Genomic signatures of optimal growth temperature in the family Colwelliaceae

• Main Author: Gentilhomme, Anais
• Other Authors: Collins, R. Eric, Hennon, Gwenn M.M., Leigh, Mary-Beth, Drown, Devin
• Format: Thesis
• Language: English
• Published: 2020
• Subjects: Proteobacteria; Master of Science in Oceanography; Arctic; Fairbanks; Sea ice; Alaska
• Online Access: http://hdl.handle.net/11122/12401

Thesis (M.S.) University of Alaska Fairbanks, 2020 The temperature range supporting growth defines a complex physiological phenotype that depends on interactions between an organism's genome and its environment. Its implications are widespread since small changes in optimal growth temperature (OGT) can alter an organism's ability to inhabit an ecological niche. Thus, organisms with extreme thermal growth traits (e.g., psychrophilic, with OGT < 15℃, or thermophilic, with OGT 60 -80℃) may be useful for identifying promising targets when searching for life on other planets, as well as predicting population dynamics in a warming Arctic. We performed comparative genomic analyses of bacteria newly isolated from Arctic sea ice that were affiliated with Colwelliaceae, a family of Gammaproteobacteria that contains many psychrophilic strains, to identify genomic factors that might be used to predict OGT. A phylogenomic analysis of 67 public and 39 newly-sequenced strains, was used to construct an updated phylogenetic tree of Colwelliaceae, of which at least two genera were well represented. To augment the previously reported OGTs of 26 strains, we measured growth rates at −1, 4, 11, and 17 ℃ to determine the OGTs of these 39 new strains of Colwelliaceae. We found that growth rates among all isolates were comparable at −1℃, but varied widely above 10 ℃, indicating higher variability in the ability to tolerate warmer temperatures. To analyze the phenotypic differences on a genomic level, we examined indices of amino acid substitutions that have previously been linked with cold adaptation via an increase in protein flexibility. We found that these indices were significantly correlated with OGT at the whole genome level, although the sign of some correlations were opposite of the predicted positive correlation between temperature and the indices. Using these data, we fit a multiple linear regression model for OGT within the Colwelliaceae family that incorporates the three most informative amino acid indices: GRAVY, ...