Development and characterization of five cell models for chemoresistance studies of human ovarian carcinoma

Platinum agents and paclitaxel (taxol) are among the most effective drugs currently available for treatment of ovarian cancer. One of the hurdles with taxol and platinum- based therapy is the clinical development of resistance to these agents. To investigate the mechanism of drug resistance in human ovarian cancer, we developed and characterized 5 cell models for chemoresistance studies of cisplatin, carboplatin and taxol. We report in this study that these human ovarian carcinoma cell model systems include 2 models for cisplatin resistance, 2 models for carboplatin resistance, and 1 model for taxol resistance. The biological and biochemical characteristics of the models showed that (i), the IC50 values of the drugs for all these resistant cell models were 3 times (or more) higher than those for the parental tumor cells. There also exist varying degrees of cross-resistance to several other chemotherapeutic agents in these systems. Moreover, the intracellular drug accumulations in these cells were significantly reduced as compared to those in the parental cells. (ii), The proliferation rates of these resistant cells were markedly decreased. However, there were no obvious changes in cell cycle distribution in these model systems. (iii), Our results for the expression of a few major drug resistance-related genes revealed that the expression of p53, lrp-1 and mrp-1 was decreased, while the expression of pkc, topo I and topo II beta was increased in the resistant tumor cells as compared with the parental cells. In contrast, no significant alterations in gst-pi and topo II alpha expression were found. Interestingly, the levels of mdr-1 expression were elevated in some models, but were reduced in others, thus suggesting that different pathways are involved in the formation of drug resistance in different cell model systems, and that different mechanisms are responsible for the development of different drug resistances in tumor cells. Taken together, our findings indicate that these models may be potentially used to assess the biochemical and genetic mechanisms of drug resistance in human ovarian cancer and to identify new drug resistance-related genes.