Researchers at New York University’s Center for Comparative Functional Genomics and the University of California, Berkeley have used computational analyses to predict a genome-wide map of microRNA (miRNA) targets in the animal model organism, Caenorhabditis elegans (C. elegans). MicroRNAs bind to messenger RNA (mRNA) in a specific section, called 3’UTR, and are known to regulate them. Parts of the predicted map were confirmed through the development of a novel in vivo method that asked whether the 3’ UTR part of mRNAs was driving regulation during development in a living organism. Their research appears in the most recent issue of Current Biology.
In mapping miRNA targets, the research team examined the function of the genome of C. elegans, the first animal species whose genome was completely sequenced and a model organism to study how embryos develop. Using PicTar, an algorithm developed at NYU, the researchers predicted miRNA functions of C. elegans genes. The researchers found that one-third of C. elegans miRNAs target gene sets have related functions. That is, it appears that miRNAs can control groups of genes that work in a specific biological process. At least 10 percent of C. elegans genes are predicted miRNA targets.
To test the computational predictions, the NYU team developed a new in vivo analysis system comparing the expression of a reporter, green fluorescent protein (GFP) carrying target 3’ UTRs with controls, that did not carry the target 3’UTRs. The laboratory results confirmed the role of specific 3’ UTRs in suppressing gene expression even more widely than predicted by the computational analysis, suggesting that 3’ UTRs contain a largely unexplored universe for gene regulation.
The thousands of genome-wide miRNA target predictions for nematodes, or roundworms, humans, and flies are available from the PicTar website (pictar.bio.nyu.edu) and are linked to a new graphical network-browsing tool developed in the NYU Center for Comparative Functional Genomics. This allows for exploration of miRNA target predictions in the context of various functional genomic data resources (gnetbrowse.org).
The research was supported by grants from the National Science Foundation and the National Institutes of Health.