Inhibition of DNA synthesis by nitroheterocycles. I. Correlation with half-wave reduction potential.

Twenty-one nitroheterocycles, including metronidazole, misonidazole and AF-2, were tested for their ability to inhibit DNA synthesis in mouse L-929 cells growing in culture. All those tested inhibited the rate of incorporation of 3H-thymidine into L cells following drug treatment for 4 h under aerobic conditions. Only 4 drugs reached their limits of solubility before the uptake of 3H-thymidine was inhibited by 50% or more. For the remaining 17, the log of the concentration producing 50% inhibition of incorporation was directly correlated with the half-wave reduction potential of the compound.

Summary.-Twenty-one nitroheterocycles, including metronidazole, misonidazole and AF-2, were tested for their ability to inhibit DNA synthesis in mouse L-929 cells growing in culture. All those tested inhibited the rate of incorporation of 3H -thymidine into L cells following drug treatment for 4 h under aerobic conditions. Only 4 drugs reached their limits of solubility before the uptake of 3H-thymidine was inhibited by 50%o or more. For the remaining 17, the log of the concentration producing 50%o inhibition of incorporation was directly correlated with the half-wave reduction potential of the compound.
NITROHETEROCYCLES are used widely in industry and medicine as food preservatives, antibacterial agents, pesticides, dye intermediates, and explosives. Correlations have been observed between the electron-affinity of many nitroheterocycles and their toxicity towards bacterial and mammalian cells (Adams et al., 1976a;Hirano et al., 1967;Sasaki, 1954), their ability to cause DNA damage (Olive & J)urand, 1978) and to act as hypoxic cell radiosensitizers and mutagens (Olive & I)urand, 1978;Adams et al., 1976b). All these diverse effects require the presence of the nitro group (Katae et at., 1967; and most appear to be related to the ability of nitroheterocycles to form toxic intermediates capable of interacting and binding to cell components (McCalla et al., 1970;Olive & McCalla, 1977).
Flagyl (metronidazole) wNas a gift from Searle Laboratories, San Juan, Puerto Rico. Misonidazole was kindly supplied by Dr C. Smithen, Roche Pharmaceutical Company, England. Dimetronidazole was obtained from Salisbury Labs, Iowa. 3,5-Dinitrobenzonitrile, 2-methyl-5-nitroimidazole 5-nitro-2-furoic acid, 4-nitroimidazole and 8nitroquinoline were purchased from Aldrich Chemical Company. Drugs were prepared before use from stock solutions in DMSO (Sigma) at a concentration of 20 to 200 mg/ml) Reduction potentials.-Half-wave reduction potentials (E-) were measured with a Princeton Applied Research Model 364 polarographic analyser with a dropping mercury electrode and standard calomel electrode (SCE). Values for the reduction potential were obtained w%Aith the differential pulse mode. Nitroheterocycles were dissolved in phosphate-buffered saline (Dulbecco formulation from GIBCO), pH 7-2. Chemical structures and reduction potentials are given in the Table. The half-wave reduction potential for 5-nitro-2-furaldehyde diacetate was -220 mV immediately after dissolving in buffer. but fell to -350 mV after standing for 24 h.
Cells.-The mouse L-cell parent line w%as purchased from Grand Island Biological Company, Grand Island, New York. A sublire was obtained after 2 years in suspension culture that was more sensitive than the parent line to nitrofurazone toxicity under aerobic incubation, and this line was used for subsequent experiments. The plating efficiency of this line was 0-50-0-58. Cells were maintained in suspension culture in Joklic modified minimal essential medium with 10% foetal calf serum from GIBCO.
Measurement of thymidine incorporation.-Approximately 4 x 105 exponentially growing L-929 cells from a suspension culture were allowed to attach to 60mm Falcon plastic Petri dishes for 1 h before the experiment. The medium on the plates was then replaced with fresh medium containing 15% foetal calf serum and 0-05 ,uCi/ml 14C-TdR (Amersham, 80 C/m.ol). After 15 min the radioactive medium was removed, the cells (attached to the plates) washed several times wN,ith fresh medium, and then incubated for 4 h with nitroaromaties in a humidified CO2 incubator. After treatment, the cells wi-ere washed free of drug and resuspended in medium containing 150/ foetal calf serum and 2 ,uCi/ml 3H-TdR (Amersham, 18 C/mol) for 15 min. After Awashing, the cells were removed from the plate with trypsin, resuspended in 1 ml medium, and the cells counted with an electronic cell counter (Coulter Electronics, Hialea, Florida). Fifty microlitres of cell suspension was then pipetted on to 4 24Imm filter-paper discs (Whatman GF/A). After drying, the discs were washed twice in cold 50o TCA followed by 2 wAashes in 950% ethanol. When dry, the discs were introduced into scintillation vials containing 5 ml of scintillation fluid, and the radioactivity determined with a Beckman 8100 liquid scintillation counter. The amount of TdR incorporated after drug treatment wz-as determined using 2 methods. First, the radioactivity incorporated per cell was calculated, and second, the ratio of incorporated 3H-TdR (given after drug treatment) to 14C-TdR (given before drug treatment) was also determined. The latter method was found to be more reproducible.

