Beclin 1 immunoreactivity was predominantly found in the cytoplasm, but also in the nuclei and on membranes (Figure 1A,B). The positive expression rate of Beclin 1 was 80.41% in ovarian epithelial cancers, 73.08% in borderline tumors, 23.08% in benign tumors, and 20.00% in normal ovarian tissues (Table 1). The positive rates of Beclin 1 were significantly higher in ovarian epithelial cancer and borderline tumor than in benign ovarian tumor or normal ovarian tissue (all p < 0.001; Table 1).

We also examined the ovarian carcinoma tissues of 20 patients, including 15 serous adenocarcinomas and five poorly differentiated adenocarcinomas, and 10 samples of normal ovarian tissues (excised during the surgical removal of cervical cancers). We used immunoblotting to detect the expression of Beclin 1 at the protein level and found that the levels of Beclin 1 were higher in the ovarian carcinoma than in the normal ovarian tissues (Figure 2A,B).

In total, 148 cases of ovarian cancer were divided into a Beclin 1 high-expression group (++/+++) and a Beclin 1 low-expression group (−/+). The rate of high Beclin 1 expression differed significantly between patients with well-moderately differentiated ovarian cancers (69.1%) and those with poorly differentiated cancers (51.9%; p < 0.05). However, the Beclin 1 levels were not associated with pathological type, surgical stage, or lymph-node metastasis (all p > 0.05; Table 2).

Cox’s regression analysis was performed using surgical stage, differentiation, pathological type, lymph-node metastasis, residual lesion size, and Beclin 1 expression as the dependent variables, and survival time as the independent variable. The results showed that surgical stage and Beclin 1 expression were independent risk factors for ovarian cancer prognosis (Table 3).

As of December 2013, the follow-up times ranged from 54 months to 103 months, and 24 deaths occurred in the Beclin 1 low-expression group (30 patients) and 35 deaths in the high-expression group (63 patients). A Kaplan–Meier analysis showed that the survival rate was significantly higher in the Beclin 1 high-expression group than in the low-expression group (p < 0.05; Figure 3A). The survival rate in patients with stage III–IV ovarian cancer (48 deaths) was significantly lower than that in patients with stage I–II disease (11 deaths; p < 0.05; Figure 3B).

Written informed consent was not given by the participants for their clinical records and tissue samples to be used in this study. However, the patient information was anonymized and de-identified before its analysis. The analysis of patient tissues was approved by the Institutional Review Board of Shengjing Hospital of China Medical University (the ethics approval code was “2013PS158K”).

Ovarian tumors (malignant, borderline, and benign) and normal ovarian tissues from patients treated between 2004 and 2009 at Shengjing Hospital, affiliated with the China Medical University, were collected retrospectively. All tissue sections were examined by specialists to make a final diagnosis. Histopathological diagnoses were made using the World Health Organization criteria. The classification of cancer stage and grade was according to the International Federation of Gynecology and Obstetrics (FIGO, 2009). There were 148 cases of primary malignant ovarian tumors, 26 borderline ovarian tumors, 25 benign ovarian tumors, and 30 normal ovarian tissues (excised during the surgical removal of cervical cancers). The clinical and pathological information about the patients was collected from their clinical records, and included their age, surgical stage, lymph-node metastasis, pathological tumor grade and subtype, and residual tumor size.

The age range (median) was 16–76 years (52.7 years) in the malignant ovarian tumor group; 22–77 years (39.3 years) in the borderline ovarian tumor group; 15–81 years (43 years) in the benign ovarian tumor group; and 37–62 years (45.5 years) in the normal ovarian tissue group. There were no statistically significant differences in the ages of these groups (p > 0.05). The specific histological types and pathological grades are shown in Tables 1 and 2.

