Long-term adaptive evolution of highly recoded E. coli strains

Advances in genome engineering and DNA synthesis technologies have facilitated the large-scale recoding of bacterial genomes, whose objectives include expanded amino acid incorporation, multi-virus resistance, and biocontainment. Our current understanding, however, of the consequences of specific codon substitutions or the deletion of translation elements, such as release factor 1 (RF1), is limited. Furthermore, highly recoded strains tend to have significantly decreased growth rates. To shed greater light on genome design and to concurrently improve strain fitness, we have performed adaptive evolution of several recoded E. coli strains for more than 1,000 generations in independent replicate populations grown in glucose minimal media. Among these strains is C321.DA, in which all instances of TAG stop codons and RF1 have been substituted and deleted, respectively. The growth rate of C321.DA is nearly half of the growth rate of its parent strain (ECNR2) in minimal media. However, the average growth rates of evolved C321.DA populations are more than four-fold the growth rate of the ancestral C321.DA strain and two-fold the growth rate of ECNR2. Additionally, the growth rates of some independent subpopulations of evolved recoded strains in minimal media approach the growth rate of the parent strain in rich media. Preliminary analysis of next-generation sequencing results on dominant clones from independent generation ~1,100 populations is ongoing. Results from this study will guide current and future recoding efforts in the lab while improving utility of existing strains for downstream applications.