The Evolutionary Ecology of Antibiotic Resistance in an Insect Model

Abstract

The fitness consequences of antibiotic resistance are commonly quantified in vitro, and fitness estimated from growth rate measurements or direct competition experiments in drug-free media, overlooking environmental factors within a live infection which may also affect bacterial fitness, such as colonisation rate, transmission potential, or the presence of other interacting microbes. This dissertation aimed to assess the effects of both in vitro and in vivo conditions on bacterial fitness post-acquiring spontaneous resistance mutation and validate earlier studies in nutrient rich media or other laboratory systems in live infections. To achieve this, I utilised the larvae of the diamondback moth, Plutella xylostella, and its natural gut symbiont, the opportunistic pathogen Enterobacter cloacae. This experimental system provided novel means of exploring multiple components of fitness in the presence and absence of antibiotics including within-host competitive ability and between-host transmissibility. First, I generated and characterised a collection of E. cloacae mutants, selected on either cefotaxime, nalidixic acid, or rifampicin and explored the pleiotropic fitness costs associated with resistance. Fitness parameters were quantified in vivo and compared to measurements in vitro. Results suggest that bacterial competitiveness can be environment-dependent, with costs generally enhanced in vivo, highlighting that in vitro measurements may be an unreliable basis for antimicrobial resistance management. Next, I explored the sociality of antibiotic resistance as a potential mechanism explaining the co-occurrence of sensitive and resistant strains under varying demographic and environmental conditions. The relative fitness of the susceptible strain was measured in broth and biofilms in vitro but also in vivo. Cooperative protection in mixed genotype infections allowed the persistence of susceptible bacteria under selection, conferring fitness benefits when susceptible bacteria were rare. Frequency-dependent fitness suggests stable maintenance of both genotypes, despite resistance costs. Overall these results emphasise the value of exploring diverse fitness components of microbes and illustrate how insect models can provide valuable systems for testing and refining hypotheses on the fitness consequences of resistance mutations.