Abstract :
[en] Introduction
In source decay (ISD) is a common phenomenon occurring very rapidly during ionization process in the source of MALDI-MS instruments and resulting in the presence of well resolved peaks of fragments in mass spectrum. While they make interpretation of spectra more complex, these fragments were shown to be useful to sequence peptides and proteins. Concerning glycans, only a few reports were published, using different matrices on various samples and therefore making it difficult to compare.
In this context, the goal of this work is to perform a systematic study allowing to define optimal conditions to induce ISD of glycans or, inversely, to minimize this phenomenon in the study of more complex mixtures.
Methods
Glycans were purchased from Sigma-Aldrich. Iodomethane was used in DMSO/NaOH to permethylate the glycans. This reaction was stopped by water and permethylated glycans were extracted by chloroform. Spectra were recorded on a Bruker Ultraflex II in positive ion mode. 2,5-dihydroxybenzoic acid (DHB) and 6-aza-2-thiothymine (ATT) were prepared at 20 mg/ml in 50 % acetonitrile, 0.1 % formic acid solution. 9-aminoacridine (9-AA) was dissolved at saturation in a 50 % acetonitrile, 0.1 % formic acid and further diluted 4 times in the same solution. α-Cyano-4-hydroxycinnamic (HCCA) acid was prepared at 20 mg/ml in 97 % acetone, 0,1 % formic acid. In some spots, LiI was added to obtain Li+ adducts instead of Na+ adducts.
Preliminary data
In source fragmentation of permethylated Lacto-N-difucoHexaose I and LS tetrasaccharide B was first studied in DHB. While the MS/MS of the Na+ adducts of these compounds (performed by LID) produces intenses B and Y fragments, those resulting from in source fragmentation are mainly oxonium ions, resulting from the cleavage of a glycosylic bond without any exchange of hydrogen atoms. These oxonium fragments were also obtained for lithium adducts. It was previously described that these fragments are produced by the cleavage of a protonated glycosidic bond. These ions carry their positive charge on a trivalent oxygen atom and are therefore not present on the spectra as sodium adducts. Since the peaks of protonated glycans are very low in MALDI spectra, it would indicate that protonation of glycosidic bonds of permethylated glycans would strongly favor a fragmentation reaction. Different matrices were tested to compare their ability to induce in source fragmentation of permethylated glycans. Interestingly, ATT gave similar results comparing to DHB while HCCA showed a lesser ability to promote in source fragmentation. However, the most striking result came from the use of 9-AA. This matrix, which is usually used in negative ion mode, was able to produce easily sodium adducts ions of permethylated glycan with a satisfying signal to noise ratio in positive ion mode. Moreover, practically no in source fragmentation was observed with this matrix. The few produced fragments were B ions but no oxonium ions were detected. Presence of these B fragments was increased for Li+ adducts. As 9-AA is the most basic of tested matrices, the absence of oxonium ions could result from its inability to transfer protons to the glycosidic bond of permethylated glycans. 9-AA could therefore become a matrix of choice to study complex mixtures of glycans, by reducing artefact peaks produced by ISD.
Novel aspect
ISD of permethylated glycans is induced by DHB while 9-AA strongly favors the presence of molecular ions.
