Site-Specific DNA Damage Recognition by Enzyme-Induced Base Flipping

One of the most interesting enigmas in enzymatic DNA recognition is the mechanism by which damaged bases buried in the DNA base stack are detected and subsequently excised. The essential enzyme combatants in this never-ending battle against genomic instability, the DNA repair glycosylases (Fig. 1A), must act specifically against a variety of base lesions that arise from oxidative reactions, deamination events, and alkylating agents, without also mistakenly removing normal nucleobases (1). This is a high stakes game: the cost of being slow is irreversible damage to the coding content of the genome, and the cost of being sloppy is the indiscriminate introduction of toxic abasic sites in DNA (2-4). Because many damaged bases are nonperturbing to the duplex structure and may differ from normal bases by only a single atom change, an extraordinary and highly conserved recognition mechanism has evolved to accomplish this task. This mechanism has been termed "base flipping," and involves the dramatic rotation of an entire damaged base and sugar from the DNA duplex using binding forces imposed by the enzyme (Fig. 1B) (5-7). The problem facing these enzymes is both enormous and unique. Consider that a damaged base may compose only one out of 10⁷ base pairs in the human genome (8, 9). Thus, a crude estimate of the specificity for excision of a specific damaged base over a normal base of (k_cat{plus 45 degree rule}K_m)^specific{plus 45 degree rule}(k_cat{plus 45 degree rule}K_m)^nonspecific ≥ 10⁷ may be assumed. Although such issues as increased accessibility of damaged vs normal bases in the densely packed chromatin may account for some of this apparently large specificity, it is difficult to dismiss the magnitude of the discrimination achieved by the recognition mechanisms of these enzymes. Equally remarkable is that specificity is achieved without forming sequence specific contacts with the DNA substrate. Indeed, any specificity mechanism that arose during natural selection involving sequence specific interactions would have been selected against, because this mechanism compromises the ability of the organism to execute unbiased repair of DNA lesions regardless of the sequence context. Accordingly, DNA glycosylases present a unique evolutionary solution to site recognition that is fundamentally different from that of restriction enzymes, repressor proteins, or transcription factors. Structural and other biophysical studies over the last four years have begun to reveal the basis for DNA glycosylase specificity and the critical role of DNA base flipping (5, 10-16). The system that has advanced most rapidly in this time span is that of the DNA repair enzyme uracil DNA glycosylase. This enzyme removes unwanted uracil residues in DNA that have arisen by spontaneous deamination of cytosine, or by misincorporation of dUTP during DNA replication (17). UDG has provided an experimentally tractable system to investigate the multistep process of base flipping, using a wide range of experimental approaches. The findings in this system should have widespread applicability given the highly conserved structural features of enzymes that flip bases (Fig. 1B). The general questions that have driven our research in base flipping over the last few years also direct the course of this chapter. To begin, we ask why has base flipping evolved as the sole mechanism for recognition and catalysis by DNA glycosylases? Second, what are the natures of the energetic barriers that an enzyme must overcome to extricate a base from its position in the DNA duplex? And finally, what temporal events occur during enzymatic base flipping, and how do the various steps contribute to catalytic specificity? We are continuing to develop new experimental tools to investigate base flipping, and we present these new approaches in the hope that they become generally useful in the investigation of other enzymes that flip bases.