Combining Shigella Tn-seq data with gold-standard E. coli gene deletion data suggests rare transitions between essential and non-essential gene functionality

Nikki E Freed,Dirk Bumann,Olin K Silander
DOI: https://doi.org/10.1186/s12866-016-0818-0
2016-09-06
Abstract:Background: Gene essentiality - whether or not a gene is necessary for cell growth - is a fundamental component of gene function. It is not well established how quickly gene essentiality can change, as few studies have compared empirical measures of essentiality between closely related organisms. Results: Here we present the results of a Tn-seq experiment designed to detect essential protein coding genes in the bacterial pathogen Shigella flexneri 2a 2457T on a genome-wide scale. Superficial analysis of this data suggested that 481 protein-coding genes in this Shigella strain are critical for robust cellular growth on rich media. Comparison of this set of genes with a gold-standard data set of essential genes in the closely related Escherichia coli K12 BW25113 revealed that an excessive number of genes appeared essential in Shigella but non-essential in E. coli. Importantly, and in converse to this comparison, we found no genes that were essential in E. coli and non-essential in Shigella, implying that many genes were artefactually inferred as essential in Shigella. Controlling for such artefacts resulted in a much smaller set of discrepant genes. Among these, we identified three sets of functionally related genes, two of which have previously been implicated as critical for Shigella growth, but which are dispensable for E. coli growth. Conclusions: The data presented here highlight the small number of protein coding genes for which we have strong evidence that their essentiality status differs between the closely related bacterial taxa E. coli and Shigella. A set of genes involved in acetate utilization provides a canonical example. These results leave open the possibility of developing strain-specific antibiotic treatments targeting such differentially essential genes, but suggest that such opportunities may be rare in closely related bacteria.
What problem does this paper attempt to address?