Linking temperature dependence in fitness effects of mutations to thermal niche adaptation

Fitness effects of mutations may generally depend on temperature that influences all rate-limiting biophysical and biochemical processes.  Earlier studies suggested that high temperatures may increase the availability of beneficial mutations (“more beneficial mutations”), or allow beneficial mutations to show stronger fitness effects (“stronger beneficial mutation effects”).  The “more beneficial mutations” scenario would inevitably be associated with increased proportion of conditionally beneficial mutations at higher temperatures.  This in turn predicts that populations in warm environments show faster evolutionary adaptation but suffer fitness loss when faced with cold conditions, and those evolving in cold environments become thermal-niche generalists (“hotter is narrower”).  Under the “stronger beneficial mutation effects” scenario, populations evolving in warm environments would show faster adaptation without fitness costs in cold environments, leading to a “hotter is (universally) better” pattern in thermal niche adaptation.  We tested predictions of the two competing hypotheses using an experimental evolution study in which populations of two model bacterial species, Escherichia coli and Pseudomonas fluorescens, evolved for 2,400 generations at three experimental temperatures.  Results of reciprocal transplant experiments with our P. fluorescens populations were largely consistent with the “hotter is narrower” prediction.  Results from the E. coli populations clearly suggested stronger beneficial mutation effects at higher assay temperatures, but failed to detect faster adaptation in populations evolving in warmer experimental environments (presumably because of limitation in the supply of genetic variation).  Our results suggest that the influence of temperature on mutational effects may provide insight into the patterns of thermal niche adaptation and population diversification across thermal conditions.

Methods
Data collected from an experimental evolution project. 

Eighteen Escherichia coli evolution lines and 18 Pseudomonas fluorescens evolution lines evolved at three temperatures as isolated cultures.  Nine incubators were used in this study, randomly assigned into three blocks.  The three incubators in each block were set 19, 26 or 31°C.  There were two replicate microcosms for each species in every incubator.  Therefore, a total of 18 evolution lines were established for each species (3 blocks × 3 temperatures × 2 within-incubator replicates). 

Population density was measured as colony formation units per ml.

Fitness of each evolved population against it ancestral strain at each of the three temperatures (19, 26 or 31°C) was measured by competition experiment.  Fitness of evolved E. coli populations and their ancestor, all of the strain REL606, relative to a reference strain, REL606Ara+, was measured by head-to-head competition assays.  Fitness of evolved and ancestral P. fluorescens populations, all of the strain SBW25EeZY6KX, was measured against a reference strain SBW25.  Relative fitness of each tested population against the reference strain was estimated from the Malthusian parameters.  Fitness of each evolved population relative to its ancestor (change in fitness during the evolution experiment) was calculated as difference between the two, analogous to the selection coefficient index.