistically significant in univariate analysis were involved in multivariate logistic regression model. In each categorical variable the first‐ordered category was considered as the reference level.
Univariate logistic regression analysis indicated that N stage N2, N3 ∼ N0, N1 (OR = 2.906, 95%CI: 1.091–7.741; p = 0.033), extracapsular nodal extension (ECE) (OR = 3.52, 95%CI: 1–12.38; p = 0.049), perineural invasion (PNI) (OR = 3.250, 95%CI: 1.217–8.676; p = 0.019), DNMT3A methylation status high meth∼ low meth (OR = 0.073, 95%CI: 0.019–0.281; p = p < 0.001), DNMT1 expression high expression∼ low expression (OR = 4.021, 95%CI: 1.505–10.741; p = 0.006) and DNMT3B expression high expression∼ low expression (OR = 3.538, 95%CI: 1.340–9.343; p = 0.011) were the important affecting features that could associated with resistance incidence in ER+ tamoxifen‐treated breast carcinoma patients.
In the second step, in order to eliminate potential confounding effects of statistically significant predictive factors in univariate analysis, multivariate logistic regression model was performed. The results showed that N stage N2, N3 ∼ N0, N1 (OR = 7.693, 95%CI: 1.608–36.808; p = 0.011), perineural invasion (PNI) (OR = 6.889, 95%CI: 1.503–31.588; p = 0.013), DNMT3A methylation status high meth∼ low meth (OR = 0.021, 95%CI: 0.003–0.156; p < 0.001) and DNMT3B expression (OR = 6.088, 95%CI: 1.344–27.572; p = 0.019) were remained as independent predictors of treatment (Table 2).
Table 2.
Univariate and multivariate logistic regression models for tamoxifen response in estrogen receptor positive breast carcinoma patients (N = 72)
| Univariate analysis | Multivariate analysis | |||||
|---|---|---|---|---|---|---|
| Factor of base model | OR | 95%CI | p‐value | OR | 95%CI | p‐value |
| Histological grade (MBR a ) | ||||||
| Grade I | ||||||
| Grade II | 2.252 | 0.780–6.505 | 0.134 | |||
| Grade III | 2.844 | 0.713–11.351 | 0.139 | |||
| T stage | ||||||
| T1 | ||||||
| T2 | 0.608 | 0.144–2.576 | 0.499 | |||
| T3 | 1.467 | 0.282–7.627 | 0.649 | |||
| T4 | 0.800 | 0.037–17.196 | 0.887 | |||
| N stage | ||||||
| N0, N1 | ||||||
| N2, N3 | 2.906 | 1.091–7.741 | 0.033 | 7.693 | 1.608–36.808 | 0.011 |
| Extracapsular Nodal Extension (ECE) | ||||||
| No | ||||||
| Yes | 3.52 | 1–12.38 | 0.049 | 5.852 | 0.872–39.283 | 0.069 |
| DCIS histology | ||||||
| Non‐Comedo type | ||||||
| Comedo type | 1.290 | 0.317–5.256 | 0.722 | |||
| Nipple Involvement | ||||||
| No | ||||||
| Yes | 1 | 0.290–3.45 | p > 0.99 | |||
| Lymphatic invasion | ||||||
| No | ||||||
| Yes | 2.2 | 0.712–6.793 | 0.17 | |||
| Perineural invasion (PNI) | ||||||
| No | ||||||
| Yes | 3.250 | 1.217–8.676 | 0.019 | 6.889 | 1.503–31.588 | 0.013 |
| PR status | ||||||
| Positive | ||||||
| Negative | 1.130 | 0.428–2.985 | 0.805 | |||
| HER‐2 status | ||||||
| Positive | ||||||
| Negative | 1.540 | 0.534–4.438 | 0.424 | |||
| p53 status | ||||||
| Positive | ||||||
| Negative | 1.909 | 0.696–5.236 | 0.209 | |||
| DNMT1 meyhylation status | ||||||
| Low meth | ||||||
| High meth | 0.691 | 0.261–1.834 | 0.459 | |||
| DNMT3A meyhylation status | ||||||
| Low meth | ||||||
| High meth | 0.073 | 0.019–0.281 | p < 0.001 | 0.021 | 0.003–0.156 | p < 0.001 |
| DNMT3B meyhylation status | ||||||
| Low meth | ||||||
| High meth | 0.244 | 0.047–1.266 | 0.093 | |||
| DNMT1 expression | ||||||
| Low expression | ||||||
| High expression | 4.021 | 1.505–10.741 | 0.006 | 1.859 | 0.478–7.232 | 0.371 |
| DNMT3A expression | ||||||
| Low expression | ||||||
| High expression | 2.200 | 0.857–5.645 | 0.101 | |||
| DNMT3B expression | ||||||
| Low expression | ||||||
| High expression | 3.538 | 1.340–9.343 | 0.011 | 6.088 | 1.344–27.572 | 0.019 |
MBR: Modified‐Bloom‐Richardson.
