Formalist features determining the tempo and mode of evolution in Pseudomonas fluorescens SBW25 : a thesis submitted in partial fulfilment of the requirements for the degree of Ph.D. in Evolutionary Genetics at Massey University, Auckland, New Zealand

Abstract

In order to explain the adaptive process, it is necessary to understand the generation of heritable phenotypic variation. For much of the history of evolutionary biology, the production of phenotypic variation was believed to be unbiased, and adaptation the primary outcome of selection acting on randomly generated variation (mutation). While true, ‘internal’ features of organisms may also play a role by increasing the rate of mutation at specific loci, or rendering certain genes better able to translate mutation into phenotypic variation. This thesis, using a bacterial model system, demonstrates how these internal features – localised mutation rates and genetic architectures – can influence the production of phenotypic variation. Previous work involving the bacterium Pseudomonas fluorescens SBW25 has shown that mutations at three loci, wsp, aws and mws, can cause the adaptive wrinkly spreader (WS) phenotype. For each locus, the causal mutations are primarily in negative regulators of di-guanylate cyclase (DGC) activity, which readily convert mutation into the WS phenotype. Mutations causing WS at other loci were predicted to arise, but to do so with less frequent types of mutation. The data presented in this thesis confirms this prediction. My work began with the identification and characterisation of a single rare WS-causing mutation: an in-frame deletion that generates a translational fusion of genes fadA and fwsR. The fusion couples a DGC (encoded by fwsR) to a membrane-spanning domain (encoded by fadA) causing relocalisation of the DGC to the cell membrane and the WS phenotype. This is one of the few examples of adaptation caused by gene fusion and protein relocation in a realtime evolution experiment. I next took an experimental evolution approach to isolate further rare WS types and characterized these, revealing a range of rarely taken mutational pathways to WS. Lastly, I describe an example of extreme molecular parallelism, in which a cell chaining phenotype is caused – without exception – by a single nucleotide substitution within the gene nlpD, despite multiple mutational pathways to this phenotype. Characterisation of different nlpD mutants suggests this molecular parallelism is caused by a high local mutation rate, possibly related to the initiation of transcription within this gene.