Background
The combination of mutation and aberrant expression of tumor suppressor genes is critical in cancer susceptibility and tumor progression. BRCA1 protein plays multiple essential functions such as tumor suppressor, transcriptional regulation and DNA repair in normal epithelial cells and stem cells [
1]. An inherited
BRCA1-mutation confers an increased risk of ovarian cancer, with lifetime risk estimates ranging from 10-60%, compared to a risk of less than 2% for the general population [
2‐
4]. About 10% of women presenting with ovarian cancer carry a BRCA-mutation. Previous publications indicate that a
BRCA1 mutation is associated with cancer progression through pathways of cell proliferation [
5], differentiation [
6], and apoptosis [
7]. It is known that loss of BRCA1 function may ultimately activate JAK-STAT pathways and stimulate cell proliferation in breast, prostate, lung and ovarian cancer [
8,
9]. But it is still unclear what molecular mechanisms and targets characterize the early molecular events of malignant transformation.
Chronic inflammatory microenvironments have been hypothesized as the major factors predisposing ovarian [
10,
11] and other cancers [
12]. Lipid mediators such as lysophosphatidic acid (LPA) and prostaglandin with their associated receptors and pathways such as COX have been shown to play a critical role in cancer initiation and progression [
13,
14]. Unfortunately, platelet activating factor (PAF, 1-O-alkyl-2-acetyl-sn-glycero-3-phosphorylcholine), as one of the most potent lipid mediators, has not yet been well studied in the regulation of early events of cancer transformation and progression [
15], particularly with the at-risk
in vitro and
in vivo models. PAFR belongs to the G protein-coupled receptor (GPCR) protein family, and transduces cell signals via the G proteins and associated protein phosphorylation cascades [
16,
17]. When cells are exposed to PAF, it induces cell proliferation, activates tyrosine kinase [
18] and protein phosphorylation [
19] in human epithelial cells, skin fibroblasts [
20], endothelial cells [
21], lung fibroblasts cells [
22], pulmonary vascular smooth muscle cells [
23] and keratinocytes [
24]. PAF plays significant roles in many biological pathways in inflammatory diseases and cancer progression [
25,
15,
26]. Upon PAF/PAFR activation, the Signal Transducers and Activators of Transcription (STAT) pathways are activated by phosphorylation changes, dimerization, and translocated into the nucleus to activate transcription of specific genes in regulation of cellular functions [
18,
27]. Our earlier study demonstrated that platelet activating factor (PAF) and PAFR play a significant role in ovarian cancer progression and invasion through activation of a set of tyrosine phosphor-EGFR/Src/FAK/Paxillin[
15]. In this study, we investigate the possibility that inflammation associated lipid mediator PAF might mediate the early BRCA-carcinogenic events using an
in vitro at-risk model employing cancerous and non-cancerous ovarian
BRCA1-mutant epithelial cells with over expression of PAFR.
Methods
Chemical reagents
DMSO, PAF, and ginkgolide B (>90% high-performance liquid chromatography grade), cell culture mediums of MCDB-105 and medium 199 were obtained from Sigma-Aldrich (St. Louis, MO). MEGM mammary epithelial cell growth medium was purchased from Lonza (Walkersville, MD) and RPMI 1640 from Invitrogen (Carlsbad, CA). CV-3988 (a selective inhibitor of PAFR) was purchased from Biomol International L.P (Plymouth Meeting, PA). Rabbit polyclonal antibody against PAFR was purchased from Cayman chemical company (Ann Arbor, Michigan). Phospho-kinase array kit and polyclonal antibody against phosphor-STAT1 was purchased from R&D (Minneapolis, MN). Rabbit polyclonal antibody against STAT1 and monoclonal antibody against FAK were purchased from Cell Signaling Technology (Boston, MA) and Invitrogen-Biosource International Inc. (Carlsbad, CA), respectively.