3.4. Predictive value of DNMTs promoter methylation on patients’ survival
Kaplan–Meier survival curves based upon DNMTs’ promoter methylation statuses using MS‐HRM results of 72 tamoxifen‐treated breast carcinoma patients revealed that DNMT1 promoter methylation status was not associated with disease free survival (p = 0.482, Supplementary Figure S3a) and overall survival (p = 0.227, Supplementary Figure S3b). In contrast, patients with high methylation status of DNMT3A had a significantly increased DFS (p <0.001, Figure 2a) and OS (p = 0.002, Figure 2b) chance. DNMT3B methylation status was not correlated with disease free survival (p = 0.097, Supplementary Figure S3c) and OS (p = 0.617, Supplementary Figure S3d) in tamoxifen‐treated breast carcinoma patients.
Figure 2.
Kaplan–Meier (KM) survival curves of breast carcinoma patients stratified by methylation status of DNA methyl transferase 3A (DNMT3A). Survival rates for 72 breast carcinoma were calculated using the Kaplan–Meier method and compared with the log‐rank test. Patients with high methylation status of DNMT3A had a significantly increased DFS (p < 0.001, a) and OS (p = 0.002, b) chance
3.5. Predictive value of DNMTs expression and patients’ survival
Kaplan–Meier survival analysis demonstrated that tamoxifen‐treated breast carcinoma patients with low expression levels of DNMT1 have better disease free survival (p = 0.001, Supplementary Figure S4a) and overall survival (p = 0.004, Supplementary Figure S4b) rate. Significant relationships was observed between low expression level of DNMT3A and better DFS (p = 0.043, Figure 3a), in contrast no significant association was revealed between DNMT3A expression and OS (p = 0.485, Figure 3b). Overexpression of DNMT3B was associated with poor DFS (p = 0.04, Supplementary Figure S4c) and OS (p = 0.02, Supplementary Figure S4d).
Figure 3.
Kaplan–Meier (KM) survival curves of breast carcinoma patients stratified by the median values of DNA methyltransferase3A (DNMT3A) expression. DNMT3A low expression correlated with better DFS (p = 0.043, a), in contrast no significant association was demonstrated between DNMT3A expression and OS (p = 0.485, b)
3.6. Univariate and multivariate Cox regression analysis
To determine whether DNMTs promoter methylation and mRNA expression could provide superior prognostic value for disease recurrence in tamoxifen‐treated breast cancer patients compared with the other known clinical or prognostic factors, univariate Cox proportional hazard analysis were performed. In each categorical variable the first‐ordered category was considered as reference level. The results indicated that DNMT3A methylation status Low meth ∼high meth (HR = 4.546, 95%CI: 2.323–8.897; p < 0.001), DNMT1 expression high expression∼ low expression (HR = 3.149, 95%CI: 1.539–6.444; p = 0.002) and DNMT3B expression high expression∼ low expression (HR = 1.994, 95%CI: 1.009‐ 3.942; p = 0.047) revealed predictive power for DFS. Besides, DNMT3A methylation status and both DNMT1 and DNMT3B expression were included in predictive factor for OS in tamoxifen‐treated breast carcinoma patients as follow: DNMT3A methylation status low meth ∼high meth (HR = 3.798, 95%CI: 1.522–9.476; p = 0.004), DNMT1 expression high expression∼ low expression (HR = 4.408, 95%CI: 1.459‐ 13.324; p = 0.009), DNMT3B expression high expression∼ low expression (HR = 3.166, 95%CI: 1.136‐ 8.821; p = 0.028). Other clinicopathological features which were significant in univariate Cox regression analysis are presented in Table 3.
Table 3.
Univariate Cox regression models for disease free survival and overall survival in estrogen receptor positive tamoxifen–treated breast carcinoma patients (N = 72)
| Univariate Cox regression model for DFS | Univariate Cox regression model for OS | |||||
|---|---|---|---|---|---|---|
| Factor of base model | HR | 95%CI | p‐value | HR | 95%CI | p‐value |
| Histological grade (MBR a ) | ||||||
| Grade I | ||||||
| Grade II | 2.034 | 0.915–4.520 | 0.081 | 3.969 | 1.104–14.261 | 0.035 |
| Grade III | 2.016 | 0.777–5.229 | 0.149 | 3.207 | 0.715–14.328 | 0.128 |
| T stage | ||||||
| T1 | ||||||
| T2 | 0.691 | 0.257–1.857 | 0.464 | 0.526 | 0.142–1.954 | 0.338 |
