Research Article: Single-molecule FRET reveals proofreading complexes in the large fragment of Bacillus stearothermophilus DNA polymerase I

Date Published: May 23, 2018


Author(s): Thomas V. Christian, William H. Konigsberg.


There is increasing interest in the use of DNA polymerases (DNA pols) in next-generation sequencing strategies. These methodologies typically rely on members of the A and B family of DNA polymerases that are classified as high-fidelity DNA polymerases. These enzymes possess the ability to selectively incorporate the correct nucleotide opposite a templating base with an error frequency of only 1 in 106 insertion events. How they achieve this remarkable fidelity has been the subject of numerous investigations, yet the mechanism by which these enzymes achieve this level of accuracy remains elusive. Several smFRET assays were designed to monitor the conformational changes associated with the nucleotide selection mechanism(s) employed by DNA pols. smFRET has also been used to monitor the movement of DNA pols along a DNA substrate as well as to observe the formation of proof-reading complexes. One member among this class of enzymes, the large fragment of Bacillus stearothermophilus DNA polymerase I (Bst pol I LF), contains both 5′→3′ polymerase and 3′→5′ exonuclease domains, but reportedly lacks exonuclease activity. We have designed a smFRET assay showing that Bst pol I LF forms proofreading complexes. The formation of proofreading complexes at the single molecule level is strongly influenced by the presence of the 3′ hydroxyl at the primer-terminus of the DNA substrate. Our assays also identify an additional state, observed in the presence of a mismatched primer-template terminus, that may be involved in the transfer of the primer-terminus from the polymerase to the exonuclease active site.

Partial Text

Structure-function relationships of high-fidelity DNA polymerases have been studied extensively by a variety of techniques with the aim of understanding the mechanism involved in base selectivity. Despite impressive advances, many problems relating to the origin of their fidelity remain unsolved. Most of the kinetic studies have been carried out at the ensemble level and have led to the development of a general mechanism for the nucleotidyl transfer and exonuclease activities of DNA pols, but fail to account for their selectivity [1,2]. Observing molecular interactions at the single-molecule level enables the identification of subpopulations involved in various processes, such as conformational changes or transient binding events, that would be difficult to characterize in bulk solution [3]. Single-molecule Forster resonance energy transfer (smFRET) has been used successfully to identify subpopulations of DNA polymerase-DNA complexes undergoing conformational changes and subdomain movements important for base selectivity and editing that would be obscured in experiments carried out at the ensemble level [4–7]. When smFRET experiments are coupled with structural and kinetic evidence, important insights into the mechanism of these enzymes are likely to emerge.

Modified and unmodified oligonucleotides were purchased from either Integrated DNA Technologies (Coralville, IA) or Keck Biotechnology Resource Laboratory (New Haven, CT). A plasmid containing the IPTG inducible cDNA for the Bacillus stearothermophilus DNA polymerase I large fragment was kindly provided by Lorena Beese. Cy3B-NHS ester was purchased from GE Healthcare Bio-Sciences (Pittsburg, PA) and Atto647N-maleimide was purchased from ATTO-TEC GmbH (Siegen, Germany). Pyranose Oxidase, Catalase, D-glucose and Trolox were purchased from Sigma Aldrich (St. Louis, MO). All other reagents were purchased from Sigma-Aldrich (St. Louis, MO).

Single-turnover pre-steady state kinetic analysis of correct dNTP incorporation reveal that the Atto647N-labeled Bst pol I LF mutant had a kpol approximately 3-fold lower than wild-type Bst pol I LF (28.4 vs 90.9 s−1) at 25 °C. However, the apparent dissociation constant for the correct dNTP incorporation was also reduced by ~3-fold compared to that of the wild-type enzyme (11.5 vs 32.9 μM). The net effect is that the catalytic efficiency (kpol/Kd,app) of the labeled Bst pol I LF mutant vs wild-type Bst pol I LF remains virtually unchanged (2.47 vs 2.76 μM−1 s−1). The catalytic efficiency of incorrect nucleotide incorporation was also determined for the Atto647N-labeled Bst pol I LF mutant which was on the order of 1 × 10−5 μM−1 s−1. Thus, the Atto647N-labeled mutant of Bst pol I LF remained both active and selective following the mutagenesis and labeling procedures.

There has been great interest in the elucidation of the mechanism by which a DNA polymerase selects the correct dNTP substrate as well as how the enzyme processes mismatched bases incorporated into the primer strand [17–20]. These polymerases have been employed in a variety of sequencing strategies that either involve the enzyme directly or exploit some aspect of the enzyme’s mechanism [21,22]. Here we have used smFRET to monitor the conformational changes involved in dNTP selection as well as conformational changes involved in the formation of a proofreading complex in the presence of mismatched bases in a primer-template substrate with Bst pol I LF.




Leave a Reply

Your email address will not be published.