Selective Nonenzymatic Formation of Biologically Common RNA Hairpins

While the prebiotic synthesis of the building blocks of life is well supported experimentally, no chemical process is known to selectively yield common biological RNA sequences and structures. RNA oligomers carrying functional information are generally believed to emerge from a large pool of random sequences as a result of Darwinian evolution. Herein, we profiled the nonenzymatic partitioning of randomized RNA overhangs into ligated products via two competing prebiotic pathways: loop-closing ligation and splint-directed ligation, using deep sequencing. Comprehensive sequence-reactivity profiles demonstrate that loop-closing ligation preferentially yields hairpin structures with loop sequences UNNG, CNNG, and GNNA (N represents A, C, G, or U). These loop sequences significantly overlap with the most common biological tetraloops. Conversely, splinted ligation products favor CG-rich overhangs. Leveraging these sequence rules, we successfully achieved high-efficiency template-free assembly of functional ribozymes from short RNAs. Our results support the notion that a one-step, prebiotic chemical process could have selectively created elementary RNA structures that are common in biology. The shared sequence selectivity of loop-closing ligation and biological tetraloops suggests that underlying chemical interactions within unpaired RNA sequences may have governed both the primordial assembly of functional RNAs and the evolutionary outcome of extant biology.