Biomedical Sciences Graduate Program | University of California, San Francisco

RNA-protein interactions are essential for many cellular processes, and we are interested in understanding the molecular details of sequence-specific recognition, how multiprotein complexes are assembled on RNAs, and in designing novel RNA-binding proteins. The structural diversity of RNA molecules makes the study of RNA-protein recognition an interesting but complex problem. Part of the laboratory has been using small peptide-RNA model systems to examine how RNA structure can help guide protein binding to specific sites and applying the principles of recognition to design novel RNA-binding molecules. Another part of the laboratory has been studying how binding specificity is established in the context of multiprotein complexes, using transcriptional activation by HIV Tat and recognition of 3Õ splice sites by the splicing factor SF1/mBBP as model systems.

In one family of RNA-binding proteins, which includes the Tat and Rev regulatory proteins from HIV, short arginine-rich RNA-binding domains are used to recognize specific RNA hairpins. Interestingly, the peptides can adopt different conformations - a-helices, b-ribbons, and extended chains - depending on the shape of the RNA site recognized. In all cases studied so far, specific interactions occur in the RNA major groove at architecturally distinct structures formed by bulges and loops. These studies show how RNA structure can provide a framework for peptide folding, much as the surrounding framework of a protein can stabilize the secondary structure of a peptide. Indeed, the RNA-binding domain of one viral Tat protein can recognize two different RNA sites, adopting a different conformation when bound to each site. Interestingly, in one of these cases a cellular protein, cyclin T1, assembles on the RNA along with Tat whereas in the other case it does not. We also have shown that the structure of an RNA-binding peptide can be stabilized by placing it within a larger protein context, in this case engineering the a-helix from Rev into a zinc finger framework, and that the stabilized helix binds with high affinity to the Rev response element (RRE). By constructing libraries of zinc finger variants, it is now possible to identify RNA-binding peptides that bind specifically to other RNA sites.

We have developed reporter systems in bacteria and mammalian cells to identify RNA-binding molecules from cDNA and combinatorial libraries. Using these tools, we have been able to design peptides that bind to the HIV RRE with affinities higher than Rev itself, and in some cases the peptides adopt different conformations than Rev. These molecules efficiently inhibit the function of Rev in transporting viral RNAs from the nucleus to the cytoplasm. The mammalian reporter system also is being used to identify novel cellular RNA-binding proteins involved in RNA transport and recognition of 3Õ splice sites. We are able to assemble large multiprotein complexes on these reporters in vivo and to dissect specific RNA-protein interactions in these contexts. We are interested in understanding how splicing complexes may be directed to alternative splice sites, a common form of regulation in mammalian cells, and in particular, how different proteins may be used to direct the branchpoint binding protein, SF1/mBBP, to alternative 3Õ splice sites.

To learn more about how amino acids are used to recognize specific RNA tertiary structures and to help in designing new RNA-binding molecules, we are developing a computer algorithm to model RNA structure and interactions with proteins. We have constructed databases of all possible hydrogen-bonding interactions between amino acids and bases, and between base pairs and base triples. These databases are being used to analyze interactions found in known structures of RNA-protein and DNA-protein complexes and to identify unique small molecule binding sites in RNAs. We anticipate that our studies of RNA-protein interactions will help us better understand the recognition process per se and will provide insight into important cellular RNA processing pathways.