Phylogeny constrains and environment modulates glycoside hydrolase architecture

Glycoside hydrolases (GHs) are central to microbial carbohydrate degradation and frequently occur as multidomain proteins containing accessory domains involved in substrate binding, localization, and retention. However, the extent to which GH multidomain architectures are shaped by phylogeny versus environmental adaptation remains poorly understood. Here, we analyzed 158,829 GH-containing proteins from 17,108 high-quality metagenome-assembled genomes (MAGs) in the GEM catalog, focusing on major GH families involved in polysaccharide and oligosaccharide degradation. We developed domain-domain adjacency networks (DANs) to quantify multidomain organization and compared network weighted architectural similarity across bacterial genera and environments. GH architectures were highly conserved within genera, with bootstrapped DANs displaying stable network properties and distinct lineage-specific architectural similarities, indicating strong phylogenetic constraints on domain organization. Accessory domains associated with substrate targeting and enzyme retention represented approximately 61% of non-catalytic partner domains, highlighting their importance in GH evolution. Although phylogeny was the dominant determinant of multidomain architecture, environmental effects were detected in cosmopolitan genera. Environmental variation had a limited and lineage-dependent influence on GH abundance but more consistently affected the prevalence of multidomain GHs, particularly within GH3, GH13, and GH18 families. These results support a hierarchical model in which phylogeny establishes a conserved framework of GH organization, while environmental pressures fine-tune multidomain architectures. Our study demonstrates that protein domain architecture represents an important and previously underappreciated dimension of microbial functional adaptation and provides a scalable network-based framework for linking enzyme organization to ecological function.