Genome-scale constraint-based metabolic model of Clostridium thermocellum

1. School of Engineering Ethanol Acetate Formate Lactate H2 CO2 Cellulose Cellobiose Fructose CREDIT: DOE Joint Genome Institute http://genome.jgi-psf.org/finished_microbes/cloth/cloth.home.html FluxB FluxA FluxC Genome-scale constraint-based metabolic model of Clostridium thermocellum Chris M. Gowen1,3, Seth B. Roberts1, Stephen S. Fong1,2 1Department of Chemical and Life Science Engineering, Virginia Commonwealth University, Richmond, VA, USA 2Center for the Study of Biological Complexity, Virginia Commonwealth University, Richmond, VA, USA 3Presenting author, contact: gowencm@vcu.edu Motivation Cellulose makes up roughly 60% of the dry weight of all plant biomass on earth and therefore represents an extremely abundant and sustainable feedstock for the production of liquid fuels. All current methods for the biochemical conversion of cellulosic biomass to ethanol for fuel fall into three main categories: Metabolic solution space for growth of C. thermocellum on cellobiose shows tradeoff between H2 and ethanol production Reaction A 2 A + B → C + D Mass balance statement Reactions Flux vector, v A B C D … 0 0 0 0 0 0 v1 . . . . . · Metabolites (X) Cellulase Enzymes Cost prohibitive due to supplemental enzymes and additional process steps Current practice in most pilot plants, enzymes are costly Obviates need for additional enzymes and maximizes process efficiencies 2 Enzymatic Saccharification Distillation Pretreatment Fermentation Typical process Simultaneous Saccharification and Fermentation Consolidated Bioprocessing[1] Any given flux state can be defined as a vector, and the reaction matrix combined with boundary constraints define the borders of a solution space within which the flux state must always fall. 1 S · v = d[X] / dt All reactions available to an organism according to genome annotation and biochemical evidence are compiled in a stoichiometric matrix, S, which is part of a genome-scale mass-balance problem. Cellulase Enzymes Combined with flux balance analysis and suitable boundary constraints, the model is able to closely match experimentally observed fermentation characteristics. For example, researchers have long noted that C. thermocellum can be forced to increase ethanol production by thermodynamically preventing H2 gas production [3]. The three-dimensional depiction of the solution space (above, right) demonstrates that maximum growth rate is achieved with no ethanol production, but as H2 secretion flux is restricted, the maximum growth rate peak shifts towards higher ethanol production. Microorganisms generally pursue (either through evolution or regulation) the maximum growth rate. This heuristic also permits the use of this model as a tool for computational strain design. The graph (below) shows single-gene deletions predicted to improve ethanol secretion at the maximum growth rate. This technique can be used to inform genetic engineering decision making. Enzymatic Saccharification / Fermentation Distillation Pretreatment 3 Flux balance analysis [2] Biological Saccharification / Fermentation Distillation Pretreatment Boundary conditions are set based on observed substrate uptake and byproduct secretion rates. Flux balance analysis is then used to probe the resulting solution space by maximizing a cellular objective such as growth rate within the given constraints. The resulting vector describes the predicted reaction fluxes throughout the model. Metabolic model of Clostridium thermocellum Clostridium thermocellum is one of a number of organisms capable of direct fermentation of cellulose to ethanol, and its cellulolytic system is one of the most efficient known to researchers. This efficiency is achieved partly because C. thermocellum assembles most of its cellulase enyzmes onto an extracellular, cell-associated scaffold. The entire assembly is termed the “cellulosome” and maximizes synergies between the different catalytic mechanisms of its cellulase arsenal and subsequent sugar uptake. Results and Discussion A constraint-based genome-scale metabolic model has been constructed based on the published annotation of Clostridium thermocellum’s genome as well as the incorporation of biochemical observation. The statistics of the resulting model are shown in the table to the right. The model is unique in its incorporation of proteomics data to account for substrate-dependent production of the cellulosome. Single gene deletions for which increased ethanol production is predicted. Comparison of model predictions to experimental observations – C. thermocellum iSR432 was used to simulate growth in multiple conditions. Actual (□) and predicted (▬) reaction flux rates are shown, and predicted fermentation product production rates are shown as ranges as determined by flux variability analysis. For each simulation, the boundary fluxes for cellobiose, acetate, and formate were constrained to match the measured fluxes during (A) chemostat growth on cellobiose and (B) fructose, and (C) batch growth on cellobiose. • References and Acknowledgements • The authors would like to thank J. Paul Brooks for his computational and operations expertise, David Hogsett and Chris Herring for discussions regarding C. thermocellum physiology, Lee Lynd for generously providing C. thermocellum cultures and Stephen Rogers and Evert Holwerda for providing valuable assistance. • Lynd, L. R., Weimer, P. J., Van Zyl, W. H., & Pretorius, I. S. (2002). Microbial cellulose utilization: fundamentals and biotechnology. Microbiology and Molecular Biology Reviews, 66(3), 506-577. • Edwards, J. S., Covert, M., & Palsson, B. Ø. (2002). Metabolic modelling of microbes: the Flux-balance approach. Environmental Microbiology, 4(3), 133-140. • Lamed, R. J., Lobos, J. H., & Su, T. M. (1988). Effects of Stirring and Hydrogen on Fermentation Products of Clostridium thermocellum. Applied and Environmental Microbiology, 54(5), 1216-1221. • Shlomi, T., Cabili, M. N., Herrgard, M. J., Palsson, B. Ø., & Ruppin, E. (2008). Network-Based prediction of human tissue-Specific metabolism. Nature Biotechnology, 26(9), 1003-1010. • Burgard, A. P., Pharkya, P., & Maranas, C. D. (2003). Optknock: a bilevel programming framework for identifying gene knockout strategies for microbial strain optimization. Biotechnology and Bioengineering, 84(6), 647-657. • Fong, S. S., Burgard, A. P., Herring, C. D., Knight, E. M., Blattner, F. R., Maranas, C. D., et al. (2005). In silico design and adaptive evolution ofescherichia coli for production of lactic acid. Biotechnology and Bioengineering, 91(5), 643-648. • Desai, S., Guerinot, M., & Lynd, L. (2004). Cloning of l-Lactate dehydrogenase and elimination of lactic acid production via gene knockout in thermoanaerobacterium saccharolyticum jw/sl-Ys485. Applied Microbiology and Biotechnology, 65(5), 600-605. The broad mixture of fermentation byproducts produced by C. thermocellum is reflective of the fact that it has evolved towards its own ends, rather than for the production of ethanol. We demonstrate here the creation of a genome-scale metabolic model of the metabolism of C. thermocellum and discuss its use for model-guided metabolic engineering to optimize production of ethanol from cellulosic substrates in C. thermocellum. Microbial Genomics 2009