Glycan uniformity within molecular variants of transferrin with distinct affinity for concanavalin A

Generation of clade- and symbiont-specific antibodies to characterize marker molecules during Cnidaria-Symbiodinium endosymbiosis

The endosymbiosis between cnidarians and dinoflagellates is responsible for the formation of coral reefs. Changes in molecules have been identified during the process of cnidaria-Symbiodinium endosymbiosis. However, the complexity of the molecular interaction has prevented the establishment of a mechanistic explanation of cellular regulation in this mutualistic symbiosis. To date, no marker molecules have been identified to specifically represent the symbiotic status. Because the endosymbiotic association occurs in the symbiotic gastrodermal cells (SGCs), whole cells of isolated SGCs were used as an antigen to generate monoclonal antibodies (mAb) to screen possible molecular candidates of symbiotic markers. The results showed that one of the generated monoclonal antibodies, 2–6F, specifically recognized clade C symbiotic Symbiodinium but not its free-living counterpart or other Symbiodinium clades. The expression levels of 2–6F mAb-recognized proteins are highly correlated with the symbiotic status, and these proteins were characterized as N-linked glycoproteins via treatment with peptide N-glycosidase F. Furthermore, their glycan moieties were markedly different from those of free-living Symbiodinium, potentially suggesting host regulation of post-translational modification. Consequently, the 2–6F mAb can be used to detect the symbiotic state of corals and investigate the complex molecular interactions in cnidaria-Symbiodinium endosymbiosis.

Construction of a Database of Collision Cross Section Values for Glycopeptides, Glycans, and Peptides Determined by IM-MS

An ion mobility quadrupole time-of-flight mass spectrometer was used to examine the gas-phase structures of a set of glycopeptides resulting from proteolytic digestion of the well-characterized glycoproteins bovine ribonuclease B, human transferrin, bovine fetuin and human α1-acid glycoprotein, the corresponding deglycosylated peptides, and the glycans released by the endoglycosidase PNGase F. When closely related glycoforms did not occur naturally, exoglycosidases were used to achieve stepwise removal of individual saccharide units from the nonreducing termini of the multiantennary structures. Collision cross sections (CCS) were calculated and plotted as a function of mass-to-charge ratio. Linear trendlines were observed for the glycoforms of individual N-linked glycopeptides, the deglycosylated peptides, and the released, deutero-reduced permethylated glycans. For the glycoforms of a given glycopeptide or set of derivatized glycans, the slope of the line connecting CCS values remained similar for the [M+3H]3+ ions observed as the glycan antennae were shortened by stepwise exoglycosidase treatments; this trend was consistent regardless of the peptide length or the saccharide removed. The results form the basis for a database of CCS values and the CCS increments that correspond to changes in glycoform compositions.

Square-wave voltammetry assays for glycoproteins on nanoporous gold

Electrochemical enzyme-linked lectinsorbent assays (ELLA) were developed using nanoporous gold (NPG) as a solid support for protein immobilization and as an electrode for the electrochemical determination of the product of the reaction between alkaline phosphatase (ALP) and p-aminophenyl phosphate (p-APP), which is p-aminophenol (p-AP). Glycoproteins or concanavalin A (Con A) and ALP conjugates were covalently immobilized onto lipoic acid self-assembled monolayers on NPG. The binding of Con A - ALP (or soybean agglutinin - ALP) conjugate to glycoproteins covalently immobilized on NPG and subsequent incubation with p-APP substrate was found to result in square-wave voltammograms whose peak difference current varied with the identity of the glycoprotein. NPG presenting covalently bound glycoproteins was used as the basis for a competitive electrochemical assay for glycoproteins in solution (transferrin and IgG). A kinetic ELLA based on steric hindrance of the enzyme-substrate reaction and hence reduced enzymatic reaction rate after glycoprotein binding is demonstrated using immobilized Con A-ALP conjugates. Using the immobilized Con A-ALP conjugate, the binding affinity of immunoglobulin G (IgG) was found to be 105 nM, and that for transferrin was found to be 650 nM. Minimal interference was observed in the presence of 5 mg mL-1 BSA as a model serum protein in both the kinetic and competitive ELLA. Inhibition studies were performed with methyl D-mannoside for the binding of TSF and IgG to Con A-ALP; IC50 values were found to be 90 μM and 286 μM, respectively. Surface coverages of proteins were estimated using solution depletion and the BCA protein concentration assay.

