DNA polymerases are crucial constituents of the cellular
machinery for replicating and repairing DNA. Discerning mechanisms of DNA
polymerase is important for revealing the origin of fidelity discrimination. We
are characterizing the catalytic details of medium to low fidelity DNA
polymerases, namely DNA polymerase β (pol β), DNA polymerase λ
(pol λ), and DNA polymerase IV (Dpo4), through dynamics simulations,
quantum mechanics (QM), and quantum mechanics/molecular mechanics (QM/MM)
calculations. By linking simulation results and experiments in an iterative
program design, we are delineating how the specific but versatile active site
of each DNA polymerase controls efficiency and fidelity mechanisms in a
tailored fashion.
Findings:
(1) QM/MM studies of Pol β
Chemical Pathways
We performed a QM study to systematically explore the
potential energy surface of the nucleotidyl-transfer reaction in pol β.
The most favorable reaction route has an initial step of slight rearrangement
in the active site (of less than 1.5 kcal/mol) followed by two major steps: (1)
proton transfer from O3′ (primer) to O2α(Pα)
and the formation of a pentavalent, trigonal-bipyramidal Pα
center via an associative mechanism at a cost of 11 ~ 16 kcal/mol, indicating
a rate-limiting chemical step for correct nucleotide insertion in pol β;
and (2) dissociation of the triphosphate bond followed by a full transfer of
the nucleotide, which requires less than 7 kcal/mol.
In addition, we explored several pathways of the chemical
reaction of pol β for correct G:C versus incorrect G:G basepairings using
QM/MM techniques. Particularly, we analyzed the possible routes for the initial
deprotonation: (i) direct transfer to a phosphate oxygen of the incoming
nucleotide, (ii) direct transfer to an active-site Asp group and (iii) transfer
to explicit water molecules. The most probable initial step corresponds to step
(iii), with an activation energy of about 15 kcal/mol. Initial steps (i) and
(ii) appear less likely as they are at least 7 and 11 kcal/mol, respectively,
higher in energy. Overall, the rate determining step for both the correct and
incorrect nucleotide cases is the initial deprotonation; however, the
activation energy for the mismatched G:G case is 5 kcal/mol higher than the G:C
complex, due to active-site distortions.
In conclusion, our QM and QM/MM studies suggest that the
active-site rearrangement (first step in the most favorable pathway) to attain
a reactive geometry is essential and closely related to the “pre-chemistry”
avenue, which we have developed recently as a key step in the overall kinetic cycles
of DNA polymerases.
(2) MD and Experiments on Pol λ's Slippage Tendency
In our collaborative work, we reported crystal structures
and MD simulations of pol λ mutants (Arg517Ala and Arg517Lys) bound to a
primer-template during strand slippage. Relative to the primer strand, the
template strand is in multiple conformations, indicative of intermediates on
the pathway to deletion mutagenesis. We compared results from MD simulations
of Arg517Lys mutant ternary complexes to those of the wild-type as well as an
Arg517Ala mutant to elucidate critical interactions that constrain the DNA in
the ternary position. We found, intriguingly, two distinct conformations of
Lys517: in one orientation, Lys517 makes significant interactions with the DNA
while limited interactions with the DNA in the other. The limited interactions
enhance DNA motion toward the inactive binary position. Arg517's electrostatic
interactions with the template and the DNA backbone therefore contribute to pol
λ's specificity and stabilizing effect on the DNA. These findings also
indicate that DNA motion in pol λ is related to its slippage tendency; the
extent of DNA movement captured in our simulations mirrors the experimental
frameshift error rates: wild-type < Arg517Lys < Arg517Ala. The results
suggest that dNTP-induced repositioning of the template strand during the
normal catalytic cycle is a key to controlling strand slippage during DNA
synthesis.
(3) QM/MM study of the nucleotidyl-transfer reaction in
Dpo4
We examined the possible chemical reaction pathways of Dpo4
for dCTP insertion opposite 8-oxoguanine using the QM/MM approach. The most
favorable reaction path involves initial deprotonation of O3˘H via two bridging water molecules to O1A,
overcoming an energy barrier of 22.0 kcal/mol. The other feasible path is
initiated by deprotonation directly to O1A, but requires higher activation
energy (24.0 kcal/mol). In each pathway, the rate-limiting step is the initial
deprotonation rather than the formation and breaking of the O3˘–Pa
and Pa–O3A bonds. The trigonal-bipyramidal
configuration for Pa is
found in both pathways, suggesting the associative nature of the chemical
reaction in Dpo4. In contrast, the Dpo4/DNA complex with imperfect active-site
geometry needs to overcome much higher energy barriers, further emphasizing the
critical role of “pre-chemistry” reorganization to assemble the catalytic
region for the chemical reaction.
Participants in these research projects, many of whom are
young women, have received extensive training in investigating biological
function of enzymatic systems, applied molecular simulations, high performance
computing, statistical mechanics, quantum and molecular mechanics, and
development of free energy methods.