UC Berkeley

Global methods for assaying translation have greatly improved our understanding of the protein-coding capacity of the genome. In particular, it is now possible to perform genome-wide and condition-specific identification of translation initiation sites through modified ribosome profiling methods that selectively capture initiating ribosomes (TIS- profiling). Here we apply this approach to meiotic and mitotic timepoints in budding yeast and show the surprising diversity of protein products that can be revealed by such methods.Chapter 2 describes our use of TIS-profiling to globally annotate translation initiation sites in yeast, with a particular focus on alternative N-terminally extended protein isoforms, which initiate from near-cognate (non-AUG) start codons upstream of annotated AUG start codons. We identified 149 genes with an extended isoform and show that these isoforms are produced in concert with canonical isoforms and are translated with high specificity, resulting from initiation at only a small subset of possible start codons. The non-AUG initiation driving their production is enriched during meiosis and induced by low eIF5A, which is seen in this context. Despite our success in identifying extended protein isoforms, high-confidence identification of new coding regions that entirely overlap annotated coding regions – including those that encode truncated protein isoforms – remained challenging. As described in Chapter 3, we developed a sensitive and robust algorithm focused on identifying N-terminally truncated proteins genome-wide, identifying 388 truncated protein isoforms. We performed extensive experimental validation of these truncated proteins and defined two general classes. The first set lack large portions of the annotated protein sequence and tend to be produced from a truncated transcript. We show two such cases, Yap5truncation and Pus1truncation, to have condition-specific regulation and functions that appear distinct from their respective annotated isoforms. The second set of N-terminally truncated proteins lack only a small region of the annotated protein and are less likely to be regulated by an alternative transcript isoform. Many localize to different subcellular compartments than their annotated counterpart, representing a common strategy for achieving dual localization of otherwise functionally identical proteins. Together these findings support the adoption of less static views of gene identity and a broader framework for considering the translational capacity of the genome.