Directed evolution of mesophilic HNA polymerases providing insight into DNA polymerase mechanisms

Detailed biochemical characterization of natural and mutant enzymes provides essential clues to understand their mechanisms. There are, however, limits to the throughput of such approaches and they are not without errors. DNA polymerases have benefited from over 50 years of detailed study and remain not fully understood. As such, methods that allow high-throughput interrogation of variants, and viable analysis pipelines to identify relevant variants, become an important tool to accelerate research. Using the DNA polymerase from B. subtilis Phi29 bacteriophage as a model, we demonstrate how coupling focused libraries, selection and deep sequencing can be combined to identify variants of interest for characterization. As selection parameters can be controlled, different areas of an enzymes mechanism can be explored. Focusing selection on faster HNA (1,5-anhydrohexitol nucleic acid) synthesis, we identified P562del as a variant of interest, enriching significantly between rounds. Characterization confirmed its faster HNA synthesis initiation but lower processivity and fidelity. P562 is a non-conserved residue, unlikely to be selected by more traditional approaches, but its deletion recapitulates knowledge on how Phi29 exonuclease, thumb and TPR2 subdomains regulate polymerase function. Our data further support the hypothesis that Phi29 shows a two-state binding to its template: a fast non-replicative complex that transitions to a replication-competent state.