Bacterial pathogens have evolved intricate mechanisms to evade the human immune system, including the production of immunomodulatory enzymes. Streptococcus pyogenes serotypes secrete two multi-modular endo-β-N-acetylglucosaminidases, EndoS and EndoS2, that specifically deglycosylate the conserved N-glycan at Asn297 on IgG Fc, disabling antibody-mediated effector functions. Amongst thousands of known carbohydrate-active enzymes, EndoS and EndoS2 represent just a handful of enzymes that are specific to the protein portion of the glycoprotein substrate, not just the glycan component. Here, we present the cryoEM structure of EndoS in complex with the IgG1 Fc fragment. In combination with small-angle X-ray scattering, alanine scanning mutagenesis, hydrolytic activity measurements, enzyme kinetics, nuclear magnetic resonance and molecular dynamics analyses, we establish the mechanisms of recognition and specific deglycosylation of IgG antibodies by EndoS and EndoS2. Our results provide a rational basis from which to engineer novel enzymes with antibody and glycan selectivity for clinical and biotechnological applications.
Abstract Bacterial pathogens have evolved diverse and intricate mechanisms to evade the human immune system, including the production of immunomodulatory enzymes. Streptococcus pyogenes secretes two multi-modular endo-β-N-acetylglucosaminidases, EndoS and EndoS2, that specifically deglycosylate the conserved N-linked glycan at Asn297 on IgG Fc, disabling antibody-mediated effector functions. Amongst thousands of known carbohydrate-active enzymes, EndoS and EndoS2 represent just a handful of enzymes that are specific to the protein portion of the glycoprotein substrate, not just the glycan component. Here, we present the cryoEM structure of EndoS in complex with the IgG1 Fc fragment. In combination with small-angle X-ray scattering, alanine scanning mutagenesis, hydrolytic activity measurements, enzyme kinetics, nuclear magnetic resonance and molecular dynamics analysis, we establish the mechanisms of recognition and specific deglycosylation of IgG antibodies by EndoS and EndoS2. Our results provide a rational basis from which to engineer novel enzymes with antibody and glycan selectivity for clinical and biotechnological applications.
Abstract Bacteria produce a remarkably diverse range of glycoside hydrolases to metabolize glycans from the environment as a primary source of nutrients, and to promote the colonization and infection of a host. Here we focus on EndoE, a multi-modular glycoside hydrolase secreted by Enterococcus faecalis , one of the leading causes of healthcare-associated infections. We provide X-ray crystal structures of EndoE, which show an architecture composed of four domains, including GH18 and GH20 glycoside hydrolases connected by two consecutive three α-helical bundles. We determine that the GH20 domain is an exo-β-1,2- N -acetylglucosaminidase, whereas the GH18 domain is an endo-β-1,4- N -acetylglucosaminidase that exclusively processes the central core of complex-type or high-mannose-type N -glycans. Both glycoside hydrolase domains act in a concerted manner to process diverse N -glycans on glycoproteins, including therapeutic IgG antibodies. EndoE combines two enzyme domains with distinct functions and glycan specificities to play a dual role in glycan metabolism and immune evasion.
N-glycosylation is one of the most abundant posttranslational modifications of proteins, essential for many physiological processes, including protein folding, protein stability, oligomerization and aggregation, and molecular recognition events. Defects in the N-glycosylation pathway cause diseases that are classified as congenital disorders of glycosylation. The ability to manipulate protein N-glycosylation is critical not only to our fundamental understanding of biology but also for the development of new drugs for a wide range of human diseases. Chemoenzymatic synthesis using engineered endo-β-N-acetylglucosaminidases (ENGases) has been used extensively to modulate the chemistry of N-glycosylated proteins. However, defining the molecular mechanisms by which ENGases specifically recognize and process N-glycans remains a major challenge. Here we present the X-ray crystal structure of the ENGase EndoBT-3987 from Bacteroides thetaiotaomicron in complex with a hybrid-type glycan product. In combination with alanine scanning mutagenesis, molecular docking calculations and enzymatic activity measurements conducted on a chemically engineered monoclonal antibody substrate unveil two mechanisms for hybrid-type recognition and processing by paradigmatic ENGases. Altogether, the experimental data provide pivotal insight into the molecular mechanism of substrate recognition and specificity for GH18 ENGases and further advance our understanding of chemoenzymatic synthesis and remodeling of homogeneous N-glycan glycoproteins.
Solution-state NMR can be used to study protein–lipid interactions, in particular, the effect that proteins have on lipids. One drawback is that only small assemblies can be studied, and therefore, fast-tumbling bicelles are commonly used. Bicelles contain a lipid bilayer that is solubilized by detergents. A complication is that they are only stable at high concentrations, exceeding the CMC of the detergent. This issue has previously been addressed by introducing a detergent (Cyclosfos-6) with a substantially lower CMC. Here, we developed a set of bicelles using this detergent for studies of membrane-associated mycobacterial proteins, for example, PimA, a key enzyme for bacterial growth. To mimic the lipid composition of mycobacterial membranes, PI, PG, and PC lipids were used. Diffusion NMR was used to characterize the bicelles, and spin relaxation was used to measure the dynamic properties of the lipids. The results suggest that bicelles are formed, although they are smaller than "conventional" bicelles. Moreover, we studied the effect of MTSL-labeled PimA on bicelles containing PI and PC. The paramagnetic label was shown to have a shallow location in the bicelle, affecting the glycerol backbone of the lipids. We foresee that these bicelles will be useful for detailed studies of protein–lipid interactions.