| T3 | 1.190 | 0.413–3.432 | 0.747 | 1.001 | 0.248–4.032 | 0.999 |
| T4 | 1.438 | 0.167–12.391 | 0.741 | 2.481 | 0.256–24.071 | 0.433 |
| N stage | ||||||
| N0, N1 | ||||||
| N2, N3 | 2.325 | 1.2–4.506 | 0.012 | 1.852 | 0.750–4.574 | 0.182 |
| Extracapsular Nodal Extension (ECE) | ||||||
| Absent | ||||||
| Present | 2.174 | 1.057–4.469 | 0.035 | 2.002 | 0.76–5.274 | 0.160 |
| DCIS histology | ||||||
| Non‐Comedo type | ||||||
| Comedo type | 1.017 | 0.395–2.620 | 0.972 | 1.840 | 0.609–5.557 | 0.280 |
| Nipple Involvement | ||||||
| Absent | ||||||
| Present | 1.417 | 0.585–3.434 | 0.441 | 1.099 | 0.317–3.810 | 0.881 |
| Lymphatic invasion | ||||||
| Absent | ||||||
| Present | 1.785 | 0.742–4.296 | 0.196 | 1.825 | 0.531–6.268 | 0.339 |
| Perineural invasion (PNI) | ||||||
| Absent | ||||||
| Present | 2.322 | 1.199–4.499 | 0.013 | 0.787 | 0.310–1.999 | 0.614 |
| PR status | ||||||
| Positive | ||||||
| Negative | 1.156 | 0.585–2.283 | 0.677 | 2.466 | 0.997–6.102 | 0.051 |
| HER‐2 status | ||||||
| Positive | ||||||
| Negative | 1.144 | 0.519–2.521 | 0.739 | 1.223 | 0.403–3.706 | 0.723 |
| p53 status | ||||||
| Positive | ||||||
| Negative | 0.570 | 0.268–1.214 | 0.145 | 1.004 | 0.381–2.645 | 0.994 |
| DNMT1 meyhylation status | ||||||
| High meth | ||||||
| Low meth | 1.268 | 0.648–2.481 | 0.488 | 1.751 | 0.696–4.406 | 0.234 |
| DNMT3A meyhylation status | ||||||
| High meth | ||||||
| Low meth | 4.546 | 2.323–8.897 | p < 0.001 | 3.798 | 1.522–9.476 | 0.004 |
| DNMT3B meyhylation status | ||||||
| High meth | ||||||
| Low meth | 1.974 | 0.861–4.525 | 0.108 | 1.368 | 0.398–4.708 | 0.619 |
| DNMT1 expression | ||||||
| Low expression | ||||||
| High expression | 3.149 | 1.539–6.444 | 0.002 | 4.408 | 1.459–13.324 | 0.009 |
| DNMT3A expression | ||||||
| Low expression | ||||||
| High expression | 1.963 | 1.001–3.850 | 0.051 | 1.381 | 0.555–3.441 | 0.488 |
| DNMT3B expression | ||||||
| Low expression | ||||||
| High expression | 1.994 | 1.009–3.942 | 0.047 | 3.166 | 1.136–8.821 | 0.028 |
MBR: Modified‐Bloom‐Richardson.
In the second step, multivariate Cox regression was performed. Factors which were identified as significant in univariate analysis were then included in multivariate models to determine whether they were independent predictor of survival after adjustment of other significant clinicopathological features in patients or not.
Data analysis revealed that N stage N2, N3 ∼ N0, N1 (OR = 2.435, 95%CI: 1.132–5.235; p = 0.023), and DNMT3A methylation status low meth ∼ high meth (HR = 3.732, 95%CI: 1.614–8.628; p = 0.002) remained as independent predictive factors for DFS after adjustment of other criteria.
Multivariable Cox regression analysis for OS demonstrated that DNMT3A methylation status low meth ∼ high meth (HR = 3.201, 95%CI: 1.153–8.886; p = 0.026), and DNMT1 expression high expression ∼ low expression (HR = 3.454, 95%CI: 1.126–10.590; p = 0.030) remained significant after adjustment of other criteria (Table 4).
Table 4.
Multivariable Cox regression models for disease free survival and overall survival in estrogen receptor positive tamoxifen–treated breast carcinoma patients (N = 72)
| Multivariate analysis for DFS | Multivariate analysis for OS | |||||
|---|---|---|---|---|---|---|
| Factor of base model | HR | 95%CI | p‐value | HR | 95%CI | p‐value |
| Histological grade (MBR a ) | ||||||
| Grade I | ||||||
| Grade II | 3.995 | 1.114–14.326 | 0.034 | |||
| Grade III | 3.383 | 0.725–15.792 | 0.121 | |||
| N stage | ||||||
| N0, N1 | ||||||
| N2, N3 | 2.435 | 1.132–5.235 | 0.023 | |||
| Extracapsular Nodal Extension (ECE) | ||||||
| Absent | ||||||
| Present | 1.028 | 0.408–2.591 | 0.953 | |||
| Perineural invasion (PNI) | ||||||
| Absent | ||||||
| Present | 2.063 | 1.021–4.167 | 0.044 | |||
| DNMT3A meyhylation status | ||||||
| High meth | ||||||
| Low meth | 3.732 | 1.614–8.628 | 0.002 | 3.201 | 1.153–8.886 | 0.026 |
| DNMT1 expression | ||||||
| Low expression | ||||||
| High expression | 1.853 | 0.844–4.072 | 0.124 | 3.454 | 1.126–10.590 | 0.030 |
| DNMT3B expression | ||||||
| Low expression | ||||||
| High expression | 1.571 | 0.729–3.383 | 0.249 | 2.453 | 0.853–7.056 | 0.096 |
MBR: Modified‐Bloom‐Richardson.
