Background
Ovarian carcinoma is the most lethal malignancy of the female genital tract, mainly due to the failure of early diagnosis, heterogeneous histology subtypes and the limitations for the conventional chemotherapies [
1,
2]. The important prognostic factors include tumor stage, age at initial diagnosis, tumor morphological subtypes and grade, optimal resection for advanced ovarian cancer, as well as the effect of chemotherapy following primary surgery [
3]. Current researches have focused on the study of various molecular signaling reactions or pathways in ovarian carcinoma to explore the molecular markers for early detection, prognosis assessment and hopefully as potential therapeutic targets.
Interferon regulatory factor 4 binding protein (IBP, also known as: DEF6) first identified in 2003 [
4], plays multiple important roles in various biological processes that involve the immune system. Loss of DEF6 in mice resulted in the development of systemic autoimmunity and developmental defects at the earliest stage of thymocyte differentiation [
5‐
10]. Besides, DEF6 plays important role in the regulation of cell motility, cytoskeletal rearrangements, focal complex/adhesion assembly, cell polarity and cell migration through the stimulation of actin polymerization [
11‐
17].
Signaling involving DEF6 has been implicated in tumorigenesis. Increased expression of DEF6 has been shown to be correlated with the malignant behavior of extra-skeletal myxoid chondrosarcoma [
18], colorectal cancer [
19], breast cancer cells [
20], and oral squamous cell carcinoma [
21]. DEF6 may serve as a potential target for anti-angiogenic intervention in renal cell carcinoma [
22]. DEF6 is also a novel target of tumor suppressor p53 and can suppress cisplatin-mediated apoptosis of breast cancer cells via a negative feedback regulation of the p53 signaling pathway [
23]. High levels of DEF6 were found to decrease cisplatin-induced growth suppression and apoptotic cell death, in association with decreased p53 activity and imbalanced expressions of the Bcl-2 family members. Recently, p16 has been shown as a prognostic indicator in ovarian/fallopian tubal high-grade serous carcinoma [
24‐
26]. Moreover, both DEF6 and p16 contribute to the regulation of cell cycle progression, and p53 plays important role in the cell cycle checkpoints. Despite their potential close interaction in cell cycle progression, the roles of DEF6, p16 and p53 have not been fully elucidated in ovarian carcinomas.
In the present study, we sought to better understand the differential expression of DEF6, its relation to the expressions of p53 and p16, and the prognostic significance in different histology subtypes of ovarian carcinoma based on a well-defined cohort of ovarian carcinomas. Clinicopathological data and survival curves were compared between patients with different scores of DEF6 to explore the potential of DEF6 as a prognostic marker.
Methods
Patient specimens
We collected cases with surgically operated ovarian carcinomas from the files of Departments of Pathology of Taipei Medical University Hospital and Wan Fang Medical Center between January 1998 and December 2011. We used the diagnostic criteria of the World Health Organization classification [
1] and tumor staging system of the International Federation of Gynecology and Obstetrics (FIGO). The pathological diagnosis was reviewed by at least two pathologists. Cases of secondary metastasis to the ovaries were excluded.
Pathologic variables included histology subtypes (i.e. low- or high-grade serous carcinoma, mucinous carcinoma, endometrioid carcinoma and clear cell carcinoma), stage, therapy, recurrence free interval (if applicable), and site of recurrent disease (if applicable). Debulking surgery was found to be optimal if the maximum diameters of the individual residual tumor deposits were all less than 1.0 cm. Only cases with optimal resection were enrolled and analyzed in this study. Patient information were de-identified and assigned a study number. Tissue microarrays (TMA) with three to four tumor regions were chosen from paraffin embedded blocks. Each tumor core measured 0.3 cm in maximal diameter. This study had been reviewed and approved by the Institutional Review Board of Taipei Medical University (TMU-IRB 99049).
Immunohistochemistry and interpretation
The DEF6 monoclonal antibody (clone 1 F2; Abnova, Taiwan; 1: 3000), p53 (clone: DO-7; Ventena; prediluted), and p16 (clone: E6H4; Ventena; prediluted) were used. Formalin-fixed and paraffin-embedded tissue sections were cut, deparaffinised and rehydrated. Autoclaved retrieval technique by using 10 mM citric acid buffer (10–20 min) and inhibited by endogenous peroxidase activity (0.3 % H2O2; 5 min) were conducted. Tissue sections were incubated with primary antibodies in an automated stainer system (Ventana, BenchMark XT). Tissue sections were then incubated with secondary antibody (dilution rate 1:100, 30 min), peroxidase-conjugated streptavidin (100 μg/mL), and 0.02 % 3,3’-diaminobenzidine tetrahydrochloride (DAB) (0.05 M Tris-HCl buffer with 0.03 % H2O2). All slides were counterstained with hematoxylin and analyzed by two pathologists (P-L Liew and C-L Chen).
