High-resolution crystal structures reveal plasticity in the metal binding site of apurinic/apyrimidinic endonuclease I

Apurinic/apyrimidinic endonuclease I (APE1) is an essential base excision repair enzyme that catalyzes a Mg2+-dependent reaction resulting in the cleavage of the phosphodiester backbone 5′ of an abasic site within double-stranded DNA.1,2 Pre-steady-state turnover rates for APE1’s endonuclease activity have been estimated to be greater than 7003 or 850 s–1,4 whereas steady-state turnover rates are approximately 2 s–1.5 Thus, substrate turnover is diffusion-limited, and the slow step of the reaction occurs after the chemistry.4 In fact, product release has been proposed to be the slow step of the reaction.6 The endonuclease reaction involves a one-step associative phosphoryl transfer mechanism, with water serving as the nucleophile7 and preference for the Rp stereoisomer during cleavage.2 However, the number of metal ions involved and their coordination in the enzyme remains controversial. A two metal ion mechanism was first proposed by Steitz and co-workers for Klenow fragment,8,9 related polymerases, and associated exonucleases.10 Since then, there has been a general assumption that most enzymes that cleave DNA, including APE1, will in fact use two metal ions.11 APE1 is most closely related in structure to Escherichia coli exonuclease III, which binds a single metal ion in the active site.12 The two metal ion assumption was challenged by Tainer and co-workers, who put forth a mechanism for APE1 involving only one metal ion based on structural and enzymatic characterizations of an APE1–DNA complex.6 In that work, they reported a 3.0 A structure with one Mn2+ ion bound to E96 in the active site with cleaved DNA. Crystal structures of APE1 with Pb2+ bound in the active site of the enzyme were then reported, and a two metal ion mechanism was proposed.13 In this mechanism, coordinating residues for the metal ions included D70 and E96 for one metal and H309, D210, and N212 for a second metal ion, despite the fact that Mg2+ strongly prefers coordinating oxygen ligands.14 The two metal ion binding sites then formed the basis of a proposed moving metal ion mechanism involving Mg2+ binding first to a site coordinated by D210 and N212 and then moving 5 A to a site coordinated by D70, E96, and D308.15,16 Finally, a 25Mg solid-state NMR study reported that APE1 binds one mole equivalent of Mg2+, which is disordered due to its coordinating ligands, suggesting plasticity in the active site.17 Recently, a 2.4 A resolution structure of an APE1–Mg2+–product complex was reported in which Mg2+ is coordinated solely to E96,18 as was shown previously for the Mn2+ complex. Although the effect of substituting E96 on the catalytic activity was characterized in this recent study,18 there is currently no solution data for direct metal binding by APE1. In the absence of substrate, Mg2+ is coordinated by D70 and E96 in crystal structures reported to date at moderate resolution.19 In this study, we present the highest resolution structure of APE1 as a complex with bound Mg2+ determined at 1.4 A, the structure of APE1 bound to Mn2+, and the first apo-APE1 structure (i.e., without bound metal). Our motivation for this study was to establish a structural basis for metal binding in the absence of substrate and to determine the contributions of specific residues on metal binding and catalysis by APE1. Our results provide new insights on the initial capture of metal ion by APE1 involving remarkable plasticity of a metal-coordinating ligand within the active site.