4. DISCUSSION
Epigenetic modifications including DNA methylation is one of the regulatory mechanisms which influence on gene expression. Accumulating data exhibited that abnormal DNA methylation associated with tumor formation, metastasis incidence and progression of drug resistance (Chen et al., 2015; Luczak & Jagodzinski, 2006). In mammals, methylation of genomic DNA facilitate by DNA methyltransferases (DNMTs). Interaction and cooperation of the most important DNMTs (DNMT1, DNMT3A, and DNMT3B) established and maintained the normal pattern of methylation in genome which disrupted during tumorigenesis (Jones & Liang, 2009). Expression of DNMTs has a key role in distribution of DNA methylation in various malignancies. Furthermore, presence of CpG islands in the promoter sequence of DNMTs gives them the potential of epigenetic regulation (http://www.ensembl.org). Tamoxifen almost included in the treatment protocol of all stages of estrogen receptor positive breast cancer patients, but unfortunately about 50% of tamoxifen‐treated patients revealed recurrence symptoms (Schiff et al., 2003). Although tamoxifen resistance progression is an intricate process, recent studies revealed that epigenetic modifications, especially DNA methylation, play crucial roles in disease recurrence which made it an attractive research field (Bae et al., 2004; Miyamoto et al., 2005).
Recent epigenetic studies have indicated that the major contributor genes in development of breast malignancies and progression of tamoxifen resistance contain CpG islands in their promoter regions which sensitized them to methylation‐mediated regulation. For example, aberrant methylation of hormone receptor promoters (estrogen and progesterone receptors) is indicated to be involved in tamoxifen refractory (Bardou, Arpino, Elledge, Osborne, & Clark, 2003; Herynk & Fuqua, 2007). Methylation also mediates down regulation of Phosphatase and tensin homolog (PTEN) (Phuong et al., 2011) in breast cancer patients who do not benefit from tamoxifen therapy. Moreover, Iorns et al. (2008) indicated that hypermethylation of Cyclin‐dependent kinase 10 (CDK10) promoter contributes to tamoxifen resistance progression in breast cancer. In the study by Widschwendter et al. (2004) silencing of CYP1B1, encoding a tamoxifen‐ and estradiol‐metabolizing cytochrome p450, by promoter methylation was observed in human breast tumors. Also, downregulation of miR‐320a and miR‐27b assist to disease recurrence and interestingly promoter methylation‐mediated downregulation of these miRs have also been reported in tamoxifen resistant breast cancer cells (He, Gu, Jiang, Jin, & Ma, 2014; Li, Wu, Liu, & Tang, 2016; Lu et al., 2015). In the present study, Methylation‐specific high‐resolution melt curve analysis (MS‐HRM) and quantitative Real Time PCR (qRT‐PCR) assays were performed to evaluate promoter methylation and mRNA expression of three functional DNMTs (DNMT1, DNMT3A, DNMT3B) in ER+ tamoxifen treated breast cancer patients and normal breast tissues. Comparison of 72 breast tumor tissues and 30 adjacent normal tissues demonstrated that expression of all three DNMTs were significantly increased in tumor tissues (N = 72) in comparison to adjacent normal tissues (N = 30). Our findings were in line with other studies that previously indicated that DNMTs' are overexpressed in tumor tissues in various malignancies in comparison to normal tissues such as lung cancer (Belinsky, Nikula, Baylin, & Issa, 1996), ovarian cancer (Bai et al., 2012), acute and chronic myelogenous leukemia (Mizuno et al., 2001), gastric cancer (Ding, Fang, Chen, & Peng, 2008), and melanoma (Deng et al., 2009). Furthermore, Nagai, Nakamura, Makino, and Mitamura (2003) indicated that the increased expression of DNMT1 and DNMT3B are involved in hepatocellular carcinogenesis. Similarly, overexpression patterns of DNMT1 and DNMT3B in ovarian cancer have been also reported by Gu et al. (2013). We also exhibited that in tumor specimens, DNMTs were significantly overexpressed in tamoxifen resistant tumors (N = 36) in comparison to tamoxifen sensitive cases (N = 36). This finding was in line with previous studies, showing that pharmacologic or genetic downregulation of DNMTs could reduce tamoxifen resistance in breast cancer (Altundag et al., 2004; Bovenzi & Momparler, 2001). It is also in line with previous studies, revealing an association between overexpression of DNMT1 and DNMT3B with drug resistance in ovarian cancer cell line and murine neuroblastoma cells (Li et al., 2009; Qiu, Mirkin, & Dwivedi, 2002). The epigenetic mechanisms that regulate expression of DNMTs are not well understood. In the current study, methylation status of the most prominent CpGs in DNMTs promoters has been analyzed to investigate if methylation process contributes to regulation of these crucial genes or not. Our MS‐HRM experiments exhibited that DNMT3A and DNMT3B promoters were significantly demethylated in breast tumor tissues (N = 72) in comparison to normal breast samples (N = 30). The same result was reported by Naghitorabi et al. (2013) who showed that DNMT3B promoter was hypo‐methylated in tumors in comparison to adjacent normal breast tissues. However, there were no reports about low methylation status of DNMT3A promoter in breast tumor tissues previously.
