Cellular and adenovirus dl312 DNA metabolism in cycling or mitotic human cultures exposed to supralethal gamma radiation.

Cellular repair of DNA damage due to lethal gamma irradiation was studied to reveal differences between strains and cell cycle stages that are otherwise difficult to detect. Cycling and metaphase-blocked cultures of normal fibroblasts and carcinoma cells were compared for repair of gamma sites (gamma radiation-induced nicks, breaks, and alkalilabile sites in DNA) at supralethal exposures ranging from 7 to 150 krad 137Cs radiation and at postirradiation incubations of 20-180 min. Fibroblasts from normal human skin or lung repaired gamma sites efficiently when cycling but did not repair them when blocked at mitosis. Bladder (253J) or lung (A549) carcinoma cells, unlike normal fibroblasts, repaired gamma sites efficiently even when blocked at mitosis. HeLa cells degraded their DNA soon after exposure at all doses tested, regardless of mitotic arrest. Whether the above differences in DNA repair between cell cycle stages and between strains result from differences in chromatin structure (cis effects) or from differences in the nuclear enzymatic environment (trans effects) could be resolved by placing an inert, extrachromosomal DNA molecule in the cell nucleus. Specifically, cis effects should be confined to the host chromosomes and would not be detected in the inert probe whereas trans effects should be detected in host chromosomes and inert probe DNA alike. Indeed, we found a suitable DNA molecule in the adenovirus deletion mutant dl312, which does not proliferate in the absence of E1A complementation. Gamma sites in 32P-labeled adenovirus dl312 DNA were repaired efficiently in all hosts, regardless of mitotic arrest. Failure of mitosis-arrested fibroblasts to repair gamma sites was therefore due to a cis effect of chromatin organization rather than to a trans effect such as repair enzyme insufficiency. In sharp contrast, chromosomes of mitotic carcinoma cells remained accessible to repair enzymes and nucleases alike. By means of these new tools, we should get a better understanding of higher-order chromatin management in normal and cancer cells.