Abstract

Many auto-immune diseases are hallmarked by the presence of auto-reactive B cells that can develop into autoantibody-secreting plasma cells. In most cases, the secreted autoantibodies have extensively been studied in their role as disease-associated biomarkers, and for some, specific pathogenic effector functions have been demonstrated supporting the use of interventions that target the plasma cell compartment. In other cases such as rheumatoid arthritis (RA), however, the pathogenicity of specific autoantibodies is less clear, and the therapeutic efficacy of B cell depleting therapy that spares the plasma cell compartment indicates that auto-reactive B cells themselves can have pathogenic effector functions that contribute to disease. In this context, it is of great interest to understand the mechanisms that allow auto-reactive B cells to emerge from the naïve repertoire, a process that marks the onset of systemic autoimmunity and frequently precedes the clinical onset of disease.

RA is characterized by a remarkable appearance of autoantibodies that target post-translational modifications of proteins, of which anti-citrullinated protein antibodies (ACPA) display the highest specificity for disease. In addition, ACPA associate with active destructive RA and pose individuals with arthralgia at-risk for progression towards arthritis. We recently found that ACPA display significant alterations with regard to the glycosylation of both the Fc-tail as well as the F(ab)-domain. While the ACPA-IgG Fc-tail loses galactose and sialic acid residues prior to the onset of arthritis, a process associated with enhanced inflammatory antibody activity that occurs potentially under the influence of IL-17 producing T cells, ACPA also carry abundant glycans in the antigen-binding region of the F(ab) domain. These latter glycans are reminiscent of F(ab)-glycans found in follicular lymphoma B cells, and were here identified as N-linked, biantennary glycans composed of a remarkably high frequency of sialic acid residues. Notably, N-glycosylation requires the presence of glycosylation consensus sequences in the protein. As such sequences are normally scarce in the germline encoded variable regions of B cell receptors (BCR), the acquisition of F(ab)-glycans requires mutations of the amino-acid sequence to generate N-glycosylation sites. This process is mediated by somatic hypermutation, which is mainly induced by T cells that provide help to B cells. As >90% of all ACPA-IgG molecules carry such F(ab)-linked N-glycans, and as protective antibodies in the same individuals and many autoantibodies in other diseases do not show this feature, it is conceivable that the acquisition of F(ab)-glycans by ACPA-IgG is a T cell –mediated process that provides a selective advantage to ACPA-expressing B cells. This notion is supported by the observation that additional F(ab)-glycans are not found on ACPA-IgM. How F(ab)-glycans facilitate the emergence and/or expansion of auto-reactive B cells in this context, however, remains unclear. Using recently developed technology to identify and isolate citrullinated antigen-specific B cells from patients, we can now address this question by studying the frequency and localisation of N-glycosylation sites in the antibody repertoire and by studying the phenotype and functional characteristics of ACPA-expressing B cells. Together with our investigations on the modulation of ACPA Fc-glycans, these studies provide a deeper understanding of mechanisms that allow the development of autoimmunity as such, and of the mechanisms that underlie the progression from systemic autoimmunity towards overt autoimmune disease.