N-glycan breakdown by bacterial CAZymes

Abstract

The modification of proteins by N-glycans is ubiquitous to most organisms and they have multiple biological functions, including protecting the adjoining protein from degradation and facilitating communication or adhesion between cells, for example. Microbes have evolved CAZymes to deconstruct different types of N-glycans and some of these have been characterised from microbes originating from different niches, both commensals and pathogens. The specificity of these CAZymes provides clues as to how different microbes breakdown these substrates and possibly cross-feed them. Discovery of CAZymes highly specific for N-glycans also provides new tools and options for modifying glycoproteins.

Keywords: CAZymes; N-glycan; human gut microbes.

Figure 1. N-glycan structure guide

(A) The structures of main types of characterised N-glycans. The core pentasaccharide of an N-glycan is outlined with an orange box, the linkages follow a linkage key and annotated with “α” for α-linkages otherwise they are β-linkages. Maturing N-glycans found in the endoplasmic reticulum have three glucose sugars on Arm A (far left). High-mannose N-glycans are composed only of α-linked mannose that have strict linkages between them as shown. Mammalian complex N-glycan is the most heterogeneous of the N-glycans, but broadly speaking their antenna is LacNAc disaccharides. For biantennary structures, these are linked to the arms through β1,2-linkages, but triantennary and tetraantennary have antenna linked through a β1,4-linkage to the α1,3 arm and then a β1,6-linkage to the α1,6 arm, respectively. The antennae are commonly capped with sialic acids, but fucose decoration is also common both on the antenna and α1,6-fucose on the core GlcNAc. There can be variations on this description, such as polyLacNAc repeats and bisecting GlcNAc linked to the core mannose via a β1,4-linkage. Hybrid N-glycans are a mix between high-mannose and mammalian complex N-glycans. Plant N-glycans have a much stricter structure, with a Lewis A epitope as antenna linked to the mannose arms via β1,2-linkages. They also have bisecting β1,2-xylose and α1,3-fucose on the core GlcNAc. Invertebrate N-glycans commonly have both α1,3- and α1,6-fucose decorating the core GlcNAc. Characterised yeast α-mannans have long α1,6-mannan extensions on the α1,3 arm, which are again decorated with more α-mannose, α-galactose, and β-mannose depending on the species [12]. (B) The enzyme activities of GH18, GH85, and PNGase enzymes that remove N-glycans from glycoproteins. (C) Useful glycan structures discussed in the review. (D) Examples of glycan structures adapted from [11] from a variety of organisms displaying nonclassical compositions [54–59].

Figure 2. Characterised PNGase enzymes

This summarises the specificities and limitations of different PNGases described so far.

Figure 3. High-mannose N-glycan breakdown by different bacterial species

This summarises how three different bacterial species breakdown high-mannose N-glycans. This predominantly focusses on the localisation of the different enzyme activities between these species. There is processing of the glycan prior to removal from the glycoprotein and import into the cell in the form of removing mannose from the nonreducing ends of the N-glycan for S. pneumoniae, but for Bifidobacterium longum and B. thetaiotaomicron, there is no processing before GH18 or GH85 activity. These models were adapted from [12,24].

Figure 4. Mammalian complex N-glycan breakdown by different bacterial species

This summarises how two different bacterial species breakdown mammalian complex N-glycans. This predominantly focusses the localisation of the different enzyme activities. There is extracellular processing of these N-glycans by both bacterial species, but in different ways. S. pneumoniae carries out sequential removal of the different sugars by exo-acting CAZymes and the sugars are all imported for use as carbon sources. B. thetaiotaomicron has a two endo-acting CAZymes, targeting different parts of the N-glycan and a sialidase. This also includes the fate of the GH18 glycan–peptide product being cross-fed to other microbes. These models were adapted from [16,24,40,44].

Figure 5. Crystal structures of GH18 family members active on N-glycans

This summarises the structures available for different enzymes and where available different products bound in the active site. The products bound to the individual structures are also summarised with a glycan diagram. The different monosaccharides in the structures are colour-coded in the same way as the diagrams. The glycan always binds with α1,3 arm and the α1,6 arm in the right and left cleft, respectively. The diagnostic motif for GH18 enzymes is DXXDXDXE and the last two D and E are the two catalytic carboxylates, which are coloured red to orient the enzymes, especially those without product bound. Information is provided throughout about the origin of the enzyme, activity, and specificity.

Figure 6. Crystal structures of GH85 family members active on N-glycans

This summarises the structures available for different enzymes and where possible different products bound in the active site. The products bound to the individual structures are also summarised with a glycan diagram. The different monosaccharides in the structures are colour-coded in the same way as the diagrams. The glycan binds with α1,3 arm and the α1,6 arm in the right and left cleft, respectively. Two residues important for catalysis are an asparagine and glutamic acids, which have been coloured red for each enzyme. Information is provided throughout about the origin of the enzyme, activity, and specificity. The active site for EndoA has also been shown from above to emphasise how the protein clasps the N-glycan substrate.