Detection of Mutations Affecting Heterogeneously Expressed Phenotypes by Colony Immunoblot and Dedicated Semi-Automated Image Analysis Pipeline

To understand how bacteria evolve and adapt to their environment, it can be relevant to monitor quantitatively phenotypic changes that occur in a population. Single cell level analyses and sorting of mutant cells according to a particular phenotypic readout can constitute efficient strategies. However, when the phenotype of interest is expressed heterogeneously in ancestral isogenic populations of cells, single cell level sorting approaches are not optimal. Phenotypic heterogeneity can for instance make no-expression mutant cells indistinguishable from a subpopulation of wild type cells transiently not expressing the phenotype. The analysis of clonal populations (e.g., isolated colonies), in which the average phenotype is measured, can circumvent this issue. Indeed, no-expression mutants form negative populations while wild type clones form populations in which average expression of the phenotype yields a positive signal. We present here an optimized colony immunoblot protocol and a semi-automated image analysis pipeline (ImageJ macro) allowing for rapid detection and quantification of clones harboring mutations that affect the heterogeneous (i.e., bimodal) expression of the type three secretion system-1 (TTSS-1) in Salmonella enterica serovar Typhimurium. We show that this protocol can efficiently differentiate clones expressing TTSS-1 at various levels in mixed populations. We were able to detect the emergence of hilC mutants in which the proportion of cells expressing TTSS-1 was reduced compared to the ancestor. We could also follow changes in the frequency of different mutants during long-term infections. This demonstrates that our protocol constitutes a tractable approach to assess semi-quantitatively the evolutionary dynamics of heterogeneous phenotypes, such as the expression of virulence genes, in bacterial populations.