RESULTS
Inhibition of DNA synthesis by nitroaromatics occurs over a wide range of concentrations (Fig. 1). With the exception of 4-nitroimidazole (E , --675 mV) and 5-nitro-2-furoic acid (EI-400 mV), which reached the limits of solubility with little evidence of inhibition of DNA synthesis, a 4h incubation of all nitroaromatics under aerobic conditions inhibited subsequent incorporation of 3H-TdR. The shapes of the curves describing the effects of the nitroheterocycles on DNA synthesis were similar for all nitroheterocycles examined (Fig. 1). In 3 experiments, niridazole at concentrations of 0 5-0 75 mm decreased incorporation to 70 0 of the control level, but not further. Similarly, 0 5-0 75mm NP-10 reduced incorporation to 60% of the control value.  Table. corporation of 3H-TdR to 50O% of the amount in untreated cells was determined for each nitroaromatic after a 4h exposure under aerobic conditions. A linear relation was obtained when the redox potential was plotted as a function of the log of this concentration (Fig. 3). The electron-affinity of nitroheterocycles has been correlated with a number of biological effects, including their ability to inhibit colony formation under aerobic conditions (Adams et al., 1976b). It is therefore not surprising that the log of the concentration producing 5000 inhibition of DNA synthesis can also be correlated with the electron-affinity of this series of nitroheterocycles. In fact, inhibition of DNA synthesis was probably the mechanism of the "cytotoxicity" observed when Chinese hamster V79 cells in culture were treated under aerobic conditions over a long term with a series of nitroaromatics (Chapman et al., 1973;Adams et al., 1976b).
Other toxic effects of nitroheterocycles are greatly enhanced by anaerobic conditions, which accelerate metabolism of the nitro group by cultured cells to toxic species (Olive & McCalla, 1977). However, inhibition of DNA synthesis by nitroheterocycles occurs even under aerobic conditions. The presence of the nitro group is required for this effect, as evidenced by the absence of effects onDNA synthesis by derivatives lacking the nitro group, as well as the correlation observed here between electron-affinity and inhibition of DNA synthesis. Also, as shown in the following paper (Olive (1979)), reduced products of nitrofurazone had no effect on DNA synthesis. This suggests that the nitro group must be intact for the compound to inhibit DNA synthesis. It seems probable that the parent compound or the nitro anion radical, which is formed under aerobic as well as anaerobic conditions (Mason & Holtzman, 1975;Wardman & Clarke, 1976;Sealy et al., 1978), may be responsible for the inhibition of DNA synthesis by nitroheterocycles. The mechanism behind this inhibition is explored further in a subsequent paper.
The author wishes to thank Barb L. Thomas for excellent technical assistance. This investigation was supportedl by Grants Number CA 245519 and CA 06973 awarde(d by the National Canicer Tnstitute, DHEWVAI.