Screening for the presence of polyglycosylceramides in various tissues: partial characterization of blood group-active complex glycosphingolipids of rabbit and dog small intestines

The SPPL3-Defined Glycosphingolipid Repertoire Orchestrates HLA Class I-Mediated Immune Responses

HLA class I (HLA-I) glycoproteins drive immune responses by presenting antigens to cognate CD8+ T cells. This process is often hijacked by tumors and pathogens for immune evasion. Because options for restoring HLA-I antigen presentation are limited, we aimed to identify druggable HLA-I pathway targets. Using iterative genome-wide screens, we uncovered that the cell surface glycosphingolipid (GSL) repertoire determines effective HLA-I antigen presentation. We show that absence of the protease SPPL3 augmented B3GNT5 enzyme activity, resulting in upregulation of surface neolacto-series GSLs. These GSLs sterically impeded antibody and receptor interactions with HLA-I and diminished CD8+ T cell activation. Furthermore, a disturbed SPPL3-B3GNT5 pathway in glioma correlated with decreased patient survival. We show that the immunomodulatory effect could be reversed through GSL synthesis inhibition using clinically approved drugs. Overall, our study identifies a GSL signature that inhibits immune recognition and represents a potential therapeutic target in cancer, infection, and autoimmunity.

Genogroup IV and VI canine noroviruses interact with histo-blood group antigens

Human noroviruses (HuNV) are a significant cause of viral gastroenteritis in humans worldwide. HuNV attaches to cell surface carbohydrate structures known as histo-blood group antigens (HBGAs) prior to internalization, and HBGA polymorphism among human populations is closely linked to susceptibility to HuNV. Noroviruses are divided into 6 genogroups, with human strains grouped into genogroups I (GI), II, and IV. Canine norovirus (CNV) is a recently discovered pathogen in dogs, with strains classified into genogroups IV and VI. Whereas it is known that GI to GIII noroviruses bind to HBGAs and GV noroviruses recognize terminal sialic acid residues, the attachment factors for GIV and GVI noroviruses have not been reported. This study sought to determine the carbohydrate binding specificity of CNV and to compare it to the binding specificities of noroviruses from other genogroups. A panel of synthetic oligosaccharides were used to assess the binding specificity of CNV virus-like particles (VLPs) and identified α1,2-fucose as a key attachment factor. CNV VLP binding to canine saliva and tissue samples using enzyme-linked immunosorbent assays (ELISAs) and immunohistochemistry confirmed that α1,2-fucose-containing H and A antigens of the HBGA family were recognized by CNV. Phenotyping studies demonstrated expression of these antigens in a population of dogs. The virus-ligand interaction was further characterized using blockade studies, cell lines expressing HBGAs, and enzymatic removal of candidate carbohydrates from tissue sections. Recognition of HBGAs by CNV provides new insights into the evolution of noroviruses and raises concerns regarding the potential for zoonotic transmission of CNV to humans.

IMPORTANCE Infections with human norovirus cause acute gastroenteritis in millions of people each year worldwide. Noroviruses can also affect nonhuman species and are divided into 6 different groups based on their capsid sequences. Human noroviruses in genogroups I and II interact with histo-blood group antigen carbohydrates, bovine noroviruses (genogroup III) interact with alpha-galactosidase (α-Gal) carbohydrates, and murine norovirus (genogroup V) recognizes sialic acids. The canine-specific strains of norovirus are grouped into genogroups IV and VI, and this study is the first to characterize which carbohydrate structures they can recognize. Using canine norovirus virus-like particles, this work shows that representative genogroup IV and VI viruses can interact with histo-blood group antigens. The binding specificity of canine noroviruses is therefore very similar to that of the human norovirus strains classified into genogroups I and II. This raises interesting questions about the evolution of noroviruses and suggests it may be possible for canine norovirus to infect humans.

Glycosphingolipid functions

The combination of carbohydrate and lipid generates unusual molecules in which the two distinctive halves of the glycoconjugate influence the function of each other. Membrane glycolipids can act as primary receptors for carbohydrate binding proteins to mediate transmembrane signaling despite restriction to the outer bilayer leaflet. The extensive heterogeneity of the lipid moiety plays a significant, but still largely unknown, role in glycosphingolipid function. Potential interplay between glycolipids and their fatty acid isoforms, together with their preferential interaction with cholesterol, generates a complex mechanism for the regulation of their function in cellular physiology.

