Revealing the glycation sites in synthetic neoglycoconjugates formed by conjugation of the antigenic monosaccharide hapten of Vibrio cholerae O1, serotype Ogawa with the BSA protein carrier using LC-ESI-QqTOF-MS/MS

Abstract

In this manuscript, we present the determination of glycation sites in synthetic neoglycoconjugates formed by conjugation of the antigenic monosaccharide hapten of Vibrio cholerae O1 serotype Ogawa to BSA using nano- liquid chromatography electrospray ionization quadrupole time-of-flight tandem mass spectroscopy (LC-ESI-QqTOF-MS/MS). The matrix-assisted laser desorption/ionization-TOF/TOF-MS/MS analyses of the tryptic digests of the glycoconjugates having a hapten:BSA ratio of 4.3:1, 6.6:1 and 13.2:1 revealed only three glycation sites, on the following lysine residues: Lys 235, Lys 437 and Lys 455. Digestion of the neoglycoconjugates with the proteases trypsin and GluC V8 gave complementary structural information and was shown to maximize the number of recognized glycation sites. Here, we report identification of 20, 27 and 33 glycation sites using LC-ESI-QqTOF-MS/MS analysis of a series of synthetic neoglycoconjugates with a hapten:BSA ratio of, respectively, 4.3:1, 6.6:1 and 13.2:1. We also tentatively propose that all the glycated lysine residues are located mainly near the outer surface of the protein.

INTRODUCTION

Glycoconjugate vaccines are obtained by covalent attachment of synthetic or bacterial carbohydrate antigens to carrier proteins. Such conjugates provide T cell-dependent immunogenicity against the saccharide hapten. With the involvement of T cells, immunological memory is invoked, avidity maturation and isotypes switching occurs, to generate complement-activating protective antibodies.^[1]

Different studies have been carried out on the analysis of glycoproteins using mass spectrometry.^[2–4] Novel soft ionization techniques such as electrospray ionization (ESI) and matrix-assisted laser desorption ionization (MALDI) were applied successfully for the determination of the glycation sites of N- and O-glycoproteins, as well as for the general carbohydrate structure determination.^[4,5] Generally, two different methods are applied. The first consists of hydrolysis of the glycans from the protein, followed by a purification step and mass spectrometry analysis of the intact carbohydrate portions.^[6,7] Sometimes, it is necessary to derivatize the released polysaccharides in order to obtain a sufficient m/z signal of the released carbohydrate.^[8] The second approach is based on the entire glycoprotein digestion with endoproteases, followed by mass spectral and tandem mass spectral analyses of the obtained digests.^[9,10] The main advantage of this method is that the product ions formed correspond to the glycopeptide fragments, which contain the exact glycation site.^[11] The determination of glycation sites is useful to verify structures of machine-synthesized large glycopeptides and/or glycoproteins where small-sized glycopeptide intermediates are used as building blocks. Thus, one of the main applications can be the quality control of neoglycoconjugate vaccines. In addition, it will permit the verification of site-specific glycation in future synthetic neoglycoconjugates.

In our previous work, we have reported the determination of the glycation sites of a series of neoglycoconjugate models by MALDI-time-of-flight tandem mass spectroscopy (TOF-MS/MS). These glycoconjugates were composed of the spacer-equipped terminal monosaccharide antigen of the O-specific polysaccharide (O-PS) of Vibrio cholerae O1, serotype Ogawa covalently linked to protein carrier BSA.^[12] We were able to determine the hapten:BSA ratios of the different synthetic glycoconjugate preparations as being 4.3:1, 6.6:1 and 13.2:1. The obtained glycoconjugates were digested with trypsin, and the resulting digests were analyzed by MALDI-TOF/TOF-MS/MS in order to determine the conjugation sites between the carbohydrate and the carrier protein. In that study, we reported that three glycation sites on the Lys 235, Lys 437 and Lys 455 residues were identified on the three analyzed glycoconjugates. We postulated that perhaps selecting the protease trypsin may have not been the correct choice for the digestion of our hapten-BSA glycoconjugates. Additionally, ion suppression effects can occur during the MALDI-TOF-MS experiments and afford a low-quality MALDI-MS spectrum.^[13,14]

We recently reported the determination of the glycation sites in Bacillus anthracis neoglycoconjugate vaccine by MALDI-TOF/TOF-collision-induced dissociation (CID)-MS/MS and liquid chromatography (LC)-ESI- quadrupole (Qq)TOF-MS/MS.^[15] Among other things, it permitted us to evaluate the hapten-to-BSA ratio which was found to be 5.4:1. The carbohydrate-protein vaccine model was then digested using two different proteases: the trypsin and the GluC V8 endoproteinase. The different digests were subsequently subjected to MALDI-TOF/TOF-MS/MS and LC-ESI-QqTOF-MS/MS analysis. The study revealed that the LC-ESI-QqTOF-MS/MS analysis of the different digests allowed identification of a higher number of glycation sites (30 were discovered), when comparing to the MALDI-TOF/TOF-MS/MS analysis of the same digests (only five glycation sites identified). We thus concluded that the LC-ESI-QqTOF-MS/MS analysis of both tryptic and GluC V8 digests was more efficient to reveal the glycation sites on our synthetic carbohydrate-protein glycoconjugates.

In the present work, we applied the foregoing strategy.^[15] Thus, we digested synthetic glycoconjugates with different hapten:BSA ratios (4.3:1, 6.6:1 and 13.2:1), formed by conjugation of the monosaccharide antigen Vibrio cholerae O1 serotype Ogawa to BSA, separately with trypsin and the GluC V8 endoproteinase. The different digests were then analyzed by LC-ESI-QqTOF-MS/MS, and the glycopeptides were sequenced manually in order to determine the glycation sites.

