Novel high-grade serous epithelial ovarian cancer cell lines that reflect the molecular diversity of both the sporadic and hereditary disease

Abstract

Few cell line models of epithelial ovarian cancer (EOC) have been developed for the high-grade serous (HGS) subtype, which is the most common and lethal form of gynaecological cancer. Here we describe the establishment of six new EOC cell lines spontaneously derived from HGS tumors (TOV2978G, TOV3041G and TOV3291G) or ascites (OV866(2), OV4453 and OV4485). Exome sequencing revealed somatic TP53 mutations in five of the cell lines. One cell line has a novel BRCA1 splice-site mutation, and another, a recurrent BRCA2 nonsense mutation, both of germline origin. The novel BRCA1 mutation induced abnormal splicing, mRNA instability, resulting in the absence of BRCA1 protein. None of the cell lines harbor mutations in KRAS or BRAF, which are characteristic of other EOC subtypes. SNP arrays showed that all of the cell lines exhibited structural chromosomal abnormalities, copy number alterations and regions of loss of heterozygosity, consistent with those described for HGS. Four cell lines were able to produce 3D-spheroids, two exhibited anchorage-independent growth, and three (including the BRCA1 and BRCA2 mutated cell lines) formed tumors in SCID mice. These novel HGS EOC cell lines and their detailed characterization provide new research tools for investigating the most common and lethal form of EOC.

Keywords: BRCA mutations; cell lines; epithelial ovarian cancer; high-grade serous.

Figure 1. Morphology of cell lines derived from patients 866, 2978, 3041, 3291, 4453 and 4485

Shown are light microscopy images of cell lines at 80% confluence at passages between 66 and 73. Cell lines had developed into predominantly small epithelial type cells, often aggregated, and fibroblast shaped cells were notably absent.

Figure 2. Characterization of the new BRCA1 intronic mutation identified in the OV4485 cell line

A) Primer design for amplification of BRCA1 cDNA surrounding the intronic mutation (black arrows, exon 13-15; white arrows, exon 14-16), and agarose gel of obtained PCR products. PCR amplification of β–actin was used as a control. B) Schematic illustration of alternative BRCA1 mRNA splicing, alternative 3′acceptor of exon 15, and predicted translated BRCA proteins obtained resultant from cDNA sequencing of the OV4485 PCR products. C) Detection of BRCA1 protein in cell lysates of the different cell lines by Western blot. β–actin was used as loading control.

Figure 3. Genomic landscape of the HGS cell lines

Total copy number of ASCAT derived segments was plotted for the genome on a scale of 0 to 15 copies using IGV. Segments with copy number above ploidy (see Table 2 for values) are indicated in red (CNV^gain), and those below ploidy are blue (CNV^loss). Regions of the genome with LOH are indicated in green and shown immediately below each CNA graph depicted for each cell line. Arrows indicate genomic locations containing amplification of KRAS and CCNE1 loci in the genomic profile of the OV866(2) cell line.

Figure 4. Immunostaining of cell lines and corresponding solid tumors

A) Immunohistochemical analysis of formalin-fixed paraffin-embedded solid tumors from patients 866, 2978, 3041, 3291, and 4485 with cytokeratin (CK7, CK8, CK18 and CK19) markers, WT1, PAX8, p53 and HER2. Brown color indicates positive staining, nuclei are counterstained with hematoxylin, and images are at 400x magnification. Topmost images are hematoxylin-eosin images of the corresponding tumor samples. B) Detection of CKs, WT1, PAX8, p53, HER2, E-cadherin and vimentin by Western blot of lysates of cell lines. Lower panel shows an agarose gel of TP53 PCR products from reversed-transcribed cell line mRNAs. β–actin was used as control in both procedures. Blk = Blank, RT = Reverse transcription

Figure 5. Analysis of p53 function in the different cell lines by RT Q-PCR

Gene expression of p53 target genes (p21, NOXA and MDM2) were quantified by RT Q-PCR (see Methods for details) at times 0h, 2h and 5h after 8Gy gamma-irradiation (white circle). Non-irradiated cells were used as controls (black square). Note the increased expression of p53 target genes after DNA damage induced by γ-irradiation in the control p53 proficient cell line TOV21G, and the absence of response for all the HGS EOC cell lines.

Figure 6. In vitro characterization of cell lines using diverse oncogenic assays

A) Spheroid formation of cell lines after 5-7 days in OSE culture media using the inverted droplet technique. Photos are representative of observations from three independent experiments. B) Anchorage-independent growth in soft agar. Pictures show representative images of colonies formed in the agar after two-weeks of culture from three independent experiments performed in triplicate. C) Migration evaluation by the wound-healing scratch assay. Photos show representative images of monolayer cell cultures at 0h and 24h after the scratch was performed. Three independent experiments were performed in triplicate. In all experiments, pictures of different cell lines were taken at the same magnification.

Figure 7. In vivo SCID mouse xenograft tumor formation

A) Tumor volume of OV4485 (white square), OV4453 (black lozenge) and OV866(2) (white triangle) cell lines injected subcutaneously in SCID mice (n=5). Points represent average ± SEM and curves were plotted up until the first animal was sacrificed when end-point limits were attained. * denotes statistical differences (p<0.05, Student's t-test). B and C) Kaplan-Meyer survival curves of mice injected subcutaneously (n=5) or intraperitoneally (n=5) in SCID mice with OV4485 (solid line), OV4453 (dashed line) and OV866(2) (assured line) cell lines. Survival was followed over a period of approximately 10 months and each event was considered as the time of sacrifice at end-point limits.