Interactions of the UVRABC endonuclease in vivo and in vitro with DNA damage produced by antineoplastic anthracyclines.

The anthracycline antineoplastic agents Adriamycin and N-trifluoroacetyl-Adriamycin-14-valerate were assayed in vivo and in vitro for ability to produce DNA lesions recognized by the UVRABC endonuclease, a DNA repair enzyme of Escherichia coli which recognizes large, bulky lesions in DNA. We found that, while both drugs produce DNA lesions, only the lesions produced by Adriamycin were toxic. Hence, anthracycline antineoplastic activity may be related to production of large, bulky lesions in DNA, while toxicity may correlate with toxicity measured in a simple E. coli DNA repair mutant test system.

Repair of haloethylnitrosourea-induced DNA damage in mutant and adapted bacteria.

The sensitivities of Escherichia coli K-12 strain AB1157, its uvrA-deficient mutant AB1886, and its recA mutant AB2463 to N,N'-bis(2-chloroethyl)-N-nitrosourea, N-(2-chloroethyl)-N-nitrosourea, and N-ethyl-N-nitrosourea have been determined. These data indicate that loss of either uvr excision repair or recA-dependent DNA repair greatly increases sensitivity to the haloethylnitrosoureas. At the same time, loss of recA-dependent DNA repair increases sensitivity to N-ethyl-N-nitrosourea significantly while loss of uvr excision repair increases sensitivity to this agent only marginally. Adapting the uvrA-deficient and recA-deficient mutants by growth in N-methyl-N'-nitro-N-nitrosoguanidine increases survival after exposure to either N-methyl-N'-nitro-N-nitrosoguanidine or N-ethyl-N-nitrosourea, but neither adapted strain loses its sensitivity to N,N'-bis(2-chloroethyl)-N-nitrosourea. Taken together, these data indicate that the haloethylnitrosoureas cause other important cytotoxic lesions in DNA in addition to those involving alkylation of the O6 position of guanine and that the uvrA and recA gene products are involved in the repair of these lesions.

Analysis of mRNA synthesis following induction of the Escherichia coli SOS system.

Escherichia coli responds to impairment of DNA synthesis by inducing a system of DNA repair known as the SOS response. Specific genes are derepressed through proteolytic cleavage of their repressor, the lexA gene product. Cleavage in vivo requires functional RecA protein in a role not yet understood. We used mRNA hybridization techniques to follow the rapid changes that occur with induction in cells with mutations in the recA operator or in the repressor cleavage site. These mutations allowed us to uncouple the induction of RecA protein synthesis from its role in inducing the other SOS functions. Following induction with ultraviolet light, we observed increased rates of mRNA synthesis from five SOS genes within five minutes, maximum expression ten to 20 minutes later and then a later decline to near the initial rates. The presence of a recA operator mutation did not significantly influence these kinetics, whereas induction was fully blocked by an additional mutation in the repressor cleavage site. These experiments are consistent with activation of RecA protein preceding repressor cleavage and derepression of SOS genes. The results also suggest that the timing and extent of induction of individual SOS genes may be different.

Properties and regulation of the UVRABC endonuclease.

This report summarizes the cloning of the uvrA, uvrB and uvrC genes of E. coli, the identification and isolation of the gene products, the regulation of the genes, and reconstitution of active UVRABC endonuclease from the individually isolated components.

Potential radiosensitizing agents. Dinitroimidazoles.

New compounds of the nitroimidazole series have been synthesized as radiosensitizers which selectively sensitize hypoxic cells to the lethal effect of radiation. The reaction of 2,4(5)-dinitroimidazole (2) with chloroethanol or hydrochloric acid yielded 4(5)-nitro-5(4)-chloroimidazole (3), which upon reaction with ethylene oxide yielded the 4-nitro-5-chloroimidazole-1-ethanol (6). Reaction of 2 with ethylene oxide resulted in a mixture of two compounds, the 2,4-dinitroimidazole-1-ethanol (4) and 2,3-dihydro-5-nitroimidazo[2,1-b]oxazole (5). The structure of the new heterocyclic compound 5 was confirmed by 1H NMR, mass spectrum, and X-ray crystallography. These agents were tested for their ability to sensitize hypoxic Escherichia coli cells to killing by ionizing radiation. Compound 4 was found to be the most active agent of this series of compounds.

Radiosensitization, pharmacokinetics, and toxicity of a 2-nitroimidazole nucleoside (RA-263).