Isolation and partial characterization of Galalpha-containing polyglycosylceramides from porcine tissues

Mammalian cell surface carbohydrate antigens are present both as glycoproteins and glycolipids. Of the glycolipids, polyglycosylceramides (PGC) have very long carbohydrate chains extending out from the cell surface. Hereto, Gal alpha-terminating xenoantigens in pig tissues have been identified in glycoproteins and short chain glycolipids but no studies of the complex PGC have been performed. In this communication, we describe the isolation and partial characterization of PGC from pig erythrocytes, small intestinal mucosa, kidney and liver. The mucosa, kidney and liver PGC fractions contained a complex pattern of Gal alpha antigens as shown by immunostaining using the Griffonia Simplicifolia isolectin B(4) while no reactivity was found with the erythrocyte PGC fractions. The mucosa PGC fractions stained strongly for blood group A antigens while the erythrocyte PGC fractions were negative. The presence of Gal alpha-terminating PGC compounds in porcine tissue adds further complexity to the distribution of this xenoantigen. Due to the long carbohydrate chains, PGC will be important targets for the Gal alpha xenoantibodies in pig to human xenotransplantation.

Characterization of blood group ABO(H)-active gangliosides in type AB erythrocytes and structural analysis of type A-active ganglioside variants in type A human erythrocytes

Several monosialogangliosides containing the type A-active epitope have been detected in type A erythrocytes on immunological analysis with a monoclonal antibody, and three of them were purified by repeated silica bead column chromatography and by scraping from the TLC plate. Two of these A-active gangliosides were characterized by methylation analysis by GC/MS, negative SIMS, MALDI-TOF/MS, proton nuclear magnetic resonance spectroscopy, and immunological assays, and their structures were concluded to be as follows. A-active ganglioside I:A-active ganglioside II:The reactivity of the purified gangliosides to the anti-A monoclonal antibodies (mAbs) exhibited enhancement after removal of the sialic acid. Therefore, the sialic residue has been shown to inhibit the binding to the terminal A-active epitope through the formation of an immune complex. To confirm the presence of A- (including S-A-I, -II and -III) and B-active gangliosides, the reactivity of anti-A and -B mAbs were investigated using total gangliosides from type A, -B and -AB erythrocytes on TLC plate. The results were that the gangliosides from types A and AB showed positive reaction to anti-A mAbs, whereas in the anti-B mAbs binding the gangliosides from types B and AB were positive. Thus, it revealed that A-active gangliosides were present in type A and -AB, and B-active gangliosides in types B and AB. As there was no difference in respective gangliosides on type AB erythrocytes of 22 individuals, both A- and B-active gangliosides are equally present in type AB erythrocytes. The biological significance of these A- and B-active ganglioside variants remains vague at present. As these molecules exhibit different reactivities to the anti-A mAbs, it is very likely that they can regulate the antigenicity of the A-epitope on the cell surface.

Pathogen-Host Protein-Carbohydrate Interactions as the Basis of Important Infections

Microbe-host protein-carbohydrate interactions are most likely the essential first step to produce an infection, although this has been strictly proven only in a few cases. Improved glycotechnology will help identification of new carbohydrate receptors and this knowledge may be used to identify microbial carbohydrate-binding proteins by affinity proteomics approaches. In some cases such conserved proteins may prove to be successful vaccine components, in other cases, like influenza, saccharide analogues may be the only rational alternative. The prognosis may be, based on these improvements, that infection medicine will make considerable progress in the near future.

Binding of rabbit hemorrhagic disease virus to antigens of the ABH histo-blood group family

The ability of rabbit hemorrhagic disease virus to agglutinate human erythrocytes and to attach to rabbit epithelial cells of the upper respiratory and digestive tracts was shown to depend on the presence of ABH blood group antigens. Indeed, agglutination was inhibited by saliva from secretor individuals but not from nonsecretors, the latter being devoid of H antigen. In addition, erythrocytes of the rare Bombay phenotype, which completely lack ABH antigens, were not agglutinated. Native viral particles from extracts of infected rabbit liver as well as virus-like particles from the recombinant virus capsid protein specifically bound to synthetic A and H type 2 blood group oligosaccharides. Both types of particles could attach to adult rabbit epithelial cells of the upper respiratory and digestive tracts. This binding paralleled that of anti-H type 2 blood group reagents and was inhibited by the H type 2-specific lectin UEA-I and polyacrylamide-conjugated H type 2 trisaccharide. Young rabbit tissues were almost devoid of A and H type 2 antigens, and only very weak binding of virus particles could be obtained on these tissues.

