Metaproteomics of cellulose methanisation under thermophilic conditions reveals a surprisingly high proteolytic activity

Cellulose is the most abundant biopolymer on Earth. Optimising energy recovery from this renewable but recalcitrant material is a key issue. The metaproteome expressed by thermophilic communities during cellulose anaerobic digestion was investigated in microcosms. By multiplying the analytical replicates (65 protein fractions analysed by MS/MS) and relying solely on public protein databases, more than 500 non-redundant protein functions were identified. The taxonomic community structure as inferred from the metaproteomic data set was in good overall agreement with 16S rRNA gene tag pyrosequencing and fluorescent in situ hybridisation analyses. Numerous functions related to cellulose and hemicellulose hydrolysis and fermentation catalysed by bacteria related to Caldicellulosiruptor spp. and Clostridium thermocellum were retrieved, indicating their key role in the cellulose-degradation process and also suggesting their complementary action. Despite the abundance of acetate as a major fermentation product, key methanogenesis enzymes from the acetoclastic pathway were not detected. In contrast, enzymes from the hydrogenotrophic pathway affiliated to Methanothermobacter were almost exclusively identified for methanogenesis, suggesting a syntrophic acetate oxidation process coupled to hydrogenotrophic methanogenesis. Isotopic analyses confirmed the high dominance of the hydrogenotrophic methanogenesis. Very surprising was the identification of an abundant proteolytic activity from Coprothermobacter proteolyticus strains, probably acting as scavenger and/or predator performing proteolysis and fermentation. Metaproteomics thus appeared as an efficient tool to unravel and characterise metabolic networks as well as ecological interactions during methanisation bioprocesses. More generally, metaproteomics provides direct functional insights at a limited cost, and its attractiveness should increase in the future as sequence databases are growing exponentially.