Discovery and Development of Promiscuous O-Glycan Hydrolases for Removal of Intact Sialyl T-Antigen

Abstract

Mucin-type O-glycosylation (O-glycosylation) is a common posttranslational modification that confers distinct biophysical properties to proteins and plays crucial roles in intercellular signaling. Yet, despite the importance of Oglycans, relatively few tools exist for their analysis and modification. In particular, there is a need for enzymes that can cleave the wide range of O-glycan structures found on protein surfaces, to facilitate glycan profiling and editing. Through functional metagenomic screening of the human gut microbiome, we discovered endo-O-glycan hydrolases from CAZy family GH101 that are capable of slowly cleaving the intact sialyl T-antigen trisaccharide (a ubiquitous O-glycan structure in humans) in addition to their primary activity against the T-antigen disaccharide. We then further explored this sequence space through phylogenetic profiling and analysis of representative enzymes, revealing large differences in the levels of this promiscuous activity between enzymes within the family. Through structural and sequence analysis, we identified active site residues that modulate specificity. Through subsequent rational protein engineering, we improved the activity of an enzyme identified by phylogenetic profiling sufficiently that substantial removal of the intact sialyl Tantigen from proteins could be readily achieved. Our best sialyl T-antigen hydrolase mutant, SpGH101 Q868G, is further shown to function on a number of proteins, tissues, and cells. Access to this enzyme opens up improved methodologies for unraveling the glycan code.

In Situ Glycan Analysis and Editing in Living Systems

“…To achieve selective editing of N -glycans, Huang’s group performed subtype-selective “delete” and “insert” operations on cell-surface glycans on the basis of the substrate selectivity of different endoglycosidases and their mutants . Withers’s group discovered that endo- O -glycan hydrolases can selectively cleave O -glycans at the cellular and protein level . Protein- and cell-specific glycan editing helps us to better understand and intervene in the functions of glycans on glycoconjugates and cell conditions.…”

Carving out a Glycoside Hydrolase Active Site for Incorporation into a New Protein Scaffold Using Deep Network Hallucination

“…The endo-α- N -acetylgalactosaminidases (EC 3.2.1.97) of the GH101 enzyme family as classified in the CAZy database () catalyzes the hydrolysis of the mucin O -glycan, Galβ(1–3)­GalNAc. , Despite the more than 1350 amino acid residues large multidomain architecture of the GH101 family, the domain constituting the actual catalytic site is composed of only approximately 300 amino acids, forming a distorted (β/α) 8 TIM-barrel-like fold sharing structural similarity with the GH13 α-amylase family. − The substrate binding pocket of endo-α- N -acetylgalactosaminidases from Streptococcus pneumoniae (EngSP) and Bifidobacterium longum (EngBF) is shaped to complement the Galβ(1–3)­GalNac substrate, whereas specific hydrogen bonds and local conformational changes involving a conserved tryptophan “lid” contribute to an occluded bound state. , The anomeric carbon of Galβ(1–3)­GalNac is positioned in close proximity to the catalytic nucleophile and acid/base that are central to the double-displacement mechanism retaining the stereochemistry of the glycan. − The remaining structural domains presumably play crucial roles in maintaining overall enzyme stability and solubility or in providing functions such as macromolecular substrate recognition. Creating a molecule with a simplified scaffold and reduced molecular size would provide a more manageable GH101 template enzyme for engineering efforts aimed at processing and modifying complex glycans.…”

“…The endo-α- N -acetylgalactosaminidases (EC 3.2.1.97) of the GH101 enzyme family as classified in the CAZy database ( ) catalyzes the hydrolysis of the mucin O -glycan, Galβ(1–3)GalNAc. 15 , 16 Despite the more than 1350 amino acid residues large multidomain architecture of the GH101 family, the domain constituting the actual catalytic site is composed of only approximately 300 amino acids, forming a distorted (β/α) 8 TIM-barrel-like fold sharing structural similarity with the GH13 α-amylase family. 17 − 19 The substrate binding pocket of endo-α- N -acetylgalactosaminidases from Streptococcus pneumoniae (EngSP) and Bifidobacterium longum (EngBF) is shaped to complement the Galβ(1–3)GalNac substrate, whereas specific hydrogen bonds and local conformational changes involving a conserved tryptophan “lid” contribute to an occluded bound state.…”

A Guide to Gene‐Centric Analysis Using TreeSAPP

“…TreeSAPP outputs can be rendered using the Interactive Tree of Life (iTOL) tool (Letunic & Bork, 2019) or with programmatic data visualization libraries to generate beautiful and informative figures. TreeSAPP has recently been used to identify glycoside hydrolase enzymes with novel blood antigen–cleaving activities (Rahfeld, Sim, et al., 2019; Rahfeld, Wardman, et al., 2019; Wardman et al., 2021) and to classify lignin‐depolymerizing enzymes into subclasses (Levy‐Booth et al., 2022).…”