Agalactosyl IgG: an aid to differential diagnosis in early synovitis

The binding of synovial tissue-derived human monoclonal immunoglobulin M rheumatoid factor to immunoglobulin G preparations of differing galactose content

There are several sites on IgG Fc that have been reported to be the epitopes for binding rheumatoid factors (RF). It is now established that there are alterations in the oligosaccharides on IgG from patients with rheumatoid arthritis and it has been suggested that these changes may enhance immune complex and cryoglobulin formation. We have used a series of IgG preparations differing in their content of oligosaccharide chains lacking galactose from 18 to 86% to determine whether changes in sugar content affect the binding of rheumatoid factor. Five of 16 monoclonal rheumatoid factors prepared from synovial tissue, from patients with juvenile or adult rheumatoid arthritis, bound better to IgG which was deficient in galactose. Six of the 16 rheumatoid factors from the same patients bound independently of the galactose content. Four of the 16 rheumatoid factors could not be absolutely grouped in this manner but seemed to demonstrate a preference for agalactosyl IgG. One rheumatoid factor bound better to fully galactosylated IgG. There was an association between enhanced binding to galactose-deficient IgG and monoreactivity and a very strong association between the functional affinity of the rheumatoid factors and the dependent binding.

IgG glycosylation in autoimmune-prone strains of mice

The relationship between increased levels of IgG oligosaccharide chains lacking galactose (G0) and the development of rheumatoid arthritis is unclear. In order to further our understanding of the observed correlation between raised serum G0 and arthritis, we have studied G0 levels in arthritis-prone and non-susceptible (i.e. non-arthritis-prone) mice and the effects on G0 of mycobacterial antigens, which have been postulated to play a role in the early events leading to the development of arthritis. We have shown that different age-matched mouse strains have characteristic 'resting' levels of G0 which (in six out of seven strains of mice) increase with age. We have also shown that these increases can be enhanced by immunization of arthritis-prone strains of mice with an adjuvant containing mycobacteria (Freund's complete adjuvant (FCA)), suggesting that deflects in the ability to regulate these G0 changes may be related to susceptibility to arthritis.

Differential B lymphocyte galactosyltransferase activity in the MRL mouse model of rheumatoid arthritis

Oligosaccharides can be of fundamental importance to glycoprotein function. Glycosylation abnormalities are present in rheumatoid arthritis (RA) and may be associated with disease pathogenesis. To determine whether similar disease mechanisms occur in the MRL-1pr/1pr autoimmune arthritic mouse, studies on B lymphocyte galactosyltransferase (GTase) have been carried out. In MRL mice, a significant reduction in peripheral blood lymphocyte (PBL) GTase activity was found when compared to their paired splenic (SP) GTase activity (-69%, p = 0.002) and histocompatible non-autoimmune control CBA/Ca mice (-67%; p = 0.002). The changes in PBL GTase activity are similar to those found in RA and on further analysis, using mixing experiments in the presence of purified human milk GTase, this reduction was shown not to be due to the presence of a soluble intracellular GTase inhibitor. Furthermore when examining MRL derived hybridoma cells producing IgG, significantly reduced GTase activity was detected in the rheumatoid factor (RF) producing hybridoma cells compared to those secreting an irrelevant antibody (-21%, p < 0.05). Together these findings suggest that the glycosylation changes observed in this study, and those reported previously in RA, are tissue-specific, may result from cells trafficking from centres of disease activity and are not the result of direct enzyme inhibition. It is now important to further understand the mechanisms controlling glycosylation and relate disease associated changes with those occurring as part of normal cellular physiology.

Second Jenner international glycoimmunology meeting

Images

Abnormal galactosylation of serum IgG in patients with systemic lupus erythematosus and members of families with high frequency of autoimmune diseases

Gas chromatographic carbohydrate analyses of IgG from 30 patients with idiopathic systemic lupus erythematosus (SLE) revealed lower content of galactose when compared to that in 36 controls of similar ages (mean +/- SD, 3.18 +/- 0.66 vs 3.82 +/- 0.41 galactose residues/mole of IgG, P < 0.001). Abnormal galactosylation was observed in 60% of SLE patients. Analyses of IgG from 58 members of five families, characterized by a high frequency of SLE and other autoimmune diseases and serological abnormalities, and 51 controls of similar age range revealed that IgG galactose deficiency was detectable not only in some members with clinical and serological abnormalities (P < or = 0.001), but also in those without evidence of autoimmune diseases or abnormal serologies (P < or = 0.001). These data indicate that abnormal galactosylation of IgG frequently occurs in asymptomatic members of families with a high frequency of SLE and other autoimmune diseases and suggests that this abnormality may be an indicator for the development of these diseases.

Changes in normal glycosylation mechanisms in autoimmune rheumatic disease

To investigate potential mechanisms controlling protein glycosylation we have studied the interrelationship between lymphocytic galactosyltransferase (GTase) activity and serum agalactosylated immunoglobulin G levels (G(0)) in healthy individuals and patients with rheumatoid arthritis and non-autoimmune arthritis. In RA there was reduced GTase activity and increased G(0). A positive linear correlation between B and T cell GTase was found in all individuals. The relationship between GTase and G(0) was found to be positive and linear in the control population and negative and linear in the RA population. Sulphasalazine therapy maintained normal levels of GTase and caused a reduction in G(0) in the RA population. IgG anti-GTase antibodies (abs) were significantly increased in the RA population, whereas IgM anti-GTase abs were significantly decreased in both the RA and the non-autoimmune arthritis groups. These data describe a defect in RA lymphocytic GTase, with associated abnormal G(0) changes, which is corrected by sulphasalazine. A possible regulatory mechanism controlling galactosylation in normal cells is suggested, in which there is parallel control of B and T cell GTase. IgM anti-GTase abs may be integrated into this normal regulatory process. This is disrupted in RA, where the positive feedback between GTase and G(0) is lost and there is an associated increase in IgG anti-GTase abs, which may result from isotype switching as IgM anti-GTase abs are reduced. We suggest that these mechanisms are of relevance to the pathogenesis of RA, and that their manipulation may form part of a novel therapeutic approach.

Abnormalities in the glycosylation of IgG in spouses of patients with rheumatoid arthritis. A family study

Forty-seven members of eight families with a rheumatoid proband were analysed for abnormal glycosylation of IgG. The results (%G(o) which is the percentage of oligosaccharide chains lacking galactose) were corrected for age and expressed as SD units about the mean for the normal population. Seven of 8 probands, 3/8 spouses, 3/5 RA relatives and 4/26 non-RA relatives had %G(o) values greater than 1SD above the age corrected mean for the normal control population (P less than 0.001, less than 0.01, less than 0.005 and greater than 0.5 respectively). A further 13 spouse pairs were studied. Ten of 13 probands and 8/13 spouses had %G(o) values greater than 1SD above the mean (P less than 0.001 and less than 0.001 respectively). Thus in total, a strikingly high number of unaffected spouses had high %G(o) values (11/21). IgM, IgA and IgG rheumatoid factors were studied. While RA patients' sera showed a correlation between IgM and IgA rheumatoid factors and %G(o), (IgM, r = 0.41 0.05 greater than P greater than 0.02, IgA, r = 0.36, P = 0.05), no correlation between IgG RF and %G(o) was noted in the RA patients. No correlation was found between any of the RF classes and %G(o) in spouses and non-RA relatives.