Immortalized BRCA1-mutant human ovarian surface epithelial cells (HOSE-636, and HOSE-642) were generated from women who underwent prophylactic oophorectomies because of predisposing BRCA1 mutation. Immortalized normal human ovarian surface epithelial cells (HOSE-E6E7 and HOSE-27) were derived from primary cultured cells of the fresh ovarian scrapings at the time of surgery for benign ovarian conditions without BRCA1 mutation. The UWB1.289 (here after referred to as UWB1) cell line is a serous-type ovarian cancer with a BRCA1 mutation (ATCC American Type Culture Collection, Manassas, VA) and OVCA429 cell line is a serous ovarian cancer without a BRCA1 mutation.
Cell lines and cell culture
HOSE-E6E7, HOSE-27, HOSE-636, HOSE-642, and OVCA429 were cultured in M199/MCDB-105 with 15% fetal bovine serum (Gemini Bioproducts, West Sacramento, CA) and 1% antibiotic (200 mmol/L L-glutamine, 10,000 units penicillin, and 10 mg/mL streptomycin). UWB1 cells were cultured in MEGM/RPMI with 3% fetal bovine serum and 200 μg/ml Geneticin (GIBCO-Invitrogen, Carlsbad, CA). Cell culture conditions were maintained at 37°C under 5% CO2 and 95% air in a high-humidity chamber. All non-commercial cell lines were constructed from material collected under IRB approved protocols for collection of "de-identified" human subjects.
BRCA1 mutation characterization analysis
BRCA1 exon sequence analysis: Genomic DNA was isolated from HOSE-636 cells with the Puregene Trial kit (Gentra Systems, Inc.), following the manufacturer's instructions. DNA was precipitated using a standard protocol using saturated NaCl and isopropanol. The DNA pellet was washed with 75% ethanol, air dried, and resuspended in Tris-EDTA (50 mM, pH6.8). Direct DNA sequencing for
BRCA1 was done for all coding exons using Big Dye Terminator chemistry and an automated 3100 DNA sequencer (Applied Biosystems, Foster City, CA). In addition, HOSE-636, HOSE-642 and UWB1 cells with
BRCA1 mutation were certified having the BRCA1 protein truncation by Western Blot according to the published method [
28].
Western blot
The cultured cells were washed twice with 1 × PBS and lysed with ice-cold cell lysis buffer (Cell Signaling Technology, Boston, MA). Protease inhibitors including phenylmethylsulfonyl fluoride (1 mM) and Protease Inhibitor Cocktail I and II (Sigma-Aldrich, St. Louis, Mo) were freshly added to the cell lysis buffer. After a brief vortexing and centrifuging at 4°C, the cell lysates were transferred to a new tube. The concentration of protein was determined with the Bradford Protein Assay Kit (Bio-Rad Laboratories). Protein lysates were subjected to 8-16% SDS-PAGE gel separation followed by transferring onto polyvinylidene difluoride membrane (Perkin-Elmer) using the Wet Transfer Cell (Invitrogen, Carlsbad, CA) and blocked with 5% fat-free dry milk at room temperature for 1 hour. After washing four times, membranes were incubated with primary antibody (Anti-BRCA1 antibody, Cell Signal Technology, Boston) in blocking solution (1:1000 dilution) at 4°C for overnight. Secondary antibody (1:1000) tagged with HRP was used to reveal the protein expression signals with chemiluminescent substrate kit (Pierce Chemical, Co.). Similar western blot protocol has been used for other protein detection including PAFR, STAT1 and FAK.
Immune staining of PAFR in human tissue
Immunohistochemical analysis was performed for detecting PAFR protein expression in sixteen wild type and eight BRCA1-mutant patients using the EnVision/AP system (DakoCytomation) which used an anti-rabbit immunoglobulin conjugated to an alkaline phosphatase polymer (Labeled polymer-AP) with a red stain as positive after the addition of the Substrate-Chromogen solution. Slides were washed with deionized water and counterstained with hematoxylin and 5% ammonium hydroxide, and mounted in Accergel (Accurate Chemical and Scientific Corp). The image was recorded by a digital camera (Optronic Inc.). Human pancreatic tissue sections were used as a positive control.
