With only a modest, 6-8 nucleotide seed region required for pairing, miRNAs theoretically aren’t too fussy when it comes to which mRNAs they pair up with. But don’t let this apparent “beer goggle” regulation fool you. When it comes down to miRNA binding in vivo, miRNAs might be a little more discriminating than previously suggested.
If miRNA therapies are ever to make it to the bedside, researchers will have to learn more about which mRNAs they target and when. Depending on the algorithm used, bioinformatics approaches suggest that miRNAs could dock with thousands of mRNA targets. As an alternative or complement to in silico methods, many researchers opt to pull down Ago proteins and the associated RNAs by co-immunoprecipitation to see what comes along for the ride. The only problem then is discerning which interactions are direct or indirect.
So, to isolate direct Ago binders (and thereby identify true miRNA/mRNA pairs and binding sites), scientists at the Howard Hughes Medical Institute and Rockefeller University opted to use high-throughput sequencing cross-linking immunoprecipitation (HITS-CLIP), a method they developed recently to study protein-RNA interactions on a genome-wide scale.
With HITS-CLIP, the investigators cross-linked miRNA–Ago and mRNA–Ago complexes in vivo and, after digesting away the riff-raff, sequenced the remaining RNA fragments. Using the HITS-CLIP approach the crew pieced together genome-wide maps of the 20 most abundant miRNAs in mouse brain with low false-positive rates. Along the way, they also noticed that there was a significant amount of miRNAs interacting with their targets outside the usual 3’ UTR, in coding sequences and introns.
For the full scoop (with a special focus on miR-124 binding sites), check out Nature, June 2009.