Background
Epithelial ovarian carcinoma is the most lethal gynecologic malignancy. Ovarian clear cell carcinoma (OCCC) is the second most common histologic subtype, accounting for 5–25% of all ovarian cancer depending on geographic location [
1,
2]. It is well acknowledged that OCCC is more commonly seen in Asia women [
3,
4]. OCCC represents a great challenge due to its disease aggressiveness and chemotherapy resistance. The grave survival and lack of effective treatment prompt us to investigate the underlying molecular mechanism of OCCC.
The two most widely reported gene mutations in OCCC are AT
-Rich Interactive Domain-containing protein 1A (ARID1A) and Phosphoinositide-3-kinase Catalytic Alpha (PIK3CA), representing around 50% [
5,
6] and 40% [
7,
8] of all cases, respectively. Further studies supported that loss of ARID1A expression frequently coexisted with PIK3CA mutations [
9]. What’s more, coexistent ARID1A-PIK3CA mutations might promote ovarian clear cell tumorigenesis through synergic effects [
10,
11]. The breakthrough finding of ARID1A mutation has renewed the interest in elucidating the molecular pathways of OCCC [
2], which still remains less well-understood than that of high-grade serous carcinoma.
Cancer cell lines are affordable models for genetic and molecular profiles, reflecting the characteristics of the origin tumor. According to documentations till now, 16 OCCC cell lines were established and their characteristics have been reported [
12‐
25], added ES-2 and TOV-21G, two commercial available cell lines, which characteristics and genetic features were obtained from ATCC database (
http://www.atcc.org) and COSMIC database (
https://cancer.sanger.ac.uk/cell_lines), respectively. The majority of the established cell lines were derived from Japanese patients. Besides, only one OCCC cell line TOV-21G was found with PIK3CA and ARID1A mutation. We describe here the establishment and characterization of a novel cell line (FDOV1) from a Chinese patient that harbors coexistent ARID1A-PIK3CA mutations.
Discussion
Ovarian clear cell carcinoma, as a subtype with distinct clinical behavior, is notorious for its poor survival and resistance to conventional platinum-based chemotherapy. An international phase III clinical trial was conducted to investigate the addition of irinotecan as adjuvant chemotherapy for patients with advanced-stage [
28]. Unfortunately, no significant survival benefit was found. Given the limited benefit from cytotoxic drugs, there is now great interest in the development of molecular targeted therapy, which requires investigation of a number of areas in the laboratory [
3]. Better understanding of the underlying molecular mechanism is of importance.
According to documentations, there were 16 ovarian clear cell carcinoma cell lines reported so far, in addition, TOV-21G, ES-2, which are commercially available from American Tissue Culture Center, and FDOV1. Their characteristics were compared in Table
3. Firstly, most of the OCCC cell lines were from Japan, while FDOV1 is the only one derived from a Chinese mainland woman. It has been demonstrated that the same kind of carcinoma from different race even country tends to have distinct characteristics, especially genetic characteristics [
29‐
31]. We postulated that genetic background might be partly the reason for the marked geographic difference in OCCC. Therefore, the establishment and characterization of a novel cell line derived from Chinese patients is of significance. Secondly, all cell lines were derived from ovary (10/19), ascites (6/19), and metastatic or recurrent site (3/19), which implied good materials for primarily culture. Lastly, as previously mentioned, ARID1A and PIK3CA mutations are the two most commonly reported. Interestingly, it is found that ARID1A–PIK3CA mutational co-occurrence (33%) is very high in ovarian CCC [
11]. Research work by Chandler et al. supported that a genetic epistasis model wherein ARID1A and PIK3CA mutations cooperate and ovarian cancer can arise only when these genes are co-mutated in the mouse ovarian surface epithelium [
11]. Coexistent ARID1A–PIK3CA mutations might be a major driver of OCCC in humans. Among 19 OCCC cell lines, only FDOV1 and TOV-21G carried coexistent ARID1A–PIK3CA mutations. The character of concurrent ARID1A–PIK3CA mutations of FDOV1 might help to better understand the underlying mechanism and even to explore the activity of dual inhibitors in vitro. However, the cell lines established before 2000 lack genetic information due to testing techniques. Besides, some new molecular features found by WES in FDOV1 requires more attentions.