Among tumor tissues DNMT3A promoter methylation status was significantly higher in TAM‐S tumors (N = 36) in comparison to TAM‐R group (N = 36). In the present study, a negative association was found between methylation levels of each DNMT and its expression, although this negative correlation was only statistically significant for DNMT3A. It has been revealed previously that gene silencing could be mediated by methylation of CpG‐rich areas (CpG islands) located in the regulatory regions include promoter sequence (Curradi, Izzo, Badaracco, & Landsberger, 2002). The role of DNA methylation pattern in regulation of DNMTs expression has been also shown previously in gliomas (Rajendran et al., 2011) and colorectal cancer (Huidobro et al., 2012). In our experiments, multivariate logistic regression analysis confirmed the role of DNMT3A low methylation status and DNMT3B overexpression in tamoxifen resistance progression after adjustment of other clinicopathological features. Among all type of DNMTs, DNMT3B has been shown previously as the most important factor in promoter methylation of tumor suppressor genes (Linhart et al., 2007). In this study, Kaplan–Meier survival analysis showed that overexpression of DNMT1 negatively correlated with disease free survival and overall survival in tamoxifen‐treated breast carcinoma patients (N = 72); it has been approved by other studies that showing the higher expression of DNMT1 is associated with increased risk of death in gastric cancer (Cao et al., 2014) and renal cell carcinoma (Li et al., 2014). In the same manner, our results exhibited that high expression of DNMT3A negatively correlated with DFS which was in accordance with the previous study that exhibited the higher expression of DNMT3A in glioblastoma significantly decreases survival (Cheray et al., 2016). In contrast to our results, low expression of DNMT3A was correlated with poor prognosis in lung adenocarcinoma patients (Husni et al., 2016). Our data revealed that there was an inverse association between DNMT3B expression and survival that was in line with other studies, showing DNMT3B overexpression correlated with poor prognosis in diffuse large B‐cell lymphoma patients (Amara et al., 2010) and primary acute myeloid leukemia (Niederwieser et al., 2015). Multivariate cox regression analysis demonstrated that low methylation status of DNMT3A promoter was remained as an unfavorable predictor of disease free survival and overall survival after adjustment of other significant clinicopathological features. Over expression of DNMT1 also significantly associated with poor overall survival after adjusting of other confounder factors
In conclusion, our findings indicated that the promoters of DNMT3A and DNMT3B were significantly demethylated in tumor tissues and all three DNMTs were significantly overexpressed in breast tumors. Furthermore, increased expression of DNMT3A and DNMT3B were observed in TAM‐R tumor tissues in comparison to TAM‐S cases. Indeed, low methylation status of DNMT3A and overexpression of DNMT3B may be used as independent resistant prognostic marker in tamoxifen‐treated breast cancer patients. More studies are needed to investigate the relationship between the expression of DNMTs and aberrant DNA methylation pattern of tamoxifen resistance mediated genes. Our knowledge of epigenetic‐mediated mechanisms of tamoxifen resistance should be extended to introduce DNMTs as a new target to prevent disease recurrence in tamoxifen‐treated breast cancer patients.
CONFLICTS OF INTEREST
The authors declare that they have no conflicts of interest.
ETHICAL APPROVAL
All experiments performed in this study were in accordance with the ethical standards of the local ethical committee at Mashhad University of Medical Sciences, Mashhad, Iran.
INFORMED CONSENT
Each recruited patient has signed written informed consent for retention and analysis of her tissue and long‐term follow‐up.
Supporting information
Additional Supporting Information may be found online in the supporting information tab for this article.
Table S1. Fold change of DNMTs mRNA expression levels of tumor tissues (N = 72) compared to adjacent normal breast tissues (N = 30).
Table S2. Fold change of DNMTs mRNA expression levels of tamoxifen–resistant tumor tissues (N = 36) compared to tamoxifen‐sensitive ones (N = 36).
Figure S3. Kaplan–Meier (KM) survival curves of breast carcinoma patients stratified by methylation status of DNA methyl transferases (DNMTs). Survival rates for 72 breast carcinoma were calculated using the Kaplan–Meier method and compared with the log‐rank test.
Figure S4. Kaplan–Meier (KM) survival curves of breast carcinoma patients stratified by the median values of mRNA expression of DNA methyltransferases (DNMTs). Patients with low expression of DNMT1 demonstrated better DFS (p = 0.001, A) and OS (p = 0.004, B) rate; low expression of DNMT3B associated with better DFS (p = 0.04, C) and OS (p = 0.02, D).
ACKNOWLEDGMENTS
This research work was conducted as a part of the PhD thesis and was financially supported by research grant (Grant No. 921853) from the Vice Chancellor of Research, Mashhad University of Medical Sciences, Mashhad, Iran. The authors are grateful to staff of Cancer Institute of Imam Khomeini hospital of Tehran for their kindly collaborations.
