(281b) Dynamic Multi-Species Model of Uranium-Bioremediation

Although microorganisms exist naturally in communities, most modeling and physiology studies focus on pure cultures. In order to devise environmental biotechnology strategies for uranium bioremediation of contaminated sites (Lovley, 2006), novel computational and modeling methods are needed. Geobacter species dominate when acetate is injected into uranium-contaminated ground to stimulate the reductive precipitation of this contaminant, but later sulphate reducers regain their dominance and cause the reduction of soluble uranium(VI) to stop (Anderson, 2003). This example outlines the importance of understanding the dynamics of microbial communities for the purpose of designing long-term bioremediation.

To investigate this process, we developed a genome-scale dynamic community model of Rhodoferax ferrireducens and Geobacter sulfurreducens, based dynamic flux balance analysis. The model is compared with field experiment data in order to understand the complex and emergent behaviours of multi-species systems.

Prior to acetate injection, the community composition is modulated by ammonium concentration. If ammonium is not limiting, R. ferrireducens significantly outcompetes G. sulfurreducens due to its higher biomass yield. However, in low ammonium environments, G. sulfurreducens has a small advantage because R. ferrireducens is hampered by the lack of nitrogen fixation. Immediately after acetate injection begins, G. sulfurreducens rapidly outcompetes R. ferrireducens, since the high acetate flux favours the organism with the higher growth rate. Rapid ammonium utilization accompanies the fast growth rate. Once ammonium is exhausted, G. sulfurreducens begins to fix N2 at an energetic cost. Nitrogen fixation leads to higher respiration in G. sulfurreducens, which explains the increased uranium reduction in the field data under low ammonium conditions. Once Fe(III) is exhausted, U(VI) becomes the sole electron acceptor for the iron reducers. However, since the uranium concentration is much lower than the iron concentration, uranium reduction cannot support the biomass concentration achieved under iron reduction, leading to the decay of both organisms. Excess acetate stimulates the growth of other acetate oxidizing organisms, and eventually causes G. sulfurreducens to lose its majority position within the community, and uranium bioremediation is stalled. The model predicted change in community composition between G. sulfurreducens and R. ferrireducens is consistent with field data.

Our model depicts the community metabolic dynamics of uranium-contaminated anoxic subsurface environment. The model explains the dominance of Geobacteraceae over Rhodoferax species during bioremediation, as well as the eventual decrease in uranium removal. The model detects the change in metabolic state of G. sulfurreducens, and explains the increased bioremediation effects in low ammonium environments. This model can be used to predict the potential outcome of a bioremediation strategy, hence aid in the design of such strategies.