Sorghum cellulose hydrolysis by the rumen bacteria, Bacteroides succinogenes

Abstract

Sweet sorghum is one of the most attractive feedstocks for production of energy and chemicals from biomass because of its high yield and carbohydrate content. The goal of this study was to evaluate the utility of the rumen bacteria, Bacteroides succinogenes, for direct conversion of the cellulose fraction of sorghum biomass to organic acids. The study consisted of three phases: (1) substrate screening to examine the effect of sorghum pretreatment by milling and ensiling, (2) fermentation of sorghum under various conditions, and (3) development of a kinetic model describing the growth of B. succinogenes on mixed substrates of glucose and sorghum cellulose. The results from the first phase of the study suggested that neither milling nor ensiling of sorghum rendered the structural components more susceptible to hydrolysis and fermentation by B. succinogenes. Therefore, only the lingnocellulosic fraction of the sorghum biomass was used as substrate for fermentations in the second phase of the study. For the fermentations, mean values of percent conversion of glucose, cellobiose, sorghum cellulose and filter paper were 96, 59, 27 and 45%, respectively. The presence of glucose, at the level occurring naturally in the sorghum variety used, or high concentrations of succinic acid, the major fermentation product, did not have any detrimental effects on cellulose utilization by the organism. It was concluded that differences in cellulose substrate (filter paper or sorghum cellulose) utilization by B succinogenes were probably due to differences in surface characteristics of the substrates. Cell mass yields from the fermentations were 0.30, 0.35, 0.23 and 0.13 g cell/g substrate for glucose, cellobiose, sorghum cellulose and filter paper, respectively. Succinate yields were 0.23, 0.35. 0.42 and 0.37 g/g substrate and acetate yields were 0.07, 0.16, 0.22 and 0.29 g/g substrate for glucose, cellobiose, sorghum cellulose and filter paper, respectively. A mix-substrate model was developed assuming that growth on either substrate was independent of the presence of the other substrate and also independent of product accumulation. The model did not provide a good fit for the data when parameters evaluated from single-substrate fermentations were used; however, it did give reasonably good results when fitted to the data as an empirical model.