Binding specificities of the lectins fromHelix pomatia, soybean and peanut against different glycosphingolipids in liposome membranes

Generation of glycans depleted decellularized porcine pericardium, using digestive enzymatic supplements and enzymatic mixtures for food industry

BACKGROUND

Xenogeneic pericardium has been used largely for various applications in cardiovascular surgery. Nevertheless, xenogeneic pericardial patches fail mainly due to their antigenic components. The xenoantigens identified as playing a major role in recipient immune response are the Galα1-3Gal (α-Gal) epitope, the non-human sialic acid N-glycolylneuraminic acid (Neu5Gc), and the porcine SDa antigen, associated with both proteins and lipids. The reduction in glycans from porcine pericardium might hinder or reduce the immunogenicity of xenogeneic scaffolds.

METHODS

Decellularized porcine pericardia were further treated at different time points and dilutions with digestive enzymatic supplements and enzymatic mixtures applied for food industry, for the removal of potentially immunogenic carbohydrates. Carbohydrates removal was investigated using up to 8 different lectin stains for the identification of N- and O-glycosylations, as well as glycolipids. Histoarchitectural changes in the ECM were assessed using Elastica van Gieson stain, whereas changes in mechanical properties were investigated via uniaxial tensile test and burst pressure test.

RESULTS

Tissues after enzymatic treatments showed a dramatic decrease in lectin stainings in comparison to tissues which were only decellularized. Histological assessment revealed cell-nuclei removal after decellularization. Some of the enzymatic treatments induced elastic lamellae disruption. Tissue strength decreased after enzymatic treatment; however, treated tissues showed values of burst pressure higher than physiological transvalvular pressures.

CONCLUSIONS

The application of these enzymatic treatments for tissue deglycosylation is totally novel, low cost, and appears to be very efficient for glycan removal. The immunogenic potential of treated tissues will be further investigated in subsequent studies, in vitro and in vivo.

MCAW-DB: A glycan profile database capturing the ambiguity of glycan recognition patterns

Glycan-binding protein (GBP) interaction experiments, such as glycan microarrays, are often used to understand glycan recognition patterns. However, oftentimes the interpretation of glycan array experimental data makes it difficult to identify discrete GBP binding patterns due to their ambiguity. It is known that lectins, for example, are non-specific in their binding affinities; the same lectin can bind to different monosaccharides or even different glycan structures. In bioinformatics, several tools to mine the data generated from these sorts of experiments have been developed. These tools take a library of predefined motifs, which are commonly-found glycan patterns such as sialyl-Lewis X, and attempt to identify the motif(s) that are specific to the GBP being analyzed. In our previous work, as opposed to using predefined motifs, we developed the Multiple Carbohydrate Alignment with Weights (MCAW) tool to visualize the state of the glycans being recognized by the GBP under analysis. We previously reported on the effectiveness of our tool and algorithm by analyzing several glycan array datasets from the Consortium of Functional Glycomics (CFG). In this work, we report on our analysis of 1081 data sets which we collected from the CFG, the results of which we have made publicly and freely available as a database called MCAW-DB. We introduce this database, its usage and describe several analysis results. We show how MCAW-DB can be used to analyze glycan-binding patterns of GBPs amidst their ambiguity. For example, the visualization of glycan-binding patterns in MCAW-DB show how they correlate with the concentrations of the samples used in the array experiments. Using MCAW-DB, the patterns of glycans found to bind to various GBP-glycan binding proteins are visualized, indicating the binding "environment" of the glycans. Thus, the ambiguity of glycan recognition is numerically represented, along with the patterns of monosaccharides surrounding the binding region. The profiles in MCAW-DB could potentially be used as predictors of affinity of unknown or novel glycans to particular GBPs by comparing how well they match the existing profiles for those GBPs. Moreover, as the glycan profiles of diseased tissues become available, glycan alignments could also be used to identify glycan biomarkers unique to that tissue. Databases of these alignments may be of great use for drug discovery.

