An Exploration of Fitness Landscapes: Theory and Experiments

Genotypes, phenotypes, and fitness are the ultimate determinants of evolution. The relationship between these three components is collectively referred to as the fitness landscape. Evolutionary biologists have been working to understand the mechanisms and processes governing the fitness landscape since the early 20th century. However, it has proved difficult to unravel due to the tremendous combinatorics of genotypes, and the complex relationships between all components of the landscape. Here I study evolution on the fitness landscape through a combination of modeling and experiments. I identify a paradox within Fisher’s Geometric Model of Adaptation, and relax the model’s assumptions in an effort to solve this problem. I find that restricting the level of pleiotropy and restricting the number of maladapted traits simultaneously solves the paradox, and maintains fits to other experimental data. To complement this modeling, I spend the second two results chapters discussing experimental results. I focus on a case study of genetic divergence in the E. coli lac operon repressor (lacI), and aim to understand the underlying processes and mechanisms that cause divergence. Divergence at this site is due to the historical contingency of mutation fitness effect on epistatic interactions with other substitutions. I then examine the underlying mechanism of hange in lacI mutation fitness effects. I find that the cost of lac expression has decreased in evolved strains, due to an increase in translational capacity. The major benefit of lacI mutations is rendered obsolete, by other mutations that provide a similar growth benefit, and they do not fix.