skip page navigationOregon State University

News

The following letter was submitted to the journal Nature on April 17, 1998.

Vitamin C has not been shown to cause oxidative damage in vivo

Sir - In the 9 April issue of Nature, Podmore et al.1 argue that 500 milligrams per day of supplemental vitamin C exhibits genotoxic pro-oxidant effects in human peripheral lymphocytes, a conclusion based on the measurement of two oxidative DNA adducts: 8-oxo-guanine (oxo8Gua) and 8-oxo-adenine. We do not agree that their conclusions, which have been publicized throughout the world, are supported by the data. We take issue with the high values they report for the oxidative DNA adducts and the claim that their values are "remarkably similar to those found in other in vivo systems"1. The values reported agree with some other measurements using gas chromatography-mass spectrometry (GC-MS), but are one to three orders of magnitude higher than those obtained by other techniques. The authors fail to acknowledge the ongoing debate in the scientific literature about artifactual DNA oxidation and the great uncertainty regarding the accurate measurement of oxidative DNA adducts2-6, upon which their entire study and conclusions are based.

The authors obtained mean values for oxo8Gua of "30 lesions per 105 guanine bases"1, and tried to support the validity of these estimates by citing similar values derived from independent GC-MS studies of DNA adducts in human lymphocytes7, cultured human cells8, and mouse brain9. In contrast, numerous studies using high-performance liquid chromatography with electrochemical detection (HPLC-EC) have reported much lower baseline values, for example: 3.74 ± 1.62 and 2.9 ± 1.5 oxo8Gua per 105 guanine bases in human lymphocytes10,11; 0.24 ± 0.13 and 0.19 ± 0.13 in human polymorphonuclear and mononuclear leukocytes, respectively12; 0.52 ± 0.25 in human lung13; and 0.11 ± 0.01 and 0.04 ± 0.002 in liver DNA from old and young rats, respectively6. A recent study measuring oxo8Gua by HPLC-EC in liver DNA of rats supplemented with ascorbate found no effect of ascorbate, with oxo8Gua levels of 1.89 ± 0.32 and 1.93 ± 0.65 per 105 guanine bases and a 59-fold difference in liver ascorbate concentrations14. In addition, as has been discussed recently by us6 and others3, even the current lowest HPLC-EC values for oxo8Gua may be overestimates. Recently developed cell and molecular biological methods for measuring oxidative DNA adducts have generated values for oxo8Gua of between 0.008-0.024 adducts per 105 guanine bases (i.e., 1250 to 3750 fold lower than those reported by Podmore et al.1), a range which includes human lymphocytes3,15.

Thus, the measurement of oxidative DNA adducts is problematic and prone to very substantial ex vivo oxidation artifacts. In our own work, we have observed oxidation during both DNA isolation and hydrolysis, and have adopted methods to minimize artifacts, such as using a recently introduced phenol-free method of DNA extraction16 and adding the divalent metal chelator desferal to prevent ex vivo Fenton chemistry (rather than EDTA, which chelates iron in a redox-active form and can act as a pro-oxidant)6. With such changes, we have arrived at our current baseline estimate (for young rat liver) of 0.04 oxo8Gua per 105 guanine bases6, approximately 750 times lower than the values of Podmore et al.1, and considerably lower than some of our own past values17, which now must be interpreted accordingly6. We cannot evaluate the methods used by Podmore et al.1, since they are not detailed in the paper, but a recent study has demonstrated artifactual formation of oxidized DNA bases during sample preparation and derivatization for GC-MS analysis5. Although only few 8-oxo-adenine measurements are available in the literature for comparison with the study of Podmore et al.1, our arguments apply equally to this DNA adduct, the abundance of which in vivo should be similar to that of oxo8Gua.

It appears, therefore, that the oxidative DNA adducts measured by Podmore et al.1 do not accurately reflect DNA damage in vivo. It is likely that >90% of the DNA adducts were formed ex vivo. The elucidation of ascorbate's in vivo effects on oxidative DNA damage has to rely on more definitive methodologies and more thorough experimentation. In addition, the value of ascorbate as an antioxidant and for human health does not rest on a few measurements of oxidative DNA damage, but rather is supported by hundreds of biochemical, cellular, physiological, nutritional and epidemiological studies18. It would be as unwise to recommend vitamin C supplementation based on a single measurement of decreased oxidative DNA damage as it is to discourage its use based upon this study1.

In conclusion, we believe that the results presented are an ex vivo artifact, given the high values obtained. In the context of the huge literature supporting the health benefits of vitamin C, the conclusions of the study are unwarranted.

Kenneth B. Beckman, Hal J. Helbock, Bruce N. Ames
Department of Molecular and Cell Biology, 401 Barker Hall, University of California, Berkeley, CA 94720, USA
e-mail: bnames@uclink4.berkeley.edu

Balz Frei
Linus Pauling Institute, Oregon State University, 571 Weniger Hall, Corvallis, OR 97331, USA
e-mail: balz.frei@orst.edu

--------------------------------------------------------------------------------

1 Podmore, I. D. et al. Nature 392, 559 (1998).

2 Beckman, K. B. & Ames, B. N. J. Biol. Chem. 272, 19633-19636 (1997).

3 Collins, A., Cadet, J., Epe, B. & Gedik, C. Carcinogenesis 18, 1833-1836 (1997).

4 Cadet, J., Douki, T. & Ravanat, J. L. Environ. Health Perspect. 105, 1034-1039 (1997).

5 Douki, T., Delatour, T., Bianchini, F. & Cadet, J. Carcinogenesis 17, 347-353 (1996).

6 Helbock, H. J. et al. Proc. Natl. Acad. Sci. U. S. A. 95, 288-293 (1998).

7 Olinski, R. et al. Cancer Lett. 106, 207-215 (1996).

8 Jaruga, P. & Dizdaroglu, M. Nucleic Acids Res. 24, 1389-1394 (1996).

9 Liu, P.K. et al. J. Neurosci. 16, 6795-6806 (1996).

10 Liu, L. et al. Mutation Research 370, 145-150 (1996).

11 Inoue, T. et al. Jpn. J. Cancer Res. 84, 720-725 (1993).

12 Nakajima, M., Takeuchi, T. & Morimoto, K. Carcinogenesis 17, 787-791 (1996).

13 Asami, S. et al. Carcinogenesis 18, 1763-1766 (1997).

14 Cadenas, S., Barja, G., Poulsen, H. E. & Loft, S. Carcinogenesis 18, 2373-2377 (1997).

15 Pflaum, M., Will, O. & Epe, B. Carcinogenesis 18, 2225-2231 (1997).

16 Nakae, D., Mizumoto, Y., Kobayashi, E., Noguchi, O. & Konishi, Y. Cancer Lett. 97, 233-239 (1995).

17 Fraga, C. G., Motchnik, P. A., Wyrobek, A. J., Rempel, D. M. & Ames, B. N. Mutation Research 351, 199-203 (1996).

18 Packer, L. & Fuchs, J., eds, Vitamin C in Health and Disease (Marcel Dekker, New York, 1996).