Defining the constraints on microbial evolution via horizontal gene transfer: Uncovering the roles of protein complexity, function and divergence

Although much of the observed prokaryotic diversity on Earth is the product of incremental accumulation of beneficial mutations with small phenotypic effects, horizontal transfer of whole genes (HGT) has shaped adaptive prokaryotic evolution as well. Understanding the evolutionary and cellular mechanisms governing horizontal gene transfer (HGT) is critical for predicting how bacterial genomes and phenotypes evolve. Experiments were used to investigate the covariance between changes in Escherichia coli (E.coli) fitness caused by horizontal gene transfer (HGT) of exogenously expressed genes and the number and complexity of the protein connectivity of the manipulated genes. Prior work investigating natural HGT among multiple bacterial taxa in combination with network theory, which provides a theoretical framework for characterizing the complexity of protein interaction (protein connectivity), shows that not all genes are transferred between bacteria equally. Surprisingly the underrepresented genes tend to occupy highly connected network positions. The manipulative experiments presented here tested the hypothesis that the protein connectivity in the recipient genome profoundly influences HGT. First, changes in relative fitness were measured in a pooled population created from approximately 4122 E.coli cell lineages, each of which expressed a single different E.coli gene transferred via an HGT. This work showed that the covariance between the protein connectivity complexity and gene transferability was more complicated than previously suggested. While the complexity of protein connectivity was important, the clustering of those interactions and the biological function of the gene also had a significant role. In a subsequent experiment, the role of sequence divergence was included into the analysis via individual fitness measurements for 178 E.coli cell linages each over expressing genetic homologs from Vibrio cholera (89 genes) and Staphylococcus aureous (89 genes). Surprisingly, the role of protein connectivity was insignificant when compared to the role of divergence. Finally, a study of the population dynamics of a large megaplasmid was used to illustrate the patterns and processes governing the acquisition and maintenance of large transfers of genetic material via HGT. In the conclusion, patterns of covariance between cell fitness and protein connectivity are discussed.