Hydroxyl radical is highly reactive and certainly capable of destroying isolated DNA, protein, or lipid in a test tube. However, a highly reactive species is not necessarily highly toxic. In the case of hydroxyl radical, the rate of reaction with every organic molecule is near the diffusion limit, ranging from 109 to 1010 M-1 ‚s-1. Thus, hydroxyl radical will randomly damage isolated DNA in a simple phosphate-buffered solution. Binding of DNA binding proteins protects sequences from hydroxyl radical, enabling molecular biologists to identify specific regions of the DNA that interact with the protein. As more and more components of a cell are added to the mixture, the chances that DNA or any critical cellular target will be hit by a highly reactive species become smaller and smaller. Radiation chemists calculate that the diffusion distance of hydroxyl radical is only 3 nm in a cell (2). Hydroxyl radical will randomly attack noncritical cellular components because its broad reactivity at diffusionlimited rates makes hydroxyl radical an indiscriminant species. The reductionist nature of biochemistry to work on clean, isolated systems exaggerates the apparent toxicity of hydroxyl radical because alternative noncritical targets are removed from the system. An additional difficulty is that natural antioxidants and scavenging enzymes like superoxide dismutase and catalase are absent from in vitro systems but are present in vivo. Still, the injection of exogenous superoxide dismutase or catalase into animals and humans can reduce injury from ischemia and inflammation, which establishes that oxidative processes are involved in many disease processes that cannot be prevented by endogenous antioxidant defenses (3, 4). However, the superoxide-driven Fenton reaction is a relatively slow reaction that is easily stopped in vitro by adding small amounts of catalase or superoxide dismutase. In vivo, up to 1% of soluble protein is superoxide dismutase (5, 6), while catalase and glutathione peroxidase are also abundant and effectively remove hydrogen peroxide. Even submicromolar concentrations of hydrogen peroxide will not exist for long in vivo. Yet, millimolar concentrations are called low doses in most in vitro experiments. To understand oxidant toxicity in vivo, one needs to look for free radical reactions that are fast enough to outcompete endogenous antioxidant defenses. The reactions must be fast enough that neither superoxide dismutase nor catalase will stop the reaction. Other oxidative mechanisms must be operating in vivo. In the present review, we will focus upon only one reactionsthe diffusion-limited reaction of superoxide with nitric oxide. More certainly remain to be investigated. At present, the only known biological molecule that is produced in high enough concentrations and can react fast enough with superoxide to outcompete endogenous superoxide dismutase is nitric oxide (6). Nitric oxide is a free radical that combines by radical-radical coupling with superoxide to form peroxynitrite anion (7):

The reaction rate is 6.7× 109 M-1 ‚s-1, which is at least 3 times faster than superoxide dismutase reacts with superoxide (8, 9). The scavenging of superoxide by superoxide dismutase is partially reduced by physiological levels of chloride ions, which screen the electrostatic field that attracts superoxide to the active site. Chloride decreases the scavenging of superoxide by a factor of 2-3 (9, 10), but should have no effect upon the radical-radical coupling of nitric oxide and superoxide. Therefore, the formation of peroxynitrite is slightly more favorable under physiological conditions compared to reactions conducted in phosphate buffer. Peroxynitrite itself is not a free radical because the …