Relationships among tumor burden, tumor size, and the changing concentrations of fibrin degradation products and fibrinolytic factors in the pleural effusions of rabbits with VX2 lung tumors

The VX2 tumor is derived from a papilloma virus-induced rabbit epithelial cell line. If VX2 tumor cells (trapped in a plasma clot) are introduced intravenously into NZW rabbits, the cells lodge in the lung capillary bed and produce tumors. Independently of the tumor burden (ie, the total tumor weight per rabbit), approximately 15% of rabbits with VX2 lung tumors accumulate an effusion in the interpleural space and this pleural effusion contains products of hemostasis. We hypothesized that these products were of intra-tumoral origin and that they changed in concentration as tumor burden increased. Interrelationships among lung-, tumor-weights, and pleural effusion volumes, and the concentrations of fibrinolytic factors, their catabolic products, and other proteins of pleural effusions were measured in rabbits with a wide range of tumor burdens. Positive correlations between tumor burden and total lung weight and between pleural effusion volume and net lung weight suggested that interstitial fluid from the stroma of tumors passed directly into the extravascular space of the lung(s) and into the interpleural space(s). Analyses of pleural effusions indicated that plasminogen-, alpha(2)-antiplasmin-, and plasminogen activator inhibitor-1-related proteins, urokinase-like- and tissue-plasminogen activator activities, and vascular endothelial growth factor increased in concentration up to a tumor burden of approximately 20-25 g. Plasmin activity and intact fibrinogen were absent. The concentration of fibrin(ogen) degradation products did not change significantly up to a tumor burden of approximately 25 g but increased substantially as tumor burdens exceeded 25 g. In conclusion, interstitial fluid from tumors enters the extravascular space of the host and may accumulate with fluid from non-tumor sources as a pleural effusion. The concentrations of fibrinolytic factors and their products in pleural effusions reflect the tumor burden of the rabbit. Conceivably, the components of a malignant effusion contain much information about the extent of tumor growth.

Fibrinogen catabolism within the procoagulant VX-2 tumor of rabbit lung in vivo: Effluxing fibrin(ogen) fragments contain antiangiogenic activity

Many types of solid tumors are known to be procoagulant environments. This is partly because a hyperpermeable vascular system within the tumor allows plasma hemostatic factors to accumulate in relatively high concentrations in the stroma, and many solid-tumor cells express tissue factor or a procoagulant factor. These circumstances appear to exist in the VX-2 lung tumor of the New Zealand White (NZW) rabbit, and they sustain a measurable turnover of stromal deposits of fibrin(ogen). We have measured the turnover of fibrinogen within tumors of the VX-2 tumor-burdened rabbit and analysed the catabolic products of fibrin(ogen) and the status of fibrinolysis in tumor-derived interpleural effusate. Using intravenously injected (125)I-labeled rabbit fibrinogen as a marker, we found that fibrinogen (approximate blood concentration 1740 microg/mL) passed from blood to VX-2 tumor stroma, saturating the tumor at a concentration of approximately 348 microg fibrinogen/g in approximately 12 hours. We measured fibrin(ogen) fragments, at a concentration of approximately 292 microg/mL, in interpleural effusates that we recovered from 13% of the VX-2-burdened rabbits. Unreduced fibrin(ogen) fragments consisted of 4 major components with a relative molecular mass of approximately 250,000 (assumed to be fragment X; approximately 9% of total fragments from densitometry of immunoblots), 200,000 (d-dimer; 41%), 110,000 (fragment D; 49%), and 50,000 to 55,000 (fragment E; 1%-2%) kD. Total fibrin(ogen) fragments immunopurified from effusates exhibited an antiangiogenic effect when subjected to a chick embryo chorioallantoic membrane procedure. Interpleural effusates were devoid of plasmin activity or active plasminogen activator inhibitor-1 but contained plasmin complexes and active urokinase-like plasminogen activator (uPA), alpha(2)-antiplasmin, and thrombin-activatable fibrinolysis inhibitor. We speculate that VX-2 cells release uPA to activate fibrinolysis within the tumor stroma. Catabolic products of hemostasis (eg, fibrinolytic fragments, angiostatin) flux from the stroma into the interpleural space, thereby providing a net antiangiogenic property to the effusate and ultimately to the lymphatic and circulatory systems.