Design of a covalently bonded glycosphingolipid microarray

Glycosphingolipids (GSLs) are well known ubiquitous constituents of all eukaryotic cell membranes, yet their normal biological functions are not fully understood. As with other glycoconjugates and saccharides, solid phase display on microarrays potentially provides an effective platform for in vitro study of their functional interactions. However, with few exceptions, the most widely used microarray platforms display only the glycan moiety of GSLs, which not only ignores potential modulating effects of the lipid aglycone, but inherently limits the scope of application, excluding, for example, the major classes of plant and fungal GSLs. In this work, a prototype "universal" GSL-based covalent microarray has been designed, and preliminary evaluation of its potential utility in assaying protein-GSL binding interactions investigated. An essential step in development involved the enzymatic release of the fatty acyl moiety of the ceramide aglycone of selected mammalian GSLs with sphingolipid N-deacylase (SCDase). Derivatization of the free amino group of a typical lyso-GSL, lyso-G(M1), with a prototype linker assembled from succinimidyl-[(N-maleimidopropionamido)-diethyleneglycol] ester and 2-mercaptoethylamine, was also tested. Underivatized or linker-derivatized lyso-GSL were then immobilized on N-hydroxysuccinimide- or epoxide-activated glass microarray slides and probed with carbohydrate binding proteins of known or partially known specificities (i.e., cholera toxin B-chain; peanut agglutinin, a monoclonal antibody to sulfatide, Sulph 1; and a polyclonal antiserum reactive to asialo-G(M2)). Preliminary evaluation of the method indicated successful immobilization of the GSLs, and selective binding of test probes. The potential utility of this methodology for designing covalent microarrays that incorporate GSLs for serodiagnosis is discussed.

Receptor-dependent and -independent axonal retrograde transport of poliovirus in motor neurons

Poliovirus (PV), when injected intramuscularly into the calf, is incorporated into the sciatic nerve and causes an initial paralysis of the inoculated limb in transgenic (Tg) mice carrying the human PV receptor (hPVR/CD155) gene. We have previously demonstrated that a fast retrograde axonal transport process is required for PV dissemination through the sciatic nerves of hPVR-Tg mice and that intramuscularly inoculated PV causes paralytic disease in an hPVR-dependent manner. Here we showed that hPVR-independent axonal transport of PV was observed in hPVR-Tg and non-Tg mice, indicating that several different pathways for PV axonal transport exist in these mice. Using primary motor neurons (MNs) isolated from these mice or rats, we demonstrated that the axonal transport of PV requires several kinetically different motor machineries and that fast transport relies on a system involving cytoplasmic dynein. Unexpectedly, the hPVR-independent axonal transport of PV was not observed in cultured MNs. Thus, PV transport machineries in cultured MNs and in vivo differ in their hPVR requirements. These results suggest that the axonal trafficking of PV is carried out by several distinct pathways and that MNs in culture and in the sciatic nerve in situ are intrinsically different in the uptake and axonal transport of PV.

Further characterization of the saccharide specificity of peanut (Arachis hypogaea) agglutinin

2-Dansylamino-2-deoxy-D-galactose (GalNDns) has been shown to bind to peanut (Arachis hypogaea) agglutinin (PNA) in a saccharide-specific manner. This binding was accompanied by a five-fold increase in the fluorescence of GalNDns. The interaction was characterized by an association constant of 0.15 mM at 15 degrees and delta H and delta S values of -57.04 kJ.mol-1 and -118.1J.mol-1.K-1, respectively. Binding of a variety of other mono-, di- and oligo-saccharides to PNA, studied by monitoring their ability to dissociate the PNA GalNDns complex, revealed that PNA interacts with several T-antigen-related structures, such as beta-D-Galp-(1----3)-D-GalNAc, beta-D-Galp-(1----3)-alpha-D-GalpNAcOMe, and beta-D-Galp-(1----3)-alpha-D-GalpNAc-(1----3)-Ser, as well as the asialo-GM1 tetrasaccharide, with comparable affinity, thus showing that this lectin does not discriminate between saccharides in which the penultimate sugar of the beta-D-Galp-(1----3)-D-GalNAc unit is the alpha or beta anomer, in contrast to jacalin (Artocarpus integrifolia agglutinin), another anti T-lectin which preferentially binds to beta-D-Galp-(1----3)-alpha-D-GalNAc and does not recognize beta-D-Galp-(1----3)-beta-D-GalNAc or the related asialo-GM1 oligosaccharide. These studies also indicated that, in the extended combining region of PNA which accommodates a disaccharide, the primary subsite (subsite A) is highly specific for D-galactose, whereas the secondary subsite (subsite B) is less specific and can accommodate various structures, such as D-galactose, 2-acetamido-2-deoxy-D-galactose, D-glucose, and 2-acetamido-2-deoxy-D-glucose.

