Reanalysis of Global Proteomic and Phosphoproteomic Data Identified a Large Number of Glycopeptides

Protein glycosylation plays fundamental
roles in many cellular processes, and previous reports have shown
dysregulation to be associated with several human diseases, including
diabetes, cancer, and neurodegenerative disorders. Despite the vital
role of glycosylation for proper protein function, the analysis of
glycoproteins has been lagged behind to other protein modifications.
In this study, we describe the reanalysis of global proteomic data
from breast cancer xenograft tissues using recently developed software
package GPQuest 2.0, revealing a large number of previously unidentified
N-linked glycopeptides. More importantly, we found that using immobilized
metal affinity chromatography (IMAC) technology for the enrichment
of phosphopeptides had coenriched a substantial number of sialoglycopeptides,
allowing for a large-scale analysis of sialoglycopeptides in conjunction
with the analysis of phosphopeptides. Collectively, combined tandem
mass spectrometry (MS/MS) analyses of global proteomic and phosphoproteomic
data sets resulted in the identification of 6 724 N-linked
glycopeptides from 617 glycoproteins derived from two breast cancer
xenograft tissues. Next, we utilized GPQuest 2.0 for the reanalysis
of global and phosphoproteomic data generated from 108 human breast
cancer tissues that were previously analyzed by Clinical Proteomic
Analysis Consortium (CPTAC). Reanalysis of the CPTAC data set resulted
in the identification of 2 683 glycopeptides from the global
proteomic data set and 4 554 glycopeptides from phosphoproteomic
data set, respectively. Together, 11 292 N-linked glycopeptides
corresponding to 1 731 N-linked glycosites from 883 human glycoproteins
were identified from the two data sets. This analysis revealed an
extensive number of glycopeptides hidden in the global and enriched
in IMAC-based phosphopeptide-enriched proteomic data, information
which would have remained unknown from the original study otherwise.
The reanalysis described herein can be readily applied to identify
glycopeptides from already existing data sets, providing insight into
many important facets of protein glycosylation in different biological,
physiological, and pathological processes.