The DEF6 cytoplasmic staining results were scored according to the percentages of positive cells: score 0 (less than 5 %), score 1 (5-24 %), score 2 (25-75 %), and score 3 (more than 75 %). The stromal lymphocytes were served as positive internal control. The results of p53 and p16 immunostains were scored according to the percentages of positive cells: score 0 (0 %), score 1 (1-24 %), score 2 (25-75 %), and score 3 (more than 75 %). Score 0 and score 3 of p53 were recorded as aberrant expression; whereas score 1 and score 2 of p53 were designated as insignificant expression.
Cell lines and cell lysates
Total cell lysates from six ovarian carcinoma cells, including A2780, ES-2, TOV-21G, TOV-112D, OVCAR3, and HBT75, were used for the detection of DEF6 expression by Western blotting. A2780 (ovarian carcinoma cell line with unknown disease type identification) cell was cultured in RPMI-1640 with 10 % fetal bovine serum (FBS). ES-2 (clear cell carcinoma), TOV-21G (clear cell carcinoma) and TOV-112D (endometrioid carcinoma) were obtained from ATCC bioresource center (Manassas, VA, USA). ES-2 was cultured in McCoy’s 5 medium with 10 % FBS. TOV-21G and TOV-112D were cultured in a 50:50 mixture of Medium 199 and MCDB105 with 10 % FBS. Cells were grown to near confluent, washed with PBS, and then lysed by RIPA lysis buffer. The lysates of OVCAR3 (serous carcinoma) and HBT75 (serous carcinoma) cells were kindly provided by Prof. Chao-Lien Liu at the Department of Medical Laboratory Science and Biotechnology, Taipei Medical University.
Western blot assay
Lysates from the six ovarian carcinoma cell lines and two positive control oral carcinoma cells, HSC-3 and SCC25, were separated in a 10 % polyacrylamide gel and transferred onto a PVDF membrane. After blocking, the blot was incubated with indicated antibodies overnight at 4 °C. Antibodies against DEF6and β-actin (Genetex, Irvine, CA, USA) were used as the primary antibodies. After washing, the blot was incubated with horseradish peroxidase-labeled goat anti-mouse/rabbit IgG (Jackson Laboratory). The expression profile of the proteins was visualized using a Western Lightening-ECL kit (PerkinElmer, Waltham, MA, USA).
Statistical analyses
We performed statistical analyses by using SPSS for Windows software (SPSS, Chicago, IL). Data were expressed as median (interquartile range), mean (SD), and percentages. Chi-square test, Mann-Whitney U tests, and Student’s t-test were used as statistical methods. Spearman rank correlation and multivariate linear regressions with stepwise variable selection were performed to assess the significant associations between ordinal or continuous predictor variables. The Kaplan-Meier method, Cox proportional hazard regression model and multifactorial Cox regression analysis were used to examine all factors found to be prognostic of survival in univariate analysis and the analysis of DFS and OS. A P-value of less than 0.05 was considered statistically significance.
Discussion
DEF6 is a conserved protein associated with the functions of lymphocytes and is highly expressed in T-cells and T-cell homing organs [
4‐
27]. DEF6 was expressed in breast cancer cells [
20], oral squamous cell carcinoma [
21], colorectal carcinoma [
19], and in tumor vessels of renal cell carcinoma [
22]. The role of DEF6 expression in human cancer is unclear, and the prognostic significance of DEF6 expression and the co-expression of p16 and p53 in ovarian carcinomas are largely unknown.
In this study of 180 cases of ovarian carcinoma using immunohistochemistry, a high expression of DEF6 was commonly found in ovarian carcinomas, and associated with different histology subtypes, advanced FIGO stage, and reduced overall survival (OS) and disease free survival (DFS). Strong expression of DEF6 was observed in high-grade serous carcinoma and endometrioid carcinoma. These findings were supported by the high expression of DEF6 in two serous, two clear cell and one endometrioid carcinoma cell lines, suggestive of the frequent expression of DEF6 in ovarian carcinoma tissues and cells.