Angiostatin II is the predominant glycoform in pleural effusates of rabbit VX-2 lung tumors

Angiostatin (AST), a polypeptide with potent antiangiogenic properties, is released proteolytically from plasminogen in vivo. Plasminogen exists naturally in plasma as two glycoforms (PLGs), I and II. Recently it was shown with the use of a chick-embryo chorioallantoic membrane (CAM) assay that rabbit PLG-I and -II yield distinct ASTs-AST-I and -II, respectively-with different antiangiogenic activities. AST glycoforms were of similar molecular weight, approximately 30 to 32,000 kD, and probably consisted of kringles 1 to 3 only. AST has now been identified in the interpleural effusate released from VX-2 lung tumors in rabbits. Effusate was collected from six rabbits with high tumor burdens and fractionated by means of lysine-Sepharose chromatography. The epsilon-aminohexanoic acid-eluted protein of all effusates contained AST (kringles 1-3) at a mean concentration of 1.2 microg/mL of effusate; with regard to AST content, 97% was AST-II. A CAM assay revealed that the lysine-Sepharose-bound fraction from all interpleural effusates contained potent antiangiogenic activity. Blood and urine from rabbits with high burdens of VX-2 contained essentially only AST-II, at mean concentrations of 145 and 4 ng/mL, respectively. AST was absent from the blood of control rabbits. In an attempt to compare their uptake by VX-2, iodine 125-labeled AST-I and iodine 131-labeled AST-II were injected intravenously into tumor-bearing rabbits. AST-I entered the tumor 1.6 times faster than AST-II. As a means of accounting for the preponderance of AST-II in the interpleural effusate, we postulate that VX-2 cells release proteolytic activity to activate plasminogen but that of the two PLGs, PLG-II may be the preferred substrate for AST formation in vivo.

Metabolism of rabbit angiostatin glycoforms I and II in rabbits: angiostatin-I leaves the intravascular space faster and appears to have greater anti-angiogenic activity than angiostatin-II

Plasminogen (PLG) exists in the circulation as two glycoforms, I and II. Angiostatin (AST) is a polypeptide that has been cleaved from the kringle region of PLG and has strong anti-angiogenic properties. AST-I and AST-II, which consisted only of kringles 1 through 3, were prepared by the action of urokinase on purified rabbit PLG-I and PLG-II, respectively, in the presence of N-acetyl cysteine, followed by affinity chromatography on lysine-Sepharose. Purified AST-I and AST-II were tested for functional activity with a chick chorioallantoic membrane (CAM) model; when similar amounts were applied to a 6-day CAM, AST-I was substantially more effective than AST-II in decreasing vascular supply to the CAM over a 72-hour period; this activity correlated with a loss of capillaries, probably through apoptosis of endothelial cells. Radiolabeled AST-I and AST-II (iodine 125 and iodine 131) were co-injected intravenously into healthy rabbits to determine their clearances from plasma measured over 3 days. Over a dose range of 0.08 to 2.7 microg/kg, the fractional catabolic rate within the intravascular space (j(3)) indicated that AST-I was cleared 3-fold to 4-fold more rapidly than AST-II (P < .001). The catabolic half-life of AST-I (2.01 +/- 0.19 days) was significantly less than that of AST-II (2.62 +/- 0.20 days). The faster clearance of AST-I from the intravascular space was matched by its more rapid passage than AST-II to the extravascular space of various organs over 60 minutes in vivo. This property of AST-I as compared with AST-II may partially explain its greater anti-angiogenic potential. From the plasma concentrations of PLG-I and PLG-II and their relative behaviors toward rabbit VX-2 lung tumors in vivo, we predict that substantially greater quantities of AST-II than AST-I may be released into the extravascular space of tumors.

