Applications of recombineering and recoded genome construction.

Abstract

Genome engineering offers the opportunity to bypass natural selection to generate novel organisms. Although multiple types of genome engineering are currently possible, work described in this dissertation occurred in the context of developing a recoded Escherichia coli strain, rE.coli-57. Here, every instance of the following seven codons were replaced in the 3.97 Mbp MDS42 genome and their corresponding translation recognition machinery removed (termed recoding): two leucine codons (UUG and UUA), two serine codons (AGC and AGU), two arginine codons (AGA and AGG), and one stop codon (UAG). Although recoding rare codons is possible by editing the target genome in vivo, larger numbers of codons currently require researchers to produce genome segments using de novo DNA synthesis and assemble these segments to produce the final genome. However, the unknown impact of specific recoded codons and non-recoding mutations caused by DNA synthesis and replication errors incentivizes researchers testing each synthesized segment’s ability to complement its corresponding wild type locus to identify recoded segments with decreased complementation fitness. In this dissertation, I describe work to repair 21 recoded segments unable to complement wild type deletion. Next-Generation Sequencing detected non-recoding mutations in all 21 recoded segments. By reviewing the size and number of segment mutations, as well as the overlap between mutations and essential genes, I focused repair efforts on ten recoded segments with a total of 22 non-recoding mutations of bp in essential genes. These mutations were repaired using recombineering targeting episomally-expressed recoded segments with the pORTMAGE plasmid system. As recombineering was performed in strains containing two or more loci with high homology to repair oligonucleotides, I optimized four recombineering protocol conditions to increase recombineering efficiency and parallelizability for a single E. coli culture. During my thesis, I also investigated the impact of recoding on strain fitness by recoding antibiotic resistance cassettes and essential genes identified as potentially contributing to decreased fitness during recoded segment complementation. I conclude this dissertation with a discussion on the future of large-scale genome engineering.