Cell treatment and proliferation assay
Cell proliferation was assessed by using a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay (Promega, Madison, WI). Monolayer cells at 60-80% confluence were enzymatically removed using trpysin/EDTA and plated in 96-well flat-bottomed plates at a concentration of 1 × 10
5 per well. After overnight plating and starving for 24 h in FBS free medium, HOSE-E6E7, HOSE-27, HOSE-636, HOSE-642, and UWB1 cells were treated with different concentrations of PAF in 0.5% FBS medium daily for 3 days, raging from 0.1 nM to 100 nM. For details of PAF treatment and MTT assay, see our previous publication [
15].
Cell cycle and apoptosis analysis
Immortalized human ovarian surface epithelial cells HOSE-E6E7and HOSE-642 were rendered at confluence by incubation in M199/MCDB105 medium with FBS free overnight. After treatment with 100 nM PAF 24 and 72 h, cells were harvested and fixed with 70% cold ethanol in PBS buffer by suspending the cell pellet and incubating at -20°C within 7 days. The cells were resuspended in PI master mix (PI:40 ug/ml and RNase:100 ug/ml in PBS buffer without Ca
2+ and Mg
2+) at a final cell density of 0.5 × 10
6 cells/ml. After transferring to a Falcon 2054 (Fisher# 149592A) tube and incubation at 37°C for 30 minutes, cells were subjected to analysis using a FACS Calibur™ Flow Cytometer (Becton-Dickson, San Jose, CA). Each experiment was repeated three times and the cell cycle profiles, apoptosis index were analyzed by ModFit LT software (Verity Software House, Inc., Topsham, ME) [
29].
Kinase microarray analysis
UWB1 cells were incubated in fresh RPMI/MEGM FBS-free medium overnight at 37°C in a humidified chamber, which were pretreated with ginkgolide B (GB,100 μM, as specific PAFR antagonist) 3 h before PAF (100 nM) treatment. Cells were harvested at certain time points after treatment by twice washing with 1 × PBS, then lysed with ice-cold cell lysis buffer (Cell Signal Technology), including protease inhibitors.
Phospho-Kinase Array kit (R&D, Minneapolis, MN) was used to detect the relative levels of phosphorylation of 46 kinase phosphorylation sites. The human Phospho-Kinase Array is divided into A and B parts with different sets of protein kinase targets, to maximize sensitivity and minimize cross-reactivity. Capture and control antibodies were spotted in duplicate on nitrocellulose membranes (A and B). First, the membranes were incubated with 1 mL of blocking buffer in 8-Well Multi-dish for 1 h on a rocking platform. About 310 μg of fresh protein lysates were diluted in 1 ml array buffer 1 (1:5) and incubated (A and B membrane) overnight at 4°C. Membranes were then removed to the same container and washed 3 times using 1 × washing buffer. A detection antibody cocktail (for A and B membrane, respectively) diluted with 1 × array buffer was incubated for 2 h at room temperature. After an additional 3 washes, 30 min for each, streptavidin-HRP (1:2000) was used to reveal protein phosphorylation by chemiluminescent kit (Pierce Chemical, Co.) and quantified via an Axon Genepix scanner (Molecular Devices, Co.) and analyzed with ArrayVision software.
