Bacterial conjugation in a heterogeneous world

Abstract

Bacteria frequently contain plasmids, circular pieces of non-chromosomal DNA that can make bacteria resistant to antibiotics or allow plasmids to be transferred to other bacteria in a process called conjugation. Conjugation plays a major role in the transmission of plasmids carrying antibiotic-resistance genes. Previous laboratory experiments and mathematical models of conjugation usually only distinguished between plasmid-free and plasmid-bearing bacteria and considered only a single homogeneous environment. This thesis elaborates this to include multispecies communities and various environments. Chapter 2 is a systematic review and meta-analysis of conjugation frequencies from plasmid-bearing Escherichia coli to various recipient species in liquid broth and filter matings. Decreasing taxonomic relatedness between donor and recipient bacteria was associated with a lower conjugation frequency in liquid broth matings but not in filter matings. Chapter 3 describes the development of mathematical models of conjugation in two connected environments, mimicking the gut lumen and wall, that differ in the conjugation efficiency. Simulations with these models show that conditions at the gut wall should be taken into account when describing plasmid dynamics in the gut. Chapter 4 describes a laboratory experiment in which it was found that products secreted by lactic acid bacteria derived from the caeca of broiler chickens decreased the growth rates of a plasmid-bearing and a plasmid-free Escherichia coli population but did not affect the conjugation rate of a narrow-host-range plasmid carrying an antimicrobial resistance gene. Calculations with a mathematical model showed that, for growth rates and fitness costs estimated from the in vitro experiment, small fitness costs of plasmid carriage can lead to plasmid loss under continuous-flow conditions mimicking the gastrointestinal tract. Chapter 5 describes the development of a mathematical model that included ecological interactions and conjugation between multiple bacterial species. Simulations with this model showed that the intraspecies conjugation rate of the initially plasmid-bearing species and the fitness costs of bearing a plasmid determined the possibility for plasmid invasion mediated by species that were already present in plasmid-free communities. In contrast, ecological interspecies interactions determined the possibility for plasmid invasion mediated by species that were not yet present in plasmid-free communities. In both scenarios, ecological interspecies interactions affected plasmid spread. Chapter 6 describes an experiment where bacteria carrying plasmids that make them carbapenemase-producing, extended-spectrum beta-lactamase-producing or chloramphenicol-resistant were fed to broiler chickens. The bacteria were excreted and spread from the environment to non-inoculated chickens. Antibiotic treatment of the chickens increased the transmission rate more than three-fold in all inoculums. The transmission rates of bacteria with plasmids that make them carbapenemase-producing were 52 – 68% lower than the transmission rates of the bacteria with the other plasmids. Chapter 7 summarizes how the key findings of this thesis relate to the aim of this thesis, discuss which extensions of mathematical models are useful to further increase our understanding of plasmid dynamics, and gives an overview how the mathematical models of, and quantitative data on, conjugation and transmission of bacteria described in this thesis contribute to the fields of plasmid-based antibiotic resistance and plasmid dynamics more generally.