This exciting cross-disciplinary project in computational biology and microbiology between the University of Strathclyde and the James Hutton Institute/James Hutton Limited aims to understand the host-specific carbohydrate-active enzyme (CAZyme) arsenals of bacterial plant pathogens using computational and experimental approaches to reveal fundamental rules of life governing how CAZymes combine and co-operate to degrade complex substrates.

Knowing these compositional rules will provide a framework to generate new combinations of co-ordinated CAZyme activities to degrade specified targets in an engineering biology context. A resulting ability to design bespoke industrial enzyme mixes for broad or targeted biomass degradation would have significant economic, environmental, and societal impact, as CAZymes are currently exploited in industrial and engineering biology for biofuel production, textile processing, and the production of prebiotics, and have potential wide-ranging impacts in, e.g. generation of more potent biopharmaceuticals, and more efficient waste valorisation.

This project will investigate co-infection using a range of Pectobacterium (Pba, Ppar) strains using cutting-edge computational and experimental techniques, including:

• sequencing of culture collections: the Hutton bacteria culture collection holds many sequenced Pba strains. New Ppar strains from the collection will be isolated and sequenced
• bioinformatics: CAZymes will be identified and catalogued across all available Pectobacteriacae genomes (e.g. cazy_webscraper, dbCAN)
• CAZyme classification & AI: state-of-the-art techniques will be used to classify (e.g. ProteinBERT) and predict structures (e.g Simplefold) for CAZymes and CAZyme complements across Pectobacteriaceae
• isolate/enzyme selection: Pba/Ppar genomes with distinctive genomic/virulence features will be selected for practical investigation. Enzymes with particular relevance to industrial application will be prioritised
• functional characterisation: Screening of Pba/Ppar genomes with distinctive genomic/virulence features, and for specific CAZyme mutants with transposon mutant libraries. Construction of defined single/multiple-gene knockouts. Characterisation of mutant phenotypes.

These lines of investigation will be combined to infer rules governing CAZyme combinations and validated activities using statistical and machine learning approaches to learn fundamental rules governing how CAZymes act co-operatively, leading to proposed novel enzyme mixtures with potential industrial application.

While microbial enzymes already convert plant waste into biofuels and commodity chemicals, current approaches rely on trial-and-error combinations with limited efficiency. The industrial challenge is stark: complex plant polysaccharides in waste streams resist degradation, creating major economic and environmental bottlenecks.

Decoding how plant pathogens orchestrate their enzyme arsenals could revolutionize industrial biomass conversion. Plant pathogens have already solved this problem – they precisely adapt their CAZyme combinations to breach specific plant barriers. By unravelling these compositional rules, we aim to create a systematic framework for designing bespoke enzyme cocktails tailored to different waste streams. This could enable targeted, efficient biomass degradation at industrial scales, unlocking significant environmental and economic value from materials currently treated as waste.