There are many molecular markers possessing prognostic value. We compared the clinicopathological parameters and prognostic factors of DEF6, p16 and p53. The p16 is a cyclin-dependent kinase inhibitor which is integral to the retinoblastoma (Rb) gene-mediated control of the G
1-S phase transition of the cell cycle [
28]. It was also shown that ectopic expression of DEF6 shortened the G
1 interval in the cell cycle, and increased cyclin D1 expression [
21]. It is therefore important to explore the cooperation between the two genes in the tumor progression. Although p16 has been widely regarded as a surrogate marker of high-risk human papillomavirus (HPV) in uterine cervical pathology, the role of p16 protein expression in ovarian carcinomas remains limited. In our study (Table
1), we found the highest percentage of score 3 p16 expression (74.7 %) in high-grade serous carcinoma, an ovarian carcinoma with notorious poor outcome, and similarly DEF6 (52 %). Our findings suggest that p16 is also a valuable prognostic marker for ovarian carcinomas. Recent studies have investigated the diagnostic role of p16 and prognostic indicator of p16 with clinical outcome in ovarian/tubal high-grade serous carcinoma. As p16 was expressed in the majority of p53-positive and p53-negative serous tubal intraepithelial carcinomas, the addition of p16 helped to compensate the practical limitations of p53 in the diagnosis of serous tubal intraepithelial carcinomas [
24]. Moreover, a recent study on ovarian/tubal high grade serous carcinomas revealed three distinct subgroups according to p16 expression and RB1 status (i.e., p16 homogenous stain/RB1-, p16 homogenous stain/RB1+, and p16 heterogeneous stain/RB1+), which possessed clinical relevance for stage and patient outcome upon multivariate analysis [
29]. We also assessed the prognostic significance of the status of DEF6/p16 co-expression. Notably, the co-expression of both proteins was associated closely with adverse clinical outcome (Table
2). The OS and DFS, in particular, were the shortest in the group with strong DEF6 and p16 co-expression, followed by DEF6+/p16- and DEF6-/p16+, while the DEF6-/p16- group had the longest survival rates (
P = 0.027 for OS and
P = 0.023 for DFS, respectively). These findings suggest that DEF6 and p16 positively interact and contribute to the tumor progression in ovarian carcinoma, and the molecular mechanisms deserve to be further studied. This suggestion is supported by the observations that both proteins play important role in the control of the G
1-S phase transition of the cell cycle [
21,
28].
DEF6 is a novel p53 target gene and negatively regulated by p53, it can suppress cisplatin-mediated apoptosis of breast cancer cells [
23]. In the present study, we found that strong DEF6 expression and aberrant p53 expression had the shortest survival, whereas co-expression of insignificant p53 expression and negative DEF6 displayed the longest survival rate, with the other two groups in between (
P = 0.031 for OS and
P = 0.028 for DFS, Table
3). Thus, the evaluation of the p53 status coupled with DEF6 might be important for risk stratification, which certainly awaits sequencing validation of the p53 status for further clarification.
We further evaluated the overall survival rates of DEF6, p16 and p53 in all patients and four different histology subtypes of ovarian carcinoma by using Kaplan-Meier analysis. Significantly shorter survival rates were seen in patients with high expressions of DEF6 (P = 0.008) and p16 (P = 0.022). Patients with aberrant p53 expression in high-grade serous carcinoma had shorter overall survival (P = 0.012). Surprisingly, high expression of DEF6 in clear cell carcinoma was significantly correlated to shorter overall survival (P = 0.001). We enrolled in total 41 cases of ovarian clear cell carcinoma (30 FIGO stage I, 4 FIGO stage II and 7 FIGO stage III) in this study. Therefore, our study suggests that patients of ovarian clear cell carcinoma with high DEF6 expression deserve a poor prognostic factor compared with other histological subtypes. Our study is the first report on the prognostic impact of DEF6 overexpression in early-stage ovarian clear cell carcinomas, however in-depth investigations are necessary.
Univariate analysis showed that FIGO stage, DEF6 and p16 were associated with shorter survival (Table
3). Upon multivariate analysis, DEF6 remained the significant prognostic value (Table
4). These findings suggest that DEF6 in ovarian carcinomas may facilitate tumor cell growth or proliferation, motility, invasion and metastasis, leading to high tumor stage and hence poor prognosis.
Abbreviations
DFS, disease-free survival; FIGO, International Federation of Gynecology and Obstetrics; H&E, hematoxylin and eosin; HR, Hazard Ratio; IHC, immunohistochemical; IBP, interferon regulatory factor 4 binding protein; OS, overall survival; TMA, tissue microarray
Acknowledgements
The authors thank Professor Hey-Chi Hsu for his assistance and critical review. We thank Professor Chao-Lien Liu at the Department of Medical Laboratory Science and Biotechnology, Taipei Medical University for providing the lysates of OVCAR3 and HBT75 cells.