Poster abstracts
Poster number 3 submitted by Christina Akirtava
Quantitative analysis of yeast Kozak variants for AUG and near-AUG start codons using a massively parallel reporter assay
Christina Akirtava (Department of Biological Sciences, Carnegie Mellon University), Hunter Kready (Department of Biological Sciences, Carnegie Mellon University), Lauren Nazzaro (Department of Biological Sciences, Carnegie Mellon University), Gemma E. May (Department of Biological Sciences, Carnegie Mellon University), Joel McManus (Department of Biological Sciences, Carnegie Mellon University)
Abstract:
Translation typically initiates at AUG codons surrounded by favorable sequences first discovered by Marilyn Kozak and, hence, termed the “Kozak context” (Kozak, 1984). Interestingly, translation initiation can also occur at near cognate codons (NCCs) that differ from AUG at one nucleotide. Translation from NCCs regulates gene expression during stress and its misregulation has been implicated in neurodegeneration and cancer (Kearse & Wilusz, 2017). While the Kozak context for AUGs and NCCs has been experimentally tested in human cells (Diaz de Arce et al, 2017; Noderer et al., 2014), limited studies have been done in yeast (Cuperus et al., 2017; Dvir et al., 2013). As such, the extent to which Kozak sequence controls translation in yeast remains unclear. We used Fluorescence-Activated Cell Sorting and highthroughput sequencing (Noderer et al., 2014) to assay expression of YFP reporters harboring all Kozak variants surrounding AUG and NCCs in S. cerevisiae. Compared to human cells, we find less initiation at NCCs in yeast grown in rich media, with no NCC context supporting more than 10% of optimal AUG expression. Among AUGs, the Kozak context influences expression over a ~20-fold range, with a strong preference for -3 A. We further tested a leaky scanning model (LSM) using hundreds of reporters with tandem AUG codons in varied Kozak contexts. Surprisingly, tandem AUGs did not increase reporter expression as compared to a single start. However, incorporating the effects of tandem start codons, increased the predictive relationship between Kozak context and initiation from 5% to 35-75%. Finally, using our Kozak strengths in an LSM we predicted translation of endogenous yeast genes and found that Kozak context alone accounts for 9% of the variation in translation efficiency. These results comprehensively define the Kozak context for AUG and NCCs in yeast. While relative Kozak strengths explained much of the expression of our reporter library, our results suggest that other factors limit the rate of translation initiation in vivo.
Keywords: Translation start codon, Kozak