N-glycans are key players mediating cell–cell communication in the immune system, interacting with glycan-binding proteins. In the present article, we discuss key themes that are emerging from the structural analysis of complex-type N-linked glycans from human and murine immune cell lines, employing high-sensitivity MALDI (matrix-assisted laser desorption ionization)–TOF (time-of-flight) MS technology. Particular focus is given to terminal epitopes, the abundance of multiply branched N-glycans and how glycosylation can affect human health in diseases such as congenital neutropenia and glycogen storage disease.
- congenital disorder of glycosylation
- congenital neutropenia
- glycogen storage disease 1b (GSD 1b)
- mass spectrometry
- matrix-assisted laser-desorption ionization–time-of-flight (MALDI–TOF)
Glycans attached to proteins and lipids on the surfaces of immune cells serve as ligands for three families of glycan binding proteins [C-type lectins, siglecs (sialic acid-binding, immunoglobulin-like lectins) and galectins] [1–3] which play vital roles in mediating cell–cell communication in the immune system . The last decade has witnessed an exponential growth in our understanding of the roles glycans play in immune events. Much of this research has been fuelled by the CFG (Consortium for Functional Glycomics) which, in 2001, received long-term funding from the NIH (National Institutes of Health)'s NIGMS (National Institute of General Medical Sciences) to “define the paradigms by which protein–carbohydrate interactions mediate cell communication”. To this end, the CFG has developed tools and resources to support worldwide glycobiology research. These include the development of open-access repositories populated with glycomics data for mouse and human immune cells (see http://www.functionalglycomics.org/glycomics/publicdata/glycoprofiling-new.jsp). In the present article, we discuss some of the key themes that are emerging from profiling the neutral and sialylated complex-type N-glycans of these cells. We also review recent research on severe congenital neutropenias which is being greatly facilitated by knowledge of neutrophil glycomes.
Immune cell expression of N-linked sialyl Lewisx
E-, P- and L-selectin are C-type lectins that mediate leucocyte trafficking and recruitment [5,6]. All three recognize the sLex [sialyl Lewisx or NeuAcα2-3Galβ1-4(Fucα1-3)GlcNAc] motif, but each has its own specific requirements for the molecular context of this motif. For example sulfate moieties are crucial for P- and L-selectin [7–10] binding, but not for E-selectin . The best understood selectin ligands, for example PSGL-1 (P-selectin glycoprotein ligand 1), display their sLex on O-glycans [11,12]. In contrast, E-selectin binding does not require sulfation, and at least two of its physiological ligands, ESL-1 (E-selectin ligand 1) and CD44, involve N-linked sLex [13,14]. It is noteworthy that sLex was discovered in humans when neutrophil N-glycosylation was first studied by MS a quarter of a century ago [15,16]. Indeed, even though the use of the term ‘glycomics’, as a descriptor for MS determination of glycan populations, only became fashionable in the early years of the 21st Century, its roots go back to the exploitation of FAB (fast atom bombardment) technology for the characterization of leucocyte glycosylation in the early 1980s [15,16]. These early FAB studies focused on a subsection of the neutrophil glycome that was enriched in extended polylactosamine antennae, and established the sugar sequences terminating these antennae. Only with the advent of MALDI–TOF (matrix-assisted laser-desorption ionization–time-of-flight) MS technology  has it become possible to define the structures of the multitude of individual glycans that constitute the neutrophil N-glycome .
As an example, partial MALDI–TOF MS data of neutral N-glycans from human and murine neutrophil cells are presented in Figure 1. Glycomic profiling suggests that the N-glycome of human neutrophil cells is composed mainly of complex bi- tri- and tetra-antennary N-glycans (Figure 1A and insets), of which the latter are significantly extended with LacNAc (N-acetyl-lactosamine, Galβ1-4GlcNAc) repeating units (for complete spectra, see the CFG website at http://www.functionalglycomics.org/). Their antennae are predominantly terminated with NeuAc (N-acetylneuraminic acid) and Lex [Lewisx or Galβ1-4(Fucα1-3)GlcNAc] epitopes, although these are almost exclusively observed on different antennae. sLex epitopes on human N-glycans were detected (m/z 3140 and 3950; Figure 1A, insets), but were only of extremely minor abundance within the N-glycome (less that 0.1% of the total N-glycan population).
