Natural selection drives adaptive evolution by spreading beneficial mutations through populations. However, the ability of selection to act on epistatic interactions between mutations at different loci depends on recombination. With high levels of recombination, such as those observed in sexual eukaryotes, genes associate with one another randomly, such that natural selection cannot effectively act on particular interacting combinations but instead acts on the average effect of each gene across all genetic backgrounds. Bacteria do not sexually reproduce but still recombine through a process called homologous recombination that occurs less frequently and involves shorter segments of DNA than recombination in eukaryotes. Within these highly linked bacterial genomes, selection may compete favorably with recombination to promote the spread of beneficially interacting mutations. The goal of this project is to advance our understanding of bacterial evolution by quantifying the ability of selection to act on epistatic fitness effects in bacterial genomes and how this selective process leaves observable signatures in genomic data. Additionally, this project will involve the development of novel genomic analyses to study the evolution of gene networks that likely harbor epistatically interacting mutations.
The first aim of this proposal uses novel population genetic computer simulations to study how epistasis drives bacterial evolution. Since the relative importance epistasis depends on the fitness effects of epistatic interactions compared to individual additive gene effects and recombination, these quantities will be varied across simulations. The sensitivity of bacterial genomes to epistasis will be measured by their tendency to form beneficial combinations of alleles that expand in the population. Once these clones consisting of beneficial combinations of alleles increase in frequency and the population reaches a local fitness optimum, recombination dynamics may change owing to recombinants having lower fitness unless a sufficiently large change enables colonization of another local fitness optimum. A similar simulation framework will be employed to study these recombination dynamics between populations at different fitness optima and quantify how the interplay between selection and recombination may create heterogeneity in observed patterns of homologous recombination, in terms of both the observed rate and tract length distributions. These simulations will test the hypothesis that heterogeneity increases with the relative strength of epistatic to additive fitness effects, and thus selection for particular allelic combinations.
The second aim explores the evolution of highly interacting genes that potentially harbor epistatically interacting mutations. Using known interaction networks from well- characterized metabolic genes, novel genomic analyses will be created to study how the structure and connectivity of network interactions explains patterns of genomic variation in the bacterial pathogen Streptococcus pneumoniae. Preliminary analyses indicate that selection may be maintaining certain allele combinations of metabolic genes, as expected under a model of fitness epistasis.
Deeper understanding of fundamental adaptive processes in bacteria will illuminate the origins of biodiversity and improve strategies to contain bacterial pathogens. The proposed research in this fellowship application will advance our understanding of bacterial evolution and adaptation using novel population genetic simulations and innovative DNA sequence analysis techniques. This work will enhance the evolutionary analysis of genomic data from bacterial infectious diseases.
|Arnold, Brian J; Gutmann, Michael U; Grad, Yonatan H et al. (2018) Weak Epistasis May Drive Adaptation in Recombining Bacteria. Genetics 208:1247-1260|