Glycosylation Patterns of Proteins Studied by Liquid Chromatography-Mass Spectrometry and Bioinformatic Tools

Due to their extensive structural heterogeneity, the elucidation of glycosylation patterns in glycoproteins such as the subunits of human chorionic gonadotropin (hCG), hCG-α and hCG-β remains one of the most challenging problems in the proteomic analysis of posttranslational modifications. In consequence, glycosylation is usually studied after decomposition of the intact proteins to the proteolytic peptide level. However, by this approach all information about the combination of the different glycopeptides in the intact protein is lost. In this study we have, therefore, attempted to combine the results of glycan identification after tryptic digestion with molecular mass measurements on the intact glycoproteins. Despite the extremely high number of possible combinations of the glycans identified in the tryptic peptides by high-performance liquid chromatography-mass spectrometry (> 1000 for hCGα and > 10.000 for hCG-β), the mass spectra of intact hCG-α and CG-β revealed only a limited number of glycoforms present in hCG preparations from pools of pregnancy urines. Peak annotations for hCG-α  were performed with the help of an algorithm that generates a database containing all possible modifications of the proteins (inclusive possible artificial modifications such as oxidation or truncation) and subsequent searches for combinations fitting the mass difference between the polypeptide backbone and the measured molecular masses. Fourteen different glycoforms of CG-α, including methionine-oxidized and N-terminally truncated forms, were readily identified. For hCG-β, however, the relatively high mass accuracy of ± 5 Da was still insufficient to unambiguously assign the possible combinations of posttranslational modifications. Finally, the mass spectrometric fingerprints of the intact molecules were shown to be very useful for the characterization of glycosylation patterns in different CG preparations. Dagstuhl Seminar Proceedings 05471 Computational Proteomics http://drops.dagstuhl.de/opus/volltexte/2006/543