MATERIAL AND METHOD

Preparation of the hapten-BSA neoglycoconjugate

The synthesis of the hapten-BSA glycoconjugates has been described previously.^[16] Briefly, the synthesis started by the preparation of the methyl 6-hydroxyhexanoyl α-glycoside form of the upstream terminal moiety of the O-PS of Vibrio cholerae serogroup O1 serotype Ogawa antigen (4-(3-deoxy-L-glycero)-2-O-methyl-α-D-perosamine). The spacer-equipped sugar was then reacted with 1,2-diaminoethane which converted the methyl ester group in the spacer into the corresponding amino amide. The latter was treated with diethyl squarate to produce squarate monoester. The hapten was finally conjugated with the BSA (Sigma Aldrich, Saint Louis, MO, USA) in 0.5 M borate buffer (pH 9.0) at the initial molar carbohydrate-protein ratio of 100:1. Isolation of the neoglycoconjugates was carried out by ultrafiltration using centrifugal filters with a molecular weight cut-off of 30,000 Da.

Neoglycoconjugates digestion with different proteases

The digestions of the hapten-BSA glycoconjugates were carried out with trypsin and GluC V8 protease (Sigma Aldrich, Saint Louis, MO, USA). Thus, 100 µg of the glycoconjugate was dissolved in a mixture of 0.1% RapiGest SF Surfactant (1 µg, Waters, USA) in 50 mM of NH₄HCO₃ (100 µL) at a pH of 8.0 and reduced by treatment with 2 µl of 10 mM dithiothreitol (Sigma Aldrich, Saint Louis, MO, USA) for 30 min at room temperature, followed by alkylation with 2 µl of a 50 mM iodoacetamide (Sigma Aldrich, Saint Louis, MO, USA) for 1 h at room temperature. A portion (50 µg) of the glycoconjugate was digested with trypsin using a 20 ng/ml of trypsin dissolved in NH₄HCO₃ (50 mM, 1 ml, pH 7.8) at a trypsin:-glycoprotein ratio of 1:25 (w/w) and incubated at 37 °C overnight with shaking. The other 50 µg of the glycoconjugate was digested using the GluC V8 endoprotease dissolved in NH₄HCO₃ (50 mM, 1 ml, pH 7.8) at a protease:glycoprotein ratio of 1:25 (w/w) and incubated at 37 °C overnight with shaking. The sample was then dried under vacuum, and the residue was dissolved in 20 µl of 1% acetic acid (Sigma-Aldrich, Oakville, ON, Canada). An aliquot of each sample (10 µl) was then cleaned up using ZipTip C18 (Millipore, Bedford, MA, USA) before mass spectral analysis.

LC-ESI-QqTOF-CID-MS/MS analysis

The peptides were separated on a DIONEX UltiMate3000 Nano LC System (Germering, Germany). 250 fmol of the digested glycoprotein was dissolved in 0.1% TFA and loaded onto a precolumn (300 µm i.d. × 5 mm, C18 PepMap100, 5 µm (LC Packing, Sunnyvale, CA)) in order to desalt and concentrate the sample. After their elution from the precolumn, the mixture of peptides and glycopeptides was separated on a nanoflow analytical column (75 µm i.d. × 15 cm, C18 PepMap 100, 3 µm, 100 A, (LC Packing, Sunnyvale, CA)) at a flow rate of 180 nl/min. The mobile phase eluents used were composed of 0.1% FA/0.01% TFA/2% ACN (A) and 0.08% FA/0.008% TFA/98% ACN (B). The elution started with 0% B for 10 min, followed by a gradient of 0 → 60% B in 55 min and 60 → 90% B in 3 min and was kept at 90% B for 3 min. The MS/MS analysis of the eluted peptides and glycopeptides was accomplished using an Applied Biosystems API-QSTAR XL QqTOF-MS/MS hybrid tandem mass spectrometer (Applied Biosystems International-MDS Sciex, Foster City, CA, USA) equipped with a nano-electrospray source (Protana XYZ manipulator) which produces the electrospray through a PicoTip needle (10 µm i.d., New Objectives, Wobum, MA, USA) carrying a voltage of 2400 V. The mass spectra were acquired from m/z 100 to m/z 2000.

RESULTS

As previously stated, the purpose of this study was to determine the glycation sites in synthetic hapten-BSA glycoconjugates with hapten-to-BSA ratios: 4.3:1, 6.6:1 and 13.2:1 prepared by conjugation of the antigenic monosaccharide hapten of Vibrio cholerae O1 serotype Ogawa to BSA.

Carbohydrate fragmentation nomenclature used throughout this manuscript to identify the various glycopeptides is that described by Domon and Costello^[4] whereas the nomenclature used to identify the true peptide fragments is the one described by Roepstorff et al. and lately modified by Johnson and coworkers.^[17,18]

Mass spectrometry determination of the glycation sites on the neoglycoconjugate possessing a hapten:BSA ratio of 4.3:1, 6.6:1 and 13.2:1

The following parameters were applied for the MASCOT library searching: carbamidomethyl (C) as fixed modifications, oxidation (M) as variable modifications, a peptide mass tolerance of ± 0.2 Da and a maximum of missed cleavages of 1.

