**4. Discussion**

*In vitro*, vitamin C is capable of acting as a pro-oxidant, particularly in combination with metal ions [14], and there are concerns that it might lead to excessive DNA oxidation in humans. There are reports of increases in oxidized bases in the DNA of white blood cells after supplementation of normal subjects with vitamin C [7] or with vitamin C plus ferrous sulfate [15]. However, these studies were carried out using GC-MS to measure oxidized bases—a technique now recognized as being particularly prone to artifact and therefore unsuitable for determining low level base oxidation [16].

The saturation plasma level for vitamin C in humans is around 80 μM, and this concentration is approached with daily intake of 200 mg/day [17]—an intake that can feasibly be attained by a recommended intake of several portions of fruits and vegetables each day. The justification for exceeding in our experiments the maximum vitamin C concentration found in plasma is that vitamin C is reportedly not stable in cell culture medium; 40% depletion was detected after 2 h of incubation [18], and only about 30% remained at 6 h.

*In vivo*, in subjects supplemented with 200 mg of vitamin C per day, active accumulation by lymphocytes results in intracellular levels saturating at 3.5 mM [17]. Welch *et al.* [19] reported a linear accumulation of vitamin C up to about 0.5 mM in 4 h in human fibroblasts incubated in 50 μM vitamin C. We found that the concentration of vitamin C inside HeLa cells after culture in medium with the vitamin was in the mM range, and was still high after 24 h. Although they have undergone many changes in their several decades of continuous culture from their origin in tumor tissue, HeLa cells evidently retain the active transport mechanisms responsible in normal cells for uptake of vitamin C [19], and so we regard them as suitable for an investigation of intracellular effects of vitamin C.

In the present study, in spite of the high intracellular concentration of vitamin C, there was no increase in SBs or FPG-sensitive sites. Others have reported no increase in SBs and/or oxidized bases in various cell types after incubation with vitamin C—in human HepG2 liver-derived cells with 10 μM vitamin C [20], in primary human fibroblasts incubated with 100 μM vitamin C [21], and most recently in HL-60 cells (human leukemic cell line) incubated with concentrations of 50–250 μM vitamin C [22,23].

In our experiments, preincubation with up to 200 μM vitamin C had no significant effect on the level of H2O2-induced breaks or 8-oxoGua induced by treatment with Ro plus light. There are mixed reports on the effect of pre-incubation with vitamin C on the DNA damage inflicted by genotoxic agents. Vitamin C pre-incubation (30 min, 10 or 50 μM) decreased the SBs caused by streptozotocin in lymphocytes or HeLa cells [24], and protected against the effect of NiCl2 in lymphocytes [25]. Duarte and Jones [21] preincubated primary human fibroblasts for 12 h with 20, 100 or 500 μM vitamin C, and—also using the comet assay—found an increase in H2O2-induced damage. Carbofuran-induced damage to human lymphocyte DNA was decreased by preincubation with 30 μM vitamin C [26]. Similar protection was afforded by 10–50 μM vitamin C pretreatment against damage induced by 2-hydroxyethyl methacrylate and urethane dimethacrylate (used in dental treatment) [27,28].

When vitamin C is present simultaneously with a genotoxic agent (rather than during a preincubation), potentiation of the damaging effect of the latter may be seen—even when the effect of preincubation is a protective one. Such a dual effect was reported by Wozniak and Blasiak [25] for NiCl2. Vanadyl sulfate led to more DNA breaks in HeLa cells and lymphocytes when they were co-incubated with vitamin C for 1 h [29], perhaps as a result of a metal ion-vitamin C interaction. Both DNA breaks and oxidized purines were induced when HL-60 cells were incubated with vitamin C and copper (II) sulfate, but no such damage was caused by vitamin C alone [23]. Vitamin C was unable to protect human submandibular gland and oral epithelial cells against DNA damage induced by bracken fern, and in fact showed a synergistic effect on DNA breakage [30]. In contrast, co-incubation of HepG2 cells with both vitamin C and *N*-nitrosamines resulted in fewer DNA breaks/oxidized bases [20]—perhaps an effect of vitamin C mediated through xenobiotic metabolizing enzymes.

