PDF, English - main document Restricted access: Repository staff only until 22 July 2025. Login+Download (32MB) | Terms of use

Abstract

Antimicrobial resistance (AMR) presents a major public health challenge threatening the foundations of modern medicine, with gaps in our understanding of the evolutionary pathways leading to AMR. Through genetic interactions with resistance genes, the genetic background offers a way to understand and modulate AMR evolution. To systematically assess how the genetic background affects the evolution of AMR, I established a high-throughput (HT) experimental evolution protocol that allowed me to grow and track thousands of lineages over time. I used this protocol to profile a genome-wide single-gene knockout (KO) collection of Escherichia coli K12 during treatment with the commonly used cephalosporin cefotaxime. By quantifying the contribution of every non-essential gene to AMR evolvability, I uncovered several genes that delayed or promoted cefotaxime resistance evolution. While several mutants with decreased evolvability (evolvability genes) were involved in peptidoglycan (PG) remodeling and metabolism, mutants with increased evolvability were enriched for phospholipid transport and enterobacterial common antigen biosynthetic process. I uncovered two evolvability genes, dpaA, which cleaves linkages between PG and outer membrane (OM), and ivy, which folds periplasmic proteins and is required for DpaA protein presence. The absence of these evolvability genes blocks evolutionary trajectories towards cefotaxime AMR via epistatic genetic interactions with the known cefotaxime resistance genes marR and ompR. Using HT microscopy, proteomics, and quantitative genetics, I uncovered that LPS biosynthesis and transport are deregulated in marR-resistant mutants, and these phenotypes are exacerbated in the absence of the evolvability gene dpaA. This negative genetic interaction is likely caused by the increased abundance of crosslinks between PG and OM in dpaA KO mutants, which could prevent the compensatory shedding of LPS via OM vesicles and blebs in marR mutants. This epistatic interaction can be relieved in mutants with reduced LPS biosynthesis or decreased abundance of PG-OM crosslinks. Uncovering deregulated LPS biosynthesis and transport in marR mutants leading to a compromised OM barrier function allowed me to identify several drugs that inhibit AMR evolution against cefotaxime, potentially by selectively preventing the evolution of marR mutations. Furthermore, to assess the specificity of evolvability genes, I evolved nine evolvability gene mutants against 12 other antibiotics. I found that all of them delayed antibiotic resistance evolution of multiple antibiotics in a drug-, target- or resistance-dependent manner. Furthermore, I developed a complementary approach using Clustered Regularly Interspaced Short Palindromic Repeats interference (CRISPRi) to measure the global epistasis of antibiotic targets or resistance genes with the KO collection in the absence or presence of an antibiotic. This approach is more scalable and can support the insights gained from the experimental evolution of a genome-wide KO library. I applied this method to probe genome-wide genetic interactions of the antibiotic target and resistance gene DNA topoisomerase IV (parC) and discovered that gene KOs with negative epistasis were enriched for binding DNA. The identified evolvability genes represent potential targets for AMR-inhibiting drugs against multiple antibiotics. They can guide combinatorial strategies with existing drugs to delay resistance evolution, resensitize resistant populations, and improve antimicrobial therapies.