Negative and positive ion matrix‐assisted laser desorption/ionization time‐of‐flight mass spectrometry and positive ion nano‐electrospray ionization quadrupole ion trap mass spectrometry of peptidoglycan fragments isolated from various Bacillus species

A general approach for the detailed characterization of sodium borohydride‐reduced peptidoglycan fragments (syn. muropeptides), produced by muramidase digestion of the purified sacculus isolated from Bacillus subtilis (vegetative cell form of the wild type and a dacA mutant) and Bacillus megaterium (endospore form), is outlined based on UV matrix‐assisted laser desorption/ionization time‐of‐flight (MALDI‐TOF) and nano‐electrospray ionization (nESI) quadrupole ion trap (QIT) mass spectrometry (MS). After enzymatic digestion and reduction of the resulting muropeptides, the complex glycopeptide mixture was separated and fractionated by reversed‐phase high‐performance liquid chromatography. Prior to mass spectrometric analysis, the muropeptide samples were subjected to a desalting step and an aliquot was taken for amino acid analysis. Initial molecular mass determination of these peptidoglycan fragments (ranging from monomeric to tetrameric muropeptides) was performed by positive and negative ion MALDI‐MS using the thin‐layer technique with the matrix α‐cyano‐4‐hydroxycinnamic acid. The results demonstrated that for the fast molecular mass determination of large sample numbers in the 0.8–10 pmol range and with a mass accuracy of ±0.07%, negative ion MALDI‐MS in the linear TOF mode is the method of choice. After this kind of muropeptide screening often a detailed primary structural analysis is required owing to ambiguous data. Structural data could be obtained from peptidoglycan monomers by post‐source decay (PSD) fragment ion analysis, but not from dimers or higher oligomers and not with the necessary sensitivity. Multistage collision‐induced dissociation (CID) experiments performed on an nESI‐QIT instrument were found to be the superior method for structural characterization of not only monomeric but also of dimeric and trimeric muropeptides. Up to MS 4 experiments were sometimes necessary to obtain unambiguous structural information. Three examples are presented: (a) CID MS n ( n = 2–4) of a peptidoglycan monomer (disaccharide‐tripeptide) isolated from B. subtilis (wild type, vegetative cell form), (b) CID MS n ( n = 2–4) of a peptidoglycan dimer (bis‐disaccharide‐tetrapentapeptide) obtained from a B. subtilis mutant (vegetative cell form) and (c) CID MS 2 of a peptidoglycan trimer (a linear hexasaccharide with two peptide side chains) isolated from the spore cortex of B. megaterium . All MS n experiments were performed on singly charged precursor ions and the MS 2 spectra were dominated by fragments derived from interglycosidic bond cleavages. MS 3 and MS 4 spectra exhibited mainly peptide moiety fragment ions. In case of the bis‐disaccharide‐tetrapentapeptide, the peptide branching point could be determined based on MS 3 and MS 4 spectra. The results demonstrate the utility of nESI‐QIT‐MS towards the facile determination of the glycan sequence, the peptide linkage and the peptide sequence and branching of purified muropeptides (monomeric up to trimeric forms). The wealth of structural information generated by nESI‐QIT‐MS n is unsurpassed by any other individual technique. Copyright © 2001 John Wiley & Sons, Ltd.