Elucidating Metabolic Variability in Isogenic Microbial Populations Arising due to Noise in Protein Expression

Abstract

Members of a population of isogenic microbes vary in their protein expression states because of stochastic gene expression. Advances in single-cell proteomics on model organisms like E. coli and S. cerevisiae are uncovering this heterogeneity in protein copy numbers. Here we use genome-scale flux balance models to study the effect of heterogeneity in protein expression on metabolic behavior. We predict wide distribution in specific growth rates among the members of the population in accordance with recent single-cell growth rate measurements. Using flux balance analysis along with principal component analysis enables us to identify sub-populations which differ in their metabolic pathway usage. In case of E. coli grown in a minimal medium, we predict presence of slow-growing acetate secreting cells, fast-growing CO2 secreting cells and shifting preference between glycolysis and ED pathway to metabolize glucose. Preliminary population-level measurements support our prediction of acetate secreting cells in aerobic growth conditions. We also find that variability in expression of few genes may be sufficient to capture most of the metabolic variability of the entire population. We extend this study to S. cerevisiae for which the requisite data is becoming available.