A total of 100 pairs of tumor and ANCT tissue samples were collected and stored in RNA Later™ stabilization solution (cat. no. AM7024; Thermo Fisher Scientific, Inc., Waltham, MA, USA) at the time of surgery from patients with breast cancer at Lady Reading Hospital (Khyber Pakhtunkhwa, Pakistan) between June 2012 and August 2015. All the patients were female. The ANCT sections were taken from an area ~2 cm from the cancer lesion site, as selected by a histo-pathologist. The patients with a history of other infectious, metabolic or familial diseases were excluded from the present study. Furthermore, prior approval was obtained from the Ethical Review Committee (approval no. CIIT-09-10-14) of the COMSTAS Institute of Information Technology (Islamabad, Pakistan) and the collaborating hospital aforementioned for the initiation of this project. In addition, informed and written consent was received from all patients enrolled in the present study, prior to sample collection.

RNA extraction from all tissue samples (tumor and ANCT) was performed using the standard TRIzol^® method (27). The extracted mRNA was converted to cDNA using a cDNA synthesis kit (Thermo Fisher Scientific, Inc.). The primers used for mRNA expression analysis were as follows: E2F4 forward, 5′-TCAGAAATCTTTGATCCCAC-3′ and reverse, 5′-AGATATAATCGTGGTCTCCC-3′; β-actin (internal control) forward, 5′-CACTCTTCCAGCCTTCCTTC-3′ and reverse, 5′-TGATCTCCTTCTGCATCGTG-3′. PCR amplification was performed using SYBR^®-Green method (28) through Step One Plus^™ Real-Time PCR system (Applied Biosystems; Thermo Fisher Scientific, Inc.). The thermocycler conditions were as follows: Initial denaturation at 95°C for 10 min, followed by 40 cycles of 95°C for 15 sec, 56°C for 60 sec and 72°C for 20 sec; and a final extension at 72°C for 1 min. mRNA expression of the E2F4 gene was calculated using the 2^−ΔΔCq method (29).

DNA from tumor and ANCT tissues was extracted using the standard phenol-chloroform method (30), and confirmed by 2% agarose gel electrophoresis. Genomic DNA (~2–5 µg) was bisulfite modified using an EpiJET™ Bisulfite Conversion kit (Thermo Fisher Scientific, Inc.), as per the manufacturer's instructions. Bisulfite treatment was used to convert non-methylated cytosine ‘C’ bases to thymine ‘T’ bases; however, the methylated cytosine ‘C’ was left unchanged.

The methylation status of the E2F4 promoter was analyzed in the tumor and ANCT tissues from all recruited patients with breast cancer. The target region is located at a site 84 bp upstream and 117 bp downstream from the transcription start site (TSS). Furthermore, methylation-specific primers were designed for the target site, according to National Center for Biotechnology Information Nucleotide database (https://www.ncbi.nlm.nih.gov/nucleotide?cmd=search) and scanned 29 CpG sites within the targeted sequence. The two sets of methylated primers were as follows: Forward, 5′-CGTTACGTTTTTTGGAAGGC-3′ and reverse, 5′-CGTACCGACTTAAAATACCCG-3′. Un-methylated primers were as follows: Forward, 5′-TAGTGTTATGTTTTTTGGAAGGTGT-3′ and reverse, 5′-TTCATACCAACTTAAAATACCCAAA-3′. Promoter methylation status was analyzed using the MethPrimer™ software (version 1.0; http://www.urogene.org/cgi-bin/methprimer/methprimer.cgi; Peking Union Medical Colege Hospital, Beijing, China) available online (31).

The targeted promoter regions around the TSS of E2F4, amplified from bisulfite converted genomic DNA, was performed using MSP reactions. Bisulfite converted DNA was amplified in a one-step MSP reaction, using two independent sets of primers aforementioned (methylated and un-methylated) and Maxima^® Hot Start Taq DNA polymerase (Thermo Fisher Scientific, Inc.). Briefly, bisulfite converted DNA (~50 ng) was amplified with 0.2 µM of each deoxynucleotide triphosphate and primer (both methylated and un-methylated forward and reverse primers) using 1 U Maxima^® Hot Start Taq DNA polymerase.

The PCR conditions for E2F4 were as follows: A single cycle at 98°C for 30 sec (initial denaturation); followed by 35 cycles at 98°C for 10 sec (denaturation), 62°C for 30 sec (annealing) and 72°C for 30 sec (extension); and a single cycle at 72°C for 7 min (final extension) using 1 U Maxima^® Hot Start Taq DNA polymerase. MSP validity was checked by amplifying CpG-methylated human genomic DNA (Thermo Fisher Scientific, Inc.) as the positive control with PCR water (Solis BioDyne, Tartu, Estonia) used as the negative control; however, for the non-methylation-specific PCR, bisulfite-unconverted human genomic DNA was used as the positive control for all primers specific to the E2F4 gene promoter. All PCR reactions were carried out using a BIOER™ Thermal Cycler 9500 (Hangzhou Bioer Technology Co., Ltd., Binjiang, China).

The PCR products from the aforementioned reactions were resolved on a 2% agarose gel and visualized under an UV illuminator (BioDoc Analyzer, Biometra; Analytik Jena, Jena, Germany) using a 100 bp DNA size ladder. To establish the degree of methylation from qualitative MSP data, the change in methylation (Δ meth) value was calculated, as previously described (25,26). The Δ meth value was calculated by subtracting the un-methylation values of a particular sample from its methylation values, when considering un-methylation as the default condition.