Comparative metabolism of plasminogen glycoforms I and II in the alloxan-diabetic rabbit

The metabolism of plasminogen glycoforms I and II was measured in alloxan-induced diabetic and in age-matched control rabbits. Radiolabeled plasminogen I and II were degraded significantly more slowly in diabetic compared with control rabbits; plasminogen II [half-time (T1/2), 1.31 days] was degraded faster than plasminogen I (T1/2), 1.86 days) in diabetic rabbits and in control rabbits (T1/2, 1.18 and 1.58 days, respectively). From the catabolic rates and relative quantities in plasma, we calculated that approximately four molecules of plasminogen II were degraded for one molecule of plasminogen I in the diabetic and control rabbits. To verify this later observation, plasminogen I and II production by diabetic rabbit livers was compared with that by the control livers in vitro. During perfusion with [3H]leucine, 3H-labeled protein was released more slowly from diabetic than from control livers, but no quantitative difference in total plasminogen yield between diabetic and control livers was found. Nevertheless, plasminogen II was produced 0.7 +/- 0.4 and 4.3 +/- 0.3 times faster than plasminogen I by diabetic and control livers, respectively. Plasminogen metabolism in the diabetic rabbit did not differ qualitatively from that in the control rabbit except that catabolism was slowed.

Purification of isotransferrins by concanavalin A sepharose chromatography and preparative isoelectric focusing

1. From pooled serum containing genetically homogeneous transferrin C1, transferrin was purified and separated in three fractions (tri-tri, bi-tri- and bi-bi-antennary transferrin C1), using Concanavalin A-Sepharose. 2. Each of these fractions was separated into its sialic acid-dependent subfractions by preparative isoelectric focusing. Sixteen iso-transferrin C1 fractions were obtained, which differed in their degree of glycan branching and/or their sialic acid content. 3. Preliminary carbohydrate analyses suggest that in some iso-transferrins the N-acetylglucosamine and the galactose content is lower than expected.

Catabolism of plasminogen glycoforms I and II in rabbits: relationship to plasminogen synthesis by the rabbit liver in vitro

The metabolisms of the two glycoforms of rabbit plasminogen have been compared in rabbits. Plasminogen I and II (ratio in plasma, 1:2.2) differ only in glycan content: plasminogen I probably possesses one N-glycan and one O-glycan, and plasminogen II only one O-glycan. New Zealand White (NZW) rabbits were injected intravenously with 125I-plasminogen I and 131I-plasminogen II, and blood samples were taken at regular intervals over 5 days. Kinetic behaviors were determined from protein-bound radioactivities using a three-compartment model. Fractional catabolic rates for plasminogen II in the vascular space (2.42 d-1) and the total body (0.56 d-1) were significantly greater than those measured for plasminogen I (1.12 and 0.45 d-1); half-lives were 1.53 and 1.23 days for plasminogen I and II, respectively (P < .01). Fractional distributions among the vascular, noncirculating vascular, and extravascular compartments were 0.41, 0.13, and 0.46 for plasminogen I, and 0.23, 0.11, and 0.65 for plasminogen II. From these data, we determined that plasminogen II was catabolized 4.8 times more rapidly than plasminogen I and was quantitatively contained largely in the extravascular space. By comparison, perfusion of rabbit livers ex corpora showed that plasminogen II was synthesized and released 5.0 times faster than plasminogen I over a 5-hour period. The possible roles for these glycoforms in vivo with respect to their different turnover rates and compartmental distributions are discussed.

Clinical applications of lectin affinity electrophoresis

Principles and several modifications of lectin affinity electrophoresis are described. The results obtained using these newly developed techniques are reviewed for individual glycoproteins, the altered lectin reactivities of which have some clinical implications, showing different lectin reactivities, which occur not only on malignant transformation but also in association with inflammatory process and hormonal action.

