Patient clinicopathological data was obtained from the Breast Surgery Department of The First Affiliated Hospital of Zhengzhou University (Zhengzhou, China) between January and June 2013, including 112 specimens from sporadic breast cancer cases and 20 normal breast specimens, with patient ages ranging from 32–78 years (median age, 58.6 years). Written informed consent was obtained from all patients and the study was approved by the ethics committee of Zhengzhou University. The histological grades of the specimens were as follows: grade I, 30 cases; grade II, 50 cases; and grade III, 32 cases. Each specimen was isolated and immediately stored on ice. One half of each section was embedded in paraffin and underwent S-P immunohistochemistry, and the remaining half was frozen in liquid nitrogen for RNA or DNA isolation. Pathological diagnoses of all specimens were clear, and patients had received no treatment prior to their surgery.

S-P immunohistochemistry. Tissue samples from breast cancer and normal breast specimens were studied using S-P immunohistochemistry. Slides were deparaffinized using xylol, rehydrated in a graded alcohol series and subsequently stained with hematoxylin and eosin (Beijing Solarbio Science & Technology Co., Ltd., Beijing, China). For immunostaining, endogenous biotin activity was blocked using the SP-9000 kit (ZSGB-BIO, Beijing, China). An immunostaining assay was performed using two primary antibodies: anti-DNMT1 mouse monoclonal antibody (cat. no. ab134148; 1:500; Abcam, Cambridge, MA, USA) and anti-ER mouse monoclonal antibody (cat. no. ab32063; 1:500; Abcam). Samples were incubated with primary antibody at 37°C for 1 h. Polyclonal goat anti-rabbit immunoglobulin G (cat. no. A0277; 1:1,000; Beyotime Biotech, Jiangsu, China) was used as a secondary antibody, and was incubated with samples for 30 min at room temperature. Slides were washed three times between steps using Tris-buffered saline. Immunoreactions were visualized by a streptavidin-biotin complex using the 3,3′-Diaminobenzidine Chromogenic kit (ZSGB-BIO). The specimens were counterstained using hematoxylin.

Total cell RNA was extracted from breast cancer and normal breast tissue specimens using TRIzol® reagent (Gibco; Thermo Fisher Scientific, Waltham, MA, USA). The A260/A280 absorption of isolated RNA was analyzed using a UV–VIS spectrophotometer (UV-2450; Shimadzu Corporation, Kyoto, Japan) to evaluate the purity and concentration of RNA. A Thermoscript™ RT-PCR System (Fermentas; Thermo Fisher Scientific) was employed to synthesize complementary DNA using 1 µg RNA. The ERα gene was amplified using the following primers: Forward, 5′-TGATGAAAGGTGGGATACGAAA-3′ and reverse, 5′-GGCTGTTCTTCTTAGAGCGTTTG-3′, to create a 168 bp product. The DNMT1 gene was amplified using the following primers: Forward, 5′-CTACCAGGGAGAAGGACAGG-3′ and reverse, 5′-CTCACAGACGCCACATCG-3′, to create a 152 bp product. The β-actin gene was amplified using the following primers: forward, 5′-AGGCATTGTGATGGACTCCG-3′) and reverse, 5′-AGTGATGACCTGGCCGTCAG-3′, to create a 301 bp product, which was utilized as an internal control. The program was monitored and processed using the GeneAmp® PCR System 9700 (Applied Biosystems; Thermo Fisher Scientific), and the thermal cycling conditions used were as follows: an initial heating cycle of 95°C for 5 min, 35 cycles of 94°C for 30 sec, annealing at 65°C (β-actin) for 45 sec or 54°C (ERα, DNMT1) for 30 sec, 72°C for 90 sec and a final 5 min extension step at 94°C. The targeted PCR products were loaded onto 1.5% agarose gels and confirmed using electrophoresis and sequencing. The relative gene expression data was analyzed using the 2^−∆∆Cq method.

