Background/Purpose:

Anti-citrullinated protein antibodies (ACPA) were first identified based on their high specificity for RA and now are commonly used as a diagnostic tool. Yet, despite the success of B cell targeted therapy, and evidence that ACPA may arise long before overt inflammatory arthritis, the roles of human B cells autoreactive for citrullinated protein antigens have been little explored.

Methods:

For comprehensive B cell profiling, we developed a 12-color flow cytometry panel that assesses diverse B-cell subset specific markers, with live/dead discrimination and myeloid/T-cell exclusions. Fluorochrome-labeled tetrameric forms were made with cyclic citrullinated peptide (CCP) used in clinical testing, as well as a panel of self-protein derived peptides implicated in RA, and control peptides with a single amino acid type substitution. ACPA labeled beads were included for internal calibration. Multivariate methods were used to identify natural divisions within B cell data sets, and Spearman coefficients to assess for correlations between cytometric data and clinical/laboratory data. 

Results:

During panel development, we performed cross-sectional studies of a total of 30 RA and 16 healthy adults. Pilot surveys demonstrated only a low background levels of CD19+ B-cell binding to control CQP tetramer in  RA and healthy adults (< 0.3% mean). While levels of B cell binding to CCP were overlapping between the groups, as some RA were indistinguishable from healthy controls, B cells from RA pts displayed significantly higher binding to CCP (7.46%+/-6.8, mean +/- SD) compared to binding of a control tetramer (p<0.0001, 2-tailed Mann-Whitney test). In RA, we found expansions of CCP-reactive B cells in diverse cellular subsets, which varied between subjects, and which included naïve mature B cells (IgD+ CD27-) and both CD27+IgD- and CD27- IgD- memory subsets. In each sample, CCP-binding by B-cell subsets was independent of relative BCR levels, as assessed by MFI. There were no significant relationships between levels of CCP-binding B cells and IgG, IgA or IgM anti-CCP or IgG to panel of peptide-specific ACPAs. Importantly, our surveys demonstrated a significant correlation between levels of CCP-binding by naïve mature (CD27- IgD+) B cells and DAS28 score (n=18, p=0.04, r=0.48). We also found correlations between CD95 expression (an activation marker) on CCP-binding B cells and DAS28 score (p=0.03, r=0.56) that was improved by exclusion of pts on biologic agents (n=7, p=0.01, r=0.89).

Conclusion:

These studies have provided the first direct measurements of trafficking disease-specific B cells to citrullinated self-antigens. Our initial surveys found no relationship between levels of circulating CCP-reactive B cells with levels of serum ACPA, which therefore may reflect the contribution of sources no longer linked to synovial disease. However, we did find evidence of an association between levels of CCP-reactive B cells, and especially activated B cells, and overall disease activity by DAS28 score. Our investigative approach may therefore help to better stratify patients and identify those in whom autoantigen-specific lymphocytes are direct drivers of pathogenesis.

---

« Back to 2012 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/characterization-of-circulating-human-b-cells-that-bind-cyclic-citrullinated-peptide-antigens-in-clinically-active-rheumatoid-arthritis/