Emergence of diverse ligase reactivities from a single RNA evolution experiment

RNA enzymes that assemble RNA from shorter pieces would have been important for primordial RNA-based biology. To explore the scope of RNA-catalyzed RNA assembly, we used directed evolution to re-engineer a ligase ribozyme that uses 5'-phosphorimidazolide RNA substrates into a ribozyme that ligates RNA substrates carrying the biologically relevant 5'-triphosphate group. Unexpectedly, this single RNA evolution experiment generated, in addition to the desired triphosphate ligase, four other distinct classes of ligase ribozymes. Three of these four ligase classes were isolated at extremely low abundances, collectively covering <0.1% of the selected RNA population. One class exhibits strict specificity for 5'-phosphorimidazolide substrates, even though these substrates were never presented during evolution. Another ligase class mediates a reaction between its 5'-triphosphate and the substrate's terminal 2'-OH, but only when the substrate carries a 3'-phosphate. This reaction resembles those catalyzed by protein enzymes such as RtcB ligase and plant RNA splicing ligases in RNA repair pathways. The remaining classes catalyze two different but related reactions, each involving the 5'-triphosphate groups on the ribozyme and a distinct internal hydroxyl group on the substrate. These reactions resemble branching reactions catalyzed by naturally occurring ribozymes such as the group II intron and the spliceosome, even though these ribozymes emerged from a synthetic RNA library with no relation to biological catalysts. The emergence of such catalytic diversity from a single RNA evolution experiment under highly constrained selection pressures and the discovery of new enzyme reactivities in rare RNA sequences highlight the catalytic flexibility and evolutionary potential of RNA. These findings strengthen the plausibility that an RNA World could have supported a wide range of chemistries essential for the emergence of life.