Principal Investigator David Bartel
Teh Bartel lab was among those to discover the abundance of miRNAs. These tiny endogenous regulatory RNAs are encoded by unusual genes whose primary transcripts form distinctive hairpin structures from which the miRNAs are processed. Each hairpin is processed to generate the mature ~22-nt miRNA, which is then incorporated into a silencing complex, where it can specify post-transcriptional gene repression by base-pairing to messages of protein-coding genes.
We previously developed cloning and computational strategies for miRNA gene discovery in animals and plants, and found, for example, that humans have hundreds of miRNA genes. We have also shown that many miRNAs are present at levels of more than 1,000 molecules per cell, with some exceeding 50,000 molecules per cell, indicating that the miRNAs, together with their associated proteins, are among the more abundant ribonucleoprotein complexes found in animal cells. Furthermore, many of the miRNAs from nematodes are related to those found in humans, suggesting important functions since early metazoan evolution. Nonetheless, some miRNAs, particularly those that are expressed in only a few cells and are less extensively conserved, were not identified by earlier cloning and computational efforts.
To explore more thoroughly the genomics of miRNAs and other endogenous silencing RNAs, we are using high-throughput sequencing methods to obtain millions of small-RNA sequencing reads from model plants and animals. Analyses of Physcomitrella (moss) and Arabidopsis sequences have revealed a layer of miRNA-based control involving miRNAs that are more diverse and evolutionarily fluid than those found previously. In plants we also find that dual miRNA complementary sites trigger the production of double-stranded RNA and endogenous siRNAs—an observation that provides potential insights into the recognition of aberrant RNA by the RNA-silencing machinery. Analysis of nematode and fly sequences has also revealed additional miRNAs, including some that had been missed earlier because their initial processing involves splicing rather than the specialized machinery that generates canonical miRNAs.