The objective of the proposed research is to define the roles of drug structure and DNA conformation in the recognition of biologically- relevant DNA targets by anticancer drugs. The research will focus on the interactions between nucleosomal DNA and four enediyne antitumor antibiotics: neocarzinostatin (NCS), esperamicin A1 (ESP A1), calicheamicin gamma1[I] (CAL), and C-1027. We have established that ESP A1 binds to DNA by intercalation of an anthranilate moiety, and the presence of a similar structure in C-1027 suggests that it too should undergo intercalation. CAL, however, may select curved DNA structures, and may induce the equivalent of helical winding in plasmid DNA. These hypotheses will be tested in the first two specific aims of the proposal. The question of intercalation by C-1027 will be addressed by characterizing DNA damage in chromatin and the effect of bound drug on the supercoiling and viscometric properties of naked DNA. The relationship between CAL binding and DNA bending or flexibility will be investigated by gel electrophoresis and DNA circle-closure assays, with CAL/ESP derivatives employed to identify critical drug structures. Such a relationship suggests that CAL damage sites may be nonrandomly distributed in the genome, an hypothesis that will be tested by statistical sequencing of damage sites in nucleosomal DNA and by analysis of the periodicity of damage sites in long tracts of defined- sequence DNA. In the third specific aim, enediyne target recognition processes will be studied in model nucleosome systems. CAL damage was found to be enhanced at a site of sharp DNA bending in nucleosomes reconstituted on 5S rDNA. The universal nature of this observation will be tested in isolated nucleosomes and nucleosomes reconstituted on a fragment of the tyrT gene of E. coli. To investigate the factors involved in this enhancement, CAL damage sites will be placed in different locations in the nucleosomal DNA. Finally, CAL, ESP A1 and NCS produce damage in different regions of the nucleosome, yet they are all potent cytotoxic agents. Based on the observed preference of NCS for damaging transcriptionally-active genes, it is hypothesized that the similar cytotoxicities of the enediynes may relate in part to the altered structure of nucleosomes present in active genes. To test this hypothesis, enediyne-mediated DNA damage will be studied in reconstituted and isolated nucleosome variants associated with transcriptional activity. The results will be compared to studies in classical nucleosomes. The results of the proposed studies should have broad implications for the design of enediyne agents, for identifying the biologic mechanisms of action of the enediynes and other DNA-directed anticancer drugs, and for the development of molecular probes of DNA structure in vivo. The importance of developing biologically-relevant model targets for the study of drug-DNA interactions is clear, in as much as they reveal features of molecular recognition not observed in other DNA models.
Showing the most recent 10 out of 12 publications
