Structures of LIG1 that engage with mutagenic mismatches inserted by polβ in base excision repair

Abstract

DNA ligase I (LIG1) catalyzes the ligation of the nick repair intermediate after gap filling by DNA polymerase (pol) β during downstream steps of the base excision repair (BER) pathway. However, how LIG1 discriminates against the mutagenic 3′-mismatches incorporated by polβ at atomic resolution remains undefined. Here, we determine the X-ray structures of LIG1/nick DNA complexes with G:T and A:C mismatches and uncover the ligase strategies that favor or deter the ligation of base substitution errors. Our structures reveal that the LIG1 active site can accommodate a G:T mismatch in the wobble conformation, where an adenylate (AMP) is transferred to the 5′-phosphate of a nick (DNA-AMP), while it stays in the LIG1-AMP intermediate during the initial step of the ligation reaction in the presence of an A:C mismatch at the 3′-strand. Moreover, we show mutagenic ligation and aberrant nick sealing of dG:T and dA:C mismatches, respectively. Finally, we demonstrate that AP-endonuclease 1 (APE1), as a compensatory proofreading enzyme, removes the mismatched bases and interacts with LIG1 at the final BER steps. Our overall findings provide the features of accurate versus mutagenic outcomes coordinated by a multiprotein complex including polβ, LIG1, and APE1 to maintain efficient repair.

Structures of LIG1 uncover the mechanism of sugar discrimination against a ribonucleotide at 3’- and 5’-end of the nick DNA

“…Our previous and present LIG1 structures revealed the features of DNA substrate and LIG1 interaction that dictate accurate versus mutagenic outcomes (53). Our structures contribute to the understanding of the mechanism by which LIG1 discriminates against non-canonical DNA intermediates during nick sealing at each individual step of the ligation reaction as well as how the ligase active site engages with nick DNA distinctly depending on the architecture of cognate (step 2), mismatched (step 1 or 2), or ribonucleotide (step 3)-containing ends (Supplementary Table 1).…”

“…In our previously solved LIG1/A:T structure, we reported that the active site can accommodate a cognate A:T base pair where DNA-AMP intermediate is formed (53). The comparison of the 2Fc-Fo maps at 3’-OH and 5’-P ends of the nick DNA demonstrated continuous and discontinuous maps in the LIG1 structures of rG:C and A:T, respectively (Figure 2A-B).…”

“…Our previously solved LIG1/mismatch structures demonstrated that the A:C is positioned at the initial step (step 1) where the ligase active site stays adenylated (LIG1-AMP) at its active site lysine residue (K568), and that the LIG1/G:T mismatch is captured during step 2 when an adenylate is transferred to the 5’-end of the nick (53), while the LIG1/rG:C structure shows a phosphodiester bond formation coupled to AMP release at step 3 of the ligation reaction (Supplementary Figure 2). We also compared the map of LIG1/rG:C (EE/AA) with the previously solved structure of LIG1/C:G (wild-type) (54,55).…”

“…In our structure studies, we use the EE/AA mutant of LIG1 carrying Glu(E)346 and Glu(E)592 mutations at the Mg 2+ -dependent high-fidelity site to crystallize the ligase in the conditions lacking Mg 2+ (53,55). In the reaction conditions that mimic the LIG1 crystals, we also observed efficient nick sealing of the DNA substrate with 3’-rG:C in the absence of Mg 2+ for longer time points of incubation (Supplementary Figure 7).…”

“…Furthermore, at atomic resolution, we recently reported that the LIG1 discriminates against the mutagenic mismatches distinctly depending on the architecture of 3’-terminus at the nick DNA. Our structures of LIG1/nick DNA complexes with G:T and A:C mismatches revealed the ligase strategies that favor or deter the ligation of base substitution errors at the final step of BER pathway (53).…”