Stable and highly active cellulolytic enzymes are essential for the efficient conversion of lignocellulosic biomass into fermentable sugars. Natural cellulolytic systems such as the bovine rumen are known to harbor fibrolytic microbes and represent promising sources of enzymes for biomass degradation. In the project presented here, pyrosequencing has been employed to identify feedstock-targeted enzymes within the transcriptome of rumen microbial communities.

Switchgrass and alfalfa were incubated for 72 hr in the bovine rumen and nucleic acids were extracted from the fiber-associated microbial communities. Based on 16S rRNA sequencing, the microbial community tightly associated with switchgrass differed significantly from that associated with alfalfa, suggesting that distinct sets of organisms are involved in degrading each of these two feedstocks.

Expression profile of the switchgrass-associated organisms was determined by 454-pyrosequencing. We identified 85 highly expressed putative glycosyl hydrolases and 201 unique glycosyl hydrolase transcripts. ~4,000 genes without assigned function were highly expressed and some of them might encode truly novel proteins involved in biomass degradation. We will expand our analysis to expression profiles of rumen microbial communities associated with other biofuel crops.

The results obtained in the course of our project indicate that the fiber-bound microbes are indeed a rich source of putative cellulolytic enzymes that might be useful for large-scale biofuel production. Currently we are developing techniques to capture the full-length sequence of selected transcripts from rumen community DNA. Expressing the recombinant proteins and subjecting them to detailed physicochemical characterization will allow us to verify the sequence-based annotation of the transcript tags.