Table 3
Cell lines of ovary clear cell carcinoma
1 | HUOCA-II (1987) | Japan | 51 | Ovary | 24,28 | 46 | Yes | |
2 | RMG-I (1988) | Japan | 34 | Ascites | 60 | 47 | Yes | |
3 | OCC1 (1990) | Hong Kong | 47 | Ascites | 36,38 | 70–77 | Yes | |
4 | RMG-II (1991) | Japan | 53 | Ascites | 58 | hypertetraploid | No | |
5 | TOV-21G(1991)a | French-Canadian descent | 62 | Ovary | 36 | 47 | Yes | ARID1A p.Q758fsa75, ARID1A p.Y551fsa72, KRAS p.G13C, PIK3CA p.H1047Y |
6 | ES-2(1991)a | Black | 47 | Ovary | 24 | 66- 88 | Yes | BRAF p.V600E, MAP2K1 p.D67N |
7 | OVISE (1995) | Japan | 40 | Metastatic tumor | 60–70 | 59–65 | Yes | |
8 | OVTOKO (1995) | Japan | 78 | Metastatic tumor | 60–70 | 76–83 | Yes | |
9 | JHOC-5 (1999) | Japan | 47 | Ovary | 52 | 74–85 | No | |
10 | JHOC-6 (1999) | Japan | 43 | Recurrent tumor | 70 | 46–49 | Yes | |
11 | SMOV-2 (1999) | Japan | 46 | Ovary | 48.2 | 85–92 | Yes | P53 mutation (−) |
12 | TAYA (2002) | Japan | 43 | Ascites | 50 | 69–74 | No | P53 mutation (codon 132 in exon 5), PTEN mutation(−) |
13 | RMG-V (2005) | Japan | 52 | Ascites | 15.5 | 77–85 | No | |
14 | TU-OC-1 (2013) | Japan | 65 | Ovary | 38.4 | 64–90 | Yes | PIK3CA E542K mutation |
15 | TU-OC-2(2016) | Japan | 68 | Ovary | 37.5 | 41–96 | No | PIK3CA mutation (−), TP53 mutation (−) ARID1A loss |
16 | HCH-1 (2016) | Japan | 67 | Ovary | 48.7, 66.4 | 39–44 | Yes | PIK3CA mutation (−), P53 mutation(−), PTEN mutation(−), KRAS mutation(−) |
17 | NOCC(2016) | Japan | 48 | Ascites | 29 | 60–83 | Yes | |
18 | HCH-3 (2017) | Japan | 41 | Ovary | 82 | 78–87 | Yes | KRAS mutation (+), TP53 mutation(+) |
19 | FDOV1(2017) | China mainland | 67 | Ovary | 37.4 | 45–90 | Yes | PIK3CA p.H1047R, ARID1A p.L2106 fs, ARID1A p.N201 fs, SPOP p.D82H ARID1B p.E1681X |
Notably, SPOP, a relatively new gene regulating DNA damage repair, is the most frequently mutated gene in prostate carcinoma [
32,
33]. Mutations in SPOP lead to genomics instability and are identified as driver events resulting in tumorigenesis of prostate carcinoma via coordination with PI3K/mTOR and AR (Androgen Receptor) pathway in mouse [
34,
35]. Studies on SPOP mutation in other solid carcinomas is reported in follicular and papillary thyroid cancers [
36,
37], endometrial clear cell carcinoma [
38,
39], gynecological carcinosarcoma [
40], lung cancer [
41], and colorectal cancer [
42]. What is worthy of mentioning is that SPOP mutation in gynecological malignant tumor is not uncommon (18% in endometrial clear cell carcinoma, 14% in carcinosarcoma) [
38,
40]. In OCCC, the mutation rate of SPOP is 11.1% (1/9) according to COSMIC (Catalogue Of Somatic Mutations In Cancer) database. However, few relevant functional studies are documented in OCCC. Thus, FDOV1, a novel cell line harbored SPOP mutation, is an ideal tool for functional study. In addition, it is suggested that SPOP mutation might represent sensitivity to DNA damaging agents such as PARP inhibitor in prostate carcinoma [
35]. Consequently, whether it is indicated a similar effect on OCCC still needs further investigation.
Chromosome 20q13 ZNF217 (Zinc Finger Protein 217) locus amplification, which has found in FDOV1, is one of common molecuclar genetic aberrations in OCCC [
43]. The frequency is reported from 20 to 36% [
10,
43,
44]. It is reported that ZNF217 amplification often coexisted with loss of ARID1A expression and PIK3CA mutation [
10]. Importantly, ZNF217 amplification has a correlation with shorter worse survival outcome in OCCC, and multivariate analysis suggest it is an independent prognostic factor of PFS (Progression-Free Survival) and OS (Overall survival) [
43]. Still, ZNF217 amplification has an adverse prognostic implication in breast and gastric cancer [
45‐
47]. However, the underlying mechanism is still tensely studied. FDOV1 will be an good cell line for mechanism research in vitro.
