Disease-specific autoantibody production in the lungs and salivary glands of anti-synthetase syndrome

Abstract

Interstitial lung disease is a common complication of anti-synthetase syndrome (ASS), and lymphocytic infiltration is often observed in the lesion. We have recently reported that disease-specific autoantibodies are produced by infiltrating lymphocytes in some autoimmune diseases. Here, we investigate the antigen specificity of B cells in the lung lesions of ASS patients. A total of 177 antibodies were produced from antibody-secreting cells in bronchoalveolar fluid (BALF) of three each of serum anti-Jo-1 and serum anti-EJ antibody-positive patients. Twelve to 30% and 50 to 62% of these antibodies were disease-specific autoantibodies, respectively. These autoantibodies recognized conformational epitopes of the whole self-antigen and had affinity maturations, indicating that self-antigens themselves are the target of humoral immunity. In addition, 100 antibodies were produced from two salivary gland tissues, obtained by chance, of ASS patients. Salivary glands are not generally recognized as lesions of ASS, but unexpectedly, ASS-related autoantibody production was also observed similar to that of BALF. Immunostaining confirmed the presence of ASS-related autoantibody-producing cells in salivary glands. Our results suggest that disease-specific autoantibody production at lesion sites is a common pathogenesis of autoimmune diseases, and that tissue-specific production of autoantibodies can provide insights regarding the distribution of organ manifestations in autoimmune diseases.

Keywords: anti-Ro52 antibody; anti-aminoacyl-tRNA synthetase antibody; anti-synthetase syndrome; autoantibody; idiopathic inflammatory myopathy; salivary gland.

Figure 1

Autoantibody production in the lungs. (A) IgG was purified from paired samples of serum and bronchoalveolar fluid (BALF), and the reactivity against Jo-1-, EJ-, Ro52m-, and Ro60-binding beads was measured as median fluorescence intensity (MFI) by antigen-binding bead assay. (B) Monoclonal antibodies were generated from antibody-secreting cells in BALF. The reactivity of these antibodies against Jo-1-, EJ-, Ro52m-, and Ro60-binding beads was measured as MFI by bead assay. Antibodies that bound to the beads are shown in red and those that did not are shown in blue. The isotype of the antibodies are shown by shapes. Pt4B73 could bind to both Jo-1 and EJ. Pt4B40 and Pt4B43 are polyreactive antibodies. (C) The proportion of ASS-related autoantibodies among the antibodies derived from each patient are shown. The center of the graph indicates the number of antibodies produced from each patient.

Figure 2

Autoantibody production in the salivary glands. (A) Monoclonal antibodies were generated from antibody-secreting cells in salivary gland tissues. The reactivity of these antibodies against Jo-1-, EJ-, Ro52m-, and Ro60-binding beads were measured as MFI. Antibodies that bound to the beads are shown in red and those that did not are shown in blue. The isotype of the antibodies are shown by shapes. (B) The proportion of ASS-related autoantibodies among the antibodies derived from each patient are shown. The center of the graph indicates the number of antibodies produced from each patient.

Figure 3

Detection of autoantibody-producing cells by immunofluorescence using autoantigens. Fresh-frozen sections of salivary glands were stained with purified green fluorescent protein (GFP) or GFP-autoantigen fusion proteins, anti-CD138 antibody (a marker of the antibody-producing cells), and DAPI. Autoantibody-producing cells are defined as GFP-autoantigen-positive and anti-CD138 antibody–positive cells and are indicated by white arrowheads. Representative single marker and overlay images of the salivary gland (A) from patient 7 (serum anti-Jo-1 antibody–positive) and (B) from patient 6 (serum anti-EJ antibody–positive) are shown. Scale bar indicates 50 μm.

Figure 4

Binding mode of autoantibodies against autoantigen. The reactivities of autoantibodies to full-length or fragmented autoantigen-binding beads were examined. The number of antibodies that can recognize each fragment (red or blue) or only the full-length antigen (black) is shown. The figure shows the three-dimensional structures of (A) Jo-1 (4X5O), (B) EJ (2PME), and (C) Ro52 (5OLM and 2IWG) reported in PDB, colored by the fragment. Note that the reported structure does not include the positions of some amino acids. Since the full-length structure of Ro52 is not known, the two structural analyses were cited separately. (D–F) We performed western blotting to determine whether autoantibodies could recognize GFP-fused denatured antigens ((D) Jo-1, (E) EJ, and (F) Ro52m) on the membrane. “pos” indicates blotted with anti-GFP antibody, and other lane names indicate the name of the antibody blotted. The antibody name indicates the patient number of origin, whether BALF (B) or salivary gland (S) derived, and the cell number of origin. The colors of the antibody names were matched to the results in (A–C). * indicates very weak binding that is rather difficult to see in the figure. ‡Since the sequences of Pt4B13 and Pt4B14 are identical, only Pt4B13 was examined.

Figure 5

Affinity maturation of the autoantibodies. Representative revertant antibodies were produced by reverting the somatic hypermutations of the disease lesion–derived antibodies to the genomic sequence. The reactivities of disease lesion–derived antibodies and revertant antibodies against corresponding antigens were compared. Disease lesion–derived Pt4B73 could bind to both Jo-1 and EJ. Wilcoxon signed-rank test.

Figure 6

The characteristics of autoantibodies. (A) H-chain and (B) L-chain isotypes for each corresponding autoantigen of BALF or salivary gland-derived antibodies were compared with other antibodies by Fisher’s exact test with Bonferroni correction. **p < 0.01, ***p < 0.001. Because there were only two IgDs, we excluded them from subsequent analysis. (C) The number of the somatic hypermutations of heavy chain of the antibodies from BALF samples were compared between autoantibodies and other antibodies by Steel test. *p < 0.05. (D) The gene usage of the antibodies are shown.