Grant: Discovery Grants Program - Individual

Protein production is often the rate-limiting step in cell replication. Efficient protein production depends on efficient translation initiation, elongation and termination, as well as mRNA stability. The translation machinery, including rRNA, ribosomal proteins, translation initiation factors, tRNA and its charging enzymes, etc., can translate some mRNAs more efficiently than others. Strong selection operates on organisms, especially rapidly proliferating ones such as Escherichia coli and Saccharomyces cerevisiae, to optimize mRNA and the translation machinery in order to achieve efficient protein production. By contrasting highly expressed and lowly expressed genes, we have been able to detect sequence and structural motifs in prokaryotic and eukaryotic mRNAs that enhance translation efficiency. In this new funding cycle, we will motivate research students to address specific questions in prokaryotic and eukaryotic translation. What features in the 5'UTR of mRNAs increase translation initiation efficiency? S. cerevisiae used to be the only eukaryotic species with large-scale characterization of the m7G-capped mRNA sequences, but 5'UTR sequences have now become available for human genes. Such data would facilitate the identification of sequence and structural motifs in the 5'UTR that enhance translation initiation. How do 5'UTR sequences interact with the downstream coding sequences through base-pairing can also be addressed with the new data. Given the observation that the nucleotide following the stop codon (i.e., the +4 site) strongly affects translation termination efficiency, how is the +4 nucleotide decoded? Is it rRNA, release factor or ribosomal protein or all of them that decode this +4 site? How does this +4 site and its decoders coevolve? Protein structures with mRNA in complex with ribosome and release factors shed light on addressing such questions because one can see which nucleotides in the 30S rRNA or which amino acid residues in release factors or ribosomal proteins are close enough to interact with the +4 nucleotide. There are more than 200 nonhomologous ribosomal proteins in extant species, but only 33 of them are shared among the three domains of life. This implies frequent loss or gain of ribosomal proteins along evolutionary lineages. What types of ribosomal proteins tend to be lost or gained in evolution? How do these loss and gain events alter the interaction among ribosomal proteins and between rRNA and ribosomal proteins? To address these questions, we need to 1) take advantage of the genomic, transcriptomic and proteomic data available in public databases, and 2) develop powerful bioinformatic tools to convert the information to knowledge. My lab has a long tradition of developing powerful but user-friendly and free software packages not only for supporting our own research but also for the wide biology community. Software development represents one key training component in our research program.