No abstract

Studies were undertaken to investigate acquired resistance to cisplatin in human ovarian cancer cells. The cell lines A2780 and A2780/CP70 were studied to assess their respective characteristics of drug accumulation and efflux, cytosolic inactivation of drug, and DNA repair. All experiments were performed using 1-h drug exposures. The A2780/CP70 cell line was 13-fold more resistant to cisplatin than A2780 cells. When studied at their respective IC50 doses, drug accumulation rates were similar for the two cell lines. However, the resistant cell line was twofold more efficient at effluxing drug, which was associated with reduced total drug accumulation for equivalent micromolar drug exposures. At equivalent levels of total cellular drug accumulation, the two cell lines formed the same levels of cisplatin-DNA damage, suggesting that cytosolic inactivation of drug does not contribute to the differential in resistance between these cell lines. Resistant cells were also twofold more efficient at repairing cisplatin-DNA lesions in cellular DNA and in transfected plasmid DNA. We conclude that in these paired cell lines, alterations in drug uptake/efflux and in DNA repair are the major contributing factors to acquired resistance to cisplatin. (J. Clin. Invest. 1991. 87:772-777.)

We have studied several aspects of DNA damage formation and repair in human ovarian cancer cell lines which have become resistant to cisplatin through continued exposure to the anticancer drug. The Cells in culture were treated with cisplatin, and the two main DNA lesions formed, intrastrand adducts and interstrand cross-links, were quantitated before and after repair incubation. This quantitation was done for total genomic lesions and at the level of individual genes. In the overall genome, the initial frequency of both cisplatin lesions assayed was higher in the parental than in the derivative resistant cell lines. Nonetheless, the total genomic repair of each of these lesions was not increased in the resistant cells. These differences in initial lesion frequency between parental and resistant cell lines were not observed at the gene level. Resistant and parental cells had similar initial frequencies of intrastrand adducts and interstrand cross-links in the dihydrofolate reductase (DHFR) gene and in several other genes after cisplatin treatment of the cells. There was no increase in the repair efficiency of intrastrand adducts in the DHFR gene in resistant cell lines compared with the parental partners. However, a marked and consistent repair difference between parental and resistant cells was observed for the gene-specific repair of cisplatin interstrand cross-links. DNA interstrand cross-links were removed from three genes, the DHFR, multidrug resistance (MDR1), and b-globin genes, much more efficiently in the resistant cell lines than in the parental cell lines. Our findings suggest that acquired cellular resistance to cisplatin may be associated with increased gene-specific DNA repair efficiency of a specific lesion, the interstrand cross-link.The metal coordination complex cisplatin has been widely used in the treatment of ovarian, testicular, head and neck, bladder, and small-cell lung carcinomas (19). DNA is thought to be the most critical target for the antitumor action of cisplatin, although the compound can also react with other biologically important molecules such as RNA and proteins (25, 29). The major lesions produced by cisplatin are DNA intrastrand adducts (IA) between adjacent purines GpG or ApG. Minor lesions include IA between two guanine residues separated by one nucleotide residue (GpXpG), interstrand cross-links (ICL) between two adjacent guanine residues on opposite DNA strands, and DNA-protein crosslinks (13,27). The lesions of most therapeutic significance are thought to be IA and ICL.Platinum-based chemotherapy is the standard treatment in the initial therapy of advanced-stage ovarian cancer. Although up to 80% of patients will have an early favorable response to therapy, less than 20% of individuals will experience long-term disease-free survival (33). As a result of primary or acquired resistance, most ovarian cancer patients will eventually relapse with cisplatin-resistant disease (5). Attempts to overcome drug resistance are central to both clinical and basic molecular research ...

We assessed the possible role of the human repair genes, ERCC1 and ERCC3, in resistance to cisplatin-induced cytotoxicity. The UV repair-deficient Chinese hamster ovary (CHO) 43:3B [designated ERCC1(-)] cell line and its paired subline 83-J5, which is stably transfected with the human DNA excision repair gene ERCC1 [designated ERCC1(+)], were used in this study. UV repair-deficient CHO 27-1 cells [designated ERCC3(-)] and its paired subline designated 'ERCC3(+)', which is stably transfected with the human DNA excision repair gene ERCC3, were also used. In each pair of cell lines, we assessed cisplatin cytotoxicity, cellular drug accumulation and platinum-DNA adduct repair after 1 h drug exposures. Drug accumulation and DNA repair were assessed by atomic absorption spectrometry with Zeeman background correction. ERCC1(+) cells (IC50 = 4.0 microM) were 5-fold more resistant to cisplatin than ERCC1(-) cells (IC50 = 0.75 microM). ERCC1(+) cells repaired 25% of DNA lesions in cellular DNA within a 6 h time period following an IC50 drug exposure and repaired 48% over 24 h. No DNA repair was observed in ERCC1(-) cells during the same time periods. Both cell lines showed similar patterns of drug accumulation. For ERCC3(-) cells (IC50 = 54 microM) and ERCC3(+) cells (IC50 = 49 microM), the profiles of cisplatin sensitivity and cellular drug accumulation were similar. When treated with 50 microM cisplatin, these cells showed similar patterns of drug accumulation, and were equally efficient at forming and repairing lesions in cellular DNA. These data show that in UV repair-deficient CHO cells, ERCC1 confers resistance to cisplatin and confers the ability to remove platinum from cellular DNA. In contrast, ERCC3 does not influence cisplatin drug sensitivity or adduct repair capability. This suggests that ERCC1 may be a determinant of cisplatin resistance, whereas ERCC3 is probably not.

No abstract