Comparison of the carbohydrate-binding specificities of seven N-acetyl-D-galactosamine-recognizing lectins

Seven plant lectins, Dolichos biflorus agglutinin (DBA), Griffonia simplicifolia agglutinin (GSA, isolectin A4), Helix pomatia agglutinin (HPA), soybean (Glycine max) agglutinin (SBA), Salvia sclarea agglutinin (SSA), Vicia villosa agglutinin (VVA, isolectin B4) and Wistaria floribunda agglutinin (WFA), known to be specific for N-acetyl-D-galactosamine-(GalNAc) bearing glycoconjugates, have been compared by the binding of their radiolabelled derivatives, to eight well-characterized synthetic oligosaccharides immobilized via a spacer on an inert silica matrix (Synsorb). The eight oligosaccharides included the Forssman, the blood group A and the T antigens, as well as alpha GalNAc coupled directly to the support (Tn antigen) and also structures with GalNAc linked alpha or beta to positions 3 or 4 of an unsubstituted Gal. The binding studies clearly distinguished the lectins into alpha GalNAc-specific agglutinins like DBA, GSA and SSA, and lectins which recognize alpha- as well as beta-linked GalNAc residues like HPA, VVA, WFA and SBA. HPA was the only lectin which bound to the beta Gal1----3 alpha GalNAc-Synsorb adsorbent (T antigen) indicating that it also recognizes internal GalNAc residues. Among the alpha GalNAc-specific lectins, DBA strongly recognized blood group A structures while GSA displayed weaker recognition, and SSA bound only slightly to this affinity matrix. In addition, DBA and SSA were able to distinguish between GalNAc linked alpha 1----3 and GalNAc linked alpha 1----4, to the support, the latter being a much weaker ligand. These results were corroborated by the binding of the lectins to biological substrates as determined by their hemagglutination titers with native and enzyme-treated red blood cells carrying known GalNAc determinants, e.g. blood group A, and the Cad and Tn antigens. For SSA, the binding to the alpha GalNAc matrix was inhibited by a number of glycopeptides and glycoproteins confirming the strong preference of this lectin for alpha GalNAc-Ser/Thr-bearing glycoproteins.

Host cell-induced differences in O-glycosylation of herpes simplex virus gC-1. I. Structures of nonsialylated HPA- and PNA-binding carbohydrates

Lectins with narrow oligosaccharide specificities were established as probes to study the host cell influence on the biosynthesis of O-linked oligosaccharides of the herpes simplex virus type 1 (HSV-1)-specified glycoprotein C (gC-1). We found that only gC-1 and no other glycoprotein bound to the peanut lectin (PNA), with main specificity for Gal(beta 1-3)GalNAc. Previously, we have shown that only gC-1 binds to the Helix pomatia lectin (HPA), with main specificity for terminal GalNAc. The O-linked oligosaccharides binding to PNA and HPA were released by alkaline borohydride treatment and characterized. A structural determination of these oligosaccharides showed that the HPA-binding carbohydrates were monosaccharides (GalNAc), and that the PNA-binding oligosaccharides were disaccharides with the structure Gal-GalNAc. The PNA- and HPA-binding oligosaccharides were arranged as Pronase-resistant clusters on gC-1, consisting of about seven individual, adjacent oligosaccharides. In addition to these disaccharides, Pronase-resistant PNA-binding glycopeptides of gC-1 also contained neutral trisaccharides. Larger O-linked oligosaccharides, binding to the wheat germ lectin, were found in gC-1, but not in proximity to the PNA-binding ones. It was concluded that the lectins mentioned should be useful probes in screening HSV-infected cells of different lineages for differences in processing of O-linked oligosaccharides.

