Background
Ovarian cancer is the leading cause of death among gynecologic malignancies worldwide, and the survival rates remain disappointing for patients suffering from advanced cancers [
1,
2]. Development of cancer metastasis is the major cause that kills patients with ovarian cancer. However, the molecular mechanisms contributing to its aggressiveness are still not fully understood.
In the past decade, it has been established that the matrix metalloproteinases (MMPs), including Membrane-type-1 MMP (MT1-MMP or MMP14) and MMP-2, play a critical role in degrading the basement membrane and the extracellular matrix (ECM), resulting in tumor cell dissemination and outgrowth of secondary cancers [
3,
4]. MT1-MMP not only directly cleaves ECM components, but also functions as the main activator of MMP-2 [
5,
6]. Conversely, tissue inhibitor of metalloproteinase-2 (TIMP-2) inhibits MMP-2 after binding with its hemopexin domain [
7,
8]. As in other solid cancers, MT1-MMP has been reported to be widely expressed in ovarian cancers and related malignant ascites of all histological types, but not in normal ovarian epithelium or benign tumors [
9‐
12]. Despite the central role of MT1-MMP in these cancer metastases, little is known about its transcriptional regulators [
13].
FBI-1 (also known as POKEMON, LRF in mouse, or OCZF in rat), which was originally identified as a factor that binds to the inducer of short transcripts (IST) element of human HIV-1 genome [
14], is a member of the POK (POZ and krǔppel) family of transcription factor. Recent reports revealed the participation of FBI-1 in NF-κB activation [
15], adipogenesis [
16], lymphocyte differentiation [
17], and oncogenesis [
18‐
21]. Emerging studies have shown that FBI-1 is strongly expressed in diffuse large B-cell lymphoma, follicular lymphoma, breast, lung, colon, prostate and bladder carcinomas [
18,
22]. However, its role in ovarian tumors has not been reported.
In the current study, we have illustrated a novel mechanism that may account for the aggressiveness and poor prognosis of human ovarian cancer. We showed that FBI-1 interacted and activated MT1-MMP, increased cell motility and invasion of ovarian cancer. FBI-1 is significantly up-regulated in advanced stages of ovarian cancer and associated with overall and disease free survival of patients with ovarian tumors.
Discussion
Since ovarian cancer is often asymptomatic until it has extensive extra-ovarian spread, the high lethality of this disease is at least partly attributed to the fact that patients are often at an advanced stage at the time of first diagnosis [
25]. Therefore, a more precise understanding of the genetic mechanisms that control cancer cell dissemination is important. In this report, we established for the first time that vast majority of ovarian cancers produced aberrantly high levels of FBI-1 when compared with benign or borderline tumors. Overexpression of FBI-1 was also observed in most of the ovarian cancer cell lines. Clinicopathological analysis indicated that high expression of FBI-1 was associated with advanced stage, shorter overall and disease-free survival. The cancer cells at metastatic foci and malignant ascites stemmed from advanced diseases expressed significantly higher levels of FBI-1 when compared with the primary cancers, supporting a role of FBI-1 in the dissemination of ovarian cancer. Using cell line models, we further provided evidence supporting that FBI-1 plays a significant role in migration, invasion, and proliferation of ovarian cancer.
FBI-1 is encoded by the
ZBTB7A gene at chromosome 19p13.3. To date, little is known about the mechanism leading to
FBI-1 up-regulation in cancer. In non -Hodgkin B-cell lymphoma, t(14;19)(q32;p13.3) translocation was reported and aberrant activation of upstream regulators was speculated to be involved in its up-regulation[
26]. Recently,
FBI-1 gene amplification was found to be a frequent event in non-small cell lung cancer [
22]. Indeed, a remarkable proportion of our ovarian cancers (86%) displayed more than two-fold amplification of
ZBTBZA gene and the gene amplification status was significantly correlated with its elevated expression at transcriptional level.
Previous reports suggested that the key role of FBI-1 in oncogenesis is to act as a potent inhibitory regulator of the tumor suppressor gene
p14
ARF
, subsequently leading to MDM2 constitutive activation and p53 degradation [
18,
27]. Taking advantage of ovarian cancer cell lines with different
p53 status, we proved that FBI-1 could enhance cancer cell invasion and migration in a p53-independent manner. This hypothesis was confirmed by wild-type p53 manipulation study. It is likely that FBI-1 has more target genes than originally anticipated. Indeed, studies have suggested that FBI-1 could repress retinoblastoma gene, another well-known tumor suppressor gene [
20] and could activate fatty-acid synthase gene (FASN) [
28].
