Maximizing Coverage of Glycosylation Heterogeneity in MALDI-MS Analysis of Glycoproteins with Up to 27 Glycosylation Sites

2008-05-01T00:00:00Z (GMT) by Ying Zhang Eden P. Go Heather Desaire
Glycosylation affects various biological functions of proteins (e.g., protein binding, inter- or intracell signaling, etc.), and it can serve as an indicator of disease. Therefore, characterization of the glycosylation in proteins is one important step in developing a comprehensive understanding of the biological significance of glycosylation and in facilitating disease diagnosis. Glycopeptide-based MS analysis has proven to be a viable tool for glycopeptide analysis. However, when glycopeptides coexist with peptides, glycopeptide signals are usually suppressed by the strongly ionizing peptides. Toward this end, it would be desirable to seek methods to improve glycopeptide detection. Herein, we performed an in-depth study of maximizing glycosylation coverage on model glycoproteins by optimizing all the aspects of glycopeptide-based analysis, including sample preparation methods, mass spectral techniques, and data analysis strategies. For sample preparation, several approaches, including reversed-phase high-performance liquid chromatography, lectin-based affinity, and hydrophilic affinity using a carbohydrate-based resin, were compared and tested individually as well as in parallel. For mass spectral techniques, profiling glycopeptides in both positive and negative ion mode is essential to obtain complete glycan profiles. For data analysis, incorporating variable modifications in the database search of GlycoPep DB enhances glycopeptide coverage. In addition, the use of PNGase F helps to confirm the presence of weakly ionized glycopeptides when they coelute with strongly ionizing species. In doing so, we created a work flow that is designed specifically to optimize the coverage of glycosylation heterogeneity in terms of the number of glycosylation sites detected and their corresponding glycan profiles. To test the effectiveness of this approach, a glycoprotein with 27 potential glycosylation sites, and an unknown glycosylation profile, was analyzed; on this protein, more than 300 glycoforms from 23 detected glycosylation sites were identified. This work demonstrates that these strategies significantly improve the glycopeptide detection, thereby, facilitating understanding the functional properties of glycans on the glycoproteins.