DNA polymerase I proofreading exonuclease activity is required for endonuclease V repair pathway both in vitro and in vivo

• Published in: DNA repair Vol. 64; pp. 59 - 67
• Main Authors: Su, Kang-Yi, Lin, Liang-In, Goodman, Steven D., Yen, Rong-Syuan, Wu, Cho-Yuan, Chang, Wei-Chen, Yang, Ya-Chien, Cheng, Wern-Cherng, Fang, Woei-horng
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01.04.2018
• Subjects: Deoxyinosine; DNA polymerase I 3′-5; DNA repair; Endonuclease V repair; exonuclease; In vivo repair assay; MALDI-TOF mass spectrometry; Reconstitution repair in vitro; Repair mechanism; Endonuclease V repair; DNA polymerase I 3′-5; Deoxyinosine; exonuclease; MALDI-TOF mass spectrometry; In vivo repair assay; Repair mechanism; Reconstitution repair in vitro; DNA repair
• ISSN: 1568-7864; 1568-7856
• DOI: 10.1016/j.dnarep.2018.02.005

•A phagemid-based deoxyinosine substrate can be repaired by minimal component reconstitute assay with endo V, polI and ligase.•The phagemid-based deoxyinosine substrate can be used for in vivo assay.•In vivo substrate with an additional C-C marker can identify strand loss associated with incomplete repair.•Using single nucleotide extension and MALDI-TOF MS mapped the 2-nucleotide excision track by pol I dI repair excision.•In vivo and in vitro results supported a complete mechanism for endo V repair pathway. Deamination of adenine can occur spontaneously under physiological conditions to generate the highly mutagenic lesion, deoxyinosine (hypoxanthine deoxyribonucleotide, dI). In DNA, dI preferably pairs with cytosine rather than thymine and results in A:T to G:C transition mutations after DNA replication. The deamination of adenine is enhanced by ROS from exposure of DNA to ionizing radiation, UV light, nitrous acid, or heat. In Escherichia coli, dI repair is initiated by endonuclease V (endo V; nfi gene product) nicking but a complete repair mechanism has yet to be elucidated. Using in vitro minimum component reconstitution assays, we previously showed that endo V, DNA polymerase I (pol I), and E. coli DNA ligase were sufficient to repair this dI lesions efficiently and that the 3′-5′ exonuclease of pol I is essential. Here we employed a phagemid-based T-I substrate mimicking adenine deamination product to demonstrate pol I proofreading exonuclease is required by the endo V repair pathway both in vitro and in vivo. In vivo we found that the repair level of an nfi mutant (11%) was almost 8-fold lower than the wild type (87%). while the polA-D424A strain, a pol I mutant defective in 3′-5′ exonuclease, showed a high repair level similar to wild type (both more than 80%). Using additional C–C mismatch as strand discrimination marker we found that the high level of dI removal in polA-D424A was due to strand loss (more than 60%) associated with incomplete repair. Thus, pol I proofreading exonuclease is the major function responsible for dI lesion removal after endoV nicking both in vitro and in vivo. Finally, using MALDI-TOF to analyze single-nucleotide extension product we show that the pol I proofreading exonuclease excises only 2-nt 5′ upstream of endo V incision site further honing the role of pol I in the endoV dI dependent repair pathway.