Polyglycosylceramides recognized by Helicobacter pylori: analysis by matrix-assisted laser desorption/ionization mass spectrometry after degradation with endo-beta-galactosidase and by fast atom bombardment mass spectrometry of permethylated undegraded material

Human erythrocyte polyglycosylceramides (PGCs) are recognized by the gastric pathogen Helicobacter pylori and are based on a successively extended and highly branched N-acetyllactosamine core linked to ceramide and substituted by fucose and sialic acid. As a step in the identification of the binding epitope we earlier characterized intact PGCs by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry, MALDI-TOF MS (Karlsson,H., Johansson,L., Miller-Podraza,H., and Karlsson,K-A. [1999] Glycobiology, 9, 765-778). In the present work, PGCs from human blood group O erythrocytes were digested with endo-ss-galactosidase (Bacterioides fragilis), an enzyme which cleaves the bond 3Galss1-4GlcNAc in linear but not branched poly-N-acetyllactosamine chains. The enzymatic digestion resulted in a mixture of neutral and sialic acid-containing glycolipids together with terminal and internal sequences of mainly neutral oligosaccharides. The products were analyzed by MALDI-TOF MS in both positive and negative ion mode which gave spectra where the ions could be assigned to structures of the neutral and acidic components, respectively. Among glycolipids found were [structure in text] where R could be H, Fuc or NeuAc. Also observed were structures as [structure in text] which indicated linear extension along both branches. Observed at higher masses were fully branched structures obtained by stepwise extension with [structure in text] where R could be H, Fuc or NeuAc. Most probably further branching may occur along both the (1-->3)- and the (1-->6)-linked branches to give a partly dendritic structure. Structures with more than one sialic acid substituted could not be observed in the MALDI spectrum. Complementary information of the terminal sequences was obtained by FAB-MS analysis of permethylated undegraded PGCs. High-temperature gas chromatography/mass spectrometry of reduced and permethylated products from enzyme hydrolysis documented that Fuc was present in a blood group O sequence, Fuc-Hex-HexN-. Fucose may be placed on short (monolactosamine) or longer branches, while sialic acid seems to be restricted to monolactosamine branches. The conclusion is that human erythrocyte PGCs display microheterogeneity within terminal and internal parts of the poly-N-acetyllactosamine chains. The first branch from the ceramide end may be located at the second or third Gal and possibly also on the first Gal. Other branches may occur on every N-acetyllactosamine unit in fully branched domains, or there may be linear extensions between branches resulting in incompletely branched structures. The extended linear sequences may be present in both 3- and 6-linked antennae. Terminal structures are based on one, two or maybe higher number of N-acetyllactosamine units.

A strain of human influenza A virus binds to extended but not short gangliosides as assayed by thin-layer chromatography overlay

A human strain of influenza virus (A, H1N1) was shown to bind in an unexpected way to leukocyte and other gangliosides when compared with avian virus (A, H4N6) as assayed on TLC plates. The human strain bound only to species with about 10 or more sugars, while the avian strain bound to a wide range of gangliosides including the 5-sugar gangliosides. By use of specific lectins, antibodies, and FAB and MALDI-TOF mass spectrometry an attempt was done to preliminary identify the sequences of leukocyte gangliosides recognized by the human strain. The virus binding pattern did not follow binding by VIM-2 monoclonal antibody and was not identical with binding by anti-sialyl Lewis x antibody. There was no binding by the virus of linear NeuAcalpha3- or NeuAcalpha6-containing gangliosides with up to seven monosaccharides per mol of ceramide. Active species were minor NeuAcalpha6-containing molecules with probably repeated HexHexNAc units and fucose branches. This investigation demonstrates marked distinctions in the recognition of gangliosides between avian and human influenza viruses. Our data emphasize the importance of structural factors associated with more distant parts of the binding epitope and the complexity of carbohydrate recognition by human influenza viruses.