Genomic DNA from breast cancer and normal breast tissues was isolated using a DNA Extraction kit (Axygen; Corning Life Sciences, Corning, NY, USA), followed by treatment with sodium bisulfite using the CpGenome™ DNA Modification kit (Epigentek Group, Inc., Farmingdale, NY, USA) according to the manufacturer's protocol. MS-PCR was conducted using the GeneAmp® PCR System 9700.

The unmethylated DNA of ERα was amplified using the following primers: Forward, 5′-GGGGTTGGATGTAGTGGTTTAT-3′ and reverse, 5′-TAAAACTACAAATACCCACCA-3′, to create a 170 bp product. Thermal cycling conditions used were as follows: An initial heating cycle of 94°C for 5 min, 35 cycles of 94°C for 30 sec, 58°C for 45 sec, 72°C for 90 sec and a final 5 min extension step at 72°C. The MS-PCR products were loaded onto 2% agarose gels and resolved by electrophoresis. The band intensities of the reaction products were examined using vision work software LS 6.6a (UVP, Inc., Upland, CA, USA).

All experiments were performed in triplicate and data were analyzed using SPSS version 17.0 (SPSS, Inc., Chicago, IL, USA). The χ² test and Spearman rank correlation coefficient analysis were engaged to assess the univariate association between the correlation of expression of DNMT1 with ERα, as well as the methylation status of ERα and its clinical significance. P<0.05 was considered to indicate a statistically significant difference.

As revealed in Fig. 1, ERα expression is primarily observed in the nucleus, and partially in the cytoplasm; however, ERα was only stained in the cytoplasm in ERα-negative breast cancer samples. The rate of ERα protein positive expression was 95.00% in 20 cases of normal breast tissue, while expression decreased significantly to 67.85% in breast cancer tissue (χ²=6.197; P=0.013).

DNMT1 was characterized by yellow or brown staining in the nucleus and cytoplasm in ERα-negative breast cancer tissue, with weaker staining observed in paracarcinoma tissue (Fig. 2). The rate of DNMT1 protein positive expression was 10.00% in normal breast tissue (n=20), 46.05% in ERα-positive breast cancer tissue (n=76) and 81.11% in ERα-negative breast cancer tissue (n=36). The difference between the three groups was statistically significant (χ²=31.960; P<0.0001).

According to the clinicopathological data of breast cancer patients (Table I), positive expression of ERα protein was associated with progesterone receptor expression (PR; P<0.05), but not with age, menopause state, tumor size, number of positive nodes, Tumor-Node-Metastasis (TNM) stage and tumor type (P>0.05). The protein and mRNA expression levels of DNMT1 were negatively correlated with the expression of ERα (P<0.05), but positively correlated with the methylation of ERα (P<0.05).

The positive expression rate of ERα mRNA was 100.00 and 70.54% in normal breast and breast cancer specimens, respectively (χ²=7.857; P=0.005; Fig. 3). The positive rate of DNMT1 expression was 15.00, 48.68 and 88.89% in the normal breast tissue group (n=20), ERα-positive breast cancer group (n=76) and ERα-negative breast cancer group (n=36), respectively, and the difference between the three groups was statistically significant (χ²=30.794; P<0.0001; Fig. 4). The mRNA expression levels of DNMT1 were negatively correlated with ERα expression in normal breast and breast cancer specimens [Coefficient of rank correlation (r_s)=−0.470; P<0.0001].

The present study used the MS-PCR method to detect the methylation status of ERα in the normal breast tissue group (n=20), ERα-positive breast cancer group (n=76) and ERα-negative breast cancer group (n=36). The results of the present study revealed that the hypermethylation rates were 10.00, 36.84 and 86.11%, respectively (χ²=36.292; P<0.0001; Fig. 5). Correlation analysis of Spearman rank revealed that DNMT1 expression and the methylation status of ERα were significantly positively correlated in breast cancer (r_s =663; P<0.0001).

In conclusion, the results of the present study indicated that DNMT1 protein and mRNA expression levels were negatively correlated with the expression of ERα, but positively correlated with methylation of ERα.