Qiuting Hong; Evan Parker; Ting Song; Carlito Lebrilla

Chemsitry, UC, Davis, Davis, CA

Novel Aspect
A comprehensive site-specific glycan map for plasma glycoproteins

Glycosylation analysis on the site-specific level provides both protein and glycosylation information to facilitate biomarker discovery. However, site-specific analysis remains a difficult task, and there is no comprehensive glycan map even for the most abundant plasma glycoproteins. We are in the process of mapping the glycosylation and the site-specific heterogeneity in plasma proteins. Trypsin produces consistent peptides for quantitation but may provide limited glycosite information. Nonspecific protease yields better glycosite coverage but data analysis is complicated. To obtain comprehensive glycan maps several protease methods are used. A software tool for examining tandem MS of glycopeptides is also proven to be highly useful. The method can be applied to any glycoprotein and will further facilitate our efforts on disease biomarker discovery.

Glycosylation analysis was performed on commercially available plasma proteins including IgG, IgM, IgA, transferrin, haptoglobin, alpha-2-macroglobulin, alpha-1-antitrypsin, C3 complement, and alpha-1-acid glycoprotein. Glycopeptides were profiled using an Agilent 6520 nano-LC-Chip/QTOF MS, and identified using in-house software, GPFinder. Both trypsin and pronase (nonspecific protease), using both in-gel and in-solution digestion techniques, were applied to obtain a more comprehensive glycan map. Proteins were treated with DTT and IAA before enzyme digestion in a 37?C water bath. Trypsin digestions were performed overnight, and the resulting mixtures were used directly for LC-MS analysis. Pronase digestions were performed 1 hour for the in-gel and 3 hours for the in-solution approaches. The resulting sample was purified using C18 or graphitized zip-tip before MS analysis.

Preliminary Results/Abstract
We have determined the site-specific glycosylation of the top 10 plasma glycoproteins using both specific and nonspecific proteolysis. Many of these proteins have only partial assignments or no site-specific assignments for protein glycosylation, despite being the most abundant proteins. For proteins, like alpha-2-marcoglobulin that have fewer lysine and arginine, trypsin produces large glycopeptides and multiple glycosites on single peptides. For alpha-2-marcoglobulin, only one glycosite (N1424) can be mapped out using trypsin. Pronase produces smaller peptides and yields site-specific glycoform information for all eight glycosites of alpha-2-marcoglobulin. It was found that glycosites N247, N410, N869 have both high mannose type and sialylated complex type glycans, whereas the other five glycosites contain sialylated and neutral complex type glycans.
For some proteins, the enzymes provide complimentary information. Take IgA as an example, trypsin digestion provides glycosylation information for three out of seven glycosites, N144(IgA1), N131(IgA2), and N205(IgA2). Pronase digestion was found to yield glycan information the same sites N144(IgA1) and N131(IgA2), but also other sites N340(IgA1). Glycoforms on glycosite N340(IgA1) and N205(IgA2) are all fucosylated, while glycoforms on N144(IgA1) and N131(IgA2) are all nonfucosylated. No glycans were detected on the remaining sites and are believed to unoccupied in our protein standard.
In summary, a comprehensive glycan map for 43 glycosites from the 10 most abundant plasma glycoproteins has been obtained using different proteolysis approaches and the in-house software, GPFinder. We showed that glycoforms are very specific on different glycosites. This study is part of a larger one where we plan to map protein glycosylation in plasma proteins beginning with the most abundant proteins. The site-specific glycosylation for low abundant plasma proteins will be investigated using protein enrichment methods to facilitate their analyses.