Position-specific codon bias within an essential gene of E. coli

View all posters

Eric Kelsic, Hattie Chung, Harris Wang, Roy Kishony

Harvard Systems Biology Department, United States

Variation of synonymous codons within a gene can strongly affect protein expression levels. Therefore, understanding which codons are optimal for a particular gene is of great importance for synthetic biology, for interpreting patterns of evolutionary conservation and for understanding gene regulation. Previous methods of codon optimization have focused on mimicking global patterns such as the average frequencies of codon usage across the genome, but with limited success. At present, we still lack a detailed understanding of which codons are optimal at varying positions within a gene. Here we present a new method for systematically making single-codon variants of an essential gene in E. coli, and measuring the resulting change in growth rates. Mutant libraries were constructed using Multiplex Automated Genome Engineering (MAGE) transformations. Mutants were pooled together in a competition assay, and the change in frequency of all mutants over time was measured using next-generation sequencing. We systematically measured growth rates of more than 4500 single-codon variants across the entire length of a gene that is essential for initiating translation, infA (IF1). We found that many single-codon synonymous changes have a significant negative effect on growth rate. I will present our analysis of this dataset, describing patterns of global and position-specific codon bias at the level of an individual gene.