Co-immunoprecipitation
A 400 μg protein lysate of HOSE-642 cells with PAF treatment were incubated with 5 μg antibodies of PAFR or phosphor-PAK with 400 μL immunoprecipitation buffer for 2 hours at 4°C. Protein A/G agarose beads were added for overnight incubation. After five times wash, the protein of interest was eluted by SDS/reducing sample buffer and boiling for 3 min and then subjected to 7.5% SDS-PAGE gel separation. The protein was transferred to PVDF membrane (Perkin-Elmer) using the wet transfer cell (Invitrogen, Carlsbad, CA) and blocked with 5% fat-free dry milk at room temperature for 1 hour. After washing four times, membranes were incubated with different primary antibodies phosphor-FAK (1:1000 dilution) for PAFR co-immuneprecipitation, and PAFR (1:1000 dilution) for FAK co-immuneprecipitation, respectively, at 4°C overnight. Following an additional three washes, protein expression signals were detected by HRP with chemiluminescent kit (Pierce Chemical, Co.) as described before. Protein IgG was used as control to confirm the immune specificity of protein-protein interaction and precipitation. A similar approach was used to define the protein-protein interaction between FAK and STAT1.
Statistical analysis
The phosphor-kinase array data were analyzed by Origin 7.0 software. All other data were analyzed by using one-way factorial analysis of variance tests. Data collected from the treated and untreated cells with different dose or at different time point were analyzed by paired t-test and ANOVA test, respectively. The significance of protein expression on western blots and the immune staining intensity of PAFR in fallopian tube epithelial cells of patients with or without BRCA1+ mutation were analyzed by paired Student t-test.
Discussion
BRCA1 is a tumor-suppressor gene in which germ-line mutations lead to a predisposition of breast and ovarian cancer [
35]. Normal BRCA1 regulates multiple nuclear processes including DNA repair and recombination, checkpoint control of the cell cycle [
36]. There is growing interest in the identification of the early events and molecular signals involved in the predisposition to cancer associated with BRCA mutations [
11]. The present study provides the first evidence to demonstrate that
BRCA1-mutation related cancer risk and malignant transition may be associated with accumulation of potent pro-inflammatory lipid PAF and over expression PAFR in non-malignant ovarian epithelial cells. Over-expression of PAFR has been observed in malignant melanoma [
37] and ovarian cancer [
15]. We have found that PAFR is also over expressed in non-cancerous
BRCA1-mutant epithelial cells of the ovarian surface and distal fallopian tube, compared to the wild type normal HOSE cells and fallopian tube surface epithelial cells. This observation may require additional confirmation with a larger set of non-malignant
BRCA1-mutant cells and tissue specimens. However, our study provides clear evidence that potent pro-inflammatory mediators such as phospholipid PAF and its receptor PAFR play significant roles in inducing anti-apoptosis and malignant potential of the at-risk epithelial cells in coordination with BRCA1 dysfunction (Fig.
5 and
6).
In addition, we found that cell proliferation was increased ~20% in all
BRCA1-mutant HOSE cell lines by PAF treatment. This was less than the induced effect observed in many ovarian cancer cell lines (>100% increase) [
15], but significantly decreased cell proliferation in normal wild type HOSE cell lines. Moreover, PAFR inhibitors blocked the PAF-induced cell proliferation and p-STAT expression in all
BRCA1-mutant cell lines. This suggests that over expression of PAFR could be an essential sensor for the
BRCA1-mutant surface epithelial cells to probe potent lipid inflammatory mediators found in the micro-environment, including the conditions of incessant ovulation cycles [
38,
39]. Correspondingly, PAF-induced differential effects in cell proliferation, cell cycle and apoptosis in the wild type and non-malignant
BRCA1-mutant HOSE cells provide additional confirmation that BRCA1-mutation and dysfunction is associated with the phenotypic PAF response, especially under inflammatory conditions through the PAFR pathway (Fig.
5). These combined findings suggest that dysfunction of BRCA1 and over expression of PAFR may contribute and account for the greater likelihood of malignant transformation of BRCA1-mutant at-risk epithelial cells, compared to wild type ovarian or tubal epithelium.
