UC San Diego

The emergence of whole-genome sequencing and other high- throughput experimental technologies in recent years has transformed life sciences research from a relatively data- poor discipline into one that is data-rich. Studies of transcriptional regulatory networks, in particular, have benefited from these developments as high throughput post- genomic data yield system-wide measurements that reveal the components, interactions, and functional states of regulatory systems in response to environmental stimuli. Accordingly, the analysis of transcriptional regulatory networks is the primary topic of this dissertation. To this end, there are dual focal points; the first involves the static modeling of transcriptional regulatory networks, and the second studies how they change during adaptive evolution. The former activity centers on developing a novel approach, called the R matrix formalism, for modeling transcriptional regulatory networks, first, in small model systems, and then by applying the methodology to Escherichia coli. In contrast, the latter activity focuses on investigating how E. coli adapts to environmental perturbations, namely growth on the relatively poor substrates glycerol- and L-lactate-minimal medium, in the laboratory using various post-genomic and traditional experimental techniques. The results presented herein constitute significant advances on several fronts. First, the R-matrix framework for modeling transcriptional regulatory networks, while largely proof-of-concept in the context of this dissertation, will likely prove useful as an integrative method for compiling and analyzing the rapidly growing body of ChIP-chip and transcriptional profiling data, in addition to providing a novel means to study other systems and pathways. Furthermore, the adaptive evolution work highlights the striking malleability of the transcriptional regulatory network when faced with environmental perturbations. Specifically, the RNA polymerase is directly mutated in the glycerol- adaptation experiments, whereas cAMP- and ppGpp-related pathways are targeted in the L-lactate case. Therefore, when taken together, this dissertation provides a novel approach to modeling regulatory networks as static entities, while additionally providing an unprecedented view into the dynamic nature of these networks during adaptation and the critical role that they play in determining the cellular phenotype