Antibodies to a strain-specific citrullinated Epstein-Barr virus peptide diagnoses rheumatoid arthritis

Abstract

Rheumatoid arthritis (RA) is a chronic systemic autoimmune disease. Anti-citrullinated protein antibodies (ACPA) are crucial for the serological diagnosis of RA, where Epstein-Barr virus (EBV) has been suggested to be an environmental agent in triggering the onset of the disease. This study aimed to analyse antibody reactivity to citrullinated EBV nuclear antigen-2 (EBNA-2) peptides from three different EBV strains (B95-8, GD1 and AG876) using streptavidin capture enzyme-linked immunosorbent assay. One peptide, only found in a single strain (AG876), obtained a sensitivity and specificity of 77% and 95%, respectively and showed high sequence similarity to the filaggrin peptide originally used for ACPA detection. Comparison of antibody reactivity to commercial assays found that the citrullinated peptide was as effective in detecting ACPA as highly sensitive and specific commercial assays. The data presented demonstrate that the citrullinated EBNA-2 peptide indeed is recognised specifically by RA sera and that the single peptide is able to compete with assays containing multiple peptides. Furthermore, it could be hypothesized that RA may be caused by (a) specific strain(s) of EBV.

Figure 1

Reactivity of rheumatoid arthritis and healthy donor sera to selected linear EBNA-2 peptides originating from three Epstein-Barr virus strains anslysed by streptavidin capture ELISA. *Specificity is calculated based on the reactivity of 10 healthy donor sera (HD) to the specific peptides. Statistical calculations are performed using the Student’s t-test, where antibody reactivity to healthy controls is used for comparison. a. Reactivity of rheumatoid arthritis sera (n = 15). b. Reactivity of healthy donor sera (n = 10).

Figure 2

Overlapping antibody reactivities to Cit-Gly-containing peptides among the three Epstein-Barr virus strains; B95-8, GD1 and AG876. The colored fields represent the number of peptides recognised for the individual sera, and hence the degree of overlapping reactivities in a serum pool.

Figure 3

Reactivity of rheumatoid arthritis sera and healthy donor sera to linear and cyclic EBNA-2 peptides and control peptides analysed by streptavidin capture ELISA. (a) Reactivity of rheumatoid arthritis sera (n = 10) to cyclic and linear N/C-terminally biotinylated EBNA-2 peptides (amino acids 313–333 of Epstein-Barr virus strain AG876). “B” represents the location of the biotin labeling in relation the “EBNA” peptide. L = linear, C = cyclic. Linear peptide: GQGRGRWRG-Cit-GRSKGRGRMH-B, cyclic peptide: GQGRCGRWRG-Cit-GRSKGRGCRMH-B. Statistical calculations were performed using the Student’s t-test, where antibody reactivity to healthy controls was used for comparison. (b) Reactivity of healthy donor sera (n = 10) to linear N/C-terminally biotinylated EBNA-2 peptides. (c) Reactivity of rheumatoid arthritis sera (n = 20) to linear and cyclic EBNA-2 peptide linked to a C-terminal biotin (amino acids 313–333 of Epstein-Barr virus strain AG876). Non-citrullinated peptides (Arg) were used at controls. Statistical calculations were performed using the Student’s t-test, where antibody reactivity to non-citrullinated peptides is used for comparison. (d) Reactivity of healthy donor sera to linear and cyclic EBNA-2 peptides linked to a C-terminal biotin (n = 20).

Figure 4

Reactivity of rheumatoid arthritis sera and healthy donor sera to substituted and truncated linear EBNA-2 peptides analysed by streptavidin capture ELISA. Statistical calculations were performed using the Student’s t-test, where antibody reactivity to healthy controls is used for comparison. (a) Reactivity of rheumatoid arthritis sera to EBNA-2 peptides (n = 10). (b) Reactivity of healthy donor sera to EBNA-2 peptides (n = 10).

Figure 5

Reactivity of rheumatoid arthritis sera and control sera to the linear EBNA-2 peptide analysed by streptavidin capture ELISA and in the commercial CCPlus and CCP3.1 assays. The following sera were selected for analysis; RA (n = 126), HD (n = 80), SLE (n = 20) and SjS (n = 40). (a) Reactivity of RA sera and control sera in the CCP3.1 assay. (b) Reactivity of RA sera and control sera in the CCPlus assay. (c) Reactivity of RA sera and control sera to the linear EBNA-2 peptide.