Logo: to the web site of Uppsala University

The pan-genome of a species is made up of all gene families that can be included in any individual isolate of the species. Escherichia coli (E. coli) has an open pan-genome including at least 128000 gene families, while only about half of the genes found in each individual isolate are common to all isolates. This indicates a great intraspecies genetic diversity that is not often considered when studying antibiotic resistance. This thesis uses a comparatively large collection of isolates to include more intraspecies genetic diversity and assess its impact on resistance.

One angle of this approach was to study the impact of the pan-genome on spontaneous resistance development. For this, we compared the development of resistance to several antibiotics in a 35-strain collection of E. coli isolates. We found that frequencies of resistant mutants varied greatly between strains, that this variation was largely independent from the initial resistance level of the isolates, and that an isolate’s frequency of mutants for one antibiotic was a poor predictor of the mutant frequencies for other antibiotics. In conclusion, there was a clear impact of genetic diversity on spontaneous antibiotic resistance development. 

Using this approach, we observed a previously undescribed pattern of resistance development for tigecycline, a last-line antibiotic, via amplifications of a known efflux pump. In addition, we found a mutated allele of the pump with a reduced level of induction that did not allow for resistance development through amplifications. We showed that a fitness advantage at low antibiotics concentrations and clonal spread were likely contributing to the high occurrence of the mutated pump among E. coli isolates. While this efflux pump is common and well-studied, the lack of pre-existing knowledge of the mutated allele highlights the value of including many isolates in studies of antibiotic resistance. 

Another angle of this thesis was to determine whether intraspecies genetic diversity also impacts plasmid-borne resistance. For this, we transferred several multiresistance plasmids into a collection of E. coli hosts and characterized the plasmid-host combinations. We observed strain- and plasmid-dependent variations in resistance as well as inconsistencies in the clinical resistance categorization of different hosts with the same plasmid.

In conclusion, this work reveals the impact of intraspecies genetic diversity on the development of antibiotic resistance, both through spontaneous mutations and the acquisition of resistance plasmids, highlighting the need to include intraspecies genetic diversity in studies of antibiotic resistance.