Interestingly, despite the mouse being widely used as a model for selectin function, no sLex epitopes have been thus far observed among the neutrophil-derived N-glycans (Figure 1B). The vanishingly small levels of fucose detected on murine PSGL-1 cell lines also suggest that caution should be used when speculating about the repertoire of glycan structures present in specific cells, particularly when such speculation is based solely on general chemical analyses of the glycans . CFG data on mouse neutrophils reveal that the N-glycans are composed mainly of complex bi-antennary, and to a lesser extent tri-antennary, structures, whereas tetra-antennary N-glycans appear to be absent. Terminal epitopes consist of the sialic acids NeuAc and NeuGc (N-glycolylneuraminic acid) and Galα1-3Gal structures. Fucose residues are presented only in the context of core chitobiose substitutions, in contrast with human neutrophils, in which they are observed in significant quantities as antennal decorations at the non-reducing end. Similar N-glycomic profiling has also been performed on murine CD4+ and CD8+ T-lymphocytes  (see below).
Immune cell expression of N-linked Lex
The Lex epitope, a branched terminal structure formed by the action of an α1-3 fucosyltransferase on the internal GlcNAc of LacNAc extensions, is a potent ligand for the C-type lectin DC-SIGN (dendritic cell-specific intracellular adhesion molecule-3 grabbing non-integrin) [21–23]. Expression of DC-SIGN is observed mainly on dendritic cells and dendritic cell-like macrophages. Murine immune cells such as eosinophils and macrophages do express Lex structures as terminal N-linked determinants, although only ever at very low abundance and on relatively simple bi-antennary structures, despite extensive expression of the trisaccharide in a wide range of tissues, including kidney, brain, intestine and testes . In contrast, human immune cell populations express significant levels of Lex, with both granulocytic cells (neutrophils, eosinophils and basophils) and agranulocytic cells (monocytes, macrophages, T-cells, B-cells and natural killer cells) displaying substantial amounts of the epitope. In cases where the immune cell glycome is relatively complex (as observed in human neutrophils, for example ), Lex structures are displayed at very high levels with up to four such epitopes present on the larger tetra-antennary glycans. The reason for such a significant disparity in expression levels of Lex between the two mammalian systems is not clear, although the additional selective pressure of an active α1,3GT (UDP-Gal:β-galactosyl α1-3-galactosyltransferase) in the murine model is likely to contribute towards the observed inequality (see below).
Immune cell expression of branched and extended N-glycans
Galectins are animal lectins that have high affinity for LacNAc residues  and, depending on their type, they can form ordered arrays of complexes when they bind to multivalent glyconjugates (lattices) . Recent research has suggested that extended antennae and branching are important factors for the multivalent presentation that is critical for stable galectin-mediated lattices. Both extended antenna-branching and the concentration or density of host lectins are tightly controlled in order to facilitate lectin–glycan cross-linking and signalling in cells .
CFG data have shown that virtually all human immune cells are abundant in multiply branched N-glycans, extended with linear LacNAc units and terminated with NeuAc and fucose residues. These immune cells regulate and constantly fine-tune the density of LacNAc units so as to meet lectin-binding requirements for signal transduction. A characteristic example is that of dendritic cells, in which, upon maturation, the cells increase the abundance of poly-LacNAc elongated glycans so as to increase binding with galectin-3 . Galectin-3–LacNAc lattices have also been shown to control T-cell activation on murine immune cells [29,30]. However, complex N-glycans derived from murine immune cells examined thus far are considerably less branched and have a much lower number of LacNAc extensions than the comparable human immune cell populations.
Experimental insights into these issues can be gained from MS/MS (tandem MS) experiments. Thus, MALDI–TOF/TOF MS/MS analysis along with specific enzymatic digestions and GC (gas chromatography)–MS linkage analysis provides relatively facile means by which complex structures from a mixture may be characterized  (Figure 2). MS approaches provide direct structural evidence, enabling questions such as how many antennae the structure possesses and whether the LacNAc extensions are on a single branch of the N-glycans to be addressed, therefore correlating structure with function.
