To assess miRNA evolution across the genus, we analyzed several billion small RNA reads across 12 fruit fly species. These data permit comprehensive curation of species- and clade-specific variation in miRNA identity, abundance, and processing. Among well-conserved miRNAs, we observed unexpected cases of clade-specific variation in 5' end precision, occasional antisense loci, and putatively noncanonical loci. We also used strict criteria to identify a large set (649) of novel, evolutionarily restricted miRNAs. Within the bulk collection of species-restricted miRNAs, two notable subpopulations are splicing-derived mirtrons and testes-restricted, recently evolved, clustered (TRC) canonical miRNAs. We quantified miRNA birth and death using our annotation and a phylogenetic model for estimating rates of miRNA turnover. We observed striking differences in birth and death rates across miRNA classes defined by biogenesis pathway, genomic clustering, and tissue restriction, and even identified flux heterogeneity among clades. In particular, distinct molecular rationales underlie the distinct evolutionary behavior of different miRNA classes. Mirtrons are associated with high rates of 3' untemplated addition, a mechanism that impedes their biogenesis, whereas TRC miRNAs appear to evolve under positive selection. Altogether, these data reveal miRNA diversity among species and principles underlying their emergence and evolution.

Circadian clocks enable organisms to anticipate and adapt to fluctuating environmental conditions. Despite substantial knowledge of central clock machineries, we have less understanding of how the central clock's behavioral outputs are regulated. Here, we identify Drosophila miR-124 as a critical regulator of diurnal activity. During normal light/dark cycles, mir-124 mutants exhibit profoundly abnormal locomotor activity profiles, including loss of anticipatory capacities at morning and evening transitions. Moreover, mir-124 mutants exhibited striking behavioral alterations in constant darkness (DD), including a temporal advance in peak activity. Nevertheless, anatomical and functional tests demonstrate a normal circadian pacemaker in mir-124 mutants, indicating this miRNA regulates clock output. Among the extensive miR-124 target network, heterozygosity for targets in the BMP pathway substantially corrected the evening activity phase shift in DD. Thus, excess BMP signaling drives specific circadian behavioral output defects in mir-124 knock-outs.

Comparative genomic analyses of microRNAs (miRNAs) have yielded myriad insights into their biogenesis and regulatory activity. While miRNAs have been deeply annotated in a small cohort of model organisms, evolutionary assessments of miRNA flux are clouded by the functional uncertainty of orthologs in related species, and insufficient data regarding the extent of species-specific miRNAs. We address this by generating a comparative small RNA (sRNA) catalog of unprecedented breadth and depth across the Drosophila genus, extending our extant deep analyses of D. melanogaster with sRNA data from multiple tissues of 11 other fly species. Aggregate analysis of several billion sRNA reads permits curation of accurate and holistic compendia of miRNAs across this genus, providing abundant opportunities to identify species- and clade-specific variation in miRNA identity, abundance, and processing. Amongst well-conserved miRNAs, we observe unexpected cases of clade-specific variation in 5′ end precision, occasional antisense loci, and some putatively non-canonical loci. We also employ strict criteria to identify a massive set (649) of novel, evolutionarily-restricted miRNAs. Amongst the bulk collection of species-restricted miRNAs, two notable subpopulations of rapidly-evolving miRNAs are splicing-derived mirtrons and testis-restricted, clustered (TRC) canonical miRNAs. We quantify rates of miRNA birth and death using our annotation and a phylogenetic model for estimating rates of miRNA turnover in the presence of annotation uncertainty. We show striking differences in birth and death rates across miRNA classes defined by biogenesis pathway, genomic clustering, and tissue restriction, and even identify variation heterogeneity amongst Drosophila clades. In particular, distinct molecular rationales underlie the distinct evolutionary behavior of different miRNA classes. We broaden observations made from D. melanogaster as Drosophilid-wide principles for opposing evolutionary viewpoints for miRNA maintenance. Mirtrons are associated with a high rate of 3′ untemplated addition, a mechanism that impedes their biogenesis, whereas TRC miRNAs appear to evolve under positive selection. Altogether, these data reveal miRNA diversity amongst Drosophila species and permit future discoveries in understanding their emergence and evolution.