Neu5Acalpha3Gal is part of the Helicobacter pylori binding epitope in polyglycosylceramides of human erythrocytes

The sialic acid dependent binding by the human pathogen Helicobacter pylori to polyglycosylceramides of human erythrocytes was investigated. Polyglycosylceramides, complex glycosphingolipids with a branched N-acetyllactosamine core, were isolated from human erythrocytes, blood group O, and subfractionated after peracetylation by anion-exchange chromatography. Three subfractions were deacetylated, analysed by matrix-assisted laser desorption ionization-time of flight MS and 2D 1H NMR spectroscopy. The observed mass ranges were m/z = 3093-7622, 3968-7255 and 3459-7987 in the mass spectra of the first, second and third fractions, respectively. The observed ions agreed with the general formula Hex(x+2)HexNAcxFucyNeu5AczCer. Two-dimensional 1H total correlation spectra of the mixtures showed that the first fraction contained 3-linked sialic acid and the second and third fractions contained both 3-linked and 6-linked sialic acid. Thin-layer chromatogram binding assays using the lectins from Maackia amurensis, specific for Neu5Acalpha3Galbeta4GlcNAc, and Sambucus nigra, specific for Neu5Acalpha6Gal/GalNAc, were used to confirm this distribution. H. pylori recognized all three fractions in the binding assay, indicating that the 3-linked, rather than 6-linked, sialic acid is essential for binding.

Helicobacter pylori and neutrophils: sialic acid-dependent binding to various isolated glycoconjugates

Helicobacter pylori has been shown to agglutinate erythrocytes in a sialic acid-dependent manner. However, very few studies have examined relevant target cells in the human stomach. Neutrophils are required for the onset of gastritis, and the inflammatory reaction may be induced on contact between bacteria and neutrophils. In the present work, glycolipids and glycoproteins were isolated from neutrophils and were studied for binding by overlay with radiolabeled bacteria on thin-layer chromatograms and on membrane blots. There was a complex pattern of binding bands. The only practical binding activity found was sialic acid dependent, since treatment of glycoconjugates with neuraminidase or mild periodate eliminated binding. As shown before for binding to erythrocytes and other glycoconjugates, bacterial cells grown on agar bound to many glycoconjugates, while growth in broth resulted in bacteria that would bind only to polyglycosylceramides, which are highly heterogeneous and branched poly-N-acetyllactosamine-containing glycolipids. Approximately seven positive bands were found for glycoproteins, and the traditional ganglioside fraction showed a complex, slow-moving interval with very strong sialic-acid-dependent binding, probably explained by Fuc substitutions on GlcNAc.

Fingerprinting of large oligosaccharides linked to ceramide by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry: highly heterogeneous polyglycosylceramides of human erythrocytes with receptor activity for Helicobacter pylori

Highly microheterogeneous polyglycosylceramides (PGCs) of human erythrocytes with an average composition of about 25 monosaccharides linked to ceramide were analyzed by matrix-assisted laser desorption-ionization time-of-flight mass spectrometry (MALDI-TOF MS). The human gastric pathogen Helicobacter pylori was earlier shown to bind this glycosphingolipid mixture by thin-layer chromatogram binding assay. The receptor activity was present along the whole nonresolved chromatographic interval. Mass spectra of intact PGCs were compared with corresponding spectra of oligosaccharides enzymatically released from the ceramides. Two subfractions of PGCs containing less than one and more than one sialic acid residue per molecule were used. MALDI-MS spectra were recorded in both linear and reflectron mode with the accuracies of </=0.08% and </=0.02%, respectively, which allowed determination of the constituent parts of the detected ions in form of ceramide and number of hexoses, N-acetylhexosamines, fucoses and sialic acids. Molecular species were found based on ceramide with mainly sphingosine and fatty acids 24:0 and 24:1 (with less amounts of 22:0), and with a total number of monosaccharides ranging from 11 (neutral, m/z = 2605 for [M+Na](+)) to 41 (one sialic acid, m/z = 8057 for [M-H](-)). The saccharide composition obtained supported a successively extended and branched N -acetyllactosamine core with substitutions of fucoses (0 up to 8) and sialic acid (0 to 1). The reliable molecular analysis of large oligosaccharides linked to ceramide using this approach will be of great help for the further structure analysis in order to define the epitope for the sialic acid-dependent binding by the bacterium.

