The synthesis of long and structurally complex RNAs remains challenging, particularly when multiple site-specific modifications are required. Current approaches typically rely on solid-phase synthesis followed by (splint-assisted) ligation using enzymes such as T4 RNA ligase, requiring extensive optimization of ligation junctions and reaction conditions.1 In contrast, RNA Dynamic Covalent Chemistry (RNA DCC) exploits the reversible cleavage and ligation activity of RNase T1 to assemble RNA hairpins through a 2′,3′-cyclic phosphate intermediate,2 routinely producing ligation yields exceeding 80%.
To establish the design principles governing RNA DCC, we systematically investigated the influence of hairpin loop identity, stem architecture and reaction conditions. We show that ligation efficiency is dictated primarily by hairpin loop stability, while selective protection of RNase-sensitive guanosine residues suppresses competing cleavage and enables efficient assembly of increasingly complex RNA hairpins.
RNA DCC is particularly attractive for the modular synthesis of functional RNAs, as many biologically important RNAs naturally contain stable hairpin motifs that can serve as ligation sites. As a proof of concept, we demonstrate efficient assembly of the first stem-loop of the CRISPR guide RNA scaffold. In the future, the constant Cas9-binding scaffold could be produced on a large scale, while only the short target-specific spacer sequence would need to be synthesized separately and assembled using RNA DCC. More broadly, RNA DCC offers a modular strategy for constructing complex functional RNAs, and the development of RNases with greater sequence specificity promises to further expand its scope.