The Nano-LC-ESI-QqTOF-MS/MS analysis of the tryptic digests of the neoglycoconjugates allowed the identification of two serum albumin protein isoforms from the Bos taurus species on the MASCOT library: serum albumin precursor (gi|1351907) with the following sequence coverage: 61% for the neoglycoconjugate possessing a hapten:BSA ratio of 4.3:1, 61% for the neoglycoconjugate with a hapten:BSA ratio of 6.6:1 and 47% for the neoglycoconjugate with a hapten: BSA ratio of 13.2:1. In addition, the MASCOT search also identified for the analysis of all the tryptic digests of the neoglycoconjugates, the serum albumin protein (gi|74267962).

The submission of the LC-ESI-QqTOF-MS/MS data of the GluC V8 digests to the MASCOT library allowed the identification of two serum albumin protein isoforms from the Bos taurus species: serum albumin precursor (gi|1351907, sequence coverage of 55% for the neoglycoconjugate with a hapten:BSA ratio 4.3:1, 45% for the neoglycoconjugate with a hapten:BSA ratio 6.6:1 and 40% for the neoglycoconjugate with a hapten:BSA ratio 13.2:1) and serum albumin (gi|74267962).

We have identified the molecular masses of the glycopeptides produced by the digested glycoconjugates by comparison of the m/z values of the obtained molecular ions which did not correspond to BSA digests in the MASCOT report, with the calculated m/z values of all the possible glycopeptides. Thus, the masses of the glycopeptides were identified by the addition of either one or two carbohydrate-hapten residues (513.2323 Da), corresponding to the m/z values of the formed glycopeptides. In other words, the m/z values of the peptides which matched to the theoretical mass of a ‘peptide + carbohydrate-linker’ were identified, extracted and subjected to low-energy CID-MS/MS for sequencing to identify the glycation sites.

Mass spectrometry determination of the glycation sites on the neoglycoconjugate possessing a hapten: BSA ratio of 4.3:1

The previously described strategy allowed us to identify 11 glycation sites through sequencing of the glycopeptides obtained from the tryptic digests of the neoglycoconjugate (Table 1).

Table 1.

Tryptic glycopeptides identified during LC-ESI-QqTOF-MS/MS analysis of the hapten-BSA glycoconjugate with a hapten:BSA ratio of 4.3:1, 6.6:1 and 13.2:1

Peptide		Missed	Hapten:BSA ratio 4.3:1		Hapten:BSA ratio 6.6:1		Hapten:BSA ratio 13.2:1
Sequence (star = glycation site)	Calculated		Observed	Deviation	Observed	Deviation	Observed	Deviation
	m/z (charge)	cleavage	m/z (charge)	(Da)	m/z (charge)	(Da)	m/z (charge)	(Da)
SLGK*VGTR (Lys 455)	665.8643 (+2)	1	665.8810 (+2)	0.0168	665.8694 (+2)	0.0052	665.8699 (+2)	0.0057
EK*VLTSSAR (Lys 211)	752.3987 (+2)	1	752.4233 (+2)	0.0246	752.4114 (+2)	0.0127	752.3971 (+2)	−0.0016
ALK*AWSVAR (Lys 235)	757.9143 (+2)	1	757.9261 (+2)	0.0118	757.9189 (+2)	0.0046	757.9248 (+2)	0.0105
K*QTALVELLK (Lys 548)	828.4770 (+2)	1	828.4992 (+2)	0.0222	828.4872 (+2)	0.0103	828.4857 (+2)	0.0087
CCTK*PESER (Lys 463)	840.3662 (+2)	1	840.3949 (+2)	0.0287	840.3788 (+2)	0.0126	840.3769 (+2)	0.0107
LK*PDPNTLCDEFKADEK (Lys 140)	845.0720 (+3)	3	845.0908 (+3)	0.0188	845.0839 (+3)	0.0119	845.0911 (+3)	0.0191
DTHK*SEIAHR (Lys 28)	853.9209 (+2)	1	853.9445 (+2)	0.0236	853.9366 (+2)	0.0158	-	-
GACLLPK*IETMR (Lys 204)	951.4892 (+2)	1	951.5091 (+2)	0.0199	951.5014 (+2)	0.0122	951.5006 (+2)	0.0114
K*VPQVSTPTLVEVSR (Lys 437)	1077.0887 (+2)	0	1077.1129 (+2)	0.0243	1077.0985 (+2)	0.0099	1077.0975 (+2)	0.0089
LAK*EYEATLEECCAK (Lys 374)	1164.5347 (+2)	2	1164.5375 (+2)	0.0028	1164.5613 (+2)	0.0266	1164.5540 (+2)	0.0193
QNCDQFEK*LGEYGFQNALIVR (Lys 420)	1014.8220 (+3)	1	1014.8462 (+3)	0.0242	1014.8390 (+3)	0.0170	1014.8534 (+3)	0.0314
HKPK*ATEEQLK (Lys 561)	607.9779 (+3)	3	-	-	607.9966 (+3)	0.0187	608.0033 (+3)	0.0254
LSQK*FPK (Lys 245)	680.8716 (+2)	3	-	-	680.8722 (+2)	0.0007	680.8755 (+2)	0.0040
LCVLHEK*TPVSEK (Lys 489)	685.0222 (+3)	2	-	-	685.0256 (+3)	0.0034	685.0310 (+3)	0.0088
CASIQK*FGER (Lys 228)	854.9142 (+2)	1	-	-	854.9284 (+2)	0.0143	854.9290 (+2)	0.0149
VTK*CCTESLVNR (Lys 498)	990.4743 (+2)	1	-	-	990.4813 (+2)	0.0071	990.4930 (+2)	0.0188
FK*DLGEEHFK (Lys 36)	881.9304 (+2)	2	-	-	-	-	881.9491 (+2)	0.0188
LKECCDK*PLLEK (Lys 304)	1023.5103 (+2)	3	-	-	-	-	1023.5348 (+2)	0.0245

