Repeated emergence of giant microRNA hairpins across invertebrates

While canonical microRNA (miRNA) hairpins typically bear 30-35 base pair (bp) stems and <15 nucleotide terminal loops, certain miRNA hairpins are much longer. While vertebrates lack long miRNA hairpins, these emerged multiple times within invertebrate lineages. Systematic assessments across >1,400 genomes provided evolutionary insights into the elongation of well-conserved miRNA precursors, which can harbor >1 kb between conserved miRNA and star species and generally form highly base-paired structures. Experiments in Drosophila and humans showed that flies were preferentially capable of maturing certain long precursors, but human cells had a partial capacity. However, neither could handle extreme miRNA hairpins. Finally, analysis of structural variants revealed that extensive stem structure and a local bulge near the dicing site are critical for biogenesis of a lengthened miRNA precursor; the latter likely represents an internal DGCR8 interaction platform. Altogether, we document unanticipated structural complexity in conserved miRNAs and emphasize bioinformatic challenges for their complete annotation.