To summarize our results relating to DNA damage, in HeLa cells—which clearly are able to take up and accumulate vitamin C—we find no sign of either pro- or anti-oxidant effects.

There are some human intervention studies that suggest a possible regulatory role for vitamin C in the repair of oxidized DNA—a topic highlighted by Sram *et al.* in their recent review [8] as deserving further attention. Cooke *et al*. [31] measured 8-oxodGuo in lymphocyte DNA, serum and urine from 30 healthy volunteers involved in a 6 week placebo and 6 week vitamin C (500 mg/day) supplementation study (plus a washout period). They found a decrease in the levels of 8-oxodGuo in DNA and an increase in urine and serum. In addition to reservations about the less than ideal study design and the likely errors in the measurement of 8-oxodGuo, the interpretation of the results is questionable, since the presence of 8-oxodGuo in serum or urine does not directly reflect OGG activity; the base, not the nucleoside, is excised during BER. Tarng *et al.* [32] supplemented chronic hemodialysis patients with 300 mg of vitamin C, and found a significant upregulation of *hOGG1* mRNA at 24 h after vitamin C administration. Astley *et al.* [33] took lymphocytes from human volunteers after 3 weeks of supplementation with the modest dose of 60 mg vitamin C per 2 days. There was no significant effect on plasma vitamin C, and no change in repair capacity (detected as repair synthesis on an oxidatively damaged plasmid template).

Supplementation for 3 weeks with green kiwifruits led to a significant increase in plasma vitamin C, decreases in endogenous base oxidation and H2O2-induced damage *ex vivo*, and an enhancement of lymphocyte BER on a substrate containing 8-oxoGua [34]; however, golden kiwifruits modestly increased plasma vitamin C and showed protection against DNA oxidation but no effect on DNA repair [35]. Guarnieri *et al.* [36] supplemented volunteers with 0.5 g/day vitamin C in normal or slow release form (together with vitamin E); only the slow release form stimulated BER capacity in lymphocytes.

Reports in the literature on effects of vitamin C on DNA repair in cells in culture are sparse. Konopacka *et al.* [37] irradiated mouse lymphocytes with 2 Gy of γ-rays, and then incubated them with or without vitamin C, vitamin E and β-carotene; rejoining of strand breaks was faster in the presence of antioxidants.

We report here that vitamin C has no effect on cellular repair of SBs or oxidized bases in HeLa cells. The *in vitro* test for 8-oxoGua DNA glycosylase (OGG) activity could not be carried out, since it was evident that vitamin C itself can cause damage to nucleoid DNA (Figure 6), presumably acting as a pro-oxidant (as mentioned in the Introduction). This cleavage could be mediated via production of hydroxyl radicals in the presence of DNA-bound transition metal such as copper [14].

The reason for native nuclear DNA being resistant to attack by vitamin C, while nucleoid DNA is susceptible, is not clear, though it may be that this apparently pro-oxidant effect of vitamin C is prevented by the high concentrations of the thiol-rich antioxidant glutathione present in the intact nucleus. Bergstrom *et al.* [22,23] report the oxidation of dGuo *in vitro* by vitamin C. To summarize, we have found no evidence in HeLa cells for pro-oxidant effects of vitamin C at extracellular concentrations up to 200 μM, but also no evidence for antioxidant protection. Nor did we see any effect on the repair of SBs or oxidized bases. Vitamin C at low concentration is, however, capable of damaging nucleoid DNA. The same strategy was used in our laboratory to study the antioxidant/pro-oxidant effects of β-cryptoxanthin [13]. Our results showed that this carotenoid does not cause DNA damage but, in contrast to vitamin C, it protects HeLa cells from damage induced by H2O2 or photosensitizer plus light. In addition, it enhances DNA repair, as measured with the same assays as were used in the present report.
