Periodic bottlenecks in experimental antibiotic resistance evolution of Pseudomonas aeruginosa
Over the past decades, the spread of antibiotic resistance among nosocomial bacterial pathogens has developed into a global problem. Population bottlenecks are an important factor for bacterial evolution. Their influence on antibiotic resistance evolution is however not yet fully understood. Bottlenecks are defined as a strong reduction of population size that can lower the population‘s genetic diversity drastically. Population bottlenecks frequently occur in nature and play a significant role in the evolutionary history of populations. Bacterial populations can evolve resistance by various adaptive paths. However, the serial bottlenecks experienced by bacteria both in nature and in experimental evolution influence the direction of adaptation. After surviving a narrow bottleneck, future adaptation is more likely influenced by selective sweeps and periodic selection, rendering the adaptive paths less predictable. In contrast, higher degrees of parallel evolution and clonal interference are expected in case of a wider bottleneck, as higher genetic diversity is likely maintained. In this thesis, I validated the influence of different bottleneck sizes at different levels of selectivity on the evolvability of resistance in populations of the pathogenic bacterium Pseudomonas aeruginosa (subclone PA14). Three different evolution experiments were performed to simulate single drug treatments with carbenicillin (beta-lactam), ciprofloxacin (quinolone) and gentamicin (aminoglycoside) against PA14, for approximately 100 generations. While high inhibitory concentrations selected for the highest resistance under large transfer sizes, the highest resistance in low inhibitory concentrations populations emerged when the transfer size was small. These different dynamics are reflected by mutational patterns in the evolving bacterial genomes. Even though the total number of mutations per population for each treatment depended on the treatment drug, the diversity of the most frequent mutations at the final growth season was higher for small transfer sizes than for large transfer sizes. Surprisingly, only few mutations have completely fixed by the final transfer. These results may indicate that clonal interference of de novo mutations occurs regularly at sub-inhibitory drug concentrations. Overall, my data suggests that bottlenecks, in combination with antibiotic-induced selective pressure, can be a key determinant of resistance evolution and can shape genetic diversity within and between populations.