Statistical analyses were performed using OriginPro 2015 software (OriginLab Corporation, Northampton, MA, USA). The results are expressed as the mean and (±) standard error of the mean to determine the descriptive statistics for qualitative data. For the analysis of transcript profiles, the expression data of the target gene (E2F4) were normalized against those of the internal control gene β-actin. The correlations between various factors were assessed using the Pearson's correlation coefficient. Depending on the experiment, the statistical significance was determined to 95% confidence intervals using Pearson's correlation coefficient, Kruskal-Wallis and Mann Whitney tests and one-way analysis of variance. P<0.05 was considered to indicate a statistically significant difference. Statistical analysis enabled an investigation of the correlations between the methylation frequency and expression levels of the E2F4 gene.

The promoter methylation status of the E2F4 gene was analyzed through an MSP approach. Of the breast tumors that were observed, ~71.5% were un-methylated at the E2F4 promoter; whereas ~21.5% of breast tumors were determined to be exclusively methylated at this promoter (Fig. 1 and Table I). Conversely, ANCT control tissues were identified as differentially methylated (~51.5% methylated and ~48.5% un-methylated) at the E2F4 promoter. The Δ meth value also exhibited a remarkable difference between the breast tumor (mean=3,450.76) and the ANCT (mean=722.29) tissue samples (Fig. 2A; Table I). This indicates a statistically significant (P<0.001) decrease in E2F4 promoter methylation among patients with breast cancer, as compared with the controls. These findings highlight a possible role for E2F4 promoter methylation in breast cancer prognosis, via its elevated mRNA expression levels (Tables II and III; Fig. 2A).

The promoter methylation status of the E2F4 gene was analyzed in two age-based study cohorts, consisting of patient's ≤45 and >45 years. The study cohort comprising patients ≤45 years of age, exhibited hypo-methylation in ~66.67% of tumor; whereas, ~46.43% of their ANCT controls were hypo-methylated, indicating that non-cancerous tissues have higher methylation levels at the E2F4 promoter (Table I). Similarly, patients in the >45 years cohort exhibited hypo-methylation in ~75% of tumors, as compared with in ~50% of control samples, which again indicated higher methylation levels in controls. In addition, the Δ meth values between each study cohort were identified to be significantly different (P=0.005). Furthermore, the Pearson's correlation coefficient values for Δ meth revealed a strong positive correlation (r=0.68 for >45 years and r=0.46 for ≤45 years) for each study cohort (Tables I–III and Fig. 2B). These findings indicated an age-independent trend in E2F4 promoter methylation status that may have implications for disease prognosis. Similar methylation profiles were also observed in pre- and post-menopausal study cohorts (Table I and Fig. 2C) with 45 years being the mean age for these cohorts. This indicated that the variable methylation signature in the cohorts may be a consequence of a hormonal imbalance in these patients, revealed a mechanistic insight into the whole process of DNA methylation. Patients with and early (≤12 years) or late (>12 years) age of menarche were observed to have statistically significant (P=0.010) variations in their promoter methylation status (Table I and Fig. 2D); the results demonstrated that there was a strong positive correlation for early (r=0.64) and late (r=0.61) age of menarche patients (Table II). This indicates that early age of menarche is a putative risk that may result in notable methylation variations.

Similarly, promoter methylation status among tumors at different disease stages revealed a gradual reduction in methylation frequency with advanced tumor stage, i.e. from SI to SIII (Table III; Fig. 2E). As the Δ meth was revealed to be statistically significant (P<0.001) among these tumors, it was suggested that promoter methylation had a notable implication for disease progression and prognosis. Furthermore, for the various histopathological types of breast cancer, including invasive ductal carcinoma, invasive lobular carcinoma and ductal carcinoma in situ, the Δ meth value was statistically significant (P=0.001; Tables I and III; Fig. 2F). This demonstrates that methylation may be controlled in a tissue specific manner.

Quantitative mRNA transcript analysis of the E2F4 gene was performed using qPCR with all biopsy samples from the enrolled patients with breast cancer. β-actin was used as the internal standard for gene amplification. It was identified that there was an almost two-fold elevation in E2F4 gene expression in tumor samples when compared with ANCT tissues (Fig. 3); this difference was statistically significant (P=0.022; Table I) and indicated a potential involvement of E2F4 expression in tumor pathogenesis. The upregulation of E2F4 in breast tumor samples was determined to be negatively correlated with its promoter methylation, as per the Pearson's correlation test. The value of Pearson's ‘r’ (Pearson's r=−0.24) indicated a negative correlation with disease outcome, which suggested that a higher E2F4 expression would lead to a poorer outcome. Notably, no correlation (Pearson's r=−0.07) was calculated for the control samples (Table II).

The E2F4 promoter methylation status may correspond to its expression; the degree of E2F4 promoter methylation has a strong positive effect on its transcript expression. Statistical analysis revealed a moderately negative correlation between promoter methylation (Δ meth) and gene expression (r=−0.30) in breast tumor tissues, as well as in the ANCT samples (r=−0.32; Table IV). This demonstrates that methylation in these samples may be the causative factor for their gene upregulation. Furthermore, there was a moderately positive association between the methylation status of controls and relative expression of E2F4 gene (r=0.41); whereas, there was no association observed between tumor methylation and their relative expression (r=−0.03), although, a moderately negative correlation (r=−0.34) between the un-methylation status of these samples and their expression was identified. This demonstrates an involvement of the hypo-methylated status of E2F4 in regulating its expression in breast tumor samples (Table IV).