Figure 1(a) displays the mass spectra obtained form the extraction of the precursor ion at m/z 752.4233 (+2) corresponding to the glycated peptide EK*VLTSSAR (Lys 211). The CID-MS/MS analysis of this glycated peptide reveals a signature of the carbohydrate moiety through the following product ions: B⁺ at m/z 262.1265, [B − H₂O]⁺ at m/z 244.1142, [B − 2H₂O]⁺ at m/z 226.1047, [^2,5A − 2H₂O]⁺ at m/z 152.0663, [B − C₄H₆O₃]⁺ at m/z 142.0815 and [^2,5A − CH₄O₃]⁺ at m/z 124.0708 (Fig. 2). Moreover, we were also able to detect the entire peptide attached to the linker portion: Y⁺ at m/z 1242.6794 and Z⁺ at m/z 1224.6785. It has also to be noticed that the product ion at m/z 336.1918 can be assigned to two different structures:${[\mathrm{B}+{\mathrm{C}}_4{\mathrm{H}}_{10}\mathrm{O}]}^{+}({\mathrm{C}}_{15}{\mathrm{H}}_{30}{\text{NO}}_7^{+})\text{ and }{\mathrm{C}}_{17}{\mathrm{H}}_{26}{\mathrm{N}}_3{\mathrm{O}}_4^{+}$ (Fig. 2). However, the product at m/z 336.1918 was also detected in our previous work on Bacillus anthracis neoglycoconjugate vaccine model and assigned to the structure: [C₁₇H₂₆N₃O₄]⁺ which corresponds to a linker portion attached to a lysine residue (Fig. 2).^[15] Similarly, the product ion [C₁₁H₁₆N₃O₂]⁺ at m/z 222.1187 was produced from a linker portion attached to a lysine residue. This latter product ion was also detected in our previous work on the glycation sites determination of Bacillus anthracis neoglycoconjugate vaccine model.^[15]

Figure 1.

LC-ESI-QqTOF-MS/MS spectra of the tryptic glycopeptides (a) EK*VLTSSAR (Lys 211) at m/z 752.4233 (+2), (b) GACLLPK*IETMR (Lys 204) at m/z 951.5091 (+2), (c) VTK*CCTESLVNR (Lys 498) at m/z 990.4813 (+2) and FK*DLGEEHFK (Lys 36) at m/z 881.9491 (+2).

Figure 2.

Different product ions involving the fragmentation of the carbohydrate hapten observed during the LC-ESI-QqTOF-MS/MS analysis of the tryptic glycopeptide EK*VLTSSAR (Lys 211) at m/z 752.4233 (+2).

In addition, the CID-MS/MS analysis of this glycated peptide EK*VLTSSAR (Lys 211), at m/z 752.4233 (+2), allowed us to observe the formation of product ions corresponding to glycopeptide fragments obtained by the loss of their carbohydrate portion: [y₈ − B]⁺ at m/z 1113.6384, [b₈ − B]⁺ at m/z 1068.5932, [b₇ − B]⁺ at m/z 997.5436, [b₆ − B]⁺ at m/z 910.5089, [b₅ − B]⁺ at m/z 823.4703, [b₅ − B − H₂O]⁺ at m/z 805.4549, [b₄ − B]⁺ at m/z 722.4164, [b₄ − B − H₂O]⁺ at m/z 704.4081, [b₃ − B]⁺ at m/z 609.3292, [b₃ − B − H₂O]⁺ at m/z 591.3171, [b₂ − B]⁺ at m/z 510.2563 and [b₂ − B − H₂O]⁺ at m/z 492.2474. The series of detected y- and b-ions, which covered the sequence of the glycopeptides, are listed in Table 2.

Table 2.

LC-ESI-QqTOF-MS/MS analysis of the glycated peptide EK*VLTSSAR (Lys 211) at m/z 752.4233 (+2)

Molecular ion	Calculated	Experimental	Deviation
	m/z	m/z	Da
Y+	1242.6683	1242.6794	0.0111
Z+	1224.6577	1224.6785	0.0208
[y8 − B]+	1113.6252	1113.6384	0.0132
[b8 − B]+	1068.5572	1068.5932	0.0360
[b7 − B]+	997.5201	997.5436	0.0235
[b6 − B]+	910.4880	910.5089	0.0209
[b5 − B]+	823.4560	823.4703	0.0143
[b5 − B − H2O]+	805.4454	805.4549	0.0095
${\mathrm{y}}_7^{+}$	733.4197	733.4340	0.0143
[b4 − B]+	722.4083	722.4164	0.0081
[b4 − B − H2O]+	704.3977	704.4081	0.0104
${\mathrm{y}}_6^{+}$	634.3513	634.3576	0.0063
[b3 − B]+	609.3243	609.3292	0.0049
[b3 − B − H2O]+	591.3137	591.3171	0.0034
${\mathrm{y}}_5^{+}$	521.2673	521.2713	0.0040
[b2 − B]+	510.2558	510.2563	0.0005
[b2 − B − H2O]+	492.2452	492.2474	0.0022
${\mathrm{y}}_4^{+}$	420.2196	420.2190	−0.0006
${\mathrm{C}}_{17}{\mathrm{H}}_{26}{\mathrm{N}}_3{\mathrm{O}}_4^{+}$	336.1918	336.1918	0.0000
${\mathrm{y}}_3^{+}$	333.1875	333.1851	−0.0024
B+	262.1290	262.1265	−0.0025
[B − H2O]+	244.1179	244.1142	−0.0037
[B − 2H2O]+	226.1079	226.1047	−0.0032
${\mathrm{C}}_{11}{\mathrm{H}}_{16}{\mathrm{N}}_3{\mathrm{O}}_2^{+}$	222.1237	222.1187	−0.0050
${\mathrm{y}}_1^{+}$	175.1184	175.1162	−0.0022
[2,5A − 2H2O]+	152.0706	152.0663	−0.0043
[B − C4H6O3]+	142.0863	142.0815	−0.0048
[2,5A − CH4O3]+	124.0757	124.0708	−0.0049

Figure 1(b) displays another example CID-MS/MS analysis of a glycated peptide detected during the analysis of the tryptic digests of the hapten-BSA neoglycoconjugate with a hapten:BSA ratio of 4.3:1: GACLLPK*IETMR (Lys 204) at m/z 951.5091 (+2).

