Main

In the search for renewable alternatives to fossil fuels, the development of biofuels from plant biomass has received a large amount of attention in recent years. One of the main focuses of the biofuel field has been lignocellulosic plant material, partly owing to its sheer abundance. As this material is mainly composed of cellulose, hemicelluloses and (to a lesser extent) lignin, the viability of industrial-scale conversion of this material is obstructed by the lack of hydrolytic enzymes that can degrade it efficiently. Therefore, investigations are ongoing, both at the individual and the metagenomic scales, into the processes governing plant biomass conversion and into methods for identifying new candidate enzymes.

Anaerobic thermophiles such as the cellulolytic Clostridium thermocellum are of specific interest, as they often display high conversion rates and produce thermostable enzymes. C. thermocellum has been particularly well studied, as it grows optimally at 60 °C and can degrade the crystalline cellulose that is found in lignocellulosic material to ethanol. Dam and colleagues1 have recently analysed the genome of another anaerobic thermophile, Caldicellulosiruptor bescii, which can grow at temperatures up to 90 °C and can break down cellulose and the hemicellulose xylan simultaneously.

The C. bescii genome sequence contains 259 genes that are predicted to be involved in carbohydrate metabolism and transport. Eighty-eight of these are predicted to encode carbohydrate-active enzymes (CAZys), many of which contain various carbohydrate-binding modules (CBMs) that are needed for the breakdown of insoluble polysaccharides. One CAZy-enriched gene cluster that contains a unique set of CBMs is of particular interest. This cluster could potentially encode 13 catalytic modules that would bind to cellulose. Other CAZy activities encoded in this gene cluster suggest that C. bescii is capable of acting against many important components of plant cell walls, including xylan, xyloglucan, pectin and mannan. In order to test the predictions of gene function, Dam et al. performed proteomic analyses on extracellular protein fractions from C. bescii cultures that were grown on cellulose and xylan. Under these conditions, they detected proteins for 84 of the predicted 259 CAZy or carbohydrate transport genes. In addition, 24 proteins that had previously been annotated as conserved or hypothetical were identified as being active specifically against either cellulose or xylan. Further evaluation of these proteins indicated that no recognizable CAZy domains were present, suggesting that novel domains may exist with roles in cellulose and hemicellulose metabolism that have yet to be determined.

Another approach for identifying biomass-degrading genes was taken by Hess and colleagues2, who analysed the microorganisms that adhered to plant biomass incubated in the rumens of two cows. Samples of microorganisms that were firmly adhered to the cellulosic energy crop switchgrass were compared with those from bulk rumen fluid using 16S ribosomal RNA gene sequencing. This preliminary data suggested that a full metagenomic analysis of microorganisms adhered to switchgrass would reveal genes that are enriched in relevant degradative activities. To this end, more than 1.5 billion read pairs were generated on the Illumina paired-end sequencing platform and, following de novo assembly, over 2.5 million ORFs were predicted. Almost 28,000 candidate genes in this set were predicted to encode enzymes containing functional domains with similarity to CBMs or other relevant catalytic domains. A direct comparison with the CAZy database showed that although 2,716 candidate genes had their most significant sequence match to a known CAZy sequence, only 1% of these displayed more than 95% sequence similarity to known CAZy genes, indicating that many of these predicted genes are not yet present in the database. Ninety candidate genes were selected for expression, and the resultant proteins were tested for biochemical activity using ten substrates, including two pre-treated biofuel crops. Of these candidates, 51 displayed enzymatic activity against at least one substrate and, furthermore, proteins with low sequence similarity to a known CAZy were just as likely to be active as those with high similarity. The use of biofuel crops as substrates highlights this metagenomic strategy as a valuable tool in the hunt for new candidate enzymes.

Detailed analysis of specific bacteria sheds light on those, like C. bescii, that can degrade multiple polysaccharides as well as plant biomass. On a broader scale, metagenomic discovery and characterization of new enzymes has the potential to facilitate more efficient pre-treatments or reveal alternative mechanisms of lignocellulosic degradation.