Oxygen is the key factor associated with the difference between in vivo and in vitro effects of antioxidants

Antioxidants have received much attention as food factors and drugs. Fox et al. (1) reported that antioxidants induced DNA damage response and cell death in cultured cells. I believe that their study has exceptional importance; however, at the same time, we must be very careful in evaluating antioxidants on the basis of experiments performed under air exposure. The oxygen (O₂) concentration in water at 37 °C is ∼200 μM, and O₂ also dissolves from the air above the culture medium. Antioxidants easily reduce O₂ and generate hydrogen peroxide (H₂O₂) almost quantitatively, as evidenced by the fact that H₂O₂ is industrially produced by the reaction between O₂ and anthrahydroquinone derivative. Resveratrol, genistein, and baicalein share the same structure with anthrahydroquinone in having two or more hydroxyl groups at the aromatic ring. Therefore, most antioxidants behave as prooxidants when exposed to air. Because the rate of H₂O₂ formation depends on the antioxidant, each antioxidant behaves differently. Many studies have shown that antioxidants, including polyphenols and even vitamin C, induced apoptosis in cultured cells. I think that H₂O₂, having extremely high biological activity, is the key factor that explains these effects.

The antioxidant concentration used in the study by Fox et al. (1) was 92 μM, which could produce nearly 100 μM H₂O₂, at which many effects, including apoptosis, could be caused. Therefore, some of the reported effects are possibly explained on the basis of the amount of H₂O₂ produced. Considering these points, I would like to propose that an in vitro study of antioxidant effects be performed to determine the contribution of H₂O₂. As an example, experiments under low-O₂ concentration are necessary in the future.

On the other hand, O₂ concentration in the cells of animal tissues is far less than 40 mmHg (O₂ pressure in the vein); furthermore, O₂ concentration is 2–5% even in the arterial walls (2). Under low-O₂ concentration, the formation of H₂O₂ is largely delayed, thus enabling an antioxidant to function as an antioxidant. It is also well known that O₂ concentration in cancer cells is much less than that in normal cells. In addition, all antioxidants used in the study (1) are xenobiotics and their tissue concentration is assumed to be far less than 92 μM. Therefore, the statement “Our findings thus raised the possibility that resveratrol, baicalein, and genistein could be better used as cancer drugs” in the discussion section of the study by Fox et al. (1) may be controversial.