Physiological significance of the marked increased branching of the glycans of human serotransferrin during pregnancy

Human serotransferrin (Tf) presents a microheterogeneity based on the existence of biantennary and triantennary glycans of the N-acetyl-lactosaminic type. By affinity chromatography on a concanavalin A-Sepharose column in well-defined conditions, human serotransferrin isolated from healthy donors was resolved into three carbohydrate molecular variants: Tf-I (less than 1%), Tf-II (17 +/- 2%) and Tf-III (82 +/- 3%) containing two triantennary glycans, one triantennary and one biantennary glycans and two biantennary glycans respectively. In addition, two 'isomers' of the triantennary glycans containing the third antenna beta-1,4-linked to the alpha-1,3-mannose residue or beta-1,6-linked to the alpha-1,6-mannose residue were characterized by methylation analysis in the ratio 1:1 in both Tf-I and Tf-II variants. On concanavalin A crossed immuno-affinity electrophoresis, the patterns exhibited by each of the three purified variants or by a mixture of these variants were compared with the patterns given by transferrin present in sera from nonpregnant and pregnant women. The results suggest that the relative proportions of transferrin carbohydrate variants was unchanged when the concentration of transferrin was increased in serum from normal donors, whereas in the serum of pregnant women, especially in the last 3 months of pregnancy, when the serum concentration of transferrin reached 4.5-5 g/l, the relative proportions of the carbohydrate variants Tf-I and Tf-II increased from 1 to 6 +/- 1% and from 17 +/- 2 to 26 +/- 3% respectively while that of Tf-III decreased from 82 +/- 3 to 67 +/- 3%. The binding of the three transferrin carbohydrate variants to the receptor of the syncytiotrophoblast plasma membranes was determined by using Scatchard-plot analysis. The number of binding sites remained constant with an increase in the number of triantennary glycans whereas a decrease up to 6-fold in the affinity constant was observed. Detection of the transferrin-receptor complex by immunoblotting in the presence of non-dissociating detergents revealed the existence of only one type of receptor or of a receptor possessing similar properties involved in the binding of each of the three serotransferrin carbohydrate variants.

Functional domains of the T lymphocyte plasma membrane: characterization of the polypeptide composition

Highly purified plasma membranes from calf thymocytes were fractionated by affinity chromatography on Concanavalin A-Sepharose into two subfractions, one eluting freely from the affinity column (MF1) and a second being specifically retained (MF2). SDS-polyacrylamide-gel-electrophoresis revealed different polypeptide patterns of the two plasma membrane subfractions. Polypeptides of apparent molecular weights of 170, 150, 110, 94, 39, and 30 kDa were several-fold enriched in the adherent fraction, MF2. In contrast, several proteins in the 55-65 kDa range were preferentially recovered in the non-adherent fraction. Five Five of the six polypeptides, preferentially recovered in MF2 proved to be glycoproteins, the 39 kDa peptide was non-glycosilated. The differences in the amounts of the polypeptides specifically enriched in the adherent fraction MF2 became even more clear-cut when plasma membranes solubilized with non-ionic detergents (lysolecithin, ET-18-2H, Triton-X-100) were separated by affinity chromatography on Concanavalin A-Sepharose. The non-glycosilated peptide of apparent molecular weight of 39 kDa was recovered together with several glycoproteins in the adherent fraction, MF2, suggesting that not single glycoproteins, but plasma membrane domains were separated by Concanavalin A-Sepharose. Although the glycoproteins of the non-adherent fraction MF1 bound significant amounts of Concanavalin A, the major Concanavalin A binding glycoproteins were recovered in the adherent fraction, MF2. The plasma membrane subfractions showed also different functional properties, the specific activities [Na+ + K+]AT-Pase, Ca2+ ATPase and lysolecithin acyltransferase were several-fold enriched in the adherent fraction, MF2, as compared to MF1. The data suggest the existence of plasma membrane domains in the plasma membranes of thymocytes consisting of a different set of proteins, among others the major Concanavalin A binding glycoproteins with some membrane bound enzymes, probably implicated in the initiation of lymphocyte activation.

