Vulnerabilities in coronavirus glycan shields despite extensive glycosylation

Abstract

Severe acute respiratory syndrome (SARS) and Middle East respiratory syndrome (MERS) coronaviruses (CoVs) are zoonotic pathogens with high fatality rates and pandemic potential. Vaccine development focuses on the principal target of the neutralizing humoral immune response, the spike (S) glycoprotein. Coronavirus S proteins are extensively glycosylated, encoding around 66-87 N-linked glycosylation sites per trimeric spike. Here, we reveal a specific area of high glycan density on MERS S that results in the formation of oligomannose-type glycan clusters, which were absent on SARS and HKU1 CoVs. We provide a comparison of the global glycan density of coronavirus spikes with other viral proteins including HIV-1 envelope, Lassa virus glycoprotein complex, and influenza hemagglutinin, where glycosylation plays a known role in shielding immunogenic epitopes. Overall, our data reveal how organisation of glycosylation across class I viral fusion proteins influence not only individual glycan compositions but also the immunological pressure across the protein surface.

Fig. 1. Compositional analysis of SARS, MERS and HKU1 glycans.

a Schematic representation of SARS, MERS and HKU1 coronavirus S glycoproteins, showing the positions of N-linked glycosylation amino-acid sequons (NXS/T, where X ≠ P) shown as branches. The domains of the S glycoproteins are illustrated: N-terminal domain (NTD), receptor-binding domain (RBD), sub-domain 1/2 (SD1/2), fusion peptide (FP), heptad repeat 1/2 (HR1/2), central helix (CH), and transmembrane domain (TM). b HILIC-UPLC chromatograms of fluorescently labelled N-linked glycans from SARS, MERS and HKU1 S. Oligomannose-type glycans (M5 to M9; Man₅GlcNAc₂–Man₉GlcNAc₂) (green) and complex-type glycans (magenta) were identified by Endo H digestion, with quantification of major glycan types summarised as a pie chart. Oligomannose-type glycans are schematically annotated with mannose residues as green circles and GlcNAc residues as blue squares.

Fig. 2. Quantitative site-specific N-linked glycan analysis of SARS and MERS coronavirus spike glycoproteins.

Purified (a) MERS and b SARS S proteins were digested. Quantitative site-specific N-linked glycan analysis of a MERS and b SARS S glycoproteins. Purified S proteins were digested with trypsin, chymotrypsin, alpha-lytic protease, Glu-C, and trypsin plus chymotrypsin, then analysed by LC-ESI MS. Glycan compositions are based on the glycan library generated from negative-ion mass spectrometry of released N-glycans. The bar graphs represent the relative quantities of each glycan group with oligomannose-type glycan series (M9 to M5; Man₉GlcNAc₂ to Man₅GlcNAc₂) (green), afucosylated and fucosylated hybrid glycans (Hybrid & F Hybrid) (dashed pink), and complex glycans grouped according to the number of antennae and presence of core fucosylation (A1 to FA4) (pink). Left to right; least processed to most processed. The pie charts summarise the quantification of these glycans. Additional compositional information regarding the distribution of fucosylation and sialylation can be found in SI Fig. 3.

Fig. 3. Structure-based mapping of N-linked glycans on MERS and SARS S proteins.

The modelling of the experimentally observed glycosylation is illustrated on the prefusion structure of trimeric a MERS S (PDB ID 5X59) and b SARS S (PDB ID 5X58) glycoproteins. Structural-based mapping of N-linked glycans on a MERS and b SARS S proteins. The modelling of the experimentally observed glycosylation is illustrated on the prefusion structure of trimeric MERS S (PDB ID 5X59) and SARS S (PDB ID 5X58) glycoproteins. The glycans are colored according to oligomannose content, as defined by the key. DPP4 receptor-binding sites and ACE2 receptor-binding sites for MERS and SARS, respectively, are indicated in light blue. The S1 and S2 subunits are colored light grey and dark grey, respectively.

Fig. 4. Amino-acid sequence diversification across SARS and MERS spikes.

a Amino-acid diversity in SARS and MERS S gene sequences. Averaged values for each domain are also shown below. b Comparison of dN/dS values between buried and exposed residues across SARS and MERS S (n = 70 and 100 for SARS and MERS, respectively). The error bars correspond to the 95% highest posterior density intervals while the circles indicate mean dN/dS values. c Mapping of the per residue amino-acid diversity shown in A onto the structures of SARS and MERS S (PDB ID 5X58 and 5X59, respectively). S proteins are presented as backbone traces with residues colored according to amino-acid diversity. Residues with elevated diversity are colored in red, and N-linked glycans are presented as white surfaces.

Fig. 5. Comparative cryo-EM analysis of SARS S and HIV-1 Env glycan shields.

a Left panel: Sharpened 3.2-Å-resolution C3-symmetric cryo-EM map of SARS S 2P ectodomain visualized at a high contour level with disordered S1 receptor-binding and N-terminal domains extending out from the central core. Middle panel: Low-pass filtered (lpf) cryo-EM map of the glycoprotein visualised at a low contour level along with a simulated peptide-only map overlaid. Right panel: SPARX 3D variability map. b Same as in (a) but for HIV-1 Env BG505 SOSIP.664 construct in complex with three copies of RM20A3 base-specific Fabs.

Fig. 6. Comparison of the glycan shields of viral class I fusion proteins.

Glycan shield densities were calculated using Proteins, Interfaces, Structures and Assemblies (PISA) analyses of fully glycosylated models of SARS S, MERS S, HKU1 S, LASV GPC, HIV-1 Env (BG505), Influenza H3N2 hemagglutinin (Victoria 2011), SIV Env (PDB ID 5X58, 5X59, 5I08, 5VK2, 4ZMJ, 4O5N, 6OHY, respectively)^,,,–. Oligomannose abundances of viral glycoproteins were ascertained by HILIC-UPLC analysis of PNGase F released N-linked glycans that were fluorescently labelled with procainamide^,, (SI Fig. 5). The number of amino-acid residues interacting with N-linked glycans was divided by the number of solvent-accessible amino-acid residues of the glycoprotein as a measure for global glycan shield density. All viral glycoproteins analysed were expressed as trimers in HEK293F cells apart from LASV GPC, which was derived from virus-like particles from Madin-Darby canine kidney II cells.