Amyloid ß Binds to Albumin-Associated Lrp-Like Plasma O-Glycoproteins: Albumin Prevents Inhibition of Binding by LDL.

BACKGROUND: Albumin was reported to engage nearly 95% of plasma Amyloid ß (Aß) and to reverse Aß fibril formation in brain. OBJECTIVE: Since O-glycosylated LRP family of receptors capture Aß in brain we compared Aß binding to electrophoretically purified albumin and to O-glycoproteins AOP1 and AOP2 that adhere noncovalently to plasma albumin. METHODS: Strength of Aß-protein interaction was measured as fluorescence increase in Fluorescentlabeled Aß (F-Aß) resulting from conformational changes. Alternatively, differential segregation of free and protein-bound Aß in Density Gradient Ultracentrifugation (DGUC) was also examined. RESULTS: Fluorescence enhancement in F-Aß was significantly greater by AOP1 and AOP2 than by known Aß reactants α -synuclein and ß -cyclodextrin, but nil by albumin. In DGUC Aß migrated with the O-glycoproteins but not with albumin. Free O-glycoproteins unlike their albumin-bound forms were blocked by LDL from capturing F-Aß. Associated albumin did not affect Aß binding of O-glycoproteins. De-O-glycosylation of AOP1/AOP2 enhanced their Aß binding showing that peptide sequences at O-glycosylated regions were recognized by Aß. Unlike albumin, AOP1 and AOP2 were immunologically cross-reactive with LRP. Albumin sample used earlier to report albumin-Aß interaction contained two O-glycoproteins cross-reactive with human LRP and equal in size to human AOP1 or AOP2. CONCLUSION: Unlike albumin, albumin-bound O-glycoproteins, immunologically cross-reactive with LRP, bind plasma Aß. These O-glycoproteins are potential anti-amyloidogenic therapeutics if they inhibit Aß aggregation as other Aß reactants do. Circulating immune complexes of albuminbound O-glycoproteins with O-glycoprotein-specific natural antibodies can bind further to LRP-like membrane proteins and are possible O-glycoprotein transporters to tissues.

Anti-α-galactoside and Anti-ß-glucoside Antibodies are Partially Occupied by Either of Two Albumin-bound O-glycoproteins and Circulate as Ligand-binding Triplets.

Two heavily O-glycosylated proteins and albumin co-purified with anti-α-galactoside (anti-Gal), the chief xenograft-rejecting antibody and anti-ß-glucan (ABG) antibody isolated from human plasma by affinity chromatography on respective ligand-bearing matrices. Both antibodies and O-glycoproteins co-purified with plasma albumin eluted from albumin-specific matrix. Using components of affinity-purified antibody samples separated by electrophoresis binding of either albumin or antibody to the affinity matrix of the other or binding of O-glycoprotein to either matrix was ruled out. Enzyme-linked immunoassay and ligand-induced fluorescence enhancement of fluorolabeled antibody showed that O-glycoproteins occupied sugar-binding sites of anti-Gal and ABG. Neither antibody recognized albumin. O-Glycoprotein-albumin complexes free in plasma, or released from antibodies by specific sugars, were captured on microwell-coated O-glycan-specific lectin jacalin and detected using labeled anti-albumin. We conclude that circulating anti-Gal and ABG form protein triplets in which either O-glycoprotein bridges between antibody and albumin by binding simultaneously to both. Bound albumin restricted O-glycoprotein occupation on antibodies enabling triplets to bind other ligands using spared binding sites. Free anti-Gal and ABG were undetectable in plasma. Jacalin treatment, but not de-O-glycosylation of O-glycoproteins abolished their recognition by anti-Gal or ABG indicating that antibodies recognized serine- and threonine-rich peptide sequences that underlie the O-glycans and are reported surrogate ligands for anti-Gal. The albumin- and antibody-binding O-glycoproteins AOP1 and AOP2 were single polypeptide proteins of size 107 kDa and 98 kDa, containing 54% and 51% carbohydrate respectively and conformed to no known plasma protein in properties. Prospects of triplet-mediated modulations in autologous tissues expressing antibody ligands are discussed.

Plasma anti-α-galactoside antibody mediates lipoprotein(a) binding to macrophages.