Unlike other epithelial carcinomas, which mainly metastasize through vasculature, "shedding" and "seeding" of ovarian cancer cells from the primary tumor into peritoneal cavity are also common [
29]. These tumor dissemination processes involve detachment of tumor cells and degradation of the basement membrane. Thus, proteolytic activity becomes extremely important in ovarian cancer metastasis because it facilitates the breakthrough and invasion of the mesothelial monolayer and collagen-rich extracellular matrix as well. MT1-MMP/MMP-2/TIMP-2 axis is considered as a central determinant in these processes. Even though the posttranscriptional regulations of MT1-MMP have been extensively studied [
30,
31], knowledge about its transcriptional regulations remain limited. The GC-rich promoter of MT1-MMP is different from the other MMPs which are composed of a conserved TATA box at promoter region [
24]. Recently, transcription factors Sp1, Sp3 and Egr-1 have been identified as potent regulators of MT1-MMP expression in prostate cancer, endothelial and glomerular mesangial cells [
24,
32,
33]. Our present findings support that FBI-1 can directly bind to the promoter of
MT1-MMP, but not
MMP-2, and induce MT1-MMP expression in both transcription and protein levels, in a dose dependent manner leading to cell migration and invasion. Induction of MT1-MMP transcription by FBI-1 by mutating the potential binding sites in the MT1-MMP promoter will be performed in our future study. On the other hand, the findings of MTT assays on the two ovarian cancer cell lines harboring different
p53 status suggested that the effect of FBI-1 on cell proliferation might be restricted in cells with intact p53 function, consistent with findings from a previous report [
34].
In conclusion, our data showed that FBI-1 promotes aggressive cancer cell phenotype and its overexpression is associated with poor clinical outcome. The role of FBI-1 in ovarian cancer development and progression also makes it a potential target for therapeutic intervention in ovarian cancer treatment.
Methods
Patient samples and cell lines
220 archival clinical samples (Table
1), including paraffin-embedded tissues of 10 ovarian benign cystadenomas (ages, 20~77 years; mean age, 32 years), 19 ovarian borderline cystadenomas (ages, 20~46 years; mean age, 30 years), 111 ovarian cancers (ages, 23~78 years; average age, 50 years) and corresponding metastatic foci in 63 cases, as well as 17 malignant ascites obtained from stage III/IV ovarian cancer patients, were collected at the time of surgical resection from 1987 to 2005 at Department of Pathology, Queen Mary Hospital, the University of Hong Kong. In addition, 36 paired samples of ovarian cancers and their corresponding normal fallopian tubes and/or contralateral ovaries were processed for snap-frozen blocks followed by storage at -80°C and total RNA and genomic DNA extractions. Both collection and the use of such patient samples were approved under the Institutional Ethics Review Board. All diagnoses have been confirmed by two pathologists using criteria of the International Federation of Gynecology Oncology (FIGO).
Culture conditions of two immortalized normal human ovarian surface epithelium cell lines (HOSE6-3 and HOSE11-12) and 13 ovarian cancer cell lines (OVCA420, OVCAR-3, SKOV-3, 2780 S, 2780CP, 2008, 2008/C13, SW626, ES2, OC316, DOV13, TOV21G and TOV112D) were described previously [
35‐
38]. Cell lines used in the present study were in culture for less than 6 months.
Immunohistochemistry and western blot
Immunohistochemical staining for FBI-1 (1:400; Abcam Inc, Cambridge, UK; ab36606) was carried out on 220 paraffin-embedded clinical samples with EnVision_Dual Link System (K4061; Dako, North America, Carpinteria, CA, USA) [
36]. For negative control, primary antibody was replaced with phosphate buffer saline (PBS). Sections were assessed for both staining intensity and percentage, and the final score was determined as described [
35].
For western blot, antibodies specific to FBI-1 (1:600; Abcam), MT1-MMP (1:1000; Sigma; MS221106), p53 (1:1000; Dako; clone DO-7), and GFP (1:1500; Santa Cruz; sc-9996) were used to detect the immunoreactivity according to standard procedures [
36]. Our plasmid expressing FBI-1 was tagged with GFP. GFP is a protein composed of 238 amino acids (around 31kDa). Endogenous FBI-1 (75kDa) and GFP-tagged FBI-1 (106kDa) could therefore be differentially detected in 7.5% SDS-PAGE gels using anti-FBI-1 antibody.
