Detailed O-glycomics of the Muc2 mucin from colon of wild-type, core 1- and core 3-transferase-deficient mice highlights differences compared with human MUC2

Kristina A Thomsson 1 Jessica M Holmn-Larsson 1 Jonas ngstrm 1 Malin EV Johansson 1 Lijun Xia 0 Gunnar C Hansson 1 0 Cardiovascular Biology Research Program, Oklahoma Medical Research Foundation , Oklahoma City, OK, USA 1 Department of Medical Biochemistry, University of Gothenburg , PO Box 440, 405 30 Gothenburg, Sweden The heavily O-glycosylated mucin MUC2 constitutes the major protein in the mucosal layer that acts as a physical barrier protecting the epithelial layer in the colon. In this study, Muc2 was purified from mucosal scrapings from the colon of wild-type (WT) mice, core 3 transferase knockout (C3Gnt/) mice and intestinal epithelial cell-specific core 1 knockout (IEC C1Galt1/) mice. The Muc2 O-glycans were released by reductive -elimination and analyzed with liquid chromatography-mass spectrometry in the negativeion mode. Muc2 from the distal colon of WT and C3Gnt/knockout mice carried a mixture of core 1- or core 2-type glycans, whereas Muc2 from IEC C1Galt1/ mice carried highly sialylated core 3- and core 4-type glycans. A large portion of NeuAc in all mouse models was positioned on disialylated N-acetyllactosamine units, an epitope not reported on human colonic MUC2. Mass spectra and proton NMR spectroscopy revealed an abundant NeuAc linked to internally positioned N-acetylglucosamine on colonic murine Muc2, which also differs markedly from human MUC2. Our results highlight that murine colonic Muc2 O-glycosylation is substantially different from human MUC2, which could be one explanation for the different commensal microbiota of these two species. Introduction Dense O-glycosylation forms a protective coat around the multimerizing MUC2 protein backbone, which constitutes the 1To whom correspondence should be addressed: Tel: +46-31-7863707; Fax: +46-31-416108; e-mail: major protein of the mucosal layers that lines the colon lumen. The outer loose mucosal layer in the distal colon harbors the symbiotic commensal microbiota, whereas the inner firmly attached layer is composed of a dense MUC2 network, impenetrable for the bacteria (Johansson et al. 2008). The importance of separating the intestinal bacteria from the epithelium is illustrated by the most commonly used colitis model in mice, 35% dextran sodium sulfate (DSS) in the drinking water. DSS inflammation is observed after 35 days, but already at the first contact of DSS with the inner mucus layer of the colon, it becomes permeable to bacteria (Johansson et al. 2010). The outer and loose layer is formed from the firm layer and is continuously renewed from the inner layer. The O-glycans constitute 80% of the mass of MUC2 and are linked to serine and threonine residues in PTS domains (rich in prolineserine threonine) forming the mucin domains. The dense glycosylation makes the protein gel network of the inner layer resistant toward proteolytic degradation; however, in the outer, loosely attached layer, the glycans can act as a energy source as well as providing binding sites for the commensal bacteria, which have glycan adhesins and enzymes that can degrade glycans. The presence of the commensal bacteria flora in the colon may have several advantages also for the host as they degrade both food and mucin saccharides (Backhed et al. 2005) and may suppress the colonization of pathogens. We have previously shown that colonic MUC2 O-glycosylation in human sigmoideum is largely blood group independent and identical between individuals (Larsson et al. 2009). We currently aim to address whether MUC2 O-glycosylation could be of importance for the colonization of the commensal flora (Rawls et al. 2006). The O-glycans are added to the protein in the Golgi, and they are made up by the concerted actions of glycosyltransferases adding one monosaccharide after the other, building up linear or branched sequences. On mucins, hundreds of different O-glycans can be found (Larsson et al. 2009). The terminal monosaccharide residues can be further modified by, for example, sulfate or acetyl groups. In order to study the impact of protein O-glycosylation, various mouse models have been designed by targeted deletion of glycosyltransferases. The O-glycan biosynthesis is initiated by the addition of an N-acetylgalactosamine (GalNAc) residue to serine or threonine in the protein backbone, forming the Tn-antigen, and is then commonly modified by core transferases, where the cores 1, 2, 3 and 4 are the most abundant (Figure 1A). These cores are formed by the core 1 1,3-galactosyltransferase (C1galT1, also known as the T-synthase) adding a Gal forming the core 1 glycan Gal1-3GalNAc-Ser/Thr, the core 3 1,3 N-acetylglucosaminyltransferase (C3GnT) forming GlcNAc1-3GalNAc-Ser/Thr and additional core 2 and core 4 transferases acting on these precursors. The IEC C1Galt1/ mouse developed spontaneous colitis (Fu et al. 2011). The core 2 1,6 N-acetylglucosaminyltransferase 2 (C2Gnt2/) and the core 3 (C3Gnt/) knockout mice both displayed increased susceptibility to colitis in the DSS colitis model (An et al. 2007; Stone et al. 2009). The sulfo-transferase GlcNAc6ST2 adds sulfate to N-acetylglucosamine (GlcNAc) residues on MUC2 O-glycans, and when deleted in mice, they also showed an increased susceptibility to DSS colitis (Tobisawa et al. 2010). All these observations suggest that MUC2 O-glycosylation is important for homeostasis in the colon. Characterization of murine MUC2 O-glycosylation in the gastrointestinal tract in wild-type (WT) and knock-out mouse models has been performed previously by us, although with older and less informative approaches (Holmen et al. 2002; Thomsson et al. 2002; Hurd et al. 2005). Among the more recent studies by other groups, where modern and sensitive mass spectrometric approaches have been applied allowing more comprehensive profiling, O-glycans were extracted from various tissues including colon from WT, three core 2 knockout mouse models and from the gastric mucosa in the Fut2 null model (Magalhaes et al. 2009; Ismail et al. 2011). The approaches used were based on the analysis of permethylated derivatives of the O-glycans by matrix-assisted laser desorption ionization-mass spectrometry (MALDI-MS) or electrospray ionization mass spectrometry (ESI-MS) in the positive-ion mode, approaches that do not allow analysis of the sulfated glycans frequently found on mucins. Our laboratory has over the last years applied a different methodology for screening mucin O-glycosylation using reversed-phase graphitized carbon chromatography-liquid chromatography-mass spectrometry (LC/MS) in the negative-ion mode (Andersch-Bjorkman et al. 2007; Karlsson et al. 2009; Larsson et al. 2009; Thomsson et al. 2010). Negative-ion mode MS promotes the detection of the negatively charged glycans (sialic acids, sulfate groups) which are abundantly found on mucins. LC/MS is preceded by a small-scale preparative approach of the glycans as these are released from mucins blotted to PVDF membrane after composite gel electrophoresis ( (...truncated)