Repeated cisplatin treatment can lead to a multiresistant tumor cell population with stem cell features and sensitivity to 3-bromopyruvate.

Cisplatin is used in treatment of several types of cancer, including epithelial ovarian carcinoma (EOC). In order to mimic clinical treatment and to investigate longterm effects of cisplatin in surviving cancer cells, two EOC cell lines were repeatedly treated with low doses. In the SKOV-3 cell line originating from malignant ascites, but not in A2780 cells from a primary tumor, this led to emergence of a stable population (SKOV-3-R) which in the absence of cisplatin showed increased motility, epithelial-mesenchymal transition (EMT) and expression of cancer stem cell markers CD117, CD44 and ALDH1. Accordingly, the cells formed self-renewing spheres in serum-free stem cell medium. Despite upregulation of mitochondrial mass and cytochrome c, and no upregulation of Bcl-2/Bcl-xL, SKOV-3-R were multiresistant to antineoplastic drugs. Cancer stem cells, or tumor-initiating cells (TICs) are highly chemoresistant and are believed to cause relapse into disseminated and resistant EOC. Our second aim was therefore to target resistance in these TIC-like cells. Resistance could be correlated with upregulation of hexokinase-II and VDAC, which are known to form a survival-promoting mitochondrial complex. The cells were thus sensitive to 3-bromopyruvate, which dissociates hexokinase-II from this complex, and were particularly sensitive to combination treatment with cisplatin at doses down to 0.1 x IC 50. 3-bromopyruvate might thus be of use in targeting the especially aggressive TIC populations.

Protein markers of cancer-associated fibroblasts and tumor-initiating cells reveal subpopulations in freshly isolated ovarian cancer ascites.

BACKGROUND: In ovarian cancer, massive intraperitoneal dissemination is due to exfoliated tumor cells in ascites. Tumor-initiating cells (TICs or cancer stem cells) and cells showing epithelial-mesenchymal-transition (EMT) are particularly implicated. Spontaneous spherical cell aggregates are sometimes observed, but although similar to those formed by TICs in vitro, their significance is unclear. METHODS: Cells freshly isolated from malignant ascites were separated into sphere samples (S-type samples, n=9) and monolayer-forming single-cell suspensions (M-type, n=18). Using western blot, these were then compared for expression of protein markers of EMT, TIC, and of cancer-associated fibroblasts (CAFs). RESULTS: S-type cells differed significantly from M-type by expressing high levels of E-cadherin and no or little vimentin, integrin-ß3 or stem cell transcription factor Oct-4A. By contrast, M-type samples were enriched for CD44, Oct-4A and for CAF markers. Independently of M- and S-type, there was a strong correlation between TIC markers Nanog and EpCAM. The CAF marker α-SMA correlated with clinical stage IV. This is the first report on CAF markers in malignant ascites and on SUMOylation of Oct-4A in ovarian cancer. CONCLUSIONS: In addition to demonstrating potentially high levels of TICs in ascites, the results suggest that the S-type population is the less tumorigenic one. Nanog(high)/EpCAM(high) samples represent a TIC subset which may be either M- or S-type, and which is separate from the CD44(high)/Oct-4A(high) subset observed only in M-type samples. This demonstrates a heterogeneity in TIC populations in vivo which has practical implications for TIC isolation based on cell sorting. The biological heterogeneity will need to be addressed in future therapeutical strategies.