We collected information on the clinical chemotherapeutic treatments received and the follow-up of 93 patients from a total 148 patients with malignant ovarian cancer. These 93 patients underwent treatment for ovarian cancer that included surgical debulking followed by 6–8 postoperative cycles of conventional chemotherapy consisting of paclitaxel (Taxol^®, Harbin pharmaceutical group, Harbin, China) plus carboplatin and were followed-up for a minimum of three years after the completion of chemotherapy. We defined the overall survival of the patients as extending from the date of surgery to the date of death or the last follow-up. After the operation, patients were observed at 6-month intervals. To determine the factors influencing survival after surgery and standard chemotherapy, conventional variables together with Beclin 1 expression were tested in 93 ovarian carcinoma patients.

Paraffin-embedded histological sections of each group of ovarian tissues were cut to 5 μm. Immunohistochemistry was used to analyze the Beclin 1 expression levels. Rabbit monoclonal anti-Beclin 1 antibody (ab51031; diluted 1:180) was bought from Abcam Company (Cambridge, UK). The staining procedure was performed according to the manual of an ultrasensitive streptavidin–peroxidase kit (MaiXin.Bio, Kit 9701, Fuzhou, China) and Harris’s hematoxylin (Sairuida.Bio, Tianjin, China) was used to stain the cell nuclei. Tissues treated with phosphate-buffered saline (PBS) instead of the primary antibody were used as the negative control. A colon cancer tissue sample was used as the positive control. Buff-colored granules in the cell cytoplasm and nucleus were considered a positive result. The tissues were rated according to their chromatic intensity: no pigmentation = 0, light yellow = 1, buff = 2, and brown = 3. We chose five high-power fields in serial sections from each slice, scored them, and estimated the mean percentage of chromatic cells: <5% chromatic cells = 0; 5%–25% chromatic cells = 1; 26%–50% chromatic cells = 2; 51%–75% chromatic cells = 3; and >75% chromatic cells = 4. We multiplied these two numbers (intensity score × percentage chromatic cells): 0–2 was considered (−); 3–4 (+); 5–8 (++), and 9–12 (+++). The total 148 cases of ovarian cancer were divided into the Beclin 1 high-expression group (++/+++) and the Beclin 1 low-expression group (−/+). Two observers read the sections to control for error.

The soluble proteins were isolated from the tissues for western blotting. The protein concentrations was measured by Bicinchoninic acid (23228, Thermo Fisher Scientific, Waltham, MA, USA). Equal amounts of protein from each sample were separated by electrophoresis on an SDS-10% polyacrylamide gel, transferred to polyvinylidene difluoride membrane, and blocked with 5% nonfat dry milk in 1× TBS plus 0.1% Tween 20 at room temperature for 2 h. The membranes were incubated overnight at 4 °C with the primary antibody in 1% albumin bovine in 1× TBS plus 0.1% Tween 20. The primary anti-Beclin 1 monoclonal antibody (diluted 1:2000) was purchased from Abcam Company (ab51031) and the anti-GAPDH (glyceraldehyde-3-phosphate dehydrogenase) monoclonal antibody (diluted 1:2000) was purchased from Boshide Biotech Company (BM1985, Wuhan, China). The membranes were washed and incubated again for 2 h at room temperature with horseradish-peroxidase-conjugated anti-rabbit or anti-mouse secondary antibody. The proteins were visualized with ECL reagent (ECL Prime Western Blotting Detection Reagent, Amersham, Pittsburgh, PA, USA). The experiments were repeated three times.

Differences in proportions were evaluated using the χ² test. The χ² test or Fisher’s exact test was used to analyse the relationship between the Beclin 1 expression and clinicopathological variables, whichever was appropriated. Survival curves were generated using the Kaplan–Meier method and compared by the log-rank test. Cox’s proportional hazard regression model was used for multivariate survival analysis of prognostic factors. All statistical analyses were performed using SPSS V17.0 software (SPSS Inc., Chicago, IL, USA). A two-tailed p-value test was used in all analyses; p-values less than 0.05 were considered statistically significant.