The LC-ESI-QqTOF-MS/MS analysis of the GluC V8 digests of the neoglycoconjugate with a hapten:BSA ratio of 4.3:1 permitted the localization of ten glycation sites through the sequencing of the observed glycopeptides listed in Table 3.

Table 3.

GluC V8 glycopeptide digests identified during LC-ESI-QqTOF-MS/MS analysis of the hapten-BSA glycoconjugate with a hapten:BSA ratio of 4.3:1, 6.6:1 and 13.2:1

Peptide		Missed	Hapten:BSA ratio 4.3:1		Hapten:BSA ratio 6.6:1		Hapten:BSA ratio 13.2:1
Sequence (star = glycation site)	Calculated		Observed	Deviation	Observed	Deviation	Observed	Deviation
	m/z (charge)	cleavage	m/z (charge)	(Da)	m/z (charge)	(Da)	m/z (charge)	(Da)
K*SHCIAE (Lys 309)	679.3189 (+2)	0	679.3320 (+2)	0.0131	679.3343 (+2)	0.0154	679.3284 (+2)	0.0095
LTKVHK*E (Lys 266)	684.3745 (+2)	0	684.3867 (+2)	0.0122	684.3851 (+2)	0.0106	684.3795 (+2)	0.0050
K*VTKCCTE (Lys 495)	769.8575 (+2)	0	769.8677 (+2)	0.0102	769.8720 (+2)	0.0145	769.8668 (+2)	0.0093
K*QEPERNE (Lys 117)	771.8678 (+2)	2	771.8852 (+2)	0.0174	771.8795 (+2)	0.0117	771.8883 (+2)	0.0205
K*SLHTLFGDE (Lys 88)	830.4093 (+2)	1	830.4932 (+2)	0.0839	830.4246 (+2)	0.0153	830.4196 (+2)	0.0103
DKGACLLPK*IE (Lys 204)	878.9556 (+2)	0	878.9783 (+2)	0.0227	878.9669 (+2)	0.0113	878.9643 (+2)	0.0087
AK*DAFLGSFLYE (Lys 346)	937.4590 (+2)	1	937.4871 (+2)	0.0281	937.4759 (+2)	0.0169	937.4798 (+2)	0.0208
KK*FWGKYLYE (Lys 156)	937.9824 (+2)	0	937.9877 (+2)	0.0053	937.9862 (+2)	0.0038	937.9648 (+2)	−0.0176
LLYYANK*YNGVFQE (Lys 183)	1118.0465 (+2)	0	1118.0711 (+2)	0.0246	1118.0587 (+2)	0.0122	1118.0511 (+2)	0.0046
VSRSLGK*VGTRCCTKPE (Lys 455)	816.7456 (+3)	0	816.7646 (+3)	0.0190	816.7583 (+3)	0.0127	816.7483 (+3)	0.0027
K*TPVSE (Lys 489)	587.2979 (+2)	0	-	-	587.3103 (+2)	0.0124	587.3041 (+2)	0.0062
GPK*LVVSTQTALA (Lys 597)	899.4959 (+2)	0	-	-	899.5080 (+2)	0.0121	899.5092 (+2)	0.0133
LLKHKPK*ATEE (Lys 561)	903.9960 (+2)	1	-	-	904.0190 (+2)	0.0230	903.9916 (+2)	−0.0044
K*QIKKQTALVE (Lys 544)	900.0117 (+2)	0	-	-	900.0234 (+2)	0.0117	900.0165 (+2)	0.0048
VTK*LVTD (Lys 256)	644.8478 (+2)	0	-	-	-	-	644.8498 (+2)	0.0020
HVK*LVNE (Lys 65)	676.3588 (+2)	0	-	-	-	-	676.3647 (+2)	0.0059
VEK*DAIPE (Lys 318)	707.3534 (+2)	2	-	-	-	-	707.3578 (+2)	0.0044
CCDK*PLLE (Lys 304)	774.3520 (+2)	1	-	-	-	-	774.3662 (+2)	0.0142
LCK*VASLRE (Lys 100)	794.9162 (+2)	0	-	-	-	-	794.9163 (+2)	0.0001
VSRSLGK*VGTRCCTK*PE (Lys 455, Lys 463)	987.8230 (+3)	0	-	-	-	-	987.8313 (+3)	0.0083
KK*FWGK*YLYE (Lys 156, 160)	1194.5985 (+2)	0	-	-	-	-	1194.5626 (+2)	−0.0359
NFVAFVDK*CCAADDKE (Lys 580)	1201.5300 (+2)	3	-	-	-	-	1201.5350 (+2)	0.0050

