Structure of DNA modified by highly mutagenic and tumorigenic agents, and their less active analogs will be computed with our program DUPLEX, in order to define those conformational features that underlie the mutagenic and carcinogenic activity. We are especially interested in modification by polycyclic aromatic amines and hydrocarbons, with emphasis on the following issues: (1) chemical nature of the adduct, including ring size and substituent effects; (2) nature of base modified; (3) position of modification on a given base; (4) effect of neighboring base sequences. Aromatic amine adducts of highest priority are aniline (AN), 4- aminobiphenyl (ABP), 4-acetylaminobiphenyl (AABP), 2-aminofluorene (AF), 2- acetylaminofluorene (AAF) and 1-aminopyrene (AP) bound to guanine C-8, as well as ABP bound to adenine C-8. Next in priority are N-acetylbenzidine (NAB) and the cooked food mutagens Trp-P-2 and IQ, which also form adducts to guanine C-8. Among the polycyclic aromatic hydrocarbon adducts, our first priority remains the N-2 guanine adduct of the prototype highly tumorigenic (+) anti-benzo(a)pyrene 7, 8-diol-9, 10-epoxide (BPDE) and that of its much less active (-) anti analog. The N-2 guanine adducts of the highly tumorigenic (+) anti 5-methylchrysene-1, 2-diol-3, 4 epoxide, (+) 5- MCDE, and the much less active (+) 6-MCDE analog make up a similar pair of interest to us. The adenine N-6 and guanine N-2 adducts of (-) benzo(c)phenanthrene 3, 4-diol-1, 2-epoxide-2, (-)B(c)PHDE-2 which are apparently both highly mutagenic, are next in priority. Base sequences that are identified mutagenic hotspots are to be examined first. Of current interest are the CGC sequence, and sequences with purines adjacent to the modification. In single stranded or duplex trimers, we can investigate all sequence combinations surrounding the lesion to ascertain whether some conformational feature unique to a given sequence might predict a mutagenic hotspot. Both duplexes and single strands will be investigated. Our energy minimized structures are to be used as starting conformations for molecular dynamics simulations with the AMBER force field to provide for explicit solvent and salt incorporation. In addition, we have established collaborations with investigators who are using NMR techniques to study some of the modified DNAs, and are computing energy minimized structures that are within bounds of the experimental data. A number of new computational approaches are also planned including continued development of a build-up technique for prediction of modified DNA structures, molecular dynamics simulations with larger time steps, an effort to compute free energy rather than potential energy differences between conformers, and employment of a simulated annealing algorithm to diminish the multiple minimum problem. In addition, it is planned to automate and document DUPLEX, and release it to the scientific community.
Showing the most recent 10 out of 127 publications
