Structural diversity of metatranscriptomic cellulases drives lignocellulose bioconversion by black soldier fly

Efficient bioconversion of lignocellulose-rich livestock manure remains a critical challenge in sustainable waste management. Here, we demonstrate that black soldier fly larvae (BSFL) complete the bioconversion of chicken manure within 9 days, achieving 59.9% substrate mass reduction with substantial lignocellulose degradation driven by substrate-derived microbiota. Dynamic microbial succession from Firmicutes-dominated to Bacteroidota-enriched communities, particularly the enrichment of lignocellulolytic specialists such as Fulvivirga and Marinimicrobium, coordinately mediates this process. Temporal metatranscriptomic analysis revealed stage-specific activation of carbohydrate-active enzymes (CAZymes), with glycoside hydrolase family 9 (GH9) cellulases showing strong abundance correlation with cellulose degradation. Phylogenetic and structure-guided functional analysis identified three distinct evolutionary clusters of GH9 cellulases with divergent enzymatic specificities: Cluster I displays endoglucanase-like activity preferring longer cellooligosaccharides, Cluster II exhibits β-glucosidase-associated activity with optimal affinity for cellobiose, and Cluster III shows cellobiohydrolase-like properties with intermediate substrate preferences. Inoculation of engineered Bacillus sp. co-expressing GH9 enzymes from Clusters I and II significantly enhanced cellulose degradation by 30.20% and increased overall bioconversion efficiency by 25.75%, whereas co-expression of Clusters I and III resulted in functional redundancy. These findings reveal functional divergence among sequence-similar GH9 enzymes and establish that targeted enhancement of β-glucosidase-associated activity is rate-limiting for lignocellulose degradation. Our integrative framework combining metatranscriptomics with structure-guided functional prediction provides a powerful strategy for mining lignocellulolytic enzymes from complex microbiomes.