RNA and DNA preparation, quantitative real-time PCR (qPCR)
According to the manufacturer's instruction, total RNA and genomic DNA were isolated from frozen tissues and cell lines by using TRIzol reagent (Invitrogen, San Diego, CA, USA) or phenol/chloroform (Invitrogen) respectively. First-strand cDNA was synthesized with oligo-dT primer and SuperScript III Reverse Transcriptase kit (Invitrogen). ABI Prism 7700 Sequence Detection System and SYBR-green PCR master mix (Applied Biosystems) were used for qPCR as described [
35].
The primers used for mRNA level evaluation are as follows: FBI-1, forward, 5'-TCTGCGAGAAGGTCATCC-3', and reverse, 5'-CGTAGTTGTGGGCAAAGG-3'[
22]; MT-MMP, forward, 5'-ACGGAGGTGATCATCATTGAGG-3' and reverse, 5'-AGATGGGGCTGGACAGACACA-3'[
37,
39]; MMP-2, forward, 5'-GGCCCTGTCACTCCTGAGAT-3' and reverse, 5'-GGCATCCAGGTTATGGGGGA-3'[
36,
40]; TIMP-2, forward, 5'-GCGGTCAGTGAGAAGGAAGTGG-3', and reverse, 5'-CTTGCACTCGCAGCCCATCTG-3'[
41]; GAPDH, forward, 5'-TCCATGACAACTTTGGTATCGTG-3' and reverse, 5'-ACAGTCTTCTGGGTGGCAGTG-3'[
35]. Each sample was run in duplicate and normalized with GAPDH.
FBI-1 gene amplification was also assessed by qPCR using the following primers: forward, 5'-GAACGAGGGTTTAGTGCA-3' and reverse, 5'-CGAGCTGTTCTGGAGAGA-3'[
22]. TRAT1, a single copy gene in ovarian cancer [
42], (forward, 5'-CATGTCAGGTAAGTGGCATT-3'; reverse, 5'-GGGTCTTCTCGTTAGGACTTAG-3'), was served as an internal control. Each sample was verified in duplicate. Scoring ≥ two-fold difference in ovarian cancer samples compared with the corresponding normal counterpart indicated gene amplification.
Plasmids
The fell-length cDNA of human FBI-1 was amplified from OVCAR3 cells by PCR using FastStart Taq DNA polymerase (Roche Applied Science, Indianapolis, USA) and the primers were forward, 5'- CCCAAGCTTGGGATGGCCGGCGGCGTGGACGGC-3' and reverse, 5'- GCCTTAAGGCTTAGGCGAGTCCGGCTGTGAA-3'. PCR product was cloned into pEGFP-C3 vector with Hind III and EcoR I. The integrity and accuracy of insert was confirmed by sequencing.
pEGFP-C1-p53 construct expressing wild-type p53 and empty vehicle were generous gifts from Dr. Wilson Ching (Department of Anatomy, The University of Hong Kong, HKSAR, China). Promoter-luciferase reporter genes, pGL3-Basic-MT1-MMP-Luc (bp -500/+112) and pGL3-Basic-shortMT-MMP-luc (bp -180/+112) were kindly provided by Professor Constance E. Brinckerhoff [
24,
43] (Department of Biochemistry and Medicine, Norris Cotton Cancer Center, Dartmouth Medical School, Lebanon, NH 03756, USA). pGL2-Basic-MMP-2-Luc (full length of promoter of human MMP-2) was a generous gift from Dr. Etty N. Benveniste (Department of Cell Biology, University of Alabama at Birmingham, Birmingham, Alabama).
