Reversible changes in serum immunoglobulin galactosylation during the immune response and treatment of inflammatory autoimmune arthritis.

OBJECTIVES: Improved DNA sequencer-aided fluorophore-assisted carbohydrate electrophoresis (DSA-FACE) technology was used to monitor the changes in the galactosylation status of serum immunoglobulins during the immune response and therapy of autoimmune arthritis. METHODS: Collagen-induced arthritis (CIA) was induced in susceptible DBA/1 mice and the undergalactosylation status (UGS) of serum immunoglobulins was determined using the improved DSA-FACE technology. Prophylactic intravenous tolerisation with type II collagen as well as semitherapeutic treatment with dexamethasone (DEX) were performed and UGS was analysed. Next, the serum immunoglobulin glycosylation profiles of patients with rheumatoid arthritis (RA) and spondyloarthropathy (SpA) were studied and changes in the UGS scores during anti-tumour necrosis factor (TNF)alpha therapy followed. RESULTS: In the longitudinal CIA study, the undergalactosylation state of immunoglobulins was found to be significantly correlated with the clinical arthritis scores. Upon collagen-specific tolerisation as well as glucocorticoid semitherapeutic treatment, improvement of the clinical arthritis scores correlated with decreased levels of UGS. It was also demonstrated that withdrawal of DEX was associated with an increased UGS score. Interestingly, reversibility in the UGS was also shown during treatment of patients with RA and SpA with anti-TNFalpha. CONCLUSIONS: These findings demonstrate that the UGS of serum immunoglobulins changes during the disease course of CIA and that this UGS is inhibited by antigen-specific and antigen-independent treatment procedures. The observation that Ig galactosylation is a reversible process is also documented during treatment of patients with RA and SpA with anti-TNFalpha.

Localization of major gangliosides in the PNS: implications for immune neuropathies.

Antibodies targeting major gangliosides that are broadly distributed in the nervous system are sometimes associated with clinical symptoms that imply selective nerve damage. For example, anti-GD1a antibodies are associated with acute motor axonal neuropathy (AMAN), a form of Guillain-Barré syndrome that selectively affects motor nerves, despite reports that GD1a is present in human axons and myelin and is not expressed differentially in motor versus sensory roots. We used a series of high-affinity monoclonal antibodies (mAbs) against the major nervous system gangliosides GM1, GD1a, GD1b and GT1b to test whether any of them bind motor or sensory fibres differentially in rodent and human peripheral nerves. The following observations were made. (i) Some of the anti-GD1a antibodies preferentially stained motor fibres, supporting the association of human anti-GD1a antibodies with predominant motor neuropathies such as AMAN. (ii) A GD1b antibody preferentially stained the large dorsal root ganglion (DRG) neurones, in keeping with the proposed role of human anti-GD1b antibodies in sensory ataxic neuropathies. (iii) Two mAbs with broad structural cross-reactivity bound to both gangliosides and peripheral nerve proteins. (iv) Myelin was poorly stained; all clones stained axons nearly exclusively. Our findings suggest that anti-ganglioside antibody fine specificity as well as differences in ganglioside accessibility in axons and myelin influence the selectivity of injury to different fibre systems and cell types in human autoimmune neuropathies.

Use of HDEL-tagged Trichoderma reesei mannosyl oligosaccharide 1,2-alpha-D-mannosidase for N-glycan engineering in Pichia pastoris.

Therapeutic glycoprotein production in the widely used expression host Pichia pastoris is hampered by the differences in the protein-linked carbohydrate biosynthesis between this yeast and the target organisms such as man. A significant step towards the generation of human-compatible N-glycans in this organism is the conversion of the yeast-type high-mannose glycans to mammalian-type high-mannose and/or complex glycans. In this perspective, we have co-expressed an endoplasmic reticulum-targeted Trichoderma reesei 1,2-alpha-D-mannosidase with two glycoproteins: influenza virus haemagglutinin and Trypanosoma cruzi trans-sialidase. Analysis of the N-glycans of the two purified proteins showed a >85% decrease in the number of alpha-1,2-linked mannose residues. Moreover, the human-type high-mannose oligosaccharide Man(5)GlcNAc(2) was the major N-glycan of the glyco-engineered trans-sialidase, indicating that N-glycan engineering can be effectively accomplished in P. pastoris.

