Carbohydrate Recognition by Large Multi-Modular Glycoside Hydrolases in Clostridium perfringens

Abstract

Clostridium perfringens is an anaerobic bacterium found ubiquitously in Nature, including as a member of the human gut microbiome, that under normal conditions asymptomatically colonizes the distal gut. However, it is also an opportunistic pathogen, and is the principal cause of gas gangrene and necrotic enteritis, and represents a leading cause of foodborne illnesses. C. perfringens produces a battery of secreted carbohydrate-active enzymes called glycoside hydrolases whose activities suggest that their substrates are complex eukaryotic glycans, such as those found in the mucosal layer of the gut. A striking feature of these enzymes is their large size and extensive modularity. The additional functions conferred by their non-catalytic ancillary modules, which include family 32 carbohydrate binding modules (CBM32), are thought to increase catalytic activity and as such are proposed to play a critical role in glycan degradation. Two of the largest and most architecturally complex of these enzymes in C. perfringens are GH31 and GH84A, which contain multiple CBM32s and several other ancillary modules. Very little structural or functional information is know regarding these enzymes, particularly with respect to glycan recognition, the role of ancillary modules, and overall structural topology. We have employed a multi-disciplinary “dissect-and-build” approach to characterize the molecular determinants of carbohydrate binding by CBM32s, and global glycan recognition by full-length C. perfringens glycoside hydrolases. Various X-ray and NMR structures of CpGH31 and CpGH84A CBM32s, both in apo form and in complex with galactose and N-acetylgalactosamine, reveal several novel modes of recognition of common mucin O-glycan monosaccharides by these modules. These results highlight that the prediction of CBM32 binding partners based primarily on sequence is insufficient, as important residue conformations required for substrate specificity cannot be predicted based solely on amino acid sequence. The full-length solution structure of CpGH84A was determined using SAXS-based techniques. CpGH84A adopts an elongated conformation spanning 320 Å in length, and contains a single central hinge region that facilitates conformational heterogeneity of this enzyme in solution. These results iii provide unprecedented insights into the three-dimensional arrangement of the catalytic and ancillary modules, the latter of which contribute several functions, including carbohydrate recognition and glycan degradation.