Dwelling in the gut of herbivores, anaerobic gut fungi (AGF) have evolved several strategies to efficiently degrade unpretreated biomass, including encoding a diverse array of carbohydrate-active enzymes (CAZymes) and enzyme complexes (cellulosomes) in their genomes. Since genetic engineering tools are still immature for AGF, we developed a predictive bioinformatic pipeline to both annotate novel putative CAZymes and validate their activity through heterologous expression. Here, 173 novel fungal proteins from Piromyces finnis associated with biomass degradation were synthesized and expressed in E coli. Of these, 9.8% were identified as being well-expressed and soluble. The functions of these 17 well-expressed proteins were predicted through a comprehensive comparison of both sequence and structure homology with other known proteins. In silico function prediction on 17 proteins demonstrates sequence annotation excels in identifying catalytic domains, while structural annotation is more effective in recognizing dockerin domains that link CAZymes to cellulosomes. By integrating both approaches, we achieve a more comprehensive and accurate prediction of protein functions, which also directs protein engineering efforts. Predicted substrates for each CAZyme were experimentally validated by screening across a variety of cellulose and hemicellulose substrates, with half of the functional predictions verified. Of these proteins, one candidate Celsome_012, shows significant robust activity against beechwood xylan and was selected for further biochemical characterization with optimal activity at 39°C and pH 5.5. Additionally, removing the flexible domains of the protein further enhances its hydrolytic activity. Together, we present the first expression screening of a panel of around 200 AGF CAZymes, identifying those compatible with E. coli overexpression and providing detailed biochemical characterization of a novel fungal xylanase.

Paul Adams

Appearances