Lipoprotein (a) [Lp(a)] is the dominant lipid in atherosclerotic plaques though it is much less numerous than LDL or HDL in circulation. Molecular mechanism of selective uptake of Lp(a) into macrophages is unclear. Lp(a) was reported to form circulating immune complexes with the IgG-dominated plasma anti-α-galactoside antibody (anti-Gal) using the serine- and threonine-rich peptide sequences ( STPS) on its apo(a) subunit as surrogate ligand but left the other binding site of antibody free. We examined if these monovalent immune complexes could bind to smaller STPS-containing molecules on macrophage surface. Using placental membrane O-glycosylated proteins (PMOP) isolated by lectin affinity chromatography as model it was shown that human cell surface glycoproteins were small enough to occupy both binding sites of anti-Gal since they increased the fluorescence of FITC label at Fc part of anti-Gal and inhibited binding of anti-Gal and Griffonia simplicifolia lectin of similar specificity to immobilized ligands. Pre-incubation with anti-Gal facilitated Lp(a) attachment to macrophages unless anti-Gal-specific sugar was present. Anti-Gal-mediated attachment of apo(a) to macrophages increased with the number of apo(a) subunits. Further, anti-Gal-mediated binding of the same sample of apo(a) increased with the specific activity of anti-Gal sample. Finally binding of anti-Gal and anti-Gal-apo(a) complex to PMOP and macrophages respectively was mostly inhibited by LDL suggesting STPS as major anti-Gal epitopes on the cell surface. Results indicated that circulating Lp(a)-anti-Gal immune complexes anchor on macrophages using STPS-bearing cell surface glycoproteins as ligands and offer a pathway for Lp(a) sequestration into macrophages.

Antigen-Induced Activation of Antibody Measured by Fluorescence Enhancement of FITC Label at Fc.

Three anti-carbohydrate antibodies of defined specificity isolated from plasma were used to demonstrate that macromolecular antigen binding caused considerable enhancement of fluorescence of FITC-labeled antibody. Mono and disaccharide antigens which could compete with the large antigens in antibody binding could not however produce any increase in fluorescence. Fluorescence enhancement in a given antibody sample increased with the size of the occupying macromolecular antigen. Conversely in antibody samples of same ligand specificity isolated from plasma of different individuals, fluorescence enhancement produced by the same antigen correlated with specific activity of the antibody sample. Removal of Fc part of antibody, confirmed by electrophoresis and Fc-specific antibody binding, caused abolition of most of the antigen-driven fluorescence increase. Since antigen binding sites of antibodies were protected during FITC labeling, the above results suggest that conformational shift in Fc produced by occupation of binding sites by large antigens resulted in the enhancement of fluorescence of FITC tags on Fc. Data provides a tool for detection and measurement of specific ligands using fluorolabeled whole antibodies.

High polymeric IgA content facilitates recognition of microbial polysaccharide-natural serum antibody immune complexes by immobilized human galectin-1.

Dextran-binding immunoglobulin (DIg) and anti-ß-glucan antibody (ABG) are naturally occurring human serum antibodies specific to α- and ß-glucoside epitopes respectively of polysaccharide antigens and heavily enriched in IgA. ABG and DIg are shown here to have much more of their IgA in polymeric form than does serum IgA in general. Cell wall ß-glucans and glycoproteins of the widely consumed yeast (Saccharomyces cerevisiae) offered several hundred fold better ligands for ABG than did small ß-glucosides. Candida albicans cell wall antigen (CCA), a commonly encountered polysaccharide-rich fungal antigen was recognized by normal human serum anti-carbohydrate antibodies to precipitate maximally at a definite stoichiometry typical of immune complexes (IC). IC formed in serum in vitro on addition of CCA contained a significantly higher percentage of IgA than did either naturally occurring IC or serum. Polymeric IgA was far better ligand than monomeric IgA for both anti-IgA antibody and the most widely expressed human tissue lectin galectin-1 which recognizes O-linked oligosaccharides characteristic of IgA, in contrast to N-linked oligosaccharides present in all immunoglobulins. Moreover, desialylation by neuraminidase, an enzyme released into circulation during many microbial infections and diabetes, increased lectin-binding activity of polymeric IgA much more than that of monomeric IgA. Human galectin-1 immobilized in active form in vitro sugar-specifically captured IgA and IgA-containing IC formed by CCA in serum but not IgG. Results suggest that while high IgA content especially in polymeric form may render polysaccharide IC more susceptible to tissue uptake, desialylation of IgA in IC could enhance the process.

Galectin-1, an endogenous lectin produced by arterial cells, binds lipoprotein(a) [Lp(a)] in situ: relevance to atherogenesis.

Lipoprotein(a) [Lp(a)], a modified LDL molecule, is implicated in atherogenesis. Mechanisms of the accumulation of [Lp(a)] in atherosclerotic vessels is lacking in literature. We sought to investigate the complementarities of the carbohydrate structures on Lp(a) and LDL with galectin-1(a carbohydrate binding protein) and whether endogenous galectin-1 binds Lp(a) in situ. We investigated T-antigen structures on Lp(a) and LDL by enzyme-linked lectin assay using T-antigen specific lectins, galectin-1 and jacalin. Both jacalin and galectin-1 bound strongly to Lp(a) and to a much lesser extent, to LDL. Galectin-1 recognition of the lipoproteins was abolished when the O-linked sugars were selectively removed. Localization of endogenous galectin-1 within histological sections of human internal mammary artery and in vitro binding of Lp(a) to the tissues was analyzed by immunohistochemical staining. The Lp(a)-binding pattern was found to overlap with the localization of galectin-1. The poor Lp(a)-binding on inhibiting tissue galectin-1 with lactose, suggested the binding of Lp(a) to galectin-1. This may be suggestive of a mechanism by which Lp(a) accumulates within arterial walls in atherogenesis.