NEW YORK (GenomeWeb) – While it has been thought that in general microRNAs only bind to targets that perfectly match their seed regions, newly published data from Thomas Jefferson University researchers suggests that non-canonical binding may be even more widespread.
As such, the miRNA field has potentially been overlooking a significant number of valid miRNA targets as it explores the biological functions of the small, non-coding RNAs, according to Isidore Rigoutsos, director of the Computational Medicine Center at Thomas Jefferson University and senior author of the report.
To date, the majority of efforts to identify the targets of miRNAs have focused on matchups between the miRNAs' seed regions and complementary sequences in the 3' untranslated regions (UTRs) of mRNA based on traditional Watson-Crick base pairing. There have been reports of non-standard interactions in which base pairing is interrupted by bulges, miRNAs bind with targets outside of the 3' UTR, and where targets are not conserved amongst species, but the prevalence of these remains uncertain.
Attempting to address this issue, Rigoutsos and collaborators several years ago developed rna22, a method for identifying miRNA binding sites and their heteroduplexes. Based on results from the computational tool and experimental analyses using luciferase assays, the researchers reported in a 2006 publication that some miRNAs may have up to several thousand targets, these interactions could accommodate non-Watson-Crick interactions in the seed region, and that many binding sites exist in 5' UTRs and DNA coding sequences in addition to 3' UTRs.
Following up on those findings, Rigoutsos and collaborators published a paper in 2012 demonstrating that a variety of miRNA seed region modifications, including combinations of multiple G:U wobbles and mismatches, could be introduced without affecting the stability of miRNA/Argonaute complexes.
The results, they wrote in that paper, "indicate that the spectrum of potential targets for a miRNA can admit a wide-spectrum of seed-less targets and thus substantially differs from what is anticipated by the canonical seed model."
Amid significant advances in CHIP-seq techniques in recent years, Rigoutsos and his team sought to extend their earlier findings by studying the sequence and structural preferences that are present in the seed region of miRNA/target heteroduplexes.
As explained in a paper appearing this month in Scientific Reports, they analyzed 34 independent Argonaute HITS-CLIP datasets from seven different cell types derived from human and mouse. In line with the earlier work, they identified miRNA/target interactions of both the standard Watson-Crick flavor, as well as ones with expanded seed-region formations including various combinations of G:U wobbles and single-nucleotide bulges.
What was unexpected, Rigoutsos said, was that the non-canonical formations accounted for about two-thirds of those identified.
The investigators also discovered that miRNA binding sites are spread across all genomic regions, with a significant portion located beyond the exonic space. At the same time, the location of these targets can vary depending on cell type, even within a species. For instance, 20 to 40 percent of binding sites identified in mouse brain cells were exonic, while that number jumped to around 75 percent in mouse CD4+ T cells. In mouse embryonic stem cells, meanwhile, nearly two-thirds of miRNA targets were located in introns.
The scientists further examined the profiles of Ago-loaded miRNAs and targets. Although they found miRNAs to be "consistently present across the replicates of a given sample," they wrote in Scientific Reports, the profiles of the targets surprisingly "exhibited a more dynamic behavior across the replicates of the same cell type."
Rigoutsos' team also found evidence of a large target repertoire for many miRNAs, which can include hundreds of distinct targets for a given miRNA in the same cellular context, according to the paper. "The number of distinct targets for a given miRNA increases further when one considers the miRNA’s targetome across cell types."
In addition to pointing to broader miRNA targetome than previously anticipated, the findings may have implications for miRNA-effected regulation, the research group noted in its paper.
"The apparent abundance of non-protein-coding miRNA targets in conjunction with the finding that several miRNAs can have many targets and that an mRNA can be targeted by many miRNAs simultaneously provides additional support to the concept of miRNA sequestration and competing endogenous RNAs," it wrote. "The diversity of involved genomic transcripts and the large number of promiscuous miRNAs encountered in [different] cell types indicate that a large number of ways exist in which sequestering of miRNAs by sponges and ceRNAs through target decoying can regulate protein-coding transcripts."