Circuits of Dynamically Interacting Sigma Factors in Single Cells

Citation

Park, Jin (2018) Circuits of Dynamically Interacting Sigma Factors in Single Cells. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/CT0B-Q853. https://resolver.caltech.edu/CaltechTHESIS:12312017-234413682

Abstract

How do cells integrate multiple, dynamic genetic circuits? I study this question in the context of the alternative sigma factors of B. subtilis.

The first project proposes a novel mode of gene regulation called timesharing. The key idea is that a limited resource is shared dynamically in time. Here we show that the alternative sigma factors of B. subtilis use dynamic sharing to share a limited supply of core RNA Polymerase (RNAP). We show that 5 alternative sigma factors activate in pulses, and that these pulses operate in a competitive regime. Interestingly, we found that pairwise correlations between these sigma factors contained a mixture of positive and negative correlations, whereas one may naively expect all correlations to be negative. We show with a mathematical model that competitive pulsing can lead to non-intuitive sets of mixed correlations.

The second project take a closer, quantitative look at sigma factor competition. Although competition between the housekeeping sigma and a single alternative sigma has been well studied, competition between alternative sigmas themselves has been relatively unexplored. To address this issue, we systematically investigated the pairwise competitive relationships between 7 alternative sigma factors in B. subtilis. The main experimental tool was a 7x7 'deletion' matrix of strains, where every matrix strain was deleted for one sigma, and reported on another sigma via a fluorescent reporter. The deletion matrix revealed that competition is highly asymmetric. Deletion of any given sigma factor increased σ^W activity, but did not affect other sigma factors. These results are recreated by a minimal mathematical model of sigma factor competition, where importantly σ^W is relatively high in abundance but weak in affinity for core RNAP. We used the model to predict how overexpressing sigma factors affect each other, and these predictions were matched by experiments.

The third project reports a novel activator for alternative sigma factors. Alternative sigmas factors are activated by many forms of stress, such as nutrient limitation, temperature shifts, and molecular stresses like antibiotics. Here we show that surprisingly, cell lysis causes adjacent cells to specifically activate σ^X. This cell lysis-σ^X response is a general phenomenon, as it is observed under multiple experimental conditions. We show this relationship between cell death and σ^X is causal, since harvested cell extract activates σ^X. Finally, we hypothesize that cell death and σ^X play an important role in biofilm wrinkle formation.

Thesis Files