Inclusion of deuterated glycopeptides provides increased sequence coverage in hydrogen/deuterium exchange mass spectrometry analysis of SARS-CoV-2 spike glycoprotein

Abstract

Rationale: Hydrogen/deuterium exchange mass spectrometry (HDX-MS) can provide precise analysis of a protein's conformational dynamics across varied states, such as heat-denatured versus native protein structures, localizing regions that are specifically affected by such conditional changes. Maximizing protein sequence coverage provides high confidence that regions of interest were located by HDX-MS, but one challenge for complete sequence coverage is N-glycosylation sites. The deuteration of peptides post-translationally modified by asparagine-bound glycans (glycopeptides) has not always been identified in previous reports of HDX-MS analyses, causing significant sequence coverage gaps in heavily glycosylated proteins and uncertainty in structural dynamics in many regions throughout a glycoprotein.

Methods: We detected deuterated glycopeptides with a Tribrid Orbitrap Eclipse mass spectrometer performing data-dependent acquisition. An MS scan was used to identify precursor ions; if high-energy collision-induced dissociation MS/MS of the precursor indicated oxonium ions diagnostic for complex glycans, then electron transfer low-energy collision-induced dissociation MS/MS scans of the precursor identified the modified asparagine residue and the glycan's mass. As in traditional HDX-MS, the identified glycopeptides were then analyzed at the MS level in samples labeled with D₂ O.

Results: We report HDX-MS analysis of the SARS-CoV-2 spike protein ectodomain in its trimeric prefusion form, which has 22 predicted N-glycosylation sites per monomer, with and without heat treatment. We identified glycopeptides and calculated their average isotopic mass shifts from deuteration. Inclusion of the deuterated glycopeptides increased sequence coverage of spike ectodomain from 76% to 84%, demonstrated that glycopeptides had been deuterated, and improved confidence in results localizing structural rearrangements.

Conclusion: Inclusion of deuterated glycopeptides improves the analysis of the conformational dynamics of glycoproteins such as viral surface antigens and cellular receptors.

FIGURE 1

Representative N-glycan microheterogeneity during HDX-MS at spike sequons N61, N122, N801, N1074, and N1134. The N-terminus (A), central region (B), and C-terminus (C) of spike are shown including HDX peptide coverage (black horizontal lines), N-glycosylation sequons (green horizontal dash within HDX peptide), signal peptide (gray horizontal box), S2′ cleavage site (vertical red line), fusion peptide (white horizontal box), and the locations of two disulfide bonds in intact spike (orange brackets marked SS). Specific glycan structures detected at each sequon are cartooned as HexNAc (blue squares), hexose (green or yellow circles), fucose (red triangles) and NeuAc (purple diamonds). Glycan structures are illustrative only and were not determined in this study [Color figure can be viewed at wileyonlinelibrary.com]

FIGURE 2

Tandem mass spectrum of glycopeptide 597–607 HexNAc(2)Hex(5) including sequon N603. Sample was unlabeled (0 s D₂O exposure), fragment mass errors are shown in the lower panel. Glycan product ions (green font) from HCD will trigger EThcD MS/MS of a fresh precursor ion packet [Color figure can be viewed at wileyonlinelibrary.com]

FIGURE 3

EICs (A) and isotopic envelopes (B) of glycopeptide 597–607 HexNAc(2)Hex(5) including sequon N603 (red N). Technical triplicate injections at each D₂O exposure time (central numbers) and both control (black) and heat treatment (blue) conditions were injected in randomized order [Color figure can be viewed at wileyonlinelibrary.com]

FIGURE 4

Box and whiskers uptake plots of 12 glycopeptides including sequon N603. The left, center, and right boxes are N = 6, 2, and 4 unique glycopeptides (sequences, glycans, or charge states), respectively. Dashed lines linking boxes indicate the same D₂O exposure time and control or heat denatured state, × indicates mean [Color figure can be viewed at wileyonlinelibrary.com]

FIGURE 5

Isotopologue ratios of glycopeptide 597–607 HexNAc(2)Hex(5) fragment ions (see Figure 1 for a typical MS/MS spectrum of this peptide). Inset: the ratio of M0 (monoisotopic peak) to M1 (first M + 1 isotopic peak) was measured for HexNAc (204/205) and ~y8 (818/819) in a representative technical replicate at each state and D₂O exposure time. As product ions shifted to a deuterated distribution the ratio decreased. N = 3 to 6 averaged MS/MS scans for each bar, error bars are one standard deviation. At H240 and H960 the ~y8 M0 ion was not detected

FIGURE 6

Heat map data (960 s) overlaid on a model based upon PDBs 6VSB and 6VXX, view of central trimer head with stalk at lower left corner. Without heat treatment (A) spike labeled between 0 and 37% indicated by white to red color. Regions with no coverage are green. Taking the difference between heat treatment and control (B) shows both decreased labeling (magenta color, one example is furin site) and increased labeling (yellow color, trimer core is an example). Regions with no coverage are blue [Color figure can be viewed at wileyonlinelibrary.com]