The Dynamic Nucleotide Selection Mechanism of DNA Polymerase I at Atomic Resolution

Abstract

Replicative DNA polymerases use complex induced fit mechanisms to distinguish substrates complementary to the template from mismatches with remarkable speed and accuracy. The transient conformations and their order in the mechanism underly the speed and accuracy of polymerases, yet are difficult to characterize due to the sub‐millisecond time scales of the transitions. We employed microsecond scale, all‐atom, explicit waters molecular dynamics to characterize and visualize the conformational changes that occur when DNA polymerase I, large fragment, from Bacillus stearothermophilus encounters a complementary dCTP or a mismatched dTTP opposite a template dG. The dCTP and the Mg2+ cofactor immediately formed a stable base pair with dG and induced fluctuations of the mobile fingers subdomain between the open conformation and a state intermediate between open and closed. The enzyme closed only upon the binding of a second cation to the active site, followed by a rotation of a conserved histidine towards the 3′‐OH of the growing primer strand. We speculate this is the deprotonation step that initiates catalysis. In the presence of the mismatched dTTP‐Mg2+, a weak base pair forms with dG, but quickly breaks, leading to near dissociation of the dTTP during the 3μs simulation. These simulations detail the complex polymerase mechanism and provide structural explanations for the behavior and high fidelity of dynamic DNA polymerases during nucleotide selection.Support or Funding InformationThe authors would like to acknowledge the Jeffress Trust Awards Program in Interdisciplinary Research for funding, as well as the University of Richmond for computing resources.This work also used Extreme Science Engineering Discovery Environment (XSEDE) allocations TG‐MCB140003 and TG‐MCB140073, which are supported by National Science Foundation grant number OCI‐1053575. Additional computing resources were utilized as part of the MERCURY Consortium supported by NSF grant number CHE‐1229354