Chang, Hui-LanHui-LanChangKANG-YI SUGoodman, Steven DSteven DGoodmanChuang, Yung-ChuYung-ChuChuangHsu, Shen-JyueShen-JyueHsuFang, Yi-KaiYi-KaiFangYu, Hsiao-PeiHsiao-PeiYuFang, Cheng-HaoCheng-HaoFangYA-CHIEN YANGSUI-YUAN CHANGWOEI-HORNG FANG2025-09-172025-09-172025-0815687864https://scholars.lib.ntu.edu.tw/handle/123456789/732158Escherichia coli DNA polymerase I (Pol I) possesses a 3' to 5' proofreading function. Using a non-inhibitory in vitro proofreading assay and MALDI-TOF MS analysis, we demonstrated the Pol I proofreading function was effective at removal of mismatches within the primer-template junction. Mismatches of 1-4 nucleotides (nt) from the primer 3' end could be completely or partially corrected, with no additional editing observed further upstream. A backward movement mechanism was proposed involving distributive backtracking of polymerase along the template to remove non-fully complemented primers in order for DNA synthesis to recover. Co-editing DNA substrates containing two mismatches, one at 1-4-nt of the primer 3' end and the other outside of normal proofreading range, confirmed our distributive backtracking hypothesis. Additionally, a time course analysis revealed proofreading of internal mismatches was a non-processive reaction. To further confirm the validity of our proofreading model, we used in vivo, phagemid-derived nicked C-C substrates. Transformation results were consistent with the notion that mismatches located less than 4-nt upstream of the 3' end could be successfully proofread. In vivo proofreading of double mismatches also supports our model of polymerase backtracking for internal mismatch editing.enfalseDNA mismatchDNA polymerase IDNA replication errorIn vivo transformation assayMALDI MS analysisProofreading[SDGs]SDG3journal article10.1016/j.dnarep.2025.103864405779922-s2.0-105009043430