Figure 3 displays two examples of glycopeptides identified during the LC-ESI-QqTOF-MS/MS analysis of the GluC V8 digests of the neoglycoconjugate with a hapten-BSA ratio of 4.3: (a) K*SHCIAE (Lys 309) at m/z 679.3320 (+2) and (b) VSRSLGK*VGTRCCTKPE (Lys 455) at m/z 816.7646 (+3). The fragmentation pathway followed by the glycopeptide VSRSLGK*VGTRCCTKPE (Lys 455) at m/z 816.7646 (+3) allowed us to identify the glycation site exclusively on the Lys 455 residue (Fig. 3(b), Table 4). The precursor ion VSRSLGK*VGTRCCTKPE (Lys 455) at m/z 816.7646 (+3) produced characteristic carbohydrate moiety product ions, identified as follows: B⁺ at m/z 262.1247, [B − H₂O]⁺ at m/z 244.1136, [B − 2H₂O]⁺ at m/z 226.1042, [^2,5A − 2H₂O]⁺ at m/z 152.0669, [B − C₄H₆O₃]⁺ at m/z 142.0824 and [^2,5A₁ − CH₄O₃]⁺ at m/z 124.0714. Furthermore, we detected the following product ion Y²⁺ at m/z 1094.0368 corresponding to the entire glycopeptide minus the loss of the entire carbohydrate portion. In addition, we also observed product ions corresponding to distinct fragmentation of the peptide backbone minus the loss of the entire carbohydrate moiety; needless to say, these product ions included the previously derivatized lysine: [y₁₂ − B]⁺ at m/z 1644.7886, [y₁₁ − B]⁺ at m/z 1587.7061, [b₉ − B]⁺ at m/z 1136.5123, [b₈ − B]⁺ at m/z 1079.5651, [b₁₆ − B]²⁺ at m/z 1020.5277, [b₇ − B]⁺ at m/z 980.5699, [b₁₅ − B]²⁺ at m/z 972.0032, [b₁₄ − B]²⁺ at m/z 907.9123 and [GK* − B]⁺ at m/z 438.2288. Therefore, these product ions allowed us to sequence and characterize this glycopeptide correctly.

Figure 3.

LC-ESI-QqTOF-MS/MS spectra of the glycopeptides obtained during the digestion of the neoglycoconjugates with the GluC V8 endoproteinase: (a) K*SHCIAE (Lys 309) at m/z 679.3320 (+2), (b) VSRSLGK*VGTRCCTKPE (Lys 455) at m/z 816.7646 (+3), (c) GPK*LVVSTQTALA (Lys 597) at m/z 899.5080 (+2) and VSRSLGK*VGTRCCTK*PE (Lys 455, Lys 463) at m/z 987.8313 (+3).

Table 4.

LC-ESI-QqTOF-MS/MS analysis of the glycated peptide VSRSLGK*VGTRCCTKPE (Lys 455) at m/z 816.7646 (+3)

Molecular ion	Calculated	Experimental	Deviation
	m/z	m/z	Da
[y12 − B]+	1644.7821	1644.7886	−0.0065
[y11 − B]+	1587.7607	1587.7061	0.0546
[b9 − B]+	1136.5422	1136.5123	0.0299
${\mathrm{y}}_9^{+}$	1108.4868	1108.4700	0.0168
Y2+	1094.0541	1094.0368	0.0172
[b8 − B]+	1079.6208	1079.5651	0.0557
${\mathrm{y}}_8^{+}$	1051.4654	1051.4701	−0.0047
[b16 − B]2+	1020.5278	1020.5277	0.0001
[b7 − B]+	980.5524	980.5699	−0.0175
[b15 − B]2+	972.0014	972.0032	−0.0018
${\mathrm{y}}_7^{+}$	950.4177	950.4802	−0.0625
[b14 − B]2+	907.9539	907.9123	0.0416
${\mathrm{y}}_6^{+}$	794.3166	794.3096	0.0070
${\mathrm{y}}_5^{+}$	634.2859	634.2643	0.0216
${\mathrm{b}}_6^{+}$	600.3469	600.3362	0.0107
${\mathrm{b}}_5^{+}$	543.3255	543.3074	0.0181
${\mathrm{y}}_4^{+}$	474.2553	474.2656	−0.0103
[GK* − B]+	438.2342	438.2288	0.0054
${\mathrm{b}}_4^{+}$	430.2414	430.2307	0.0107
${\mathrm{y}}_3^{+}$	373.2076	373.2188	−0.0112
${\mathrm{b}}_3^{+}$	343.2094	343.2028	0.0066
${\mathrm{C}}_{17}{\mathrm{H}}_{26}{\mathrm{N}}_3{\mathrm{O}}_4^{+}$	336.1918	336.1787	0.0131
B+	262.1290	262.1247	0.0043
${\mathrm{y}}_2^{+}$	245.1126	245.1127	−0.0001
[B − H2O]+	244.1179	244.1136	0.0043
[B − 2H2O]+	226.1079	226.1042	0.0037
${\mathrm{C}}_{11}{\mathrm{H}}_{16}{\mathrm{N}}_3{\mathrm{O}}_2^{+}$	222.1237	222.1242	−0.0005
${\mathrm{b}}_2^{+}$	187.1083	187.1204	−0.0121
[2,5A − 2H2O]+	152.0706	152.0669	0.0037
[y1 − H2]+	146.0453	146.0706	−0.0253
[B − C4H6O3]+	142.0863	142.0824	0.0039
[2,5A1 − CH4O3]+	124.0757	124.0714	0.0043

Finally, the following product ions, characteristic to a spacer portion attached to a lysine residue were also identified as follows: [C₁₇H₂₆N₃O₄]⁺ at m/z 336.1787 and [C₁₁H₁₆N₃O₂]⁺ at m/z 222.1242. The bidirectional sequence of the glycated peptide was also determined through the identification of the y- and b-ions (Fig. 3(b), Table 4).

