Adaptive landscapes in evolving populations of Pseudomonas fluorescens in simple environments

Abstract

The adaptive landscape heuristic can be used to answer the question "how predictable is evolution?" because its topology will impact the repeatability of evolution. In my Masters research I addressed this question in two ways: (1) I reviewed empirical adaptive landscape studies in the fields of directed protein evolution and microbial experimental evolution and (2) I performed a selection experiment to characterize adaptive landscape topology by measuring variance in fitness and metabolic phenotype within and among genetically distinct Pseudomonas fluorescens strains in two environments. Empirical studies have found that protein level landscapes are generally smooth, however, population level landscapes are rugged even in simple environments. Experimentally I found that the pattern of variance in fitness and metabolic phenotype was unique to the selection environment. The response to selection was highly repeatable at the level of fitness, but the underlying genetic routes taken were different for each environment and more variable in xylose than in glucose, suggesting a more rugged underlying landscape. More generally, my research suggests that making statements about the predictability of adaptive evolution at the population level may be challenging and wi11likely depend on the specifics of the environment in which selection occurs.