This research is based on compelling evidence that the variable sensitivity of different mammalian cell types to bifunctional antineoplastic drugs correlates with extent of formation or rate of removal of DNA interstrand cross-links. The major aim of this proposal is to identify, isolate and characterize human-cell enzymes potentially involved in repair of cross-link lesions. In particular, we will test the hypothesis that excision of monoadduct reactive intermediates of cross-links, renders cells relatively resistant by suppressing cross-link formation. Substrates for in vitro enzyme assay will be made by reacting DNA with a range of bifunctional agents, including the nitrosoureas (e.g., BCNU); nitrogen, phenylalanine and phosphoramide mustards; mitomycin-C; and cisplatin. Cross-linking will be determined in vitro by measuring the renaturability of DNA with fluorometric methods. Initially, known DNA repair enzymes such as DNA-glycosylases or alkyltransferase will be tested for their capacity to remove cross-link intermediates. Secondly, new activities for cross-link-precursor excision will be sought in human cell extracts. In addition, all these enzymes and extracts will be tested for their ability to cleave existing cross-links. In parallel with in vitro enzyme studies, human cell lines with varied sensitivity to each drug will be characterized with respect to formation and removal of cross-links in vivo, as determined by DNA renaturability, alkaline elution and buoyant density centrifugation methods. The ultimate goal is to correlate repair-enzyme activity, cross-link formation and cross-link persistence with cytotoxicity. Such information will allow predictions of drug response based on direct enzyme assay in normal or neoplastic tissue. Finally, intervention in the repair process by enzyme inhibition might provide a means to overcome drug resistance.
Showing the most recent 10 out of 18 publications