Characterization of sialyltransferase mutants using surface plasmon resonance.

Sialyltransferases are enzymes responsible for the important sialylation of glycoconjugates. Since crystal structures are not available, other tools are needed to study enzymatic mechanisms. As a model, we used human alpha2,6-sialyltransferase. A putative acceptor-binding domain containing the small and the very small sialyl motifs was randomly mutated. This resulted in enzymes with altered enzymatic activity. Affinity chromatography demonstrated that their binding to donor substrate was maintained. To illustrate the role of the mutated domain in acceptor binding, a method based on surface plasmon resonance was set up. Only at low salt and high acceptor concentration was association of wild-type ST6GalI with asialofetuin demonstrated. As expected, this interaction was affected by cytidine 5'-monophospho-N-acetylneuraminic acid, the donor substrate, which proves the specificity of the interaction. Different types of mutants were found. For some, the drop in activity could be explained by loss in affinity for the acceptor. For others, the catalytic center, but not the acceptor-binding site, was affected. Neither acceptor binding nor catalytic activity were limited to the sialyl motifs. To our knowledge, this is the first example in which surface plasmon resonance is successfully used to demonstrate the binding of a glycosyltransferase to its natural acceptor.

Cloning of Trypanosoma cruzi trans-sialidase and expression in Pichia pastoris.

Trypanosoma cruzi, the agent causing Chagas' disease, expresses an enzyme that transfers sialic acids among glycoproteins and glycolipids both from the host cell surface and its own surface. This enzyme, called trans-sialidase, is different from higher eukaryotic sialyltransferases in that it does not accept cytidine 5'-monophospho-N-acetylneuraminic acid as a donor substrate. Also, the common glycosyltransferase structure is not present. To study this enzyme, an active member was cloned and expressed in higher eukaryotic cells. Expression of recombinant enzyme was achieved in the methylotrophic yeast Pichia pastoris. The N-terminal fusion of a secretion signal and the C-terminal addition of an epitope tag resulted not only in high expression levels, but also enabled easy detection and purification. Using P. pastoris, we obtained about 5 mg of enzymatically active trans-sialidase per liter of induced culture medium.

In vitro conversion of the carbohydrate moiety of fungal glycoproteins to mammalian-type oligosaccharides--evidence for N-acetylglucosaminyltransferase-I-accepting glycans from Trichoderma reesei.

To investigate the potential of filamentous fungi to synthesize N-glycans that are convertible to a mammalian type, in vitro glycosylation assays were performed. Recombinant human N-acetylglucosaminyltransferase I, human beta-1,4-galactosyltransferase and rat alpha-2,6-sialyltransferase were successively used to mimic part of the mammalian glycosylation synthesis pathway. High-mannose carbohydrates on Trichoderma reesei cellobiohydrolase I were converted to a hybrid mammalian-type structure. Successful modification varied markedly with the strain of T. reesei used to produce cellobiohydrolase I. In vitro pretreatment of fungal glycoproteins with Aspergillus saitoi alpha-1,2-mannosidase improved subsequent hybrid formation. It was, however, not possible to trim all fungal oligosaccharides to an acceptor substrate for mammalian glycosyltransferases. With T. reesei RUTC 30, capping glucose residues and phosphate groups were shown to be responsible for this lack of trimming. N-glycan processing in T. reesei apparently involves different steps, including alpha-1,2-mannosidase trimmings, and thus resembles the first mammalian glycosylation processes. The alpha-1,2-mannosidase trimming steps can be exploited for further in vitro and/or in vivo synthesis of complex oligosaccharides on (heterologous) glycoproteins from filamentous fungi.