Host cell-induced differences in the O-glycosylation of herpes simplex virus gC-1. II. Demonstration of cell-specific galactosyltransferase essential for formation of O-linked oligosaccharides

By using a lectin-based screening method for cell-dependent variations of O-glycosylation of viral glycoprotein, we found that O-linked oligosaccharides of herpes simplex virus type 1 (HSV-1) glycoproteins in virus-infected mouse neuroblastoma (C1300) cells differed from those of HSV glycoproteins produced in other cells. Thus, O-linked oligosaccharides of HSV-1-specified glycoprotein C (gC-1), produced in GMK cells and a number of other cells, occurred mainly as trisaccharides or larger structures. In contrast, gC-1, produced in C1300 cells, contained O-linked monosaccharides and very few, if any, larger oligosaccharides of this class. A structural comparison between O-linked oligosaccharides of gC-1 from HSV-1-infected C1300 cells and from GMK cells showed that biosynthesis was interrupted prior to formation of a core disaccharide with terminal galactose, indicating a major early defect in O-glycosylation of glycoproteins in C1300 cells. A comparison of the content of galactosyltransferases between C1300 and GMK cells showed that C1300 cells lacked galactosyltransferases, including the specific enzyme engaged in formation of the core O-linked disaccharide mentioned, while other glycosyltransferases adding terminal sugars to O-linked oligosaccharides were present in equal amounts in both cell lines. These results indicated that HSV-1 is strictly dependent on host cell-specified factors for biosynthesis of O-linked oligosaccharides associated with viral glycoproteins.

Binding specificities of the lectins PNA, WGA and UEA I to polyvinylchloride-adsorbed glycosphingolipids

The binding specificities of the lectins PNA (peanut agglutinin), WGA (wheat germ agglutinin), and UEA I (Ulex europeus agglutinin I) against glycosphingolipids were investigated using an enzyme-linked immunosorbent assay (ELISA), utilizing the biotin-avidin system for detection of bound lectin. PNA showed the highest affinity to GA1, but also bound, though less strongly, to GM1 and GD1b. WGA bound to 3'-nLM1 and 6'-nLM1, the former twice as strongly as the latter, but not to any sialic acid containing glycolipid of the gangliotetraose series. UEA I showed a high affinity for the Lea glycolipid which has an alpha 1-4 linked fucose but not for the glycolipids with alpha 1-3 or alpha 1-2 linked fucose. Interestingly, 3'-nLM1 and nLA1, glycolipids lacking fucose, also bound UEA I. The results show that lectins should be used with caution for establishing terminal sugar sequences in glycosphingolipids.

Specific binding of concanavalin A to free inositol and liposomes containing phosphatidylinositol

We previously reported that concanavalin A could bind specifically to liposomes containing phospholipids and lacking glycoconjugates (Biochem. Biophys. Res. Comm. 74, 208, 1977). In the present study we show that the binding of concanavalin A to the liposomes was greatly increased (up to 5 fold) by the presence of phosphatidylinositol in the liposomes. Furthermore, the binding of concanavalin A to phosphatidylinositol-liposomes was specific and could be inhibited by either alpha-methyl mannoside or by myo-inositol. We also found that concanavalin A-induced lymphocyte mitogenesis could be inhibited either by alpha-methyl mannoside or by myo-inositol. Simultaneous addition of both inhibitors to concanavalin A and liposomes showed that inhibition was non-competitive: alpha-methyl mannoside was more inhibitory to liposomes lacking phosphatidylinositol, and myo-inositol was more inhibitory to liposomes containing phosphatidylinositol. This suggests that the binding site for inositol might be different than that for mannose. Equilibrium dialysis and Scatchard plots revealed 4 binding sites each for inositol and mannose at neutral pH. The binding constants of concanavalin A were 0.13 X 10(4) and 0.25 X 10(4) liters/mole respectively for inositol and mannose. We conclude that concanavalin A binds specifically to the inositol portion of phosphatidylinositol.

Different populations of herpes simplex virus glycoprotein C discriminated by the carbohydrate-binding characteristics of N-acetylgalactosamine specific lectins (soybean and Helix pomatia). Brief report

From the herpes simplex virus specified glycoprotein C two fractions were isolated with affinity either for Helix pomatia lectin (HPA) or soybean lectin (SBA). The data indicated that HPA and SBA, despite their mutual main specificity for N-acetylgalactosamine, recognize structurally different gC populations.