Engineering Escherichia coli as a general genome synthesis platform

Abstract

Writing DNA at genome scale provides unprecedented control in engineering biological systems. This includes the ability to reprogram the genetic code to expand biological functionality and equips biologists with tools to ask previously unanswerable questions on genome organisation and regulation. To date, the synthesis of megabase-sized genomes is costly, operates with custom, laborious methods and requires months to years. A comprehensive suite of methods for the synthesis of genomes at the gigabase scale – such as the human genome – is yet to be developed. Central challenges for genome synthesis are genome design, chemical synthesis of short DNA pieces and the biological assembly of these pieces at the scale of hundreds of kilobases to megabases. In this thesis, I describe methodological advances to accelerate the total synthesis of the Escherichia coli genome and the engineering of E. coli as a chassis for the synthesis of megabase-sized human genome sections. Bacterial conjugation permits the direct transfer of megabase-scale DNA segments between cells while preserving the structural integrity of the DNA, a key step in genome synthesis workflows. However, delivery of large helper plasmids encoding the conjugative function is required to convert E. coli strains to conjugation-competent donor strains. I present orthogonal conjugation, a method that leverages conjugative plasmid orthogonality to enable the efficient and simple delivery of such helper plasmids by conjugation. Total genome synthesis so far has only been realised in prokaryotic systems. To generalise and extend genome writing beyond E. coli to higher organisms with larger genomes at the gigabase scale, new genome synthesis methods are required. For the synthesis of the E. coli genome, small synthetic DNA (1–10 kb) pieces are assembled to bacterial artificial chromosomes (BACs, 50–100 kb), which are then used to replace wildtype sequence in the E. coli genome. Conjugation coupled with programmed excision for enhanced recombination (CONEXER) accelerates the E. coli genome synthesis workflow. CONEXER enables high-fidelity delivery of 100 kb BACs by conjugation coupled directly to recombination, obviating the need for intermediate verification steps. Here, we improved on CONEXER, identifying that deletion of recA in E. coli increases the fidelity of recombination. Extending the principles of CONEXER for large-scale organism-agnostic genome synthesis, we developed BAC stepwise insertion synthesis (BASIS), a method for episome-to-episome recombination for in vivo assembly of megabase-scale constructs in E. coli. Using BASIS, we assembled the full-length 189 kb CFTR gene and a 1.1 Mb section of human chromosome 21 in a BAC – the largest bacterially assembled human genome section to date. BASIS establishes E. coli as a chassis for the synthesis of large human genome sections. BASIS in its first iteration was based on λ-red mediated recombination and incompatible origins of replication of the used BACs which limited its efficiency. We developed BASIS with recBCD-mediated recombination, increasing efficiency 39-fold. We tested variants of compatible origins of replication to uncouple delivery and recombination. These improvements suggest the potential for BASIS as a method to assemble large genome sections of diverse organisms. Addressing biological assembly and synthesis of hundreds of kilobases to megabases, this thesis presents methods that position E. coli as a general platform for genome synthesis – both for the accelerated synthesis of its own genome to expand biological functionality and as a host organism to assemble sections of genomes of diverse organisms.