Glyco

Title : Identification of N- and O-linked glycopeptides by Mascot

Abstract :

The feasibility of the Mascot search engine for automated annotation of both N- and O-linked glycopeptides was validated using two standard bovine glycoproteins ( alpha-1-acid glycoprotein and fetuin )
When searched against the custom glycoprotein database , the MS2 spectrum shown in Fig. 1A is annotated as a di-sialylated bi-antennary N-glycopeptide of alpha-1-acid glycoprotein , with a Mascot ion score of 24 (Fig. 2A)
As theoretically expected, Mascot annotated the intense peaks to a series of y ions starting from y13 (peptide + HexNAc) until y20 (peptide + HexNAc(O)3 − Hex(J)4 − Neu5Ac(U)1).
Together with the precursor mass, and the b1, b2 and b3 ions, the presence of additional HexNAc(O)1 − Hex(J)1 − Neu5Ac(U)1 residues was confirmed, thereby providing 100% sequence coverage of the glycan (Fig. 2A)
However, no other information in the spectrum confirmed the peptide sequence except the precursor mass
The lack of peptide fragmentation information in the MS2 spectrum might create difficulties in differentiating glycopeptide sequences resulting in similar Mascot ion scores
However, the fragmentation of the glycopeptides can be fine-tuned by the NCE values used for HCD fragmentation
As an example, the tryptic peptides of alpha-1-acid glycoprotein were fragmented at different NCE values
At NCE values of 15 and 25, the di-sialylated bi-antennary glycopeptide MS2 spectra displayed glycosidic fragment ions (Fig. 2A,B)
However, the same glycopeptide contained a series of peptide cleavage type y ions (y4 to y9) at an NCE value of 35 with almost no information about the glycan structure
Hence, Mascot annotated mostly the peptide part of the glycopeptide sequence (Fig. 2C)
Consequently, a single NCE value might not provide enough information about both the glycan and peptide sequence
With the stepped NCE option of quadrupole-orbitrap mass spectrometers, the instrument can acquire fragmentation data of the precursors at multiple collision energies
With this option, up to three different NCE values can be selected to generate a composite MS2 spectrum as shown in Fig. 2D, combining 15, 25 and 35 as NCE values
This MS2 spectrum revealed near to complete information about the glycan sequence and the peptide y ions (y4, y5, y8 and y12) were detected as well
Mascot unambiguously annotated this MS2 spectrum with an ion score of 34 (Fig. 2D)
The feasibility of the Mascot search engine for the analysis of O-linked glycopeptides was validated by analyzing the mass spectrometry data of bovine fetuin against a custom O-glycoprotein sequence of fetuin
Mascot annotated mono- and di-sialylated core-1 O-glycans on two different peptide sequences
A series of y ions (y5 to y18) and b ions (b1, b2,b3) covering the most intense peaks (Fig. 3A,B) clearly confirmed that these MS2 spectra correspond to the O-linked glycopeptides
These spectra were unambiguously annotated with an excellent Mascot ion score of more than 40
Similar to N-linked glycopeptides , the b1, b2 and b3 ions at m/z values of 292.102 (Neu5Ac) , 454.155 (Neu5Ac-Hex) and 657.233 (Neu5Ac-Hex-HexNAc) covered the low mass glycan fragment ions and provided an additional layer of confirmation about the O-glycopeptide spectra
A similar fragmentation behavior was observed for two other O-glycopeptides of the same protein (Fig. 3C,D)
To further display the feasibility of Mascot, analysis of bacterial O-glycosylation was performed on a purified PilE protein
The PilE protein contains a di-N-acetyl-bacillosamine (diNAcBac) and galactose based glycans with a potential acetylation on the galactose residue
Mascot was able to annotate the diNAcBac (Supplementary Fig. 3A), diNAcBac-Gal (Supplementary Fig. 3B) residues as well as the monoacetylation (Supplementary Fig. 3C) and diacetylation (Supplementary Fig. 3D) on galactose residues
For the complex O-glycosylation study, we re-analyzed the previously published mass spectrometry data from the immunoaffinity purified fractions and whole cell extract of a Neisseria gonorrhoeae strain
The Mascot annotated glycopeptides were compared to the previously published data, where software assistance and manual data analysis was performed and the majority (11/13 glycopeptides ) of the previously confirmed O-glycopeptides were identified automatically (Supplementary Table 2)
Taken together all these results clearly indicates the potential of the described approach i.e. using stepped NCE values, a custom linearized glycoprotein database and the Mascot search engine for automated glycopeptide annotation

Output (sent_index, trigger, protein, sugar, site):

0. glycopeptides, , -, -, glycopeptides
1. glycopeptides, , -, -, glycopeptides
1. glycoprotein, , glycoprotein, -, -
1. glycoprotein, , glycoproteins, -, -
1. glycoproteins, , fetuin, -, -
1. glycoproteins, , glycoprotein, -, -
1. glycoproteins, , glycoproteins, -, -
10. glycopeptide, , -, -, glycopeptide
11. glycopeptide, , -, -, glycopeptide sequence
17. O-glycoprotein, , O-glycoprotein, -, -
17. glycopeptides, , -, -, glycopeptides
19. glycopeptides, , -, -, glycopeptides
2. N-glycopeptide, , glycoprotein, -, N-glycopeptide
2. di-sialylated, , -, -, N-glycopeptide
2. glycoprotein, , glycoprotein, -, -
21. O-glycopeptide, , -, -, O-glycopeptide
21. glycopeptides, , -, -, glycopeptides
22. O-glycopeptides, , protein, -, O-glycopeptides
27. O-glycopeptides, , -, -, O-glycopeptides
27. glycopeptides, , -, -, glycopeptides
28. glycopeptide, , -, -, glycopeptide
28. glycoprotein, , glycoprotein, -, -
6. glycopeptide, , -, -, glycopeptide sequences
7. glycopeptides, , -, -, glycopeptides
8. glycoprotein, , glycoprotein, -, -
9. glycopeptide, , -, -, glycopeptide

Output(Part-Of) (sent_index, protein, site):

17. O-glycoprotein, sequence
17. fetuin, sequence
2. MS2, N-glycopeptide
2. glycoprotein, N-glycopeptide
22. protein, O-glycopeptides
8. glycoprotein, peptides

*Output_Site_Fusion* (sent_index, protein, sugar, site):