31.3 N&v 569 Nr

their ability to search through vast tracts of DNA to find subtle anomalies in the structure. The human repair enzyme 8-oxoguanine glycosylase (hOGG1) is particularly impressive in this regard because it efficiently removes 8-oxoguanine (oxoG), a damaged guanine (G) base containing an extra oxygen atom, and ignores undamaged bases. Verdine and colleagues now report the structure of hOGG1 bound to undamaged DNA (Banerjee et al., page 612 of this issue), revealing a unique strategy that allows faithful removal of damaged bases but not their normal counterparts. The information content of the DNA double helix is preserved by a crew of DNArepair enzymes that defend the genome from the harmful effects of DNA damage. A specialized pathway known as base-excision repair (BER) plays a primary role in rectifying base damage. The damaged base is removed by BER glycosylases, followed by excision of the remaining sugar fragment and installation of an undamaged nucleotide by a DNA polymerase. Notably, inherited defects in BER have recently been linked to colorectal cancer. A variety of agents cause oxidative damage to DNA, including oxygen radicals and ionizing radiation. Oxidation of G to form oxoG produces a subtle structural transformation that results in deleterious mutations because DNA polymerases misread oxoG as a thymine (T) base when the genome is being duplicated during cell division. The human oxoG repair enzyme (hOGG1) catalyses excision of oxoG in the first step of BER. Structural studies of glycosylases involved in the repair process reveal common features of damaged-base recognition that include enzyme-initiated DNA distortion and bending to flip the damaged base out from the DNA double helix for recognition within a base-specific cavity of the enzyme. Verdine and colleagues previously determined the structure of an inactive hOGG1 variant bound to DNA containing an oxoG– cytosine (C) base pair.Surprisingly,the only obvious mechanism by which hOGG1 discriminates between oxoG and G is through a single hydrogen bond to oxoG. It seemed unlikely that this single interaction would be sufficient to allow discrimination between the two bases, particularly given the millionfold excess of G over oxoG within the human genome. In addition, hOGG1 makes extensive contacts with the orphaned cytosine base, which ensures that oxoG is removed only when in the appropriate base-pairing context. Although extensive biophysical and structural studies intimate that there are general features of damaged bases that signal their presence to repair enzymes, the steps involved in finding damaged bases in a sea of normal ones are still unclear. Most mechanisms invoke the enzyme sliding or hopping along the DNA duplex until a damaged site is detected.A particularly intriguing question is whether normal bases are also extruded from the helix during the search process. Banerjee et al. reasoned that a structural snapshot of the enzyme encountering a G should provide insight into how the normal base is distinguished from its damaged oxoG counterpart. However, because the enzyme does not specifically recognize G,encountering and rejecting G is a transient process. Indeed, much like a train that stops only at certain locations, hOGG1 probably pauses only when close to oxoG. To capture the enzyme at an undamaged G–C base pair, the authors used an innovative ‘covalent news and views