REFERENCES
- Altundag, O. , Altundag, K. , & Gunduz, M. (2004). DNA methylation inhibitor, procainamide, may decrease the tamoxifen resistance by inducing overexpression of the estrogen receptor beta in breast cancer patients. Medical Hypotheses, 63(4), 684–687. [DOI] [PubMed] [Google Scholar]
- Amara, K. , Ziadi, S. , Hachana, M. , Soltani, N. , Korbi, S. , & Trimeche, M. (2010). DNA methyltransferase DNMT3b protein overexpression as a prognostic factor in patients with diffuse large B‐cell lymphomas. Cancer Science, 101(7), 1722–1730. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bae, Y. K. , Brown, A. , Garrett, E. , Bornman, D. , Fackler, M. J. , Sukumar, S. , … Gabrielson, E. (2004). Hypermethylation in histologically distinct classes of breast cancer. Clinical Cancer Research: an Official Journal of the American Association for Cancer Research, 10(18 Pt 1), 5998–6005. [DOI] [PubMed] [Google Scholar]
- Bai, X. , Song, Z. , Fu, Y. , Yu, Z. , Zhao, L. , Zhao, H. , … Wei, M. (2012). Clinicopathological significance and prognostic value of DNA methyltransferase 1, 3a, and 3b expressions in sporadic epithelial ovarian cancer. PLoS ONE, 7(6), e40024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bardou, V. J. , Arpino, G. , Elledge, R. M. , Osborne, C. K. , & Clark, G. M. (2003). Progesterone receptor status significantly improves outcome prediction over estrogen receptor status alone for adjuvant endocrine therapy in two large breast cancer databases. Journal of Clinical Oncology: Official Journal of the American Society of Clinical Oncology, 21(10), 1973–1979. [DOI] [PubMed] [Google Scholar]
- Belinsky, S. A. , Nikula, K. J. , Baylin, S. B. , & Issa, J. P. (1996). Increased cytosine DNA‐methyltransferase activity is target‐cell‐specific and an early event in lung cancer. Proceedings of the National Academy of Sciences of the United States of America, 93(9), 4045–4050. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bovenzi, V. , & Momparler, R. L. (2001). Antineoplastic action of 5‐aza‐2'‐deoxycytidine and histone deacetylase inhibitor and their effect on the expression of retinoic acid receptor beta and estrogen receptor alpha genes in breast carcinoma cells. Cancer Chemotherapy and Pharmacology, 48(1), 71–76. [DOI] [PubMed] [Google Scholar]
- Cao, X. Y. , Ma, H. X. , Shang, Y. H. , Jin, M. S. , Kong, F. , Jia, Z. F. , … Jiang, J. (2014). DNA methyltransferase3a expression is an independent poor prognostic indicator in gastric cancer. World Journal of Gastroenterology: WJG, 20(25), 8201–8208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chen, C. C. , Lee, K. D. , Pai, M. Y. , Chu, P. Y. , Hsu, C. C. , Chiu, C. C. , … Leu, Y. W. (2015). Changes in DNA methylation are associated with the development of drug resistance in cervical cancer cells. Cancer Cell International, 15, 98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cheray, M. , Pacaud, R. , Nadaradjane, A. , Oliver, L. , Vallette, F. M. , & Cartron, P. F. (2016). Specific inhibition of DNMT3A/ISGF3gamma interaction increases the temozolomide efficiency to reduce tumor growth. Theranostics, 6(11), 1988–1999. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Curradi, M. , Izzo, A. , Badaracco, G. , & Landsberger, N. (2002). Molecular mechanisms of gene silencing mediated by DNA methylation. Molecular and Cellular Biology, 22(9), 3157–3173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dannenberg, L. O. , & Edenberg, H. J. (2006). Epigenetics of gene expression in human hepatoma cells: Expression profiling the response to inhibition of DNA methylation and histone deacetylation. BMC Genomics, 7, 181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Das, P. M. , & Singal, R. (2004). DNA methylation and cancer. Journal of Clinical Oncology: Official Journal of the American Society of Clinical Oncology, 22(22), 4632–4642. [DOI] [PubMed] [Google Scholar]
- De la Grange, P. , Dutertre, M. , Correa, M. , & Auboeuf, D. (2007). A new advance in alternative splicing databases: From catalogue to detailed analysis of regulation of expression and function of human alternative splicing variants. BMC Bioinformatics, 8, 180. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Deng, T. , Kuang, Y. , Wang, L. , Li, J. , Wang, Z. , & Fei, J. (2009). An essential role for DNA methyltransferase 3a in melanoma tumorigenesis. Biochemical and Biophysical Research Communications, 387(3), 611–616. [DOI] [PubMed] [Google Scholar]
- Ding, W. J. , Fang, J. Y. , Chen, X. Y. , & Peng, Y. S. (2008). The expression and clinical significance of DNA methyltransferase proteins in human gastric cancer. Digestive Diseases and Sciences, 53(8), 2083–2089. [DOI] [PubMed] [Google Scholar]
