Objective The value of cell lines for pre-clinical work lies in choosing those with similar characteristics. Selection of cell lines is typically based on patient history, histological subtype at diagnosis, mutation patterns, or signaling pathways. Although recent studies established consensus regarding molecular characteristics of ovarian cancer cell lines, data on in vivo tumorigenicity remains only sporadically available, impeding translation of in vitro work to xenograft models. Methods We introduced 18 ovarian cancer cell lines into athymic nude mice through subcutaneous, intraperitoneal, and ovary intrabursal routes, and observed tumor development over 6 weeks. We also profiled cell line gene expression and identified differentially expressed gene sets based on their ability to form tumors in the subcutaneous or intraperitoneal locations. Representative cell lines were further subjected to proteomic analyses. Results Ovarian cancer cell lines showed variable ability to grow in mice when implanted subcutaneous, intraperitoneal, or intrabursal. While some cell lines grew well in both SC and IP locations, others showed a strong propensity to grow in one location only. Gene expression profiles suggested that cell lines showing preference for IP growth had gene expression patterns more similar to primary tumors. Conclusions We report the tumorigenicity of 17 human ovarian cancer cell lines and one mouse cell line in three distinct anatomical locations, and associated gene networks. Growth patterns and histopathology, linked to molecular characteristics, provide a valuable resource to the research community, and better guide the choice of cell lines for in vitro studies to translate efficiently into xenograft testing.

BackgroundTopotecan (TPT) is a therapeutic option for women with platinum-resistant or -refractory ovarian cancer. However, the dose-limiting toxicity of TPT is myelosuppression. This led us to seek a combination treatment to augment TPT anti-cancer activity in a cancer-targeted manner. Ovarian serous cancers, a major subtype, show dysregulated DNA repair pathway and often display a high level of CHEK1 (CHK1), a cell cycle regulator and DNA damage sensor. CHEK1 inhibitors are a novel approach to treatment, and have been used as single agents or in combination chemotherapy in many cancers.MethodsWe evaluated the cellular effects of TPT in a panel of high grade serous (HGS) and non-HGS ovarian cancer cells. We then determined IC50s of TPT in the absence and presence of CHEK1 inhibitor, PF477736. Synergism between TPT and PF477736 was calculated based on cellular viability assays. Cytotoxic effect of the combined treatment was compared with apoptotic activities by Caspase3/7 activity assay and Western blotting of cleaved-PARP1 and γH2AX.ResultsNon-HGS ovarian cancer cells were generally more sensitive to TPT treatment compared to HGS ovarian cancer cells. When combined with CHEK1 inhibitor, TPT potently and synergistically inhibited the proliferation of HGS ovarian cancer cells. This dramatic synergism in cellular toxicity was consistent with increases in markers of apoptosis.ConclusionsOur findings suggest that the addition of CHEK1 inhibitor increases the response of ovarian cancer cells to TPT. Furthermore, reduced dosages of both drugs achieved maximal cytotoxic effects by combining TPT with CHEK1 inhibitor. This strategy would potentially minimize side effects of the drugs for extended clinical benefit.Electronic supplementary materialThe online version of this article (doi:10.1186/s12885-015-1231-z) contains supplementary material, which is available to authorized users.

Ovarian cancer (OC) is extremely heterogeneous, implying that therapeutic strategies should be specifically designed based on molecular characteristics of an individual's tumor. Previously, we showed that IKKε promotes invasion and metastasis in a subset of OCs. Here, we identified CHEK1 as an IKKε-dependent lethal gene from shRNA kinome library screen. In subsequent pharmacological intervention studies, the co-inhibition of IKKε and CHEK1 was more effective in killing OC cells than single treatment. At the molecular level, co-inhibition dramatically decreased pro-survival proteins, but increased proteins involved in DNA damage and apoptosis. IKKε-knockdown increased p21 levels, while overexpression of wild-type IKKε, but not a kinase dead IKKε mutant decreased p21 levels. We further demonstrated that the depletion of p21 rendered OC cells more resistant to cell death induced by co-inhibition of IKKε and CHEK1. In conclusion, we revealed a novel interplay between IKKε, CHEK1 and p21 signaling in survival of OC. Our study provides a rationale for the clinical development of specific IKKε inhibitor and for usage of IKKε as an exploratory marker for resistance to CHEK1 inhibitors in the clinic. The interplay provides one potential explanation as to why very few clinical responses were achieved in patients treated with single-agent CHEK1 inhibitors.