2017 ESA Annual Meeting (August 6 -- 11)

PS 6-78 - Bacterial carbon-use- efficiency predicted from genome-specific metabolic models varies phylogenetically and correlates with genome traits

Monday, August 7, 2017
Exhibit Hall, Oregon Convention Center
Mustafa Saifuddin and Adrien C. Finzi, Department of Biology, Boston University, Boston, MA
Background/Question/Methods

Respiration by soil bacteria accounts for one of the largest fluxes of CO2 from the land surface to the atmosphere, yet bacterial physiology and community composition are poorly represented in C cycle models. A single, constant parameter, carbon use efficiency (CUE), is often used in models to determine the partitioning of C between losses as respiration and incorporation into microbial biomass, which can ultimately enter the soil organic C pool. This parameter has a large impact on estimates and projections of soil C pool sizes and greenhouse gas emissions from soil; however, it is not well-understood how CUE varies between bacterial taxa and whether this variation makes microbial community composition an important factor in predicting C cycle fluxes. Metabolic models representing bacterial physiology have previously been used to estimate detailed metabolic fluxes in silico, but this approach has not yet been applied to better understanding CUE. We generated genome-specific metabolic models for over 200 bacterial species using the Department of Energy’s Knowledgebace (kBase) and calculated CUE under varying degrees of resource limitation.

Results/Conclusions

In the absence of resource limitation, CUE varied widely between individual taxa, suggesting that intrinsic variation in CUE between taxa may be as large as that previously attributed to abiotic factors like temperature. Additionally, we observed a significant phylogenetic signal in CUE estimates and variation in these values was significantly related to genome traits, including genome size. For example, CUE declined by 4% per mega base pair (Mbp) increase in genome size. Finally, we explored the degree to which particular metabolites imposed constraint on CUE, and found that CUE responded most strongly to a subset of specific amino acids and carbohydrates, including lysine and arabinose. These findings provide a method for estimating CUE, a critical microbial physiological parameter in C cycle models and for exploring relationships between microbial community composition, bacterial physiology, and C cycling.