Immune cell expression of truncated N-glycans
The recently described C-type lectin, designated LSECtin (liver and lymph node sinusoidal endothelial cell C-type lectin) , demonstrates high binding selectivity towards N-glycan ligands exhibiting terminal GlcNAcβ1-2Man structures. Indeed, the affinity of the molecule for this disaccharide is approximately 3.5 μM, one of the most specific and strongest interactions of any known lectin or lectin-like molecule to date . Experiments involving the intravenous introduction of recombinant LSECtin have demonstrated that it can have a strong anti-inflammatory effect in cases of acute liver injury, binding to CD44 displayed on T-cells within the liver, suppressing their activation . With LSECtin expression having been described on human peripheral blood and thymic dendritic cells, as well as monocyte-derived macrophages , it is entirely possible that similar immunomodulatory functions could extend beyond the liver . The expression among immune cell populations of unusually high levels of truncated or part-processed N-glycans bearing such exposed GlcNAc structures is therefore of significant interest. Murine immune cell populations profiled thus far have uniformly low levels of GlcNAcβ1-2Man structures, with the exception of T-cells and some types of non-circulating B-cells. The human immune system generally exhibits a similar trend towards low levels of truncated structures, with distinct exceptions in the cases of principle antigenic effector cells (basophils and eosinophils, although, interestingly, not mast cells) and T-cells where very substantive quantities of terminal GlcNAc are observed. Interactions of these ligands on human immune cells with potential binding partners such as LSECtin could dictate cellular function and have important implications for the pathogenesis of allergic or inflammatory diseases.
Immune cell expression of bisected N-glycans
Bisecting GlcNAc, a branching modification to complex type N-glycans carried out by the action of GnT (N-acetylglucosaminyltransferase)-III (or mgat3), catalysing the addition of a GlcNAc residue to the 4 position of the mannose residue at the base of the trimannosyl core, has a profound effect upon the conformation of glycans so modified . This has implications in both the further processing and elongation of the putative glycans, since bisected N-glycans are no longer recognized as substrates for other glycosyltransferases, suppressing further branching formation by GnT-IV or GnT-V and the addition of core fucose via the action of Fut8 (fucosyltransferase 8) . These conformational alterations will of course also influence the presentation of ligands in the context of glycan-binding proteins. Examples of such behavioural modifications upon introduction or overexpression of the GnT-III gene include the blocking of natural killer cell cytotoxicity , the up-regulation of EGFR (epidermal growth factor receptor) endocytosis , protection of mammary epithelial cells from tumour progression  and the inhibition of tumour metastasis via enhancement of integrin-modulated cell–cell adhesion . Immune cells that express some bisected structures as revealed by glycomic analyses include B-cells and T-lymphocytes in the mouse, and B-cells, T-lymphocytes, natural killer cells, basophils, mast cells and eosinophils in the human system.
Sialylation and α-galactose on murine immune cells
There is considerable commonality between the N-glycan biosynthetic pathways of humans and mice. However, there are two prominent exceptions. The first is that murine N-linked glycans can be capped by two types of sialic acids: NeuAc and NeuGc. This is due to the fact that the enzyme which converts NeuAc into NeuGc, CMP-NeuAc hydroxylase, in humans contains an inactivating mutation . The second is that murine N-linked glycans can be capped by the Galα1-3Gal epitope. The lack of this epitope in humans is caused by an inactivation mutation in α1,3GT . Both NeuGc- and Galα1-3Gal-capped N-glycans are abundantly expressed by murine immune cells such as neutrophils, macrophages, eosinophils, B-cells and T-cells. As both NeuGc terminal sialylation and α-Gal capping compete for the same terminal position in N-glycans, evidence has emerged that implicates the dynamic interchange of these substituents with immune cell function. Both murine CD4+ and CD8+ T-lymphocytes undergo a dramatic remodelling of their N-glycans upon cytokine activation. The most striking feature of this remodelling is a reduction in terminal NeuGcα2-6Gal sequences with a concurrent increase in Galα1-3Gal. These structural changes were mediated by a reduction in the expression of the sialyltransferase ST6Gal I and an increase in the expression of α1,3GT . As stated above, such functional changes in immune cell glycosylation will not occur in humans because of the lack of expression of NeuGc and Galα1-3Gal structures. It therefore remains one of the most enigmatic questions in glycoimmunology as to what mechanism human cells utilize as an alternative to NeuGc and Galα1-3Gal structures. One possibility is the increased levels of fucosylated structures such as Lex and sLex which are present at much higher levels in human immune cells than murine immune cells (see above).
Congenital neutropenia, GSD (glycogen storage disease) and glycosylation
Comparison of N-glycomic profiles from immune cells can be a valuable tool for detecting various anomalies in cell metabolism. Two characteristic cases include that of SCN (severe congenital neutropenia), a disorder associated with life-threatening bacterial infection affecting neutrophil function [45,46] and GSD type 1b [47,48]. Both of these disorders have been associated with neutropenia and various organ malformations. Analysis of the N-glycomes of neutrophils from the above cases provides a novel, potentially unifying, mechanism for most of the neutrophil dysfunctions recognized to date.
