MASS SPECTROMETRIC GLYCAN REARRANGEMENTS Manfred Wuhrer,* Andre ´ M. Deelder, and Yuri E.M. van der Burgt Leiden University Medical Center, Biomolecular Mass Spectrometry Unit, Department of Parasitology, Leiden, The Netherlands Received 27 October 2010; revised 28 February 2011; accepted 28 March 2011 Published online 10 May 2011 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/mas.20337 Mass spectrometric rearrangement reactions have been reported for a large variety of compounds such as peptides, lipids, and carbohydrates. In the case of carbohydrates this phenomenon has been described as internal residue loss. Resulting fragment ions may be misinterpreted as fragments arising from conven- tional glycosidic bond cleavages, which may result in incorrect structural assignment. Therefore, awareness of the occurrence of glycan rearrangements is important for avoiding misinterpreta- tion of tandem mass spectra. In this review mass spectrometric rearrangements of both derivatized and underivatized (native) oligosaccharide structures are discussed. Similar phenomena have been reported for glycopeptides, labeled glycan structures and other biomolecules containing a carbohydrate part. Rear- rangements in oligosaccharides and glycoconjugates have been observed with different types of mass spectrometers. Most of the observed carbohydrate rearrangement reactions appear to be linked to the presence of a proton. Hence, tandem mass spectro- metric analysis of alkali adducts or deprotonated ions often prevents rearrangement reactions, while they may happen with high efficacy with protonated glycoconjugates. # 2011 Wiley Periodicals, Inc., Mass Spec Rev 30:664–680, 2011 Keywords: carbohydrate; derivatization; glycopeptide; inter- nal residue loss; oligosaccharide I. REARRANGEMENTS IN MASS SPECTROMETRY In 1959, McLafferty stated in his paper entitled ‘‘Mass spectrometric analysis—molecular rearrangements’’ that ‘‘Structure determination by mass spectrometry (MS) has been hampered by the relatively unpredictable possibilities of mol- ecular rearrangement.’’ In this work electron impact (EI) frag- mentation behavior of small organic molecules such as ketones, alcohols and ethers was analyzed and it was con- cluded that ‘‘examination of a large number of spectra shows that rearrangements are very common. There are few molecules (other than diatomic) that do not exhibit this phenomenon in their mass spectra to some extent.’’ Many of these rearrangements are a result of hydrogen transfer upon heterolytic cleavage of covalent bonds. One example is the transfer of a g-hydrogen atom in carbonyl compounds during EI-induced carbon–carbon bond cleavage to yield an enol and an alkene species, which is commonly known as the McLafferty rearrangement (McLafferty, 1959; Nibbering, 2004). After this first report, mass spectrometric rearrange- ments have been observed in various compounds, such as fatty acid methyl esters (McLafferty, 1959; Ryhage & Stenhagen, 1960; Vidavsky et al., 1994; Nibbering, 2004), polyethylene glycols (Cerda et al., 2002) and peptides (Paizs & Suhai, 2005; Polfer et al., 2007; Bleiholder et al., 2008). It is of great importance to determine the amount of this so-called scram- bling in peptide fragmentation since MS-based proteomics largely relies on peptide sequencing for protein identifications. Peptide sequence scrambling often is overlooked, but upon cursory examination it is clearly found (Harrison, 2008; Molesworth, Osburn, & Van Stipdonk, 2009; Good, Marin- Vicente, & Zubarev, 2010; Saminathan et al., 2010). Likewise, rearrangements during fragmentation of carbohydrates have been observed for a large variety of structures. Sometimes the rearrangement processes even dominate the fragmentation pathways in tandem MS. Therefore, in tandem MS of glyco- conjugates it is important to take the possibility of rearrange- ments into account. McLafferty emphasized the difference between a random and a specific molecular rearrangement. The random or ‘‘reshuffling process’’ is not useful for structure determination since the outcome cannot be predicted. However, the specific process that involves a predictable rearrangement may yield valuable structural information. Here we review specific re- arrangement processes that have been reported and rational- ized in mass spectrometric determinations of carbohydrate structures. On the one hand knowledge of these carbohydrate rearrangements contributes to fragment assignments in tandem MS spectra and on the other hand it prevents misinterpretation and incorrect sequencing. II. EARLY OBSERVATIONS OF MASS SPECTROMETRIC CARBOHYDRATE REARRANGEMENTS The loss of internal monosaccharide residues in MS of carbo- hydrates was first reported by McNeil (1983). In chemical ion- ization (CI)-MS spectra of oligosaccharide derivatives, he observed fragment ions that were potentially misleading for sequence purposes (McNeil, 1983 and references cited therein). In CI-MS spectra of permethylated oligosaccharide alditols ions were observed at m/z-values that corresponded to those of pseudomolecular ions [M þ H] þ of smaller perme- thylated oligosaccharide alditols. Thus, one could (wrongly!) conclude that the sample was impure, that is, a mixture con- taining also smaller oligosaccharide alditols (McNeil, 1983). By studying partially ethylated, per-O-methylated oligosac- charide alditols that varied in ethylation positions, McNeil elegantly showed that these ions were actually fragment ions *Correspondence to: Manfred Wuhrer, Leiden University Medical Center, Biomolecular Mass Spectrometry Unit, Department of Para- sitology, PO Box 9600, 2300 RC Leiden, The Netherlands. E-mail: m.wuhrer@lumc.nl Mass Spectrometry Reviews, 2011, 30, 664–680 # 2011 Wiley Periodicals, Inc.