Background
Ovarian cancer accounts for approximately 3% of all cancers in women and has the highest mortality of all cancers of the female reproductive system. This reflects, in part, a lack of early symptoms and proven ovarian cancer screening tests. Malignant surface epithelial tumors (carcinomas) are the most common ovarian cancers, accounting for 90% of cases. These tumors differentiate during malignant transformation into four major histotypes: serous, mucinous, endometrioid and clear cell. The overall picture suggests that ovarian cancer, like other cancers, is a spectrum of diseases and not a single disease entity [
1]. Recently, gene expression profiling studies have indicated that the transcription factor PAX8 is a potential diagnostic marker for ovarian carcinoma [
2].
PAX8 is a member of the PAX gene family, consisting of nine well-described transcription factors (PAX1-9). The temporal and spatial expression patterns of PAX genes are tightly regulated, and their expression is observed primarily during fetal development [
3]. In most cases, PAX gene expression attenuates when development is complete, but in a few tissues, it persists into adult life. However, abnormal cell growth and proliferation is often associated with high expression levels of PAX genes [
4]. Nevertheless, the precise role that PAX genes play in cancer is still unclear. Cancer-promoting PAX genes might not themselves be responsible for cells shifting to a tumorigenic state, but following tumor onset, PAX genes clearly play important roles in the acquisition of characteristics that define malignancy. In fact, overexpression of PAX proteins
per se does not appear to be an initiating or transforming molecular event in tumor pathogenesis, but it facilitates malignant development through the effects of PAX genes on apoptosis resistance, tumor cell proliferation and migration, and repression of terminal differentiation [
4].
PAX8 plays a key role in thyrocyte differentiation [
5]. It is expressed during the organogenesis of the thyroid gland, Mullerian tract, and kidney, as well as in the adult thyroid and kidney [
6]. Knockout mice lacking PAX8 have a smaller thyroid, with normal calcitonin-producing parafollicular C cells but no follicular cells; thus, they suffer from severe hypothyroidism [
7]. Congenital hypothyroidism is caused by several genetic defects, and among these there are mutations of the PAX8 gene [
8]. In addition to hypothyroidism, PAX8 plays a role in the progression of follicular thyroid carcinomas and adenomas [
9] and is overexpressed in the majority of gliomas, Wilms tumors and well-differentiated pancreatic neuroendocrine tumors [
10‐
12]. Interestingly, aberrant expression of PAX8 has been reported in epithelial ovarian cancer [
13], and it was described as one of the top 40 genes specifically upregulated in different types of ovarian carcinomas [
14]. PAX8 is not expressed in the surface epithelial cells of the ovary; however, recently its expression was found in 96% of serous ovarian carcinomas, in 89% of endometrioid and 100% of clear cell carcinomas, whilst was not detected in mucinous carcinomas [
9]. Recently, it has been demonstrated that high-grade serous carcinoma (HGSC) originates in fallopian tubal secretory epithelial cells, which are positive for PAX8 expression [
15].
Our studies provide strong evidence that PAX8 plays an important role in the tumorigenicity of ovarian cancer cells both in vitro and in vivo and identify PAX8 as a major biomarker and target for ovarian cancer.
Methods
Cell culture and DNA transfection
The human ovarian carcinoma cell lines SKOV-3, TOV-21G, OVCAR-3, TOV-112D and A2780 were obtained from the CEINGE Cell Culture Facility (Naples, Italy) and were grown in RPMI medium (Euroclone) containing 10% fetal bovine serum (Euroclone). The medullary and cortical cells were kindly provided by Prof. Lucio Nitsch (University of Naples, Italy) and were maintained in CHANG MEDIUM C lyophilized kit (Irvine Scientific). The nontumorigenic ovarian cells IOSE-80PC were obtained by Canadian Ovarian Tissue Bank and were grown in medium 199:MCDB 105 (Sigma-Aldrich) containing 10% fetal bovine serum. For stable transfection experiments, cells were plated at 5 × 105 cells/100-mm tissue culture dish 24 h prior to transfection. Transfections were carried out with the Lipofectamine (Invitrogen) and FUGENE reagent (Promega) for SKOV-3 and IOSE-80PC cells, respectively, according to the manufacturer's directions. Forty-eight hours later, transfected cells SKOV-3 and IOSE-80PC were selected in the presence of 0.4 μg/ml of puromycin (Sigma-Aldrich) and 0.2 mg/ml of G418 (Gibco), respectively.
Total RNA was extracted using TRIzol reagent (Invitrogen) and cDNA was synthesized using iScript cDNA Synthesis kit according to the manufacturer’s instructions (Biorad). Subsequently, cDNA was used for each PCR reaction with each primer pair. The PAX8 specific primers designed to detect PAX8 splice variants were previously described [
13]. Real time RT–PCR analysis was performed using IQ™ SYBR Green PCR Master Mix (Biorad) in a iCycler IQ™ real-time detection system (Biorad).