Transient transfection and luciferase reporter assay
For ectopic expression of FBI-1 and p53, or knockdown of MT1-MMP, cells (OCVA 420 and SKOV-3) were seeded at 90% confluence one day before transient transfection with various amounts of pEGFP-FBI-1, pEGFP-p53, siMT1-MMP (Ambion, USA), or controls using Lipofectamine™2000 (Invitrogen) according to manufacturer's instruction. Twenty-four hours after transfection, cells were collected followed by migration and invasion assays and proliferation analysis. To assess the transcriptional activities of MT1-MMP and MMP-2, 1 × 105 cells were grown in 24-well plates and transiently co-transfected with indicated dosage of pEGFP-FBI-1 construct and pGL3-Basic-MT1-MMP-Luc or pGL2-Basic-MMP-2-Luc reporter plasmids. pRL-SV40-Luc was used as internal control, whereas corresponding empty vectors were served as negative controls. After 48 h incubation, cells were lysed for luciferase activity analysis using the Dual-Luciferase Reporter Assay System (Promega, Madison, WI). Renilla luciferase activities were used to normalize the transfection efficiency. Each experiment was repeated twice in duplicate wells.
Establishment of FBI-1 stable knockdown cells
For stable silencing, human FBI-1-specific shRNA (SureSilencing shRNA constructs) and control shRNA vector (pGeneClip™ puromycin vector) were purchased from SuperArray (SABioscicences Corporation, Frederick, USA), transfected into OVCA 420 and SKOV-3 cells, and selected with puromycin as published previously[
36]. Knockdown efficiency was confirmed by qPCR and immunoblotting analysis.
In Vitro migration and invasion assays
To determine the impact of FBI-1 on cell mobility, transient overexpressing or stable knockdown FBI-1 cells and their control cells were counted and equally plated on the upper compartments of the 24-wells, 8-μm pore size Transwell chambers (BD Biosciences, San Jose, CA). Cell migration and invasion assays were performed with self-coated Gelatin and Matrigel, respectively. After 24 h, the migrated and invaded cells on the lower surface of membrane were fixed and stained with methanol mixed crystal violet. All assays were run in independently three times.
Proliferation assay
Effect of FBI-1 with or without wild type p53 on cell proliferation was determined by 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl-tetrazolium bromide (MTT) assay as previously described [
36]. Every experiment was repeated twice in triplicate wells, separately.
Chromatin immunoprecipitation (ChIP)
Confluent OVCA 420 and SKOV-3 cells grown on 10-cm dish were cross-linked by adding 1% formaldehyde for 10 min. After stopping the reaction with 1.25 M glycine solution and washing with PBS, cells were suspended in RIPA buffer (150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM Tris, and 5 mM EDTA) containing protease inhibitor cocktail and sonicated until most of the DNA fragments ranged from 500 to 1000bp. 50 μl of soluble lysate was saved as 'input sample'. The rest of the lysate was pre-cleared with protein A/G beads and salmon sperm DNA. Anti-FBI-1 antibody and 30 μl protein A/G beads were then added to the pre-cleared samples. Samples with no antibody incubation were used as negative control. Beads were centrifuged and washed with RIPA buffer, IP buffer (0.5 M LiCl, 0.1 M Tris, 1%NP-40, and 1% sodium deoxycholate), and then RIPA buffer. The bound material was eluted from the beads at 85°C for 10 min in elution buffer (0.1 M NaHCO
3 and 1%SDS) and the RNA and protein was digested using RNase A and proteinase K, respectively. DNA was precipitated and purified by using phenol/chloroform. The following primers were used to detect MT1-MMP promoter sequences, which is designed using Primer3 (v.0.4.0) software based on the sequence of luciferase reporter construct [
24]: forward, 5'-CCGACAGCGGTCTAGGAAT-3', reverse, 5'- AGACAACGGGAGGGTCTTG-3'.
Statistical analysis
The clinical-pathological features and the data from in vitro assays were analyzed using Statistical Package for Social Science 15.0 for windows (SPSS Inc., Chicago, IL, USA) software. Spearman's rho test, Kruskal-Wallis rank test, Mann-Whitney test, Kaplan-Meier method (using log-rank test), Cox's regression model were used to assess correlation, difference of multiple groups, difference between two groups of non-parametric data, survival probability and multivariate survival analysis, respectively. P value <0.05 was considered to be significant.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
LLJ designed and performed the experiments, interpreted the results and prepared the manuscript under the supervision of ANYC, MKYS and OGWW who also conceived the study and critically revised the manuscript. ANYC, KFT and HYSN provided the clinical material while ANYC also facilitated laboratory settings for the experiments. WFL and XFL provided cell lines and contributed to the manuscript editing. ESYW and HYC provided the technical assistance for experiments. All authors approve the final version of the manuscript.