PURPOSE: To propose an improved version of RADACK, a stochastic simulation of radiolytic attack on DNA, that takes into account the reactivity of each amino acid of a specifically bound protein with hydroxyl radicals. To apply it to the natural lactose operator-repressor complex taking advantage of recently reported structures. To compare the obtained probabilities of DNA strand break induction with those calculated with the previous versions and with an experimental pattern of strand break probabilities. MATERIALS AND METHODS: Models of complexes close to the natural ones, derived from crystallography- and NMR-based structures recently available in the PDB databank, were used. The specific chemical reactivity of each amino acid was introduced in the new version of RADACK (the reactivity model). The probabilities of strand break induction by the irradiation of the complex were calculated with this new version as well as with previous ones. RESULTS: The patterns of probabilities of strand break induction calculated with the improved version of RADACK were partially different from those obtained with previous versions. The patterns obtained for both, using putative models of natural complexes, were consistent with the experimental results, but some discrepancies were suggestive of slight differences between these structures and the real natural system. The crystallographic structure agreed best with the experimental results. CONCLUSIONS: A new version of RADACK was validated that took into account the reactivity of atoms in both DNA and protein. The putative modelled structures of a natural lactose operator-repressor complex were discussed.
PURPOSE: To propose an improved version of RADACK, a stochastic simulation of radiolytic attack on DNA, that takes into account the reactivity of each amino acid of a specifically bound protein with hydroxyl radicals. To apply it to the natural lactose operator-repressor complex taking advantage of recently reported structures. To compare the obtained probabilities of DNA strand break induction with those calculated with the previous versions and with an experimental pattern of strand break probabilities. MATERIALS AND METHODS: Models of complexes close to the natural ones, derived from crystallography- and NMR-based structures recently available in the PDB databank, were used. The specific chemical reactivity of each amino acid was introduced in the new version of RADACK (the reactivity model). The probabilities of strand break induction by the irradiation of the complex were calculated with this new version as well as with previous ones. RESULTS: The patterns of probabilities of strand break induction calculated with the improved version of RADACK were partially different from those obtained with previous versions. The patterns obtained for both, using putative models of natural complexes, were consistent with the experimental results, but some discrepancies were suggestive of slight differences between these structures and the real natural system. The crystallographic structure agreed best with the experimental results. CONCLUSIONS: A new version of RADACK was validated that took into account the reactivity of atoms in both DNA and protein. The putative modelled structures of a natural lactose operator-repressor complex were discussed.