To sum it up, a total of 20 glycation sites were identified during the LC-ESI-QqTOF-MS/MS analysis of the tryptic and GluC V8 digests of the neoglycoconjugate with a hapten:BSA ratio of 4.3:1 (Fig. 4(a)) for total sequence coverage of 89%. Figure 5(a) displays the 3D structure of the BSA represented using the Swiss-Pdb Viewer software, where the glycated lysine residues are highlighted in red.^[19,20] A quick glance at Fig. 5(a) allowed us to conclude that the glycated lysine residues are located on the outer surface of the protein.

Figure 4.

Neoglycoconjugate model sequences where the glycation sites are indicated by an asterix (red = identified on tryptic digests, blue = identified on GluC V8 digests and red and underlined = identified on both tryptic and GluC V8 digests) for the neoglycoconjugates with a hpaten-BSA ratio of (a) 4.3:1. (b) 6.6:1 and (c) 13.2:1.

Figure 5.

3D-structure of the neoglycoconjugate model vaccines. The glycated lysine residues are highlighted in red (Swiss-Pdb Viewer software) for the neoglycoconjugates with a hapten:BSA ratio of (a) 4.3:1, (b) 6.6:1 and (c) 13.2:1. This figure is available in colour online at wileyonlinelibrary.com/journal/jms

Mass spectrometry determination of the glycation sites on the neoglycoconjugate possessing a hapten: BSA ratio of 6.6:1

The LC-ESI-QqTOF-MS/MS analysis of the tryptic digests of the neoglycoconjugate with a hapten:BSA ratio of 6.6:1 allowed identification of 16 glycation sites on through sequencing of the observed glycopeptides (Table 1).

Figure 1(c) displays the following example of the glycopeptide which was identified during the CID-MS/MS analysis of the tryptic digests of the neoglycoconjugate with a hapten:BSA ratio of 6.6:1: VTK*CCTESLVNR (Lys 498) at m/z 990.4813 (+2).

The neoglycoconjugate with a hapten:BSA ratio of 6.6:1 was also digested with the GluC V8 endoproteinase and subsequently analyzed by LC-ESI-QqTOF-MS/MS. The glycated peptides were thus discovered after the manual sequencing of the peptides which were suspected to be glycated (same mass than a ‘peptide + carbohydrate linker’) and listed in Table 3.

The LC-ESI-QqTOF-MS/MS analysis of the GluC V8 digests of the neoglycoconjugate with a hapten:BSA ratio of 6.6:1 allowed us to identify 14 glycation sites. Moreover, the combination of the identified glycation sites from both the tryptic and GluC V8 digests gives us a total of 27 glycation sites with total sequence coverage of 90% (Fig. 4(b)).

In addition, the Fig. 5(b) which shows the 3D structure of the BSA with the glycated lysine residues highlighted in red enabled us to conclude that the glycated lysine residues are located on the outer surface of the BSA.

Mass spectrometry determination of the glycation sites on the neoglycoconjugate possessing a hapten: BSA ratio of 13.2:1

The neoglycoconjugate with a hapten:BSA ratio of 13.2:1 was also digested with trypsin and LC-ESI-QqTOF-MS/MS analysis of the digests allowed identification of 17 glycopeptides (Table 1), corresponding to 17 glycation sites.

Figure 1(d) shows the following example of identified glycopeptide, which was only discovered on the neoglycoconjugate with a hapten:BSA ratio of 13.2:1: FK*DLGEEHFK (Lys 36) at m/z 881.9491 (+2).

Finally, the analysis by LC-ESI-QqTOF-MS/MS of the GluC V8 digests permitted identification of 20 glycopeptides (Table 3). Moreover, we also observed glycopeptides carrying two carbohydrate-linker residues: VSRSLGK*VGTRCCTK*PE (Lys 455, Lys 463) at m/z 987.8313 (+3) and KK*FWGK*YLYE (Lys 156, 160) at m/z 1194.5626 (+2).

It is important to note in this rationale that we have identified glycopeptides which contained two diagnostic glycation sites. Unquestionably, two glycopeptides with the following peptide sequence VSRSLGKVGTRCCTKPE were detected containing one glycation site on the Lys 455, VSRSLGK*VGTRCCTKPE (Lys 455) at m/z 816.7483 (+3), and also with two glycation sites on the Lys 455 and Lys 463 residues: VSRSLGK*VGTRCCTK*PE (Lys 455, Lys 463) at m/z 987.8313 (+3). This observation supports the fact that our sample is composed of glycoforms. Figure 3(d) displays the spectra obtained during the CID-MS/MS analysis of the glycated peptide VSRSLGK*VGTRCCTK*PE (Lys 455, Lys 463) at m/z 987.8313 (+3), while the observed product ions and their mass-to-charge ratios are listed in Table 5.

Table 5.

