Global epistasis and some side effects of adaptation

Over time, adaptive evolution can lead to profound changes in the phenotypes and behaviors of biological systems but predicting these changes is difficult. One major challenge is that how new mutations alter phenotypes and fitness of organisms often depends on the genetic background in which they arise (G×G interactions or "epistasis"), the environment (G×E interactions), or both (G×G×E interactions). To interrogate these interactions in budding yeast, we measured the effects of about 100 transposon-insertion mutations on growth rate in about 40 "background" strains in six laboratory environments. We found that almost all of these mutations exhibit global epistasis, meaning that their average effects linearly decline with the growth rate of the background strain into which they are introduced. The slopes of these linear trends are largely invariant across environments, and most mutations switch from beneficial to deleterious around the same background growth rate, which we call the "pivot" growth rate. Finally, we investigate how such global epistasis affects the pleiotropic consequences of adaptation to one ("home") environment on growth rate of the organism in other ("non-home") environments. We find that in many non-home environments, growth rate is expected to converge to the pivot growth rate, suggesting that this growth rate is a stable evolutionary attractor.