Sequence features associated with microRNA strand selection in humans and flies

Here, we assess miRNA strand selection bias in humans and fruit flies by analyzing the sequence composition and relative expression levels of the two strands of the precursor duplex in these species. We find that the sequence elements associated with preferential miRNA strand selection and/or rejection differ between the two species. Further, we identify another feature that distinguishes human and fly miRNA processing machinery: the relative accuracy of the Drosha and Dicer enzymes.Our result provides clues to the mechanistic aspects of miRNA strand selection in humans and other mammals. Further, it indicates that human and fly miRNA processing pathways are more distinct than currently recognized. Finally, the observed strand selection determinants are instrumental in the rational design of efficient miRNA-based expression regulators.MicroRNAs are small single-stranded endogenous RNAs, approximately 22 nucleotides in length, which are involved in posttranscriptional gene regulation in a wide variety of species [1-4]. miRNAs function as a component of an RNA-Induced Silencing Complex (RISC) by guiding it to specific targets through base-pairing interaction between the miRNA seed region and a complementary sequence in the 3'-UTR of a target transcript [5,6]. In humans and other animals, the seed region normally extends from the second to eighth positions of mature miRNA [7].All known miRNAs originate from a single-stranded RNA precursor shaped as a hairpin loop structure [8,9]. In animals, the hairpin is excised from a longer precursor by Drosha and subsequently cut by Dicer, which produces an RNA duplex with 3' overhanging ends, each 2 nucleotides long [8,10-13]. This duplex structure is shared between miRNA and small interfering RNA (siRNA) processing pathways [14,15]. For both miRNAs and siRNAs, one of the duplex strands is then incorporated into the RISC and the other eliminated. Since the selected strand determines the functional specificity of an RISC, this is