Methods
Medical history
A 67-year-old woman was admitted into Fudan University Shanghai Cancer Center with the chief complaint of palpable abdominal mass in May 2016. Ultrasound revealed a large pelvic mass (14 × 10 cm) with mixed components and a hyperechoic mural nodule (4 × 3 cm). Serum tumor markers including carbohydrate antigen (CA) 125 and CA199 were not elevated. Intraoperative findings were only significant for a cystic-and-solid mass (15 cm in diameter) in the left adnexa. Frozen pathology was highly suspicious of ovarian clear cell carcinoma. The patient underwent comprehensive staging surgery and the final stage was FIGO (International Federation of Gynecology and Obstetrics) Ia. She gave written informed consent before surgery according to institutional guideline (FUSCC 050432–4-1212B). Six cycles of TC regimen (paclitaxel + carboplatin) chemotherapy were administered. She was followed up every 3 months after treatment. The patient was admitted into outside hospital with chief complain of intestinal obstruction in May 2017. She developed multiple sites of venous thromboembolism and cerebral infarction. Unfortunately, she died 1 month later with no chance of anti-cancer treatment.
Establishment of cell line
The fresh tissue (1 cm3) from primary ovarian tumor was obtained during comprehensive staging surgery, and suspended in phosphate buffer solution (PBS) to get rid of necrotic and connective tissue. The material was then finely minced into 1-mm3 tissue blocks and placed in 10-cm cell culture dishes (Corning, NY, USA) containing Medium 199 medium (HyClone, Thermo Scientific, USA) with 10% fetal bovine serum (Gibco, Life technologies, USA), 100 IU/ml penicillin (HyClone, Thermo Scientific, USA) and 50 μg/ml streptomycin (HyClone, Thermo Scientific, USA). The cells were then incubated in a humidified atmosphere containing 5% CO2 at 37 °C. After repeated passage, a new cell line of FDOV1 was successfully established with more than 80 generations. FDOV1 cells have been sent to China General Microbiological Culture Collection Center for preservation (No. 13812) in March 2017.
Growth characteristics
Cell proliferation and growth curve was determined by Cell Counting Kit-8 (CCK-8) test. FDOV1 cells (2 × 103 cells/100 μl) were seeded in 96-well plates in sextuplicate. Ten days later, 100 μl mixture (CCK-8/medium 1:9) were added into the wells for incubation by 2 hours. We then measured the Optical Density (OD) at 450 nm (OD450) by a microplate reader (Synergy H4, Bio-Tek).
Cell cycle analysis by flow cytometry
Resuspended FDOV1 cells (1 × 106 cell) were kept overnight in 3 ml 75% ethanol at − 20 °C. After treatment, the cells were collected and stained with propidium iodide solution for 20 min at 4 °C in darkness. We then assessed the cell cycle by flow cytometer (BD Biosciences, San Jose, CA, USA).
Chromosome analysis
The cells in exponential phase were treated with 0.25 μg/ml colchicine for 6 hours and kept overnight at 37 °C. We then treated the cells with 0.1% trypsin solution for 15 s at room temperature and stained them with 3% Giemsa for G-band karyotyping. After examing 45 metaphases, the histograms of the chromosomal distribution were determined.
Tumor markers
A total of 2 × 106 cells were cultured for 4 days in total. The medium was changed and collected every 2 days, then were sent for tumor marker detection by chemiluminescence immunoassay. Detected markers included CA 125, CA 199, CA 153, CA 724, AFP, CEA and HE4.
Heterotransplantation
All the procedures were approved by the Department of Laboratory Animal Science in Fudan University. For in vivo studies, 4–6 weeks old NOD/SCID mice (Shanghai SLAC Laboratory Animal, Shanghai, China) were maintained under sterile conditions. FDOV1 cells (1 × 107 cells, passage 18) were injected subcutaneously into the dorsal flanks of two mice. Visible tumors were noted in both mice 20 days after transplantation.
Histological analysis and immunohistochemistry
Transplant tumors were fixed in 10% buffered formalin for 24 h and embedded in paraffin. Hematoxylin and eosin (HE) were used to stain 4-μm sections. The markers used for immunohistochemistry were listed as follows: HNF-1β (Sigma, CA, USA), Napsin A (Abcam, Cambridge, UK), Pax-8 (Cellmarque, CA, USA), ER (Roche, Basel, Swiss) and PR (Roche, Basel, Swiss). Microscopic slides were reviewed by a senior gynecology-dedicated pathologist (Prof. Yang).
Whole-exome sequencing (WES) of FDOV1 cell and patient’s tumor tissue
The WES was performed by 3D Medicines Corporation and the specific method was the same as a previous publication [
48].
Sanger sequencing
Mutations of the specific genes in FDOV1 cells were validated by Sanger Sequencing. The RNA was extracted by Trizol Methods, then reverse transcription polymerase chain reaction (RT-PCR) was performed as our previous study [
49]. The primers were showed in Additional file
1: Table S1. PCR amplification was performed as follows: denaturation at 98 °C for 5 min, 30 cycle of 98 °C for 10s, 58 °C for 30s, and 72 °C for 1 min, then holding at 72 °C for 1 min, finally incubation at 4 °C. Sequencing was performed at Boshang Biotechnology Co.Ltd. (Shanghai, China).