Date of Completion


Embargo Period



iron metabolism, siderophores, Bacillus anthracis, genetic algorithm, flux balance analysis, GAFBA, computational biology, metabolic modeling

Major Advisor

Dr. Ranjan Srivastava

Associate Advisor

Richard Parnas

Associate Advisor

Yu Lei

Associate Advisor

Kenneth Noll

Associate Advisor

Peter Karp

Field of Study

Chemical Engineering


Doctor of Philosophy

Open Access

Open Access


Bacillus anthracis, the causative agent of anthrax, has to be able to adapt to different physical conditions encountered during the beginning and establishment of the infection process inside the host. One of these conditions is the availability of nutrients like iron. B. anthracis secretes siderophores capable of competing with the iron transport/storage proteins of the host. Previous studies have been dedicated to the analysis of these systems. However, the impact of iron availability on B. anthracis metabolism from a global perspective remains as unclear. In this dissertation, I combined an in silico genome-scale metabolic analysis with experimental studies for B. anthracis grown under different iron conditions in order to study the impact of iron availability on the metabolism of the bacterium.

To develop the first genome-scale metabolic model for B. anthracis, I designed and implemented a new methodology for the curation of genome scale metabolic models that combined a genetic algorithm and flux balance analysis (GAFBA). GAFBA finds metabolites whose mass balance constraints cannot be satisfied, because the metabolite is participating in additional reactions that were not taken into account during the reconstruction process. This resulted in an unfeasible solution to the optimization problem. GAFBA facilitates the curation of genome-scale metabolic networks, and aids in fundamental studies in metabolism.

During the experimental studies of the growth of B. anthracis under ironpoor/ rich conditions, the carbon metabolism increased under iron-poor conditions, although this did not directly support growth. Thus, the B. anthracis model was used to characterize the metabolic response of the bacteria under iron-poor/rich conditions. Under iron-poor conditions, a down regulation in the TCA cycle reactions resulted in the reduced uptake of some amino acids. Also, the production of siderophores increased. A higher glucose uptake rate was found to fulfill amino acids requirements for siderophore biosynthesis and growth. For the first time, I found a correlation between the biosynthesis of autoinducer-2, a quorum sensing signal molecule, and the growth of B. anthracis under iron-poor conditions.