Background/Purpose

Disease-associated HLA-DR molecules are the greatest genetic risk factor for rheumatoid arthritis (RA) and for antibiotic-refractory Lyme arthritis (LA). We have developed a novel approach to identify disease-relevant HLA-DR-presented peptides in synovial tissue using discovery-based proteomics and translational research. Using this approach, we previously identified endothelial cell growth factor (ECGF) as the first autoantigen known to be a target of T and B cell responses in antibiotic-refractory or antibiotic-responsive LA, but not in RA. In this study, we continued to characterize the repertoire of naturally presented HLA-DR peptides in synovial tissue in patients with RA or LA.

Methods

HLA-DR-presented peptides were eluted from synovia, identified by tandem mass spectrometry, synthesized, and tested for reactivity with the matching patient’s PBMC. Immunoreactive peptides or their source proteins were then tested for T cell reactivity by IFN-g ELISpot assay or for antibody responses by ELISA. All RA patients met the 2010 ACR/EULAR criteria for RA and the LA patients met the CDC criteria for Lyme disease.

Results

In one RA patient who lacked positive tests for rheumatoid factor (RF) or anti-citrullinated protein antibodies (ACPA), 1 of 86 non-redundant HLA-DR-presented peptides identified from her synovial tissue induced her PBMC to secrete IFN-g. The peptide was derived from the protein annexin A2 and others had shown that ~10% of RA patients make autoantibodies against this self-protein. We tested serum samples from our cohort of RA patients for anti-annexin A2 autoantibodies, and for comparison, from healthy control subjects or from patients with antibiotic-responsive or antibiotic-refractory LA.  In our RA cohort, sera from 24% of 91 patients had antibody responses to annexin A2 that were >3SD above the mean value in healthy control subjects (Figure). Surprisingly, about 20% of the patients with antibiotic-responsive or antibiotic-refractory LA also had antibody reactivity with this autoantigen. In annexin A2-positive RA patients, the magnitude of antibody responses to ACPA or RF were less than in annexin A2-negative patients. Studies to look for linked T and B cell responses to annexin A2 are currently in progress in both the RA and LA cohorts.

Conclusion

We confirm that annexin A2, a phospholipid-binding protein that protects damaged endothelial cells, is a potential autoantigen in a subgroup of patients with RA. Moreover, we report for the first time that this protein may serve as an autoantigen in a subgroup of patients with LA. As with reactivity to ECGF, autoantibody responses to annexin A2 in Lyme disease seem to occur as a part of the immune response to the infection, whereas additional factors, such as immune dysregulation, are required for refractory arthritis.