RSS Feeds

MicroRNA targets: How big is the iceberg?

October 14, 2003

In line with the dogma of molecular biology "DNA makes RNA makes protein", RNA molecules have largely been thought of as intermediaries between the information encoded in the genome and the proteins that do the work. More recently, however, it has become clear that RNAs play far more active roles in most if not all plant and animal species.

We know that small RNAs that do not encode proteins themselves can regulate messenger RNA molecules that do. They bind to their specific targets, which are then degraded rather than translated into protein. Hundreds of the small microRNAs have been identified in plants, worms, flies, mice, and humans, but their specific functions have remained elusive because it has been very difficult to identify their targets.

Up until now, targets were known for only three animal microRNAs, a tiny tip of what most experts expect will be a vast "iceberg" of RNA-regulated genes. Stephen Cohen and colleagues at the European Molecular Biology Laboratory in Heidelberg have taken a bioinformatics approach to identify additional targets in the fruit fly Drosophila. Their method and the resulting lists of predicted targets represent an important step towards our understanding of how microRNAs affect protein composition in animal cells.

Cohen and colleagues started by determining features that were shared by the few previously validated targets of animal microRNAs and then systematically searched for similar features in the genome databases. To reduce the rate of "false-positives", Cohen and colleagues determined whether the target sequences were similar between two closely related Drosophila species, as it is reasonable to assume that "real" microRNA-target pairs would be evolutionary conserved. Where the information was available, they also incorporated information from the partially completed mosquito genome sequence.

This resulted in lists of predicted targets for over 60 Drosophila microRNAs, which the authors ordered such that those targets that they considered most likely to be real were at the top. This strategy correctly identified all previously validated targets from a large database, and puts them near the top of the ordered lists. Moreover, Cohen and colleagues selected six additional predicted targets and were able to experimentally validate all of them.

This not only more than doubles the size of the iceberg's tip, but the fact that Cohen and colleagues publish their full lists and invite other researchers to check whether their favorite genes are likely to be microRNA targets, also suggests that we will soon get a better sense of what's under the water.

Public Library of Science