The shapes of biantennary and tri/tetraantennary alpha 1 acid glycoprotein by small-angle neutron and X-ray scattering

Two forms of alpha 1 acid glycoprotein (orosomucoid) have been studied using small-angle neutron and X-ray scattering techniques; in one form all the five glycan chains were biantennary, while in the other they were either triantennary or tetraantennary. The radius of gyration RG was found to be sensitive to salt for the biantennary form, but to be unchanged up to an ionic strength of 3 M for the triantennary and tetraantennary forms. Conformational heterogeneity is thus associated with carbohydrate heterogeneity. Hydrodynamic frictional coefficients <f> confirm these findings. Simple models of alpha 1 acid glycoprotein were developed to account for the RG and <f> values. These show that the compact conformation is slightly more elongated than a globular protein and that the expanded biantennary conformation has a most extended carbohydrate structure. Up to half of the surface of the compact shape can be covered by carbohydrate residues.

External labelling of glycoproteins from first-trimester human placental microvilli

The brush-border glycoproteins of first-trimester human placentas were investigated by using two external labelling techniques: (1) sequential digestion with neuraminidase and galactose oxidase, followed by reduction with NaB3H4, which 3H-labels terminal galactose and galactosamine residues; and (2) sequential treatment with periodate and NaB3H4, which 3H-labels terminal sialic acid residues. The labelling procedures were performed on intact tissue so that the results would more closely approximate the topography of the brush border in vivo. The microvilli were isolated, subjected to sodium dodecyl sulphate/polyacrylamide-gel electrophoresis, and the [3H]glycoproteins detected by fluorography. Densitometer scans of the fluorograms of the [3H]galactoproteins showed that, under reducing conditions, 90% of the protein-associated radioactivity was incorporated into two glycoproteins. The major [3H]galactoprotein of early placental microvilli had an estimated molecular mass of 92 kDa (desialylated) and migrated as a diffuse band. A minor 180 kDa glycoprotein was less consistently labelled. No change in the apparent molecular mass of either component was detected in the absence of beta-mercaptoethanol, suggesting that the 180 kDa component was not a dimer of the 92 kDa glycoprotein. The remaining 10% the radioactivity was equally distributed among several minor membrane components. Densitometer scans of the fluorograms of the [3H]sialoproteins showed that, under either reducing or non-reducing conditions, 90% of the 3H was preferentially incorporated into the 92-110 kDa region of the gel. Although no distinct bands were visible, the higher-molecular-mass region of this area was always most heavily labelled. A minor 180 kDa glycoprotein was also 3H-labelled. The pattern of brushborder [3H]glycoproteins from first-trimester placentas differed markedly from that of term placental microvilli and from placental fibroblast plasma membranes that were 3H-labelled by identical external labelling techniques. These results indicate that: (1) the glycoprotein determinants of brush-border topography change during pregnancy; (2) within the placenta, the major 92 kDa (desialylated) determinant, which has not been previously described, is unique to the trophoblastic cells.

Characterisation of Concanavalin A-binding glycoproteins from mouse splenic leukocytes by two-dimensional electrophoresis: preferential binding of incompletely glycosylated forms of H-2 antigen to the lectin

Concanavalin A (Con A)section-binding proteins from mouse spleen leukocytes were characterised by two-dimensional electrophoresis of material precipitated, by Con A plus anti-Con A, from lysates of biosynthetically-labelled cells. Although most cell surface (iodinatable) proteins are known to bind Con A, some of the major Con A-binding proteins detected by immunoprecipitation, after a four-hour biosynthetic labelling period, are not iodinatable and are probably intracellular. Thus the major biosynthetically labelled Con A-binding species are: (i) a non-iodinatable, high molecular weight glycoprotein (C-145); (ii) intracellular precursors of secretory immunoglobulins (IgM and, probably, IgA); (iii) immature (not fully-sialylated) forms of H-2 D and K antigens; and (iv) Ia antigens. In the case of the H-2 antigens, (and possibly of other cell surface proteins) the selection of immature forms by Con A is not due to lack of biosynthetic labelling of mature products, but to preferential binding of Con A to incompletely glycosylated molecules.