Bacterial cellular heterogeneity in inducible gene expression and drug susceptibility

Abstract

Biological processes are inherently noisy, and such stochasticity can result in significant cell to cell variability in isogenic bacterial populations. Functionally, cellular heterogeneity can have significant impact on the survival of the bacterial population in nature, as well as in our ability to understand, predict and control them. Since most of our knowledge in biology comes from ensemble experiments which measure average behaviour, there is a need to design studies that can identify and characterize variability present in the population. In this thesis, cellular heterogeneity and its consequences in bacterial inducible gene expression systems and in bacterial response to antimicrobial treatment was studied, using a combination of complementary bulk and single cell assays. In the first part of the thesis, antimicrobial action on bacterial populations was examined. Bacterial resistance to antibiotics is a major threat to public health today. Significant efforts in research and health policy have been directed towards understanding and combatting this issue. While genetic mechanisms of resistance development are well understood, there are several phenotypic responses such as persistence, tolerance and heteroresistance resulting due to transient variations within sub populations of cells that allow them to survive an antimicrobial stress. Clinically, these phenomena have been linked to recurrence of various infections such as tuberculosis, cystic fibrosis and urinary tract infections. To overcome these issues, several alternatives to antibiotics are being tested, including antimicrobial peptides (AMPs). These molecules are part of the innate immune system, known for their broad-spectrum activity and immunomodulatory effects. While they have shown promising results in combatting resistant bacteria, the efficacy of these peptides on persistent and heteroresistant cells haven’t been thoroughly investigated. With this background, the action of antibiotics and AMPs on bacterial populations was investigated for their ability to kill such phenotypically different cells. Significant difference was observed in the efficacy of these antimicrobials, when tested on actively growing and growth arrested E. coli. While, both are equally efficient on exponentially growing cells, diverse responses were observed with stationary phase cells. Variability existed even amongst the peptides. Particularly, colistin was found to induce heteroresistant behaviour. This observation was further studied and found to have link with hypermutations in the bacterial genome. In addition, cross resistance to other antimicrobials was also observed. Finally, single cell imaging of this heteroresistant population confirmed the variability in response of individual cells to colistin treatment. Overall, deeper understanding of the interaction between bacterial cells and these antimicrobials was obtained, which can form a basis for designing efficient treatment strategies in the future. In the second part of the thesis, heterogeneity of gene expression from the inducible arabinose operon in E. coli was studied. The arabinose operon is known for its graded response to arabinose concentrations at the population level. At the individual cell level, however, it exhibits an all-or-none response due to stochasticity in the numbers of the arabinose import proteins, AraE and AraFGH. By using different mutant strains of E. coli with varying levels of these transporters, the gene expression response from this operon was characterized for its uniformity, tunability and sensitivity to a range of arabinose concentrations. Varying the levels of each transporter was found to influence the system in a different way, and maximising for one property usually led to a trade-off with another property. In addition, analysis of the temporal behaviour of the gene expression process using a fast-degrading green fluorescent protein (GFP) reporter revealed a dependency on the physiological state of the cells. Overall, information obtained from this study will aid in engineering of inducible gene expression systems for better control