Cell extracts and Western blot
Cell extracts and western blot were carried out as previously described [
16].
shRNA, plasmid and antibodies
Five shRNA targeting PAX8, Mission shRNA lentiviral plasmids (SHCLNG-NM_003466, Sigma-Aldrich) and MISSION Non-targeting shRNA control vector (SHC002, Sigma-Aldrich) were used. The 3XFLAG-Pax8 expression vector was previously described [
16]. The antibodies used for immunofluorescence and immunoblotting were: Pax8 (kindly provided by Prof. R. Di Lauro), fibronectin clone IST4 (Sigma-Aldrich), vimentin, twist, vinculin, actin and α-tubulin (Santa Cruz Biotechnology).
Cell proliferation and invasion assay
To evaluate cell growth, SKOV-3, SKOVCtrl-, siCl32, and siCl48 cells were plated at 8 × 104 cells per 60-mm plate. The medium was changed every 24 h, and every 24 h cells were collected and counted. Cell invasion assay was examined using a reconstituted extracellular matrix (Matrigel; BD Biosciences). Filters (8 μm pore size) on the bottoms of the upper compartment of the transwells (6,5 mm; Corning) were coated with 2 mg/ml of matrigel. 2 × 105 cells were suspended in 100 μl of RPMI with 0.2% FBS. The cells were then plated onto the coated wells and incubated at 37°C for 16 h. Medium in the lower compartment was supplemented with 5% FBS as a chemoattractant. Noninvading cells were removed from the top of the wells with a moistened cotton swab. Cells that penetrated the membrane were fixed with 11% glutaraldehyde and stained with 0.1% crystal violet. The concentration of solubilized crystal violet in 10% acetic acid was evaluated as absorbance at 590 nm. Results are ± SD of three independent experiments.
Immunofluorescence and Confocal Laser Scanning Microscopy
Cells were grown directly on glass coverslips for 48-72 h, fixed in 4% paraformaldehyde in PBS for 20 min at room temperature, permeabilized for 10 min in 0.2% Triton X-100 in PBS, and incubated for 60 min in 0.5% BSA (bovine serum albumin) in PBS. The coverslips were subsequently incubated at 4°C for 1 h with rabbit polyclonal anti-PAX8 diluted 1:1000 in 0.5% BSA in PBS and, after PBS washing, incubated for 30 min with Alexa Fluor-594 goat anti-rabbit IgG (Vinci Biochem) diluted 1:200 in 0.5% BSA in PBS. After final washings with PBS, the coverslips were mounted on a microscope slide using a 50% solution of glycerol in PBS. Images were collected with a Zeiss LSM 510 confocal laser scanning microscope, equipped with a 543 nm HeNe laser, and a Plan-Apochromat 63/1.4 oil immersion objective. Emitted fluorescence was detected using LP 560 long pass filter for TRITC.
Wound-healing assay
Confluent SKOVCtrl-, siCl32, and siCl48 cells plated on tissue culture dishes were wounded by manual scratching with 200-μl pipette tip, washed with PBS and incubated at 37°C in complete media. At the indicated time points, phase contrast images at specific wound sites were captured.
Anchorage-independent growth in soft agar
Cells were mixed in RPMI 2X (Sigma-Aldrich Aldrich), tryptose phosphate buffer, and 1.25% of Noble Agar (Difco Laboratories Inc.) and plated in 60-mm dishes on the top of 1% agar base. The colonies were allowed to grow in incubator at 37°C, 5% CO2 for 2 to 3 weeks. The images of cell colonies were captured with an inverted microscope.
Animal experiments
All animal studies were conducted at Biogem Scarl Ariano Irpino, AV (Italy), Preclinical Research and Development Service. Nude female NOD-SCID mice (NOD-CB17/PRKDC/J) were purchased from Charles River Laboratories International, Wilmington, MA. Animals have been housed and used following the rules of the Italian laws (DL.vo N° 116 - 27/01/1992 and related) and of the EU directive (2010/63/UE - 22/09/2010) on the protection of animals used for experimental purposes. All the in vivo procedures were in compliance with the Guide for the Care and Use of Laboratory Animals (United States National Research Council, 1996). All the in vivo experimental activities were evaluated and approved by the Committee for the Ethics of the Experimentations on Animals (CESA) of Biogem (ID code 4215) and were authorized by the Italian Minister of Health.