A 2-nitroimidazole nucleoside, 1-(2',3'-dideoxy-alpha-D-erythro-hex-2'-enopyranosyl)-2-nitroimida zole (RA-263), has been investigated for its radiosensitization, pharmacokinetics, and toxicity properties. The in vitro radiosensitization tests against hypoxic Chinese hamster (V-79) cells demonstrated that RA-263 was a more potent radiosensitizer than misonidazole and at 2 mM concentration approached the oxic curve. Significant in vitro radiosensitization activity was also observed in EMT6 mammary tumor cells. The in vitro cytotoxicity data suggested that RA-263 is considerably more toxic to hypoxic cells than misonidazole. The increased cytotoxicity may be related to its higher depletion of nonprotein thiols (NPSH) than misonidazole. The combined effects of radiosensitization and hypoxic cell toxicity were measured by preincubation of the V-79 cells for 4 h under hypoxic conditions before irradiation. The results demonstrated a synergistic response by causing a significant decrease in the extrapolation number with loss of shoulder of the radiation survival curves. The in vivo radiosensitization experiments conducted by the in vivo-in vitro cloning assay with the EMT6 mammary tumor indicate that RA-263 is an effective sensitizer. Pharmacokinetic data suggested that RA-263 was eliminated from plasma by a rapid alpha phase and a slower beta phase with T 1/2 of 36 and 72 min, respectively. The concentration in the brain was approximately one-sixth of tumor concentration, suggesting that RA-263 is excluded from the CNS. Moreover, RA-263 was two times less toxic than misonidazole on equimolar basis by acute LD50 tests. This agent was also significantly less mutagenic than misonidazole in a strain of Escherichia coli.

Cross-linking studies with the uvrA and uvrB proteins of E. coli.

The interactions of the uvrA and uvrB proteins with DNA have been investigated using a DNA-protein cross-linking technique. It is demonstrated that hydrolysis of ATP by the uvrA protein facilitates cross-linking of this protein to single-stranded DNA, whether the DNA is UV irradiated or not. In contrast, cross-linking to unirradiated double-stranded DNA is not facilitated by ATP hydrolysis and is in fact increased by the substitution of the non-hydrolysable analogue aTP gamma S for ATP. In the presence of ATP, a dose-dependent increase is observed in the amount of uvrA protein which can be cross-linked to UV-irradiated double-stranded DNA. Binding of uvrB protein to puvrA-DNA complexes has a stabilising effect and increases the number of complexes which can be cross-linked whether the substrate is single- or double-stranded DNA. We can find no evidence that ATP hydrolysis by uvrA protein results in unwinding of UV-damaged DNA.

Repair of plasmid DNA damaged in vitro with cis- or trans-diamminedichloroplatinum(II) in Escherichia coli.

Plasmid pBR322 was modified in vitro with the antitumor compound cis-diamminedichloroplatinum(II) (cis-DDP) or the isomeric trans-DDP. The numbers of platinum adducts were determined by atomic absorption spectrophotometry. DNA-repair-proficient and various DNA-repair-deficient (uvrB, uvrD, recB and recA) strains of Escherichia coli were transformed by the damaged plasmids and the ratios of the transformation frequencies of cells by damaged plasmids relative to those by untreated plasmids were determined. Results of transformation assays indicated that the uvrB gene function was essential for repair of plasmid DNA damaged with cis-DDP. A functional recA gene product seemed to be of minor importance for repair of plasmids damaged with cis-DDP. trans-DDP had a different effect on plasmid DNA. trans-DDP-modified DNA was better able to transform cells than cis-DDP-modified DNA, and the DNAs appeared to be repaired differently. Prior induction of SOS functions increased the survival of plasmids treated with cis-DDP in wild-type and uvrD mutants, but did not increase the survival of plasmids damaged with trans-DDP in these strains. In in vitro repair experiments, plasmid DNA modified with cis-DDP was more readily incised by the UVRABC excinuclease than that modified with trans-DDP.

A novel repair enzyme: UVRABC excision nuclease of Escherichia coli cuts a DNA strand on both sides of the damaged region.

The uvrA, uvrB, and uvrC proteins of Escherichia coli were purified from strains that greatly overproduce these proteins. Using the purified proteins, the UVRABC nuclease was reconstituted in vitro. The reconstituted enzyme acted specifically on DNA damaged with UV, cis-platinum, and psoralen plus near UV. When UV-irradiated DNA was used as substrate, the enzyme made two cuts on the damaged DNA strand, one on each side of the damaged region. The enzyme hydrolyzed the eighth phosphodiester bond on the 5' side of pyrimidine dimers. On the 3' side of pyrimidine dimers, the UVRABC nuclease cut the fourth or the fifth phosphodiester bond 3' to pyrimidine dimers. The oligonucleotide with the damaged bases that is generated by these two cuts was released during treatment with the enzyme. We have also obtained evidence suggesting that the enzyme acts by the same mechanism on PydC photoproducts which are thought to be of primary importance in UV-induced mutagenesis.