Signal transducers and activators of transcription (STATs) are a family of cytoplasmic proteins that function as signal messengers and transcription factors involved in cellular responses to cytokines and growth factors [
40]. A line of evidence supports that STAT pathways play a significant role in malignant transformation [
8,
41]. For example, a recent report showed that STAT1 phosphorylation determines Ras oncogenicity through p27 protein [
42]. Activated STAT1, STAT3 and STAT5 are often observed in solid and liquid tumors, and are required for tumor transformation and progression which involves a set of oncogenes such as EGFR, Ras, Src, and FAK [
8,
43‐
45]. Here we found that phosphorylation of STAT1, as well as STAT4 and STAT6 (not shown) were simultaneously activated by PAF treatment in
BRCA1- mutant ovarian cancer cells (UWB1). Activated STAT1 phosphorylation was further confirmed in two additional
BRCA1- mutant HOSE cell lines, but not shown in wild type HOSE-E6E7 and ovarian cancer (OVCA429) cells (Fig.
4). In addition, PAF-induced STAT1 phophorylation was partially blocked by PAFR inhibitors in
BRCA1- mutant HOSE-642 and UWB1 cells. This suggests that PAFR is involved in STAT1 activation in
BRCA1- mutant HOSE cells. In the BRCA1-wild type cells, functional BRCA1, as a tumor suppressor, acts in concert with STAT1 to activate transcription of a subset of IFN-γ gene targets and growth inhibition by cytokine [
46]. Here we provide independent evidence that coordination of abnormal function of BRCA1 and activation STAT1 phosphorylation could induce HOSE cell proliferation and anti-apoptosis in
BRCA1- mutant cells, particularly under inflammatory conditions.
Focal Adhesion Kinase (FAK) is a key mediator of signaling induced by integrins that play an instrumental role in many cellular functions including cell survival, proliferation [
47] and stem cell signaling [
48]. FAK is often associated with cancer transformation [
44], progression, and metastasis [
15] through the site specific phosphor-activation. Our findings are consistent with many other observations that FAK directly interacts with STAT1 [
33,
34], through phosphorylation of FAK to regulate the cellular functions such as cell apoptosis, migration, invasion and metastasis [
49‐
51]. Although further investigation is required to confirm this finding in different BRCA1-mutant cell lines, our data suggests that FAK may have distinct activation pathways in non-malignant
BRCA1-mutant epithelial cells and in wild type malignant progressive cells [
51]. Nevertheless, the evidence for the assignation of the roles of inflammatory lipids, including PAF through paracrine regulation in cancer initiation, particularly in BRCA-mutant epithelial cells is likely as one of key mechanisms, but require further investigation. It may be restricted to specific situations, depending on mutation sites, tissue type and microenvironment. Here we provided a unique BRCA1-mutant at-risk
in vitro model and demonstrated an evidence that coordinated activation of PAF-PAFR, and FAK and STAT pathways, with the abnormal BRCA1 functions may contribute to the significant increases of malignant potential and early events of tumor transformation in at-risk ovarian epithelium (Fig
6E). This
in vitro at-risk ovarian epithelium may represent a valuable model to understand the early molecular events of malignant transformation and development. Certainly, it requires further molecular approaches including networking pathway and target knock-down
in vitro and
in vivo models to confirm that the signal axis of PAF/PAFR-FAK-STAT pathway is significantly mediated in the early events of ovarian epithelial malignant development.
Author details
1 Department of Obstetrics and Gynecology, Brigham and Women's Hospital, Boston, MA, USA
2 Channing Laboratory, Brigham and Women's Hospital, Boston, MA, USA
3 Obstetrics and Gynecology Department, Peking University People's Hospital, Beijing, China
4 Department of Gynecologic Oncology, University of Texas M. D. Anderson Cancer Center
Houston, TX, USA
5 Department Pathology, Brigham and Women's Hospital, Boston, MA, USA
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
LZ, DW, JW, DE, LMB, JR, LJ, VS carried out the experiments, protocol design, data analysis and interpretation of data and drafting the manuscript. WQ, AV, CC, DWC, JW, SCM, and BY are involved in experiment design, acquisition, interpretation and manuscript preparation. All authors read and approved the final manuscript.