Clostridium thermocellum is a fermentative, Gram-positive, thermophile capable of cellulosome-mediated breakdown of hemicellulose and simultaneous biofuels (ethanol and H2) production, and is thus an excellent candidate for consolidated bioprocessing. However, ethanol and/or H2 production yields are below theoretical maxima due to branched product pathways. Biofuel yields may be improved by manipulation of fermentation conditions or implementation of rational metabolic engineering strategies. However, the latter relies on a thorough understanding of gene content, gene product expression, enzyme activity, and intracellular metabolite levels, which can all influence carbon and electron flux. The thesis work represents the first large-scale attempt in combining bioinformatic, enzymatic, proteomic, and culture perturbation approaches to systematically understand these interactions.
C. thermocellum was used to investigate how these parameters affect end-product yields. Enzyme activities involved in conversion of pyruvate to end-products were consistent with end-product profiles and draft genome annotation. NADH and NADPH-dependent alcohol dehydrogenase (ADH) activities were comparable, whereas NADPH-dependent hydrogenase activities were higher than NADH and ferredoxin-dependent hydrogenase activities. While product yields changed in response to exogenous end-product additions, most core fermentative enzyme activities did not, suggesting that these changes may be governed by thermodynamics. The lack of major changes (>2-fold) in expression in response to growth and gas sparging was further confirmed by proteomics and RT-qPCR, respectively, although the latter revealed that ADH expression changes in response to gas sparging.
Improved genome curation allowed refinement of metabolic pathways. A genomic and end-product meta-analysis of ethanol and/or H2 producing fermentative bacteria revealed that presence/absence of genes encoding hydrogenases and aldehyde dehydrogenases/ADHs had the greatest impacts on biofuel yields. However, genome content alone did not necessarily explain end-product yields.
Given that genomic analysis of C. thermocellum revealed the presence of redundant genes encoding enzymes with analogous functions, shotgun and multiple reaction monitoring proteomics was used to refine which proteins are expressed. Absence/low expression of aldehyde dehydrogenase, ferredoxin-dependent hydrogenase and NADH:ferredoxin oxidoreductase suggest that these enzymes may not play a significant role in metabolism. An alternative electron flow pathway is proposed to explain end-product synthesis patterns in response to pyruvate addition or presence of protein inhibitors (CO, hypophosphite).