156 | Mol. Omics, 2020, 16, 156--164 This journal is © The Royal Society of Chemistry 2020 Cite this: Mol. Omics, 2020, 16, 156 Novel O-linked sialoglycan structures in human urinary glycoproteins† Adam Pap, ab Ervin Tasnadi, cd Katalin F. Medzihradszky a and Zsuzsanna Darula* a Glycopeptides represent cross-linked structures between chemically and physically different biomolecules. Mass spectrometric analysis of O-glycopeptides may reveal the identity of the peptide, the composition of the glycan and even the connection between certain sugar units, but usually only the combination of different MS/MS techniques provides sufficient information for reliable assignment. Currently, HCD analysis followed by diagnostic sugar fragment-triggered ETD or EThcD experiments is the most promising data acquisition protocol. However, the information content of the different MS/MS data is handled separately by search engines. We are convinced that these data should be used in concert, as we demonstrate in the present study. First, glycopeptides bearing the most common glycans can be identified from EThcD and/or HCD data. Then, searching for Y 0 (the gas-phase deglycosylated peptide) in HCD spectra, the potential glycoforms of these glycopeptides could be lined up. Finally, these spectra and the corresponding EThcD data can be used to verify or discard the tentative assignments and to obtain further structural information about the glycans. We present 18 novel human urinary sialoglycan structures deciphered using this approach. To accomplish this in an automated fashion further software development is necessary. Introduction Golgi-derived protein glycosylation is among the most common post-translational modifications (PTMs). In the past three decades mass spectrometry has become the method of choice for PTM analysis, even in a high throughput manner. 1 Glyco- peptides represent crosslinked molecules from two biopolymer families with unique chemical and physical features that lead to different fragmentation behaviors depending on the type of MS/MS activation. Glycosidic bonds are weaker than peptide bonds. Thus, depending on the type of collisional activation, either only glycan fragments are formed mostly via single bond cleavages (ion trap CID), 2,3 or some peptide fragmentation and smaller, less informative glycan fragments are observed (beam-type CID/HCD). 3–5 In both cases, glycan fragmentation yields characteristic Y ions ([Nomenclature 6 ]), the most abundant ones tend to be Y 1 for N-glycopeptides 7 and Y 0 for O-glycopeptides. 4 However, in beam-type CID peptide backbone fragments typi- cally undergo gas phase deglycosylation without leaving any mark on the formerly modified Ser or Thr residue(s). 4 Thus, these spectra may contain sufficient information for sequence identification, but assigning the glycosylation site is usually not possible. In the alternative, electron-transfer dissociation (ETD) activation almost exclusively peptide backbone fragments are formed, 8 and very limited glycan fragmentation is observed (mostly sialic acid losses). 9 In typical intact glycopeptide analysis mixtures of different complexity, depending on the enrichment method applied, are submitted to automated LC/MS/MS analyses, and the resulting data files are interpreted by different search engines that were more or less attuned to the specific task of glycopeptide identification. 10–13 These searches produce a long list of glycopeptide assignments. Unfortunately, these lists may not be very reliable. As is the case with cross- linked peptides, and particularly because of reasons outlined above, one frequently does not get sufficient information in a single spectrum for the unambiguous assignment of both com- ponents or sometimes all components, since in O-glycopeptides multiple modifications frequently occur in a single sequence and the decorations may represent different glycans. In most cases, the search engine considers all permutations from the combination of peptides and glycans, and picks glycopeptide candidates based on the observed precursor mass and any observed peptide fragments, with the glycan assigned solely based on the mass difference between the precursor mass and a Laboratory of Proteomics Research, Biological Research Centre, Temesvari krt. 62, H-6726 Szeged, Hungary. E-mail: darula.zsuzsanna@brc.hu b Doctoral School in Biology, Faculty of Science and Informatics, University of Szeged, Kozep fasor 52, H-6726 Szeged, Hungary c Systems and Synthetic Biology Unit, Biological Research Centre, Szeged, Hungary, Temesvari krt. 62, H-6726 Szeged, Hungary d Doctoral School of Computer Science, Faculty of Science and Informatics, University of Szeged, Kozep fasor 52, H-6726 Szeged, Hungary † Electronic supplementary information (ESI) available. See DOI: 10.1039/ c9mo00160c Received 29th October 2019, Accepted 27th January 2020 DOI: 10.1039/c9mo00160c rsc.li/molomics Molecular Omics RESEARCH ARTICLE Published on 29 January 2020. Downloaded on 9/29/2021 8:28:46 AM. View Article Online View Journal | View Issue