To generate xenografts, human ovarian cancer cells were cultured in DMEM with 10% heat-inactivated FBS. 24 six-week-old nude female NOD-SCID mice (NOD-CB17/PRKDC/J) (Charles River Laboratories International) were randomly assigned to four groups: the SKOV3 group (n = 6), SKOV3Ctrl- group (n = 6), siCl32 group (n = 6) and siCl48 group (n = 6). They were injected subcutaneously in the both flanks with 7×106 cells suspended in 0.2 ml PBS/Matrigel Matrix GF (1:1) (BD Biosciences). Mice were daily monitored for clinical signs and mortality. Body weight recordings were carried out weekly. Tumor growth was measured twice a week with a Mitutoyo caliper. The formula TV (mm3) = [length (mm) × width (mm)2]/2 was used. At the end of the study mice were sacrificed by cervical dislocation.
Discussion and Conclusions
Epithelial Ovarian Cancer (EOC) is a morphologically and biologically heterogeneous disease and remains a leading cause of morbidity and mortality. It accounts for approximately 3% of all cancers in women and despite considerable efforts to improve early detection and advances in chemotherapy, the highest mortality rate of ovarian cancers has markedly increased worldwide.
There is ample evidence that dysregulated expression and/or activation of specific members of the PAX family appear to play a major role in the progression of specific cancers arising in those organ systems in which PAX proteins exert their developmental functions during embryogenesis [
21], but their precise role in cancer is still unclear. Recently, a genome-scale analysis of 102 cancer cell lines identified PAX8 as a lineage-specific survival gene, highly expressed in ovarian cancer lines and amplified in a substantial fraction of primary ovarian tumors [
22]. It was initially hypothesized that epithelial ovarian cancer derives from the epithelial cells covering the ovary, but recent evidences showed that high-grade serous carcinoma (HGSC) originates in the fallopian tubal secretory epithelial cells, which are positive for PAX8 expression [
15]. Newly, we have demonstrated that PAX8 plays a critical role in cell cycle progression and cell survival of differentiated epithelial cells [
23], reinforcing the crucial involvement of this transcription factor in different biological processes. To investigate the role of PAX8 in ovarian cancer we performed studies
in vitro and in
vivo. We analyzed the expression level of PAX8 in a series of ovarian cancer cell lines, thus we have chosen SKOV-3 cell lines to assess PAX8 involvement in ovarian tumorigenesis. We selected stable cell clones constitutively silenced by sh-PAX8 to detect the function of this transcription factor in an epithelial ovarian cancer cell line. Our results indicated that PAX8 knock-down elicited a dramatic effect on SKOV-3 cell growth, inhibited the invasion rate of these cells through the Matrigel, and reduced the migration rate in wound-healing assay. To assess the ability of PAX8 to inhibit tumor growth
in vivo, we injected SKOV-3 cells constitutively silenced by shPAX8 into nude mice. The results obtained in this study show for the first time that PAX8 is capable of inducing
in vivo tumor growth. The size of palpable lesions well correlates with PAX8 expression level of single clones, confirming the role of PAX8 as oncogene
in vivo.
Recently, consistent with our findings, it was reported that PAX8 transcriptionally regulates E2F1, a key regulator of the G1/S phase of the cell cycle [
24]. To date, the essential role of PAX8 in the development and differentiation of the thyroid gland has been extensively described; nonetheless, its role in other contexts has not been addressed. Our data provide the first evidence of a clear involvement of PAX8 in the
in vivo tumorigenesis of ovarian cancer cells. Given the enormous heterogeneity of ovarian cancer and the enhanced expression of PAX8 only in some epithelial subtypes, it will be very interesting to see if a subset of novel PAX8 target genes is relevant for cancer initiation and/or maintenance, in order to identify novel targets for ovarian cancer therapy.
To further characterize PAX8 effects on cell migration and invasion we analyzed the expression level of some master regulators of EMT [
19] in normal ovarian cell line IOSE-80PC stable transfected with PAX8. Here, we reported that the expression of PAX8 significantly induces SNAIL and MMP13 while inhibits TIMP3, reinforcing the idea that this transcription factor might represent a potential new target for preventing ovarian tumor invasion and metastasis. In conclusion, we believe that the role of PAX8 in cancer is emerging as an exciting research area and promises to deliver many new insights into the onset and growth of ovarian epithelial carcinomas.
Acknowledgments
We sincerely thank Anna Conti and Lucio Nitsch for the primary ovarian cell lines and for helpful discussion. We also thank the Preclinical Research and Development Service of Biogem (AV, Italy) for the in vivo experiments.
This work was supported by grants from the Italian Ministry of Education, University and Research (MIUR-PRIN 2009), from the Italian Ministry of Economy and Finance to the CNR for the Project FaReBio di Qualità and by the grant Medical Research in Italy (MERIT) RBNE08YFN3_001.
Competing interests
The authors declare no conflict of interest.
Authors’ contributions
DPT and ZM designed research; DPT, LV, dCT and FMG performed research; DPT and ZM analyzed data, DPT and ZM wrote the paper. All authors read and approved the final manuscript.