N-Linked Glycopeptide Identification Based on Open Mass Spectral Library Search.

Zhiwu An,Fuquan Yang,Yan Fu,Jifeng Wang,Lian Shu,Qingbo Shu,Hao Lv

doi:10.1155/2018/1564136

Abstract

Confident characterization of intact glycopeptides is a challenging task in mass spectrometry-based glycoproteomics due to microheterogeneity of glycosylation, complexity of glycans, and insufficient fragmentation of peptide bones. Open mass spectral library search is a promising computational approach to peptide identification, but its potential in the identification of glycopeptides has not been fully explored. Here we present pMatchGlyco, a new spectral library search tool for intact N-linked glycopeptide identification using high-energy collisional dissociation (HCD) tandem mass spectrometry (MS/MS) data. In pMatchGlyco, (1) MS/MS spectra of deglycopeptides are used to create spectral library, (2) MS/MS spectra of glycopeptides are matched to the spectra in library in an open (precursor tolerant) manner and the glycans are inferred, and (3) a false discovery rate is estimated for top-scored matches above a threshold. The efficiency and reliability of pMatchGlyco were demonstrated on a data set of mixture sample of six standard glycoproteins and a complex glycoprotein data set generated from human cancer cell line OVCAR3.

Highlights

Glycosylation plays a key role in most biological processes such as glycose metabolism, signaling, cell adhesion, and cell proliferation [1]
Based on its inherited algorithm, pGlyco only considers matching of Y fragment ions when scoring high-energy collisional dissociation (HCD)/collision-induced dissociation (CID) MS/MS spectra; it is unable to evaluate matching of peptide fragment ions for each GPSM
On the same data set, GPQuest identified 4,562 GPSMs at 1% false discovery rate (FDR) [24]. Among these GPSMs, we found that 1,162 spectra were theoretically unable to be consistently identified by two software tools, because (1) after raw data conversion to mgf by GPQuest workflow, 614 spectra had different precursor masses from those converted from pParse, which was used by pMatchGlyco workflow, (2) in the peptide library that pMatchGlyco built based on deglycosylated peptides, 497 GPSMs from GPQuest workflow did not have the corresponding peptides, and (3) 51 spectra had different precursor masses and lacked corresponding peptides

Summary

Introduction

Glycosylation plays a key role in most biological processes such as glycose metabolism, signaling, cell adhesion, and cell proliferation [1]. Two main types of glycans have been reported: N-linked and O-linked glycans. N-linked glycans are attached to the Asn residue within a peptide sequence sequon of Asn-Xaa-Ser/Thr/Cys, where Xaa can be any amino acid except Pro. N-linked glycans own a five-monosaccharide core structure. The structures of O-linked glycans are more complex. We here focus on the analysis of N-linked glycopeptides

Methods

Results

Discussion

Conclusion