Esau L. Medina, PhD Candidate
DNA polymerases are an ancient class of enzymes responsible for maintaining and replicating genetic information in living organisms. Their ability to replicate long DNA polymers with high fidelity and efficiency has led to widespread applications in biotechnology. However, the strict substrate specificity exhibited by most naturally occurring DNA polymerases limits their utility in applications requiring greater chemical diversity. To address this limitation, directed evolution technologies have been developed, enabling the creation of engineered polymerases capable of synthesizing modified or unnatural nucleic acid polymers, known as xeno-nucleic acids (XNAs).. These non-cognate nucleic acid substrates, which feature alterations to the sugar, backbone, or nucleobase moieties, follow Watson-Crick base-pairing rules and exhibit enhanced stability across thermal, biological, and chemical conditions.
In this dissertation, we describe a directed evolution campaign aimed at converting a family of highly selective DNA polymerases into an unnatural homolog with strong RNA synthesis activity. Using a homologous recombination strategy, we diversified the library by shuffling four B-family DNA polymerase homologs. Polymerase variants were expressed in E. coli and encapsulated in water-in-oil droplets, which contained triphosphates and a fluorescent polymerase activity sensor. Through a compartmentalization-based microfluidic selection strategy (DrOPS), we rapidly identified the engineered polymerase C28, a homolog capable of synthesizing RNA at a rate of ~3 nt/sec and >99% fidelity. C28 demonstrated the ability for long-range synthesis, reverse transcription, and chimeric DNA-RNA amplification. Although the enzyme readily accepts a broad range of RNA analogs, it discriminates strongly against most XNAs. The approach taken to evolve polymerase C28 may be generalizable to expand polymerase applications for synthetic biology.
