**1. Introduction**

Vitamin C (ascorbic acid) is one of the best known and most studied dietary antioxidants. Its antioxidant credentials are well-established *in vitro*, since it can readily be demonstrated to prevent oxidation of lipids, DNA and other biological molecules. However, it is also possible for vitamin C to act as a pro-oxidant, through its ability to reduce transition metal ions, thus promoting the Fenton reaction which, acting on peroxides, produces highly reactive hydroxyl radicals [1]. The effect of vitamin C *in vivo* has been investigated in human trials with DNA oxidation as the measured endpoint. They include single dose intervention and longer trials, typically of a few weeks with daily supplementation. The first single-dose study was carried out by Green *et al.* [2], who administered 35 mg/kg of vitamin C to six healthy volunteers after an overnight fast and found a protection against γ radiation-induced DNA breaks in white blood cells, using the comet assay (described below). Since then, more than 20 intervention studies with vitamin C have been carried out, and they have been critically reviewed by Duarte and Lunec [3] who find that most of the studies show either a decrease in DNA oxidation or no effect, while there are some that show an increase in DNA lesions. Møller and Loft [4] also describe conflicting results with vitamin C, concluding that more studies with better designs to avoid bias should be made in order to understand the role of vitamin C in protecting against DNA oxidation. In a subsequent supplementation trial, Møller *et al.* [5] found that a slow release formulation of vitamin C had a more pronounced and sustained protective effect on the steady state level of DNA base oxidation than did a normal release formulation.

Herbert *et al.* [6] carried out a double-blind placebo-controlled trial giving different doses of vitamin C (0, 80, 200 and 400 mg/day) to four groups of 40 healthy volunteers over a period of 15 weeks (plus a 10 week washout period). They showed that vitamin C did not affect the intracellular level of 8-oxodGuo measured by HPLC. (This challenged an earlier report that vitamin C caused oxidation of bases in DNA [7]).

Sram *et al.* [8] recently reviewed human studies based on measurement of a variety of biomarkers of genetic damage—including some trials aimed at specific groups with occupational/environmental exposure to genotoxins. Generally either a protective effect or no effect was seen, depending—they suggest—on factors such as individual diet-derived vitamin C concentrations, levels of exposure to xenobiotics, and oxidative stress.

Bjelakovic *et al*. [9] published a meta-analysis of randomized trials of antioxidant supplements aimed at primary and secondary prevention of various diseases, with mortality as the endpoint. They concluded that antioxidant supplements have no beneficial effects on mortality, though in the case of vitamin C the number of individuals sampled is still relatively small and a definitive statement cannot be made.

Reviews of this subject conclude that we still need more studies to clarify the effect of vitamin C in humans [1,8–10]. We considered that it would be useful to investigate effects of vitamin C on genetic stability in cell culture, and so have examined, in the human tumor cell line HeLa, potential pro-oxidant, DNA-damaging effects of the vitamin; protection against DNA breakage caused by H2O2; and protection against base oxidation induced by photosensitizer plus visible light. In addition, we investigate the possibility that vitamin C might influence the capacity of cells for DNA repair.

We applied the comet assay, as used in many human biomonitoring studies, to the measurement of DNA damage, both strand breaks (SBs) and oxidized bases (employing the enzyme formamidopyrimidine DNA glycosylase (FPG) which converts 8-oxoGua and breakdown products of damaged purines to SBs). We used two approaches to measure repair of DNA damage [11]. If cells are subjected to damage (SBs or base oxidation), and then incubated to allow cellular repair, the residual lesions can be measured at intervals to show the kinetics of damage removal. Alternatively, in an *in vitro* assay, a cell extract is incubated with substrate DNA containing specific damage. The ability of the extract to introduce DNA breaks in the substrate cells reflects the activity of the enzymes responsible for the initial steps of repair, *i.e.*, removing the lesion.
