Glycopeptide inclusion improves sequence coverage.
Addition of EThcD to facilitate the detection and identification of glycopeptides during HDX-MS significantly improved sequence coverage and conformational dynamic analysis of SARS-CoV-2 614G spike protein (Figure S2). Glycopeptides were confidently identified at 9 out of the 22 potential occupied N-glycosylation sequons per monomer in spike 614G variant (Table S1, 41% sequon coverage and 84% overall coverage) after manual data processing. At 8 N-glycosylation sequons with confident peptide coverage only glycopeptides were identified, with the exception, N1074, covered by one non-glycosylated peptide (Figure 1C). Therefore, without the inclusion of EThcD for glycopeptide detection and identification, the sequon coverage would have dropped to 1 (5%) and overall spike coverage to 76% (Figures 1 & S2). Additionally, because nine N1074 glycopeptides had HexNAc(2)Hex(5) and HexNAc(2)Hex(6) glycans (Figure 1), the detection of only the non-glycosylated peptide may not represent the true conformational dynamics at N1074.
On-line digestion with nepenthesin-2 and pepsin allows analysis of deuterated peptides in acidified conditions, but the non-specific nature of digestion, along with the need for rapid digestion and LC separation, brings concerns of lowered signal per peptide, especially when layered with potential glycan structural diversity at each sequon. Despite these potential issues with sensitivity, N-glycosylation sequon-specific combinations of peptide and glycan diversity were observed (Figure 1). Seven glycopeptides (one in 2+ and 3+ charge states) with unique amino acid sequences including N234 displayed only one glycan type HexNAc(2)Hex(9). In contrast, 13 unique glycopeptide amino acid sequences including N603 displayed five or more glycan types, while the only glycopeptide amino acid sequence including N1134 displayed ten glycan types. Table S1 summarizes N-glycosylation sequon coverage and occupancy. All glycan groups at the 9 sequons with coverage were of similar types (oligomannose, complex, and hybrid) as described in previous publications on spike microheterogeneity (34, 39, 46).
Evidence for deuterated glycopeptides is shown in Figures 2 to 4 using glycopeptide 597-607 HexNAc(2)Hex(5) as an example. We obtained high-confidence (Byonic score = 508.5) MS/MS identification (Figure 2), reproducible extracted ion chromatograms (Figure 3A), and high-quality isotopic envelopes (Figure 3B). We observed that each unique amino acid sequence for a glycopeptide showed consistent uptake plots in both control and heat denatured states although they had different N-glycan groups. For example, glycopeptides with sequence597VITPGTN TSNQ607 had highly consistent deuterium uptake (Figure 4, left peptide) even though 5 different glycan groups were identified on N603. However, overlay of composite uptake plots for all twelve N603 overlapping glycopeptides (Figure 4) indicated that unique amino acid sequences had different uptakes, and increasing glycopeptide length affected estimated deuteration. For example, C-terminally extending the above 597-607 glycopeptide by 2 amino acids decreased uptake in the control state but increased uptake after heat-treatment (Figure 4, center peptide). Our interpretation was that glycopeptide amino acid sequence at sequon N603 was a stronger determinant of deuteration than N-glycan identity.
Supporting this interpretation, glycopeptides covering sequons N61 and N234 showed two levels of deuteration in the heat denatured state on composite uptake plots, also associated with the length of glycopeptides. For example, the N61 sequon’s proximity to the sequence66HAIH69 indicated that inclusion of this histidine-rich “cool” spot of deuteration decreased the percentage labeling of 7 longer glycopeptides ~15% after heat treatment compared to 4 overlapping but shorter glycopeptides that did not include the 66HAIH69sequence (Figure S3). This result is consistent with the slow deuteration of His residues (47, 48), and amino acid sequence being a stronger determinant of deuteration than N-glycan identity.