- Gu, Y. , Yang, P. , Shao, Q. , Liu, X. , Xia, S. , Zhang, M. , … Shao, Q. (2013). Investigation of the expression patterns and correlation of DNA methyltransferases and class I histone deacetylases in ovarian cancer tissues. Oncology Letters, 5(2), 452–458. [DOI] [PMC free article] [PubMed] [Google Scholar]
- He, D. X. , Gu, X. T. , Jiang, L. , Jin, J. , & Ma, X. (2014). A methylation‐based regulatory network for microRNA 320a in chemoresistant breast cancer. Molecular Pharmacology, 86(5), 536–547. [DOI] [PubMed] [Google Scholar]
- Herynk, M. H. , & Fuqua, S. A. (2007). Estrogen receptors in resistance to hormone therapy. Advances in Experimental Medicine and Biology, 608, 130–143. [DOI] [PubMed] [Google Scholar]
- Huidobro, C. , Urdinguio, R. G. , Rodriguez, R. M. , Mangas, C. , Calvanese, V. , Martinez‐Camblor, P. , … Fraga, M. F. (2012). A DNA methylation signature associated with aberrant promoter DNA hypermethylation of DNMT3B in human colorectal cancer. European Journal of Cancer (Oxford. England: 1990), 48(14), 2270–2281. [DOI] [PubMed] [Google Scholar]
- Husni, R. E. , Shiba‐Ishii, A. , Iiyama, S. , Shiozawa, T. , Kim, Y. , Nakagawa, T. , … Noguchi, M. (2016). DNMT3a expression pattern and its prognostic value in lung adenocarcinoma. Lung Cancer (Amsterdam, Netherlands), 97, 59–65. [DOI] [PubMed] [Google Scholar]
- Iorns, E. , Turner, N. C. , Elliott, R. , Syed, N. , Garrone, O. , Gasco, M. , … Ashworth, A. (2008). Identification of CDK10 as an important determinant of resistance to endocrine therapy for breast cancer. Cancer Cell, 13(2), 91–104. [DOI] [PubMed] [Google Scholar]
- Jiang, Q. , Zheng, S. , & Wang, G. (2013). Development of new estrogen receptor‐targeting therapeutic agents for tamoxifen‐resistant breast cancer. Future Medicinal Chemistry, 5(9), 1023–1035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jones, P. A. , & Liang, G. (2009). Rethinking how DNA methylation patterns are maintained. Nature Reviews Genetics, 10(11), 805–811. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Li, M. , Balch, C. , Montgomery, J. S. , Jeong, M. , Chung, J. H. , Yan, P. , … Nephew, K. P. (2009). Integrated analysis of DNA methylation and gene expression reveals specific signaling pathways associated with platinum resistance in ovarian cancer. BMC Medical Genomics, 2, 34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Li, M. , Wang, Y. , Song, Y. , Bu, R. , Yin, B. , Fei, X. , … Wu, B. (2014). Aberrant DNA methyltransferase 1 expression in clear cell renal cell carcinoma development and progression. Chinese Journal of Cancer Research = Chung‐kuo Yen Cheng Yen Chiu, 26(4), 371–381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Li, X. , Wu, Y. , Liu, A. , & Tang, X. (2016). MiR‐27b is epigenetically downregulated in tamoxifen resistant breast cancer cells due to promoter methylation and regulates tamoxifen sensitivity by targeting HMGB3. Biochemical and Biophysical Research Communications, 477(4), 768–773. [DOI] [PubMed] [Google Scholar]
- Lin, J. H. , Zhang, S. M. , & Manson, J. E. (2011). Predicting adherence to tamoxifen for breast cancer adjuvant therapy and prevention. Cancer Prevention Research (Philadelphia, Pa), 4(9), 1360–1365. [DOI] [PubMed] [Google Scholar]
- Linhart, H. G. , Lin, H. , Yamada, Y. , Moran, E. , Steine, E. J. , Gokhale, S. , … Jaenisch, R. (2007). Dnmt3b promotes tumorigenesis in vivo by gene‐specific de novo methylation and transcriptional silencing. Genes and Development, 21(23), 3110–3122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lu, M. , Ding, K. , Zhang, G. , Yin, M. , Yao, G. , Tian, H. , … Sun, F. (2015). MicroRNA‐320a sensitizes tamoxifen‐resistant breast cancer cells to tamoxifen by targeting ARPP‐19 and ERRgamma. Scientific Reports, 5, 8735. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lu, R. , Hu, X. , Zhou, J. , Sun, J. , Zhu, A. Z. , Xu, X. , … Guo, L. (2016). COPS5 amplification and overexpression confers tamoxifen‐resistance in ERalpha‐positive breast cancer by degradation of NCoR. Nature Communications, 7, 12044. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Luczak, M. W. , & Jagodzinski, P. P. (2006). The role of DNA methylation in cancer development. Folia Histochemica Et Cytobiologica, 44(3), 143–154. [PubMed] [Google Scholar]
- Miyamoto, K. , Fukutomi, T. , Akashi‐Tanaka, S. , Hasegawa, T. , Asahara, T. , Sugimura, T. , & Ushijima, T. (2005). Identification of 20 genes aberrantly methylated in human breast cancers. International Journal of Cancer, 116(3), 407–414. [DOI] [PubMed] [Google Scholar]
- Mizuno, S. , Chijiwa, T. , Okamura, T. , Akashi, K. , Fukumaki, Y. , Niho, Y. , & Sasaki, H. (2001). Expression of DNA methyltransferases DNMT1, 3A, and 3B in normal hematopoiesis and in acute and chronic myelogenous leukemia. Blood, 97(5), 1172–1179. [DOI] [PubMed] [Google Scholar]
