Identification of novel citrullinated autoantigens of synovium in rheumatoid arthritis using a proteomic approach

Abstract

Recently, autoantibodies to some citrullinated autoantigens have been reported to be specific for rheumatoid arthritis (RA). However, an entire profile of and autoimmunity of the citrullinated proteins have been poorly understood. To understand the profile, we examined citrullinated autoantigens by a proteomic approach and further investigated the significance of citrullination in antigenicity of one of the autoantigens. Specifically, we detected citrullinated autoantigens in synovial tissue of a patient with RA by two-dimensional electrophoresis and Western blotting by using pooled sera from five patients with RA and anti-citrulline antibodies. After identifying the detected autoantigens by mass spectrometry, we investigated the contribution of citrullination to autoantigenicity by using a recombinant protein with or without citrullination on one of the identified novel citrullinated autoantigens. As a result, we found 51 citrullinated protein spots. Thirty (58.8%) of these spots were autoantigenic. We identified 13 out of the 30 detected citrullinated autoantigenic proteins. They contained three fibrinogen derivatives and several novel citrullinated autoantigens (for example, asporin and F-actin capping protein alpha-1 subunit [CapZalpha-1]). We further analyzed the contribution of citrullination to autoantigenicity in one of the detected citrullinated autoantigens, CapZalpha-1. As a result, frequencies of autoantibodies to non-citrullinated CapZalpha-1 were 36.7% in the RA group tested, 10.7% in the osteoarthritis (OA) group, and 6.5% in healthy donors. On the other hand, those to citrullinated CapZalpha-1 were 53.3% in the RA group, 7.1% in the OA group, and 6.5% in the healthy donors. This shows that autoantigenicity of citrullinated or non-citrullinated CapZalpha-1 is relevant to RA. The antibody titers to the citrullinated CapZalpha-1 were significantly higher than those to the non-citrullinated CapZalpha-1 in 36.7% of patients; however, the other patients showed almost equal antibody titers to both citrullinated and non-citrullinated CapZalpha-1. Therefore, the autoantibodies would target citrulline-related and/or citrulline-unrelated epitope(s) of CapZalpha-1. In conclusion, we report a profile of citrullinated autoantigens for the first time. Even though citrullination is closely related to autoantigenicity, citrullination would not always produce autoantigenicity in RA. Citrullinated and non-citrullinated autoantigens/autoepitopes would have different pathological roles in RA.

Figure 1

Detection of citrullinated synovial autoantigens in rheumatoid arthritis (RA). Proteins extracted from synovium of a patient with RA were separated by two-dimensional electrophoresis, and protein spots were stained by SYPRO Ruby gel stain (a). Then, the proteins were transferred onto nitrocellulose membranes and reacted with (c) anti-modified citrulline (anti-MC) antibodies after modification, (d) pooled serum samples from five patients with RA (diluted at 1:500 per person), and (e) pooled serum samples from five healthy donors (diluted at 1:500 per person). For these experiments, the capability of the anti-MC antibodies to detect citrullinated proteins was confirmed by response to the cell lysate of Escherichia coli treated or not treated with peptidylarginine deiminase (PAD) (b). Protein spots that reacted with the RA sera and the anti-citrulline antibodies, but not with the sera from healthy donors, were thought to be candidates for citrullinated synovial autoantigens, indicated by the numbers 1–30 in (c) and (d).

Figure 2

Citrullination of F-actin capping protein α-1 subunit (CapZα-1) in synovial tissues of patients with rheumatoid arthritis. The extracts from synovial tissues from three patients (RA103, 107, and 109) were separated by two-dimensional electrophoresis. Proteins were stained with SYPRO Ruby protein gel stain (left column). After transfer, citrullinated proteins were detected by Western blotting with anti-modified citrulline (anti-MC) antibodies (middle column). The protein spots of CapZα-1 confirmed by Western blotting with anti-CapZα-1 antibodies (right column) are indicated by arrows.

Figure 3

Preparation of recombinant F-actin capping protein α-1 subunit (CapZα-1) with His-tag and citrullination of the recombinant CapZα-1. (a) cDNA for the entire protein coding region of human CapZα-1 was amplified by reverse transcription-polymerase chain reaction and then inserted into the pETBlue-2 vector. The full-length CapZα-1 was produced in Escherichia coli and purified by using histidine-Ni⁺ affinity (lane 1, produced CapZα-1 before purification; lane 2, CapZα-1 after purification). (b) The recombinant CapZα-1 was reacted with 20 U/mg of peptidylarginine deiminase in various concentrations of CaCl₂. Then, the citrulline residues were detected by the anti-modified citrulline (anti-MC) antibodies.

Figure 4

Detection of the autoantibodies to citrullinated or non-citrullinated F-actin capping protein α-1 subunit (CapZα-1) by enzyme-linked immunosorbent assay (ELISA). The autoantibodies to citrullinated or non-citrullinated CapZα-1 were detected by ELISA in patients with rheumatoid arthritis (RA) and osteoarthritis (OA) (a), systemic lupus erythematosus (SLE) (b), and in healthy donors (age- and gender-matched to RA, OA, or SLE groups). The dotted line indicates the cutoff point for positive reaction of 100 binding units. Serum samples diluted at 1:400 were used. The numbers in square brackets indicate the numbers of the antibody-positive serum samples/the numbers of the tested serum samples in each category. The numbers in parenthesis indicate percentages of the antibody-positive serum samples in each category.

Figure 5

The autoantibody titers to non-citrullinated and citrullinated F-actin capping protein α-1 subunit (CapZα-1) in the individual patients with rheumatoid arthritis. The patients were classified into four groups according to the reactive patterns. (a) The autoantibodies both to non-citrullinated and citrullinated CapZα-1 were negative (group A). (b) The autoantibodies to non-citrullinated CapZα-1 were negative, and the autoantibodies to citrullinated CapZα-1 were positive (group B). (c) The autoantibodies to non-citrullinated CapZα-1 were positive, and the autoantibodies to citrullinated CapZα-1 were further increased (group C). (d) The autoantibodies both to non-citrullinated and citrullinated CapZα-1 were positive with similar titers (group D).

Figure 6

Cross-reactivity of autoantibodies to citrullinated F-actin capping protein α-1 subunit (CapZα-1). To investigate cross-reactivity to other synovial proteins, the autoantibodies purified with citrullinated CapZα-1 (lane 1), control human immunoglobulin G (lane 2), polyclonal anti-CapZα-1 antibodies (lane 3), and rheumatoid arthritis serum samples (lane 4) were reacted with synovial tissue proteins separated by SDS-PAGE. The purified autoantibodies cross-reacted with five protein bands (arrows). Three bands (white arrows) were not reacted with polyclonal anti-CapZα-1 antibodies. On the other hand, two bands (black arrows) were reacted with polyclonal anti-CapZα-1.