LC-ESI-QqTOF-MS/MS analysis of the glycated peptide VSRSLGK*VGTRCCTK*PE (Lys 455, Lys 463) at m/z 987.8313 (+3)

Molecular ion	Calculated	Experimental	Deviation
	m/z	m/z	Da
[b14 − B]+	1814.9000	1814.8000	0.1000
[b11 − B]+	1393.7910	1393.8366	−0.0456
[y9 − B]+	1360.5973	1360.5399	0.0574
Y2+	1350.6700	1350.6416	0.0284
[b10 − B]+	1237.6899	1237.6267	0.0632
[Y − B]2+	1220.1091	1220.0900	0.0191
[y7 − B]+	1202.5282	1202.6109	−0.0827
[b9 − B]+	1136.6422	1136.6248	0.0174
[b8 − B]+	1079.6208	1079.5776	0.0432
[b7 − B]+	980.5524	980.5371	0.0153
[b14 − B]2+	907.9539	907.9442	0.0097
[b12 − B]2+	777.4148	777.4165	−0.0018
[b11 − B]2+	697.3994	697.3653	0.0341
[CTK* − B]+	642.2910	642.2561	0.0349
[y3 − B]+	625.3181	625.3205	−0.0024
${\mathrm{b}}_6^{+}$	600.3469	600.3531	−0.0062
${\mathrm{b}}_5^{+}$	543.3255	543.3561	−0.0306
[K*VG − B]+	537.3026	537.2682	0.0344
[TK* − B]+	482.2604	482.2529	0.0075
[GK* − B]+	438.2342	438.2285	0.0057
${\mathrm{b}}_4^{+}$	430.2414	430.2185	0.0229
[K* − B]+	381.2127	381.1820	0.0307
${\mathrm{b}}_3^{+}$	343.2094	343.2214	−0.0120
${\mathrm{C}}_{17}{\mathrm{H}}_{26}{\mathrm{N}}_3{\mathrm{O}}_4^{+}$	336.1918	336.1890	0.0028
B+	262.1290	262.1259	0.0031
[B − H2O]+	244.1179	244.1156	0.0023
[B − 2H2O]+	226.1079	226.1058	0.0021
${\mathrm{C}}_{11}{\mathrm{H}}_{16}{\mathrm{N}}_3{\mathrm{O}}_2^{+}$	222.1237	222.1170	0.0067
${\mathrm{b}}_2^{+}$	187.1083	187.1118	−0.0035
[2,5A − 2H2O]+	152.0706	152.0673	0.0033
[y1 − H2]+	146.0453	146.0758	−0.0305
[B − C4H6O3]+	142.0863	142.0837	0.0026
[2,5A − CH4O3]+	124.0757	124.0733	0.0024

In the product ion scan of the precursor ion at m/z 987.8313 (+3), we observed the product ion Y²⁺ at m/z 1350.6416 formed by the loss of the carbohydrate moiety from one of the glycated lysine residues. The product ion [Y − B]²⁺ at m/z 1220.0900 was given rise to by the loss of the carbohydrate portions from the two glycated lysine residues. It is important to mention that the following series of diagnostic product ions: B⁺ at m/z 262.1259, [B − H₂O]⁺ at m/z 244.1156, [B − 2H₂O]⁺ at m/z 226.1058, [^2,5A − 2H₂O]⁺ at m/z 152.0673, [B − C₄H₆O₃]⁺ at m/z 142.0837, [^2,5A − CH₄O₃]⁺ at m/z 124.0733, [C₁₇H₂₆N₃O₄]⁺ at m/z 336.1890 and [C₁₁H₁₆N₃O₂]⁺ at m/z 222.1170 were formed specifically from both the carbohydrate portions and the spacer moieties of the glycopeptide precursor ion. Furthermore, we have observed during this CID-MS/MS analysis ‘proper’ peptide product ions that were identified as follows: [b₁₄ − B]⁺ at m/z 1814.8000, [b₁₁ − B]⁺ at m/z 1393.8366, [y₉ − B]⁺ at m/z 1360.5399, [b₁₀ − B]⁺ at m/z 1237.6267, [y₇ − B]⁺ at m/z 1202.6109, [b₉ − B]⁺ at m/z 1136.6248, [b₈ − B]⁺ at m/z 1079.5776, [b₇ − B]⁺ at m/z 980.5371, [b₁₄ − B]²⁺ at m/z 907.9442, [b₁₂ − B]²⁺ at m/z 777.4165, [b₁₁ − B]²⁺ at m/z 697.3653, [CTK* − B]⁺ at m/z 642.2561, [y₃ − B]⁺ at m/z 625.3205, [K*VG − B]⁺ at m/z 537.2682, [TK* − B]⁺ at m/z 482.2529, [GK* − B]⁺ at m/z 438.2285 and [K* − B]⁺ at m/z 381.1820. All of these product ions were formed by the loss of the carbohydrate moiety exclusively, which allowed us to localize unambiguously the two different glycation sites.

Thus, the CID-MS/MS sequencing of the previously mentioned glycopeptides allowed us to identify 22 glycation sites. Moreover, the combination of the CID-MS/MS analysis of the tryptic and GluC V8 digests enabled us to discover a total of 33 glycation sites with total sequence coverage of 88% (Fig. 4(c)). As previously observed for the neoglycoconjugates with hapten:BSA ratios of 4.3:1 and 6.6:1, the glycated lysine residues for the hapten-BSA glycoconjugate with a ratio of 13.2:1 are positioned on the outer surface of the protein (Fig. 5(c)).

CONCLUSION

The LC-ES-QqTOF-MS/MS analysis of the tryptic and GluC V8 digests allowed identification of a higher number of glycation sites than previously reported for the neoglycoconjugates with a hapten:BSA ratio of 4.3:1, 6.6:1 and 13.2:1.^[12] As expected, the number of glycation sites in the neoglycoconjugates increases with the hapten:BSA ratio. Indeed, 20 glycation sites were observed for the neoglycoconjugate with a hapten:BSA ratio of 4.3:1, 27 for the neoglycoconjugate with a hapten:BSA ratio of 6.6:1 and 33 for the neoglycoconjugate with a hapten:BSA ratio of 13.2:1. Moreover, it was found that the glycated lysine residues are situated on the outer surface of the BSA.

It is also interesting to note that most of the observed glycation sites were at positions similar to those observed for the Bacillus anthracis neoglycoconjugate vaccine.^[15]