- Nagai, M. , Nakamura, A. , Makino, R. , & Mitamura, K. (2003). Expression of DNA (5‐cytosin)‐methyltransferases (DNMTs) in hepatocellular carcinomas. Hepatology Research: the Official Journal of the Japan Society of Hepatology, 26(3), 186–191. [DOI] [PubMed] [Google Scholar]
- Naghitorabi, M. , Mohammadi Asl, J. , Mir Mohammad Sadeghi, H. , Rabbani, M. , Jafarian‐Dehkordi, A. , & Javanmard, H. S. (2013). Quantitative evaluation of DNMT3B promoter methylation in breast cancer patients using differential high resolution melting analysis. Research in Pharmaceutical Sciences, 8(3), 167–175. [PMC free article] [PubMed] [Google Scholar]
- Niederwieser, C. , Kohlschmidt, J. , Volinia, S. , Whitman, S. P. , Metzeler, K. H. , Eisfeld, A. K. , … Bloomfield, C. D. (2015). Prognostic and biologic significance of DNMT3B expression in older patients with cytogenetically normal primary acute myeloid leukemia. Leukemia, 29(3), 567–575. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Phuong, N. T. , Kim, S. K. , Lim, S. C. , Kim, H. S. , Kim, T. H. , Lee, K. Y. , … Kang, K. W. (2011). Role of PTEN promoter methylation in tamoxifen‐resistant breast cancer cells. Breast Cancer Research and Treatment, 130(1), 73–83. [DOI] [PubMed] [Google Scholar]
- Qiu, Y. Y. , Mirkin, B. L. , & Dwivedi, R. S. (2002). Differential expression of DNA‐methyltransferases in drug resistant murine neuroblastoma cells. Cancer Detection and Prevention, 26(6), 444–453. [DOI] [PubMed] [Google Scholar]
- Rajendran, G. , Shanmuganandam, K. , Bendre, A. , Muzumdar, D. , Goel, A. , & Shiras, A. (2011). Epigenetic regulation of DNA methyltransferases: DNMT1 and DNMT3B in gliomas. Journal of Neuro‐Oncology, 104(2), 483–494. [DOI] [PubMed] [Google Scholar]
- Rofaiel, S. , Muo, E. N. , & Mousa, S. A. (2010). Pharmacogenetics in breast cancer: Steps toward personalized medicine in breast cancer management. Pharmacogenomics and Personalized Medicine, 3, 129–143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schiff, R. , Massarweh, S. , Shou, J. , & Osborne, C. K. (2003). Breast cancer endocrine resistance: How growth factor signaling and estrogen receptor coregulators modulate response. Clinical Cancer Research: An Official Journal of the American Association for Cancer Research, 9(1 Pt 2), 447S–454S. [PubMed] [Google Scholar]
- Tanic, N. , Milovanovic, Z. , Dzodic, R. , Juranic, Z. , Susnjar, S. , Plesinac‐Karapandzic, V. , … Dimitrijevic, B. (2012). The impact of PTEN tumor suppressor gene on acquiring resistance to tamoxifen treatment in breast cancer patients. Cancer Biology and Therapy, 13(12), 1165–1174. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Widschwendter, M. , Siegmund, K. D. , Muller, H. M. , Fiegl, H. , Marth, C. , Muller‐Holzner, E. , … Laird, P. W. (2004). Association of breast cancer DNA methylation profiles with hormone receptor status and response to tamoxifen. Cancer Research, 64(11), 3807–3813. [DOI] [PubMed] [Google Scholar]
- Williams, K. E. , Anderton, D. L. , Lee, M. P. , Pentecost, B. T. , & Arcaro, K. F. (2014). High‐density array analysis of DNA methylation in Tamoxifen‐resistant breast cancer cell lines. Epigenetics, 9(2), 297–307. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wojdacz, T. K. , Borgbo, T. , & Hansen, L. L. (2009). Primer design versus PCR bias in methylation independent PCR amplifications. Epigenetics, 4(4), 231–234. [DOI] [PubMed] [Google Scholar]
- Wu, S. T. , Sun, G. H. , Cha, T. L. , Kao, C. C. , Chang, S. Y. , Kuo, S. C. , & Way, T. D. (2016). CSC‐3436 switched tamoxifen‐induced autophagy to apoptosis through the inhibition of AMPK/mTOR pathway. Journal of Biomedical Science, 23(1), 60. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Additional Supporting Information may be found online in the supporting information tab for this article.
Table S1. Fold change of DNMTs mRNA expression levels of tumor tissues (N = 72) compared to adjacent normal breast tissues (N = 30).
Table S2. Fold change of DNMTs mRNA expression levels of tamoxifen–resistant tumor tissues (N = 36) compared to tamoxifen‐sensitive ones (N = 36).
Figure S3. Kaplan–Meier (KM) survival curves of breast carcinoma patients stratified by methylation status of DNA methyl transferases (DNMTs). Survival rates for 72 breast carcinoma were calculated using the Kaplan–Meier method and compared with the log‐rank test.
Figure S4. Kaplan–Meier (KM) survival curves of breast carcinoma patients stratified by the median values of mRNA expression of DNA methyltransferases (DNMTs). Patients with low expression of DNMT1 demonstrated better DFS (p = 0.001, A) and OS (p = 0.004, B) rate; low expression of DNMT3B associated with better DFS (p = 0.04, C) and OS (p = 0.02, D).