Interaction of cholera toxin and Escherichia coli heat-labile enterotoxin with glycoconjugates from rabbit intestinal brush border membranes: relationship with ABH blood group determinants

The capacity of cholera toxin (CT) and type I heat-labile enterotoxin produced by Escherichia coli isolated from human intestine (LTh) to interact with glycoconjugates bearing ABH blood group determinants from rabbit intestinal brush border membranes (BBM) was studied. On the basis of the type of intestinal compounds related to the human ABH blood group antigens, rabbits were classified as AB or H. Toxin binding to the intestinal glycolipids and glycoproteins depends on the blood group determinant borne by the glycoconjugate and on the analyzed toxin. LTh was capable of interacting preferentially with several blood group A- and B-active BBM glycolipids compared to those isolated from animals lacking these antigens (H rabbits). Also, LTh preferably bound to several BBM glycoproteins from AB rabbit intestines compared to those from H ones. One of these glycoproteins, the sucrase-isomaltase complex (EC 3.2.1.48-10) isolated from AB and H rabbits showed the same differential LTh binding. Conversely, CT practically did not recognize either blood group A-, B-, or H-active glycolipids and glycoproteins. These results may be relevant for carrying out in vivo experiments in rabbits in order to disclose the role of ABH active-glycoconjugates in the secretory response induced by LTh in rabbit intestine.

Direct matrix-assisted laser desorption/ionization mass spectrometric analysis of glycosphingolipids on thin layer chromatographic plates and transfer membranes

Results are reported for analysis by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOFMS) of native glycosphingolipids (GSLs) after development on thin layer chromatographic plates and after heat transfer of the GSLs from the plates to several types of polymer membranes. The spectral quality is better for membrane-bound analytes, in terms of sensitivity, mass resolution and background interference. The sensitivity gain compared with liquid secondary ion mass spectrometry (LSIMS) of GSLs on thin layer plates is 1-2 orders of magnitude (detection limits of 5-50 pmol vs. 1-10 nmol). Resolution and mass accuracy (0.1%) are limited by the irregular membrane surfaces and this effect cannot be entirely compensated by delayed extraction. The best results were obtained with a polyvinylidene difluoride (PVDF) P membrane, with irradiation from a nitrogen laser. Although the Nafion membrane could not be used for molecular weight profiling, its acidic character led to sample hydrolysis at the glycosidic linkages, thus yielding a series of fragments that could be used to determine the sequence of carbohydrate residues. Structural information could also be obtained by post-source decay (PSD) experiments on mass-selected precursor ions. Samples containing both neutral and acidic components were characterized in a 1:1 combination of 2, 5-dihydroxybenzoic acid and 2-amino-5-nitropyridine. GSLs that exhibited binding to antibodies in an overlay assay on the TLC plate were transferred to membranes and analyzed by MALDI-TOFMS without interference from the antibody or the salts and buffers used during the binding and visualization steps. Taking advantage of the insights into sample preparation gained from these studies, future research will extend this approach to analysis by matrix-assisted laser desorption/ionization Fourier transform ion cyclotron resonance mass spectrometry (MALDI-FTICRMS) with an external ion source.

Selective binding by Helicobacter pylori of leucocyte gangliosides with 3-linked sialic acid, as identified by a new approach of linkage analysis

The human gastric pathogen Helicobacter pylori has been shown to bind to glycoconjugates of human leucocytes in a sialic acid-dependent way. In order to improve the identification of the binding epitope, a new technique was developed to analyze the ketosidic linkage position between a terminal sialic acid and the consecutive monosaccharide. Permethylation and reduction with LiAlH4 followed by trifluoroacetolysis in 1000:1 trifluoroacetic anhydride:trifluoroacetic acid (24 h, 100 degrees C) results in the cleavage of glycosidic but not ketosidic bonds. The disaccharide products were analyzed by gas chromatography-mass spectrometry and sialyl-3 or -6 position and NeuAc or NeuGc are identified by their separate retention times and mass spectra. The method was worked out on model saccharides and applied on five-sugar gangliosides (sialylparaglobosides) of human leucocytes. Radiolabeled Helicobacter pylori was shown to bind to the upper part, but not to the lower part, of the five-sugar interval of a mixture of gangliosides separated on a thin-layer chromatogram. Using a membrane blotting procedure the active and inactive bands were isolated and shown to be NeuAcalpha2-3- and NeuAcalpha2-6-paraglobosides, respectively.

Meaning and therapeutic potential of microbial recognition of host glycoconjugates

The microbe-host interface is currently in focus because of attempts to develop infection therapy in humans based on either natural receptor saccharide (respiratory and gastrointestinal disease) or sophisticated sialic acid analogues designed from crystal structures (influenza). Most of the known host receptors for microbes are glycoconjugates, and the diversity and selectivity of host tissue glycosylation allow for the tropism of infections. However, among the many binding specificities detected so far, the biological role has been proven only in a few infectious model systems. The existence of multiple specificities of a single microbe is both a complicating factor and a challenge, requiring expanded research with a special demand on glycoscience.