N-glycomic profiling indicated that neutrophils from patients with mutations in either the G6PC3 (glucose-6-phosphatase catabolic 3)  or G6PT (glucose-6-phosphate translocase)  gene exhibit a major common defect of glycosylation mainly on their N-glycans . Their complex type N-glycans were found to be agalactosylated (truncated glycosylation), indicating a failure of galactosylation (Figure 3A). As a result, they were profoundly deficient in LacNAc repeats, sialylation and Lex epitopes. The lack of such structural features, as discussed above, reduces the number of glycan ligands available for binding with selectins, siglecs and galectins, with profound implications for cellular function .
The biochemical cause of agalactosylation remains unknown, and further studies are required to delineate this phenomenon. However, these dysfunctions appear to be a part of the same metabolic pathway, meaning that, in either case, both mutated proteins affect the cell metabolism of G6P (glucose 6-phosphate) in neutrophils (Figure 3B). In healthy neutrophils, once G6P has entered the ER (endoplasmic reticulum) and the G6PC3 enzyme has produced glucose and Pi, the latter exits the ER via the G6P transporter (which has been suggested to be an antiporter ) exchanging Pi for a new G6P (Figure 3B). When either of the G6PC3 or G6PT proteins are dysfunctional, then it appears that the above metabolism is disrupted, resulting in the agalactosylated structures mentioned above.
These dysfunctions, suggested as a novel CDG (congenital disorder of glycosylation), would be impossible to detect with the classic diagnostic blood test which involves checking the glycosylation status of serum glycoproteins such as transferrin . Since blood glycoproteins are mainly produced by the liver, which expresses a different isoenzyme (G6PC1) of the G6P, serum glycosylation will read as normal, even in the patients with G6PC3 mutation.
Our greater appreciation of the increasingly important role that glycosylation plays in the immune system is being underpinned by more detailed structural knowledge. The driver for these advances in glycan structural definition has been the application of improved mass spectrometric analytical methodologies such as MALDI–TOF MS. Data emerging from such studies are opening up new areas of research and providing new insights into both the human and mouse immune systems.
This work was supported by the Analytical Glycotechnology Core (Core C) of the Consortium for Functional Glycomics [grant number GM 62116] and the Biotechnology and Biological Sciences Research Council [grant number BBF0083091].
Joint Sino–U.K. Protein Symposium: a Meeting to Celebrate the Centenary of the Biochemical Society: A Biochemical Society Focused Meeting held at Shanghai University, Shanghai, China, 5–7 May 2011. Organized by Tom Blundell (Cambridge, U.K.), Zengyi Chang (Peking University, China), Ian Dransfield (Edinburgh, U.K.), Neil Isaacs (Glasgow, U.K.), Glenn King (University of Queensland, Australia), Sheena Radford (Leeds, U.K.), Zihe Rao (Nankai University, China), Yi-Gong Shi (Tsinghua University, China), Chihchen (Zhizhen) Wang (Institute of Biophysics, Chinese Academy of Sciences, China), Jiarui Wu (Shanghai Institute of Biological Sciences, China) and Xian-En Zhang (Ministry of Science and Technology, China). Edited by Zengyi Chang and Neil Isaacs.
Abbreviations: CFG, Consortium for Functional Glycomics; DC-SIGN, dendritic cell-specific intracellular adhesion molecule-3 grabbing non-integrin; ER, endoplasmic reticulum; FAB, fast atom bombardment; G6P, glucose 6-phosphate; G6PC3, glucose-6-phosphatase catabolic 3; G6PT, glucose-6-phosphate translocase; GC, gas chromatography; GnT, N-acetylglucosaminyltransferase; GSD, glycogen storage disease; α1,3GT, UDP-Gal:β-galactosyl α1-3-galactosyltransferase; LacNAc, N-acetyl-lactosamine; Lex, Lewisx; LSECtin, liver and lymph node sinusoidal endothelial cell C-type lectin; MALDI–TOF, matrix-assisted laser-desorption ionization–time-of-flight; MS/MS, tandem MS; NeuAc, N-acetylneuraminic acid; NeuGc, N-glycolylneuraminic acid; PSGL-1, P-selectin glycoprotein ligand 1; siglec, sialic acid-binding, immunoglobulin-like lectin; sLex, sialyl Lewisx
- © The Authors Journal compilation © 2011 Biochemical Society