Engineering chimeric DNA polymerases for DNA movable type storage

Abstract DNA-based information storage offers a promising alternative to conventional media due to its high density, long-term stability, and low energy requirements. However, its application remains hindered by synthesis costs, limited sequence length and poor scalability. DNA polymerase is a critical enzymatic tool in the DNA storage systems by enabling high-fidelity data writing and targeted sequence amplification. In this study, we engineered chimeric DNA polymerases by fusing the high-fidelity 9°N DNA polymerase with double-stranded DNA binding proteins derived from thermophilic archaea. These fusions significantly enhanced processivity, thermal stability, and salt tolerance by stabilizing enzyme-template interactions, mimicking sliding clamps while preserving catalytic efficiency. Leveraging these properties, we demonstrated precise file retrieval from a mixed oligonucleotide pool using orthogonal barcode primers. Compared with wild-type 9°N, the chimeric polymerases, particularly PLS, exhibited reduced substitution error rates and improved read accuracy. We then applied these enzymes to a DNA movable type storage system, where prefabricated DNA modules were assembled into encoding blocks. Using engineered polymerases, these blocks were recombined to enable flexible data rewriting without de novo DNA synthesis. Moreover, a multi-enzyme assembly strategy enabled the construction of kilobase-scale DNA sequences encoding a classical Chinese poem, achieving complete data recovery. All assembled fragments remained stable in E. coli over 100 generations, exhibiting the potential for in vivo storage. Collectively, our findings demonstrated the role of engineered DNA polymerases for DNA-based information storage. Moreover, this system reduced synthesis demands, supported scalable rewriting, and ensured long-term preservation, offering a practical route to sustainable, high-fidelity DNA data storage.